Wanted: a material that acts as a brilliant insulator against heat and cold, a structure able to withstand massive amounts of pressure, a substance resistant to acids and toxins. Although this could be a want ad for a tile for the space shuttle, a fiber-optic wire, or a biohazard suit, it also happens to characterize the attributes of a tooth. When a tooth is worn, broken or defective a replacement is necessary to restore its form and function. Synthetic replacements are available to restore enamel and dentin, but to date there is nothing like the real thing in appearance or functional capacity.

The goal of biomimetics research is to create a material through the use of biologic information that will copy to some degree biologic structure, in this case a tooth. The impetus to engage in such a scientific pursuit requires participation from material science, engineering, biophysics and biology experts as well as healthy support from the leadership at the NIH. As Dr. Malcolm Snead from the CCMB explains, The NIH leadership decided that a pan-institutional effort should be made to understand how tissues organize and how replacement tissues either complemented, compromised or improved the status of the original tissues.

            Synthesizing a biologic structure is a bold endeavor to contemplate whether the structure being crafted is complex or seemingly simple. The tooth for example may outwardly appear to be a simple structure. A cross section of the tooth however, reveals an elegant organization of complex structures, each of which provide the tooth with necessary adaptive properties.

The average tooth is assaulted daily with a range of adverse environmental conditions. The teeth are exposed to extreme physical forces when they bite down on hard materials, be they bones or jawbreakers. Teeth are bathed constantly in enzyme-rich saliva and coated with toxin-producing bacteria that would erode even the best of Italian marble statues. They are exposed to acid, alkali, and sugar day after day, year after year and still most survive much longer than the person whose mouth they inhabit. The durability of a tooth is easy to demonstrate by the persistence of teeth in the oldest human remains.

To maintain form and function capable of resisting such physical and chemical offenses several fascinating tooth structures have evolved. These structures, which are unique to teeth, include enamel, dentin, a discreet interface between the two called the dentin-enamel junction (DEJ), a protective coating between bone and tooth called cementum, and a cushioning ligament between the tooth and the jawbone called the periodontal ligament. Each tissue requires a precise series of cellular and molecular events to produce the final structure that provides resistance to the numerous stressors affecting the teeth.

Several strategies could be devised to restore lost or defective tooth structure with a natural substitute. Scientists could simulate enamel by assembling individual proteins and minerals in vitro, in a test tube so to speak, to replicate the events that occur naturally during enamel biomineralization. They could devise a strategy to coax the cells that make these structures to give a repeat performance in the controlled environment of a research laboratory. Or researchers could attempt to make the entire tooth by recapitulating the entire developmental program in a carefully controlled location within the jaw. 

Each strategy requires that scientists understand how these tissues are formed during development. Dental scientists have had a descriptive story of tooth development for decades. But to be able to harness the powers of development at the biochemical, cellular or whole organ level scientists needed to know much more than the organization of the cells contributing to the formation of the tooth. Important information was needed to discover what cells were involved in forming the various parts of the tooth, how these cells communicated with one another, what protein products the different cell types made and how all these components worked together to create highly-organized, mineralized tissues such as enamel, dentin and the DEJ.

In recent years dental scientists have charted what genes and gene products (proteins) were involved with the development of tooth structures. Discovering the what, where, and when of the gene expression relevant to tooth development was an important hurdle for dental scientists. But as many scientists will explain, the explosion of genetic knowledge in biologic systems is still followed by an enormous gap in knowledge in how the genes and their proteins actually function to put an organ together.

At the University of Southern California School of Dentistry’s, Center for Craniofacial Molecular Biology (CCMB), several scientific groups study tooth development to understand how specific genes and their respective proteins come together to make this 3-D organized structure. The molecular biology group headed by Dr. Malcolm Snead and including Dr. Michael Paine, focuses on the genetic control of enamel matrix proteins such as amelogenin. The biochemistry group, that includes Dr. Janet Oldak-Moradian and Dr. Alan Fincham, concentrates on the process of enamel biomineralization. Dr. Shane White, a dentist and dental materials scientist, works to understand why dental materials, both natural and synthetic, fail. Scientists under the charge of Dr. Margarita Zeichner-David focus their efforts on analyzing the genetics of cementum formation.

