Characterization of human adipose tissue derived adult multipotent precursor cells

CHARACTERIZATION OF HUMAN ADIPOSE DERIVED ADULT MULTIPOTENT PRECURSOR CELLS LEONG TAI WEI DAVID Bachelor of Chemical Engineering, Hons A THESIS SUBMITTED FOR THE DEGREE OF DOCTOR OF P

CHARACTERIZATION OF HUMAN

ADIPOSE DERIVED ADULT MULTIPOTENT PRECURSOR CELLS

LEONG TAI WEI DAVID

NATIONAL UNIVERSITY OF SINGAPORE

2006

CHARACTERIZATION OF HUMAN

ADIPOSE DERIVED ADULT MULTIPOTENT PRECURSOR CELLS

LEONG TAI WEI DAVID

(Bachelor of Chemical Engineering, Hons)

A THESIS SUBMITTED FOR THE DEGREE OF

DOCTOR OF PHILOSOPHY DEPARTMENT OF BIOLOGICAL SCIENCES NATIONAL UNIVERSITY OF SINGAPORE

2006

PREFACE

This thesis is submitted for the degree of Doctorate of Philosophy in the Department of Biological Sciences at the National University of Singapore No part of this thesis has been submitted for any other degree or equivalent to another university or institution All the work in this thesis is original unless references are made to other works Parts of this thesis had been published or presented in the following:

International Refereed Journal Publications

Cover pages of some of the following papers are found in the appendix

1 Leong TW, Chew FT, Hutmacher DW Isolating bone marrow stem cells using sieve technology Stem Cells 2004; 22:1123-1125

2 *Ng KW, *Leong DT, Hutmacher DW The challenge to measure cell proliferation in three dimensions Tissue Engineering 2005; 11:182-191

3 Leong DT, Hutmacher DW, Chew FT, Lim TC Viability and adipogenic potential of human adipose tissue processed cell population obtained from pump-assisted and syringe-assisted liposuction Journal of Dermatological Science J Dermatol Sci 2005 37:169-176

4 Leong DT, Khor WM, Chew FT, Lim TC, Hutmacher DW Characterization of osteogenically induced adipose tissue derived precursor cells in 2-dimensional and 3-Dimensional environments Cells Tissues Organs 2006; 182(1):1-11

5 Leong DT, Abraham MC, Rath SN, Lim T-C, Chew FT and Hutmacher

DW Investigating the effects of preinduction on human adipose derived

6 Leong DT et al Absolute quantification of gene expression with real time PCR Biomaterials 2007; 28(2):203-210

7 *Gupta A, *Leong DT, Bai HF, SB Singh, Chew FT, Hutmacher DW The combined action of 1α, 25–dihydroxyvitamin D3, β–glycerophosphate and ascorbic acid is essential for the osteo – maturation of adipose derived stem cells Manuscript under preparation

8 Leong DT, Hutmacher DW, Chew FT Genome wide gene expression revealed activating transcription factor 5 (ATF5) to be responsible for maintaining stemness in adipose derived stem cells Manuscript under prep

* contributed equally

Book Chapters

1 Hutmacher DW, Leong DT, Chen Fulin Polysaccharides in Tissue Engineering Applications, Handbook of Carbohydrate Engineering In press

Intellectual Competition

1 Finalist in Tan Kah Kee Young Inventors Awards 04 Making bone from fat

International and Local Conferences Presentations and Awards

1 Poster Presentation Preliminary studies on human adipose derived stem cells obtained from pump- assisted liposuction 4th Sino-Singapore Conference on Biotechnology, National University of Singapore, Singapore, November 11-13th, 2003 Leong TW, Chew FT, Hutmacher

DW, Lim TC

2 Oral Presentation Preliminary studies on human adipose derived stem cells

obtained from pump- assisted liposuction 2nd prize for 8th Biological

Singapore, December 3-5th, 2003 Leong TW, Chew FT, Hutmacher DW, Lim TC

3 Oral Presentation Osteogenesis of human adipose derived stem cells The Asian Surgical Association - 14th Biennial Congress & Scientific Meeting, Kota Kinabalu, Malaysia December 4-6 2003 Leong TW, Chew FT, Hutmacher DW, Lim TC

4 Poster Presentation: Viability of human adipose tissue stem cells from pump- vs syringe-assisted liposuction The Asian Surgical Association - 14th Biennial Congress & Scientific Meeting, Kota Kinabalu, Malaysia December 4-6 2003 Leong TW, Chew FT, Hutmacher DW, Lim TC

5 Poster Presentation The behavior of human adipose tissue derived precursor cells seeded on polycaprolactone-tricalcium phosphate scaffolds Tissue Engineering Society International, 6th Annual International Conference and Exposition, Orlando Florida, United States

of America December 10-13, 2003 Leong TW, Khor WM, Chew FT, Hutmacher DW

6 Poster Presentation Osteoprogenitor Cells induced from Adipose derived Mesenchymal Cells 8th National University of Singapore-National University Hospital Annual Meeting, October 7-8, 2004 Lim TC, Leong

TW, Chew FT, Hutmacher DW

8. Poster Presentation In vivo characterization of adipose derived

osteoprogenitors within polycaprolactone-tricalcium phosphate scaffolds Regenerate 2005, June 1-3 2005, Atlanta Georgia Leong TW, Lim TC, Chew FT, Hutmacher DW

Acknowledgements

“Therefore, since we are surrounded by such a great cloud of witnesses, let us throw off everything that hinders …, and let us run with perseverance the race marked out for us.”

