1

MICRORNAS AS BIOMARKERS IN GINGIVAL CREVICULAR FLUID

By

KELSEY CRONAUER WAHL

A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT

OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE

UNIVERSITY OF FLORIDA

2020

2

© 2020 Kelsey Cronauer Wahl

3

To my husband and family for their never-ending love, encouragement, and support

4

ACKNOWLEDGMENTS

I would like to thank my mentor, Dr. L. Shannon Holliday, for his guidance and

support with this project. I would like to thank Dr. Calogero Dolce and Dr. Robert Caudle

for their assistance. I would like to thank all faculty, staff, and residents in the University

of Florida Orthodontic Department for their support during residency. I would also like to

thank my wonderful husband and my family for their love, support, and encouragement

to pursue my passion.

5

TABLE OF CONTENTS page

ACKNOWLEDGMENTS .................................................................................................. 4

LIST OF TABLES ............................................................................................................ 6

LIST OF FIGURES .......................................................................................................... 7

LIST OF ABBREVIATIONS ............................................................................................. 8

ABSTRACT ................................................................................................................... 10

CHAPTER

1 INTRODUCTION .................................................................................................... 12

2 MATERIALS AND METHODS ................................................................................ 22

Participants and Eligibility ....................................................................................... 22

Study Design .......................................................................................................... 22 Sample Collection ................................................................................................... 23

PerioPaper ....................................................................................................... 23

Durapore Filter Membrane ............................................................................... 24 Microcapillary Tube .......................................................................................... 24

RNA Isolation .......................................................................................................... 25 cDNA Synthesis ...................................................................................................... 26

Real-Time PCR ....................................................................................................... 26 Statistical Considerations ........................................................................................ 27

3 RESULTS ............................................................................................................... 31

4 DISCUSSION ......................................................................................................... 42

5 CONCLUSIONS ..................................................................................................... 47

LIST OF REFERENCES ............................................................................................... 48

BIOGRAPHICAL SKETCH ............................................................................................ 54

6

LIST OF TABLES

Table page 2-1 Sample Collection Pattern for Subjects .............................................................. 27

2-2 Sample Identifiers ............................................................................................... 28

3-1 Ct Values for PerioPaper Samples ..................................................................... 39

3-2 Ct Values for Durapore Samples ........................................................................ 39

3-3 Ct Values for Microcapillary Tube Samples ........................................................ 39

3-4 ANOVA Descriptive Statistics for Full Data Set (Ten Samples) .......................... 40

3-5 ANOVA Descriptive Statistics for Subset of Data (Six Samples) ........................ 40

3-6 t-test for Treated and Untreated Subjects ........................................................... 41

7

LIST OF FIGURES

Figure page 2-1 Microfuge® 18 Centrifuge ................................................................................... 28

2-2 2720 Thermal Cycler .......................................................................................... 29

2-3 96-well Plate ....................................................................................................... 29

2-4 C1000 Thermal Cycler ........................................................................................ 30

3-1 Amplification Plot of A1 and B1 Samples ........................................................... 35

3-2 Amplification Plot of C1 and D1 Samples ........................................................... 35

3-3 Amplification Plot of E1 and F1 Samples ............................................................ 36

3-4 Amplification Plot of A2, F2, M1 and M2 Samples .............................................. 36

3-5 Bar Graph Showing Full Dataset (Green) and Subset (Blue) of Mean Ct Values for miRNA-146a ...................................................................................... 37

3-6 Bar Graph Showing Full Dataset (Green) and Subset (Blue) of Mean Ct Values for miRNA-103-3p ................................................................................... 37

3-7 Bar Graph Showing Mean Ct Values for miRNA-146a (Blue) and miRNA 103-3p (Green) and ΔCt (Gold) in Treated and Untreated Subjects .......................... 38

8

LIST OF ABBREVIATIONS

ALP Alkaline phosphatase

ANOVA Analysis of variance

cDNA Complementary deoxyribonucleic acid

Ct Cycle threshold

DNA Deoxyribonucleic acid

EV Extracellular vesicle

GCF Gingival crevicular fluid

HGF Human gingival fibroblast

IL-1β Interleukin one beta

IL-6 Interleukin six

IL-8 Interleukin eight

IRAK1 Interleukin one receptor associated kinase one

miRNA Micro-ribonucleic acid

mRNA Messenger ribonucleic acid

NF-κB Nuclear factor kappa B

PB Processing body

PCR Polymerase chain reaction

pre-mRNA Precursor messenger ribonucleic acid

pri-miRNA Primary micro-ribonucleic acid

qPCR Quantitative polymerase chain reaction

RANK Receptor activator nuclear kappa B

RANKL Receptor activator nuclear kappa B ligand

RT-PCR Real-time polymerase chain reaction

9

RISC Ribonucleic acid induced silencing complex

RNA Ribonucleic acid

SLE Systemic lupus erythematosus

TNF-α Tumor necrosis factor alpha

TRAF6 Tumor necrosis factor receptor associated factor six

UTR Untranslated region

ΔCt Delta cycle threshold

ΔΔCt Delta delta cycle threshold

10

Abstract of Thesis Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Master of Science

MICRORNAS AS BIOMARKERS IN GINGIVAL CREVICULAR FLUID

By

Kelsey Cronauer Wahl

May 2020

Chair: Lexie Shannon Holliday Major: Dental Sciences – Orthodontics

In medicine and dentistry there is a significant focus on improving patient

diagnosis and outcomes. With the identification of exosomes as methods of intercellular

communication and genetic exchange, another avenue for biomarker discovery

presented itself. Released from osteoclasts and inflammatory cells, extracellular

vesicles carry genetic information, such as microRNAs (miRNAs), which can alter gene

transcription in target cells and affect the inflammatory and bone remodeling processes.

Gingival crevicular fluid (GCF) is a commonly used fluid for detection of biomarkers,

including miRNAs. The objective of this study was to determine if miRNA-146a could be

sufficiently collected by GCF methods and which method was best for collecting this

miRNA. This study compared three different methods which are the following:

Periopaper strips, Durapore filter membrane, and micropipettes. Twelve adult

volunteers were sampled and sampling took place over two visits. For each subject

three teeth were sampled, each tooth with one of the proposed methods. A pattern was

established so that collection method varied per tooth but repeated every fourth subject,

for comparison. Collections were taken on the mesial and distal surfaces of each

specified tooth. RNA isolation, cDNA synthesis, and real-time polymerase chain

11

reaction (RT-PCR) were performed on each sample to assess expression of miRNA-

146a. A reference gene, miRNA-103-3p, was also used for assessment and

normalization of data. Samples from seven subjects were analyzed and the results

showed that there was no significant difference in method of collection for identifying

and expressing miRNA-146a. However, a secondary analysis of samples showed that

there was a significant difference in expression of miRNA-146a between subjects who

were undergoing orthodontic treatment and those who were untreated. This study

concluded that there is no advantage to using one of the three collection methods over

another for expression of miRNA-146a.

12

CHAPTER 1 INTRODUCTION

For many years, miRNAs have been studied to determine their roles in health

and disease.1 Comprising 1-2% of all genes in mammals, nearly all biological processes

are affected by miRNA control.2 Structurally, miRNA is a single-stranded RNA, 21-25

nucleotides in length.2 miRNAs are present in multiple organisms, including plants,

worms, and humans, and many are conserved across species.3 Remarkably,

components of miRNA machinery have been found in archaea and eubacteria, showing

their lasting presence from ancient times to today.4 Originally found to regulate

developmental timing in worms, the function of miRNA was determined to be even

greater, with the discovery of its involvement in networking and coordination of gene

expression.5 miRNAs have been shown to participate in cell cycle control, cardiac and

skeletal muscle development, and neurogenesis.6 Additionally, they have been

implicated in a number of diseases including cancers, heart disease, and neurological

diseases.7,8 Consequently, miRNAs are studied as markers of disease progression and

for diagnostic purposes.8

Production of miRNA takes places in the nucleus and cytoplasm of a cell.9 In the

nucleus, different areas of a gene are transcribed by RNA Polymerase II into long

primary miRNAs (pri-miRNA).10 These pieces are either inserted into noncoding RNAs,

fixed into introns of protein-coding genes, or are clustered in polycistronic transcripts.11

