Structural Support for Damaged Tissue Repair

Zachary D Blount. American Scientist; Research Triangle Park. Volume 105, Issue 3,  May/June 2017.

It all began with an experience one of us (Arinzeh) had more than two decades ago. In 1991, a summer research experience at the University of California at Berkeley demonstrated how engineering could improve the lives of patients. Instead of working in a more traditional area such as automobile design, Arinzeh spent the summer after her junior year of college working in a rehabilitation laboratory. Engineers there were designing new prosthetic devices for patients who had lost limbs, and new assistive devices to help paralyzed patients move. The engineers would then collaborate with clinicians at a rehabilitation center to test their developments. Before that summer she hadn’t connected traditional engineering principles with the opportunity to solve biomedical problems. But by the end of those short months, Arinzeh was hooked on the promise of using mechanical engineering to help people move better.

Tissue engineering, a budding field at that time, offered a chance to move beyond building prosthetics. Damage to musculoskeletal tissues, such as bone and cartilage, and nervous tissue, such as the spinal cord, can be debilitating and can severely limit a person’s quality of life. In addition, such tissues cannot fully regenerate after a severe injury or in response to disease. Tissue engineers aim to fully repair and regenerate that tissue so that it regains complete function, but at that time researchers still had a lot to learn about cells and their support structures to solve these problems.

The earliest successes were with skin, in which researchers used dermal cells to generate grafts, leading to the first commercial products in the late 1990s. Researchers imitate nature, using cells as building blocks and developing strategies to guide the cells to form the appropriate tissue. Because stem cells are precursors to almost all tissue types, such cells are a promising source of these critical building blocks. But cells don’t grow and differentiate on their own. The cell’s microenvironment can influence stem-cell function in critical ways. Engineered microenvironments, or scaffolds, can effectively promote stem cells and other cell types to form tissues. To construct such scaffolds, some important tools are what are called functional biomaterials. These materials respond to environmental changes such as pH, enzymatic activity, or mechanical load, and their composition can mimic or replicate components of native tissue.

One of us (Arinzeh) wanted to use functional biomaterials to create three-dimensional tissuelike structures where cells can grow, proliferate, and differentiate, ultimately forming and regenerating tissue. Our group’s work started with bone studies in the 1990s, eventually moving into cartilage and the spinal cord over the past decade. The overall goal is to produce structures that could someday help patients struggling with severe injuries and movement disorders to move freely. For bone repair, our group has studied composite scaffolds consisting of polymers and ceramics that provide both mechanical and chemical cues to repair bone. Piezoelectric materials, which respond to mechanical stimuli by generating electrical activity, are used to encourage the growth of nerve tissue as well as cartilage and bone. Glycosaminoglyatns (GAGs), a major component of native cartilage tissue, provide growth factors to promote tissue formation, and Arinzeh has designed biomimetic scaffolds that incorporate these molecules. After all these years, the promise that seemed so enticing in 1991 is becoming a practical reality, with huge implications for human health.

Alternatives to Donated Tissue

In the United States, people sustain nearly 8 million bone fractures each year. The body can naturally heal small, simple bone fractures. However, 10 percent of these injuries-such as large-scale bone defects or severe injuries from a car accident-are more challenging. These patients often need multiple surgeries and tissue implants to complete the job. Allogeneic (or donated) bone grafts offer an important source of tissue, but they aren’t always the best choice. Chemical processing of donated tissues can lead to poor healing in patients, and such tissues can harbor a small risk of disease transmission.

Tissue engineers are working on alternative scaffold materials that can fill in the bone defects and also stimulate tissue repair. Bone scaffold materials can sustain two different tissue-building strategies. Osteoconductive materials support the attachment of bone cells and bone tissue in-growth, whereas osteoinductive materials promote stem cells to differentiate to form bone synthesizing cells, known as osteoblasts. Osteoinductive scaffolds often rely on the interplay between mesenchymal stem cells- multipotent stem cells found in bone marrow that can produce a range of tissues, including bone, cartilage, and fat-and growth factors, such as bone morphogenetic protein. This factor signals to local stem cells to become osteoblasts. The scaffold can be used alone and rely on the patient’s own cells within the defect to populate the scaffolds, or researchers can seed scaffolds with cells prior to implantation.