Hebrews Chapter 12, verse 1, The Holy Bible

This thesis and its corresponding research work was a race, a race that would have been impossible for me to finish if not for the people that God had placed in my life I would like to thank them with my deepest gratitude

Prof Chew Fook Tim, I will always remember how you believed in a stranger to the field

of biology I hope that the many life lessons you have passed down to me had been and would continue to be useful to me Thank you for teaching me some skills that would ensure that I survived in this competitive arena Thank you for these many years of patience and huge funding for my experiments I hope that I would be as good as you, or even better, in research work, as you have once so expressed Thank you for making this experience fondly memorable

Prof Dietmar Hutmacher, you are a well-spring of innovative ideas I hope that some of your creativity had rubbed off on me Thank you for looking out for me and nurturing me Thank you for your encouraging support to listen to my crazy ideas and actually funding some of them You have taught me so much over these four years I know that I will be able to give to my students in the future just as you have given to me Thank you

Prof Lim Thiam Chye, many thanks for providing us with the adipose tissues Without which, it would be impossible for me to work Many thanks for giving your clinical inputs

in my project

To my good friends in TE laboratory, Kee Woei, Mohan, Suman and Amy Thanks for the many hours of tea sessions discussing our ideas and aspirations I will always treasure those times we had Kee Woei thank you for teaching me at the beginning of my thesis work It would not be as smooth if not for your help

To my colleagues in TE laboratory, many thanks for your help over these four years Last but certainly not least, to the dearest people in my life,

my wife, Shirley Ting – you are my light at the end of the long tunnel Thank you for your long-suffering patience I would never have made it without your constant encouragement Thank you love

and my mom, Teo Siok Lew who never failed to support me in what I love to do Thank you for teaching me perseverance in my growing up years Thank you mom

Both of you share the degree with me Thank you for your support, sacrifice, patience and love during this period I love you

This thesis is dedicated to the memory of my late father, Leong Fook Wing

Above all, to God be the glory

Chapter 1 – Introduction

1.1.1 Clinical need requiring bone healing or regenerative

1.1.2 The use of stem cells in regeneration of bone 1-2

Chapter 2 – Literature Review

2.2 Adipose tissue – a possible alternative to bone marrow as a source of

2.3 Adult mesenchymal stem cells or precursor cell populations 2-4

2.5 Other potential applications in regenerative medicine 2-16

Chapter 3 – Research Program

3.2.1 Phase I: Procurement of adipose tissue and their viability

3.2.2 Phase II: Characterization of ADSC in culture plastic and

polycaprolactone environments (Chapter 6 and Chapter 7) 3-3

3.2.3 Phase III: Characterization of ADSC in in vivo environments

3.2.4 Phase IV: Genome-wide transcriptome analysis of ADSC under

induction gave insights into differentiation behavior of ADSC

(Chapter 9)

3-5

Chapter 4 – General Materials and Methods

Chapter 5 – Viability and adipogenic potential of human adipose

tissue derived stem cells isolated from tissue aspirated with a vacuum

Chapter 6 – The combined action of 1,25–dihydroxyvitamin D3, β–

glycerophosphate and ascorbic acid was essential for the osteogenic

Chapter 7 – Characterization of osteogenically induced adipose tissue

derived precursor cells in 2D and 3D environments

Chapter 8 – Investigating the effects of preinduction on human

adipose derived precursor cells in an athymic rat model

Chapter 9 – Genome wide analysis revealed ATF5 as potential

transcription factor for downstream expression of a stemness gene in

adipose derived stem cells

Summary of Thesis

This thesis was motivated by the hypothesis that there is a subpopulation of cells in human adipose tissue which possesses multipotential differentiation properties for applications in regenerative medicine applications

A sub-population of cells termed by Zuk et al (2002) as “adipose derived stem cells” (ADSC) adhered to tissue culture plastics and was able to proliferate well in standard tissue culture conditions Further characterizations were carried out on several fronts, from their responses to osteogenic induction on tissue culture conditions to 3-D environments in an athymic mouse model The characterization

of ADSC for this thesis ended with the eludication of a potential stemness marker based on robust global gene expression array analysis from an in-house depository

of donors’ adipose derived stem cells

In this thesis, it was shown that ADSC isolated from adipose tissue aspirated with the vacuum pump were viable for subsequent cellular studies The osteogenic differentiation was chosen as a proof of principle of the multipotential properties of ADSC It was shown that 1,25–dihydroxyvitamin D3 worked synergistically with β-glycerophosphate and ascorbic acid to induce ADSC towards the osteogenic lineage ADSC were able to express bone matrix proteins and mineralization of its matrix under such conditions Progressing from culture plastics systems to 3-dimensional studies using polycaprolactone-tricalcium phosphate scaffolds, ADSC were still able to express bone characteristics when induced Similar characteristics

of ADSC were confirmed in an in vivo model

After characterizing ADSC from the in vitro to in vivo systems, the study

progressed to the molecular level Using microarray technologies, novel subgroups

of donors samples were discovered Analyzing the downregulated genes and looking for possible candidates of genes important to maintaining stemness, activating transcription factor 5 (ATF5) was determined Further validation using real time reverse-transcription PCR showed a highly significant and consistent downregulation of ATF5 in samples but no such downregulation of ATF5 in the negative control cells samples This placed ATF5 on the line-up as a candidate for

a gene important in maintaining stemness of ADSC or at least as a novel precursor gene target involved in controlling osteo-differentiation of ADSC

List of Tables

Table 2.1 Differentiation potentials of ADSC

Table 4.1 Supplements in osteogenic and adipogenic induction cocktail

Table 6.1 Nomenclature of the various study groups in induction cocktail

study

Table 6.2 Comparison table between various groups and their

corresponding results

Table 7.1 Summary table listing details of R-phycoerythrin (R-PE) or

phycoerythrin-Cy5 (PE-cy5) conjugated primary antibodies used for cell surface marker profiling