Processing of the pri-mRNAs occurs by two Ribonuclease-III enzymes, Drosha and

Dicer.12 In the nucleus, the enzyme Drosha attaches to a pri-mRNA and excises the

next precursor, pre-mRNA.10 The pre-mRNA is about seventy nucleotides long and can

fold into a stem-loop structure.10 The pre-mRNA is exported from the nucleus into the

13

cytoplasm by Exportin-5.13 Once in the cytoplasm, it is cleaved by the enzyme Dicer,

producing a miRNA duplex intermediate that is about twenty base pairs in length.14 The

duplex contains one strand of mature miRNA, which becomes bound by a large

Argonaute protein and becomes part of the RNA-induced silencing complex (RISC).15

The mechanism of action of miRNA is achieved through the RISC and is

attributed mainly to the Argonaute core.15 The Argonaute protein not only binds the

miRNA in forming the RISC, but it also joins with other Argonaute proteins to form the

center of the RISC complex.16 This Argonaute core is highly conserved amongst

species and is responsible for the slicing activity that cleaves messenger RNA (mRNA),

as instructed by the miRNA.17 The RISC binds to a 3’ untranslated region of the target

mRNA (mRNA-3’-UTR), regulating the expression of the mRNA, and subsequently, the

expression of genes.18

Once bound, the RISC cleaves the target mRNA, inhibiting protein synthesis and

degrading the target mRNA.19 If destined for degradation, the cleaved mRNA associates

with cytoplasmic processing bodies (PBs).20 The PBs localize to the RISC’s Argonaute

core and contain enzymes needed for mRNA degradation, such as decapping enzymes

and exonucleases.20 Together, the PBs and RISC carry out the actions determined by

their associated miRNA.9 Furthermore, a single miRNA can regulate multiple targets,

emphasizing the widespread effect of miRNAs throughout the body.21

miRNAs have been studied as potential biomarkers of health and disease.1

Biomarkers are considered any structure or substance objectively measured in a

sample to indicate health and disease.22 Cytokines, such as interleukin-1β (IL-1β) and

tumor necrosis factor alpha (TNF-α) have been studied as biomarkers in the medical

14

and dental fields.23 Studies of tooth movement have also observed interleukin-6 (IL-6),

interleukin-8 (IL-8), dentine sialophosphoprotein and alkaline phosphatase (ALP) as

biomarker candidates in orthodontic tooth movement.24-26 Furthermore, receptor

activator of nuclear factor kappa B (RANK), RANK ligand (RANKL), and osteoprotegerin

(OPG) have been studied as biomarkers of bone resorption.27 Studies show the

presence of RANKL, a transmembrane protein, and RANK in extracellular vesicles

(EVs) and suggest that EVs and their components can be detected in gingival crevicular

fluid.28,29

The term “exosome” was initially suggested to identify extracellular vesicles of

endosomal origin.30 Once regarded as cellular trash, exosomes are now known as a

form of intercellular communication that can alter functions of target cells.31 microRNAs

are found in exosomes and function by regulating gene expression post-

transcriptionally.32 Some studies have observed the role of miRNAs in bone remodeling

and inflammation.33,34 One miRNA proposed as a key regulatory molecule of the

inflammatory response is miRNA-146.35

The miRNA-146 family consists of miRNA-146a and miRNA-146b.35 Both can be

induced by pro-inflammatory cytokines, such as IL-1β and TNF-α.35 It is also reported

that miRNA-146a negatively regulates the innate immune response by repressing IL-1

receptor-associated kinase 1 (IRAK1) and tumor necrosis factor receptor-associated

factor 6 (TRAF6).36 IRAK1 and TRAF6 are significant because they are two crucial

molecules in the nuclear factor kappa B (NF-κB) pathway.36 IRAK1 and TRAF6 are

significant because they are two crucial molecules in the nuclear factor kappa B (NF-κB)

pathway.36 They act by increasing NF-κB activity, resulting in an increased expression

15

of the pro-inflammatory cytokines IL-6 and IL-8, which are key mediators of

inflammation.36

In medicine, some chronic inflammatory diseases are shown to be related to

miRNA-146a dysregulation.35 Psoriasis, systemic lupus erythematosus (SLE), and

rheumatoid arthritis are a few of the most significant.37-39 In dentistry, studies have

proposed miRNA-146a as an inflammatory biomarker by observing its expression in

patients with periodontitis.40,41 A study by Motedayyen et al. assessed miRNA-146a

expression in subjects with healthy periodontium versus chronic periodontitis.41 They

found that subjects with chronic periodontitis had significantly higher levels of miRNA-

146a.41 Furthermore, they found that elevated miRNA-146a was accompanied by a

reduction in TNF-α and IL-6, two pro-inflammatory cytokines.41

These results are similar to another study by Xie et al. which analyzed miRNA-

146a expression in the inflammatory response in human gingival fibroblasts (HGFs).41,42

In this study, they used polymerase chain reaction (PCR) to measure the expression of

miRNA-146a after Porphyromonas gingivalis (P. gingivalis) lipopolysaccharide

stimulation.42 Their results showed a significant difference in miRNA-146a expression in

the P. gingivalis stimulated HGFs, with a greater expression of miRNA-146a in these

HGFs.42

MicroRNA-146a expression in bone has also been of interest.43 A study by

Nakasa et al. aimed to determine whether overexpression of miRNA-146a inhibits

osteoclastogenesis.44 In their animal study, they administered an intravenous injection

of double-stranded miRNA-146a into mice with arthritis.44 They observed that while not

completely suppressed, the mice injected with miRNA-146a showed a reduction of pro-

16

inflammatory cytokines TNF-α, IL-1β, and IL-6.44 Radiographically and histologically,

those injected with the microRNA showed less destruction of cartilage and bone.44

Further assessing the role of miRNA-146a in bone remodeling, a study by

Holliday et al. looked at osteoclast exosomes, exosomal components and bone

remodeling processes.34 They discussed candidate molecules identified in exosome-

based regulation of bone remodeling.34 Two of these candidates were miRNA-146a and

miRNA-214.34 Studies by Li et al. and Sun et al. detected miRNA-214 as enriched in

osteoclast exosomes compared with exosomes from precursors.33,45 The Holliday et al.

study found similar results to Sun et al. regarding miRNA-214, which Sun et al. reported

to be enriched 4-fold in osteoclasts.33,34 Significantly, Holliday et al. found that miRNA-

146a was enriched over 80-fold in osteoclasts compared with precursors, making it a

potential biomarker for bone remodeling and providing a rationale for further study of

miRNA-146a as a candidate biomarker for orthodontic tooth movement.34

A recent study of orthodontic tooth movement by Atsawasuwan et al. analyzed

gingival crevicular fluid (GCF) samples of orthodontically treated and untreated subjects

to observe the expression of microRNA-29.46 GCF samples were collected five times

throughout treatment and the results showed the greatest expression of microRNA-29 in

the subjects who underwent canine retraction.46 Furthermore, they found that the

highest concentration of their miRNA of interest was present in the fraction of GCF

containing EVs.46 Their results showed an increased expression of miRNA-29 in the

treated subjects, revealing a correlation between this miRNA, orthodontic tooth

movement, and osteoclast function.46 Furthermore, this study suggests that miRNAs in

GCF can be candidate biomarkers for orthodontic tooth movement.46

17

Of the many miRNAs that impact physiologic and disease processes there are

some deemed “reference” miRNAs.47 One such reference miRNA is miRNA-103-3p.47

While found to promote human gastric cancer cell proliferation, it is an endogenous

miRNA present in serum, plasma, urine, and cerebrospinal fluid.48 A study by Song et

al. identified suitable reference genes for quantitative PCR (qPCR) analysis of serum.49