Arinzeh has developed a scaffold with both osteoconductive and osteoinductive properties, without any added growth factors. In addition, the material is a composite, with both a ceramic component and a polymer or plastic component. The composite material combines both the beneficial biochemical properties of the bioactive ceramic and the mechanical properties of the polymer. Such ceramics are based on calcium phosphate, which can chemically bond to bone tissue and degrade over time. At the same time they spur the stem cells to differentiate into bone cells. Two types of conventional bioactive ceramics we use are ß-tricalcium phosphate and hydroxyapatite. Both are osteoconductive: Tricalcium phosphate degrades quickly, whereas hydroxyapatite is more stable. The latter has a similar composition to bone mineral and can more easily form chemical bonds with bone tissue.

To coax stem cells to become bone cells, Arinzeh wanted a scaffold in which she could both control the rate of degradation and have a stable but chemically reactive surface for the cells. So we mixed the hydroxyapatite with tricalcium phosphate to form a biphasic ceramic. We constructed these scaffolds with different ratios of hydroxyapatite to tricalcium phosphate to investigate the optimal ratio for promoting stem-cell differentiation. These structures had pore sizes of 300 to 600 micrometers to allow cells to infiltrate the structures and to make room for tissue growth. We then seeded the scaffolds with human mesenchymal stem cells. First, we studied the scaffolds with stem cells over 28 days under cell culture conditions, in either general growth media or media that would promote the stem cells to become bone cells. The stem cells on scaffolds comprising 60/40 and 20/80 ratios (by weight) of hydroxyapatite to tricalcium phosphate showed signs of differentiation: increased activity of alkaline phosphatase, an enzyme expressed by bone cells. Only the 20/80 scaffold, however, produced osteocalcin, a marker of mature bone cells, in the presence of general growth media. Those results made it clear that the 20/80 scaffold included the optimal mix of biomaterials.

We then implanted 100 percent hydroxyapatite, 100 percent tricalcium phosphate, and 20/80 scaffolds loaded with mesenchymal stem cells underneath the skin of severe combined immunodeficient mice to evaluate cell differentiation and tissue formation. These research animals have compromised immune systems, which prevents them from rejecting implanted human cells. We performed two tests, one with six weeks and one with 12 weeks of implantation, after which, in both cases, the porous structure of the 20/80 scaffolds was filled with bone tissue. Both scaffolds made from the single bioceramic material showed little to no bone formation, demonstrating that the optimal ratio of hydroxyapatite to tricalcium phosphate was critical to induce the differentiation of stem cells. We suspect that the 20/80 scaffold releases more phosphate and presents a more stable scaffold surface, two features that allowed the cells to attach and produce a uniform bone matrix. As a result, Arinzeh pursued scaffolds incorporating the 20/80 version for bone-tissue engineering applications.

Although these combined ceramics facilitate bone-tissue growth, they are brittle, which limits their use for treating bone defects in tissues that bear weight. Therefore, Arinzeh wanted to produce composite scaffolds that combined the 20/80 ceramics with polymers. In addition to making the scaffolds stronger, we could use these polymers to create fibrous structures that more closely mimic the mineralized collagen within the bone’s extracellular matrix, a stew of bioactive molecules and support structures surrounding cells in the body’s tissues. We can form fibers of the composite with diameters in the nanometer to micrometer range using a technique called electrospinning, which applies an electric field to a polymer solution and “spins” the material into tiny, threadlike structures. The structural fibers in the bone extracellular matrix are similar in diameter to electrospun fibers, making this technique ideal for producing scaffolds.

Arinzeh chose a polymer called polycaprolactone because it supports cell attachment, is biocompatible, and is mechanically flexible at room and body temperatures. Its bonelike strength makes it useful for weightbearing defects, and its compatibility with surgical rods and pins allows it to be easily inserted into defects. For the next round of scaffolds, we electrospun polycaprolactone fibers and embedded them with nanoparticles made of our 20/80 ceramics. We tried several formulations that varied the amount of embedded ceramic and the solvent used. Those formulations led to various fiber shapes and changes in the distribution of the 20/80 nanoparticles within the fibers. The fiber surfaces were rougher for scaffolds containing higher amounts of 20/80 ceramics, which could mean that these nanoparticles are clustering within the fibers.