Table 9.1 Primer sequences used for real time PCR

Supplemental

Table 11.1

46 significantly expressed genes, comparing I+ with I- with at

least p-value < 0.05

List of Figures

Fig 2.1 Microscopic image of the human adipose stroma

Fig 2.2 Illustration representing essential components of a tissue engineered

construct

Fig 5.1 Image showing the initial centrifugal separation of the different

layers of pump-aspirated adipose tissue in a 50-ml Falcon tube

Fig 5.2 Colonies of ADSC in a heterogeneous cell population

Fig 5.3 alamarBlue assay study indicating thatmetabolic activity of cell

population arising from both methods of aspiration

Fig 5.4 Oil Red O staining for fat vacuoles present in adipogenically

induced ADSC

Fig 5.5 Semi-quantitative measure of extent of Oil Red O Stain of

adipogenic induced cells over uninduced cells for the two methods

of aspiration

Fig 6.1 Alizarin red S staining of ADSC cultures subjected to different

culture conditions

Fig 6.2 Immunostaining of ADSC cultures for collagen type 1

Fig 6.3 Immunostaining of ADSC cultures for osteonectin

Fig 6.4 Immunostaining of ADSC cultures for osteopontin

Fig 6.5 Immunostaining of ADSC cultures for osteocalcin

Fig 7.1 Flow cytometric data: Grey histograms refer to isotypic controls

Black histograms refer to various CD markers

Fig 7.2 Light microscopy pictures of a heterogeneous cell population and

positive osteonectin staining

Fig 7.3 Bridging cells and cell sheets formation in scaffolds

Fig 7.4 Graphs showing DNA content, alamarBlue reduction and alkaline

phosphatase concentration changes over time

Fig 7.5 Summary of RT-PCR

Fig 8.2 Microscope pictures of ADSC in culture

Fig 8.3 Histology and immunohistochemical images of cells on culture

plastics and induced for 28 days with osteogenic media

Fig 8.4 Scanning electron microscope images

Fig 8.5 Histology and immunostaining images

Fig 8.6: Semi-quantitative stain counting data for Alizarin Red S and

Masson Trichrome stains

Fig 8.7 Semi-quantitative stain counting data for Collagen Type I,

Osteopontin and Osteonectin staining

Fig 8.8 Mechanical testing results

Fig 9.1 Flow diagram describing the procedure underwent in analysing the

Genechip data

Fig 9.2 Normalized data presented in a box plot for 54,675 genes of each

sample

Fig 9.3 Condition tree of groups based on significantly expressed genes

Fig 9.4 Alizarin Red S staining for mineralization

Fig 9.5 Immunostaining of various osteo-related proteins

Fig 9.6 Graph summarizing gene expression of Ataxin 1, Muscleblind-like

1, EH1 domain and eukaryotic translation initiation factor 4E binding protein 1

Fig 9.7 Graphs summarizing normalized ATF5 mRNA data from

microarray analysis for each sample

Chapter 1 Introduction

CHAPTER 1

1 INTRODUCTION

1.1 Background

This section provides background information on the following two topics:

• The current clinical need shrouding bone repair and regeneration

• Adipose tissue as a potential depository of stem cells to answer this need

1.1.1 Clinical need requiring bone healing or regenerative intervention

The Healthcare Cost & Utilization Project stated that 12,700 craniotomies and

craniectomies were performed in 2001 and 20,616 procedures to correct facial

trauma defects The total costs of these surgical procedures were estimated to be

US$949 million (Panagiotis M., 2005) These statistics however do not include

those from orthopedic fractures and bone tumor osteoectomies and excision In

such cases, the in situ and surrounding bone could not provide good structural,

cosmetic and protective functions It then becomes clear that the ability to repair

bone defects can help in the restoration of function to the affected area

One possible treatment for non-unions of bone fractures is using autologous bone

grafts Autologous bone grafts enhance osseous union by contributing

osteoconductive and possibly osteoinductive materials to the fracture site and

remains a common graft material (Panagiotis M., 2005) Often, bone grafting is

employed in patients’ bone defects where existing internal or external fixation was

ineffective and must be replaced Vascularized bone autografts, had been used to

enhance the biologic environment and fill bone defects Unfortunately, a finite

Chapter 1 Introduction

could be up to 30% (Younger EM 1989) This limitation was further aggravated in

post excised bone tumor cases where the amount of available autologous bone

would not be enough to fill large bone defects Therefore there is a need to look for

alternative strategies for bone

1.1.2 The use of adult stem cells in regeneration of bone

Stem cells showed much promise in curing diseases like diabetes, Parkinson’s

disease, neurological degeneration, and congenital heart disease The existence of

stem cells is a matter of public discussion where religious, ethical, political, and

economic implications interest coincide Concerns plaguing the field are

theoretical and religious, such as defining when human life begins, which reflect

beliefs and philosophies These concerns and the technical problem of customizing

donor stem cells which would not elicit an immunological response from the host,

remained unsolved Much of these impediments had made the progress of

embryonic stem cells applications from the bench to the bedside, difficult These

concerns therefore had motivated more intensive research into characterization of

other potential sources of stem cells which might avoid these issues altogether

1.1.2.1 Adult stem cells

Stem cells are loosely defined as self-renewing progenitor cells that can generate

one or more specialized cell type

The edge that adult stem cells have over their embryonic counterparts in that the

ethical and immuno-incompatibility issues are sidestepped since the use of adult

Chapter 1 Introduction

stem cells in recent years, especially the multipotent bone marrow mescenchymal

stem cells (BMSC) (Gundle R et al., 1995; Haynesworth SE et al., 1996) Several

elegant studies have documented the ability of BMSC to differentiate into

non-hematopoietic cell types, including brain (Brazelton TR et al., 2000; Zhao LR et al.,

2002), skeletal muscle (Ferrari G et al., 1998; Gussoni E et al., 1999), liver and

epithelium (Petersen BE, et al., 1999; Theise ND et al., 2000), heart (Tomita S et

al., 1999; Orlic D et al 2001, Kucia M et al., 2004) and bone (Prockop DJ 1997)

There are clinical applications of stem cells which include the reconstitution of

blood lineage cells (Lagasse E et al., 2001, Morrison SJ et al.1994, Weissman IL et

al., 2000a) using hematopoietic stem cells, and bone reconstruction (Petite H et al

2000, Krebsbach PH et al., 1998, Quarto R et al., 2001)

