Self-Defensive Antimicrobial Biomaterial Surfaces

Abstract

Self-defensive biomaterial surfaces are being developed in order to mitigate infection associated with tissue-contacting biomedical devices. Such infection occurs when microbes colonize the surface of a device and proliferate into a recalcitrant biofilm. A key intervention point centers on preventing the initial colonization. Incorporating antimicrobials within a surface coating can be very effective, but the traditional means of antimicrobial delivery by continuous elution can often be counterproductive. If there is no infection, continuous elution creates conditions that promote the development of resistant microbes throughout the patient. In contrast, a self-defensive coating releases antimicrobial only when and only where there is a microbial challenge to the surface. Otherwise, the antimicrobial remains sequestered within the coating and does not contribute to the development of resistance. A self-defensive surface requires a local trigger that signals the microbial challenge. Three such triggers have been identified as: (1) local pH lowering; (2) local enzyme release; and (3) direct microbial-surface contact. This short review highlights the need for self-defensive surfaces in the general context of the device-infection problem and then reviews key biomaterials developments associated with each of these three triggering mechanisms.

Graphical abstract

1. Context

Biomaterials-associated infection, often referred to simply as implant infection, is now recognized as one of the principal failure mechanisms of tissue-contacting biomedical devices [1]. All devices - e.g. hip/knee prostheses [2–4]; heart valves [5]; pacemakers [6]; cochlear implants [7]; shunts [8]; surgical mesh [9]; sutures [10]; as well as next-generation tissue-engineering constructs [11]; among many others - are susceptible to this problem. Implant infections are a consequence of microbes first colonizing a foreign device surface and then continuing to develop into a biofilm [12, 13]. Because bacteria growing as biofilms are highly tolerant to antimicrobials [12, 14–18], these infections can typically be eliminated only by removing the device, followed by resolving the infection, and possibly re-implanting a new device as part of one or more subsequent revision surgeries [19, 20]. The consequences are substantial. In the case of an infected joint prosthesis, for example, revision can take as long as a year with the patient in a compromised condition throughout, and the cost is amplified by a factor of ten or more. Significantly, the five-year mortality rate associated with the infection of periprosthetic joint replacement currently exceeds that associated with breast cancer [21 ]. The cost of health-care associated infections, the majority of which are in one way or another related to a tissue-contacting biomedical device - has been estimated to be in the range of tens of billions of dollars ($US) annually [22–24].

Avoiding the initial colonization of a device is a key intervention point where biomaterials-associated infection can be prevented [25, 26]. While many different strategies have been explored, including ones that alter surface topography and antifouling properties [27–29], some of the most successful ones incorporate antimicrobials such that an antimicrobial elutes from the device over time [30, 31]. Gentamycin-eluting bone cement and beads, approved by the United States Food and Drug Administration (FDA) for several decades, are prime examples [32–34]. However, while often effective in vitro and in small-animal models, few other antimicrobial-based approaches have achieved clinical impact, largely because of concerns about systems that couple a continuously eluting antimicrobial with an implanted medical device [35, 36]. In many clinical cases, infection is not an issue, and unnecessary antimicrobial exposure from an eluting device promotes the development of antimicrobial resistance by bacteria, which at some later date can enable an untreatable infection. This is a very real concern, and the increasing awareness of a potentially devastating post-antibiotic era is driving changes in antimicrobial use and strategy in medical devices and beyond [37–40].

2. Self-Defensive Surfaces

Self-defensive biomaterial surfaces offer a compelling solution. In the context of infection resistance, the term self-defensive was introduced in 2013 by Boulmedais et al. [41, 42]. It refers to a surface that responds to some form of challenge mounted by microbes such that the surface response inhibits colonization and, consequently, reduces the threat of infection. Most research thus far has concentrated on bacterial challenges, but many of the concepts apply equally well to fungal infections.

Research on self-defensive surfaces and coatings has accelerated in the short time since the idea was conceived. The original approaches concentrated on coatings that responsively release antimicrobials, but the concept has since been expanded to include a range of surfaces which responsively interact with bacteria based on changes in local hydrophobicity and topography. A variety of phrases have been used to describe self-defensive surfaces including: smart; bacteria responsive; on demand; triggered; intelligent; and stimuli responsive; among others. Many of these advances and various approaches have recently been detailed in two different reviews [43, 44]. This review concentrates specifically on coatings that responsively release antimicrobials and describes them by the phrase “self-defensive.”

