Assessment of nonreleasing antifungal surface coatings bearing covalently attached pharmaceuticals

A. Surface-attached echinocandin drugs effectively prevent biofilm formation on substrates

The ability of the surfaces to kill attaching yeast cells and prevent biofilm formation was measured with static biofilm assay. Figure 2 indicates that Ani, Cas, and Mic, when grafted onto surfaces, eliminate the ability of Candida albicans to form colonies (zero colony counts) on these surfaces, as evidenced by the static biofilm assay. (The limit of detection of the assay is estimated at 54 CFU cm−2 for the probability of collecting one colony off samples for transfer to the assay.)

B. Antifungal surface coatings remain effective after multiple challenges with fresh inoculum

To investigate the contact-killing ability of surfaces, antifungal surface coatings were reused multiple times in microbiological assays. Figure 3 shows static biofilm assays conducted on Ani- and Cas- and Mic-grafted samples for five repeat assays.

C. Investigation of surface washing protocols

Figure 4 shows the antifungal testing result for samples washed by the rigorous washing protocol (Acetone and SDS 70 °C). When more rigorous washing was used, surfaces conjugated with Ani and Cas were effective in eliminating >106 CFU of C. albicans. However, the antifungal effect from Mic was eliminated (i.e., indistinguishable from the control).

D. Surface analysis reveals the fate of surface-attached antifungal drugs after different washing protocols

In the three-step coupling protocol, the first step involved functionalizing surfaces with carboxylate groups by plasma polymerization of PrA. This is a substrate-independent method for depositing a functional thin film interlayer to many bulk materials commonly used for medical device manufacture.28,3428. B. R. Coad, T. Scholz, K. Vasilev, J. D. Hayball, R. D. Short, and H. J. Griesser, ACS Appl. Mater. Interfaces 4, 2455 (2012). https://doi.org/10.1021/am300128n34. H. J. Griesser and R. C. Chatelier, J. Appl. Polym. Sci. 46, 361 (1990). https://doi.org/10.1002/app.1990.070460018 Surfaces with carboxylate groups were then activated using carbonyldiimidazole chemistry.3535. G. T. Hermanson, Bioconjugate Techniques, 3rd ed. (Academic, Boston, 2013), p. 259. Subsequently, incubated echinocandins form covalent ester bonds to surfaces through carboxylate groups (Fig. 1). The successful execution of the intended multistep coating scheme was verified by surface analyses using ellipsometry, XPS, and ToF-SIMS.Table I shows the measured thickness of surface coatings after various washing procedures. The initial PrApp surface coating was 19.6 nm thick and had a refractive index of ∼1.56 which is typical for polymeric layers.36,3736. N. Kehagias, S. Zankovych, A. Goldschmidt, R. Kian, M. Zelsmann, C. S. Torres, K. Pfeiffer, G. Ahrens, and G. Gruetzner, Superlattices Microstruct. 36, 201 (2004). https://doi.org/10.1016/j.spmi.2004.08.00537. J.-g. Liu and M. Ueda, J. Mater. Chem. 19, 8907 (2009). https://doi.org/10.1039/b909690f CDI activation increased the thickness slightly due to swelling of the polymer layer. After grafting of echinocandins and washing samples with acetone, water, and SDS, the layer thickness may have decreased slightly, but experimental errors make this hard to prove conclusively. However, it is evident that large thickness changes were not observed, revealing that the binding and washing conditions did not cause extensive damage or delamination to the thin polymeric surface coating.Table icon

TABLE I. Thickness measurements by ellipsometry of fresh PrApp, after activation, and after echinocandin immobilization followed by various washing treatments.