Of six candidate reference microRNAs, microRNA103-3p was found to be moderately

abundant in serum samples, suggesting it as a suitable miRNA for comparison.49

For many years, GCF has been collected and analyzed for potential

biomarkers.50 Rather than collecting and examining saliva or other oral fluids, GCF is

preferred due to its site-specific nature and better ability to express the metabolic status

of localized tissue.51 Its inclusion of bacterial and host cell molecules makes it ideal for

evaluating cellular metabolism.51 Considered a serum transudate or inflammatory

exudate, GCF accumulates in the gingival crevice by leakage through postcapillary

venules in tissue.51,52 When the lymphatic system is unable to remove the fluid, due to a

greater filtrate volume, leakage occurs and fluid escapes into the gingival sulcus.52 It is

in the sulcus or at the gingival margin that GCF is collected.53 Experimentally, this

process is supported by an animal study by Del Fabbro et al.54 They observed gingival

fluid flow by measuring osmotic pressure at different points in the gingival crevice.*

They also observed the thickness of epithelium to distinguish absorption or fluid

release.54 These methods of measurement are an improvement from past techniques,

where quantities were primarily based on estimates.55

Collection of available GCF is a sensitive process because contamination of any

kind can affect analysis and results.53 If blood contaminates the sample it must be

18

discarded, and plaque and saliva must be removed from the area.56 Furthermore,

collecting GCF samples is a challenge when fluid is sparse.53 Enough fluid must be

gathered for adequate analysis, but increasing collection time can increase the risk for

contamination.53 It has been shown that GCF can be isolated from a healthy sulcus,

although only in small amounts.51 It has been clinically demonstrated that GCF

production greatly increases when gingival inflammation is present.57 Damage to the

epithelium leads to localized periodontal inflammation and increased vascular

permeability, which increases the flow.57 A healthy periodontium releases about 3

μl/hour of fluid into the gingival sulcus, while sites with periodontal disease can release

up to 44 μl/hour.58 Regardless of these limitations, many studies in health care have

successfully used GCF collection techniques to obtain adequate fluid samples for

analysis.59,60

The gingival washing method, use of micropipettes, and use of filter paper strips

are three main techniques for collecting GCF.53 Each technique has its advantages and

disadvantages, with the washing method best suited for samples that need cells

collected.53 This technique involves injecting and aspirating a balanced salt solution of

known amount into the gingival sulcus repeatedly, before collecting the final sample.53

By mixing the fluids in this way it is difficult to determine the sample dilution and to

control the final volume.53 As described, the gingival washing method is not the most

efficient system for gathering GCF but can be useful if sample dilution is not of

concern.53

Using capillary tubes to collect GCF is more beneficial than gingival washing, if

standardization of the GCF volume is desirable.53 Because the micropipettes have a

19

known diameter, GCF volume is easy to determine.53 Tubes of varying diameters are

available for GCF collection, and filling the tube fully with GCF would provide an

accurate measurement of sample volume.61 The tubes are placed in the gingival crevice

and fluid flows into the micropipettes by capillary action.53,61,62 The disadvantage of this

method is that it may require more time to obtain an adequate volume of GCF.53

Using absorbent paper strips is the favored technique for GCF collection

because it is fast and least traumatic.63 Many types and sizes of paper strips are

available, such as Whatman chromatography paper and PerioPaper™.64 When using

the strips the sample tooth is cleaned with a cotton pellet, isolated with cotton rolls,

dried, and then the paper strip is placed.56 The strip is typically placed into the gingival

sulcus until slight resistance is felt, or at the coronal edge of the gingival sulcus.53

Placement of the strip causes little to no irritation, and GCF is collected by fluid

migration through the paper.53

To observe miRNAs from GCF, RT-PCR is often used.65 PCR is a method for

amplifying fragments of DNA.65 This technique allows for segments of specific

chromosomes to be amplified more than a million-fold, and for testing of very small

sequences of DNA, which could not otherwise be tested.66 RT-PCR is the ability to

monitor the progress of the PCR as it occurs, since data is collected throughout the

PCR process.65 RT-PCR records the point during cycling when the target is first

detected and amplified, rather than reporting the amount of target gathered after a fixed

number of cycles, as in conventional PCR.67

The real-time PCR process consists of twenty-five to fifty cycles, and each cycle

undergoes three phases at different temperatures.65 The initial phase is denaturation,

20

when the DNA helix is separated into two strands.65 Denaturation occurs at a high

temperature.65 The second phase, annealing, occurs when the sample is cooled and the

added primer attaches to the 3’ end of the target DNA to be amplified.65 The third phase

is amplification, which occurs at a high temperature and is when elongation of the

primer occurs, creating a complementary DNA strand to the target.65

Significantly, a fluorescent dye is used in qPCR and binds to double-stranded

DNA molecules by inserting between the base pairs.68 The fluorescence is measured

with each amplification cycle to determine relatively how much DNA has been

amplified.68 With the use of dye in PCR, the higher the starting amount of the target

DNA, the sooner an increase in fluorescence is observed.68 qPCR data is represented

in an amplification plot, which shows each sample as a curve using cycle number and

fluorescent signal.69

At first glance, an amplification plot allows for basic interpretation of the data

without numerical values.69 In a plot, the curve farthest to the left signifies the highest

concentration of the target and earliest signal amplification.69 The curve farthest to the

right signifies the lowest initial concentration of the target and subsequent latest signal

amplification.69 The data can be understood more specifically by observing key points in

the amplification plot.70 A horizontal line is present in each plot and denotes a threshold

value, set by the machine.70 The section of the curve below the threshold represents

fluorescence that cannot be distinguished from background fluorescence.70 The point at

which fluorescence can be sufficiently detected is when the curve just passes the

threshold value.70 This point is easily identified by the intersection between the curve

21

and threshold line and is referred to as the cycle threshold, or Ct.70 The ability to

determine Ct values provides an easy method for comparison between samples.70

22

CHAPTER 2 MATERIALS AND METHODS

Participants and Eligibility

The University of Florida Institutional Review Board for the Protection of Human

Subjects approved this project (UF-IRB-01, IRB201801700). Additional funding for this

project was provided by the Southern Association of Orthodontists (SAO). Gingival

crevicular fluid samples were collected from twelve adult volunteers using the inclusion

and exclusion criteria. The inclusion criteria consisted of the following: an age range of

18 to 40 years old, the presence of all six maxillary and mandibular anterior teeth, and

the presence of all four maxillary premolars. Exclusion criteria included the presence of

extremely poor oral hygiene or a history of smoking or tobacco use within the past year.

The patient population was representative of the patient population of the University of

Florida College of Dentistry Department of Orthodontics, with patients of all races,

genders, and ethnicities included in the study.

Study Design

Collections of all GCF samples were carried out in two visits for each subject.

The second visit was at least twenty-four hours after the first visit. At the initial visit, all

subjects confirmed that they did not eat, drink, brush their teeth, or use mouthwash at

least two hours prior to sample collection. Subjects confirmed this again at the second

visit, prior to collection. At enrollment and prior to sample collection, all subjects

provided written informed consent. Three methods of GCF collection were used and are

the following: PerioPaper, Durapore filter membrane, and microcapillary tubes. Teeth

numbers 5, 7, and 9, were used as collection sites. For each participant, the collection

method varied per tooth, following the pattern listed in Table 2-1.

23

Sample Collection

Each sample collection Eppendorf tube was labeled with a specific subject

identifier. The subject was assigned a letter, with the first subject being “A.” Directly

following that letter was a number designating the first visit, “1,” or the second visit, “2.”

A hyphen then separated a label designating which method was performed, either “P”

for PerioPaper, “D” for Durapore filter membrane, or “M” for microcapillary tube,

followed by the tooth number that was collected. Table 2-2 shows the identifiers listed

for the samples that were analyzed.

For each subject at each visit six total samples were collected, with two samples

per technique. Plaque was removed from the buccal surface of each tooth along the

gingival margin and samples were collected on the mesial and distal of each tooth. If

any sample was visually contaminated with blood it was discarded and a new sample

was taken. The specific collection techniques for each method are as follows:

PerioPaper

The tooth designated for sampling was cleaned of any plaque on the buccal

surface using a cotton pellet. The tooth was then isolated with cotton rolls to prevent

contamination by saliva and was air dried for five seconds. A PerioPaper® GCF strip

(Oraflow Inc., Plainview, NY) was inserted about 2mm into the gingival sulcus, until mild

resistance was felt, on the mesial aspect of the tooth. The strip was left in place for thirty

seconds and was removed using a sterilized cotton plier. It was placed in a sterilized

Eppendorf tube with the appropriate subject/sample label and 100 μl of phosphate

buffered salien (PBS; pH 7.4, RNase free). The same process was repeated with a new

PerioPaper strip on the distal aspect of the tooth. The samples were immediately placed

in a freezer and kept at -80°C for further analysis.