Traditional antimicrobial delivery is based on elution (Fig. 1 dashed line), where antimicrobial is continuously released with a profile that can vary over time to produce local concentrations that can both dramatically exceed and later dramatically fall below the minimum inhibitory concentration (MIC) [45]. Notably, within an elutive mechanism, antimicrobial is delivered whether it is needed or not. However, since the infection rates associated with many, though not all, devices tend to fall within the range of about 0.1% to 5%, the release of antimicrobial is, more often than not, unnecessary. Under such a condition, antimicrobial release simply promotes the development of resistance within the patient. In contrast to elutive release, self-defensive release happens only when and where a challenge occurs (Fig. 1 solid line). In areas where there is no bacterial challenge, the antimicrobial remains sequestered within the coating. Furthermore, because of the microscale dimensions of individual bacteria and bacterial biofilms relative to the overall device, the vast majority of a device surface is unaffected and most of the antimicrobial remains sequestered (Fig. 2). A self-defensive approach thus offers a means with which to kill bacteria using antimicrobials but does so in a manner that minimizes antimicrobial release consistent with good stewardship.

Fig. 1:

An elutive material (dashed) continuously releases antimicrobial whether it is needed or not. A self-defensive coating (solid) releases antimicrobial only when there is a microbial challenge.

Fig. 2:

A self-defensive coating responds to a bacterial challenge by locally releasing antimicrobial. Antimicrobial elsewhere in the coating remains sequestered (light blue).

Three general mechanisms have thus far been identified for triggering antimicrobial release in self-defensive coatings. The first was pioneered by Sukhishvili et al. [46, 47]. They recognized that the local pH would be lowered in the presence of metabolizing bacteria as they secrete acidic metabolic products [48, 49] and designed pH-responsive polyelectrolyte hydrogel coatings to exploit this effect. The second was introduced by Cado, Boulmedais et al. [41]. Boulmedais and her colleagues took advantage of the fact that metabolizing microbes secrete specific enzymes. Coatings can be designed that incorporate substrates for those particular enzymes so that responsive coating degradation will release antimicrobial. Most recently, Liang, Libera, et al. [50] have shown that bacterial contact alone, in the absence of bacterial metabolism and the associated diffusion fields, is sufficient to drive responsive antimicrobial release from polyelectrolyte hydrogel coatings because of the high concentration of negative charge and hydrophobicity associated with the microbial envelope. The remainder of this review discusses each of these self-defensive approaches in greater detail.

3. Self-defensive surfaces based on local pH changes

Bacterial metabolism leads to the secretion of a broad array of substances meant to control the local environment. Control of the local pH is among these functions. More generally, there are bacteria that flourish most effectively in acidic environments and thus secrete acidic products to lower the pH while others prefer basic environments and thus secrete alkaline products [51, 52]. Acid secretion and its consequences are well established in the context of dental biofilms that cause caries (i.e. cavities) [53, 54] and in structural applications where microbial-induced corrosion can be an application-limiting failure mechanism [55]. The bacteria most often implicated in device-associated infection - staphylococci, pseudomonads, and enterococci, among others [56] - fall in this same category of acid producers. They typically secrete monocarboxylic acids such as lactate, acetate, and pyruvate, among others [48, 49, 57–62]. In the context of changes to the surrounding culture medium, both Handke et al. [63] and Pavlukhina et al. [46], for example, have experimentally shown that the average pH of TSB-based growth medium decreases to about 4.5 during S. epidermidis proliferation in batch cultures.

