Some reports have described the anti-bacterial effects of DLC-based hybrid or metal atom-doped coating against S. aureus [11,12,13, 24]. However, as far as we searched, there are only two studies describing the anti-bacterial effects of DLC by itself against S. aureus. Del Prado et al. showed that DLC coating decreased S. aureus attachment on the surface of ultra-high molecular weight polyethylene (UHMWPE) [8]. Levon and Myllymaa et al. showed that DLC coating inhibited S. aureus colony and biofilm formation on silicon surfaces [9, 10]. However, no studies have yet reported the effects of DLC coating of polyurethane surfaces in preventing S. aureus adhesion and biofilm formation. In this study, DLC coating was proven to suppress the adhesion and biofilm formation of S. aureus on polyurethane surface as demonstrated with UHMWPE and silicon.

Previously, we reported that DLC coating made by the same method as in this study inhibited the adhesion of Pseudomonas aeruginosa and Escherichia coli, as well as the biofilm formation of P. aeruginosa on the inner surface of the silicon tube under flow conditions with artificial urine. However, these effects could not be obtained against S. aureus [14]. This is probably because artificial urine was used as the culture medium and the substrate also differed from this study. The differences of flow condition and length of incubation time also might affect the outcomes. This kind of variability in results is seen in studies investigating the antimicrobial properties of DLC against Candida albicans, whether it is polyurethane-based or silicone-based [25, 26]. Establishing protocols to evaluate anti-infective properties according to use will be important for accelerating the development of antimicrobial coatings for implantable medical devices.

This study showed decreased adhesion and biofilm formation of S. aureus using DLC-coated polyurethane tubes, compared with uncoated polyurethane tubes. These effects were observed under both static and perfused conditions, to replicate exposure to the human bloodstream. These kinds of antibacterial tests are usually performed under static conditions; however, in the static condition, there is a difference in bacterial concentration between the deep and surface layers, and the air may affect the bacteria activity and the ability of bacteria to attach to the testing material surface [19]. In addition, solution flow may affect bacterial adhesion under the perfused conditions [24, 27]. As we envisioned that this coating will be in contact with the bloodstream, we added the experiment assessing flow conditions. Subsequently, the results of the flow studies reinforced the validity of the results under static conditions.

The physical properties of polyurethane surface were significantly changed by DLC coating. The DLC-coated polyurethane surface had a significantly decreased water contact angle, smoother surface, and a lower zeta-potential than the uncoated surface. The decreased water contact angle from 90.0° to 74.5° indicates increased hydrophilicity and surface free energy. Generally, a surface with moderate hydrophilicity increases bacterial attachment [28, 29]. Yuan et al. described that a water contact angle of 90° was the least optimal for bacterial attachment prevention [30]. The increased hydrophilicity on DLC-coated polyurethane tube reduced the S. aureus adhesion. In addition, DLC-coated polyurethane surface was smoother than that of the uncoated surface, as determined by calculating the PV, Sq, and Sa values. Increased smoothness is associated with low friction [17], whereas a larger surface area provides more area for bacterial adherence, eventually supporting bacterial growth [31,32,33,34]. Because of this, it is possible that the increased surface smoothness suppresses the adhesion of S. aureus. Furthermore, the zeta-potential of DLC-coated polyurethane was lower than uncoated polyurethane. In general, for most bacteria, the net surface charge is negative and balanced by oppositely charged counter-ions present in the surrounding media [35]. In fact, the average zeta-potential of the untreated S. aureus was found to be − 35.6 mV [36]. DLC-coated polyurethane may prevent bacterial adhesion because of the repelling effect of the more negative zeta-potential on the negatively charged S. aureus. Although the individual effects may not be conclusive, it is likely that the combination of these beneficial properties of DLC-coated surfaces facilitate the inhibition of S. aureus adhesion and biofilm formation.

Clinically, polyurethane is often used for catheters and vascular grafts. Polyurethane and ePTFE are both the main biomaterials as vascular grafts for hemodialysis; therefore, they have sometimes compared the features in this function. Polyurethane vascular grafts are reportedly more vulnerable to bacterial infection than ePTFE artificial blood vessels [2.3]. Thus, the usefulness of polyurethane for catheter and vascular prostheses would increase with improved antimicrobial properties. As S. aureus is one of the most prominent bacterial species in catheters or artificial vascular graft-related infections [7], DLC coating could prove beneficial. In addition, DLC can be used in metallic parts of blood contact devices [37, 38]; therefore, the DLC for resin tubes is also expected to be bio-compatible enough to be used in artificial blood vessels and catheters. We are currently conducting a demonstration experiment on this topic.

Current thought is that the effects of DLC coating will change depending on the properties and steric structure of the substrate. Thus, detailed investigations according to use are indicated for each combination of material and bacteria to confirm the effects of DLC. Further study will be necessary to determine if these results have clinical significance. However, as DLC itself is non-toxic and hypoallergenic for humans, DLC may prove superior in terms of safety for developing new coatings to enhance the clinical performance of biomaterials.

This study has six main limitations. First, the number of polyurethane tubes in each group was small. However, the results of this experiment were largely determined by the engineering conditions under which the DLC was made which can be strictly controlled. In addition, we used the paired comparisons for each experiment and obtained highly reproducible results. Second, as the concentrations of prepared bacterial solutions were not measured, they may have differed between experiments. To minimize this limitation, we used the same bacterial solution for each paired comparison of DLC-coated and uncoated; therefore, this limitation is not anticipated to affect the results substantially. Third, we performed in vitro studies only; S. aureus attachment or biofilm formation may behave differently in vivo. Future in vivo studies are warranted before commercialization of DLC coatings. Fourth, the perfusion studies used steady flow rather than pulsatile flow. Bacterial attachment and biofilm formation in pulsatile flow conditions might differ from those in continuous-flow conditions. Fifth, under flow conditions, the outer surface of the tube not coated with DLC is in contact with the bacterial solution, so the difference in absorbance values for biofilm measurements should be underestimated. However, this limitation does not affect the outcome. Finally, the DLC-coated sheet was used to verify the water contact angle, smoothness, and zeta-potential because the flat surface was needed for these examinations. As the DLC coating method for the polyurethane sheets was the same as that of DLC-coated tubes, the influence of this limitation on the results should be minimal because if the structure of DLC is the same, even if the steric structure of the substrate changes, it is expected that the direction of change will be the same.