Scaffolds produced with a solvent called methylene chloride had both nanometer- and micrometer-sized diameters and rough surfaces. For the scaffolds produced with a combination of this solvent with another, called N,N-dimethylformamide, the fibers were all on the micrometer scale and had smooth surfaces, with the ceramics completely embedded within the fibers. Despite the differences in appearance, all these fibrous scaffolds had similar mechanical properties to trabecular bone, the spongy, shock-absorbing tissue found at the ends of large bones and within the pelvis and vertebrae. However, the scaffolds containing 30 percent by weight of the 20/80 nanoparticles, produced with either solvent, had the most uniform distributions of ceramic within the fiber and were ideal for further biological studies.

To study the biological response to the scaffolds produced using different solvents, we cultured human mesenchymal stem cells on the scaffolds for up to 21 days. Both scaffolds supported cell growth. The scaffolds produced with methylene chloride, however, had larger spaces between the fibers. The cells could infiltrate deep within these structures and developed into stronger, vascularized, three-dimensional tissue.

These initial tests also demonstrated the potential of the scaffold to regenerate bone. In the scaffolds formed from methylene chloride solutions, more ceramic sits on the surface of fibers. When we placed these structures in simulated body fluid-buffered solutions that simulate blood but lack proteins-this feature triggered changes that resembled the natural formation of bone. Under these conditions, the ceramic nanoparticles dissolved from the fibers and deposited bonelike minerals on the surface of the fibers. When these scaffolds were cultured with mesenchymal stem cells, the cells produced and mineralized a bonelike extracellular matrix, infiltrated the scaffold, and showed osteogenesis.

Arinzeh has also tested whether this electrospun composite could heal a large bone defect in the thigh bone of an adult sheep. We implanted the scaffolds either alone or in combination with the sheep’s whole bone marrow as a source of stem cells. Harvesting a patient’s whole bone marrow at the time of surgery and immediately implanting it is a faster, simpler approach than isolating and growing a patient’s own stem cells from the bone marrow, which must be performed in a laboratory setting.

With microcomputed tomography, a high-resolution x-ray technique, we saw that a significant amount of bone filled into the defect within eight weeks after implantation. Interestingly, both the scaffold alone and the one loaded with bone marrow had similar amounts of bone fill. Both groups of treated subjects regenerated significantly more bone compared with the healing that occurred with untreated subjects. Over 16 weeks, the scaffold degraded as it was replaced by bone, a process that helps restore normal function and the ability to bear weight. These scaffolds provided the necessary support to help large bone defects heal more like smaller defects would normally. Our group plans to move these scaffolds into clinical testing soon.

Smarter Tissue Replacements

The composite scaffolds are both biocompatible and bioactive, which means they are chemically active. They interact and integrate with their surrounding environment through chemical reactions occurring at their surface. But we would like to have even greater control over tissue healing. Smart materials, which can be changed or controlled in precise ways with external stimuli, offer additional ways to direct the healing process. More recently Arinzeh has investigated piezoelectric materials, which can generate electrical charge when they are mechanically deformed. Electric fields can guide the development and regeneration of many tissues, including cartilage, bone, and nerves.

Clinicians already use electrical stimulator devices for bone repair and neural conditions. In addition, collagen, which is found in many tissues, exhibits piezoelectric properties. Few tissue-engineering researchers, however, have studied piezoelectricity, even though such materials could boost tissue regeneration for cells and tissues that rely on electrical stimuli.

One advantage of piezoelectric scaffolds is that they can generate electrical activity through body movement, cell attachment, and scaffold contraction rather than via an external power source. As piezoelectric materials are deformed, molecular groups within the material shift asymmetrically, which can produce an electric charge or voltage.

