In Vitro and In Vivo Assessment of the Infection Resistance and Biocompatibility of Small-Molecule-Modified Polyurethane Biomaterials

Abstract

Bacterial intracellular nucleotide second messenger signaling is involved in biofilm formation and regulates biofilm development. Interference with the bacterial nucleotide second messenger signaling provides a novel approach to control biofilm formation and limit microbial infection in medical devices. In this study, we tethered small-molecule derivatives of 4-arylazo-3,5-diamino-1H-pyrazole on polyurethane biomaterial surfaces and measured the biofilm resistance and initial biocompatibility of modified biomaterials in in vitro and in vivo settings. Results showed that small-molecule-modified surfaces significantly reduced the Staphylococcal epidermidis biofilm formation compared to unmodified surfaces and decreased the nucleotide levels of c-di-AMP in biofilm cells, suggesting that the tethered small molecules interfere with intracellular nucleotide signaling and inhibit biofilm formation. The hemocompatibility assay showed that the modified polyurethane films did not induce platelet activation or red blood cell hemolysis but significantly reduced plasma coagulation and platelet adhesion. The cytocompatibility assay with fibroblast cells showed that small-molecule-modified surfaces were noncytotoxic and cells appeared to be proliferating and growing on modified surfaces. In a 7-day subcutaneous infection rat model, the polymer samples were implanted in Wistar rats and inoculated with bacteria or PBS. Results show that modified polyurethane significantly reduced bacteria by ~2.5 log units over unmodified films, and the modified polymers did not lead to additional irritation/toxicity to the animal tissues. Taken together, the results demonstrated that small molecules tethered on polymer surfaces remain active, and the modified polymers are biocompatible and resistant to microbial infection in vitro and in vivo.

Graphical Abstract

1. INTRODUCTION

With the rapid increase of applications of biomedical implants, such as catheters, prosthetics, valves, and ventricular assist devices in modern medicine, the risk of infection caused by these devices is gaining more attention.^1-3 Biomaterial-associated microbial infections are one of the main complications for long-term implantable medical devices. Device-centered infections are caused by pathogenic bacteria in the form of biofilms.^4,5 It is difficult to treat such infections due to the formation of the biofilm, an extracellular matrix that shields the bacteria from both natural immune responses and antibiotic therapy. Surgical removal and reimplantation of these infected medical devices thereby becomes the only treatment, leading to increased patient morbidity and mortality and placing a significant financial cost on healthcare in the USA.^6,7 The development of strategies to eliminate bio-material-associated biofilms and to make those that do form treatable using standard antibiotic therapy is a high research priority in combating device-associated infection.⁸

The development of biofilms on biomaterial surfaces initiates with cellular attachment, followed by the formation of microcolonies, biofilm maturation, and finally dispersion.⁹ After bacterial cells have attached onto a surface, they start to replicate and secrete extracellular polymeric substances (EPS), contributing to the formation of a biofilm matrix; a “shelter” where all of the adherent cells are living. Although many factors have been found to influence bacterial adhesion and EPS production,¹⁰ the molecular mechanism of biofilm formation appears to involve bacterial intracellular nucleotide second messenger signaling, which allows the bacteria to monitor and respond to changing environments.^11-18 Many bacterial behaviors, including motility, biofilm formation, and virulence factor production are controlled by nucleotide second messenger signaling molecules such as cyclic dimeric guanosine monophosphate (c-di-GMP), cyclic dimeric adenosine monophosphate (c-di-AMP), cyclic adenosine monophosphate (cAMP), and cyclic guanosine monophosphate (cGMP).^19,20 Interference with these nucleotide signaling molecules and therefore regulation of bacterial behaviors provides a novel approach to control pathogenic biofilm formation on biomaterial surfaces.^21-24 Different from traditional antibiotic therapies, most virulence traits during nucleotide regulations are not essential for bacterial survival, and therefore, this approach will decrease the opportunities to develop resistance.

C-di-GMP is a key bacterial intracellular signaling molecule in many bacterial species that regulates the transition between production of EPS and formation of biofilms or a planktonic lifestyle.^25-30 High levels of c-di-GMP in bacterial cells generally promote the production of EPS, leading to biofilm formation while low levels of c-di-GMP increase bacterial cell motility and disperse the established biofilms.³¹ In the cell, the levels of c-di-GMP are controlled by extracellular cues that regulate c-di-GMP synthesis by diguanylate cyclases (DGCs) and/or hydrolysis by phosphodiesterases (PDEs).^16,32 Small molecules that are DGC inhibitors or PDE activators have been developed in an effort to reduce c-di-GMP levels and interrupt biofilm formation as a means to combating infections.^21,33,34 However, these molecules act as drugs to combat ongoing infections. Short half-lives, systemic toxicity, and increased susceptibility to bacterial resistance are potential issues for these drugs against biofilms.

