Implementation of a drug discovery tool for the evaluation of anti fibrotic compounds application in fibrovascular disorders

List of Tables Table 3.1 The filter configurations for Olympus IX-71 fluorescence microscope………27 Table 3.2 3-5% polyacrylamide gel composition……….34 Table 4.1 LOD and LOQ of the cell en

IMPLEMENTATION OF A DRUG DISCOVERY TOOL

FOR THE EVALUATION OF ANTI-FIBROTIC

GRADUATE PROGRAM IN BIOENGINEERING

NATIONAL UNIVERSITY OF SINGAPORE

2006

Acknowledgements

I would like to express my sincere gratitude to my supervisor, Associate Professor

Michael Raghunath, for his supervision and for sharing his invaluable experience

during the course of my graduate study I greatly appreciate his guidance in the

research works as well as our informal discussios Sincere thanks to my co-supervisor,

Dr Phan Toan Thang, for his valuable feedback in this project

I would like to extend my gratitude to the TML members (Ricardo Rodolfo Lareu,

Dimitrios Zevgolis, Wong Yuensy, Wang Zhibo, Harve Subramhanya Karthik, Kou

Shanshan and the attachment students: Shriju Joshi, Amelia Ann Michael, Natasha

Lee, Brenda Lim, Rosanna Chau, Srividia Sundararaman, Zhang Lei) and the Skin

Cells Research Group members (Anandaroop Mukhopadhyay and Audrey Khoo)

Without their help, support and fruitful discussions, this thesis would not be possible

I would also like to acknowledge the final year students: Lin Gen, Yin Jing, Yanxian,

Choo Liling, for making my stay in the laboratory enjoyable

Last but not least, I would like to thank all the support staff in National University of

Singapore Tissue Engineering Program (NUSTEP), Tissue Engineering Laboratory

and Tissue Repair Laboratory (TRL) for their assistance in this project and Rainbow

Instrument (Singapore) for their excellent service for the microplate readers

Table of Contents

Acknowledgement……… i

Table of Contents ………ii

Summary………v

List of Tables………vii

List of Figures………x

Chapter 1 Introduction 1

1.1 Project background and significance……… 1

1.2 Objectives……… 3

Chapter 2 Literature Review… ……….4

2.1 Establishment of the cell enumeration assay……… 4

2.1.1 Various cell enumeration assays………4

2.1.2 Quantification of cell numbers with DAPI……….8

2.2 Enhancement of the collagen matrix in fibroblast cultures………… 10

2.2.1 Collagen properties……… 11

2.2.2 Collagen biosynthesis……… 12

2.2.3 The challenge in the enhancement of collagen matrix formation in fibroblast culture……… 16

2.2.4 Macromolecular crowding……… 18

2.2.5 The ideal crowding agent……….21

2.2.6 Macromolecular crowding in collagen matrix formation…….23

2.3 Exploration of quantitative immunocytochemistry for collagen quantification……… ……… 24

Chapter 3 Experimental Details………25

3.1 Equipments……… 25

3.2 Materials……… 29

3.3 Experimental procedures……… 33

3.3.1 General methods……… 33

3.3.2 Cell enumeration assays……… 37

3.3.3 Enhancement of the collagen matrix in fibroblast culture 38

3.3.4 Collagen quantification assay based on immunocytochemistry ……… ………44

Chapter 4 Results and Discussion……….45

4.1 Establishment of the cell enumeration assay………45

4.1.1 Cell enumeration with DAPI………45

4.1.2 Comparison with MTT cell viability assay……… 49

4.2 Enhancement of the collagen matrix in fibroblast cultures………….54

4.2.1 Collagen isolation with pepsin digestion……….54

4.2.2 The trial of DexS on various fibrogenic fibroblast cell lines 55

4.2.3 Optimization of the DexS concentration in the fibroblast culture……… 61

4.2.4 Conversion of procollagen to collagen in the presence of DexS……….66

4.2.5 Collagen crosslinking in the presence of DexS………68

4.2.6 The cell surface influence on the collagen matrix deposition 69

4.3 Exploration of quantitative immunocytochemistry for collagen quantification………75

Chapter 5 Conclusions……… 78

References………80

Appendix A The calculation for Limit of Detection (LOD) and Limit of Quantification (LOQ)……… 85

Appendix B The establishment of the cell enumeration assays (the complete results)……… 86

Appendix C Enhancement of the collagen matrix in fibroblast cultures (the

complete results)……… 98 Appendix D Exploration of quantitative immunocytochemistry for collagen

quantification (the complete results)……… 109

Summary

In this project, the principle of macromolecular crowding was applied in a fibroblast

culture system to enhance the formation of collagen matrix We have successfully

demonstrated that dextran sulfate (DexS), a polyanionic macromolecule, creates a

volume exclusion effect in the culture medium, and thus accelerates the enzymatic

processing of procollagen to collagen, and its subsequent deposition

Gel electrophoresis and Western blotting revealed that in normal fibroblast culture,

most collagen remained in the culture medium in its unprocessed form The addition

of DexS resulted in the conversion of procollagen to collagen, and the subsequent

association of collagen with the cell layer This observation was confirmed with

immunocytochemistry Remarkably, the crowding effect did not seem to alter the

expression level of fibronectin, one of the ECM components However, we observed

re-arrangement of ECM and co-localization of fibronectin with collagen, as compared

to conventional culture system without DexS The optimum concentration of DexS

was found to be 50-100 µg/ml

In addition, we were able to show the presence of intensified collagen crosslinking in

our culture system This demonstrated that the specific collagen crosslinking enzyme,

lysyl oxidase, was accelerated resulting in the formation of crosslinked collagen

matrix in the presence of DexS

This project also aimed to develop an anti-fibrotic drug discovery tool that was

fluorometric-based and employed a microplate reader as the quantification device

This tool integrated the cell enumeration assay and the collagen quantification assay

on one plate We have successfully developed a cell enumeration assay that was based

on the measurement of DAPI-stained nuclei The fluorescence detected by the

microplate readers, FLUOstar and PHERAstar, was then correlated with the cell

seeding density The calibration curves from both readers showed good linearity

throughout the tested concentration range; 5000 to 200,000 cells/well for 24-well and

