Stem Cells and Tissue Engineering, American Society of Gene Therapy 7th Annual M

[Guest guest]

MedGenMed Hematology-Oncology

Conference Report - Stem Cells and Tissue Engineering -- Dreaming

Things That Never Were

Highlights of the American Society of Gene Therapy 7th Annual Meeting;

June 2-6, 2004; Minneapolis, Minnesota

Posted 08/02/2004

Sara M. ni, MD, PhD

http://www.medscape.com/viewarticle/482317_1

Introduction

You see things; and you say, " Why? "

But I dream things that never were; and I say, " Why not? "

Bernard Shaw

Stem cell biology and gene therapy might have, at first sight, little

in common. They refer to the study of completely different systems:

cells vs viral vectors and gene fragments. Investigators in either

field often have different interests: The biologists toil in their

laboratories, while the gene therapists yearn to treat patients as

soon as possible. Yet, they come together when stem cell biologists

need to devise new ways to enhance cells they are grafting in vivo.

And the efforts of gene therapists often require targeting of specific

cell types and detailed knowledge of cell development and

differentiation.

Ways to intertwine these 2 branches of investigation were highlighted

at the plenary session of the Annual Meeting of the American Society

of Gene Therapy, recently held in Minneapolis, Minnesota, by leading

researchers, to foster collaborations and an exchange of expertise. In

this report, we outline presentations that addressed, during the

conference, the most forward-looking and, at times, controversial

research in tissue regeneration and bioengineering.

Still in their infancy, these research fields are getting more

enticing by the year. New in vitro and in vivo technologies, as well

as new insights into how cells and tissue work, are unveiling new

possibilities. They are uncharted grounds, shrouded by uncertainty and

perhaps fraught with risks. But because we don't know where we are

treading, we can't foresee what benefits we might be missing. So, " Why

not? "

Searching for Multipotent Adult Progenitor Cells

Might it be that adult stem cells (ASCs) hold the same clinical

potential of embryonal stem cells (ESCs), thus allowing researchers to

bypass most of the ethical and practical issues related to the

preparation and use of ESCs for cell and tissue replacement?

Dr. Verfaillie[1] of the University of Minneapolis,

Minnesota, addressed this question while reviewing data from the last

10 years of research, wondering herself whether the greater potency of

ASCs is indeed a well-founded hope or just hype.

Stem cells, by definition, are cells able to self-renew, with a single

cell able to differentiate into multiple, functional cell types. Stem

cells also fulfill a second requirement: They can functionally

reconstitute a given tissue in vivo.[2] Tissue-specific stem cells of

hematopoietic, keratinocytic, or neural origin fulfill this

definition, although it is not quite well understood how neural stem

cells do it, and some uncertainty still surrounds the potency of

keratinocytic stem cells.

What are the differences between ESCs and ASCs? ASCs have been thought

for a long time to be far less potent in terms of renewal and

differentiation potential than ESCs. Hematopoietic stem cells (HSCs),

for example, undergo senescence with telomere shortening. They were

considered multipotent, as they can give origin to all blood cells,

but not pluripotent, as they cannot give origin to nonblood cells.

This was the tenet held until 1997-1998.

In the past few years, in fact, new results from in vitro cultures and

in vivo experiments have raised questions on the true plasticity of

ASCs.[2,3] HSCs, for instance, have been reported to differentiate

into muscle fibers able to reconstitute dystrophin expression in

models of muscular dystrophy, and into other cell types of mesodermic

origin, such as bone and endothelium. Even more difficult to

understand has been the observation of differentiation into neurons

and astrocytes of HSCs grafted from bone marrow.[4]

Approximately 500-600 studies are now reporting observations of this

kind, with HSCs induced to differentiate into a variety of nonblood

cells. Hence, some have started nurturing the idea that a bone marrow

transplant, which can be performed with relative ease, might cure all

ills. Reconstitution of lost cell types, as in patients with

Parkinson's disease, might be accomplished, along these lines, simply

with an autologous transplant (without any histoincompatibility issue)

completely bypassing all the difficulties associated with the

preparation and use of ESCs. Is it indeed so?[3-9]