Though their work is individually funded and conducted, they have a tight web of collaborations with one another as well as outside the CCMB. Their team efforts have helped to produce a more cohesive picture of tooth development. It’s a picture that brings them one step closer to producing artificial tooth structures that might one day be suitable for use by dentists to restore teeth.

A Crown of Crystals Fashioned by Genes 

CCMB researcher Dr. Malcolm Snead brings attention to a point he finds fascinating about enamel. He argues that what is particularly amazing about tooth enamel is even though it’s made almost entirely of hydroxyapatite, it has properties that are not consistent with hydroxyapatite crystals found in nature. When we chew we generate hundreds of thousands of pounds per square inch of masticatory force. If you were chewing on something that was just hydroxyapatite crystals, it would shatter like glass. But because of the way that the cells that make up the tooth are able to weave bundles of very small hydroxyapatite crystals, the final structure of mammalian enamel is remarkably tough and strong in all planes of orientation, explains Snead.

Professor Snead’s collaborator, Dr. Shane White, has utilized fracture mechanics techniques to show that the biological organization of hydroxyapatite not only renders enamel strong, but makes it wear-resistant as well. The organization of enamel’s hydroxyapatite rods and interrods makes it two to three times as strong as its constituent hydroxyapatite crystals. This interweaving arrangement of rod to interrod also provides enamel with a minimal degree of anisotropy thus preventing enamel from cracking down the length of its hydroxyapatite rods.

Neither the mechanical feature of strength or wear-resistance is replicated outside biology in quite the same way. Their co-existence in the tooth speaks to the importance of its biological manufacturing. Cell and protein-driven organizational efforts, if you will, act in all three axes to assemble the unique structure of each tooth. On one side there are the cells that make the enamel, the ameloblasts. Snead, White and Paine all describe the ameloblasts as performing a type of orchestrated dance when they deposit proteins into the enamel matrix. The ameloblasts have different vectors of movement that help give rise to a 3-dimensional continuum of protein deposition and subsequent crystal growth. Thus, even if the appropriate proteins are in the right place at the right time, hydroxyapatite won’t organize in a tight, interlacing network with out the movements of the ameloblasts.

Then there are the proteins. The developing tooth doesn’t start out as a mineralized tissue. Mineralized tissues, particularly teeth, are unique because the protein matrix, which is genetically controlled, is only a scaffolding that gets largely removed, says White. The chief protein in the enamel matrix, amelogenin, is not the end product but rather a necessary means to it. Although amelogenin makes up over 98% of the protein secreted by the ameloblasts, it is systematically removed from the matrix and replaced with hydroxyapatite crystals. 

But what role does amelogenin serve as the majority player in the enamel matrix? Is it an organizer, a crystal nucleator, a space-filler? Whatever the determined role, amelogenin is a certain player in enamel formation. This much has been shown both in the laboratory and in nature. Genetic defects in the amelogenin gene result in an enamel defect in humans called amelogenesis imperfecta (AI). This disease is characterized by soft, rapidly worn tooth enamel. With AI we know the specific defect. If we can understand the mechanism it would advance the understanding of families and of groups with similar phenotypes but different genetic defects, says White.

Snead and Paine have looked at amelogenin from a genetic perspective to determine what properties of the protein make it essential to enamel formation. Building on biochemical observations made by Fincham and Oldak that amelogenin self-assembles into discrete spheres, Snead and Paine systematically deleted parts of the amelogenin protein to see if they could interrupt this biologic activity. They were successful with their in vitro efforts and located domains in amelogenin that could disrupt its self assembly. Under the powerful view of atomic force microscopy done by Oldak and Fincham, Snead and Paine saw that their truncated amelogenins altered the formation of amelogenin nanospheres.

