Facile modification of medical grade silicone for antimicrobial effectiveness and biocompatibility: a potential therapeutic strategy against bacterial biofilms

Abstract

A one-step modification of biomedical silicone tubing with N,N-dimethyltetradecylamine, C14, results in a composition designated WinGard-1 (WG-1, 1.1 wt% C14). A surface active Silicone-Amine-Phase (SAP) is proposed to account for increased wettability and increased surface charge. To understand the mechanism of antimicrobial effectiveness, several procedures were employed to detect whether C14 leaching occurred. An immersion-growth (IG) test was developed that required knowing bacterial Minimum Inhibitory Concentrations (MICs) and Minimum Biocidal Concentrations (MBCs). The C14 MIC and MBC for Gm- uropathogenic E. coli (UPEC), commonly associated with catheter associated urinary tract infections (CAUTI), were 10 μg/mL and 20 μg/mL, respectively. After prior immersion of WG-1 silicone segments in growth medium from 1 to 28 d, the IG test for the medium showed normal growth for UPEC over 24 h indicating the concentration of C14 must be less than the MIC, 10 μg/mL. GC-MS and studies of the medium inside and outside a dialysis bag containing WG-1 silicone segments supported de minimis leaching. Consequently, a 5 log UPEC reduction (99.999% kill) in 24 h using the shake flask test (ASTM E2149) cannot be due to leaching and is ascribed to contact kill. Interestingly, although the MBC was greater than 100 μg/mL for P. aeruginosa, WG-1 silicone affected an 80% reduction via a 24 h shake flask test. For other bacteria and Candida albicans, greater than 99.9% shake flask kill may be understood by proposing increased wettability and concentration of charge illustrated in the TOC. De minimis leaching places WG-1 silicone at an advantage over conventional anti-infectives that rely on leaching of an antibiotic or heavy metal such as silver. The facile process for preparation of WG-1 silicone combined with biocidal effectiveness comprise progress toward the goals of device designation from FDA for WG-1 and clearance.

Graphical Abstract

Introduction

The body has an abundance of microbial systems in dynamic balance, ideally in homeostasis, that is, a steady state in physical and chemical balance for external and internal systems. When physiological balance is disturbed by disease, surgery, medication, a wound, a foreign body, immunodeficiency, or other circumstance, some commensal organisms coexisting with their human host become opportunistic pathogens.¹ Examples of commensal bacteria that are opportunistic pathogens include Streptococcus pneumoniae and Staphylococcus aureus, uropathogenic Escherichia coli (UPEC, motile from the intestine to the urethra),² and those from the environment such as Pseudomonas aeruginosa.³ Opportunistic pathogens readily adhere to hydrophobic silicone and form biofilms.^4–5 Protected by biofilm formation, infections result in increased morbidity and mortality and constitute a serious threat to patients in hospitals and care facilities.⁶

Selected examples of silicone medical devices and Adverse Events (AEs) are briefly summarized below. The physical location of an implanted silicone device is an important factor in determining predominant silicone-associated pathogens and other AEs.

- Urinary catheters may be implanted and held in place in the bladder by a balloon (Foley) or may be tube-like devices for intermittent use.⁷ About 80% of Catheter Associated Urinary Tract Infections (CAUTI) are due to Escherichia coli including motile uropathogenic E. coli (UPEC) usually originating from the intestine.^{2, 7}

- Endotracheal tubes (ETs) provide life-saving oxygen to the lungs from mechanical ventilation but become coated rapidly with biofilms from host oral and environmental pathogens. Thus, despite the life-saving benefits of intubation, direct access to the lungs often results in Ventilator-Associated Pneumonia (VAP) leading to increased morbidity and high mortality (33 – 50%).⁸ VAP pathogens in the endotracheal microbiome depend on the patient, geographical region, and other factors, but Gm- Pseudomonas aeruginosa is common and deadly.⁹