24

Durapore Filter Membrane

For each sample collection a Durapore filter membrane was cut to the same

length and width as a PerioPaper strip, 2x8mm. The tooth designated for sampling was

cleaned of any plaque on the buccal surface using a cotton pellet. The tooth was then

isolated with cotton rolls to prevent contamination by saliva and was air dried for five

seconds. A cut, sterilized Durapore filter membrane strip (pore size of 0.22μm; Millipore

Corp., Bedford, Mass., USA) was inserted about 2mm into the gingival sulcus, until mild

resistance was felt, on the mesial aspect of the tooth. The strip was left in place for thirty

seconds and was removed using a sterilized cotton plier. It was placed in a sterilized

Eppendorf tube with the appropriate subject/sample label and 100 μl of PBS (pH 7.4,

RNase free). The same process was repeated with a new Durapore filter membrane

strip on the distal aspect of the tooth. The samples were immediately placed in a freezer

and kept at -80°C for further analysis.

Microcapillary Tube

The tooth designated for sampling was cleaned of any plaque on the buccal

surface using a cotton pellet. The tooth was then isolated with cotton rolls to prevent

contamination by saliva and was air dried for five seconds. A 1 μL microcapillary tube

(Drummond Scientific Co., Broomall, Pennsylvania, USA) was placed on the mesial

aspect of the tooth at the entrance to the gingival sulcus. The micropipette was held in

place by the operator for five minutes. GCF was taken up into the micropipette by

capillary action. After five minutes the GCF was eluted from the micropipette using a

bulb provided by the company. The bulb is attached to one end of the micropipette and

produces gentle air pressure, dispensing the GCF into a sterilized Eppendorf tube

containing 100 μL of PBS (pH 7.4, RNase free). The same process was repeated with a

25

new microcapillary tube on the distal aspect of the tooth. The samples were immediately

placed in a freezer and kept at -80°C for further analysis.

RNA Isolation

A total of seven subjects and samples from ten visits were used for analysis, as

shown in Table 2-2. The QIAGEN kit protocol, miRNeasy Serum/Plasma Advanced Kit,

for RNA isolation was used and began by thawing each sample and placing them on

ice. 60μl Buffer RNA Proximity Ligation (RPL) was added to each Eppendorf tube

containing thawed sample with PBS. Each tube was closed and vortexed for more than

five seconds. The samples then rested at room temperature for three minutes. The

samples were centrifuged at 13,000 times gravity (×g) for three minutes at room

temperature, to precipitate the pellet, as shown in Figure 2-1. The supernatant from

each was transferred to a new, labeled, reaction tube. 1 volume isopropanol was added

and mixed by vortexing. Each entire sample was transferred to an RNeasy UCP

MinElute column, the lids were closed, and the columns were centrigured for fifteen

seconds at greater than 8,000 ×g. The flow-through was discarded. 700 μl Buffer RWT

was pipetted into the Rneasy UCP MinElute spin column. The lids were closed and they

were centrifuged for fifteen seconds at greater than 8,000 ×g. The flow-through was

discarded. 500μl Buffer RPE was placed onto the RNeasy UCP MinElute spin column.

The lids were closed and they were centrifuged for fifteen second at greater than 8,000

×g. The flow-through was discarded. 500μl of 80% ethanol was added to the Rneasy

UCP MinElute spin column. The lids were closed and they were centrifuged for two

minutes at greater than 8,000 ×g. The flow-through and collection tubes were discarded.

The Rneasy UCP MinElute spin columns were placed in new 2 mL collection tubes,

supplied in the kit. The lids of the spin columns were kept open and they were

26

centrifuged at full speed for five minutes to dry the membrane. The flow-through and

collection tubes were discarded. The Rneasy UCP Min elute spin columns were placed

in new 1.5 mL collection tubes, supplied in the kit, and 20 μl Rnase-free water was

added diretly to the center of each spin column membrane. Incubation then occurred for

one minute. The lids were closed and they were centrifuged for one minute at full

speed. The RNA was then eluted. If ready for cDNA synthesis it was kept on ice for the

next step, otherwise it was frozen at -80°C for further analysis.

cDNA Synthesis

cDNA synthesis was accomplished using the QIAGEN kit protocol, specifically

the miRCURY LNA Universal RT microRNA PCR protocol. If frozen, the RNA samples

were thawed and kept on ice. From the kit, the 5x Reaction Buffer, Enzyme mix, and

RNA spike were thawed in the fridge. A master mix was made by adding 2 μl of the 5x

Reaction Buffer, 1 μl of the enzyme mix, and 0.5 μl of the RNA spike to an Eppendorf

tube. This tube was kept on ice. New, Eppendorf tubes were labeled with

subject/sample identifiers and each RNA sample was added to the appropriately labeled

tube. Master Mix was added to each tube and then each tube was ready for cDNA

synthesis. The samples were incubated in a 2720 Thermal Cycler (Figure 2-2) for 65

minutes with the following specifications: 60 minutes at 42°C, 5 minutes at 95°C, at

least 3 minutes at 4°C.

Real-Time PCR

To prepare for qPCR, a 96-well plate was used. A total of four 96-well plates

were used and four runs of qPCR were completed. The first round of qPCR was for

subjects A and B. The second round was for subjects C and D. The third round was for

subjects E and F. The fourth round was for subjects A, F, and M. An example of a

27

planned and labeled 96-well plate used for qPCR is shown in Figure 2-3. The QIAGEN

kit protocol, miRCURY LNA Universal RT microRNA PCR protocol, was used for

sample preparation in the PCR plate and for setting the Thermal Cycler machine. In

each labeled well the appropriate cDNA sample was added, along with nuclease-free

water, PCR master mix, and SYBR green dye. A total of 10 μl was present in each well.

The plate was sealed, spun in a Microplate Spinner for a few seconds and placed into

the C1000 Touch Thermal Cycler (Figure 2-4) to undergo real-time PCR. The settings

were as follows: 39 amplification cycles at 95°C for 10 seconds, 60°C for 1 minute, and

a ramp-rate of 1/6 C/s^7 optical read. qPCR data was retrieved using the Bio-Rad

qPCR Analysis Software. From this software, raw data was retrieved as Ct values and

amplification plots.

Statistical Considerations

One-way ANOVA and student t-test were performed to determine statistical

significance, with statistical significance represented by a p-value less than 0.05.

Table 2-1. Sample Collection Pattern for Subjects

Subject Tooth #5 Tooth #7 Tooth #9

A PerioPaper Durapore Micropipette B Durapore Micropipette PerioPaper C Micropipette PerioPaper Durapore D PerioPaper Durapore Micropipette

28

Table 2-2. Sample Identifiers

Subject PerioPaper Durapore Micropipette

A A1-P5 A1-P5 A1-D7 A1-D7 A1-M9 A1-M9 A2-P5 A2-P5 A2-D7 A2-D7 A2-M9 A2-M9 B B1-P9 B1-P9 B1-D5 B1-D5 B1-M7 B1-M7 C C1-P7 C1-P7 C1-D9 C1-D9 C1-M5 C1-M5 D D1-P5 D1-P5 D1-D7 D1-D7 D1-M9 D1-M9 E E1-P9 E1-P9 E1-D5 E1-D5 E1-M7 E1-M7 F F1-P7 F1-P7 F1-D9 F1-D9 F1-M5 F1-M5 F2-P7 F2-P7 F2-D9 F2-D9 F2-M9 F2-M9 M M1-P7 M1-P7 M1-D9 M1-D9 M1-M5 M1-M5 M2-P7 M2-P7 M2-D9 M2-D9 M2-M5 M2-M5

Figure 2-1. Microfuge® 18 Centrifuge. Courtesy of Dr. Kelsey Cronauer Wahl.