Thickness of surface coatings by ellipsometryPrApp19.6 ± 0.1 nmPrApp + CDI + washed by acetone22.8 ± 0.1 nm{…above+Ani+washedbyacetone…above+Cas+washedbyacetone…above+Mic+washedbyacetone23.4 ± 0.1 nm23.5 ± 0.1 nm22.9 ± 0.1 nm{…(Ani)above+washedbywater…(Cas)above+washedbywater…(Mic)above+washedbywater22.3 ± 0.1 nm22.7 ± 0.1 nm21.8 ± 0.1 nm{…(Ani)above+washedbySDS70°C…(Cas)above+washedbySDS70°C…(Mic)above+washedbySDS70°C20.2 ± 0.1 nm21.3 ± 0.1 nm19.9 ± 0.1 nmXPS was used to analyze the amount of nitrogen on surfaces during activation, immobilization, and washing (Fig. 5). The PrApp layer is composed only of C, H, and O elements. For the PrApp layer, XPS C1s high-resolution spectra showing the presence of a significant percentage of COOH groups are shown in the supplementary material,4343. See the supplementary material at https://www.scitation.org/doi/suppl/10.1116/6.0001099 for XPS spectra of PrApp coating. whereas the absence of COOH groups on the surface of ethanol plasma polymers produced in the same reactor has previously been reported.3838. B. R. Coad, K. Vasilev, K. R. Diener, J. D. Hayball, R. D. Short, and H. J. Griesser, Langmuir 28, 2710 (2012). https://doi.org/10.1021/la204714pSurfaces overlayed with imidazole carbamate and echinocandins possess these elements but also introduce N. Therefore, the presence of N in the XPS spectrum is diagnostic for the derivative reactions. For CDI-activated surfaces, 1.2% nitrogen was detected from PrApp by XPS after washing samples with acetone to remove excess CDI. These values are typical for thin overlayer coatings. The N signal by XPS increased to 5.4%, 3.8%, and 2.5% for PrApp-Cas, PrApp-Ani, and PrApp-Mic, respectively, reflecting the increased abundance of N in their chemical structures and an overlayer coverage greater than one monolayer. After conjugation, two different washing protocols have been used. First, after washing with acetone and water, the intensity of the N signal for PrApp-Cas, PrApp-Ani, and PrApp-Mic was reduced by ∼ 59%, 47%, and 40%, respectively, indicating the desorption of weakly bound molecules (Fig. 5). Additionally, after washing with acetone, the samples were further washed with PBS, SDS at 70 °C, and water. After washing with SDS and XPS, an extra 8% and 13% reduction in intensity of the N signal for PrApp-Cas and PrApp-Ani, respectively, was indicated, and there was no signal for the PrApp-Mic samples. This result showed that the washing step using SDS removed Mic from the samples, but not Ani and Cas, and this is the reason for the result presented in .Due to poor solubility of Mic in acetone and good solubility in water,44. D. W. Denning, Lancet 362, 1142 (2003). https://doi.org/10.1016/S0140-6736(03)14472-8 it is likely that acetone washing leaves some Mic physiosorbed onto the surface, and this becomes gradually desorbed during reuse assay ().

XPS survey spectra did not show any residues of elements such as Na, Cl, K, P, and S from the buffers used in binding and washing steps, after final rinses with water. Likewise no Si signals indicated that adventitious silicone contamination was not present.

E. ToF-SIMS data confirm the presence of structural fragments of surface-attached echinocandins

ToF-SIMS is a highly surface-sensitive analytical technique for molecular characterization of the top few nanometers of a surface interface. Surface molecules are bombarded with an ion source, resulting in the ejection of charged molecular fragments, which are collected by a mass analyzer. While XPS provides semiquantitative information about the abundance of chemical elements, ToF-SIMS provides complementary qualitative data about the partial molecular structure of surface molecules. Interpretation of the complex ToF-SIMS mass spectral data is assisted by PCA.

In this study, echinocandins were linked to the surface on one of two different plasma polymer coatings, EtOHpp and PrApp, using CDI as before. EtOHpp was used to explore different chemical ways to link echinocandins to polymer coatings but not optimized for biological studies. Nevertheless, similar results were obtained for either plasma polymer coating, and, thus, we report results from more detailed investigation from the EtOHpp coatings.