100 to 40,000 cells/well for 96-well plate In addition, the comparison with MTT

assay, an established cell viability assay, showed that the DAPI staining method is

comparable or even superior in sensitivity These results indicated that this method

was a suitable cell enumeration assay for the drug discovery tool for adherent cells in

monolayer culture in a screening setting

The collagen quantification assay was based on an immunocytochemistry technique

The ECM proteins (collagen and fibronectin) were labeled with specific antibodies

that were indirectly conjugated with fluorochromes The fluorescence measurement

showed confirmatory results to that obtained with the gel electrophoresis and

immunocytochemistry staining This preliminary result demonstrated that this method

is a potentially suitable collagen quantification assay as a building block of a

discovery tool for anti-fibrotic drugs

List of Tables

Table 3.1 The filter configurations for Olympus IX-71 fluorescence

microscope………27 Table 3.2 3-5% polyacrylamide gel composition……….34

Table 4.1 LOD and LOQ of the cell enumeration assays using DAPI staining and

MTT assay………49 Table 4.2 Comparison between DAPI staining method and MTT assay…… 52

Table 4.3 The densitometry analysis for the optimization of DexS concentration

in WI-38 fibroblast culture……… 64

Table 4.4 The densitometry analysis for the optimization of DexS concentration

in HSF culture……… 64

Table B.1 The PHERAstar reading result for DAPI-stained WI-38 cells plated on

a lumox™ 24-well plate (first experiment)……… 86 Table B.2 The PHERAstar reading result for DAPI-stained WI-38 cells plated on

a lumox™ 24-well plate (second experiment)……… 87 Table B.3 The FLUOstar reading result for DAPI-stained WI-38 cells plated on a

lumox™ 24-well plate (first experiment)……… 88 Table B.4 The FLUOstar reading result for DAPI-stained WI-38 cells plated on a

lumox™ 24-well plate (second experiment)……….89 Table B.5 The PHERAstar reading result for DAPI-stained WI-38 cells plated on

a lumox™ 96-well plate (first experiment)……… 90 Table B.6 The PHERAstar reading result for DAPI-stained WI-38 cells plated on

a lumox™ 96-well plate (second experiment)……… 91 Table B.7 The FLUOstar reading result for DAPI-stained WI-38 cells plated on a

lumox™ 96-well plate (first experiment)……… 92

Table B.8 The FLUOstar reading result for DAPI-stained WI-38 cells plated on a

lumox™ 96-well plate (second experiment)……….93

Table B.9 The absorbance reading result for WI-38 cell density, plated on

24-well plate, as determined with MTT assay (first experiment)……… 94 Table B.10 The absorbance reading result for WI-38 cell density, plated on 24-

well plate, as determined with MTT assay (second experiment)…….95

Table B.11 The absorbance reading result for WI-38 cell density, plated on

96-well plate, as determined with MTT assay (first experiment)……… 96 Table B.12 The absorbance reading result for WI-38 cell density, plated on 96-

well plate, as determined with MTT assay (second experiment)…….97 Table C.1 The densitometry quantification results for the SDS-PAGE of the cell

layer and the medium fraction from pepsin digested HSF culture… 98 Table C.2 The densitometry quantification results for the SDS-PAGE of the cell

layer and the medium fraction from pepsin digested WI-38 fibroblast culture……… 99 Table C.3 The FLUOstar reading result for DAPI-stained WI-38 cells after 5 days

of culture in the presence of 100 mM AscP and DexS at various concentrations (first experiment)………100

Table C.4 The FLUOstar reading result for DAPI-stained WI-38 cells after 5 days

of culture in the presence of 100 mM AscP and DexS at various concentrations (second experiment)……… 101

Table C.5 The densitometry quantification results for the optimization of DexS

concentration in WI-38 fibroblast culture……… 102

Table C.6 The densitometry quantification results for the optimization of DexS

concentration in HSF culture……… 104

Table C.7 The densitometry quantification result for the inhibition of lysyl

oxidase with β-APN (Sample 1)………105 Table C.8 The densitometry quantification result for the inhibition of lysyl

oxidase with β-APN (Sample 2)………106

Table C.9 The densitometry quantification result (based on α1(I) intensity) for the

deposition of collagen following the addition of the DexS to the fibroblast culture (Sample 1)……… 107

Table C.10 The densitometry quantification result (based on α1(I) intensity) for the

deposition of collagen following the addition of the DexS to the fibroblast culture (Sample 2)……… 108 Table D.1 The PHERAstar reading result for collagen, fibronectin and DAPI

staining on WI-38 fibroblasts that were cultured in the presence of 100

µM AscP and DexS at various concentrations for 5 days (first experiment)……….109 Table D.2 The expression level of collagen and fibronectin of WI-38 fibroblasts

normalized with the cell population (first experiment)……… 111

Table D.3 The PHERAstar reading result for collagen, fibronectin and DAPI

staining on WI-38 fibroblasts that were cultured in the presence of 100

µM AscP and DexS at various concentrations for 5 days (second experiment)………112 Table D.4 The expression level of collagen and fibronectin of WI-38 fibroblasts

normalized with the cell population (second experiment)………….114

List of Figures

Figure 2.1 The molecular structure of MTT and the corresponding reaction

product, formazan……… 5 Figure 2.2 The chemical structure of DAPI……….9

Figure 2.3 Excitation and emission profiles of DAPI bound to dsDNA………….9

Figure 2.4 The structure of type I procollagen……… 11

Figure 2.5 The collagen synthesis, processing and assembly………16

Figure 2.6 The illustration of the crowding condition in eukaryotic cytoplasm…18

Figure 2.7 The schematic drawing to illustrate the concept of exclusion volume.19