With a battery of in vitro support and the existence of the AI genetic defect in mind, Snead and Paine took an important final step with the strategically shortened amelogenin sequences. They introduced them into transgenic mice. The enamel latticework in these transgenic animals was different from normal enamel in many respects. The rod-interrod architecture was not as uniform as the enamel from the regular mice. The crystals don’t reliably organize in a parallel fashion and look, as Paine describes, as if there’s fusion amongst crystallites. Crystal disorganization varied depending on what part of the amelogenin molecule had been deleted.

Detecting the impact these changes have on enamel’s structural integrity has proved difficult. White, who studies enamel’s mechanical properties at the micro-scale doesn’t observe significant differences between the genetically altered mice and the wild-type mice. Measuring differences may be a matter of scale however. Collaborators from Seattle have taken the mechanical analysis of enamel to the nanoscale, three levels of magnitude smaller than the micro level. At this level, enamel from the genetically altered amelogenin mice is significantly weaker. As Paine explains, the nanoscale measurements reflect the integrity of individual crystallites, It’s a phenotype related to individual ameloblasts secretory products. But the sum of the whole isn’t dramatically different with respect to mechanical properties.

White agrees that, Chasing the phenotype has been more elusive. It seems we’ve found some mineralization differences that can be seen by electron microscopy and in mechanical behavior. But these differences are not as big as we expected them to be. The lack of an outstanding defect in the mechanical properties could be due to a number of factors, explain Paine and White. The genetically altered mice express the defective protein, but there is a question as to what level this expression represents. The group also points to the possibility that the transgenic animals are mosaic and that some ameloblasts express high levels of the mutant protein while others do not.

Understanding why particular enamel configurations fail and others do not would give dental scientists valuable insight into what amelogenin’s normal function is during tooth development. As such the group of collaborators continues to manipulate the genetics of enamel matrix proteins in hopes of producing an enamel facsimile that will fail.

As the CCMB researchers have pointed out previously, making enamel will take more than getting the right proteins in the right place. Making enamel would also entail getting the cells to recap their movements of coming together and moving apart. What that really means, says White, is the idea of producing cloned enamel will be much more distant than has been previously thought. But Snead, Paine, White and their collaborators have a new model about enamel formation that advances previous models. While the previous model has accurate and insightful parts, White adds enthusiastically, We think we can go a step beyond that and relate our findings to the cells and to the cellular and molecular mechanisms.

The Chemistry Behind Bioinspiration 

CCMB researchers Janet Oldak and Alan Fincham analyze enamel from a biochemist’s points of view. When the molecular biology advances, the genes are cloned and the corresponding recombinant proteins are expressed, It’s time again to do some chemistry, explains Oldak.

When Oldak was trained as a graduate student she was taught to look at biology as something that was to be understood, not copied. But people are bolder now in what they believe is possible, at least in words. Of course, producing a copy of a biologic structure is easier said than done. There are endless questions. I think we study enamel for the sake of better understanding how biology allows it to be produced so that someday we can make a synthetic material, explains Oldak.

Oldak and Fincham’s approach is not to let ameloblasts carry on the business of making their own enamel but rather to control the individual steps of biomineralization at the biochemical level. The two scientists utilize a cell-free system to analyze how individual molecules and ions like amleogenin, calcium and phosphate interact to make hydroxyapatite crystals. Cells are nowhere to be found.

Mineralization doesn’t happen just anywhere. It requires certain conditions. The cells that are responsible for making the enamel and the dentin in the teeth, for example, create a defined space by secreting a protein laden organic matrix that contains, what scientists call, a nucleator. A supersaturated solution of calcium and phosphate ions then fills this space. Under these conditions, the nucleator initiates crystal growth and might even serve to control crystal orientation.

Something controls crystalization, says Oldak. But the question is what? Is it the cells and how they time the secretion of inorganic ions and directive proteins like amelogenin, enamelin, proteinases and ameloblastin? Or is it the bioactive proteins themselves and their complementary interactions with inorganic ions? Does crystal orientation start with nucleation or is it controlled post-nucleation? Could it be a combination of these factors?