Small-molecule derivatives of 4-arylazo-3,5-diamino-1H-pyrazole (designated as SP02 and SP03 in this study, Scheme 1a,b) were first reported to be PDE activators in Pseudomonas aeruginosa cells and to decrease the c-di-GMP levels, resulting in biofilm formation inhibition or biofilm dispersal.^35,36 In a previous study, we synthesized these molecules and found that these substituted pyrazole molecules significantly reduced the levels of c-di-AMP in Staphylococcus epidermidis cells and inhibited biofilm formation,³⁷ suggesting that small molecules SP02 and SP03 regulated the c-di-AMP synthesis and influenced the S. epidermidis biofilm formation. This finding demonstrated that the bacterial intracellular second messenger nucleotide signaling is not only limited to the c-di-GMP signaling pathway in regulating bacterial behaviors, but bacterial species can also use other second messengers, including cAMP, cGMP, and c-di-AMP, in the transduction of signals to regulate the motility and biofilms.^38-42 Further, we covalently bound SP02 or SP03 on polyurethane (PU) biomaterial surfaces using hexamethylene diisocyanate (HMDI) as a linker (Scheme 1c) and demonstrated that the modified surfaces inhibited the S. epidermidis biofilm formation and increased the efficacy of antibiotics.³⁷ Different from the approach of using small molecules as drugs, this approach eliminates the issues of general use of systemic drugs, noted above, by limiting the diffusion of the small molecules throughout the body, decreasing the general selective pressure, reducing potential off-site targeting and activity, and keeping the molecules from being degraded in the liver. Thus, this tethering strategy provided a novel approach to interrupt biofilm development and increase the efficacy of antibiotics for combating microbial infections by interfering with nucleotide second messenger signaling.

Scheme 1.

Chemical Structures of (a) SP02, (b) SP03, and (c) Polyurethane Surface Tethered with SP02 or SP03 Using HMDI as a Linker³⁷

In this study, we further explored the nucleotide second messenger signaling in S. epidermidis biofilm cells formed on modified surfaces and tested the hemocompatibility and cytocompatibility of modified PU biomaterial surfaces in vitro; then, we characterized the antibacterial properties and biocompatibility of small-molecule-modified biomaterials in a 7-day subcutaneous infection rat model. The work demonstrated the feasibility of this novel technique for modification of biomaterials as suitable clinical implants.

2. MATERIALS AND METHODS

2.1. Synthesis of Small Molecules SP02 and SP03, and Covalently Bonding SP02 and SP03 on PU Surfaces.

Small molecules SP02 and SP03 were synthesized through a two-step procedure that includes diazotization of arylamines and condensation with malonodinitrile, followed by cyclocondensation with hydrazine. The detailed method was described in the literature.^35,36 A thermoplastic silicone polycarbonate polyurethane (Carbosil 20 80A, DSM Biomedical Inc., Exton, PA) consisting of silicone and hydroxyl-terminated polycarbonate as the soft segment and methylene bis-phenyl isocyanate as hard segments⁴³ was used as the substrate for chemically binding small molecules SP02 and SP03. First, PU films were fabricated by spin-casting PU solution (~18% (w/v) in dimethylacetamide, DMAC, BDH chemicals) on a smooth poly-(dimethylsiloxane) (PDMS) mold with multiple layers to reach the thickness of ~300 μm with a curing step in a vacuum oven overnight between layers. A two-step procedure was then used to tether small molecules SP02 or SP03 on PU surfaces, as described in our previous publication.³⁷ Briefly, PU films were first grafted with hexamethylene diisocyanate (HMDI) in the presence of the catalyst triethylamine and in toluene at 50 °C for 2 h under a nitrogen flow in order to introduce ─NCO groups on PU surfaces. Small molecules SP02 or SP03 were then covalently bound to the PU surfaces through the reaction of the second ─NCO group in HMDI and the ─NH₂ group in the small molecules utilizing toluene at 40 °C for 20 h. After reactions, unreacted SP02 or SP03 were removed by rinsing polymer films in toluene 3 times. The polymer films were dried in air and stored at room temperature prior to use.

A ramé-hart contact angle goniometer (Succasunna, NJ) was used to measure the advancing water contact angles of polymer samples to indicate the surface wettability. The measurement of contact angles was performed at random locations (at least 8 locations) for each type of polymer film using an ~4 μL drop of purified Millipore water (18.2 MΩ) at each location. Results are expressed as the mean ± standard deviation.

