Kenneth E. Gonsalves, Craig R. Halberstadt, Cato T. Laurencin, Lakshmi S. Nair. Biomedical Nanostructures

Подождите немного. Документ загружается.

CHAPTER 11

Cellular Behavior on Basement

Membrane Inspired Topographically

Patterned Synthetic Matrices

JOSHUA Z. GASIOROWSKI, JOHN D. FOLEY, PAUL RUSSELL,

SARA J. LILIENSIEK, PAUL F. NEALEY, and CHRISTOPHER J. MURPHY

11.1 INTRODUCTION

The basement membrane (BM) is a specialized layer composed of a complex

web of extracellular matrix (ECM) proteins. It is through the basement

membrane that a wide array of cells (e.g., epithelial, vascular endothelial cells)

interact with their underlying stromal elements. Basement membranes have a

number of intrinsic characteristics that can serve to regulate the behaviors of

cells associated with them. These features include the presentation of ligand

sequences for speciﬁc cell receptors, serving as a reservoir for soluble cytoactive

factors such as polypeptide growth fact ors, and possessing physical character-

istics that can modulate cell behaviors. Two key physical characteristics of the

basement membrane shown to dramatically inﬂuence cell behaviors are

compliance and the inherently complex three-dimensional topography that

consists of feature types in the nano- through submicron-size scale [1].

This chapter will brieﬂy review the biology of basement membranes and

their biochemical and physical attributes. We then address the primary focus of

this chapter, namely, the development of basement membrane inspired

topographically patterned synthetic matrices for in vitro cell and tissue culture

systems. A summary of cellular behaviors inﬂuenced by the topographic

features of synthetic matrices is provided with an emphasis on the use of

anisotropically grooved surfaces. Such surfaces allow for the rapid evaluation

of cellular responses to topographic cues.

BiomedicalNanostructures, Edited by KennethE.Gonsalves, CraigR. Halberstadt,CatoT. Laurencin,

and Lakshmi S. Nair

297

11.2 THE BASEMENT MEMBRANE

The evolution of the BM is likely driven by the need for the simple cell to

interact with its environment. The earliest life forms on the planet were most

likely acutely sensitiv e to changes in their environment. If the primordial milieu

of simple molecules from which the ﬁrst living cells arose did not supply

enough nutrients at the appropriate pH, the earliest cells could not have

survived. However, in order for life to evolve into complex organisms, cells

needed to develop the ability to not only survive in their provided environment,

but also alter and shape it to be more advantageous for cell growth and

development. For this reason, secretion would appear to be a major step in

evolution. With secretion, cells could utilize chemical signals to co mmunicate

with each other as well as deposit proteins onto surfaces to create attachment

sites. By laying a protein coat down on a surface to creat e an anchor site, cells

began to shape their environment to provide them with a selective advantage.

Adherence to surfaces and to other cells could, in many instances, improve

survival, change cellular behaviors, and lead to the evolution of multicellular

organisms. This is still evident today as an individual bacterium that exists in

our current world has different resistances and expression proﬁles than the

same bacterium living within a thick, densely populated bioﬁlm [2–4]. Without

a doubt, the extracellular matrix and basement membrane have evolved to

become integral to development and disease.

11.2.1 Signiﬁcance of Basement Membranes in Disease

Besides providing cellular structural support, the functional impor tance of

basement membranes is readily evident by their role in development and

maintaining tissue homeostasis. One of the most well-known BM diseases is

thin basement membrane disease, which is exactly what it implies, a decrease in

basement membrane thickness. Thin basement membrane disease is character-

ized by a 100200 nm glomeruli BM in the kidney, which ranges from half to a

quarter of the normal size [5]. The decreased thickness can lead to blood in the

urine as well as blood pressure complications [6].

The ECM and BMs also play roles in the orchestration of normal wound

healing processes. Humans and rats with diabetes tend to have abnormally

thick basement membranes in the cornea and glomerulus [7, 8]. Though not

thoroughly investigated, thicker BMs may have altered topographical features

and compliance. These features could contribute to improper wound healing, a

common pathologic feature of diabetic patients. There are other cases,

especially within the skin and bone, where extracellular matrix proteins are

mutated or autoimmune reactions arise against matrix epitopes [9–12]. The

result of these mutations and immune reactions are alterations of basement

membrane function that can lead to disease. As such, it is important to

characterize and understand the biochemical and physical properties of

basement membranes.