Figure 2.8 Schematic depiction of the predicted dependence of reaction rate on

the concentration of crowding agent………21 Figure 2.9 The structure of DexS with sodium salt……… 22

Figure 3.1 The schematic drawing of the first experiment set-up to study the cell

surface influence on the collagen deposition……… 42 Figure 3.2 The schematic drawing of the second experiment set-up to study the

cell surface influence on the collagen deposition……….44 Figure 4.1 Nuclear staining with DAPI observed under fluorescence

microscope………45 Figure 4.2 The calibration curves for WI-38 cell density, plated on a Lumox™ 24-

well plate, as quantified with fluorescence microplate readers………46 Figure 4.3 The calibration curves for WI-38 cell density, plated on a Lumox™ 96-

well plate, as quantified with fluorescence microplate readers………47 Figure 4.4 The calibration curves for WI-38 cell density, plated on 24-well plate,

as determined with MTT assay………50

Figure 4.5 The calibration curve for WI-38 cell density, plated on 96-well plate,

as determined with MTT assay………51 Figure 4.6 Pepsin digested all proteins except collagen………54

Figure 4.7 SDS-PAGE of the cell layer and the medium fraction from pepsin

digested HSF culture showing collagen bands and the corresponding densitometry analysis……….……… 55

Figure 4.8 SDS-PAGE of the cell layer and the medium fraction from pepsin

digested WI-38 fibroblast culture showing collagen bands and the corresponding densitometry analysis……… 56

Figure 4.9 Immunocytochemistry for collagen I (green) and fibronectin (red) on

HDF……… 58

Figure 4.10 Immunocytochemistry for collagen I (green) and fibronectin (red) on

WI-38 fibroblasts……… 59

Figure 4.11 WI-38 cell number as quantified with DAPI staining method, after 5

days of treatment with 100µM AscP and DexS at various concentrations……… 61

Figure 4.12 SDS-PAGE of the cell layer and the medium fraction of pepsin

digested WI-38 culture……….63 Figure 4.13 SDS-PAGE of the cell layer and the medium fraction of pepsin

digested HSF culture………64 Figure 4.14 The WI-38 fibroblasts morphology observed under phase contrast

microscope………65 Figure 4.15 The HSF morphology observed under phase contrast microscope …66

Figure 4.16 Western blotting of the medium and the cell layer fraction of HSF

culture……… 67

Figure 4.17 The schematic drawing of the effect of volume exclusion in the

distribution of procollagen and proteinases in the solution………… 68 Figure 4.18 The inhibition of the lysyl oxidase by β-APN……….69

Figure 4.19 The effect of DexS on the deposition of exogenous collagen……… 71

Figure 4.20 The deposition of collagen following the addition of the DexS to the

fibroblast culture……… 73 Figure 4.21 The densitometry analysis and the corresponding graphs of the

deposition of collagen following the addition of the DexS to the fibroblast culture……… 74 Figure 4.22 The expression level of collagen and fibronectin and the cell

population of WI-38 fibroblasts as detected with immunocytochemistry and quantified with PHERAstar microplate reader………77 Figure B.1 The calibration curve for WI-38 cell density, plated on a lumox™ 24-

well plate, as quantified with PHERAstar (first experiment)……… 86

Figure B.2 The calibration curve for WI-38 cell density, plated on a lumox™

24-well plate, as quantified with PHERAstar (second experiment)…… 87 Figure B.3 The calibration curve for WI-38 cell density, plated on a lumox™ 24-

well plate, as quantified with FLUOstar (first experiment)………….88 Figure B.4 The calibration curve for WI-38 cell density, plated on a lumox™ 24-

well plate, as quantified with FLUOstar (second experiment)……….89 Figure B.5 The calibration curve for WI-38 cell density, plated on a lumox™ 96-

well plate, as quantified with PHERAstar (first experiment)……… 90 Figure B.6 The calibration curve for WI-38 cell density, plated on a lumox™ 96-

well plate, as quantified with PHERAstar (second experiment)…… 91 Figure B.7 The calibration curve for WI-38 cell density, plated on a lumox™ 96-

well plate, as quantified with FLUOstar (first experiment)………….92 Figure B.8 The calibration curve for WI-38 cell density, plated on a lumox™ 96-

well plate, as quantified with FLUOstar (second experiment)……….93

Figure B.9 The calibration curve for WI-38 cell density, plated on 24-well plate,

as determined with MTT assay (first experiment)……… 94

Figure B.10 The calibration curve for WI-38 cell density, plated on 24-well plate,

as determined with MTT assay (second experiment)……… 95

Figure B.11 The calibration curve for WI-38 cell density, plated on 96-well plate,

as determined with MTT assay (first experiment)……… 96

Figure B.12 The calibration curve for WI-38 cell density, plated on 96-well plate,

as determined with MTT assay (second experiment)……… 97

Figure C.1 The WI-38 cell viability after 5 days of treatment with 100µM AscP

and DexS at various concentrations (first experiment)……… 100

Figure C.2 The WI-38 cell viability after 5 days of treatment with 100µM AscP

and DexS at various concentrations (second experiment)………… 101

Figure D.1 The expression level of collagen and fibronectin and the cell

population of WI-38 fibroblasts as detected with immunocytochemistry and quantified using PHERAstar microplate reader (first experiment)……….110 Figure D.2 The expression level of collagen and fibronectin of WI-38 fibroblasts

normalized with the cell population (first experiment)……… 111 Figure D.3 The expression level of collagen and fibronectin and the cell

population of WI-38 fibroblasts as detected with

immunocytochemistry and quantified using PHERAstar microplate reader (second experiment)………113 Figure D.4 The expression level of collagen and fibronectin of WI-38 fibroblasts

normalized with the cell population (second experiment)………….114

1 Introduction

In this chapter, the background and significance of the project is covered in the first

section The second section presents the objectives of the project and the outline of