The difference between positive mass spectra recorded on the control EtOHpp surface and the surfaces of Ani and Mic is demonstrated by the scores plot shown in Fig. 6. The scores plot on the principal components axes PC1 and PC2 show that the positive mass spectra across the pairs EtOHpp-Ani (trace A) and EtOHpp-Mic (trace B) separate into two distinct clusters. This indicated that the surfaces were chemically distinct, consistent with the XPS data that indicated successful grafting. Figure 7 shows the loadings of positive mass spectra for the pairs EtOHpp-Ani (trace A) and EtOHpp-Mic (trace B). Data correlated in a logical manner with the relevant scores plots (Fig. 7). The C2H3O+ and C3H5O+ fragments are characteristic of the EtOHpp, whereas all the CH–N+ and CH–NO+ ions are associated with EtOHpp-Ani and EtOHpp-Mic. The principal positive loadings on PC1 are C5H10NO+ and C4H8NO+ fragment ions for both drugs and specific C24H23O2+ and C18H21N3O4+ fragment ions that can be derived from Ani and Mic structures, respectively (Table II). Notable in this data were structural fragments that originated from the cyclic polyamide ring structure and lipid tails characteristic of Ani and Mic.Table icon

TABLE II. Structural elements that could result from a conjugation of echinocandins with EtOHpp and PrApp. C4H8NO+ and C5H10NO+ ion fragments are shown in blue and green, respectively, and also C24H23O2+ and C18H21N3O4+ and C21H20NO3+ fragment ions are shown in red.

For PrApp-echinocandin surfaces, it has shown the same CH–N+ and CH–NO+ ions associated with Ani and Cas molecules21,2321. B. R. Coad, S. J. Lamont-Friedrich, L. Gwynne, M. Jasieniak, S. S. Griesser, A. Traven, A. Y. Peleg, and H. J. Griesser, J. Mater. Chem. B 3, 8469 (2015). https://doi.org/10.1039/C5TB00961H23. J. Naderi, C. Giles, S. Saboohi, H. J. Griesser, and B. R. Coad, J. Antimicrob. Chemother. 74, 360 (2019). https://doi.org/10.1093/jac/dky437 (Table II). However, a distinct large molecular fragment (C21H20NO3+) of lipid tail region has been detected from PrApp-Mic surfaces (Table II).

F. Discussion: Surface-immobilized echinocandins can play a role in preventing fungal adhesion and biofilm formation on surfaces

Human fungal pathogens associated with implanted medical devices cause infections that are difficult to combat and cause devastating consequences.7,247. C. Giles, S. J. Lamont-Friedrich, T. D. Michl, H. J. Griesser, and B. R. Coad, Biotechnol. Adv. 36, 264 (2018). https://doi.org/10.1016/j.biotechadv.2017.11.01024. G. Ramage, J. P. Martinez, and J. L. Lopez-Ribot, FEMS Yeast Res. 6, 979 (2006). https://doi.org/10.1111/j.1567-1364.2006.00117.x The problem is inextricably linked to biofilms and the inability of many drugs to penetrate the extracellular matrix to eliminate protected fungal cells on the surface of the material implant. With biomaterials science, it is possible to explore new strategies to prevent microbial colonization of materials, thus circumventing adhesion, the first stage in biofilm formation. In contrast to approaches that have been tried such as antifouling materials, or materials that release antibiotic agents, the approach here was to investigate whether antimicrobial agents can be chemically (covalently) immobilized on surfaces. Since immobilized agents are no longer able to diffuse freely through solution, the demonstration of a surface-mediated antimicrobial effect would provide valuable insight into how microbial pathogens explore surfaces, and could be potentially affected by agents that are toxic to them. This approach investigates potentially new mechanisms of action for antimicrobial compounds, ones that do not rely upon diffusion through solution to mediate their effect. Surface immobilization would also eliminate the circulation of compounds through the body and accumulation in sensitive excretory organs. Thus, surface formulations of antimicrobials could represent a way to extend antimicrobial drug lifetimes through novel use and repurposing them as contact-killing surfaces.Using the static biofilm method, we observed (Fig. 2) that antifungal surface coatings incorporating all three of the approved echinocandins were highly effective at preventing C. albicans adhesion on surfaces and complete prevention of biofilm formation, owing to their antimicrobial properties.