- An airway stent is a tube that is placed in the windpipe to open a narrowed area and restore normal breathing.^10–11 From an examination of explanted airway stents, high percentages of Candida spp., were found as well as Gm+ Staphylococcus aureus and Gm- P. aeruginosa.¹² In addition to biofilms, silicone airway stents are subject to adhesion of mucus and airway plugging. The use of 3D printing technology to make patient-specific silicone stents results in fewer AEs due to migration and reduced need for stent changes but infections and mucus adherence are not addressed by shape.^{11, 13}

- Implanting a silicone cerebrospinal fluid (CSF) shunt is a treatment for hydrocephalus to reduce intracranial pressure by redirecting CSF from the subarachnoid cavities of the brain to an alternative region for absorption.¹⁴ For neonates, the failure rate for CSF shunts is about 50% during the first two years, with 25% due to infection. CSF shunt tubing has a 1–2 mm ID which makes blockage of the lumen and entry ports by neural cell growth an AE responsible for the other 25%.¹⁵

- Voice prostheses often fail within the first few months due to the growth of Candida albicans.^16–17 While the virulence potential remains controversial,^18–19 C. albicans growth causes early valve failure, compromised speech, and aspiration pneumonia.²⁰

For these devices and others,^21–22 the emerging threat of drug resistance among pathogens has resulted in higher hospital costs, longer hospital stays, increased hospital mortality, and even morbidity. Biofilms provide a protective shield from host immunity resulting in chronic and recurrent infections that defy common antibiotics.^23–24

To begin the implementation of a strategy aimed at minimizing biofilm formation, we turned to experience with silicone surface chemistry²⁵ and explored modification of as-manufactured biomedical silicone tubing. We report herein a facile one-step modification of silicone catheter tubing with the tertiary C14 amine CH₃(CH₂)₁₃(CH₃)₂N that is designated WG-1 silicone. This facile modification is the subject of a recent patent.²⁶ We describe composition, physical surface characterization, and in vitro experiments for antimicrobial effectiveness and biocompatibility. In vitro experiments suggest pathogen contact with the surface is the primary mode of antimicrobial action, which stands in contrast to medical devices that leach silver, antibiotics, or other biocidal agents.

Experimental

Materials and Preparative Methods

Materials:

N,N-dimethyltetradecylamine (C14) was purchased from TCI America; ethanol was from Fisher Scientific. C14 amine was distilled (100 °C, 1 Torr) and characterized by ¹H-NMR spectroscopy (Figure S1). Bard All-Silicone Foley catheters (SKU 165822, 22FR) were purchased and used as medical grade silicone substrates for modification and often designated “silicone tubing”. For laboratory experiments, the tubing was cut into 1 cm segments and then cut in half lengthwise creating 1 cm long half-pipe samples.

Silicone Surface Modification.

In a typical experiment, silicone tubing half-pipe segments were placed in a 3.0 wt% C14-ethanol solution at 40 °C and stirred for 1 h (Figure 1). The concentration of C14 in ethanol solution for the treatment was defined as “feed concentration” or “feed”. There was no visual change after this treatment. Samples were removed, washed with deionized (DI) water three times, and placed in sterilized phosphate buffer (PB, 0.15 mM, pH 7).

Figure 1.

The surface modification process employing silicone catheter segments. C14 is the tertiary amine CH₃(CH₂)₁₃(CH₃)₂N. No change in appearance is observed.

Sterilization.

Two methods were employed for sterilization of WG-1 silicone: (1) After storage in sterilized phosphate buffer (PB, pH 7), samples were placed under UV overnight in a biosafety cabinet, or (2) Samples were placed in an autoclave-safe container with sterilized phosphate buffer (PB, pH 7) and autoclaved at 18 PSIG, 121 °C for 20 min. Stability was assessed by extracting C14 from the autoclaved segments using methanol and finding identical results by ¹H-NMR analysis to controls not autoclaved.

Physical Characterization.

