Impact of Side Chain Polarity on Non-Stoichiometric Nano-Hydroxyapatite Surface Functionalization with Amino Acids

Abstract

In this study the affinity of three amino acids for the surface of non-stoichiometric hydroxyapatite nanoparticles (ns-nHA) was investigated under different reaction conditions. The amino acids investigated were chosen based on their differences in side chain polarity and potential impact on this surface affinity. While calcium pre-saturation of the calcium-deficient ns-nHA was not found to improve attachment of any of the amino acids studied, the polarity and fraction of ionized functional side groups was found to have a significant impact on this attachment. Overall, amino acid attachment to ns-nHA was not solely reliant on carboxyl groups. In fact, it seems that amine groups also notably interacted with the negative ns-nHA surface and increased the degree of surface binding achieved. As a result, glycine and lysine had greater attachment to ns-nHA than aspartic acid under the reaction conditions studied. Lastly, our results suggest that a layer of each amino acid forms at the surface of ns-nHA, with aspartic acid attachment the most stable and its surface coverage the least of the three amino acids studied.

Figure 1

Zeta potential analysis of functionalized ns-nHA. Pre-conditioning of blank ns-nHA and addition of any of the amino acids studied notably increased the zeta potential (i.e. made it less negative). Of particular note is the near-neutral zeta potential of aspartic acid functionalized ns-nHA prepared under calcium-saturated and pH 7.8 conditions (indicated by *). This balanced charge likely results from a more balanced binding of both carboxyl and amine groups to the ns-nHA surface (compared to the other amino acids). Data reported as average ± one standard deviation (n = 5). ns-nHA unsaturated with calcium is given as US, while that pre-saturated with calcium is given as CaS. Horizontal bars indicate statistically detectable difference (p < 0.05).

Figure 2

ATR-FTIR analysis of COO⁻ peak area normalized to PO₄³⁻ peak area for blank ns-nHA. Increasing the reaction pH notably increased the ratio of COO⁻ symmetric stretch peak area to PO₄³⁻ vibration peak area. The saturation of ns-nHA with calcium (CaS) did not result in a detectably different COO⁻ peak area compared to unsaturated (US). Data reported as average ± one standard deviation (n = 3). Horizontal bars indicate a statistically detectable difference (p < 0.05).

Figure 3

ATR-FTIR analysis of the amino acid functionalized ns-nHA COO⁻ peak area normalized to PO₄³⁻ peak area and peak area of blank ns-nHA subtracted for each condition. Aspartic acid-functionalized ns-nHA was found to have the greatest symmetric COO⁻ peak area. However, there are two COO⁻ groups per aspartic acid molecule (glycine and lysine only have one COO⁻ group per molecule). As shown by the black horizontal dashed line on the aspartic acid column, the COO⁻ peak area analysis suggests that there is a comparable amount of aspartic acid attached to the ns-nHA surface as achieved with the other amino acids. Data reported as average ± one standard deviation (n = 3). ns-nHA unsaturated with calcium is given as US, while that pre-saturated with calcium is given as CaS. Horizontal bars indicate statistically detectable difference (p < 0.05).

Figure 4

ATR-FTIR analysis of the amino acid functionalized ns-nHA NH₂-H⁺ peak area normalized to PO₄³⁻ peak area. Lysine-functionalized ns-nHA was found to have the greatest symmetric NH₂-H⁺ peak area. However, there are two NH₂-H⁺ groups per lysine molecule (aspartic acid and glycine only have one NH₂-H⁺ group per molecule). As shown by the white horizontal dashed line on the lysine column, the NH₂-H⁺ peak area analysis suggests that there is a comparable amount of lysine attached to the ns-nHA surface as glycine. Data reported as average ± one standard deviation (n = 3). ns-nHA unsaturated with calcium is given as US, while that pre-saturated with calcium is given as CaS. Horizontal bars indicate statistically detectable difference (p < 0.05).

Figure 5

Molar ratio of total sites grafted on the ns-nHA surface using different amino acids (AA) following the ninhydrin protocol. Aspartic acid attached to fewer available graft sites on ns-nHA than glycine or lysine; this is likely indicative of the more balanced layer of aspartic acid on the ns-nHA surface. Data reported as average ± one standard deviation (n = 5). ns-nHA unsaturated with calcium is given as US, while that pre-saturated with calcium is given as CaS. Horizontal bars indicate statistically detectable difference (p < 0.05).

Figure 6

Molar ratio of total sites grafted on the ns-nHA surface using different amino acids (AA) following the fluoraldehyde protocol. Aspartic acid was again found to attach to fewer available graft sites on ns-nHA than the other two amino acids. Data reported as average ± one standard deviation (n = 5). ns-nHA unsaturated with calcium is given as US, while that pre-saturated with calcium is given as CaS. Horizontal bars indicate statistically detectable difference (p < 0.05).

Figure 7

A proposed schematic for the dominant mode of adsorption of the different amino acids at the ns-nHA surface. Both amine and carboxyl functional groups contributed to the attachment of the amino acids to the ns-nHA surface. This schematic generally encompasses each reaction condition, owing to little dependence observed on reaction condition (i.e. pH or pre-saturation with calcium); however, it is important to recognize that the fraction of amine groups which are ionized will change in the schematic as pH changes. Overall, the ease of approach towards the negatively charged ns-nHA surface is greatest for Lysine and Glycine, and more challenging for the negatively charged Aspartic acid. As a result, the attachment of Aspartic acid to the ns-nHA surface is more limited compared to the other two amino acids.

Figure 8

Molecular structure of Aspartic acid, Glycine, and Lysine with pK_a values of amine and carboxyl groups represented near corresponding functional groups (pK_a values from Nelson et al.).