the report

1.1 Project background and significance

Fibrovascular disorders are diseases that are characterized by an increase in the

formation of fibrous tissues and its vascularization This disease can occur in any part

of the body, both external and internal, such as skin, eye, joints, lung or liver The

severity of the disease may vary from merely pain and pruritis to functional disability

or even fatality The widespread disposition of this disease thus calls for the search of

anti-fibrotic drugs

Fibrovascular disorders are marked with excessive proliferation of fibroblasts that

produce massive amount of connective tissues, particularly collagen It is usually

preceded with inflammatory reaction and followed by vascularization of the affected

tissues This phenomenon is similar to that found during wound healing process,

which is essentially reflected in the in vitro culture of fibroblasts Therefore, the

assessment of anti-fibrotic drugs usually employs fibroblast culture to determine the

effect of the drug on the cell proliferation and collagen deposition Unfortunately, the

assessment of collagen deposition hitherto depends on the quantification of the

soluble procollagen (collagen precursor) in the culture medium This is due to the

slow processing of procollagen to collagen in in vitro environment that results in

minute amount of collagen matrix and abundance of unprocessed procollagen The

amount of this precursor protein is not necessarily equal to the amount of collagen

fibrils in the extracellular matrix that would be the proper measure of fibrosis

Therefore, this project aims to enhance the formation of collagen matrix in in vitro

fibroblast culture, to allow an accurate assessment of anti-fibrotic drugs on collagen

deposition

As another aspect, the enhancement of the collagen matrix in fibroblast culture may

be a valuable tool in tissue engineering The presence of collagen fibrils in tissue

constructs is essential in maintaining the mechanical strength and defining the shape

and form of the tissues Unfortunately, due to the above mentioned procollagen

processing setback in vitro, the construction of engineered tissues using fibroblasts

has been carried out with suboptimal amounts of endogenous collagen matrix

Therefore, the accomplishment of this project may open an avenue to create fully

functional tissue constructs In addition, collagen has been used as a scaffold or

coating in tissue engineering This collagen mainly comes from various animal

origins In the light of increasing concerns over animal-transmitted diseases, a

collagen scaffold fabricated from human fibroblasts might offer an alternative to

animal-originated collagen

This project also aimed to develop an anti-fibrotic drug discovery tool that allows the

integration of cell enumeration and collagen quantification Quantification of collagen

has been classically done using a radioactive method (metabolic labeling) or

colorimetric assays, both having several disadvantages The first method can be

laborious and involves hazardous materials and wastes, whereas the latter is

non-specific to collagen Therefore, an alternative collagen quantification assay that is

based on immunocytochemistry was explored It was integrated with a cell

enumeration method that measures the DNA content of the cell population on test

Both assays are fluorometry-based and the measurement was done using a microplate

reader

1.2 Objectives

This project has a main objective of establishing a drug discovery tool that can be

applied for the assessment of anti-fibrotic compounds It can be divided into three

sub-aims that provide comprehensive assessments for the fulfillment of the main

objective:

1 to establish a rapid cell enumeration assay that is suitable for the quantification of

fibroblast population and suitable for a microplate reader analysis,

2 to enhance the formation of collagen matrix on the fibroblast culture,

3 to develop collagen quantification assay that is based on a non-radioactive method

and suitable for a microplate reader analysis

2 Literature Review

This chapter covers the theoretical background for the project, which is divided into

three sections; the first is related to the establishment of the cell enumeration assay,

the second is related to the enhancement of the collagen matrix on the fibroblast

culture, and the third is related to the exploration of the collagen quantification assay

2.1 Establishment of the cell enumeration assay

There are currently many cell enumeration assays, which are based on either cell

viability or cell proliferation, available in the market Much of them will be discussed

in the first part of this section However, despite being convenient, these assays may

be implicated with several disadvantages particularly in the application for an

anti-fibrotic drug discovery tool Therefore, this project aims to establish a reliable cell

enumeration assay that is suitable to quantify the fibroblast population and can be

incorporated into the anti-fibrotic drug discovery tool This assay will be based on the

quantification of the nuclear content using DNA-binding dye, DAPI, and fluorescent

measurement using a microplate reader

2.1.1 Various cell enumeration assays

Conventional cell counting method using a hemacytometer

A hemacytometer is a simple device that consists of two fields, each of which is

divided into nine 1.0 mm2 squares A cover glass is placed on top creating a chamber

with a depth of 0.1 mm and a volume of 0.1 mm3 (= 10-4 ml) for each square

This method is the most commonly used method to determine the number of viable

cells Usually the dead cells are stained with trypan blue dye, leaving the cells with

uncompromised membrane integrity unstained The cell suspension is introduced into

the hemacytometer chamber and subsequently placed under a microscope for cell

counting Unfortunately, this method has a significant accuracy error due to its

subjective nature Different persons analyzing the same cell population will obtain

varying results This error will be more significant when the number of cells to be

counted is small Therefore this method usually only applies in determination of the

cell concentration in batch cultures Furthermore, counting the cells manually can be

laborious and time consuming

Assays that measure metabolic activity

Metabolic activity can be an indication of cell viability There are several metabolic

based assays available in the market, MTT assay being the commonly used This

assay is based on the reduction of tetrazolium salts to a colored, water-insoluble

formazan that can be quantified in a conventional ELISA plate reader at 570 nm

(maximum absorbance) after solubilization There are currently modified tetrazolium

salts, for instance XTT and WST-1 that will be converted by the viable cells to

water-soluble formazan, therefore eliminating the solubilization step

Figure 2.1 The molecular structure of MTT and the corresponding reaction product,

formazan (taken from Apoptosis, cell death and cell proliferation, Roche)

This assay is relatively simple and convenient The complete assay starting from the

cell culture to the absorbance measurement can be carried out on the same microplate