ATR-IR. Attenuated Total Reflectance Infrared spectra were obtained with a Nicolet iS 5. ATR-IR is a rapid diagnostic technique with a depth of penetration of ~300 nm (Ge crystal). Ultraviolet-visible (UV-VIS) spectra were obtained with a Thermo Scientific UV spectrometer Genesys 10S.

Contact angles (CAs) were obtained using a Rame-Hart Model 210 Goniometer / Tensiometer equipped with an LCD camera. Deionized water (~18.2 MΩ·cm) was used as the probe liquid. A water drop was placed on the coated surface and the image was captured immediately for static contact angles. Advancing θ_A and receding θ_R Water Contact Angles (WCAs) were obtained while the water droplets were being added to or withdrawn from the surfaces. The captured images were analyzed using Dropview image software version 1.4.11. Average values were obtained from 5 observations.

Zeta potentials were obtained using a SurPASS Electrokinetic Analyzer (Anton Paar). The method of obtaining zeta potentials from streaming potentials has been described.²⁵ In brief, the zeta potential at the solid/liquid interface is based on measurements of streaming potentials and streaming currents. The NaCl electrolyte (1mM) flows through the measuring cell that contains a coated microscope slide as depicted in Figure 2. The gap created between sample and polypropylene reference is adjusted to generate a pressure difference between the inlet and outlet of the measuring cell.

Figure 2.

A schematic that depicts measurement of the zeta potential from the streaming potential.

During a measurement, pressure increases continuously and the pressure difference across the measuring cell (Δp) and streaming potentials (ΔU) are used to calculate ζ according to the method by Fairbrother-Mastin²⁷ using Equation 1,

$$\zeta =\frac{\Delta U}{\Delta p}\times \frac{\eta }{{\epsilon \times \epsilon }_0}\times K_B$$ Eq 1

with

ζ = zeta potential

ΔU/Δp = slope of streaming potential vs differential pressure

η = electrolyte viscosity

ε = dielectric coefficient of solvent

ε₀ = vacuum permittivity

K_B = electrolyte conductivity.

Gas Chromatography/Mass Spectrometry (GC/MS) employed an Agilent Technology 5973 GC/MS instrument equipped with a J&W DB-1ms GC Column (30 m, 0.25 mm, 0.25 μm, 7 inch cage). Mass spectra were obtained for octane, a negative control, and C14 (m/e = 241) in octane for a positive control. A series of C14 solutions in octane were made by serial dilution. As shown in Figure 3 (Left), the GC detection limit was ~1–5 mg/L or 1–5 ppm. For the study of leaching, a WG-1 silicone tubing segment was placed in a 2 mL centrifuge tube containing 0.6 mL phosphate buffer (PB). This was repeated to obtain 10 segments in PB. The samples were shaken for 1, 2, and 7 days respectively at 37 °C. PB was sampled by removal of 0.5 mL from each tube to give a total of 5 mL. This was added to a 15 mL centrifuge tube with 1 mL of octane followed by vortexing for 5 min. After separation, 0.6 mL octane was removed and used for C14 analysis.

Figure 3.

(Left) Calibration data for C14 GC showing the minimum detection limit of 1–5 mg/L. (Right) Octane extract after WG-1 immersion in PB for 1, 2, and 7 d at 37 °C. Vertical scale has the same arbitrary intensities.

Biophysical Tests

MBC and MIC for C14.

A stock emulsion of C14 in DI water (10 mg/mL) was prepared by sonication for 30 min. Dilutions were obtained by adding LB medium. The mixture was initially turbid but became clear when diluted to about 75 μg/mL indicating C14 solubility. Serial dilutions of C14 in media from 100 μg/mL to 0.8 μg/mL were added to a 24-well plate (2 mL per well). Dilutions in each well were inoculated with E. coli (UPEC) to achieve 10⁵ CFU/mL. The plate was shaken at 37 °C for 24 h with high humidity followed by measurement of optical density at 600 nm by using a UV-VIS spectroscopy. Aliquots were taken from the dilutions and were plated on LB agar and incubated at 37 °C for 24 h. A minimum of 1 log reduction in optical density determined MIC. The MBC was indicated by undetectable growth on LB agar. Experimental data for uropathogenic E. coli are shown in Figure S5.