2.2. Biofilm Formation and Analysis of Bacterial Nucleotide Second Messenger Signaling in Biofilm Cells.

All bacterial culture medium, PBS (phosphate-buffered saline, 0.01M, pH 7.4, Sigma) and glass containers were autoclaved at 121 °C for 20 min prior to use. Bacterial strain S. epidermidis RP62A (ATCC 35984) was used to study the biofilm formation on small-molecule-modified PU surfaces and nucleotide signaling in biofilm cells. S. epidermidis RP62A was cultured in tryptic soy broth (TSB, Bacto, BD) in a shaker at 250 rpm and 37 °C for ~20 h and harvested by centrifugation at 1500g for 10 min. The pellet was resuspended in PBS and diluted in TSB or 50%TSB + 50% human platelet-poor plasma (PPP, see Section 2.3) to an OD₆₀₀ of 0.01, corresponding to a bacteria concentration of ~5.4 × 10⁶ CFU/mL. Biofilm growth was performed in a 12-well plate. Polymer samples with a diameter of 10 mm were sterilized in 70% ethanol for 30 min, then rinsed with PBS, and soaked in PBS in the 12-well plate for 1 h. After the removal of PBS, each well was filled with 3 mL of culture medium containing bacterial suspension and incubated in a shaker at 150 rpm and 37 °C for 24 h. The culture medium was then replaced with 3 mL of sterile PBS three times to remove nonadherent cells. The polymer sample was then transferred into a 15 mL centrifuge tube containing 10 mL of sterile PBS, vortexed for 30 s, and ultrasonicated (35 kHz, 180 W) for 15 min to remove adherent cells and biofilms on polymer film surfaces. The bacterial cells in bacterial solution samples were quantified by the agar plating method and presented by the colony-forming units per surface area (CFU/cm²).

To collect cells for the analysis of nucleotides in biofilms formed on polymer surfaces, polymer samples were incubated in a TSB medium containing the same initial concentration of bacteria at 37 °C for 3 days. Every 24 h, 1 mL of culture medium was replaced with the same volume of fresh culture medium to supplement nutrients. After 3 days, the culture medium was replaced with sterile PBS, and polymer samples were gently rinsed with PBS three times. Polymer film samples with the biofilm were transferred to 15 mL centrifuge tubes containing 2 mL of sterile PBS, vortexed, and ultrasonicated, same as described above, to remove the cells on polymer film surfaces. The bacterial solutions were vortexed again and aliquoted for nucleotide extraction and protein content analysis. The aliquoted bacterial samples were centrifuged, and the pellets were stored at −80 °C prior to analysis. The extraction and analysis of nucleotide signaling molecules by HPLC-MS/MS as well as quantification of bacterial cell proteins were described in our previous publication.³⁷ The nucleotide molecule concentration in biofilm cells is expressed as pmol/mg protein of cells.

2.3. Hemocompatibility Assessment of Small-Molecule-Modified PUs.

Human blood was collected in ACD Vacutainer blood collection tubes (Becton Dickinson, Franklin Lakes, NJ) from healthy volunteers in accordance with Institutional Review Board policy. The blood was centrifuged at 200g for 20 min, and the supernatant was collected as platelet-rich plasma (PRP) for the platelet adhesion assay. A hematology analyzer (Sysmex KX-21N, Japan) was used to measure the platelet concentration. The remaining blood was further centrifuged at 1500g for 20 min to obtain the platelet-poor plasma (PPP) for the plasma coagulation assay.

The plasma coagulation activity induced by biomaterials was assessed as coagulation time (CT), which is defined as the time from activation of the coagulation cascade to the appearance of a visible clot.^44,45 Briefly, polymer films with a diameter of 8 mm were added to 2 mL polystyrene tubes containing 500 μL of human PPP and 400 μL of PBS. 100 μL of 0.1 M CaCl₂ was then added to the samples to start the coagulation process. The polystyrene tubes were capped with parafilm and rotated at 8 rpm on a hematology mixer until a visible clot appeared, and the corresponding time was recorded as CT.

Platelet adhesion on material surfaces was performed in a 12-well plate using PRP with the platelet concentration of 2.5 × 10⁸ platelets/mL. PRP was first recalcified by adding 2 mM CaCl₂. Then, 2 mL of PRP was added to each well on top of the polymer samples (10 mm diameter), and platelets were allowed to adhere at 37 °C for 1 h with shaking at 150 rpm. PRP was then gently replaced with 2 mL of PBS five times to remove unadhered platelets, followed by fixation in 1% paraformaldehyde (PFA) for 1 h. The polymer samples were labeled with APC Mouse Anti-Human CD61 (BD Biosciences) for 1 h to label adherent platelets. Samples were transferred to microscopy slides, and 50 μL of antifading solution (FluoGel with DABCO, Electron Microscopy Science) was added before applying a glass coverslip. Samples were examined under a fluorescence optical microscope (Nikon, Eclipse 80i) using a 40× objective lens. ImageJ program was used to analyze images to count the platelets per unit area.