Centrifugation of Samples for RNA Isolation

29

Figure 2-2. 2720 Thermal Cycler. Courtesy of Dr. Kelsey Cronauer Wahl. cDNA

Synthesis

Figure 2-3. 96-well Plate Organized and Labeled. Courtesy of Dr. Kelsey Cronauer

Wahl. Plate Prepared for Fourth Run of qPCR

30

Figure 2-4. C1000 Thermal Cycler. Courtesy of Dr. Kelsey Cronauer Wahl. Plate

Loaded and Machine Prepared for qPCR

31

CHAPTER 3 RESULTS

Samples from seven of the twelve subjects were analyzed, and samples from

only the first visit were analyzed for all seven subjects. For three of the seven subjects,

their second visit samples were analyzed. Samples were collected using the three

techniques of PerioPaper strips, Durapore filter membrane, and microcapillary tubes.

qPCR was performed four times and amplification plots from all four runs were obtained

using the Bio-Rad qPCR Analysis Software and are shown from Figure 3-1 to Figure 3-

4. Using this software, numerical raw data was also obtained as Ct values. Table 3-1 to

Table 3-3 shows the Ct values obtained for each subject and organized based on GCF

collection technique and miRNA. These tables also include the calculated ΔCt values

for each subject, which uses miRNA-103-3p to normalize the data of miRNA-146a. ΔCt

values were calculate by subtracting the Ct value of miRNA-146a from the Ct value of

miRNA-103-3p.

For each of the three collection techniques, the mean Ct value for miRNA-146a

and miRNA-103-3p and the mean ΔCt value were calculated. A series of one-way

ANOVAs were conducted on the entire dataset of seven subjects (ten sample

groupings) and the descriptive statistics are presented in Table 3-4. This was done in

order to determine whether there are significant differences in the mean values of

miRNA-146a, miRNA-103-3p, and ΔCt on the basis of condition, which consisted of

Periopaper, Durapore, and microcapillary tubes. Significant mean differences were not

found with regard to miR146a, F(2, 27) = 1.539, p = .233, miR103-3p, F(2, 27) = 1.103,

p = .347, or ΔCt, F(2, 27) = 1.978, p = .158. All p-values were above 0.05. As shown,

32

means differed slightly by condition, while these mean differences were not significantly

different.

A series of one-way ANOVAs were conducted on a subset of the data, which

excluded four sample groupings. Four sample groupings were excluded because the

subjects were undergoing orthodontic treatment at the time of collection, as noted at the

visit. Six sample groupings were run for this one-way ANOVA series and descriptive

statistics are presented in Table 3-5. These analyses also failed to find significant mean

differences on the basis of miR146a, F(2, 15) = 1.970, p = .174, miR103-3p, F(2, 15) =

1.347, p = .290, or ΔCt, F(2, 15) = .529, p = .600. All p-values were above 0.05. Mean

differences were again found here on the basis of condition, while these mean

differences also failed to achieve statistical significance. Graphical representations of

data from the full dataset (ten sample) and subset (six sample) are shown in Figure 3-5

and Figure 3-6. Based on the statistical analysis we therefore fail to reject our null

hypothesis and conclude that there is no significant difference in the recovery of

miRNA-146a in GCF using Periopaper, Durapore filter membrane, or microcapillary

tubes.

An additional method of comparison was completed to determine the fold change

in expression of miRNA-146a between the GCF collection methods.71 The Livak Method

was used to calculate the fold change of gene expression, referred to as 2^-(ΔΔCt).71 In

this method, a reference is needed to compare conditions. PerioPaper was used as the

reference condition. Therefore the 2^-(ΔΔCt) values were calculated to compare fold

change expression of miRNA-146a for Durapore in relation to PerioPaper and for

microcapillary tubes in relation to PerioPaper. For Durapore the 2^-(ΔΔCt) was a 1.63

33

fold change decrease and the 2^-(ΔΔCt) for microcapillary tubes was a 3.12 fold change

decrease.

Since a few subjects were undergoing orthodontic treatment, the data was

analyzed further to assess whether there is a significant difference in miRNA-146a

expression between two groups, defined as treated and untreated. GCF collection

technique was not a factor of interest in this analysis. The treated and untreated groups

each consisted of five sampling groups. It is necessary to note that for one subject, the

samples from the initial visit were included in the untreated group, while the samples

from the second visit were included in the treated group. This was done since

orthodontic treatment did not begin until the second visit for that subject.

Independent-sample t-tests were conducted specifically comparing the untreated

subjects, with healthy periodontium, and treated subjects, with inflamed periodontium,

on their mean values of miRNA-146a, miRNA-103-3p, and ΔCt. In these analyses,

significant differences between these two groups were found with respect to miR146a,

t(22.103) = 2.253, p < .05, and ΔCt, t(28) = 4.844, p < .001, though not with respect to

miR103-3p, t(27.115) = .047, p = .963. As stated, p-values were less than 0.05 for

miRNA146a and the ΔCt. Table 3-6 presents the descriptive statistics associated with

these analyses, which indicate small mean differences on the basis of untreated

subjects and subjects undergoing orthodontic treatment. In the two cases in which

statistical significance was achieved, significantly higher means were found in the case

of untreated subjects. This data is graphically shown in Figure 3-7. Furthermore, the

Livak Method was used to calculate the fold change of gene expression between the

treated and untreated groups.71 For this calculation the mean ΔCt for the target miRNA-

34

146a and the mean ΔCt for the reference miRNA-103-3p were used to calculate the fold

change of gene expression of miRNA-146a between the two groups. The 2^-(ΔΔCt) was

an increase of 5.81.

35

Figure 3-1. Amplification Plot of A1 and B1 Samples

Figure 3-2. Amplification Plot of C1 and D1 Samples

36

Figure 3-3. Amplification Plot of E1 and F1 Samples

Figure 3-4. Amplification Plot of A2, F2, M1 and M2 Samples

37

Figure 3-5. Bar Graph Showing Full Dataset (Green) and Subset (Blue) of Mean Ct Values for miRNA-146a Based on GCF Collection Method

Figure 3-6. Bar Graph Showing Full Dataset (Green) and Subset (Blue) of Mean Ct

Values for miRNA-103-3p Based on GCF Collection Method

38

Figure 3-7. Bar Graph Showing Mean Ct Values for miRNA-146a (Blue) and miRNA

103-3p (Green) and ΔCt (Gold) in Treated and Untreated Subjects

39

Table 3-1. Ct Values for PerioPaper Samples

Subject microRNA-146a microRNA-103-3p ΔCt

A1 29.2545

27.3836 1.87092

A2 25.375

26.36 -0.985

B1 32.1364

28.8240 3.3124

C1 31.8422

27.8092 4.03296

D1 25.4571

23.2879 2.1692

E1 29.6349

24.7220 4.9128

F1 35.6996

31.2679 4.4317

F2 34.49

37.40 -2.91

M1 28.935

28.45 0.485

M2 29.03

27.66 1.37

Table 3-2. Ct Values for Durapore Samples

Subject microRNA-146a microRNA-103-3p ΔCt

A1 27.0819 25.4094 1.672 A2 26.685 25.12 1.565 B1 28.1666 24.7352 3.431 C1 34.0511 30.8830 3.168 D1 26.9159 23.8072 3.109 E1 29.4507 25.5267 3.924 F1 29.4714 24.7691 4.702 F2 28.765 26.9 1.865 M1 28.69 27.7 .990 M2 29.01 27.72 1.290

Table 3-3. Ct Values for Microcapillary Tube Samples

Subject microRNA-146a microRNA-103-3p ΔCt

A1 30.639 27.5900 3.049 A2 30.555 28.06 2.495 B1 28.531 23.3338 5.197 C1 33.996 33.1189 .877 D1 37.628 31.9406 5.687 E1 29.951 25.3635 4.588 F1 37.303 32.0536 5.249 F2 28.925 25.86 3.065 M1 26.97 24.49 2.480 M2 28.59 26.18 2.410

40

Table 3-4. ANOVA Descriptive Statistics for Full Data Set (Ten Samples)

Measure Condition Mean Standard Deviation

Standard Error

95% Confidence Interval Lower Limit

95% Confidence Interval Upper Limit

miRNA- 146a

PerioPaper 30.186 3.417 1.080 27.741 32.630

Durapore 28.829 2.105 .666 27.323 30.335 Micropipette 31.309 3.735 1.181 28.637 33.981 Total 30.108 3.223 .589 28.904 21.311 miRNA- 103-3p