Specifically in repairing and regenerating bone, the osteogenic capacity of bone

marrow has been demonstrated (Bruder SP et al., 1994) Bone marrow aspirated

from the iliac crest contains progenitor cells that can be used to augment the

osteogenic response of the implanted allografts or to heal a non-union when

injected percutaneously into the non-union fracture

Though BMSC is well characterized, the amount of bone marrow that can be

safely procured from the patient is limited In addition, it is estimated that the

concentration of stem cell is in the range of 0.001 to 0.01% of the total population

(Pittenger MF et al., 1999) This meant that extensive culture is necessary to

achieve therapeutic relevant cell numbers In addition, the effects of long term

culture on the differentiation potential of these cells remained largely unknown

Chapter 1 Introduction

1.1.2.2 Adipose tissue as a potential depot of stem cells

Studies have shown the differentiation of ADSC along mesodermal lineages

(osteogenic, chondrogenic, adipogenic and myogenic lineages) (Mizuno H et al.,

2002; Zuk PA et al., 2002)

Phenotypically, ADSC under specific induction conditions, expressed osteogenic

genes like osteonectin, osteopontin, osteocalcin, alkaline phosphatase (Zuk PA et

al., 2002) and chondrogenic markers like collagen II, chondroitin-4-sulphate and

keratan sulphate, collagen type X and aggrecan (Huang JI et al., 2004; Ogawa R et

al., 2004) and histological resemblance of hyaline cartilage (Dragoo JL et al.,

2003a)

ADSC seeded on apatite coated poly-lactic-co-glycolic acid scaffolds were used to

repair critical size murine calvarial defects which otherwise would be impossible

for the calvarium to heal itself (Cowan CM et al., 2004) There was also one

reported clinical case where iliac crest bone was mixed with isolated ADSC in an

autologous fibrin glue delivery system to repair a critical sized calvarial defect in a

7 year old girl (Lendeckel S et al., 2004) However, in the above two cases, the

isolated in vivo osteogenic properties of ADSC was not confirmed because the

chosen site of ADSC implant was in a bone region It is possible that ADSC

merely assisted in bone healing rather than a direct differentiation from ADSC to

osteoblasts in vivo

Chapter 1 Introduction

ADSC when induced towards the adipogenic lineage also expressed adipogenic

genes like leptin, peroxisome-proliferating activated receptor γ, aP2 and

lipoprotein lipase (Zuk PA et al., 2002; Bernlohr DAet al., 1985)

ADSC cells also expressed neuron-specific proteins, neuronal nuclei protein and

neuron specific enolase Late markers neuron-specific enolase and neurofilament

were expressed by 60-85% of the cell population (Yang L et al., 2003)

With these phenotypic evidences of specific tissue commitment, ADSC might

make a good candidate for stem cell intervention in reconstruction of bone,

cartilage and adipose tissue and even neuropathlogic diseases

Raising hope for genetic diseases, genetically engineered autologous stem cells

might be used to replace dysfunctional cells Genetic bone diseases like

osteogenesis imperfecta and fibrous dysplasia of bone may be improved with the

implantation of engineered stem cells at the site of clinical trauma caused by the

diseases (Bianco P, Robey PG 2001) Bone morphogenetic protein 2 transfected

ADSC have increased bone precursors with a faster onset of calcified extracellular

matrix than transfected bone marrow derived stem cells (Dragoo JLet al., 2003b)

since usually transfected with retroviral delivery systems, ADSC was able to

maintain transgenic expression even after differentiation (Morizono Ket al., 2003)

Chapter 2 Literature Review

CHAPTER 2

LITERATURE REVIEW

2.1 Description of the adipose tissue

Adipose tissue consists of adipocytes embedded in a vascular loose connective tissue, which is divided into lobules by stronger fibrous septa carrying the larger blood vessels, in order that each lobule receives an independent blood supply Within the lobules the cells are round or polygonal (Fig 2.1) Fat deposits serve as energy stores, sources of metabolic lipids, thermal insulation (subcutaneous fat), mechanical shock-absorbers (soles of feet, palms of hands, gluteal fat, synovial membrances) It occurs in abundance in subcutaneous tissue, around the kidneys,

in the mesenteries and omenta, in the female breast, in the orbit behind the eyeball,

in the marrow of bones to the plantar skin of the foot, and as localized pads in the synovial membrane of many joints

The adipose tissue is capable of considerable changes in volume during the course

of life of mammals Although relatively small increases in volume can be accommodated by changes in the amount of lipid stored in hypertrophic adipocytes, larger changes are mediated by the hyperplasic growth of new adipocytes accompanied by coordinated remodeling of the adipose vasculature (Rupnick MA et al., 2002; Hausman DB et al., 2001) It can be thought that the dynamism is mediated by resident stem cells These stem cells can be enzymatically isolated from the stroma of the adipose tissue and separated from the buoyant lipid filled adipocytes (Fig 2.1) by centrifugation A more homogeneous population emerges in culture under conditions supportive of MSC growth

Chapter 2 Literature Review

The most important features of adipose tissue as a cell source might be the easy expendability of its MSC and the consequent ease of procurement of large quantities with minimal risk to the patient Liposuction is a common surgical procedure with 478 251 elective liposuction surgeries performed in the USA during 2004 (American Society for Aesthetic Plastic Surgery 2005) It is also safe:

an American Society for Dermatologic Surgery study of outpatient cosmetic liposuction performed between 1994 and 2000 showed zero deaths on 66 570 procedures and a serious adverse event rate of 0.68 per 1000 cases (Housman TS et al., 2002)

Fig 2.1: Microscopic picture of cells within the stroma of adipose tissue Adipocytes were filled with lipids and adopted globular or polygonal morphologies Stem cells lie within this stroma