Reviewing the Evidence

Many of these reports, as noted by Dr. Verfaillie,[1] have not been

confirmed by independent groups; most do not fulfill the criteria of

plasticity, and some outright violate such " rules.[3-9] When

independent investigators cannot reproduce reported data, a number of

questions arise: Is it due to the use of different models or different

cell populations? Do the cell-manipulation and transplantation

techniques affect the results obtained in different sets of

experiments?

Analysis of the cells used in these studies has become more

sophisticated over time, with a wider application of Y chromosome

detection by fluorescence in situ hybridization (FISH), the

determination of tissue-specific expression of the beta-galactosidase

or green fluorescent protein markers, and finally adoption of the

Cre/lox system that allow more accurate cell tracking. Yet,

cell-fusion events are still quite difficult to analyze, as they can

be bypassed by the Cre/lox system.

If studies are not performed at the single-cell level, with transplant

of 1 cell, then the possibility of 2 different cell types with

different potencies still haunts the interpretation of the results

obtained. Upon review, some cases of contamination with multiple types

of ASCs have ruled out such high plasticity and appeared to confirm

the old dogmas.[1]

Cell-fusion events, between an ESC and a blood cell or between an ESC

and a neuronal precursor (with the formation of tetraploid

precursors), have also misled some investigators. The frequency of

such fusion events is thought to be about 1/100,000 cells and it thus

requires an intensive selection process to be uncovered. Reported

initially as occurring during in vitro experiments, fusion events have

been observed also in vivo under selective pressure.[1]

If, however indeed, a single ASC can shift from a mesodermic (eg,

blood cell) to an endodermic (eg, neuron) type, how is this possible,

since it defies all the tenets of tissue differentiation seen during

embryonal development? And if true, why would ASCs retain such a high

plasticity in the adult?

Transdifferentiation (de-redifferentiation) and pluripotency are the 2

explanations currently entertained to explain the puzzling findings of

cells apparently differentiating " across " the classic laws of

embryology. Pluripotency is best illustrated by an experiment that

many ran when they were little boys (or girls): from a flat worm cut

in 2 parts, out came 2 worms -- a case of tissue reconstitution

through the activity of pluripotent cells carried by the worm itself.

If these are the 4 possibilities underlying the plasticity data

reported so far[5] -- contamination with multiple stem cells,

cell-fusion events, de-redifferentiation, and pluripotency – are they

all clinically relevant? Although transdifferentiation and the

existence of pluripotent stem cells in the adult would indeed help in

cell and tissue engineering, it is still not clear what could be the

outcome of strategies relying, knowingly or unknowingly, on the use of

multiple stem cells or cell-fusion products.[1]

Characteristics of ASCs

Examples of pluripotent ASCs have been described by a number of groups

using different cell types: mesangioblasts (Cossu and colleagues),

muscle-derived stem cells (Huard and colleagues), skin-derived stem

cells ( and colleagues), and multipotent adult stem cells

(MAPCs) (Verfaillie and colleagues).[10-14]

As illustrated by Dr. Verfaillie,[1] these are some of the

characteristics so far known for MAPCs:

*

MAPCs can be isolated from cartilage, bone, muscle, and fat

tissues;

*

They can be found in human, mouse, rat, and swine (and probably

monkey) bone marrow;

*

In the mouse, MAPCs can be isolated also from the muscle and

brain;

*

They are optimally cultured in defined media, containing low

concentrations (2%) of fetal calf serum, and other factors (eg,

dexamethasone and linoleic acid/bovine serum albumin); and

*

They are responsive to the platelet-derived growth factor and

the epidermal growth factor.

Human MAPCs are among the most difficult to grow with the current

techniques, murine MAPCs being the easiest (see Table), with a

2-3-fold higher doubling efficiency in vitro.