Next, we investigated whether these surfaces could be repeatedly challenged by fresh fungal inoculum by testing their efficacy, washing, and reusing the same materials in subsequent challenges. A strong case could be argued for the claim covalent attachment of a drug to a surface if it can be demonstrated that antimicrobial surface activity is maintained after repeated challenges with high doses of pathogenic cells, and extensive washing and cleaning of the surfaces in between challenges. We observed that surfaces with attached Ani resisted five challenges and no surviving cells could be colonized from surfaces. With Cas, starting with trial 4, some living cells could be cultured from surfaces; however, the percentage of viable cells was very low (<0.1% compared with control surfaces), demonstrating that the Cas surface coatings can be considered highly antifungal for five challenges. With Mic, however, starting from trial 3, there was no antifungal surface effect.

To understand this result in the context of chemical (irreversible) and physical (reversible) bonding to surfaces, the surfaces were extensively characterized by analytical instrumentation after employing various washing protocols. Biological and instrumental results show the effect that different washing protocols have in removing physisorbed compounds from surfaces. Here, it can be deduced that the effect of washing with acetone, water, and SDS at 70 °C reduces the amount of echinocandins on the surface (Fig. 5, nitrogen is decreased on the surface), and the thickness of the surface coating is decreased slightly, up to 3 nm (Table I). Notably, the nitrogen on the Mic surface goes to zero, meaning that no echinocandin remains. Thus, the strong antifungal performance of the Mic surface after 1 or 2 trials (Figs. 2 and 3) can be explained by the gradual release of Mic from the surface by the culture medium and rinses between trials, which eventually eluted the compounds from surfaces. It must be emphasized that, while it initially appeared that Mic was covalently bound to the surface, it was actually physisorbed, albeit quite strongly, and might have been mistaken for covalent attachment had detailed analytical and biological studies not been performed. Additionally, it is also notable that the surface coatings with Ani and Cas lost about 60% and 67% of the total bound amount after washing, showing the importance of accounting for the majority proportion of physisorbed compound initially present, capable of release.

After SDS washing, all traces of Mic were removed from the surface, while Cas and Ani remained as confirmed by XPS and ToF-SIMS analyses. Thus, SDS washing was the only way to ensure that strongly physisorbed compounds were removed; otherwise, there is potential for leaching. Cas and Ani’s presence on the surface, and covalent attachment, explains their sustained antifungal performance of these two surface coatings after multiple reuses. All of our surface and analytical data are consistent with the view that these compounds form a permanent antifungal layer that is surface-active, i.e., covalently attached compounds, and supported by analytical evidence, suggesting a monolayer.

For long-term stability, the chemical nature of the bond that links antibiotic compounds to surfaces is important to consider. In this study, compounds were linked through ester bonds, and over the course of reuse experiments, if hydrolysis of compounds was occurring, the extent was not great enough to show diminished performance. However, biological and analytical studies of surface activity should be evaluated to investigate the long-term stability of covalently attached compounds.