For assessing platelet activation in blood in contact with polymers, polymer films with a diameter of 10 mm were incubated with 1 mL of human whole blood and rotated at 8 rpm on a hematology mixer at 37 °C for 1 h. Total platelets in blood were labeled with APC Mouse Anti-Human CD61 (BD Biosciences, Cat. No. 564174), and the activated platelets were stained with BV421 Mouse Anti-Human CD62P (BD Biosciences, Cat. No. 564038). Activation of platelets was assessed by flow cytometry (10 color BD FACSCanto, BD Bioscience). The ratio of activated/total platelets is calculated as % activation.

The hemolytic property of small-molecule-modified materials was assessed in accordance with ASTM F756–17.⁴⁶ Briefly, polymer films with the dimension of 1 cm × 2 cm were incubated in 15 mL centrifuge tubes containing 1 mL of human whole blood and 7 mL of PBS for 3 h at 37 °C. The tubes were inverted to mix red blood cells (RBCs) homogeneously in the suspension every 30 min. The complete lysis of RBC by pure H₂O was used to define the 100% lysis point, and the RBC suspension without polymer contact was the negative control (minimal lysis). After 3 h, the blood suspension was centrifuged at 10,000 rpm for 10 min and the free hemoglobin concentration in the supernatant was measured by the Harboe direct spectrophotometric method.^47,48 The hemolytic index is calculated as follows.

$$\begin{gathered} \text{hemolysis}(\%)=\frac{\text{free hemoglobin concentration in supernatant}}{\text{total hemoglobin concentration}} \\ \times 100\% \end{gathered}$$

2.4. Cytocompatibility Assessment of Small-Molecule-Modified PUs.

Primary human dermal fibroblast cells (PCS-201–012, ATCC, Manassas, VA) were used to test the cytocompatibility of biomaterials using a resazurin-based dye method in this study. The surfaces were sterilized using 70% ethanol for 1 h. The ethanol was removed, and the surfaces were washed two times with 1× PBS for 5 min. They were placed in a 96-well plate and incubated for 1 h with Dulbecco’s minimum essential medium (DMEM; vWR, Radnor, PA) supplemented with 10% fetal bovine serum (FBS, Atlanta Biologics, Flowery Branch, GA) and 1% penicillin–streptomycin (vWR, Radnor, PA). Cells were lifted with 500 μL of trypsin for 5 min. Five × 10³ fibroblast cells were added onto each PU biomaterial (disc shape with 5 mm diameter) surface placed in the 96-well plate. After 2 h, 300 μL of media was added, and the cells were cultured at 37 °C for 7 days. On days 1, 3, and 7, cell viabilities were measured by the PrestoBlue Cell Viability Reagent (Thermo Fisher, Carlsbad, CA). The PrestoBlue dye was mixed with cell medium in a 1:10 ratio. The media in the culture dish was removed and replaced with 300 μL of the PrestoBlue solution and incubated for 2 h. The plate was then placed on a plate reader, and based on the increase in fluorescence signal, the cell metabolic activity was measured.⁴⁹ Collected data were normalized against cell-free dye fluorescence intensity and presented as relative fluorescence units (RFU).

2.5. Assessment of Antibacterial Properties and Biocompatibility of Small-Molecule-Modified PUs In Vivo.

A 7-day subcutaneous infection rat model was used to evaluate infection resistance and initial safety of materials based on ref 50, and the related animal procedures were approved by the Pennsylvania State University College of Medicine Institutional Animal Care and Use Committee (IACUC). Briefly, bacterial strain S. epidermidis RP62A was precultured in the TSB medium and harvested by a centrifuge using the same method as described in Section 2.2. The bacterial cells were resuspended in sterile PBS and the concentration was measured by an agar plating count method. Wistar rats were purchased from Envigo (Indianapolis, IN). The health record of rats showed that all tests of viruses, bacteria, mycoplasma, fungi, parasites, and pathological lesions were negative except for the Staphylococcus aureus test within 18 months before delivery. The animals were randomly distributed into five groups (n = 3 for each group): Group 1: sham surgery; Group 2: unmodified PU with PBS inoculation; Group 3: unmodified PU with bacteria inoculation; Group 4: SP03/PU with PBS inoculation; and Group 5: SP03/PU with bacteria inoculation.