298 BIOMEDICAL NANOSTRUCTURES

11.2.2 Biochemical Attributes

Cells secrete many different ﬁbrous proteins into the extracellular space and are

constantly shaping and remodeling their surrounding environment. Collectively,

this mix of secreted proteins is known as the extracellular matrix. A

specialization of ECM proteins on the basal surface of many cells within a

given tissue becomes the basement membrane. This layer provides tensile

support, serves as a reservoir for signaling molecules, and presents biochemical

and physical cues for cellular structures and tissues (Fig. 11.1). Transmission

electron micrographs reveal that the basement membrane is comprised of an

electron dense (lamina densa) and an electron lucid layer (lamina lucida) [13–15].

The biochemical composition of these layers will vary from tissue to tissue, but

most BMs share the same ﬁbrous protein components such as collagen, laminin,

and proteoglycans [13, 16–18]. Type IV collagen is the main building block of all

basement membranes and is highly conserved across species from C. elegans to

humans. Collagen will self-assemble with itself, nidogen, the laminin family, and

other proteins to form BMs of various functions and thicknesses that usually

range from 50 to 150 nm in most tissues and up to 300400 nm in glomerular

BM or 8 mm in the human lens capsule [19–22, 55]. Many ECM proteins have

whole domains or short conserved amino acid sequences that act as signaling

antagonists or adherence ligands for cell membrane receptors. The slight

biochemical differences that make up each BM account for many of the changes

seen in cellular structures from one tissue to the next.

The majority of studies to date that have analyzed the impact of BMs on cell

behavior have focused on the cellular response to the biochemical composition

of the BM. The biochemical properties of ECMs and BMs that regulate cell

behavior are better understood than the physical propert ies because they have

historically been easier to manipulate and test. Typical experimental designs

fall into three broad categories. First, most tests on ECM proteins have

characterized changes in cells after they were transiently transfected with the

FIGURE 11.1 Schematic of cells on a basement membrane. The cartoon depicts cells

adhering, proliferating, and migrating on a basement membrane. As cells expand in a

basal to apical direction, they often begin to differentiate, as indicated by the arrow. The

enlargement on the right is a 30,000 scanning electron micrograph of a rhesus aorta

basement membrane.

CELLULAR BEHAVIOR ON BASEMENT 299

protein in question, or subjected to various doses of recombinant protein

added to the cell media. These experiments are often performed with cells

plated onto ﬂat surfaces and can reveal visual phenotypic changes. They also

allow researchers to work out cell signaling pathways, such as how the

proteolytic release of endostatin from collagen inhibits angiogenesis pathways

[23–25]. Second, studies have been performed to characterize the effect of a

growth factor or peptide on cellular migration using a Boyden chamber or

experiments involving wound healing. These experiments typically measure the

chemotactic potential of a signal protein to control immunocyte invasion or

direct cells into a wound bed [26, 27]. Third, studies to measure the effect of a

peptide signal on adherence are usually performed by seeding cells onto a

surface coated with the protein being tested. Perhaps the best characterized

ECM peptide signal from these types of experiments that has been shown to

elicit a cellular adherence response is the arginine glycineasp artic acid

(RGD) signal peptide. The RGD signal motif, found within many different

ECM proteins, is a common ligand for several integrins, and can alter the

ability of different cell types to adhere to a given surface [28, 29]. Experiments

are commonly designed to measure RGD-dependent changes in cellular

adhesion by coating a plastic surface with RGD signal peptides or full-length

ECM proteins that contain the RGD domain, like ﬁbronectin. However, while

each of the aforementioned experimental strategies is designed to characterize

the biochemical effects of individual ECMs, they rarely take into account the

impact of the physical micron, submicron, and nanotopographic environments

that cells are exposed to both in vivo and in vitro.

In addition to biochemical differences, BMs have unique physical

characteristics: thickness, compliance, pore size, ﬁber width, and three-

dimensional organization. These are all physical features that deﬁne a BM.

Provided below is a concise summary of compliance followed by a presentation

of the focus of this chapter, the topographic features of basement membranes,

and their cellular consequences.