However, the cellular metabolic activity is not always equal to the number of viable

cells The metabolic activity of different cell lines may differ resulting in the variation

of the cells’ response to tetrazolium salts In addition, even for a certain type of cell,

this response may vary depending on the metabolic state of the viable cell that is

influenced by the culture condition, such as pH or D-glucose concentration in the

culture medium or the presence of additional substance such as drugs (Shappell,

2003) Therefore, metabolic-based cell viability assay may not be suitable for the drug

discovery tool since the drug tested may cause alteration in the metabolic state of the

cells

Assays that measure cell proliferation

There are several methods to measure cell proliferation, and DNA synthesis is the

common method since cellular proliferation requires the replication of cellular DNA

Labeled nucleotides are added to the culture and will be incorporated into the DNA of

the dividing cells Traditionally, this assay involves the use of radiolabeled nucleotide,

tritiated thymidine ([3H]-TdR) Alternatively, thymidine analogues, for instance

5-bromo-2’-deoxy-uridine (BrdU), are used The incorporated BrdU is detected

immunochemically using a specific ELISA

The complete assay from the start of the cell culture to the ELISA measurement can

be performed in the same microplate, making it a convenient assay Unfortunately,

this cell proliferation assay can only capture the cells that replicate within the time

window of incubation This assay fails to include quiescent cells, therefore the result

does not represent the whole population of a cell culture

ATP-based assays

The nucleotide adenosine 5'-triphosphate (ATP) plays a dominant role in energy

exchange processes in biological systems The presence of ATP is also a useful

marker for cell proliferation An increase in the ATP level is associated with cell

proliferation, whilst cell death exhibits decrease in the ATP level The commonly

used ATP detection method is the chemiluminescent detection of luciferase

Luciferase is the catalyst for the reaction between luciferin and ATP This reaction

produces light as a side product that can be measured using a luminometer

This assay can be performed on the same plate as the culture plate and it is relatively

fast However, this assay has to be completed immediately since ATP does not

survive long storage

ATP-based cell proliferation assay might not be suitable when drug treatment is

involved The increase or decrease in the ATP concentration may not necessarily

related to the cell number in this case since the drug might alter the biological

function of the cells

Assays that measure the cellular DNA quantity

The determination of DNA concentration is a reasonable indicator of cell number,

since the levels of DNA and RNA in cells are tightly regulated (Frankfurt, 1980)

Although the levels of DNA and RNA in individual cells can vary significantly over

time, the overall amount of nucleic acid in a given cell population will not change, as

long as the cells are asynchronous Additionally, assays based on nucleic acid binding

are generally independent of changes in cellular metabolism The most common

technique to measure nucleic acid concentration is the absorbance determination at

260 nm This method is simple and easy, however, it is insensitive with a detection

limit of double-stranded DNA (dsDNA) in µg/ml range (Rengarajan et.al., 2002)

Moreover, it does not distinguish nucleotides, single-stranded DNA, contaminants and

RNA

Recently, the use of DNA-binding dyes to measure DNA concentration has gained

popularity recently because it is simple and potentially more sensitive than

absorbance measurement (Noites et.al., 1998) There are many DNA-binding dyes

available, for instance picogreen, Hoechst, DAPI, ethidium bromide, SYBR and many

more Several studies have shown that picogreen is an ultrasensitive fluorescent

nucleic acid stain for quantification of dsDNA in solution, with a detection limit in the

range of pg/ml dsDNA (Rengarajan et.al., 2002; Singer et.al., 1997)

Unfortunately, DNA quantification method usually involves trypsinization and

complete disruption of the cells to get the nuclear contents out It is not desirable in

the anti-fibrotic drug discovery tool, since trypsinization and cell lysing does not

allow further processing of the culture such as collagen extraction

2.1.2 Quantification of cell numbers with DAPI

A novel cell enumeration assay based on the DNA quantification method was

developed in this project Nucleic acid dye, 4’,6-Diamidino-2-phenylindole (DAPI),

was used to stain the cellular DNA in situ and the fluorescence signal was quantified

by a microplate reader This method is extremely rapid and simple Moreover, the cell

layer remains intact and fixed on the plate allowing further processing of the samples

DAPI is a popular nuclear counterstain when multicolor fluorescent probes are used to

stain cellular structures It emits blue fluorescence that stands out in vivid contrast to

red or green fluorescent probes The maximum excitation and emission of DAPI when

it is bound to dsDNA is 358 nm and 461 nm respectively

Figure 2.2 The chemical structure of DAPI

Figure 2.3 Excitation and emission profiles of DAPI bound to dsDNA (taken from MolecularProbes DAPI datasheet)

DAPI preferentially stains dsDNA and appears to associate with AT

(Adenine-Thymidine) clusters in the minor groove (Kubista et.al., 1987) Binding of DAPI to

dsDNA produces a ~20-fold fluorescence enhancement, apparently due to the

displacement of water molecules from both DAPI and the minor groove (Barcellona

et.al., 1990)

DAPI can also bind RNA However, it is thought that DAPI/RNA binding mode

diverse from that of DAPI/dsDNA The DAPI/RNA involves AU-selective

intercalation instead of binding at the AT cluster (Tanious et.al., 1992) In addition,

the DAPI/RNA complex exhibits a longer-wavelength fluorescence emission

maximum (~500 nm) than the DAPI/dsDNA complex (~460 nm) and a quantum yield

that is only about 20% as high (Kapuscinski, 1990)

2.2 Enhancement of the collagen matrix in fibroblast cultures

Among other important properties, the extracellular matrix (ECM) provides a

scaffolding structure to which cells are attached within tissues Collagen is the major

component of the ECM and the most abundant protein in human body Collagen

fibrils play an important role in maintaining the mechanical strength in tissues and

define the shape and form of tissues in which they occur Appropriate mechanical

strength is also essential for tissue-engineered constructs, especially when it is

intended to resist significant mechanical stresses upon implantation into the body,

such as tissue-engineered arterial conduit (Johnson and Galis, 2003)