A self-defensive approach to antimicrobial release requires a local trigger, and Albright, Sukhishvili, et al. [64] have recently shown that the pH surrounding small and metabolizing bacterial colonies adhered to a surface locally reduce the pH. While acidification within bacterial biofilms has previously been demonstrated and continues to both be studied and invoked to justify some of the remarkable properties of bacteria in the biofilm state [65–69], these recent experiments address the local environment outside of the nascent bacterial colony. Figure 3 presents one of their results where S. aureus was cultured on a continuous hydrogel thin film (~50 nm thick) of poly(methacrylic acid) (PMAA) created via layer-by-layer (LbL) deposition. A pH-sensitive fluorophore was covalently attached post deposition. After calibrating the pH dependence of the fluorescent intensity and removing the contribution of autofluorescence from the bacteria themselves, these imaging experiments showed a local pH decrease from 7.4 to about 5.5 within the underlying PMAA film. This decrease was sufficient to trigger the release of either gentamycin or polymyxin B and self-defensively inhibit S. aureus and E. coli colonization, respectively.

Fig. 3:

A hydrogel thin-film coating conjugated with pH-sensitive dye indicates local acidification around S. aureus as a consequence of bacterial metabolism. See Albright el al. for details (64).

Since the original conception of the pH-triggered self-defensive approach [46, 47], there have been a number of groups pursuing different strategies based on this same mechanism. Lee, Composto, et al. [70], for example, designed a bilayer with a tobramycin-loaded poly(acrylic acid) (PAA) inner layer and a chitosan brush outer layer. Biofilm-induced acidification was able to swell the outer chitosan layer and trigger the release of tobramycin from the inner PAA layer. A somewhat similar approach has been used by Wang, Chen, et al. [71] who made LbL multilayers between triclosan-loaded PEG-PCl micelles and PAA. Zhou et al. [72] created an analogous bilayered structure but with a pH-responsive polycationic outer layer and an inner layer of silk fibroin containing both silver nanoparticles and electrostatically complexed gentamycin. Extending their own work in this area, Pavlukhina, Sukhishvili, et al. [73] showed that multilayers of exfoliated montmorillonite and PAA can be loaded with gentamycin and, importantly, that gentamycin can be sequestered under physiological conditions for periods exceeding 45 days yet still resist colonization by various gram-positive and gram-negative bacteria. And, in contrast to the pH-lowering mechanisms, Irwin, McCoy, et al. [74, 75] have devised a system that enables the triggered release of nalidixic acid from poly(HEMA)-based gels when the pH is increased.

4. Self-defensive surfaces based on local enzyme secretion

In addition to influencing the local pH, microbes produce and secrete an array of biomolecules including various enzymes. Staphylococcal bacteria, in particular, secrete a variety of polysaccharide lyases that can degrade components of soft-tissue extracellular matrix (ECM) such as chondroitin sulfate, heparin, heparan sulfate, and hyularonic acid (HA), among a number of others [76–78]. In earlier years, these enzymes were often referred to as spreading factors, because local ECM degradation provides both energy and space for a growing microbial colony to spread [79]. One of the secreted enzymes is hyaluronidase, which can degrade HA [80]. Hyaluronidase refers to a family of enzymes secreted by many microbes. Microbial hyaluronidases cleave the 1-4 glycosidic linkages in hA to ultimately produce small-molecule disaccharides [81, 82]. Fungi, such as C. albicans, are also known to secrete hyaluronidases [1,83, 84].

Boulmedais et al. [41] took advantage of the release of ECM-degradative enzymes as a self-defensive trigger by designing LbL thin-film coatings that release antimicrobial in response to local enzymatic degradation of ECM components (Fig. 4). They constructed multilayer thin films consisting of polycationic chitosan (CHI) and polyanionic hyaluronan (HA). A cysteine-terminated 14-residue (RSMRLSFRARGYGFR), cationic antimicrobial peptide derived from bovine cateslytin, referred to as CTL, was covalently grafted to the HA (HA-CTL). Grafting was done using a maleimide chemistry by linking the terminal thiol of CTL to a carboxyl group attached to the HA. Furthermore, CTL is active against both bacteria and fungi [85, 86], and Boulmedais et al. showed that these films substantially inhibited the surface colonization by two-different gram-positive bacteria (M. luteus and S. aureus) and one fungus (C. albicans).

Fig. 4:

(A) A multilayered LbL thin film comprising alternating layers of peptide (CTL)-modified hyaluronan and chitosan is challenged by live (green) microbes. (B) Metabolically secreted enzymes (hyaluronidases) locally degrade the layer-by-layer (LbL) film. (C) Released antimicrobial locally kills the microbes (red). The honeycomb structure in (D) shows the local degradation of a FITC-labeled LbL film by C. albicans where the dark circular areas correspond to missing thin film due to enzymatic degradation. After Cado, Boulmedais et al. [41].