Cell-surface markers are helpful to identify MAPCs, in which some of

the " negative " results obtained are used to rule out the presence of

other cell types:

* Not mesodermal stem cells: eg, VCAM-negative and CD44-negative;

* Not epidermal stem cells: eg, CD31-negative;

* Not hematopoietic stem cells: eg, CD45-negative, CD34-negative,

and Thy1-negative; and

* MAPCs, however, are SCA+, Flk1+ AC133 low, SSEA1 low, Oct-4+,

nanog+, and Rex-1+.

As reported by Dr. Verfaillie and other researchers, a single MAPC is

able to differentiate in a multilineage way, giving rise to

osteoblasts, chondroblasts producing transforming growth factor-beta,

myeloblasts secreting platelet-derived growth factor, endothelial

cells releasing vascular endothelial growth factor, neurons producing

brain-derived neurotrophic factor, or hepatocyte expressing hepatocyte

growth factor and fibroblast growth factor. A single MAPC is thought

to be able to contribute to most, if not all, somatic tissues.[1-14]

Grafts of 10-12 MAPCs gave rise to 80% of the animal's chimeric

tissues, grafts of a single cell to approximately 30% of chimeric

tissues. MAPCs do not form embryoid bodies, and they yielded moderate

levels of engraftment in nonobese diabetic/severe combined

immunodeficient mice, with acquisition of tissue-specific phenotypes.

MAPCs generated from the bone marrow and other sources contributed to

most of the somatic cells in the chimeras. Efficiency of engraftment

was substantially increased by irradiation of recipient animals.

In vivo corrective studies were performed in MPS mice that have a

toxic accumulation of heparan sulfate secondary to an enzymatic

deficit. MAPCs were successfully engrafted in multiple organs,

including the lungs, muscle, and spleen following bone marrow

transplant, with the production of the missing

glycosaminoglycan-modifying enzyme and cross-correction of the

metabolic defect.[1]

Organoids formed by MAPCs in vivo are also being investigated to

evaluate their potential in tissue regeneration and organ replacement.

As mentioned by Dr. Verfaillie,[1] investigators are conducting

preliminary experiments with MAPCs in liver regeneration by " writing "

capillaries with laser-guided writing, with the formation of small

vascular tubes of endothelial cells. MAPC cells contributed to the

formation of small sinusoids and hepatocyte function were present as

long as the sinusoid-like structures maintained their 3-dimensional

structure.[14]

What is the explanation for all of these findings? Contamination by

multiple MAPCs has been ruled out. Cell-fusion events have been ruled

out in vitro, but not yet in vivo. A de-redifferentiation process or

reactivation of remnant pluripotent stem cells is still being

investigated. And of course the tantalizing question, " If indeed MAPCs

are pluripotent stem cells, why don't we usually repair ourselves? "

remains open. Loss of MAPCs with aging, the lack of appropriate

mobilization, and insufficient growth support are among the many

possibilities to be addressed in the future.[1]

Is it perhaps, as in outstanding orchestral executions, a matter of

talent and cues, with a high rate (at least so far) of one-of-a-kind,

historic performances?

As recognized by many, stem cells, whether of embryonic or adult

origin, will be, in the coming years, a new, untapped source to

support drug discovery, toxicity screenings, and basic science

studies. Clinical applications for the treatment of single-gene

disorders and degenerative diseases, and the preparation of

bioengineered tissues will follow, if all goes well and technologies

keep pace with our ambitions.

Bioengineering Tissues to Bypass Grafts

Vascular surgeons achieve the replacement of blood vessels with

autologous tissues regrafted to damaged sites or synthetic materials.

Replacements are therefore often limited to vessels with a diameter >

6 mm, and they are not available for about one third of the patients

who need one.

Challenges in designing new vascular replacements are the risk of

thrombogenicity and occlusion, related to the lack of endothelium, and

the difficulty to achieve suitable mechanical properties. Most

experimental replacements undergo bust failure because of the low

density of the matrixes and the insufficient degree of crosslinking

achieved in the constructs.