In our efforts to build novel piezoelectric tissue scaffolds, Arinzeh has focused on a polymer called polyvinylidene fluoride-co-trifluoroethylene (PVDFTrFE) because it is both biocompatible and has well-characterized piezoelectric properties. We have processed this polymer into soft or flexible fibrous scaffolds that can mimic the fibrous structure of the native extracellular matrix in a wide range of applications, most notably spinal injury.

When damaged, the spinal cord can regenerate only in limited ways. In addition, an inflammatory response can impede wound healing, which can prompt the formation of a glial scar. This type of structure acts as both a mechanical and biochemical barrier to nerve regrowth across the injury site. To repair the damaged tissue, various researchers have investigated several strategies to reduce a glial scar, including the use of growth factors and implanted cells, such as neural stem cells and glial cells. Neural stem cells are multipotent stem cells found within the nervous system. They require proteins, such as fibroblast growth factor, to differentiate into neurons. As the neural stem cells mature, they differentiate to form neurons and glial cells. Piezoelectric materials could advance neural repair by forming a bridge across damaged sites while using electrical stimuli to guide the growth of axons.

Arinzeh wanted to investigate piezoelectric polymers as scaffold materials for neural regeneration. So we initially electrospun fibrous PVDFTrFE scaffolds and examined whether they could support neural cells. We also assessed how the orientation of fibers affects neurite formation (the projection of new axons or dendrites from a neuron). We fabricated these scaffolds with an array of features: fibers aligned or arranged randomly, and with diameters at the nanometer or micrometer scale. Some of these fibers were annealed, a heat treatment that prompts crystal growth, whereas others were left untreated.

After characterizing their structural and mechanical features, we studied these scaffolds to see whether they could support neurite outgrowth. Within our scaffolds, we used dorsal root ganglions, well-studied sensory neurons that are a good cell model for examining nerve growth on biomaterials. The cells attached to the scaffolds readily. Neuntes elongated along the fibers, and the fibers provided a physical cue for directing the neuntes. We observed the greatest neurite extension with micrometer-sized annealed scaffolds, which have higher crystallinity and thus increased piezoelectric activity. On the aligned fiber scaffolds, the neurites from dorsal root ganglions extended in the same direction as that of the fibers.

Our group has recently studied these fibrous scaffolds as conduits in a rodent model for investigating spinal cord repair. In those experiments, the piezoelectric scaffolds enhanced axonal regeneration depending upon the design of the conduit. We are continuing to study these scaffolds in animal models with the hope that they can eventually be tested in the clinic.

In addition, piezoelectric scaffolds may prompt neural stem/progenitor cells to form neurons, which could someday be useful in spinal cord repair. As in previous experiments, Arinzeh varied the fiber diameter, fiber alignment, and annealing of these materials. We then cultured the stem cells over the course of nine days on these scaffolds with media that should either induce or control cell differentiation. After nine days, most of the cells in both media conditions expressed the protein ß-f// tubulin, indicating that they were neurons. We observed the highest number of apparent neurons when cells had been cultured in induction medium on aligned, annealed scaffolds. Interestingly, neural stem cells cultured on tissue culture polystyrene controls primarily expressed nestin, which is a protein marker for stem cells. These findings suggest that piezoelectric scaffolds can promote neuronal differentiation and guide neurite growth.

Piezoelectric materials can also support the growth of bone and cartilage, tissues that respond to an electrical stimulus and are activated by physiological movement or mechanical deformation. Osteocytes can sense deformations to the extracellular matrix, which can serve as a signal to remodel that tissue. Type I collagen has piezoelectric properties and is a major component of the bone extracellular matrix; therefore it could be the source of such signals. Thus, Arinzeh investigated piezoelectric scaffolds to induce human mesenchymal stem cells to form bone cells.

We electrospun polyvinylidene fluoride, which is also piezoelectric, using a range of applied voltages to determine the effect of processing methods on the piezoelectric properties of the scaffold. When fibers were electrospun using higher voltages, the resulting fibers became more piezoelectric. The scaffolds produced with the higher voltages prompted the greatest number of stem cells to differentiate into bone cells. As in earlier studies, Arinzeh measured stem-cell differentiation into bone cells based on comparisons of apatite mineralization and higher activity of alkaline phosphatase.