Currently, there are no standard methods for validating the performance of antimicrobial surface coatings with covalently attached drugs. Yet, it is clear that a system of assessment is needed to validate the nonrelease of compounds. Otherwise, as shown here and elsewhere,3939. J. Naderi, C. Giles, S. Saboohi, H. J. Griesser, and B. R. Coad, Biointerphases 13, 06E409 (2018). https://doi.org/10.1116/1.5050043 how it could be quite easy to mistakenly attribute the mechanism to a covalently attached drug, when a release mechanism was actually operating. We have shown that for surface-attached echinocandins, three pieces of evidence were required to make a strong claim for covalent-only attachment of molecules. (1) After binding antimicrobial compounds to the surface, rigorous washing is required (using organic solvents, aqueous buffers, and surfactants at 70 °C) to remove physisorbed compounds. (2) Additionally, sensitive instrumental analysis after washing is required to confirm that the compounds were still present, and that data are quantitatively and qualitatively consistent with a view that the remaining compounds form a monolayer. (3) Finally, biological assessments of the surfaces demonstrate that they remain antimicrobial after many multiple challenges using fresh inoculum, followed by cleansing washes in between. The five challenges used in this study were appropriate for our system comprised of very thin layers and relatively large molecules and probably represent a lower limit. It is unknown, however, whether these three criteria, including number of challenges, will be valid in other systems, particularly for thicker conjugating layers, different substrate materials, and the diffusivity of different antimicrobial molecules. Regardless, we advance the idea that these three criteria could be used as a starting point for developing a validated procedure for assessing antimicrobial surface coatings that are claimed to operate through contact with covalently attached molecules, and we invite researchers to investigate other systems with a hope that a standardized procedure can be developed and adopted.Finally, we conclude with a discussion of how covalently attached echinocandins (Cas and Ani) are presented toward interfacing cells and their mechanism of action. Previously, it was shown that Cas can be covalently immobilized through one or both of its pendant amine group onto material surfaces bearing aldehyde and epoxide functional groups.21,33,4021. B. R. Coad, S. J. Lamont-Friedrich, L. Gwynne, M. Jasieniak, S. S. Griesser, A. Traven, A. Y. Peleg, and H. J. Griesser, J. Mater. Chem. B 3, 8469 (2015). https://doi.org/10.1039/C5TB00961H33. T. D. Michl, C. Giles, P. Mocny, K. Futrega, M. R. Doran, H. A. Klok, H. J. Griesser, and B. R. Coad, Biointerphases 12, 05G602 (2017). https://doi.org/10.1116/1.498605440. S. S. Griesser, M. Jasieniak, B. R. Coad, and H. J. Griesser, Biointerphases 10, 04A307 (2015). https://doi.org/10.1116/1.4933108 In this study, both Cas and Ani were covalently immobilized to surfaces using CDI chemistry to their more numerous hydroxyl groups, thus presumably producing many more possible orientations at surfaces and alterations to the drugs’ chemical structure (Fig. 1). By comparing the results of CDI-conjugated Cas (Fig. 4) with those obtained previously where it was bound through amine groups, we see no qualitative or quantitative difference in antifungal susceptibility testing. The results with Ani are similar to Cas, suggesting that the surface activity of this drug is also not affected by chemical conjugation or orientation. An observation that does help to explain their surface activity comes from the ToF-SIMS analytical data, where it was shown that intact molecular fragments from the lipid tails of Ani and Cas could be liberated upon ion bombardment, and they contained no structural deletions or adducts. This suggests that these lipid tails were accessible at the interface and unconjugated. The importance of the lipid tail to the echinocandin’s antifungal activity is consistent with what is currently known from structure-activity studies conducted in solution,4141. M. B. Kurtz and J. H. Rex, in Advances in Protein Chemistry, edited by E. Scolnick (Academic, Cambridge, MA, 2001), Vol. 56, p. 423. and our results are consistent with this idea. These findings are important because they suggest that structurally similar “next generation” echinocandins that are currently undergoing clinical trials, such as Rezafungin (CD101),4242. A. K. Sofjan, A. Mitchell, D. N. Shah, T. Nguyen, M. Sim, A. Trojcak, N. D. Beyda, and K. W. Garey, J. Global Antimicrob. Resist. 14, 58 (2018). https://doi.org/10.1016/j.jgar.2018.02.013 may also have surface activity.

留言 (0)

沒有登入
gif