An ~3 cm midline incision was made on the back of each Wistar rat, and a pocket was made in the subcutaneous tissue to house a 2 cm × 1 cm polymeric film. A 200 μL dose of 5 × 10⁸ CFU/mL S. epidermidis or sterile PBS was placed on the polymer film surfaces, and the incision was sutured closed. After recovering from anesthesia, the animals were returned to their cages with normal food. To control postprocedural discomfort, 1.2 mg/kg of buprenorphine-extended release (Wedgewood Pharmacy, Swedesboro, New Jersey) was administered subcutaneously. After 7 days, animals were euthanized, and implanted polymer samples with the adhered tissues were removed and homogenized in 10 mL of sterile PBS for bacteria counting by the agar plating method. The adjacent muscle, subcutis, and overlying skin at the implantation site were also sampled and scored by gross and histopathological examination. Multiple parameters for inflammation and wound healing were scored in a routine semiquantitative manner between 0 (within normal limits) and 4 (severe change) by a board-certified veterinary pathologist. Initial scoring was performed blind to group identity, and the means for these scores were later compared between groups.

2.6. Statistical Analysis.

All experiments were performed in multiple replicates, as noted. Results are presented as mean ± standard deviation and analyzed by one-way ANOVA using OriginPro 2020b software. Differences between samples were considered statistically significant if p < 0.05. Significance was denoted with symbols (*), with one symbol denoting p < 0.05, two symbols denoting p < 0.01, and three symbols denoting p < 0.001, NS = no significance.

3. RESULTS AND DISCUSSION

3.1. Biofilm Formation Inhibition and the Cyclic Nucleotides in Biofilm Cells on Small Molecule Covalently Bonded PU Surfaces.

Small molecules SP02 and SP03 were covalently tethered on PU surfaces, and their surface chemical and physical properties have been characterized and reported in our previous publication.³⁷ In this study, we additionally measured the water contact angles of materials to present the surface wettability, and results show that the PU surface is hydrophobic with a water contact angle of 99.8 ± 1.4° and the tethering with SP02 and SP03 on PU surfaces produces no noticeable changes in wettability (Figure S1 in the Supporting Information). This most likely arises from the hydrophobic property of the small molecules that contain fluorine in the chemical structure. This study focuses on the bacterial biofilm resistance and initial indications of the biocompatibility of newly modified surfaces in in vitro and in vivo settings.

In our prior study,³⁷ we measured S. epidermidis biofilms grown on unmodified PU and small-molecule-modified PU (SP02/PU and SP03/PU) in the TSB culture medium for 24 h. These prior results are depicted in Figure 1 and show that SP02/PU and SP03/PU significantly reduced the biofilm cells by ~1 log unit compared to unmodified PU. Our prior study also used fluorescence microscopy to characterize the biofilms on polymer surfaces produced by incubation in a CDC biofilm reactor for 7 days, and results showed that both SP02/PU and SP03/PU dramatically reduced biofilm formation compared to unmodified surfaces.³⁷ To assess the potential effect of the presence of plasma proteins on the activity of small-molecule-functionalized polymer surfaces in the bloodstream, in this study, we added 50% (v/v) human plasma (PPP) to the TSB culture medium containing the same initial concentration of bacteria and cultured for 24 h. Results show that the presence of plasma in the TSB medium significantly reduced the biofilm cells on both unmodified PU and SP03-modified PU (SP03/PU) surfaces by ~0.7 log unit compared to the biofilms formed in the TSB medium without plasma. Similarly, plasma also decreased the CFU counts by ~0.4 log unit on SP02/PU surfaces, although it was not a statistically significant change (Figure 1). These new results with TSB + plasma can be compared to our prior biofilm formation inhibition results in TSB medium alone, suggesting that plasma protein adsorption on surfaces may reduce biofilm formation. This trend is consistent with our previous findings that the presence of plasma significantly reduced both S. epidermidis and S. aureus adhesion on textured surfaces compared to the adhesion in media without plasma.⁵¹ Reduction in bacterial adhesion and biofilm formation in plasma-containing media is believed to be due to the adsorption of albumin on surfaces, which minimizes bacterial adhesion and reduces biofilm formation.^52,53

Figure 1.

Agar plating counts of S. epidermidis cells in biofilms on unmodified PU and small-molecule-modified PU surfaces for 24 h in TSB and 50% TSB + 50% human plasma (n = 4). The agar plating count data for TSB are from our previous publication.³⁷ (**: p<0.01; ***: p<0.001)

Although plasma proteins reduced biofilms on both unmodified and modified PU surfaces, a significant reduction of S. epidermidis biofilm cells (0.7–0.9 log unit) was still observed on both SP02- and SP03-modified surfaces compared to unmodified PU surfaces in plasma-containing media, demonstrating that even in the more complex plasma environment, these small molecules inhibit biofilm growth. It should be noted that there was no significant difference in biofilm cell counts observed between SP02/PU and SP03/PU in either TSB or TSB + plasma, suggesting that the location of the F atom on the benzene ring of 4-arylazo-3,5-diamino-1H-pyrazole derivatives, SP02 and SP03 molecules, does not have a major effect on the inhibition of biofilm formation.