In tissues, collagen is synthesized and secreted by fibroblasts forming a matrix of

insoluble crosslinked collagen fibrils These mesenchymal cells are thus tightly

surrounded by ECM In in vitro culture, however, fibroblasts do not produce

sufficient collagen matrix Instead, these cells continuously secrete a large amount of

soluble collagen precursors into the culture medium This culture condition is

evidently not an ideal system to investigate the regulation of collagen production by

anti-fibrotic drugs

This project therefore aimed to enhance the formation of collagen matrix in fibroblast

cultures by using the principle of macromolecule crowding In this project, a

polyanionic macromolecule, dextran sulfate (DexS), was characterized with regards to

its potential to facilitate the extracellular collagen deposition

2.2.1 Collagen properties

Collagen is a molecule that comprises of three polypeptide chains (α-chains) Each

chain consists of a repeating glycine-X-Y (Gly-X-Y) triplet, in which X and Y can be

any residue, but are usually proline and hydroxyproline respectively This triplet motif

results in left-handed helices that can intertwine with each other forming a

right-handed triple-helical structure

Figure 2.4 The structure of type I procollagen (taken from Kielty and Grant, 2002)

Fibroblasts synthesize collagen as soluble procollagen, which consists of triple helical

section(s) and propeptides at both ends (C- and N-terminals) These propeptides will

be cleaved by specific proteinases, leaving the triple-helical domain with non-helical

telopeptides at both ends that is a site for collagen crosslinking The telopeptides are

susceptible to proteolytic attack, whereas the intact triple-helical domain is resistant to

most proteolytic enzymes However, it undergoes helix-to-coil transition and becomes

susceptible to degradative enzymes when it is heated to above its melting threshold

To date, there are 27 different collagen types that have been identified Collagen type

I, II and III are quantitatively the most important, accounting for over 70% of the total

collagens in human body (Kielty and Grant, 2002) The focus of this project is on

Cleavage by N-proteinase Proα2

collagen type I, which can be found throughout the body except in cartilaginous

tissues It is also synthesized in response to injury and in the fibrous nodules formed

as the consequence of fibrotic disease (Kadler et.al., 1996) In type I collagen, the

helical molecule is a heterotrimer that consists of two identical α1(I) chains and one

α2(I) chain

2.2.2 Collagen biosynthesis

The biosynthesis of collagen begins with the transcription of the individual collagen

genes and concludes in the maturation of the collagen fibrils in the ECM This

biosynthesis process is characterized by a number of co- and post-translational

modifications, some of which are unique for collagen

• Intracellular processing of procollagen

Translation of the mRNA encoding the pre-proα(I) chains occurs on the free

ribosomes and begins with the synthesis of the N-terminal Soon after, these

polypeptides begin to fold into appropriate secondary and tertiary structures As the

polypeptide chains are translocated across the endoplasmic reticulum (ER) membrane,

intrachain disulfide bonds are formed within the N- and C- terminal propeptides, and

hydroxylation of proline and lysine residues occurs within the collagenous domains

(Kielty and Grant, 2002) These chains then associate to form heterotrimeric

molecules that fold in a C- to N-terminal direction The C-propeptide plays an

important role in the association of the monomeric procollagen chain and in

determination of the chain selectivity (Bulleid et.al., 1997; Less et.al., 1997)

The stability of the triple-helix formation is dependent not only on the presence of

glycine as every third residue, but also requires the presence of 4-hydroxyproline in

the Y position at a high proportion Enzyme prolyl-4-hydroxylase (P4H) is

responsible for the hydroxylation of this imino acid The presence of hydroxyproline

at this specific position appears to favor a specific conformation of the imino acid

necessary for the packing of the collagen triple helix (Vitagliano et.al., 2001) In

addition, hydroxyproline coordinates an extensive network of water molecules with

the triple helix such that water bridges occur within and between the collagen chains

(Bella et.al., 1995)

The hydroxylation of proline residues also increases the denaturation temperature of

procollagen molecules (Berg and Prockop, 1973) There is a positive correlation

between the melting temperature of the triple helix and the extent of hydroxylation of

proline residues, as well as with the physiological temperature of an organism

(Burjanadze, 1979; Privalov, 1982)

The hydroxylation process requires ascorbic acid (vitamin C) as a cofactor for P4H In

the absence of this enzyme, procollagens are unable to leave the ER, and therefore

new collagen fibrils fail to form This will result in scurvy, that is associated with a

long-term dietary deficiency in vitamin C

There are several other post-translational modification enzymes and chaperones in

collagen biosynthesis, such as lysyl hydroxylase, prolyl-3-hydroxylase, chaperone

HSP47, etc However, they will not be discussed in this report

After the procollagen has achieved their correctly folded conformation, this protein

will leave ER and will be secreted out to the extracellular space The mechanism of

the procollagen secretion is still poorly understood, however it is known that

procollagen follows the classical secretion route for extracellular proteins, passing

through the Golgi complex Upon release from the Golgi apparatus, the bundles of

procollagen form secretory vacuoles that will exit to the plasma membrane

• Extracellular processing of procollagen to form collagen matrix

A key step in collagen fibril formation is the specific enzymatic removal of the N- and

C-propeptides from procollagen in the extracellular space by N- and C-proteinases

Both enzymes are members of the zinc-binding metalloproteinase family (Prockop

et.al., 1998) The N-proteinase has a unique property of only cleaving N-propeptide of

procollagen type I that is in native conformation Under this conformation,

N-propeptide is folded back in a hair-pin configuration such that it binds to the first part

of the major triple helix of the monomer, and leaving the N-telopeptide in a hair-pin

conformation even after the cleavage The C-proteinase specifically cleaves native

and denatured procollagen type I, as well as a precursor of lysyl oxidase This enzyme

cleaves the Ala-Asp bonds, however does not cleave similar bonds in the triple helical

domain, suggesting that the specificity depends on the sequences that flank the

cleavage site

The subsequent fate of the propeptides has also been studied These cleaved

propeptides have been suggested to play a role in the feedback control of collagen

synthesis and there is possibility that they deposit in the matrix (Risteli and Risteli,