PerioPaper 28.316 3.874 1.225 25.545 31.088

Durapore 26.257 2.085 .659 24.766 27.748 Micropipette 27.799 3.446 1.090 25.334 30.264 Total 27.458 3.238 .591 26.249 28.667 ΔCt PerioPaper 1.869 2.491 .788 0.087 3.651 Durapore 2.572 1.257 .397 1.673 3.471 Micropipette 3.510 1.578 .499 2.381 4.639 Total 2.650 1.912 .349 1.936 3.364

Table 3-5. ANOVA Descriptive Statistics for Subset of Data (Six Samples)

Measure Condition Mean Standard Deviation

Standard Error

95% Confidence Interval Lower Limit

95% Confidence Interval Upper Limit

miRNA- 146a

PerioPaper 30.671 3.437 1.403 27.064 34.278

Durapore 29.190 2.625 1.071 26.435 31.944 Micropipette 33.008 3.893 1.403 28.923 37.093 Total 30.956 3.546 0.836 29.193 32.719 miRNA- 103-3p

PerioPaper 27.216 2.865 1.170 24.209 30.222

Durapore 25.855 2.538 1.036 23.191 28.519 Micropipette 28.900 4.054 1.170 24.645 33.155 Total 27.324 3.284 0.774 25.691 28.957 ΔCt PerioPaper 3.455 1.233 0.503 2.162 4.749 Durapore 3.335 1.007 0.411 2.278 4.391 Micropipette 4.108 1.831 0.748 2.186 6.030 Total 3.632 1.362 0.321 2.955 4.310

41

Table 3-6. t-test for Treated and Untreated Subjects

Measure Group Mean Standard Deviation

Standard Error

miRNA-146a Untreated 28.866 2.099 0.542 Treated 31.349 3.717 0.960 miRNA-103-3p Untreated 27.486 2.983 0.770 Treated 27.430 3.580 0.920 ΔCt Untreated 1.381 1.574 0.406 Treated 3.919 1.282 0.331

42

CHAPTER 4 DISCUSSION

Candidate biomarkers have been studied in dentistry for many years, with a long-

term goal of identifying biomarkers to aid in early detection of disease or undesired

outcomes.72 Periodontal studies have assessed GCF for potential biomarkers to

indicate periodontal inflammation and periodontal disease.73 Some studies have

identified potential biomarkers including cytokines and miRNAs.72 Significantly, miRNAs

have been detected within extracellular vesicles, revealing their ability transfer genetic

information between cells.74 As shown in a study by Atsawasuwan et al., miRNA-29 was

detected in GCF of orthodontically treated subjects and untreated subject.46

Furthermore, the greatest amount of this miRNA-29 was detected in a portion of the

sample containing EVs.46 Other studies also identified miRNA in exosomes, with studies

by Sun et al. and Li et al. showing the presence of miRNA-214 in exosomes of

osteoclasts.33,45 Holliday et al. showed that miRNA-146a was enriched over 80-fold in

osteoclasts, providing a foundation for further study of miRNA-146a as a candidate

biomarker for bone remodeling processes and orthodontic tooth movement.34

Identifying biomarkers of tooth movement can improve orthodontic diagnosis and

treatment planning. Using a non-invasive method to collect and identify constituents of a

patient’s GCF could aid practitioners in providing the best and most efficient treatment

for the patient. This could be especially beneficial for patients at risk for adverse

treatment outcomes, such as root resorption during orthodontic treatment or ankylosis

of impacted teeth. Furthermore, determining the most cost-effective and reliable method

for sample collection would be beneficial so that in the future these practices can be

easily implemented in a clinical setting and available to practitioners.

43

GCF collection has been done for many years and shows local processes at the

site of collection.51 Studies have used GCF to identify miRNA in EVs and proteins in

EVs.34,75 While many studies use PerioPaper for GCF collection, other techniques have

been suggested, with advantages and disadvantages to each. Therefore, we performed

our study to determine whether miRNA-146a could be adequately identified within GCF

samples and whether a difference in miRNA-146a expression is noted between

collection techniques.

The results of our study show that there is no statistically significant difference in

miRNA-146a expression between the PerioPaper, Durapore, or Microcapillary tube

techniques. Using one-way ANOVA we analyzed the data twice, first with the entire

dataset (Table 3-4) and second with a subset of data (Table 3-5). Statistical analysis of

the full dataset and the subset of data did not yield p-values less than 0.05 for

expression of miRNA-146a and miRNA-103-3p between the three techniques.

Furthermore, the p-value for the normalized ΔCt was not less than the significance

value of 0.05 for the full dataset or subset of data.

Two data sets were analyzed for significance because some subjects were noted

to be in orthodontic treatment. For completion of data analysis, the subset of data

(Table 3-5) did not include samples taken from subjects that were noted at one or both

of their collection visits to be in active orthodontic treatment. Analysis of both data sets

suggests that there is no significant difference in the expression of miRNA-146a

between the three collection methods. With studies often using multiple strips or

micropipettes in one sulcus to obtain GCF, we designed our study as described to be

more clinically applicable.

44

Many studies using GCF for analysis use PerioPaper. These strips come cut,

sterilized, packaged and are easy to insert and remove from the gingival sulcus. If

interested in determining the volume of a sample, a Periotron 8000® (Oraflow,

Plainview, NY, USA) machine can be purchased and used. Durapore filter membrane is

less widely used and must be cut into strips to be used. While this requires more time

for the operator, it is a more economic material. Both PerioPaper and Durapore are

easy to use and comfortable for the subject but due to the nature of the materials,

elution of the sample can lead to particles from both strips entering the media.

Contrasting, a microcapillary tube would collect and elute a more pure sample.

Depending on the volume needed, microcapillary tubes can be purchased with different

diameters. The microcapillary tube has a defined volume and should be an efficient way

to calculate volume. However, GCF volume is minimal and often the tube is not filled

entirely, making it difficult to determine sample volume. Additionally, this technique is

tiring for both the operator and the subject. Collection at each gingival sulcus occurred

for five minutes, which is much longer than the thirty second collection time for

PerioPaper and Durapore strips. The calculated fold change in gene expression of

miRNA-146a was 1.63 for Durapore strips and 3.12 for microcapillary tubes. These

values represent a decrease in fold change expression of miRNA-146a compared to

Periopaper, with the microcapillary tubes showing the largest decrease in fold change of

gene expression. While our results showed no significant difference in miRNA-146a

expression between the three techniques, the micropipettes are more challenging to

use, require more time for GCF collection, and are less clinically applicable than the

PerioPaper and Durapore strips.

45

With some subjects undergoing orthodontic treatment, an additional analysis was

performed to determine whether there was a difference in miRNA-146a expression

between treated and untreated subjects. The data is shown in Table 3-6 as the average

Ct values for each group for miRNA-146a, miRNA-103-3p and ΔCt. Independent-

sample t-tests were performed and the p-values for miRNA-146a and ΔCt are p < .05

and p <.001, respectively. Both p-values are less than the significance value of 0.05,

meaning there is a significant difference in expression of miRNA-146a between treated

and untreated subjects. These results show a significantly greater expression of

miRNA-146a in the treated subjects. The p-value for miRNA-103-3p is greater than 0.05

and, as seen in Table 3-6, the mean Ct values for both groups only differ by 0.056. With

mean values so similar for miRNA-103-3p, it suggests that it is stable and a good choice

for a reference gene. Furthermore, the calculated fold change in gene expression of

miRNA-146a, 5.81, signifies a greater expression of miRNA-146a in subjects

undergoing orthodontic treatment. This study agrees with a study by Holliday et al.,

which detected an 80-fold change in miRNA-146a expression in osteoclasts.34 The

results of this study suggest miRNA-146a as a potential biomarker for orthodontic tooth

movement and is promising for future studies.

A limitation of our study is the small sample size. Increasing the number of

subjects would be beneficial and provide more power to the study. While we aimed to

be more practical in sample collection, the use of two collection strips or tubes per tooth

is still not ideal for a clinical scenario. The use of one strip, or tube, would be more

clinically applicable. RNA isolation, cDNA synthesis, and preparation of the PCR plate

wells are technique sensitive and prone to error and contamination. Error in these

46

processes can affect the quality and quantity of target miRNA expressed. Our study

does include subjects having orthodontic treatment. In retrospect, making treatment an

exclusion criteria would make the data set more homogenous for comparison of GCF

collection techniques.

