While it is clear from these studies that high concentrations or intakes of ascorbate can reduce tissue concentrations of Cu in a variety of species, there is a less clear consensus and uniformity with respect to possible functional effects of ascorbate overload on Cudependent processes in viva Several early studies on aortic rupture in chicks (Starcher et al. 1964;Carlton & Henderson, 1965;Hill & Starcher, 1965) obtained strong evidence for a gross failure of elastin cross-linking, attributed to reduced activity of the key enzyme, lysyl oxidase, in birds that were exposed simultaneously to excessive dietary ascorbate, and to low dietary Cu levels. Studies in mammalian species, however, have not reproduced such a dramatic effect. Nevertheless, in view of the potential importance of this interaction, for human subjects who choose to take daily megadoses of ascorbic acid, there is a need for further studies in relevant animal models such as the guinea-pig, which like humans has an absolute requirement for dietary ascorbate, as well as for Cu. The possibility that high levels of circulating vitamin C may increase the likelihood of Fe acting pro-oxidatively, and thereby influence outcome, has been illustrated by recent studies of human pre-term infants (Silvers et al. 1994;Powers et al. 1995). Deleterious effects of high ascorbate intakes may thus involve more than one transition metal.Another separate but related question arises from the need for better biochemical tests for suboptimum v. optimum status, with respect to those functionally-critical processes which are dependent on specific micronutrients. Collagen cross-linking is such a process. A severe lack of Cu can impair the activity of the Cu-dependent enzyme lysyl oxidase, responsible for the initiation of cross-linking (Farquharson et al. 1989;Robins, 1994). Impairment of cross-linking may then result in loss of tensile strength. Lack of ascorbic acid is well known to affect the hydroxylation of collagen lysyl residues (Kivirikko & Myllyla, 1982;Yeowell et al. 1995), and this could, in theory, alter the pattern of collagen cross-links, by altering the ratio of deoxypyridinoline (derived from lysine) to pyridinoline (derived from hydroxylysine) (Robins, 1994).The dual purpose of the present study was: first to re-examine the interactions of ascorbate and Cu, by dietary modulation in young guinea-pigs, and second, to test the hypothesis that Cu and/or ascorbate status might alter the ratio of deoxypyridinoline collagen-derived cross-links in bone and in urine.
MATERIALS AND METHODS
Animals and dietsThe purified guinea-pig diet contained the following components (g/kg): sucrose 331, maize starch 50, ovalbumin 300, cellulose powder 150, maize oil 73, potassium acetate 25, choline chloride 2, magnesium oxide 5 , inositol2, salt mixture 60, vitamins (see later). The salt mixture was based on that of Greenfield et al. (1969), except that the CuS04 therein was omitted. It contained (g/kg): CaC03 205, CaHP04 325, Na2HP04 185, KC1 205, MgS04 70, MnS04 4-5, Fe-citrate...