Oldak considers multiple parameters when thinking about the regulation of hydroxyapatite crystal growth. The enamel protein matrix elaborated by the ameloblasts and perhaps, in small part, the nearby, dentin-producing odontoblasts may somehow confine space for crystal growth. Likewise, proteins from the enamel matrix are hypothesized to have specific charged residues that interact with particular faces of the growing hydroxyapatite crystals. This type of interaction would promote unidirectional growth of crystals. Combine this growth pressure with polarized ameloblasts that pump calcium into the protein matrix. The crystals, whose growth was initiated near the DEJ, grow toward this source of calcium. If the predominant matrix protein, amelogenin, surrounds the growing crystals then one could also rationalize that the ions only contact one face of the nascent crystal further promoting directional growth.

One of the strengths of Oldak’s in vitro system is that she can isolate specific parameters or construct a multi-parameter system and determine what promotes crystal formation. In the recent past, she and Fincham analyzed strategically shortened amelogenin proteins, to see how amelogenin assembly and crystal formation properties were affected in vitro. This is something that we’d like to do in the future with both amelogenin and other enamel matrix proteins like enamelin, says Oldak. The group would like to look at not only how these enamel matrix proteins interact and influence crystalization they’d also like to add an additional parameter. They want to include a cationic selective membrane in the mix to emulate the directional supply of calcium ions these proteins would encounter in the in vivo environment.

There are additional configurations of enamel matrix proteins that Oldak would like to study utilizing this system. All possibilities speak to the importance of finding what molecule initiates hydroxyapatite crystal formation whether this molecule is made by ameloblasts or odontoblasts. Oldak contemplates several approaches to the question. She proposes to fix enamelin and amelogenin proteins to a surface, bathe them in superstaurated calcium and phosphate solution and monitor them with atomic force microscopy for nucleation and subsequent crystal formation. A similarly constructed scenario could be performed with amelogenin and dentin phosphoprotein or with amelogenin itself. The idea is to isolate components and see how their interaction affects hydroxyapatite formation.

Oldak is hesitant to claim that a biomimetic version of enamel is near on the horizon. But that doesn’t mean that important developments can’t come from experiments being done in the here and now. The idea is to learn scientific details and use them in the future to develop a material that is not necessarily enamel but may have some of its properties, says Oldak. She refers to a recent Material Research Society Meeting where she met someone from Bell laboratories. The company doesn’t conduct biological research. They do R & D on wires and electronic components in hopes of producing better conductive materials. But it’s amazing, she says, They’re adapting ideas from nature and thinking about adding pieces of peptides to certain crystals to improve their properties. In this case, scientists want to see if they can control the nature of the material, it’s conductivity, it’s orientation. Oldak takes comfort with the approach of taking inspiration from nature rather than trying to copy it. That’s what we’re talking about, she explains, They found out that in biology a peptide can control orientation. So let’s take advantage of this and try to control what we are interested in.

A Biologic Bumper- the DEJ

Talking about failure is difficult for some. But for Dr. Shane White, the topic of failure is constantly on his mind. The USC-trained scientist and clinical faculty alum has devoted his dental research career to discovering why, how and where materials, both man-made and natural, fail. As White explains, Most artificial crowns are made of porcelain and metal. Porcelain looks good and provides wear resistance whereas the metal provides toughness and strength. Unfortunately, the bond between the two substances is prone to failure and the porcelain often falls off the metal substructure.

But there is a remarkable natural bond that exists between the hard, wear-resistant enamel and its underlying soft yet tough dentin that resembles the bond between porcelain and metal. Both interfaces are between a brittle, weak material (porcelain and enamel) and a tough, resilient material (metal and dentin). But there’s one important difference. The natural bond is much more reliable. Scientists like Shane White and Malcolm Snead want to know why.