11.2.3 Physical Characteristics: Compliance

To date, the majority of studies on cells have been conducted on rigid surfaces

such as silicon, polystyrene, and polyurethane. These surfaces have been coated

with various ligands such as the RGD peptide to increase cell attachment or to

alter cell behavior. It has beco me clear that mechanical compliance of cellular

matrix may be as signiﬁcant as ligand functionalization in impacting

cell behavior [30, 31]. The compliance of a material relates the extent of

deformation (strain) of the material to an imposed stress (force/unit area). A

more rigid substratum has been shown to promote cell spreading and induce

phosphorylation pathways [32]. The effects of compliance on different iation

have also been demonstrated with breast epithelial cells [33]. When these cells

are cultured on ﬂoating gels, the cells differentiate into tubules. If the gels are

attached to a substrate, this differentiation does not occur. Similarly, work on

300 BIOMEDICAL NANOSTRUCTURES

human umbilical vein endothelial cells grown on polyacrylamide/gelatin

supports has shown that by decreasing the rigidity of the support, cellular

differentiation into tube-like structures will occur [34]. Increases in the traction

forces of ﬁbro blasts and the contractility of smooth muscle cells have been

reported by altering the compliance of the substrata [31, 35–39]. Cortical

neurons and glial cells are greatly inﬂu enced by compliance [40]. On hard gels,

astrocytes will outgrow neurites and possess actin stress ﬁbers. On soft matrix,

astrocytes were rounded and showed few stress ﬁbers similar to results

observed with ﬁbroblasts [31]. These characteristics are very different from

their behavior on hard surfaces. However, neurites have robust actin ﬁlaments

and enhanced F-actin protrusions on soft as well as hard surfaces.

Compliance of the matrix is probably vital in the organization of cells and

tissues in living organisms. Neurons have been shown to grow into hyrdogels in

injured areas while astrocytes will not [41]. Certainly, the results with the brain

cells suggest that the distribution and population sorting in the central nervous

system may be closely related to compliance in particular areas. Work with

stem cells has indicated that soft matrices similar to that of the brain induced

neurogenic phenotypes [42]. Stiffer matrices similar to muscle were myogeni c

and rigid supports were osteogenic. Thus, the physical properties of cellular

matrices as well as the biochemical ones are extremely important determinants

of cell behavior. Despite the recent recognition that compliance is an essent ial

property on the ECM, the compliance of native basement membranes has yet

to be reported.

11.2.3.1 Physical Characteristics: Topography Basement membranes

possess a complex ‘‘felt-like’’ three-dimensional topography. Individual

features can range in scale from approximately 10 to 400 nm with the majority

of features smaller than 100 nm [21, 22, 43, 44] (Fig. 11.2). The diameter of a

primary epithelial cell in culture can average between 20 and 50 mm, depending

FIGURE 11.2 Electron micrographs of native basement membrane topography.

(a) An SEM micrograph of cells covering the ﬁbrous basement membrane. (b) An SEM

magniﬁed image of the window in (a) showing complex topography of the basement

membrane.

CELLULAR BEHAVIOR ON BASEMENT 301

on the cell type. This represents several orders of magnitude difference in scale

between a cell and the basement membrane structural components and ensures

that a single cell interfaces with thousands of nanoscale and submicron

topographic features. Thus, in addition to a wide range of biochemical cueing,

epithelial and endothelial cells are exposed to a complex infor mation-rich

topographic environment. Altogether, the integration of biochemical and

physical cues provided by a BM plays a large role in development and disease

by regulating cell behavior.

The physical features of BMs are believed to exert great control over cell

behavior and development; however, it has been technically challenging to

design controlled experiments to test these hypotheses. The ability to separate

inherent chemical signals from purely physical ones for in vitro topography

experiments has been a difﬁcul t task for cell biologists to overcome.

Nevertheless, there has been nearly a century’s worth of studies investigating

the inﬂuence of topography on cellular behavior. As early as 1911, R.G.

Harrison was using spider webs to study cellular responses to complex surface

topography [45, 46]. What Harrison lacked was the ability to create controlled

surfaces with precisely deﬁned physical features for his experi ments.

The interdisciplinary merging of engineering and cell biology has lead to

revolutionary bioengineering techniques, such as the use of photolithography

and reactive ion etching, to create substrates to test the effects of topography

on cell behavior. Due to recent innovative fabrication strategies, researchers

can now produce the large number of nanostructured surfaces with controlled

feature sizes required to conduct statistically robust cell behavior studies [1, 47].

