Grafting Strategies of Oxidation-Prone Coiled-Coil Peptides for Protein Capture in Bioassays: Impact of Orientation and the Oxidation State

Abstract

Many biomedical and biosensing applications require functionalization of surfaces with proteins. To this end, the E/K coiled-coil peptide heterodimeric system has been shown to be advantageous. First, Kcoil peptides are covalently grafted onto a given surface. Ecoil-tagged proteins can then be non-covalently captured via a specific interaction with their Kcoil partners. Previously, oriented Kcoil grafting was achieved via thiol coupling, using a unique Kcoil with a terminal cysteine residue. However, cysteine-terminated Kcoil peptides are hard to produce, purify, and oxidize during storage. Indeed, they tend to homodimerize and form disulfide bonds via oxidation of their terminal thiol group, making it impossible to later graft them on thiol-reactive surfaces. Kcoil peptides also contain multiple free amine groups, available for covalent coupling through carbodiimide chemistry. Grafting Kcoil peptides on surfaces via amine coupling would thus guarantee their immobilization regardless of their terminal cysteine's oxidation state, at the expense of the control over their orientation. In this work, we compare Kcoil grafting strategies for the subsequent capture of Ecoil-tagged proteins, for applications such as surface plasmon resonance (SPR) biosensing and cell culture onto protein-decorated substrates. We compare the "classic" thiol coupling of cysteine-terminated Kcoil peptides to the amine coupling of (i) monomeric Kcoil and (ii) dimeric Kcoil-Kcoil linked by a disulfide bond. We have observed that SPR biosensing performances relying on captured Ecoil-tagged proteins were similar for amine-coupled dimeric Kcoil-Kcoil and thiol-coupled Kcoil peptides, at the expense of higher Ecoil-tagged protein consumption. For cell culture applications, Ecoil-tagged growth factors captured on amine-coupled monomeric Kcoil signaled through cell receptors similarly to those captured on thiol-coupled Kcoil peptides. Altogether, while oriented thiol coupling of cysteine-terminated Kcoil peptides remains the most reliable and versatile platform for Ecoil-tagged protein capture, amine coupling of Kcoil peptides, either monomeric or dimerized through a cysteine bond, can offer a good alternative when the challenges and costs associated with the production of monomeric cysteine-tagged Kcoil are too dissuasive for the application.

Figure 1

Ecoil binding to grafted Kcoil. SPR sensorgrams corresponding to Ecoil peptide injections over 200 RU of grafted Cys-free Kcoil by amine coupling (A), 15 RU of grafted Cys-Kcoil by thiol coupling (B), and 350 RU of grafted dimeric Cys-Kcoil by amine coupling (C). Sensorgrams were linearized and normalized (right panels, black). The slopes of the red fits estimate the apparent dissociation rate for the Cys-free Kcoil (D), the Cys-Kcoil (E), and the dimerized Cys-Kcoil (F) surfaces.

Figure 2

Repeatability of SPR biosensing experiments relying on coiled-coil-mediated capture of the ligand for different Kcoil grafting schemes. Ecoil-tagged ligand (FcγRIIIa V158) capture levels are evaluated with respect to the cycle number for two batches of experiments. The results are given for the amine-coupled Cys-free Kcoil (A), thiol-coupled Cys-Kcoil (B), and amine-coupled dimeric Cys-Kcoil (C). Overlaid sensorgrams of the interaction between TZM and FcγRIIIa_V158 are given for the same concentration of TZM injected in each cycle for amine-coupled Cys-free Kcoil (D), thiol-coupled Cys-Kcoil (E), and amine-coupled dimeric Cys-Kcoil (F).

Figure 3

Similarity scores for the analysis of the interaction between TZM and Ecoil-FcγRIIIa (V158), where Ecoil-FcγRIIIa is captured by amine-coupled Cys-free Kcoil (blue bars), thiol-coupled Cys-Kcoil (orange bars), or amine-coupled dimeric Cys-Kcoil (yellow bars). The similarity scores were calculated independently both for the analyte injection phase (A) and the dissociation phase (B). A mean sensorgram was obtained for each of the two 16-sensorgram batches of experiments. All 16 sensorgrams within a batch were then compared to the mean sensorgram. Results show the average similarity score for both batches of all grafting methods, with a comparison window varying from 0.1 to 10 RU. The results obtained for each batch are shown in Figure S1.

Figure 4

Repeatability of the Ecoil–Kcoil system for different Kcoil grafting strategies. Ligand (EGF) capture levels are evaluated with respect to the cycle number for two batches of experiments. The results are given for the amine-coupled Cys-free Kcoil (A), thiol-coupled Cys-Kcoil (B), and amine-coupled dimeric Cys-Kcoil (C). Overlaid sensorgrams of the interaction between EGFR and EGF are given for the same concentration of EGFR injected in each cycle for amine-coupled Cys-free Kcoil (D), thiol-coupled Cys-Kcoil (E), and amine-coupled dimeric Cys-Kcoil (F).

Figure 5

Similarity scores for the analysis of the interaction between EGFR-Fc and Ecoil-EGF, where Ecoil-EGF is captured by amine-coupled Cys-free Kcoil (blue bars), thiol-coupled Cys-Kcoil (orange bars), or amine-coupled dimeric Cys-Kcoil (yellow bars). The similarity scores were calculated independently both for the analyte injection phase (A) and the dissociation phase (B). A mean sensorgram was obtained for each of the two 16-sensorgram batches of the experiment. All 16 sensorgrams within a batch were then compared to the mean sensorgram. Results show the average similarity score for both batches of all grafting methods, with the comparison window varying from 0.1 to 1 RU. The results obtained for each batch are shown in Figure S2.

Figure 6

(A) Bioactivity of grafted Ecoil-EGF based on the coupling strategy used for the functionalization of the surfaces with Kcoil. Bioactivity is measured by Western blot in terms of relative EGFR phosphorylation levels. The signal is normalized to the signal obtained for the transient phosphorylation (after 5 min of exposure) of EGFR by 1 nM hEGF in solution and compared to the signal obtained when hEGF is non-specifically adsorbed on the surfaces. Values are mean ± standard deviation (n = 3). *p < 0.05 indicates statistically significant differences in bilateral t-test pairwise comparisons. (B) Representative western blot showing protein levels of phosphorylated EGFR (p-EGFR, ∼180 kDa) and β-actin (∼42 kDa) when surfaces are functionalized using amine coupling.