As Snead explains, the interface, known as the dentin enamel junction (DEJ) somehow links structurally dissimilar minerals so they function as a similar unit. When the teeth come together to chew, mechanical loads are transferred from enamel onto dentin in such a way that nothing breaks down, not the DEJ, not the underlying dentin, or the overlying enamel. We’re interested in trying to understand how that interface forms because we believe the DEJ is the critical component to restoring any dental structure, explains Snead.

Despite the extremely unusual and embryological and mechanical significance of the DEJ it’s received remarkably little attention, says White. Relatively little was known about the structural properties of the junction between the dentin and the enamel besides what had been produced from descriptive light and scanning electron microscope images. With these visualization tools scientists had noted a sharp boundary separating the dentin from the enamel. White used a simple test to show that a gradient of hardness existed at the DEJ between dentin and enamel. Rather than an abrupt line between the two materials, a gradual transition existed.

White’s physical characterization of the interface demonstrated the mechanical importance of the DEJ. What White and Snead now endeavor to do is understand the biology behind the DEJ’s mechanical integrity. They start their inquiry with the background that tissue interactions between epithelial and mesenchymal cells of the developing tooth result in the exchange of numerous genetic signals. These signals come in the form of bone morphogenetic proteins (BMPs) and transcription factors. Their presence at the right time and place initiates a cascade of gene regulation and results in the production of the tooth, dentin, enamel and all other tooth specific structures.

Once enamel and dentin were made scientists thought the conversation between these tissues stopped, remarks Snead. But better molecular tools have allowed scientists to assess early patterns of gene expression at the forming interface. Interestingly, as enamel and dentin begin to form, ameloblasts and odontoblasts produce proteins that were originally thought to be specific to one cell type and not the other. That is, for a brief period of time, ameloblasts make odontoblast-specific dentin products like dentinsialophosho protein (DSPP) and odontoblasts make ameloblast-specific enamel products like ameloblastin. We think that this shuffle of molecules back and forth is not only responsible for setting up the functions of the cells in terms of what they make, but in fact this temporary admixture of proteins may be the formation of the DEJ, explains Snead. The DEJ may be an embryologic mixing of gene expression parameters whose regulation is controlled simply by how many proteins are there, when they get there and how they mix, he continues.

Snead’s group uses a variety of approaches to explore the significance of this unique embryologic gene expression profile. They explore the relationship between the various proteins in vitro to determine what portions of the proteins facilitate or prevent interactions. They then introduce the altered gene into animals to either remove the gene’s function or enhance it. That becomes an important way to ask what functional significance a protein has in the process of DEJ formation, explains Snead, We also overexpress a particular protein to see if we can increase the amount of DEJ that’s formed with a candidate critical component. Should the candidate protein be important to DEJ formation, Snead and his collaborators believe that forcing cells to increase production of the protein would increase the size of the interface. We want to be able to remake that DEJ by reintroducing the appropriate proteins so that when a biomimetic or synthetic material is put onto a restoration site we have a structure that’s very similar to the original product, says Snead.

Scientists have spent decades searching for a sound way to bind restorative materials to dentin. White admits, We’re still not there yet. He and Snead agree however that producing something similar to the DEJ will allow natural and artificial materials to be bound soundly to existing tooth structures. A design that’s been engineered by millions of years of evolution is going to be difficult to improve upon, says Snead, So we’re trying to take a scintillation from nature and reapply it to make a material that will permit enhanced restorations to be produced. It’s one of the most complex clinical challenges we’ve got, says White.

A Biological Super-Glue 

Teeth don’t just sit embedded in bone. They’re attached via a shock-absorbing ligament called the periodontal ligament (PDL) and coated with a mineralized material called cementum that keeps them in firm contact with the bone. Together these tissues form the periodontum. The advantage of the periodontum is that it creates a cushion. So when there’s stress from biting and chewing on the teeth this buffers the impact, explains researcher, Dr. Margarita Zeichner-David.

The problem with this otherwise well-designed tissue is that it is susceptible to the ravages of bacteria associated with periodontal disease. The onset of periodontal disease is marked by destruction of the cementum, detachment of the PDL and pocket formation between the tooth and the bone. As the disease progresses, the bone that surrounds the tooth continues to break down and recedes down to the tooth root resulting in less support. With nothing to support them, the teeth eventually become too mobile to resist the forces of chewing and ultimately can be lost.