Different surface features (e.g., surface roughness, grooves, pores, etc.) with

dimensions ranging from tens to hundreds of nanomete rs have been reported

to affect proliferation, alignment, adhesion, and cell viability [48–53]. However,

the effects of physical features on cellular behaviors remain poorly understood

for a number of reasons. First, the topographic layouts of many BMs have not

been quantitatively analyzed. To date, the best characterized native BMs are

skin, cornea, urothelium, vascular endothelium, and the glomerulus [19, 21, 22,

54, 55]. While some morphologic aspects of the BMs from other tissues have

been characterized, a detailed quantitative description of their surface

topographic features has not be en reported. Second, research groups have

been experimenting with different manufacturing protocols to produce the

same type of topographical surfaces, with techniques such as UV or electron

beam lithography, to fabricate nanogrooves from titanium or silicon chips.

There are drawbacks and limitations for each and the result can lead to

inconsistent surface types and feature sizes. This ultimately makes it difﬁcult to

draw direct conclusions between earlier studies. Third, bioengineers still face

challenges in developing surfa ces that reach the lower limit of physical features

needed to recreate the full nanometer to submicron range of topographic

features that deﬁne native basement membranes. Despite these limitations and

shortcomings, biologists have made signiﬁcant advances in understanding the

interactions between the basement membrane and the cell. However, there is

302 BIOMEDICAL NANOSTRUCTURES

still a strong need for further development of artiﬁcial surfaces that incorporate

deﬁned topographic features in order to fully characterize the role of the

physical characteristic s of BMs in development, disease, and maintenance of

cell and tissue hom eostasis.

11.3 HISTORY OF BIOMIMETIC SYNTHETIC MATRIC ES

Biological systems are comprised of entities whose scales range at least 10 orders

of magnitude from the size of organisms down to the individual molecules that

regulate cellular behaviors. At the level where cells interact with a basement

membrane, most information is found at the submicron, nanoscale, and

subnanoscale levels. Early attempts to create biomimetic synthetic matrices were

limited by fabrication methods so that the manufactured surfaces had only

micron-scale features. With the aid of new nanofabrication techniques, we have

now been able to develop a generation of artiﬁcial surfaces that incorporate

nano- through micron-scale topographies. The nano- and submicron features

closely mimic the biologically relevant scale of physical features found in living

tissues and the micron-scale features provide a connect to the bulk of the

literature of topographic cueing.

To understand whether the scale of topographic features can inﬂuence

cellular responses, one must ﬁrst determine the relevant feature sizes of a cell’s

native microenvironment. Quantitative measurements of BM features such as

pore diameter and ﬁber width are similar in scale for many tissues studied and

also conserved across different species (Table 11.1). As more tissues have been

analyzed, some differences have emerged. One of the best known exceptions to

TABLE 11.1 SEM Physical Feature Measurements of Several Basement Membranes

Species Tissue BM

Mean feature

height

(nm)

Mean ﬁber

diameter

(nm)

Mean pore

diameter

(nm) Reference

Rhesus Cornea 191  72 77 44 72 40 [22]

Human Cornea 182  49 46 16 92 34 [21]

Human Descemet 131 41 31  93815 [21]

Human Foreskin  24 840 17

Porcine Aortic valve  27 12 38 24

Rhesus Aorta  30 11 62 37

Rhesus Carotid  31  11 60  42

Rhesus Saphenous  27  838 16

Rhesus Bladder 178 57 52 28 82 49 [43]

Bovine Glomerulus  9 3 [20]

Synthetic Matrigel 162  52 69 35 105  70 [44]

Most features are consistent in scale across different basement membranes and species. One

noticeable difference is the average pore size of the glomerulus BM, which has a specialized role

regulating solute transport.

Liliensiek SJ, Murphy CJ, personal communication.

CELLULAR BEHAVIOR ON BASEMENT 303

date is the glomerulus BM, which has smaller pore diameters [19, 20]. This

difference may be attributed to its specialized role as a selective diffusion

barrier. In order to test the roles these physical features play in guiding cellular

behavior, the features must be altered and arranged in a controlled fashion to

produce a useful analytic end point. Unfortunately, the ability to manipulate

the native BM of a tissue in vivo is limited with current biological techniques,

hence the value of artiﬁcial surfaces that can be designed with varying physical

features to exact speciﬁcations.