Significantly, Boulmedais et al. show convincing data consistent with enzymatic film degradation. One variation of the experiment used FITC-labeled CTL-modified HA. Using confocal fluorescence imaging (Fig. 4D) they observed a honeycomb structure defined by the positions of individual C. albicans cells, a morphology they attributed to local enzymatic degradation of the film. Such degradation was even more extensive in surfaces exposed to S. aureus or to M. luteus. Hyaluronidase action ultimately destabilizes the LbL thin film as HA is degraded and the molecular weight decreases. Microbial HA degradation coincided with the uptake of FITC-labeled HA degradation products by the microbes, which was manifested by an increase in fluorescent signal from within the microbial cytoplasm. Notably, this enzymatic degradation of HA-CTL does not release pure CTL but rather CTL covalently linked to an HA fragment. Boulmedais et al. [41] separately showed that CTL remains active when covalently linked to homopolymer HA.

Enzyme-responsive self-defensive surfaces continue to be developed. Wang, Chen, et al. [87] created LbL thin films comprising alternating layers of montmorillonite and hyaluronan plus gentamycin sulfate and showed that these degrade in response to hyaluronidase exposure. Work from the same group [88] demonstrated similar behavior using a two-layer composite structure of degradable HA/poly(L-lysine) multilayers over HA/chitosan multilayers. Yuan et al. [89] created enzymatically sensitive multilayers of modified HA and modified chitosan over a surface of nanoporous TiO2 that released vancomycin contained within TiO2 nanopores upon enzymatic film degradation. Similarly, Wang, Fu, et al. [90] created a nanovalve structure able to release either or both of two antimicrobials - cinnamaldehyde and ampicillin - contained within surface-patterned nanopores. The nanovalves were responsive to both pH decreases and enzymatic degradation. Francesko et al. [91] made multilayers of antimicrobial (highly cationic) aminocellulose nanoparticles and HA on silicone to mimic a urinary catheter and showed that these were effective at inhibiting colonization via P. aeruginosa.

5. Self-defensive surfaces based on direct microbial-surface contact

The fact that contact with bacteria can create a self-defensive response is a consequence of the high levels of negative charge and hydrophobicity in the cell envelope [92–94]. Indeed, these characteristics of the bacterial envelope are responsible for much of the uptake of antimicrobials delivered systemically from solution [95–97]. Negative charge is present because of components such as teichoic acids, lipoteichoic acids, and lipopolysaccharides, which comprise a significant fraction of the outer membrane(s) or cell wall of common Gram-negative and Gram-positive bacteria [95, 96]. Teichoic acid, for example, can comprise as much as 60 wt% of the wall in Gram-positive bacteria [98]. This high concentration of negative charge is in part manifested by zeta-potential measurements, which find values ranging from about −10 to −40 mV, depending on the bacterium and the medium [99–102]. In short, the nature of the bacterial cell envelope is such that the thermodynamic driving force for electrostatic complexation with a variety of cationic antimicrobials is large and can drive the uptake of many cationic compounds.

Liang et al. [50] have recently shown that contact with a bacterium can enable the transfer of cationic antimicrobials complexed within an anionic microgel surface coating to the bacterium. Such contact transfer is a consequence of the fact that the equilibrium constant associated with antimicrobial complexation within the bacterial envelope is greater than that associated with complexation within the anionic microgel. Therefore, when a bacterium comes within sufficient proximity to the loaded microgel, a threshold gradient in antimicrobial chemical potential drives antimicrobial transfer to the bacterium.

Liang et al. [50] worked with poly(acrylic acid) (PAA) microgels. At physiological pH these microgels are negatively charged and can be electrostatically deposited onto a surface. The deposited microgels were loaded by electrostatic complexation [50, 103–105] with colistin or with an antimicrobial peptide - either L5 or Sub5. Because of the lysine and/or arginine residues they contain, these three antimicrobials have electrostatic charges of +5, +6, and +7, respectively, at physiological pH. Specific conditions were identified where the complexation strength was sufficient to sequester the antimicrobials within the microgels for weeks or more, despite the fact that a traditional elutive process would have depleted the microgel reservoir over a time scale of minutes or less because of the small distances associated with diffusion out of a microgel.