The intracellular nucleotide second messengers, c-di-GMP, c-di-AMP, cGMP, and cAMP, may be involved in the molecular mechanism of biofilm development. These molecules monitor and respond appropriately to changing environments such as regulation of motility and adhesion.³⁸ In a previous study,³⁷ we analyzed these cyclic nucleotide second messengers in planktonic S. epidermidis cells cultured with different doses of small molecules SP02 and SP03 present in solution and found that c-di-AMP played a key role in regulating biofilm formation and that both SP02 and SP03 significantly reduced c-di-AMP levels in S. epidermidis, thereby interrupting biofilm formation or inducing biofilm dispersal. In this study, we analyzed the cyclic nucleotides in biofilm cells, which grew on polymer surfaces for 3 days. The extended growth time is to ensure the reaction of bacterial cells with small molecules tethered on PU surfaces and to collect sufficient numbers of biofilm cells for the extraction of nucleotides. We found that the levels of c-di-AMP in biofilm cells were dramatically higher than the c-di-GMP levels, while the other two nucleotides, cAMP and cGMP, were not detected or were within the noise level (see the representative chromatogram of nucleotides Figure S2 in the Supporting Information). The c-di-AMP level in biofilm cells on unmodified PU surfaces was 109.0 ± 1.9 pmol/mg protein of cells, significantly higher than the c-di-AMP levels in cells on SP02- and SP03-modified surfaces (SP02/PU and SP03/PU), while no significant difference was observed between SP02/PU and SP03/PU samples (Figure 2). The c-di-GMP levels in S. epidermidis biofilm cells were in the range of 1–2 pmol/mg protein of cells, and there were no significant differences among the biofilm cells on unmodified and modified PU samples. These results confirmed the previous conclusion that c-di-AMP played an important role in regulating S. epidermidis biofilm formation,³⁷ and small molecules SP02 and SP03 tethered on PU surfaces remained active. Similar to the molecules applied in solution, SP02 and SP03 influenced the nucleotide c-di-AMP metabolism³⁷ and regulated the levels of c-di-AMP in biofilm cells, resulting in inhibiting biofilm formation (Figure 1). Nucleotide second messenger signaling c-di-AMP is important in the transduction of signals in the regulation of the motility and adhesion of bacteria. It was reported that c-di-AMP mediated the biosynthesis of bacterial exopolysaccharides and regulated the biofilm formation.⁴² The reduction of c-di-AMP levels by SP02 and SP03 molecules was expected to inhibit the biosynthesis of EPS during S. epidermidis biofilm growth and reduce biofilm formation.

Figure 2.

Intracellular nucleotides c-di-AMP and c-di-GMP levels in S. epidermidis biofilm cells on polymer surfaces (n = 4, **: p<0.01; ***: p<0.001).

3.2. Hemocompatibility of SP02- and SP03-Modified PU Surfaces.

Hemocompatibility is another important aspect to consider for the use of biomaterials in blood-contacting medical devices. As a multicomponent process, the hemocompatibility of biomaterials is often assessed by the formation of blood clots and thrombus in contact with blood, which involve the activation of the intrinsic pathway of the plasma coagulation cascade and the platelet-mediated reactions.^54,55 Analysis of plasma coagulation time, platelet adhesion and activation, and the hemolysis of red blood cells (RBC) can directly address the important aspects of the hemocompatibility of biomaterials in contact with blood.

Figure 3a shows the coagulation time (CT) of human plasma incubated with polymer films. The clots in control plasma (blank without polymers) were visualized at ~21.5 min after addition of 10 mM Ca²⁺ to trigger the plasma coagulation cascade, while the clots were formed significantly faster in plasma in the presence of polymer films, as expected, since the foreign materials will induce plasma protein activation, e.g., FXII contact activation, resulting in the production of the proteolytic enzyme FXII_a, which in turn triggers a series of zymogen to enzyme conversions that eventually lead to the generation of thrombin.⁵⁶ Comparing the coagulation times of plasma in contact with polymers, the modified PU materials, SP02/PU and SP03/PU, significantly slow the coagulation of plasma compared to the unmodified polyurethane, suggesting that small-molecule-modified surfaces reduced the plasma coagulation.

Figure 3.

Blood responses to polymers: (a) coagulation time, (b) platelet adhesion, (c) platelet activation, and (d) hemolysis of red blood cells contacted with polymers (n = 4). (***: p<0.001:p < 0.001 compared to all other materials).