11.3.1 Matrigel and Randomly Ordered Arrays

Cells behave differently on ﬂat tissue culture plastic than they will on a

complex basement membrane in vivo. There is currently a demand for a more

realistic in vivo-like environment that can be used for in vitro studies; however,

this has proved to be a technically challenging task. Isolation of extracellular

matrices from living tissue is difﬁcult, often does not produce useful amounts

of experimental material, and often contains contaminating cells. The ﬁrst

major breakthrough to address this prob lem was the development of Matrigel.

Researchers discovered a tumor that was rich in basement membrane

components and grew rapidly, but remained benign in mice [56, 57]. Extracts

of that tumor were isolated, puriﬁed, and sterilized for in vitro use. It was

found that the resultant gel would alter the morphology, behaviors, and

expression patterns of cells that were plate d onto it [58, 59]. For example,

endothelial cells form a monolayer on tissue culture plastic, but arrange

themselves into tube-like structures on Matrigel [59]. Although Matrigel is a

commonly used reagent, the actual components cannot be easily manipulated

and deﬁned, demonstrating a need for alternative artiﬁcial BMs. Matrigel still

remains one of the most useful artiﬁcial basement membrane-like complexes to

study behaviors, but new nanotechnology manufacturing techniques are

beginning to provide different options for cell biologists searching for in vitro

ECM alternatives.

There have been other attempts to recreate the topographical environment

of a basement membrane, with silicon grooves, ridges, and tubes replacing

protein ﬁbers and pores (Fig. 11.3). These projects have utilized micro- and

nanomanufacturing techniques to produce surfaces that have randomly

ordered arrays, such as carbon ﬁbers, nanotubes, or nanocolumns. These

arrays have been shown to inﬂuence the morpholog y and expression of certain

cell types [60–62]. Artiﬁcial matrices that have isotropic randomly ordered

arrays of physical features are valuable research tools to create tissue culture

environments that more closely resemble in vivo environments. They are ideal

for experiments where cells can respond to a drug or protein in question within

a background more similar to in vivo conditions. Commercial stochastic

surfaces are available using electrospun ﬁbrillar structures as well as

membranes containing speciﬁed feature dimensions that have been shown to

impact cell behaviors [63, 64]. A stochastic surface has the advantage of better

304 BIOMEDICAL NANOSTRUCTURES

mimicking the surface order of the native basement membrane. These systems,

however, are not ideal for studying the direct impact that physical ECM

features have on cells. The drawback to these artiﬁcial systems is that there are

too many variables to control in an attempt to isolate the effect of one

individual feature. Pore sizes, ﬁber sizes, and depth are all randomly ordered

and cannot be easily changed in these systems. With stochastic surfa ces it is not

possible to observe anisotropic cellular behaviors (such as alignment) that

provide a rapid assessment of the cellular consequences of a surface structure.

In order to answer speciﬁc questions about the individual physical features of

BMs and their effects on cell behavior, new artiﬁcial systems had to be created.

For this reason, several laboratories have focused on the modulation of cellular

behaviors by nanogrooves.

Nanogrooves are repeating anisotropic ridges and channels that can be

created with speciﬁed lengths and depths. While they do not simulate the

surface order of the native BM, nanogrooves can be used to test a cellular

response to a single anisotropic feature. For example, they can be used to test

whether or not cells will align parallel to a topographic feature of a speciﬁc

ridge size. The advantage of the nanogroove system is that it allows biologists

to characterize cell behavior in response to a single topographic variable, thus

deﬁning basement membrane features through a reductionist approach.

11.3.2 Nanogroove Synthesis

Several laboratories have developed slightly different modiﬁcations to similar

protocols in order to create grooved surfaces [65–72]. Most surface designs

FIGURE 11.3 Scanning electron micrographs of native basement membrane,

Matrigel, and bioinspired synthetic matrices. The topographical features of the corneal

basement membrane compared to artiﬁcial surfaces. Matrigel biochemically and

topographically mimics native BMs, but it is difﬁcult to alter individual physical

features. Synthetic isotropic surfaces, such as gold-coated ‘‘grass-like’’ ﬁbers, can have

some physical features altered, but still retain the random array appearance of native

BMs. The physical features of anisotropic nanogrooves can be controlled and sized to

study the effects of individual topographic features on cell behavior. Scale

bars = 600 nm.

CELLULAR BEHAVIOR ON BASEMENT 305