Cartilage cells, known as chondrocytes, also respond to electrical stimuli. Similar to the environment surrounding bone tissue, cartilage extracellular matrix contains glycosaminoglycans (GAGs), long, sulfated polysaccharides that naturally induce changes in the electric potential in the tissue when subjected to physiological loading. We also studied whether PVDF-TrFE scaffolds could promote differentiation of stem cells into cartilage cells. These scaffolds supported greater cell differentiation compared with other conventional biomaterials, indicating that these materials may be good candidates for cartilage repair applications.

Arinzeh is among the first researchers to demonstrate that scaffolds constructed from piezoelectric materials can effectively regenerate tissue. In bone, cartilage, and nerves, we have begun to understand how the electromechanical stimulus promotes tissue formation. As a following step in developing scaffolds that could be implanted in patients, we are preparing the next generation of piezoelectric materials that degrade as tissue grows.

Mimicking Tissues’ Chemical Cues

Stem cells and other cells within growing tissues use chemical cues to respond to their environment as they develop and grow. Therefore, Arinzeh is also developing functional biomaterials that mimic that adaptability. We went back to GAGs, which are decorated with a variety of negatively charged sulfate groups, a feature that helps them interact with a variety of proteins and other factors that boost cell growth and development. However, the native polysaccharides can be difficult and costly to extract in sufficient quantities from biological sources. Arinzeh overcame this challenge by using molecules whose structure resembles native GAGs.

The GAG mimetic we came up with, cellulose sulfate, is a semi-synthetic derivative of cellulose that is inexpensive and simple to produce. Cellulose sulfate can be synthesized to produce structures similar to chondroitin sulfate, a native polysaccharide in the cartilage extracellular matrix. Arinzeh incorporated cellulose sulfate into tissue-engineering scaffolds to construct a microenvironment for stem cells that resembles the native tissue environment and attracts growth factors that induce differentiation.

Before constructing tissue scaffolds, however, we wanted to examine whether cellulose sulfate can enhance expression of cartilage-specific genes (namely aggrecan and collagen type II) and matrix molecules from mesenchymal stem cells. To model the development of cartilage tissue, we grew these cells in pellet cultures: Cells were spun down with a centrifuge into a dense collection and allowed to grow together. When cellulose sulfate was present in the media, the cells expressed more cartilage matrix-specific molecules and exhibited uniform cell shapes consistent with that of cartilage. We then incorporated this material into a tissue-engineering scaffold, hoping to further steer the differentiation of stern cells toward chondrocytes.

To incorporate cellulose sulfate into a fibrous scaffold, we blended it with gelatin, which promotes cell adhesion. Many studies have indicated that the degree of sulfation within these extracellular polysaccharides can influence cell behavior. Therefore, we investigated cellulose sulfate scaffolds that incorporated carbohydrates that were either fully or partially sulfa ted. We seeded stem cells onto these scaffolds-made of partially sulfated polysaccharides and gelatin, fully sulfated polysaccharides and gelatin, or gelatin alone-and looked-for signs of chondrogenesis. In particular, we measured the amount of collagen type II, which is found only in cartilage tissue and comprises a major portion of the fibrous extracellular matrix of cartilage.

In the scaffolds made from partially sulfated cellulose and gelatin, the amount of collagen type II secreted by the cells greatly exceeded all other scaffold combinations. Interestingly, partially sulfated cellulose has a similar chemistry as chondroitin sulfate, the naturally occurring polysaccharide in cartilage. This finding suggests that these synthetic scaffolds could lead to robust cartilage tissue formation. We are currently evaluating these scaffolds in cartilage defects. The hope is that someday we can use these structures to help patients renew their own cartilage after an injury, using their own cells and tissues.

Repairing a car is still a lot easier than repairing the human body. But after a quarter century of research, the idea of swapping in cells and support scaffolds, rather than relying on the body’s mechanisms alone, is finally a reality for both skin and cartilage. With continued work on scaffolds and studies with stem cells, we have a lot of optimism for helping patients move better and even for repairing the spinal cord.