Many researchers and dentists, including Zeichner-David, believe that cementum plays an integral role in protecting the tooth from the ravages of bacterial action and thus the prevention of periodontal disease. So the question is, can cementum be replaced once it has been destroyed? Margarita Zeichner-David and her group have recently been awarded a grant to study this very prospect. The chief obstacle in addressing this question lies in the fact that at present nobody is certain what cells make cementum or what proteins these cells make that are specific to the structure itself. As Zeichner-David explains, There is a lot of controversy about what cells deposit the cementum layer. It’s a mineralized tissue that has similarity to bone but isn’t exactly bone. It has similarity to dentin but isn’t exactly dentin. It has some similarity to enamel but it’s not exactly enamel.

Part of what makes identifying the cementum’s origin difficult is that the proteins found in cementum seem to be found in a variety of other tissues. The cells that are in the embryologic vicinity during root development are epithelial in nature, the so-called HERS cells (Hertwig’s epithelial root sheath). So one might expect they’d make enamel proteins. At first, they do. They make some enamel specific proteins such as ameloblastin. The cementum, however, also contains bone specific proteins like osteopontin and osteocalcin. But it doesn’t necessarily make sense that epithelial cells would make non-epithelial proteins. Zeichner-David speculates that HERS cells transform from an epithelial phenotype to a mesenchymal phenotype. This specialization results in the production of a mesenchymal-like cell called a cementoblast that is capable of producing mesenchymal proteins such as osteopontin and osteocalcin.

What is also difficult about studying a theoretically temperamental cell type is that it’s hard to tell when you’ve actually got it in your grasp. When a cell produces a unique protein it’s much easier to track down. Fortunate for those who study cementum formation, a candidate product has been identified. Though a role for the cementum attachement protein, or CAP, has not been completely defined, it has been isolated from cementum and bears no resemblance to either bone or tooth specific proteins. Encouraging, as Zeichner-David explains, is the fact that, The HERS cells make the CAP after they’re maintained in culture. But there’s still controversy as to whether other cell types can make CAP, like PDL, dental follicle or even bone cells. Until we find a specific cementum protein, demonstrate which cells make it and trace back the lineage of those cells this dispute won’t be resolved, says Zeichner-David.

As Zeichner-David targets the HERS cells as the originators of cementum, her efforts to re-engineer degenerated cementum are focused around the activities of these cells. Her group has worked with a periodontal disease model that entails creating a bacteria-free, artificial wound in the periodontum of rats. We create an opening between the tooth and the bone, scrape the cementum, remove some of the PDL and some of the bone, she explains. The Zeichner-David group then reintroduces components they think are essential to the generation of cementum. One strategy, she hypothesizes, is to make use of the HERS cells as a regenerative stem cell population. We take HERS cells we’ve isolated, place them into the surgically created wound and follow them to see what they do. We can follow them because they’ve been marked with a retroviral b-gal construct that causes them to turn blue, says Zeichner-David. 

Experiments such as these allow the group to address what happens when the cells are reinserted into a compromised periodontum. Zeichner-David and her group can assess whether the HERS cells attach to the dentin, whether the cells continue growing and proliferating, and whether they transform into mesenchymal cells. We want to know if they regenerate the cementum and, if so, can they regenerate the PDL and reattach to the bone, she asks. If the HERS cells are capable of re-stimulating cementum formation then cells could be used rather than compounds to repair the periodontum. 

The idea of using the cells alone is appealing. This is, after all, the drive behind tissue engineering, explains Zeichner-David, pushing stem cells to do what they’d done before, using your own system for regeneration. If HERS cells happen to be the driving force behind cementum formation, why not take advantage of what they already know how to do?

Please visit the National Library of Medicine's Pub-Med ( www.ncbi.nlm.nih.gov ) for a comprehensive list of recent publications by authors featured in this article.