The Structure and Function of the Bone Protein Osteocalcin

Submitted by Richard V. Prigodich

Department of Chemistry

ABSTRACT: Osteocalcin is a small protein found in bone tissue. Osteocalcin is the seventh most abundant protein in the human body and the second most abundant protein in bone tissue after collagen. The sequence of amino acids that comprise the primary structure of osteocalcin has been carefully conserved by evolution. This means that the identity and sequence of the 20+ types of amino acids found in the protein have changed little in evolutionary terms and it hints at an important role for osteocalcin in bone metabolism; a role that has as yet to be discovered. Recent work in our laboratory has shown that osteocalcin binds to both of the major components of bone: collagen, a flexible fibrous large protein, as well as apatite, the hard mineral component. The studies described in this proposal concern the fundamental three-dimensional structure of osteocalcin and how and where it may bind to collagen and apatite. They will elucidate the function of this small protein in the regulation of bone metabolism and test its potential use as a drug to promote the bonding of prosthetic devices to healthy bone.

BACKGROUND: If you have ever cursorily examined a piece of scrimshaw or the leg bone from your barbecued chicken, the bone tissue appears to be marvelously uniform and homogeneous. However, if you look more closely, especially at a fracture, you can begin to discern structure. What is most clear is that bone appears to be made of long, very fine, rod-like units. Indeed bone does have structural features. These structures cover distances as large as centimeters and as small as micrometers (one millionth of a meter; Currey, 1984). However, all these structures are the direct result of the structure and organization of the molecules that comprise bone tissue. The most abundant molecules in bone are apatite (HAP) and collagen.

Collagen is a protein. Proteins are polymers of amino acids. They are like strings of beads. The beads are the amino acids. There are about 20 or so different amino acids, or colors of bead. Each differently colored bead (amino acid) has a different set of chemical and physical properties. The particular combination and sequence of amino acids forces the protein to adopt a specific three-dimensional shape. The shape allows the protein to perform its particular function. Some proteins fold up into balls and increase the rates of biochemical reactions, trigger cellular events, participate in the immune response, etc. Some proteins form long rod-like structures and serve as structural supports. Collagen is one of these proteins. It is found in non-mineralized tissues such as cartilage, tendons, ligaments and skin. However in bone tissue, bone cells mineralize the collagen by surrounding it with the mineral HAP. Collagen is formed by 3 separate protein chains winding around each other like strands of a rope to form a rod 300 nanometers (billionths of a meter) long called tropocollagen. These rods have a head and a tail with a small gap at the junction. Lines of tropocollagen molecules lie next to each other in a staggered fashion to form a complex pattern. (Hodge and Petruska, 1963; Please see Figure 1 — Appendix V).

Ca₁₀(PO₄)₆(OH)₂ is apatite or hydroxyapatite (HAP), a mineral or salt composed of calcium (Ca²⁺), phosphate (PO₄^3-) and a form of water called hydroxide (OH^1-) (Posner, 1969). There are actually several other salts of calcium and phosphate and some are found in immature bone tissue. They differ in the ratio of Ca²⁺ to PO₄^3- and in the numbers and types of water molecules. The different compositions lead to different structures with different properties.

How bone tissue forms is poorly understood, but briefly this is what is known (Lowenstam and Weiner, 1989). The organic matrix, collagen, is synthesized first in the nascent bone tissue. Bone cells then concentrate calcium and phosphate and package these ions within small spheres called matrix vesicles that bud off from the cells and migrate toward the collagen. The vesicles release the ions near the surface of the collagen and calcium/phosphate minerals begin to form. As the mineral phase matures from an amorphous mixture of Ca²⁺ and HPO₄^2- into HAP, the HAP crystals grow into the gaps between the tropocollagen molecules. What is not understood is what triggers HAP deposition onto collagen at a specific site or what controls HAP crystal growth. Recently it has been shown that negatively charged acidic proteins, such as osteocalcin, are involved in similar processes. Furthermore, bone is a dynamic tissue. It is constantly being remodeled; resorbed and layed down again. It is known that the initial events in bone resorption involve a shrinkage in the size of cells covering bone tissue. This exposes the tissue to enzymes that digest collagen molecules and release osteocalcin (Bilezikian, et al. , 1996). It is also known that osteocalcin is then exposed to enzymes that can fragment osteocalcin structure. After these initial events some unknown agent mobilizes bone cells known as osteoclasts to migrate toward and then to excavate the solid bone tissue. Can osteocalcin play a role in this process as well?

Osteocalcin (Lian and Gundberg, 1988), or BGP, has only 49 amino acids, a 9- charge and it is the seventh most abundant protein in the body. The sequence of amino acids is highly conserved from swordfish to humans. Bovine and human BGP only differ by two amino acids, making bovine BGP a good model for human BGP. There appears to be no defined structure for osteocalcin in solution (Prigodich et al. , 1987), while the more relevant solid state structure to be found in bone tissue remains undetermined. These facts hint at an ancient, important, and possibly complex but as yet unknown role for BGP in bone metabolism and they underscore the importance of elucidating the structural details of osteocalcin.