47

CHAPTER 5 CONCLUSIONS

GCF collection is non-invasive and identification of biomarkers in GCF can

improve treatment planning, assist in early identification of adverse outcomes, and may

lead to therapeutic treatment approaches. The objective of this study was to determine

if miRNA-146a could be sufficiently collected by GCF methods and which method was

best for collecting this miRNA. Results from this study suggest there is no difference in

the PerioPaper, Durapore filter membrane, or microcapillary tube techniques for

expression of miRNA-146a in GCF.

Additional analysis of the data shows a significantly greater expression of

miRNA-146a in subjects undergoing orthodontic treatment compared with untreated

subjects. These results suggest there is promise for identification of miRNA-146a as a

candidate biomarker for orthodontic tooth movement.

48

LIST OF REFERENCES

1. Velu VK, Ramesh R, Srinivasan AR. Circulating MicroRNAs as Biomarkers in Health and Disease. J Clin Diagn Res 2012;6:1791-1795. 2. Vidigal JA, Ventura A. The biological functions of miRNAs: lessons from in vivo studies. Trends Cell Biol 2015;25:137-147. 3. Padmashree D, Ramachandraswamy N. Identification and characterization of conserved miRNAs with its targets mRNA in Trichinella Spiralis. Bioinformation 2016;12:279-284. 4. Pillai RS. MicroRNA function: multiple mechanisms for a tiny RNA? RNA 2005;11:1753-1761. 5. Bhaskaran M, Mohan M. MicroRNAs: history, biogenesis, and their evolving role in animal development and disease. Vet Pathol 2014;51:759-774. 6. Ardekani AM, Naeini MM. The Role of MicroRNAs in Human Diseases. Avicenna J Med Biotechnol 2010;2:161-179. 7. Ha TY. MicroRNAs in Human Diseases: From Cancer to Cardiovascular Disease. Immune Netw 2011;11:135-154. 8. Kamal MA, Mushtaq G, Greig NH. Current Update on Synopsis of miRNA Dysregulation in Neurological Disorders. CNS Neurol Disord Drug Targets 2015;14:492-501. 9. Leung AKL. The Whereabouts of microRNA Actions: Cytoplasm and Beyond. Trends Cell Biol 2015;25:601-610. 10. Macfarlane LA, Murphy PR. MicroRNA: Biogenesis, Function and Role in Cancer. Curr Genomics 2010;11:537-561. 11. Dar D, Sorek R. Bacterial Noncoding RNAs Excised from within Protein-Coding Transcripts. mBio 2018;9. 12. Johanson TM, Lew AM, Chong MM. MicroRNA-independent roles of the RNase III enzymes Drosha and Dicer. Open Biol 2013;3:130144. 13. Wu K, He J, Pu W, Peng Y. The Role of Exportin-5 in MicroRNA Biogenesis and Cancer. Genomics Proteomics Bioinformatics 2018;16:120-126. 14. Sontheimer EJ. Assembly and function of RNA silencing complexes. Nat Rev Mol Cell Biol 2005;6:127-138.

49

15. Wilson RC, Doudna JA. Molecular mechanisms of RNA interference. Annu Rev Biophys 2013;42:217-239. 16. Wei K, Wu L, Chen Y, Lin Y, Wang Y, Liu X et al. Argonaute protein as a linker to command center of physiological processes. Chin J Cancer Res 2013;25:430-441. 17. Ronemus M, Vaughn MW, Martienssen RA. MicroRNA-targeted and small interfering RNA-mediated mRNA degradation is regulated by argonaute, dicer, and RNA-dependent RNA polymerase in Arabidopsis. Plant Cell 2006;18:1559-1574. 18. Pogue AI, Clement C, Hill JM, Lukiw WJ. Evolution of microRNA (miRNA) Structure and Function in Plants and Animals: Relevance to Aging and Disease. J Aging Sci 2014;2. 19. van den Berg A, Mols J, Han J. RISC-target interaction: cleavage and translational suppression. Biochim Biophys Acta 2008;1779:668-677. 20. Aizer A, Shav-Tal Y. Intracellular trafficking and dynamics of P bodies. Prion 2008;2:131-134. 21. Shivdasani RA. MicroRNAs: regulators of gene expression and cell differentiation. Blood 2006;108:3646-3653. 22. Kinney JS, Ramseier CA, Giannobile WV. Oral fluid-based biomarkers of alveolar bone loss in periodontitis. Ann N Y Acad Sci 2007;1098:230-251. 23. d'Apuzzo F, Cappabianca S, Ciavarella D, Monsurrò A, Silvestrini-Biavati A, Perillo L. Biomarkers of periodontal tissue remodeling during orthodontic tooth movement in mice and men: overview and clinical relevance. ScientificWorldJournal 2013;2013:105873. 24. Alhashimi N, Frithiof L, Brudvik P, Bakhiet M. Orthodontic tooth movement and de novo synthesis of proinflammatory cytokines. Am J Orthod Dentofacial Orthop 2001;119:307-312. 25. Kereshanan S, Stephenson P, Waddington R. Identification of dentine sialoprotein in gingival crevicular fluid during physiological root resorption and orthodontic tooth movement. Eur J Orthod 2008;30:307-314. 26. Farahani M, Safavi SM, Dianat O, Khoramian Tusi S, Younessian F. Acid and Alkaline Phosphatase Levels in GCF during Orthodontic Tooth Movement. J Dent (Shiraz) 2015;16:237-245. 27. Lerner UH. Inflammation-induced bone remodeling in periodontal disease and the influence of post-menopausal osteoporosis. J Dent Res 2006;85:596-607.

50

28. Huynh N, VonMoss L, Smith D, Rahman I, Felemban MF, Zuo J et al. Characterization of Regulatory Extracellular Vesicles from Osteoclasts. J Dent Res 2016;95:673-679. 29. Deng L, Wang Y, Peng Y, Wu Y, Ding Y, Jiang Y et al. Osteoblast-derived microvesicles: A novel mechanism for communication between osteoblasts and osteoclasts. Bone 2015;79:37-42. 30. Harding CV, Heuser JE, Stahl PD. Exosomes: looking back three decades and into the future. J Cell Biol 2013;200:367-371. 31. H Rashed M, Bayraktar E, K Helal G, Abd-Ellah MF, Amero P, Chavez-Reyes A et al. Exosomes: From Garbage Bins to Promising Therapeutic Targets. Int J Mol Sci 2017;18. 32. Oliveto S, Mancino M, Manfrini N, Biffo S. Role of microRNAs in translation regulation and cancer. World J Biol Chem 2017;8:45-56. 33. Sun W, Zhao C, Li Y, Wang L, Nie G, Peng J et al. Osteoclast-derived microRNA-containing exosomes selectively inhibit osteoblast activity. Cell Discov 2016;2:16015. 34. Holliday LS, McHugh KP, Zuo J, Aguirre JI, Neubert JK, Rody WJ. Exosomes: novel regulators of bone remodelling and potential therapeutic agents for orthodontics. Orthod Craniofac Res 2017;20 Suppl 1:95-99. 35. Tahamtan A, Teymoori-Rad M, Nakstad B, Salimi V. Anti-Inflammatory MicroRNAs and Their Potential for Inflammatory Diseases Treatment. Front Immunol 2018;9:1377. 36. Saba R, Sorensen DL, Booth SA. MicroRNA-146a: A Dominant, Negative Regulator of the Innate Immune Response. Front Immunol 2014;5:578. 37. Shams K, Kurowska-Stolarska M, Schütte F, Burden AD, McKimmie CS, Graham GJ. MicroRNA-146 and cell trauma down-regulate expression of the psoriasis-associated atypical chemokine receptor ACKR2. J Biol Chem 2018;293:3003-3012. 38. Tang Y, Luo X, Cui H, Ni X, Yuan M, Guo Y et al. MicroRNA-146A contributes to abnormal activation of the type I interferon pathway in human lupus by targeting the key signaling proteins. Arthritis Rheum 2009;60:1065-1075. 39. Pauley KM, Cha S. miRNA-146a in rheumatoid arthritis: a new therapeutic strategy. Immunotherapy 2011;3:829-831. 40. Olsen I, Singhrao SK, Osmundsen H. Periodontitis, pathogenesis and progression: miRNA-mediated cellular responses to. J Oral Microbiol 2017;9:1333396.