Chapter 2 Literature Review

2.2 Adipose tissue – a possible alternative to bone marrow as a source of adult mesenchymal stem cells

Using bone marrow mescenchymal stem cell (BMSC) is one of the most promising strategies in regenerative medicine applications However, the actual volume of bone marrow that can be procured from the patient is limited Therefore, another potential source of stem cells that could be procured in large amounts and had a reasonably high concentration of stem cells is adipose tissue These multipotent stem cells were capable of differentiating into the adipogenic, chondrogenic and osteogenic lineages when subjected to appropriate induction stimulants (Zuk PA et al., 2002) Donor site morbidity also limited the amount of marrow that can be obtained and thereby extended the time in culture required to generate a therapeutic cell dose Thus, the volume of human marrow taken under local anaesthesia is generally limited to no more than 40ml and yielded approximately 2×109 nucleated cells (Bacigalupo A et al., 1992) Obtaining a larger volume of bone marrow would necessitate the use of general anaesthesia, increase donor site morbidity (Auquier P et al., 1995; Nishimori M et al., 2002) and further dilute the stem cell fraction with stem cell-free blood ((Bacigalupo A et al., 1992) By contrast, a typical harvest of adipose tissue, under local anaesthesia, could easily exceed 1,000 ml and yield at 2×108 nucleated cells per 100 ml of lipoaspirate (Aust

L et al., 2004)

There were attempts to enrich the “true” stem cell population from amongst a

heterogeneous adult stem cell population Friedenstein et al were the first to

identify in the adult bone marrow a cell population with strong osteogenic potential

Chapter 2 Literature Review

the stromal cells of the bone marrow by the BMSC adherence to culture plastics Subsequent studies by other groups on these adherent cells showed that these cells were able to differentiate into the mesenchymal lineages Even within the adherent cell population, it was still heterogeneous in cell types Therefore more sophisticated methods like cell sorting were utilized to purify the true BMSC from amongst this adherent cell population Gronthos S et al showed that cells with surface expression of STRO-1 and CD106 had clonogenic and differentiation potential to osteoblastic, adipogenic and chondrogenic lineage (Gronthos S et al., 2003) Another novel method was purification based on cellular size (Hung SC et al., 2002) This proposed method of “sieving” out cells was an interesting idea but the results were inconclusive (Leong TW et al., 2005) Nonetheless, the question of whether cells should be purified from their niche cells remained unanswered

2.3 Adult mesenchymal stem cells or precursor cell populations

Caplan was among the first who introduced the term “mesenchymal stem cells” widely into the scientific community (Caplan AI 1994) Jiang Y et al (2002) showed at the single cell level, that 0.1% of CD45- TER119- murine bone marrow monocular cells differentiated not only into mesenchymal cells, but also into cells

with visceral mesoderm, neuroectoderm and endoderm characteristics in vitro

However, that work was based on the murine model and has yet to be proven to be correspondingly true in humans There is a possibility of a difference within the human and mouse stem cell systems, as seen in the ability of murine leukaemia inhibitory factor (LIF) in maintaining stemness in mouse embryonic stem cells but human LIF did not have similar maintenance effect (Sato N et al., 2004)

Chapter 2 Literature Review

Therefore, it could be argued that in reality there was no true MSC in human marrow The multipotential phenotype observed was due to various lineage-specific precursor populations, creating an illusion that the bone marrow cell population possessed multipotent phenotypes

Zuk et al (2002) working with human adipose derived stromal cells showed with clonal studies that there existed a subpopulation of cells capable of differentiating into at least three lineages Guilak et al (2006) also reported similar observations in

a recent clonal study However, there was a lack of conclusive function evidences

using cloned ADSC to confirm in vitro observations Therefore whether MSC

really exist in adipose tissue currently remained controversial

Not hampered by the controversy, this thesis used precursor and stem cells terms interchangeably

2.3.1 ADSC characteristics

A general misconception associated with ADSC is that the donor must be overweight or obese to have sufficient adipose tissue available for harvest Comparing adipose tissue harvested from obese BMI patients with that from normal BMI patients however, did not show any significant increased amounts of stem cells (Morizono K et al., 2003) In addition, there were no significant differences between ADSCs from patients with and without type II diabetes in terms of their mesenchymal stem cell characteristics and osteogenic and adipogenic potential (Strem BM and Hedrick M, 2005) Besides, assays performed

Chapter 2 Literature Review

on the factors secreted by ADSC also revealed the presence of multiple angiogenic

and anti-apoptotic cytokines (Rehman J et al., 2004)

2.3.2 Cell Surface Characterization

The cell surface phenotype of human ADSC was quite similar to BMSC (Strem

BM et al., 2005a) Both stem cell types were positive for CD105, STRO-1 and CD166 CD117 (the stem cell factor receptor), also expressed by embryonic stem cells, hematopoietic stem cells, MSCs and ADSCs (Aye MT et al., 1992; Lemoli

RM et al., 1993; Ogawa M et al., 1993) In addition to these multipotent markers, ADSCs and MSCs, both expressed numerous other molecules including CD29 (beta-1 integrin, which plays a critical role in therapeutic angiogenesis (Li TS et al., 2005), CD44 (a hyaluronate receptor) and CD49e (important for cell adhesion

to fibronectin) ADSCs also expressed high levels of CD54 (ICAM-1) when compared with BM-MSCs (De Ugarte DA et al., 2003a) ICAM-1 is a member of the immunoglobulin supergene family and can be up-regulated in response to numerous inflammatory mediators and cytokines (Roebuck KA and Finnegan A 1999) ADSC did not express the HLA-DR protein and the majority express MHC Class I molecules (Aust L et al., 2004) suggesting their potential for allogeneic transplantation (Lee RH et al., 2004) One difference in the surface marker expression appears to be the reciprocal expression of VLA-4 (CD49d/CD29) and its cognate receptor VCAM-1 (CD106) It was observed that there was expression

of VLA-4 but not VCAM-1 by ADSC from the majority of donors (Strem BM et al., 2005a) but the trend is reversed in BMSC (De Ugarte DA et al., 2003a) Since these molecules were involved in hematopoietic stem and progenitor cell homing