ASTM-E2149.

The “shake flask test” was conducted according to the ASTM E2149–01 standard. This test provides antimicrobial activity of WG-1 coatings in dynamic conditions against both Gm- and Gm+ bacteria and Candida albicans. A stock solution of respective pathogens was cultured in sterile medium from a single colony prior to experimentation. Optical density at 600 nm was used to establish the initial concentration of the bacterial solution. This stock solution was then used to inoculate sterilized flasks containing a test medium (either phosphate buffer solution (PB) or M9), by dilution to ~2.0×10⁵ CFU/mL and a final total volume of 20 mL. Bard All-Silicone segments (control) were sterilized in the same manner as WG-1 samples with either UV or autoclave in PB. Eleven treated segments and 11 control samples (untreated segments) were then placed in prepared flasks filled with 20 mL test medium to achieve ~1.16 cm²/mL surface area to volume ratio. A flask containing just the pathogen and medium was also introduced as an additional control. The flasks were then subjected to gentle rotation at 37 °C in an incubator-shaker. Afterwards, 100 μL aliquots from each flask were removed, serially diluted, and plated on agar. The plates were incubated for ~24 h at 37 °C. Colonies were then counted and compared. Sequential tests for antimicrobial effectiveness of WG-1 were conducted by simply rinsing samples with DI, sterilized under UV overnight and then retesting according to the above procedure.

Dialysis bag leaching test.

A test for C14 leaching was devised by modifying a procedure described by Mohorcic.²⁸ Testing was performed on two common pathogens responsible for CAUTIs, uropathogenic Escherichia coli (ATCC# 700928, Gm-) and Enterococcus faecalis (ATCC#29212 Gm+).²⁹

Eight (20 cm in length) dialysis tubes (SnakeSkin 3.5 KDa) were knotted on one end and autoclaved at 120 °C and 18 PSIG for 15 min. Under sterile conditions, the tubes were then filled with 10 mL of M3 broth, inoculated to ~2000 CFU/mL with one of the pathogens. Each dialysis tube was placed in a 50 mL centrifuge tube containing 20 mL of M3 broth inoculated to ~2000 CFU/mL of the same pathogen. Duplicates were made by placing the following in each of two bags: 20 μL N,N-dimethytradecylamine (positive controls); 15 untreated silicone segments (negative controls); 15 Lubri-Sil segments (silver comparison); and 15 WG-1 segments. The labeled 50 mL centrifuge tubes were partially closed to allow for gas exchange while holding the dialysis tubes in place and then placed in an orbital shaker at 37 °C for 24 h. Afterwards, UV-Vis was used to establish optical density of the solution inside and outside the bags for calculation of bacterial concentrations. Additionally, aliquots were removed from inside and outside the dialysis tubes and plated onto M3 agar plates in duplicates. Plates were then incubated for 24 h at 37 °C. CFUs were determined, and plate images were obtained; images of the dialysis bags and surrounding media in the 50 mL centrifuge tubes were taken. Data are shown in Table 1, Figure 4, and Figures S6 and S7.

Table 1.

Summary of Results from the Dialysis Bag Experiments^a

pathogen	CFU/mL	Bard silicone	C14 amine	Bard Lubri-Sil	WG-1
uropathogenic E. coli (Gm–, ATCC# 700928)	inside	1.17 × 109	0	1.08 × 109	0
	outside	1.06 × 109	0	8.6 × 108	1.09 × 109
E. faecalis (Gm+, ATCC#29212)	inside	1.2 × 107	0	1.5 × 106	1.5 × 101
	outside	3 × 106	0	1.4 × 106	1 × 107

^aConcentrations are CFU/mL after 24 h incubation. Duplicate samples with four measurements for all tests. Values are estimated ±0.6 log.

Figure 4.