GOALS, SIGNIFICANCE AND FIT WITH LONG-TERM RESEARCH PROGRAM: We have shown that osteocalcin binds collagen as well as apatite (Prigodich and Vesely, 1997). We must now attempt to find the specific collagen binding site (GOAL 1). It is my hypothesis that osteocalcin is bound to collagen and lies near the site of HAP crystal growth and that it controls the orientation of the crystal growth in bone (GOAL 2). This could be accomplished by osteocalcin preferentially binding to one crystalline face of HAP (Prigodich and Zager, 1996). Furthermore, I believe different regions of osteocalcin structure have been designed by evolution to perform these different tasks (GOAL 3). This is unusual for such a small protein (It could be the first documented case.), but it has significance beyond biochemical novelty. It is my contention that a fragment of osteocalcin may be released during the initial stages of bone resorption to act as the chemotactic (cell mobilizing) agent for osteoclasts. This would mean that small easily synthesized protein fragments could be used to facilitate the bonding of implants (artificial joints) into bone tissue by stimulating bone remodeling.

I have been researching metal ion binding and osteocalcin structure for nearly my entire scientific career. Our laboratory has made some of the most significant contributions to the understanding of the structure and behavior of this protein (Prigodich, et al., 1985, 1996, 1997). We will continue this important work for some time to come.

METHODS: Goals 1 and 2 rely on the use of atomic force microscopy (AFM). This instrumentation is available to me in MCEC thanks to Dr. Broadbridge. AFM uses the electrostatic repulsion between the electrons of the atoms comprising a surface and of the atoms in a very fine (10-50 atoms across) probe. In essence the AFM "feels" the contour of the surface with a resolution of several nanometers. High enough resolution to easily "feel" a protein molecule. The idea is to bind collagen to a mica surface, image the fibers with the AFM and then to repeat the experiment in the presence of osteocalcin to find where the osteocalcin binds on the collagen fiber. AFM has enough sensitivity to easily make such images. I would need $500 per year for these fine tips, mica surfaces and osteocalcin. In goal 3 I would need to pay for fragments of osteocalcin structure to be synthesized by an academic lab that provides such service at a reasonable rate, since we have no facilities at Trinity to carry out such syntheses. I would need $2100 per year for this part of the project. It costs $700 to have 100 milligrams of a peptide (protein fragment) to be synthesized. This would be enough for sufficient quantities of three judiciously chosen fragments per year. Fragment 1 would be studied by circular dichroism (CD), a spectroscopic technique now available at Trinity, that yields global structural information for molecules.. This would be a portion of osteocalcin that has a collagen-like sequence of amino acids (Please see Figure 3 — Appendix V). Several variable lengths of this region would need to be studied. We would compare the CD of the osteocalcin fragment to the CD of collagen fragments to see if this portion of osteocalcin adopts a structure similar to that of the collagen fiber. If the CD data are similar, it is strong evidence that this portion of osteocalcin may bind to collagen fibers the way collagen fibers interact with each other. Fragment 2 would be the part of osteocalcin that is the likely candidate for binding to apatite (Please see Figure 3 — Appendix V). This could be easily tested by simple methods we have used before (Prigodich and Vesely, 1996). Fragment 3 would be that portion of osteocalcin that is known to be cleaved in the initial stages of bone remodeling. With these peptides in hand I would seek a collaboration with a lab which studies chemotaxis, the movement of cells in response to a chemical agent. Many different peptides may need to be tested to narrow dwon the essential region of the osteocalcin structure to induce chemotaxis. Without the peptides, I of course have nothing to offer to initiate a collaboration (The chicken and the egg problem.). There are several active groups at the UCONN Health Center engaged in this type of research, such as Dr. L G. Raisz, who would be interested in such a cooperative study.

FUTURE GOALS: With data collected in the studies funded by this grant, and the collaboration that I will be able to establish with funds from this grant. I will be in a position to apply for a substantial grant from the NIH: AREA (received two already), an NIH High Risk/High Impact Grant, or an R03.

REFERENCES

Currey,J. (1984) in The Mechanical Adaptations of Bone,Princeton Univ. Press, Princeton, N.J.

Posner,A.S., (1969) Physiological Reviews 49, 760-792.

Lowenstam,H.&Weiner,S. (1989) in On Biomineralization,pp 144-188, Oxford Press, N.Y.

Lian J.B.&Gundberg,C.M. (1988) Clinical Orthopaedics and Related Res. 226, 267-291.

Bilezikian, et al., (1996) in Principles of Bone Biology, Chapter 18, Academic Press, N.Y.

R.V. Prigodich, T. O'Connor and J.E. Coleman,"1H,113Cd and 31P NMR of Osteocalcin

6298.

R.V. Prigodich and M. Zager, "Indexing Crystal Faces", Powder Diffraction, (1995) 10 (2), 127-

128.

R.V. Prigodich and M. Vesely,"Characterization of the Complex Between Bovine Osteocalcin

and Type I Collagen", Archives of Biochemistry and Biophysics (1997) 345 (2), 339-341.