51

41. Motedayyen H, Ghotloo S, Saffari M, Sattari M, Amid R. Evaluation of MicroRNA-146a and Its Targets in Gingival Tissues of Patients With Chronic Periodontitis. J Periodontol 2015;86:1380-1385. 42. Xie YF, Shu R, Jiang SY, Liu DL, Ni J, Zhang XL. MicroRNA-146 inhibits pro-inflammatory cytokine secretion through IL-1 receptor-associated kinase 1 in human gingival fibroblasts. J Inflamm (Lond) 2013;10:20. 43. Kuang W, Zheng L, Xu X, Lin Y, Lin J, Wu J et al. Dysregulation of the miR-146a-Smad4 axis impairs osteogenesis of bone mesenchymal stem cells under inflammation. Bone Res 2017;5:17037. 44. Nakasa T, Shibuya H, Nagata Y, Niimoto T, Ochi M. The inhibitory effect of microRNA-146a expression on bone destruction in collagen-induced arthritis. Arthritis Rheum 2011;63:1582-1590. 45. Li D, Liu J, Guo B, Liang C, Dang L, Lu C et al. Osteoclast-derived exosomal miR-214-3p inhibits osteoblastic bone formation. Nat Commun 2016;7:10872. 46. Atsawasuwan P, Lazari P, Chen Y, Zhou X, Viana G, Evans CA. Secretory microRNA-29 expression in gingival crevicular fluid during orthodontic tooth movement. PLoS One 2018;13:e0194238. 47. Chen J, Li K, Pang Q, Yang C, Zhang H, Wu F et al. Identification of suitable reference gene and biomarkers of serum miRNAs for osteoporosis. Sci Rep 2016;6:36347. 48. Yao Y, Suo AL, Li ZF, Liu LY, Tian T, Ni L et al. MicroRNA profiling of human gastric cancer. Mol Med Rep 2009;2:963-970. 49. Song J, Bai Z, Han W, Zhang J, Meng H, Bi J et al. Identification of suitable reference genes for qPCR analysis of serum microRNA in gastric cancer patients. Dig Dis Sci 2012;57:897-904. 50. Gupta S, Chhina S, Arora SA. A systematic review of biomarkers of gingival crevicular fluid: Their predictive role in diagnosis of periodontal disease status. J Oral Biol Craniofac Res 2018;8:98-104. 51. Barros SP, Williams R, Offenbacher S, Morelli T. Gingival crevicular fluid as a source of biomarkers for periodontitis. Periodontol 2000 2016;70:53-64. 52. Lamster IB. Evaluation of components of gingival crevicular fluid as diagnostic tests. Ann Periodontol 1997;2:123-137. 53. Griffiths GS. Formation, collection and significance of gingival crevice fluid. Periodontol 2000 2003;31:32-42.

52

54. Del Fabbro M, Galardi E, Weinstein R, Bulfamante G, Miserocchi G. Fluid dynamics of gingival tissues. J Periodontal Res 1998;33:328-334. 55. Pashley DH. A mechanistic analysis of gingival fluid production. J Periodontal Res 1976;11:121-134. 56. Rody WJ, Wijegunasinghe M, Holliday LS, McHugh KP, Wallet SM. Immunoassay analysis of proteins in gingival crevicular fluid samples from resorbing teeth. Angle Orthod 2016;86:187-192. 57. Bevilacqua L, Biasi MD, Lorenzon MG, Frattini C, Angerame D. Volumetric Analysis of Gingival Crevicular Fluid and Peri-Implant Sulcus Fluid in Healthy and Diseased Sites: A Cross-Sectional Split-Mouth Pilot Study. Open Dent J 2016;10:131-138. 58. Drummond S, Canavarro C, Perinetti G, Teles R, Capelli J. The monitoring of gingival crevicular fluid volume during orthodontic treatment: a longitudinal randomized split-mouth study. Eur J Orthod 2012;34:109-113. 59. Rody WJ, Wijegunasinghe M, Wiltshire WA, Dufault B. Differences in the gingival crevicular fluid composition between adults and adolescents undergoing orthodontic treatment. Angle Orthod 2014;84:120-126. 60. Khurshid Z, Mali M, Naseem M, Najeeb S, Zafar MS. Human Gingival Crevicular Fluids (GCF) Proteomics: An Overview. Dent J (Basel) 2017;5. 61. Subbarao KC, Nattuthurai GS, Sundararajan SK, Sujith I, Joseph J, Syedshah YP. Gingival Crevicular Fluid: An Overview. J Pharm Bioallied Sci 2019;11:S135-S139. 62. Sueda T, Bang J, Cimasoni G. Collection of gingival fluid for quantitative analysis. J Dent Res 1969;48:159. 63. Guentsch A, Kramesberger M, Sroka A, Pfister W, Potempa J, Eick S. Comparison of gingival crevicular fluid sampling methods in patients with severe chronic periodontitis. J Periodontol 2011;82:1051-1060. 64. Johnson RB, Streckfus CF, Dai X, Tucci MA. Protein recovery from several paper types used to collect gingival crevicular fluid. J Periodontal Res 1999;34:283-289. 65. Garibyan L, Avashia N. Polymerase chain reaction. J Invest Dermatol 2013;133:1-4. 66. Goate AM. Molecular Biology. Alcohol Health Res World 1995;19:217-220. 67. Espy MJ, Uhl JR, Sloan LM, Buckwalter SP, Jones MF, Vetter EA et al. Real-time PCR in clinical microbiology: applications for routine laboratory testing. Clin Microbiol Rev 2006;19:165-256.

53

68. Ruijter JM, Lorenz P, Tuomi JM, Hecker M, van den Hoff MJ. Fluorescent-increase kinetics of different fluorescent reporters used for qPCR depend on monitoring chemistry, targeted sequence, type of DNA input and PCR efficiency. Mikrochim Acta 2014;181:1689-1696. 69. Pabinger S, Rödiger S, Kriegner A, Vierlinger K, Weinhäusel A. A survey of tools for the analysis of quantitative PCR (qPCR) data. Biomol Detect Quantif 2014;1:23-33. 70. Karlen Y, McNair A, Perseguers S, Mazza C, Mermod N. Statistical significance of quantitative PCR. BMC Bioinformatics 2007;8:131. 71. Livak KJ, Schmittgen TD. Analysis of relative gene expression data using real-time quantitative PCR and the 2(-Delta Delta C(T)) Method. Methods 2001;25:402-408. 72. Branca F, Hanley AB, Pool-Zobel B, Verhagen H. Biomarkers in disease and health. Br J Nutr 2001;86 Suppl 1:S55-92. 73. Jaedicke KM, Preshaw PM, Taylor JJ. Salivary cytokines as biomarkers of periodontal diseases. Periodontol 2000 2016;70:164-183. 74. Di Liegro CM, Schiera G, Di Liegro I. Extracellular Vesicle-Associated RNA as a Carrier of Epigenetic Information. Genes (Basel) 2017;8. 75. Rody WJ, Chamberlain CA, Emory-Carter AK, McHugh KP, Wallet SM, Spicer V et al. The proteome of extracellular vesicles released by clastic cells differs based on their substrate. PLoS One 2019;14:e0219602.

54

BIOGRAPHICAL SKETCH

Kelsey Cronauer Wahl was born and raised in South Florida by her parents

Edward and Julie Cronauer. She completed her undergraduate education at Vanderbilt

University where she majored in History of Art with a Pre-Dental concentration. She

pursued a career in dentistry and was accepted into the University of Florida College of

Dentistry. Upon graduating, she began her specialty training in orthodontics at the

University of Florida College of Dentistry, Department of Orthodontics. In April 2019 she

married her wonderful husband, Tyler Wahl. Her husband, brother, Edward, and

parents, Edward and Julie, have supported her every step of the way. Kelsey plans to

practice orthodontics in Florida after graduation in May 2020.