Images of centrifuge tubes and dialysis bags for one of two identical sample sets after incubation for 24 h, E. coli (ATCC# 700928): A, C14 amine, positive control, B. Bard silicone segments, negative control, C, Bard Lubri-Sil (silver comparison) and D, WG-1 silicone segments.

- Black arrows point to clear media, corresponding to ~9 log kill.

- White arrows point to turbid media due to normal E. coli growth (~1.1 × 10⁹ CFU/mL, Table 1).

Biofilm resistance.

WG-1 silicone and control samples were placed into individual wells of a sterile 24-well cell culture plate (triplicates). Each well was filled to cover samples with a bacterial or yeast suspension (1.5 mL, ~10⁵ CFU/mL in growth medium) followed by incubation for 24 h at 37 °C. After removing the plate, samples were gently rinsed 3 times with 1 mL of PB, placed in sterile 15 mL centrifuge tubes with 2 mL of trypsin (0.25%, 1X) to facilitate the detachment of cells from the surfaces,³⁰ held for 2 min at 37 °C, sonicated at 120 W output for 10 min at 25 °C, and vortexed for 30 s. Serial dilutions (100 μL) were plated on M3 agar plates and incubated for 24 h at 37 °C. Images were obtained and colonies counted.

In vitro biocompatibility.

A cytotoxicity study was conducted by following ISO 10993–5, a standard test that determines if toxic substances leach from samples. Cell growth and survival are assessed after extraction of as-received silicone tubing and WG-1 silicone. Human Dermal Fibroblast cells (HDFs) were employed in this study. In brief: (1) WG-1 silicone tube segments and untreated controls were extracted (respectively) in cell growth medium (Fibroblast Medium, Sciencell Research Laboratories) at 37 °C for 24 h., (2) Separately, HDFs were grown to 70–80% confluence in a 24-well cell culture plate, (3) The cell growth medium was replaced with the extraction medium (or blank cell growth medium as control) followed by incubation at 37 °C with 5% CO₂ for 48 h. Triplicates were made for each sample. Cell morphology after 48-h incubation was examined by optical microscopy.

Quantitative measurement of cell death used the alamarBlue assay.^31–32 After 48-h incubation, the extraction medium in the 24-well plates was replaced by 10% (v/v) alamarBlue reagent (alamarBlue dye, Thermo Scientific) mixed with fresh cell growth medium and the samples were incubated for 2 h at 37 °C with 5% CO₂. After 2 h incubation, the medium containing alamarBlue reagent was transferred to a 96-well plate (100 μL / well) to obtain fluorescence intensity (excitation / emission: 530 / 590 nm, triplicate measurements). Cell viability is calculated via Eq. 2.

$$\text{Cell viability}=\frac{{\text{Fluorescence}}_{\text{sample}}-{\text{Fluorescence}}_{\text{blank}}}{{\text{Fluorescence}}_{\text{control}}-{\text{Fluorescence}}_{\text{blank}}}\times 100\%$$ Eq. 2

Results and Discussion

Composition.

Initially, a number of amines and amine feed concentrations were explored. Based on comparative test results, N,N-dimethyltetradecylamine, CH₃(CH₂)₁₃N(CH₃)₂ or C14, with 3 wt% feed became a preferred composition due to a favorable combination of antimicrobial effectiveness and biocompatibility (Figure 1). Four successive extractions of amine with deuterated methanol from a known mass of modified samples followed by ¹H-NMR spectroscopy provided a composition 1.1 wt% amine for 3 wt% amine feed (Table 1S). The designation WG-1 is used for C14–3, that is, Bard silicone catheter tubing using 3 wt% amine feed. As noted below, WG-1 tubing half-pipe segments were used for most physical and biophysical studies.

ATR-IR.