Contact transfer was demonstrated using a nanomanipulator to contact loaded microgels with bacteria (Fig. 5 schematic). Microgels were electrostatically deposited on the cleaved end of a glass fiber and then loaded with antimicrobial. The fiber tip was lowered through buffer to contact bacteria - either E. coli or S. epidermidis - inoculated on the surface of a glass substrate. The buffer contained no nutrients, so the metabolic processes associated with local acidification and enzyme secretion were suppressed. The medium did, however, contain live/dead stain to visual bacterial killing via fluorescence imaging. Fig. 5A shows live E. coli (green) on the glass surface prior to contact, and Fig. 5B shows the same region after 30 minutes of contact with colistin-loaded microgels. Local killing is manifested by the appearance of red (dead) bacteria, and this occurs only in the area under the fiber tip where there is microgel-bacteria contact.

Fig. 5:

Contact transfer of colistin kills E. coli. Colistin-loaded poly(acrylic acid) microgels on the end of a cleaved glass fiber were brought into contact with E. coli deposited on the surface of a glass substrate. (A) No contact; (B) after 30 min of contact. The medium was 0.01 M phosphate buffer pH 7.4 containing live (green)/dead (red) stain. Images C1-C3 show a time-resolved sequence (1 min intervals) of an individual E. coli cell (short arrow) responding to contact with a colistin-loaded microgel (longer angled arrow). See Liang et al. for details [50].

The fact that the contact-transfer approach does not involve metabolism differentiates it from the pH-based and enzymatic-based mechanisms. For example, exposure to microbes can occur under non-metabolic conditions, such as those associated with a surgery. Simple contact from the initial adhesion event promotes microbial killing even before adherent microbes have an opportunity to proliferate within the nutrient-rich in vivo environment. Importantly, however, a coating material such as PAA with complexed cationic antimicrobial can release by contact as well as by local pH triggering, and such redundancy can afford an even higher level of surface protection against colonization by potentially interfering with both the initial adhesion events as well as, if necessary, subsequent metabolic growth.

6. Concluding discussion

The current generation of biomaterials has to a great extent been defined by a major paradigm shift from the pre-1990’s perspective of creating biologically inert synthetic surfaces to the current perspective of creating surfaces that controllably interact with host tissue. From this shift, a new international community has emerged that marries skills from cell and molecular biology with those from physical sciences and engineering to an extent almost unthinkable twenty-five years ago. The biomaterials community is now following a similar paradigm shift where the next-generation of biomaterials will be intentionally designed to resist microbial colonization by extending the scientific synergy to include cellular and molecular microbiology.

Within this emerging umbrella of infection-resisting biomaterials, self-defensive surfaces provide a new means with which to inhibit bacterial colonization using antimicrobial-based strategies in a manner that preserves good antimicrobial stewardship. This review has highlighted three specific mechanisms - local pH change; local enzyme secretion; and direct microbial contact - that are being exploited by self-defensive coating designs. These coatings have all shown an ability to resist the colonization of one or more important microbes by locally and responsively releasing antimicrobial. Furthermore, while not emphasized within this focused review, the majority of research on self-defensive surfaces also demonstrates, either by in vitro or by in vivo experiments, an ability of these coatings to still sustain excellent degrees of cytocompatibility. Some of the scientific challenges facing this field going forward include identifying broader combinations of coating material and antimicrobial to both address a wider range of possible infections and facilitate approval through regulatory processes as well as determine the fate of antimicrobial sequestered for long periods in the absence of a microbial challenge. Nevertheless, self-defensive biomaterials hold substantial promise for controlling both tissue-cell/material interactions as well as microbial-cell/material interactions in a way that can simultaneously promote healing and resist infection.

Highlights.

• Self-defensive surfaces release antimicrobial when challenged by bacteria

• Triggers include pH changes and enzyme secretion due to bacterial metabolism

• Contact between a bacterium and a self-defensive surface can trigger release