Chapter 2 Literature Review

al., 2002) It is interesting that these molecules represent part of a receptor–ligand pair that has an important role in hematopoietic stem cell homing to, and mobilization from, bone marrow (Papayannopoulou T et al., 1998; Kronenwett R

2.3.3 General information on ADSC

For many years researchers principally studied the adipogenic potential of preadipocytes within the stromal vascular fraction of adipose tissue (Ailhaud G et al., 1992; Dixon-Shanies D et al., 1975, Hausman GJ and Martin RJ 1989) Subsequently, it was reported that adipose tissue contains a multipotent cell population with properties that were similar to those of bone marrow MSC

Chapter 2 Literature Review

Several groups estimated the frequency and yield of bone marrow MSCs by applying clonogenic assays for either fibroblastoid-like colonies (CFU-F) or colonies expressing alkaline phosphatase (CFU-AP) (Castro-Malaspina H et al., 1984; D’Ippolito G et al., 1999; Oreffo RO et al., 1998; Muschler GF et al., 2001) Clonogenic assays had typically been used to quantify MSCs in marrow (Zuk PA

et al., 2002 Wickham MQ et al., 2003; Rodriguez AM et al., 2005b) Using these assays, the number of MSCs in bone marrow was generally found to be approximately 1 in 25,000 to 1 in 100,000, (D’Ippolito G et al., 1999; Muschler GF

et al., 2001; Banfi A et al., 2001) The frequency of CFU-F and CFU-AP of ADSPC is in the order of 1 in 100 – some 500-fold more than that found in marrow (Fraser JK et al., 2006)

ADSC and MSC both possess the ability to suppress a mixed lymphocyte reaction

in a dose-dependent and timedependent fashion (Puissant B et al., 2005; Tse WT et al., 2003) Furthermore, Rodriguez et al., 2005b have demonstrated that clonally

derived, multipotent cells from adipose tissue are immunoprivileged, both in vitro and in vivo This suggested that, similar to MSC, ADSC might have potential as

immunoprivileged universal donor cells with the capacity to be used in the allogeneic setting (Tse WT et al., 2003)

2.3.4 Specific differentiation capacity of ADSC

Adipose-derived cells could differentiate into several cell types (Table 2.1) It was not confirmed if a single adipose-derived cell could differentiate into all of these lineages; however, some groups generated ADSC clones from single cells capable

Chapter 2 Literature Review

and neuro-) (Zuk PA et al., 2002; Rodriguez AM et al., 2005a; Guilak F et al.,

2006) in vitro, thereby demonstrating the possible presence of multipotential cells

within adipose tissue

Chapter 2 Literature Review

Table 2.1 Differentiation potentials of ADSC

et al., 1987; Rodriguez AM et al., 2005a

CM et al 2004; Huang JI et al., 2002; Hicok KC et al., 2004)

et al., 2004; Planat-Benard V et al.,

2004

et al., 2002; Dragoo JL et al., 2003; Awad HA et al., 2004

al., 2003; Safford KM et al., 2002

et al., 2004; Mizuno H et al., 2002

Chapter 2 Literature Review

Based on the literature, it was decided that ADSC showed a strong capability to differentiate into the adipogenic and osteogenic lineage Based on this background, the discussion was limited to the adipogenic and osteogenic lineage

2.3.4.1 Adipogenic lineage

Given the origin of ADSCs it was not surprising that, when cultured in adipogenic medium, ADSCs expressed several adipocytic genes including lipoprotein lipase, aP2, PPARγ2, leptin, GLUT4, developed prominent lipid filled intracellular vacuoles - the definitive marker of adipogenesis (Zuk PA et al., 2001 and 2002; Safford KM et al., 2002; Wickham MQ et al., 2003) Despite certain donor-to-donor qualitative differences in adipogenic potential (Safford KM et al., 2002), the pattern of adipocytic gene expression is similar to that of BMSC (Pittenger MF et

al., 1999; Peister A et al., 2003) The in vivo potential of ADSC to differentiate

into cells of the adipocytic lineage had also been demonstrated in studies involving implantation of ADSC seeded in scaffolds made from natural biomaterials like collagen (von Heimburg D et al., 2001a and 2001b) and hyaluronic acid (Halbleib

M et al., 2003); or synthetic bioresorbable polylactic acid (Patrick CW Jr et al., 1999) or polyglycolic acid scaffolds (Lee JA et al., 2003) These studies generally

agreed that robust ectopic in vivo adipogenesis required prior in vitro

pre-differentiation of ADSCs

2.3.4.2 Osteogenesis

The ability of BMSCs to give rise to osteoblasts is well known (Owen M 1988; Bennett JH et al., 1991; Hicok KC et al., 1998; Nuttall ME et al., 1998) A disorder

Chapter 2 Literature Review

subcutaneous adipose depots, provided physiological evidence that cells capable of mineralization exist in adipose tissue (Kaplan FS et al., 1994; Shore EM et al., 2002) In the recent years, several groups had isolated cells from adipose tissue of

humans and other species capable of differentiating into osteoblasts in vitro (Zuk

PA et al., 2002; Dragoo JL et al., 2003b; Halvorsen YD et al., 2001; Tholpady SS

et al., 2003, Huang JI et al., 2002; Ogawa R et al., 2004; Hattori H et al., 2004; Kang SK et al., 2004;) Under osteogenic conditions, ADSC were observed to express genes and proteins associated with an osteoblastic phenotype, including alkaline phosphatase, collagen type I, osteopontin, osteonectin, osteocalcin, bone sialo protein, RUNX2, BMP2, BMP4, BMP receptors I and II, PTH-receptor (Zuk

PA et al., 2001; Halvorsen YD et al., 2001)

2.4 Scaffolds based bone engineering

Several studies showed that adult stem cells by themselves would likely fail to augment a bone defect Scaffold-based tissue engineering concepts involved the combination of viable cells, biomolecules and a structural scaffold combined into a