Platelet adhesion on polymer surfaces was performed in PRP solution for 1 h, and representative images of platelets are shown in Figure 4. Results showed that more platelets were observed on unmodified PU surfaces, and the platelets were counted at 42 ± 13 per 10⁴ μm² on unmodified PU surfaces, significantly higher than the adhesion on both SP02/PU and SP03/PU surfaces (Figure 3b). The small-molecule-modified surfaces reduced platelet adhesion by ~60% compared with unmodified surfaces.

Figure 4.

Representative fluorescence microscopy images of platelet adhesion on polymer surfaces: (a) unmodified PU, (b) SP02/PU, and (c) SP03/PU. Scale bar = 50 μm.

The platelet activation was measured by flow cytometry after the polymers were incubated with whole blood for 1 h at 37 °C. Results show that the platelet activation levels after contacting with either unmodified or modified polymers were in the range of 1–2%, similar to the levels of platelets in the blank control without the polymer (Figure 3c). There was no significant difference among polymer samples, suggesting that tethering small molecules on surfaces did not induce platelet activation.

For the assessment of the hemolytic properties of materials, polymer films were incubated with human blood for 3 h at 37 °C. The analysis of free hemoglobin shows that the hemolysis of RBCs was low for all samples at ~0.12% (Figure 3d). No significant differences were observed between the blank control (without polymer) and the polymer samples, demonstrating that all materials were nonhemolytic according to ASTM F756–17 and smal-molecule-modified materials did not cause hemolysis. Taken together, we can conclude that SP02- and SP03-tethered PU surfaces are hemocompatible with reduced plasma coagulation and platelet adhesion and without causing platelet activation or hemolysis.

3.3. Cytocompatibility of SP02- and SP03-Modified PU Surfaces.

The cytocompatibility of small-molecule-modified PU biomaterials was measured using human dermal fibroblast cells using a resazurin-based fluorescent dye method with the PrestoBlue Cell Viability Reagent based on the increase in the fluorescence signal. The polymer samples were incubated with cells for 7 days, and the cell viability was measured on days 1, 3, and 7 (Figure 5). Results showed that there was no difference observed in cell metabolic activity on unmodified PU surfaces among different culture periods, suggesting that PU was nontoxic to fibroblast cells, as expected. It is interesting to see that the metabolic activity on SP02- or SP03-modified PU surfaces was significantly higher than that on PU surfaces on days 3 and 7. Furthermore, the metabolic activity significantly increased on SP02- and SP03-modified surfaces on day 3 or 7 compared to day 1. The PrestoBlue method allows the development of live-cell assays for monitoring cell metabolism and viability in real time and is frequently applied for determining cytotoxicity in tissue engineering. Sonnaert et al.⁵⁷ demonstrated a quantitative correlation between the PrestoBlue results and the cell number for human periosteal cells during the expansion phase and found that the metabolic activity assay was correlated well to cell proliferation in a 3D perfusion bioreactor system. Therefore, our metabolic activity data strongly suggested that SP02 and SP03 were not only nontoxic to fibroblast cells, but instead, we can hypothesize that cells were proliferating and growing on SP02/PU and SP03/PU surfaces. The higher metabolic activity on SP02- and SP03-modified surfaces also suggests that SP02 and SP03 molecules are nontoxic to cells.

Figure 5.

Fibroblast cell metabolic activity on PU and modified PU surfaces (n = 3; *: p<0.05; **: p<0.01).