As illustrated by Dr. West[15] of Rice University, Houston,

Texas, a number of factors may contribute to the success or failure of

an experimental vascular replacement: the cell source, genetic makeup

of the cell itself, choice of scaffold material, presence of bioactive

factors, and mechanic conditions.[16]

For example, transfection of smooth muscle cells (SMCs) with the

nitric oxidase enzyme reduced adhesion of platelets and thus decreased

thrombogenicity.[17] Expression of lysyl oxidase, on the other hand,

was found to increase the mechanical properties.[18] Collagen gels

seeded with SMCs expressing lysyl oxidase showed an increase in

elasticity and tensile strength.

The material(s) used to build a scaffold should:

* Be biodegradable and made up of nontoxic products;

* Be biocompatible and permeable;

* Be endowed with the appropriate strength and flexibility;

* Acquire and maintain the correct shape; and

* Allow easy seeding by cells.

The polymers used may be synthetic (eg, polyesters) or of natural

origin (eg, collagen gels). Although synthetic polymers offer better

control, greater safety, and easier processing, natural polymers allow

better tissue formation. The polymers, in fact, support the tissue

while it is forming, promoting adhesion of the desired cells, avoiding

attachment of undesired cells, and overall favoring cell migration.

A new polymer, poly(ethylene glycol) diacrylate, is being evaluated as

a bioactive scaffold built through photopolymerization. Poly(ethylene

glycol) units are added to the scaffold under construction and

polymerized by exposure to ultraviolet light to build the main frame.

Collagenase or elastase-degradable peptides can also be added to

target the scaffold for degradation by specific proteolytic

enzymes.[16]

The addition of biospecific cell-adhesion molecules (REDV for

endothelial cells and RGDS, more ubiquitous) ensures interaction with

specific receptors and the incorporation of specific ligands or cell

types -- eg, adhesion of SMCs to an RGDS hydrogel.[19]

Migration of endothelial cells or fibroblasts through a

collagen-mimetic hydrogel can by measured with a transwell assay, with

the determination of a migration index. The extent of cell migration

can be enhanced adding tethered molecules of epidermal growth factor,

whereas increased deposits of extracellular matrix may be achieved by

the addition of interspersed molecules of transforming growth

factor-beta.[20]

To engineer a vascular graft, an annular mold is filled and exposed to

ultraviolet light to obtain a solid structure. A suspension of

endothelial cells in polymer solution contributes to the formation of

the intimal layer upon interfacial polymerization. The construct is

then passed in a bioreactor chamber to allow for a greater cell

proliferation and exposure to sheer stress. The mechanical

conditioning achieved in the bioreactor, in fact, improves the

mechanical properties of the bioengineered vessel.

As noted by Dr. West, such a procedure is more easily applicable to

the preparation of microvessels, but its use may be feasible also for

larger vascular structures.[15]

ESCs and Neuronal Repair

Fecundation of an egg by sperm yields a zygote and then a blastocyst,

rich in pluripotent ESCs, which are capable of self-renewal and

differentiation in all somatic cells of the body. ESCs thus represent

the cells of choice to repair cellular and tissue lesions brought

about by genetic or degenerative diseases.

In particular, as discussed by Dr. Lorenz Studer[21] of Cornell

University, Ithaca, New York, a lot of interest is being raised by the

possibility of using ESCs to repair neural tissues in patients with

Parkinson's or Alzheimer's disease.[22,23] And a similar strategy is

being advocated to ensure adequate supplies of beta-islet cells for

transplantation in patients with type 1 diabetes.

A National Institutes of Health (NIH) registry of available human ESCs

has been established, but Dr. Studer[21] noted that this is a registry

of cell lines for in vitro use: " None will be acceptable for clinical

use. "

At present, there are close to 1 million patients with Parkinson's

disease in the United States. At diagnosis, most of the irreversible

damage has occurred: 80% of the dopamine-producing cells have already

been lost. Neuronal cells of fetal origin have been investigated for

possible use in grafting experiments. Availability and practical

issues (eg, low efficiency), in addition to ethical concerns, are,

however, making it difficult to use them as a source of replacement

cells.