1987)

Following the cleavage, collagen molecules will self-assemble and align in a

quarter-staggered array This assembly is an entropy-driven process, in which the surface of

the protein molecules experiences loss of solvent molecules that will result in

assemblies with a circular cross-section to minimize the surface area/volume ratio of

the final assembly (Kadler et.al., 1996)

The spontaneous aggregation of collagen molecules into fibrils is followed almost

immediately by the formation of covalent cross-links within and between the collagen

molecules These cross-links, that are formed from specific lysine and hydroxylysine

residues, are essential in providing the tensile strength and mechanical stability of the

collagen fibrils which their structural roles demand (Knott and Bailey, 1998) A

specific collagen crosslinking enzyme, lysyl oxidase, plays a main role in the

oxidative deamination of lysine and hydroxylysine residues in the nonhelical

telopeptide regions to form the corresponding aldehydes (Hong et.al., 2004) These

aldehyde residues can spontaneously condense with vicinal peptidyl aldehydes or with

peptidyl lysine to generate covalent crosslinkages between the newly formed collagen

polymers (Smith-Mungo and Kagan, 1998) It is noteworthy that in collagen I fibers,

disulfide bonding does not play a part in the crosslinking because of the absence of

cysteine residues (Kielty and Grant, 2002)

The catalytic activity of lysyl oxidase is dependent on a strict steric requirement, that

is the quarter-staggered alignment of collagen molecules It also depends on the

sequence of amino acids surrounding the target lysyl/hydroxylysyl residues

(Smith-Mungo and Kagan, 1998) Lysyl oxidase binds to the fibril surface and cannot

penetrate to its inner domains (Cronlund et.al., 1985) It implies that the oxidation of

lysyl groups must occur at the early stage of fibrillogenesis and/or that cross-links are

continuously manufactured at the surface of growing fibrils

Figure 2.5 The collagen synthesis, processing and assembly

2.2.3 The challenge in the enhancement of collagen matrix formation in

fibroblast culture

In culture, fibroblasts are maintained in a monolayer culture that is highly

unphysiological Under this condition, fibroblasts have little associated matrix and are

bathed in a large volume of culture medium It is in contrast with the in vivo

environment where the cells are tightly surrounded with ECM that is dominated by

collagen The fibroblasts in culture will continuously secrete a large amount of

procollagen into the culture medium However, due to the dilute setting in the

medium, the processing of these precursor proteins to mature collagen fibrils occurs

very slowly resulting in the accumulation of procollagen in the culture medium

This issue has been underestimated in both tissue engineering and in anti-fibrotic drug

discovery The construction of engineered tissues using fibroblasts is carried out with

sub-optimal amount of endogenous collagen matrix It may compromise the shape and

the mechanical strength of the constructed tissues Furthermore, the amount of

procollagen in the culture medium has been quantified to assess the effect of

anti-fibrotic drugs This measurement is not satisfactory in this case since it only captures

the inhibition of procollagen secretion by the drugs It does not, however, capture the

inhibition of the deposition of collagen on the cell layer, that depends on the

procollagen conversion to collagen This issue becomes significantly challenging

when a specific group of anti-fibrotic drugs that prevents the procollagen conversion

is involved, for instance C-proteinase inhibitors Early experiment at BAYER AG

showed that it was not possible to demonstrate the effects of C-proteinase inhibitors in

fibroblast cultures due to the absence of procollagen conversion Remarkably, nobody

at BAYER was able to solve this conversion problem (Dr Elmar Burchardt, formerly

at R&D with BAYER, anti-fibrosis programme, personal communication, March

2005)

This project therefore aimed to solve the above mentioned issue by applying the

principle of macromolecular crowding in fibroblast cultures to enhance the formation

of collagen matrix

2.2.4 Macromolecular crowding

• The principle

In living systems, biochemical processes occur in a medium containing high

concentration of macromolecules Even though one species may not be present in

high concentration, the overall species occupy a significant part of the total volume,

typically 20-30% (Ellis, 2001a) This fraction is thus physically unavailable to other

molecules Such condition in the living cell has been termed “macromolecular

crowding” Crowded environment occurs in both intracellular and in the extracellular

matrix In Escherichia coli, the concentration of total protein inside the cells is in the

range of 200-300 mg/ml, whereas that of RNA is in the range of 75-150 mg/ml,

making up the total concentration of 300-400 mg/ml (Ellis, 2001b) In in vitro culture,

the concentration of the macromolecule is estimated to be only 1-10 mg/ml Figure

2.6 illustrates the crowding condition in the eukaryotic cytoplasm In the ECM of

tissues, polysaccharides also contribute to crowding, for instance collagen

Figure 2.6 The illustration of the crowding condition in eukaryotic cytoplasm (taken from Ellis, 2001b)

Macromolecular crowding causes an excluded volume effect that leads to a

non-specific repulsive interaction between solute molecules This interaction is always

present regardless of any other attractive or repulsive interactions that may occur

between the solute molecules (Ellis, 2001a) How much of the volume that is

unavailable to other macromolecules depends on the numbers, sizes and shapes of all

the molecules present in each compartment The concept of excluded volume is

illustrated in figure 2.7 The squares outline the volume containing spherical

macromolecules (black) that occupy ~30% of the total volume, a value typical of

intracellular compartments The volume available to another molecule (yellow) is

defined as the fraction that can be occupied by the centre of that molecule If the

introduced molecule (red) is small relative to the macromolecules, it can access

virtually all of the remaining 70% of the space (figure 2.7a) However, if the

introduced molecule (blue) is similar in size to the macromolecule, the available

volume is much less than might be expected because the centre of the introduced

molecule can not approach the macromolecule less than the distance at which the

surfaces of two molecules meet, that is indicated by the open circle around each

macromolecule (figure 2.7b) The available volume thus defines an effective

concentration of the introduced molecule, which can be much higher than the actual

concentration in the total volume

Figure 2.7 The schematic drawing to illustrate the concept of volume exclusion

(a) The introduced molecule (red) is small relative to the macromolecules (black) (b) The introduced molecule (blue) has a similar size to the macromolecules (black) (taken from Ellis, 2001a)

