stress was greater in magnitude than the opposite polarity signal produced by the release of the stress,
leading to the postulate that some rectification of the one signal was occurring in the bone matrix.
In a search for the source of this possible rectification we studied the properties of each matrix
component separately and together in their normal configuration. We found that both collagen and
apatite had some properties similar to semiconductivity, with collagen appearing to be an "N" type
material and apatite a "P" type. The junction between the two in whole bone was then found to have
some electrical and photoelectric properties similar to those of a rectifying PN junction (53, 54). In this
view, the piezoelectric property of collagen generated the electrical signals upon the application and
release of mechanical stress, with the signal of one polarity rectified to some extent by the PN junction
of the collagen-apatite relationship. In bending stress therefore, the concave side of the bone
demonstrates an overall negative polarity under stress, while the convex side is primarily positive in
polarity. Since bone growth occurs according to Wolff's law on the concave side and bone resorption
on the convex side, negative potentials were postulated to produce stimulation of the osteoblasts and
osteocytes, while positive potentials were presumed to either facilitate bone resorption by stimulating
specific cells that destroy bone (osteoclasts) or merely not to produce any stimulation of bone growth
on the convex side. In a test of this hypothesis, Bassett, Pawluk and Becker were able to demonstrate
that bone growth did occur in the vicinity of a negative electrode with currents of less than 3 µamp,
while growth was absent around the corresponding anode (55). It should be noted that the currents were
not strictly analogous to those produced by the piezoelectric effect in whole bone, being continuous
rather than intermittent as would occur in normal usage.
The control system schematic for the growth of bone in response to bending stress may be
expanded as follows:
Later we explored the apatite-collagen junction in whole bone with quite different
techniques. With trace element analysis techniques, Spadaro, Becker, and Bachman
demonstrated that the bone collagen fibril possessed a surface site capable of absorbing
specific metallic cations depending upon the radius of the hydrated ion (56). Elements
with radii between 0.65-0.75 A and between 1.2-1.4 A were bound tightly to the fiber.
The copper ion in its cupric state (CuII) with a radius of 0.65 A was found to bind to
both collagen and apatite. We later utilized this property to explore the electronic state
of the binding site using electron paramagnetic resonance (EPR) techniques. This ion
normally has a simple EPR signal, but when bound to either collagen or apatite, it
demonstrated a complex resonance spectrum with the resonances in each case being
identical, indicating that the binding sites on both apatite and collagen were identical in
electronic configuration (57). This is of course a rather unique situation considering the
great difference between these two materials, one being fibrous protein and the other an
inorganic mineral crystal. We proposed that this structural similarity of the two materials
could be involved in the initial mineralization process (the deposition of the first apatite
crystals directly on the fibers) (58).
ELECTROMAGNETISM & LIFE - 41