“construct” to promote the repair and/or regeneration of tissues (Fig 2.2) (Hutmacher DW and Garcia AJ 2005) The construct was intended to support cell migration, growth and differentiation and guide tissue development and organisation into a mature and healthy state

Scaffolds must provide sufficient initial mechanical strength and stiffness to substitute for the mechanical function of the diseased or damaged tissue with the aim of repair or regeneration Scaffolds might not necessarily be required to

Chapter 2 Literature Review

strength should be sufficient to at least support and transmit forces to the host tissue site For example, in skin tissue engineering, the construct should be able to withstand the wound contraction forces In the case of bone engineering, external and internal fixation systems might be applied to support the majority of the load bearing forces until the bone had matured (Hutmacher DW, 2000)

Fig 2.2: Illustration representing essential components of a tissue engineered construct A scaffold or matrix, living cells and/or biologically active molecules contributed the cellular, bioactive and support parameters for repair and regeneration of tissues Blue and red lines represented the extracellular matrix produced by the cells in such constructs (Illustration adapted from Hutmacher DW and Garcia AJ 2005)

Tissue engineered constructs

Chapter 2 Literature Review

Cell and tissue remodelling is important for achieving stable biomechanical conditions and vascularization at the host site Hence, the 3-D scaffold/tissue

construct should maintain sufficient structural integrity during the in vitro and/ or

in vivo growth and remodelling process (Hutmacher DW 2000),

Bioactive materials incorporated into scaffolds could also provide the necessary biomolecular cues to enhance the osteogenic potential of the cells Scaffolds tagged with BMP2 increased the proliferative and angiogenic properties of the cells-scaffold construct (Rai B et al 2005) This strategy could be briefly explained

as follows Stem cells could either be cultured in vitro and then differentiated before seeding into the scaffold or first seeded into and then differentiated in situ

After the appropriate progress of growth or differentiation of the constructs, they were implanted to replace diseased or damaged tissues With time the scaffolds are resorbed and replaced by remodelling host tissues that included a viable blood supply and nerves The ideal viable tissue-engineered constructs should adapt to the immediate bone physiology and provide long-term repair (Hench LL and Polak

JM 2002)

Investigators had used polylactic-co-glycolic acid (PLGA) coated with apatite as the scaffold material (Cowan CM et al., 2004) The apatite coating rendered the scaffold more osteoconductive This apatite coating with the proper pore sizes might have the additional effect of directing the ADCs to bypass the chondrogenic stage Kuboki et al recently reported that hydroxyapatite-coated scaffolds would induce either endochondral or intramembranous ossification They also showed

Chapter 2 Literature Review

endochondral ossification whereas the larger pore sized scaffolds (350 µm) drive intramembranous ossification (Kuboki Y et al., 2001)

Using a variety of scaffolds and matrices, human ADSC could form bone in immunodeficient rodent ectopic bone models (Lee JA et al., 2003; Hicok KC et al., 2004) In a murine critical size calvarial defect model, mouse ADSCs were able to regenerate cranial bone, which spanned the defect site, within 8 to 12 weeks of implantation (Cowan CM et al., 2004) The authors delivered ADSCs on an apatite-coated, resorbable scaffold into a critical size (4mm) calvarial defect and demonstrated about 80% closure of the defect within 12 weeks, with radio-opacity equivalent to 90% of uninjured bone at 8 weeks (Cowan CM et al., 2004) With a lack of chondrogenic matrix, Cowan CM et al proposed that bone formation occurred via intramembranous ossification This proposal was further supported by that the calvarium normally develops through this mechanism

Work on ADSC was carried further to the clinic when a seven year old girl was presented with a critical sized calvarial defect Repair was attempted with a mixture of autologous adipose-derived cells, iliac crest bone and fibrin glue, combined with a resorbable mesh (Lendeckel S et al., 2004) Computer tomography scans taken 3 months post surgery showed significant ossification throughout the defect In that study, the isolated effect of the autologous ADSC was not known as iliac crest bone would also contribute significantly to the healing process Nonetheless, authors of that case report suggested that ADSC might have contributed to healing of a critical sized calvarial defect

Chapter 2 Literature Review

ADSCs transduced with adenovirus–BMP2 and seeded into standard collagen type

I sponges formed a radio-opaque, bone-like material when implanted, intramuscularly, into immuno-compromised mice (Dragoo JL et al., 2003b) However, many other cell types showed similar results after transfection with BMP

or RUNX2 (Huang YC et al., 2005; Gersbach CA 2006; Phillips JE et al., 2006; Byers BA and Garcia AJ 2004a)

2.5 Other potential applications in regenerative medicine

Restoring blood flow to ischemic cardiac tissue had proven instrumental in the treatment of patients with acute myocardial infarctions Therapeutic angiogenesis, potentially from ADSC, would have an extremely broad range of clinical applications, such as diabetic retinopathy peripheral vascular disease, ischemic stroke and ischemic cardiomyopathy ADSC had been shown to increase angiogenesis to ischemic tissue; however the underlying mechanism remains unclear ADSC could secrete substantial quantities of angiogenic growth factors under normoxic condition like vascular endothelial cell growth factor (VEGF), hepatocyte growth factor, transforming growth factor-β However, there was a 5-fold increase in VEGF under hypoxic conditions (Rehman J et al., 2004) ADSC besides secreting angiogenic factors could also differentiate to cells of endothelial characteristics (Planat-Benard V et al., 2004b; Al Khaldi A 2003) Miranville et al had presented data supporting the presence of cells within adipose tissue that differentiated into endothelium (Miranville A et al., 2004) Amongst the population obtained from adipose tissue, CD34+/CD31- cells expressed after induction with VEGF and insulin-like growth factor (IGF) both CD31 and von Willebrand factor,

Chapter 2 Literature Review

the CD34+/CD31- ADSC to improve blood flow and capillary density in an athymic murine model of hind limb ischemia These results were confirmed by another study showing that delivery of ADSC to immunodeficient animals following induction of severe hind limb ischemia resulted in accelerated restoration of perfusion (Rehman J et al., 2004)