Three different strategies can be used to induce neural

differentiation in ESCs:

* Embryoid body-based differentiation;

* Stromal feeder-based differentiation; and

* Default differentiation.

Neurons arising from ESCs cultured in vitro can be recognized by:

* Neuronal phenotype and characteristics;

* Neurotransmitter specificity;

* Regional specificity; and

* Functional characteristics.

Embryoid body-based differentiation, studied at the end of the 1990s,

although successful in generating dopaminergic neurons, requires quite

lengthy procedures, does not work well for some cell types, and often

yields heterogeneous outputs.[21]

Stromal feeder-based differentiation, proposed by Tada and

colleagues[22] in 2003 and further investigated by Barberi and

colleagues[24] in 2003, gives rise to nestin-positive neuronal

colonies. Differentiation into dopamine-producing neurons and

motoneurons occurs in approximately 6 days. Use of stromal feeders

thus helps in increasing cell yields and in reducing output

variability.[24]

Evidence of their functional properties was given with

proof-of-principle experiments in mice affected by a chemically

induced Parkinson's-like disease, with correction of the neurologic

disease by neurons generated from ESCs by nuclear transfer.[24]

Parthenogenetic Stem Cells

Parthenogenetic stem cells are being investigated as an alternative

and easily available source of pluripotent ESCs for treatment in

primates. Parthenogenesis occurs when embryologic development is

initiated without contribution of genetic or cellular material from a

male.[25]

Differently from male pronuclei that gave rise predominantly to muscle

cells, female pronuclei yielded, far more easily, brain cells, noted

Dr. Studer: a statement that, on a light note, seemed to meet with

laughs of knowing approval by only part of the audience. This is a

fact, however, that might be put to good use for neurorepair with the

use of autologous, parthenogenetic stem cells in women.[21,25]

Neurulation and neurogenesis occurring in the primates'

parthenogenetic blastocysts followed a neuronal differentiation

process that paralleled the in vitro and in vivo developmental steps

known for these tissues. Characterization of phenotypic markers and

functional studies showed successful differentiation and growth of

midbrain dopamine neurons, with phenotype maintenance in vivo.

Oligodendrocytes, motoneurons (HB9+), and gamma-aminobutyric acid

(GABA)ergic neurons (GABA+) were also present in these

cultures.[21,25]

Feeder cells supporting neural induction were derived from ESCs, with

differentiation into mesenchymal stem cells and the formation of

preadipocytic stroma. They were, in fact, able to undergo further

adipocytic (Oil Red staining), chondrogenic (Alcian blue staining),

osteogenic (Alizarin staining), and sarcoid differentiation.[21]

The ability of these cultures to yield complex mixes of neuronal

cells, concluded Dr. Studer, raises the possibility to investigate

global cell-replacement strategies for the adult brain.

In vivo experiments in monkeys (Macaca fascicularis) showed that the

age of the recipient being transplanted did not affect efficiency of

engrafting and neuronal differentiation in vivo. No teratomic growth

has been detected, so far, in transplanted animals. Considering that

approximately 1-2 million dopaminergic neurons are present in healthy

primates, the successful transfer of 1-5 million in vitro

differentiated neurons underscores the feasibility of this approach.

In vitro growth rates, cell-transfer procedures, and in vivo grafting

are, however, still presenting significant challenges that need to be

addressed.[21]

In the discussion, Dr. Studer[21] mentioned that the toxic effects

(dyskinesia) seen in 2 previous trials[26] that required surgical

intervention were related to the number, distribution, and

contamination of the cellular populations used. This is why we need

the help of gene therapy, he concluded, to achieve better, long-term

control of dopamine production and growth of grafted neurons.