3.4. In Vivo Study of Infection Resistance and Biocompatibility of Small-Molecule-Tethered PU Surfaces.

A 7-day subcutaneous rat model was used to evaluate infection resistance and assess the initial biocompatibility of materials in the more complex biological environment (Figure 6a-d). Both SP02/PU and SP03/PU demonstrated similar properties in inhibition of biofilm formation in TSB and TSB + plasma media (Figure 1), similar hemocompatibility (Figure 3), and similar cytocompatibility (Figure 5), but SP03/PU showed slightly greater reductions in a number of viable cells in the more complex TSB + plasma system. SP03/PU was therefore selected to test the initial biocompatibility of modified biomaterials in in vivo settings and compare against the unmodified PU (control) in this study. In animal experiments, the general appearance and clinical response of animals after surgery show that all animals followed a normal course of postoperative recovery and showed normal eating, drinking, urination, and movements throughout the study and survived until sacrifice. On day 8, animals were euthanized and polymer films with tissue were removed. For the animals inoculated with PBS only, polymer films were easily separated from tissue and removed. All SP03/PU films appeared the same color as in the beginning, and no apparent clotting or infection was observed (Figure 6d). The polymer film samples were soaked in sterile PBS and sonicated to remove the cells. The agar plating count shows that bacterial cells were present on both unmodified PU and SP03/PU surfaces, even when they were inoculated with PBS only. This small number of bacterial cells is believed to come from the animal itself, as the animal’s bacterial test records show a positive test for S. aureus. The agar plating count shows that the cells were detected at 3.8 ± 0.2 log (CFU/g) on unmodified PU surfaces, and a significant bacteria reduction of ~1.5 log units was measured on SP03/PU surfaces compared to PU (Figure 6e). For the animals inoculated with bacteria, all polymer films were found to be integrated with tissue. To count the bacterial cells, the polymer film and tissue were homogenized by cutting and sonication in sterile PBS. The agar plating counts show 6.0 ± 0.9 log (CFU/g) on PU samples, while the bacterial cells were only 3.4 ± 0.5 log (CFU/g) on SP03/PU samples. SP03/PU significantly reduced bacteria cells by ~2.5 log units (>99% reduction) compared to PU films. Results clearly demonstrated that small-molecule-modified films were more resistant to bacterial infection than unmodified PU films in this in vivo setting.

Figure 6.

Infection resistance evaluation by a subcutaneous rat model. (a) Polymer film was inserted in a pocket in the subcutaneous tissue and (b) inoculated with bacteria or PBS, (c) the incision was sutured closed, (d) polymer film was removed after 7 days, and (e) bacterial cells counted by agar plating (n = 3, *: p<0.05).

The tissues at the implantation site were also sampled and scored by gross and histopathological examination (Figures S3 and S4 in the Supporting Information). Scores between 0 (within normal limits) and 4 (severe change) were given for multiple parameters, including inflammatory cells, granulation tissue, and both acute and chronic hemorrhage, and the means for these scores were compared between groups (Figure 7). The most common inflammatory cells seen were low to moderate amounts of lymphocytes and histiocytes, which could be expected at sites of wound healing at this time point for all samples, while the population of neutrophils mildly increased in the groups exposed to bacteria. Granulomatous inflammation (characterized by activated macrophages and multinucleated giant cells) was associated with the embedded foreign material such as suture material and small fragments of hair shafts and was an expected response to breaking down these materials. Granulation tissue comprised small, newly formed blood vessels and proliferating fibroblasts. The amounts were consistent with normal wound healing in response to the surgery. It was interesting that there were no bacteria seen within either the H&E or Gram-stained sections, even in the groups experimentally implanted with bacteria. This most likely arises because bacteria within the tissues were not present at high enough levels to be identified histologically on the tissue slides or due to the rat’s immune systems being able to clear the amount of inoculated bacteria by this time point. However, the bacterial cells were clearly found in the agar plating results (Figure 6e). This suggests that the bacterial cells or biofilms were mostly present at the implanted polymer surfaces, and SP03/PU significantly reduced bacterial infection. Overall, the study showed that the SP03/PU samples did not appear to cause any additional irritation/toxicity to the animal tissues or to impair wound healing, and there was not a significantly increased inflammatory reaction against the experimental polymer. It was concluded that all experimental polymers were similar to the normal polyurethane material in the rat model with regard to tissue response.

Figure 7.

Tissue gross and histopathological examination scores in the subcutaneous rat model after 7 days (n = 3).

4. CONCLUSIONS

This study demonstrated the infection resistance and initial biocompatibility and safety of the polyurethane surfaces modified with small-molecule derivatives of 4-arylazo-3,5-diamino-1H-pyrazole in in vitro and in vivo settings. Small molecules SP02 and SP03 were covalently bound on polyurethane surfaces and showed a significant reduction of S. epidermidis biofilm formation in culture media either with or without plasma. SP02 and SP03 tethered on polyurethane surfaces reduced the nucleotide second messenger c-di-AMP levels in sessile biofilm cells, suggesting that the tethered small molecules remain active and interfere with intracellular nucleotide signaling and inhibit biofilm formation. The in vitro hemocompatibility and cytocompatibility assays demonstrated that the small-molecule-modified surfaces were hemocompatible with regard to reducing plasma coagulation and platelet adhesion, and the modified surfaces were nontoxic to fibroblast cells and were not hemolytic. The 7-day subcutaneous infection rat model showed that the modified polyurethane significantly reduced S. epidermidis infection by ~2.5 log units over unmodified films, and the modified polymers did not lead to additional irritation/toxicity to the animal tissues.

Supplementary Material

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acsami.3c18231.

Water contact angles of unmodified and modified PU materials; representative chromatogram of nucleotides extracted from biofilm cells; representative histological images of tissue samples; and heat map of histopathological parameters of tissue samples with scores (PDF)