• The consequences of macromolecular crowding

The main effect of crowding on biochemical equilibria is to favor the association of

macromolecules Equilibrium constants for macromolecule associations may be

increased by two to three orders of magnitude, depending on the relative sizes and

shapes of macromolecular reactants and products, as well as the concentration of

crowders (Ellis, 2001b) This thermodynamic effect arises from the reduction in

excluded volume when macromolecules bind to one another, which leads to a

decrease in the total free energy of the solution The more solute molecules present in

a solution and the larger they are, the less randomly they can be distributed Thus, as

the total concentration of macromolecule rises, the configurational entropy of each

macromolecule species becomes smaller and its contribution to the total free energy

of the solution increases In other words, the most favored state excludes the least

volume to the other macromolecules present This conclusion also applies to all

biochemical processes in which a change of excluded volume occurs, for instance the

formation of oligomeric structures such as fibrin, collagen and multienzyme

complexes in metabolic pathways (Ellis, 2001a) It is important to note that crowding

only enhances the inherent tendency of macromolecules to bind to one another, but it

does not create this tendency de novo

The effect of crowding on reaction rate is complex and may depend on hydrodynamic

as well as thermodynamic properties of the system Consider a substrate (S) and

enzyme (E) association to form a product (P) with SE* as the transition state:

S + E SE* P + E

If the overall rate-limiting step is the encounter rate of S with E, the reaction is

diffusion-limited This encounter rate is proportional to the sum of the translational

diffusion coefficients of the associating species, which are generally reduced in the

presence of substantial concentration of crowding agents (Minton, 2005) Therefore,

crowding is expected to decrease the rate of diffusion-limited association

(short-dashed curve in figure 2.8) In contrast, if the rate of association is limited by the rate

of conversion of SE* to P, such reaction is referred to as chemically rate-limited In

this case, the association of S and E to SE* can be treated as being at equilibrium

(Ellis, 2001b) But crowding increases association such that this equilibrium is

displaced to the right and the overall reaction rate will increase as the concentration of

crowding agent rises (long-dashed curve in figure 2.8) However, even if the reaction

is chemically rate limited, the rate of encounter of reactants decreases monotonically

with increasing volume occupancy of crowder Therefore, ultimately crowding will

reach a point at which the reaction is no longer chemically rate limited, and further

increases in crowder concentration are expected to result in a decrease in association

rate (continuous curve in figure 2.8)

Figure 2.8 Schematic depiction of the predicted dependence of reaction rate on the concentration of crowding agent (taken from Ellis, 2001b)

2.2.5 The ideal crowding agent

To be an ideal crowder, the macromolecule should have a molecular weight in the

range of 50 – 200kDa, be highly water-soluble and not be prone to self-aggregation

diffusion

–limited

reaction

chemically –limited reaction

Overall

reaction rate

The molecular shape has to be globular, rather than extended, to prevent solutions

becoming too viscous to handle (Chebotareva et.al., 2004) The agent should be easily

available in highly purified form so that the use of high concentrations does not

introduce problems associated with contaminants Most importantly, the agent should

not interact with the system under test, except via steric repulsion This requirement is

the most difficult to meet It is essential to establish that any effects observed when

using crowding agents are not the result of inadvertent changes in other factors, such

as pH, ionic strength or redox potential (Ellis, 2001a) This requirement abolishes the

possibility of using highly concentrated cell extracts as crowding agent, since any

interpretation will be complicated by specific interactions, hydrolase activity and the

presence of denatured proteins

Commonly used synthetic crowding agents include Ficolls, dextrans, polyethylene

glycol and polyvinyl alcohol (Ellis, 2001a) In this project, we have chosen dextran

sulfate (DexS) with a molecular weight of 500 kDa This macromolecule is a

polyanionic derivative of dextran and mimics natural mucopolysaccharide (for

example chondroitin sulphate, dermatan sulphate) It is freely soluble in water

forming clear solutions, readily degradable by ecological systems, possesses high

purity and good stability

Figure 2.9 The structure of DexS with sodium salt

DexS has a widespread application in medical industry as an anti-coagulant agent and

as ingredients of cream and ointment for treating thrombophlebitis and for cosmetic

applications It is also used to accelerate the hybridization rate of DNA fragments and

to stabilize proteins such as fibroblast growth factor, alcohol oxidase and yeast

alcohol dehydrogenase when stored in solution

2.2.6 Macromolecular crowding in collagen matrix formation

The principle of macromolecular crowding has been tested for the collagen matrix

formation Dermal fibroblasts that are cultured in the presence of neutral polymers,

such as dextran T-40, polyvinylpyrrolidone (PVP) and polyethylene glycol (PEG),

show association of collagen on the cell layer, as detected by gel electrophoresis

(Bateman et.al., 1986; Jukkola et.al., 1991) It is in contrast to the conventional

fibroblast culture where procollagen is found accumulating in the medium The

processing of type I collagen in the presence of macromolecules occurs by the initial

removal of the C-propeptide that results in the transient accumulation of a

pN-collagen intermediate It is then followed by a slower cleavage of the N-propeptide to

produce completely processed collagen molecules (Bateman et.al., 1986; Bateman

and Golub, 1990)

The increase of C- and N-proteinase activity has also been observed in the presence of

DexS and PEG (Hojima et.al., 1994) In a cell-free system, the rate of procollagen

cleavage by C-proteinase from chick embryo tendons is increased by 10- to 15-fold,

whereas N-proteinase activity is increased by 2- to 4-fold in the presence of DexS

With PEG, the C-proteinase activity is increased by 5- to 20-fold, whereas the

N-proteinase is increased by 2- to 5-fold