2.6 Gene delivery vehicles

Owing to the high proliferation rate, ADSC might be a source of cells capable of enhanced gene delivery A number of investigators had transduced ADSC in order

to facilitate tracking or to elicit a therapeutic effect Leo et al used luciferase to allow non-invasive, real-time tracking of ADSCs in rat spine (Leo

Ad-CMV-BM et al., 2004) Similarly, Dragoo et al (Dragoo JL et al., 2003b) infected both BMSCs and ADSCs with E1A-deleted-type 5 adenovirus constructs containing the BMP2 gene or the bacterial β-galactosidase (lacZ) gene LacZ gene transduction efficiency was 35% for BMSCs and 55% for ADSCs Ad-BMP2 infection of ADSC resulted in levels of expression of BMP2 protein that were three-fold higher

than those derived from BMSC Ad-BMP2 infected ADSCs exhibited in vitro

osteoblastic differentiation in the absence of exogenous osteogenic factors They

also exhibited robust ectopic in vivo production of bone when cells were implanted

into a collagen sponge within the subcutaneous space (Dragoo JL et al., 2003b)

Given the success of unmodified MSCs in treatment of osteogenesis imperfecta (Horwitz EM et al., 1999; 2001, 2002), these data support the potential for

Chapter 2 Literature Review

skeletal system Kang SK et al., 2003a had also used an E1A-deleted type 5 adenovirus to infect ADSCs As described above, these studies employed transduction of a tracking gene (lacZ) and a potentially therapeutic gene Brain-Derived Neurotrophic Factor (BDNF) achieving 100% and 94% transduction efficiency, respectively Transduced cells were implanted into areas of the brain that had undergone 90 minutes of transient ischemia-reperfusion injury Donor cells capable of continued expression of the transgene were maintained for at least

30 days Studies using lentiviral-infected cells in which transduction efficiency was 98% at day 3 and >95% at day 100 allowed examination of gene expression during

in vitro differentiation Retention of marker gene (EGFP) expression was observed following both adipogenic and osteogenic differentiation (Aust et al., 2004; Hicok

KC et al., 1998)

It could be concluded from the literature review that ADSC has the potential to differentiate down the osteogenic lineage However, a large majority of groups working on the topic of ADSC performed their studies only with small donor

numbers In addition, few studied the osteogenic potential of ADSC in a 3-D in vitro environment The in vivo model still remained largely unexplored Currently

much of the work in the ADSC field was limited to “proof of principle” studies There is now a great need to study at the gene level to understand the adult stem cell biological reasons underlying the phenotype observed Hence these were the motivations for this thesis A detailed description and explanation of the research plan and rationale could be found in Chapter 3

Chapter 3 Research Program

“tissue engineering construct (TEC)” to promote the repair and/or regeneration of tissues In the special context of bone healing, there is a need to provide the structural and mechanical stability to ensure repair / replacement bone defect site

In addition, scaffolds also simulated partially the in vivo 3-dimensional

environment that cells actually thrive in

One of the promising adult stem cell sources is a culture plastic-adherent cell population from the human adipose tissue In order to bring adipose derived stem cells (ADSC) a step closer to the routine clinic application, it has to at least fulfill the following

• Reliable procurement of autologous ADSC

• Characterization of cells in in vitro culture plastic environment

• Characterization of cells in in vitro 3-D environment

• Characterization of cells in in vivo animal models

• An in-depth analysis of the transcriptome of ADSC to further understand

Chapter 3 Research Program

However, the current literature lags behind in addressing the preceding issues Hence, this thesis addresses these issues and aims at providing new knowledge in the field of ADSC in the context of bone tissue engineering

3.2 Four study phases

3.2.1 Phase I: Procurement of adipose tissue and their viability (Chapter 5)

The two general methods of providing the suction pressure, necessary for the aspiration of adipose tissue, either through the vacuum pump or the syringe Most

if not all groups working on ADSC assumed that the high vacuum is not detrimental to the ADSC in terms of its viability and adipogenic potential Conversely, plastic surgeons raised the issue, based on their clinical observation, that pump-aspirated adipose tissue were not viable To address the discrepancy from a cellular point of view, the following question was asked,

“Does pump-assisted lipoaspiration result in stem cells with lower proliferation and differentiation potential compared to those aspirated with the syringe?”

Equal volumes of adipose tissue aspirated with the pump and with the syringe were

compared

Chapter 3 Research Program

3.2.2 Phase II: Characterization of ADSC in culture plastic and polycaprolactone environments (Chapter 6 and Chapter 7)

With certainty that cells obtained from pump-assisted liposuction are viable and possess differentiation potential, this thesis work proceeded to characterize these cells on their potential to differentiate into other tissues besides its native tissue type The non-adipogenic lineage was focused on the osteoblastic lineage as the aspect of study for these multipotent cells Firstly, the proper use of 1,25–dihydroxyvitamin D3, β–glycerophosphate and ascorbic acid as osteogenic induction factors were confirmed Further characterization of the cells in terms of their cell surface markers and morphologies were carried before proceeding from the 2-dimensional (-D) to the 3-D environments Bone is a 3-D tissue that provides mechanical protection and support in the body Therefore to further characterize ADSC towards the osteoblastic lineage, progressing from the 2-D to the 3-D environment is inescapable The question asked to consolidate the thoughts is

“what are the characteristics of ADSC in the 2-D (culture plastics) and 3-D environment (polycaprolactone-tricalcium phosphate scaffolds) and what are their differentiation potential in these two different in vitro environments?”

3.2.3 Phase III: Characterization of ADSC in in vivo 3-D environments

(Chapter 8)

3.2.3.1 Phase IIIa: Induced versus uninduced cells-scaffold constructs

In vitro cell culture studies in 2D environments, though a good and controllable environment for scientific study of ADSC from a basic biology point of view, proved insufficient as a clinically and physiologically relevant study environment