Attenuated Total Reflectance Infrared Spectroscopy is a rapid diagnostic technique with a depth of penetration of ~300 nm (Ge crystal³³. WG-1 silicone spectra were obtained by subtraction from those for untreated silicone. Silicone catheter tubing has numerous, intense peaks that overlap with C14, but an absorption for CH stretching for WG-1 occurred in a window at 2920 cm⁻¹ and was assigned to amine (Figure 5). Though not understood at present, storing samples in phosphate buffer (PB) led to consistent results for biophysical tests described below. ATR-IR spectra revealed three broad peaks characteristic of PO₂ bands (Figure S3).³⁴ A peak at 1150 cm⁻¹ which had the least interference from a broad PDMS shoulder was assigned to ʋ_sym PO₄ (lit. 1062 cm⁻¹).³⁴ In summary, ATR-IR spectra support the near-surface presence of C14 amine and a [H₂PO₄]⁻ counter ion.

Figure 5.

ATR-IR spectra: a, C14; b, silicone spectrum subtracted from WG-1; dashed line marks the CH stretch at 2920 cm⁻¹.

Contact angles (CAs).

Hydrophobicity facilitates bacterial colonization^35–36 while hydrophilicity favors resistance to attachment.^37–38 Figure 6 shows that as a drop expands or retracts from a surface, CAs distinguish between hydrophobic, unmodified silicone (Figure 6-A) and WG-1 (Figure 6-B) which has a 20° lower advancing CA and a 37° lower receding CA. The decrease of 37° in receding CA for WG-1 compared to unmodified silicone confirms a change from hydrophobic to more hydrophilic character.^39–40 Other relevant observations concerning CA measurements are:

1. The WCAs reported in Figure 6-A were obtained after immersion of Bard silicone catheter segments in ethanol for 1 h at 40 °C as a control to match conditions for formation of WG-1 silicone. After the ethanol immersion, WCAs remain unchanged in air for months (117°, θ_A and 67°, θ_R).

2. As above, WG-1 silicone was stored in PB prior to tests. The high CA hysteresis (~ 67°) is attributed to rapid molecular motion that results in protonated nitrogen in water (polar) taking the place of relatively non-polar (CH₃)₂N- in air.

3. The CAs reported for WG-1 in Figure 6-B fall into the range of “PEGylated” polymers known for resistance to biofouling.⁴¹

Figure 6.

Contact angles by the advancing and receding drop method: A, unmodified silicone catheter; B, WG-1 silicone. Three phase contact lines are shown in red.

Zeta potentials.

Kirby has written an authoritative discussion of zeta potential measurements for silicones and related polymers.⁴² In brief, zeta potentials evaluate surface charge in a flow of dilute electrolyte (e. g., 1 mM NaCl used here). Zeta potentials provide a metric for charge at the outermost substrate surface and insight concerning prospects for bacterial membrane disruption.^{25, 43–44}

As configured, our Anton-Paar SurPASS flow cell for measuring zeta potentials requires flat samples. Therefore, Sylgard-184 coated microscope slides were employed as proxies for silicone tubing. The method for obtaining zeta potentials from streaming potentials is summarized in the Experimental Section, while details have been published.²⁵

Using the same procedure shown in Figure 1, Sylgard 184 coatings with increasing C14 wt% were prepared by increasing C14 feed. Results in Table 2 show that zeta potentials ζ increase (less negative) with increasing feed until a maximum of about −34 mV is reached for C14–3. Unexpectedly, beyond 3 wt% C14 feed, ζ decreases.

Table 2.

Zeta Potentials for the C14-Modified WG-1 Surrogate: Sylgard 184

C14 feed (wt %)	ζ (mV)
0 (Sylgard 184)	−88
0.5	−82
1.0	−75
2.0	−65
3.0	−34
4.0	−37
5.0	−42

Quaternary moieties such as those based on trimethylammonium, [Me₃N-]⁺ employed by Gottenbos⁴³ and Wang²⁵ gave positive zeta potentials. In contrast, less negative charge is generated for C14 by protonation due to the amine pKa of ~10. Thus, an unexpected feature for the Sylgard-184 proxy is that, although ζ does not become positive, it increases about 50 mV compared to unmodified silicone. Accordingly, [(CH₃)₂NH]⁺ charge is invoked for WG-1 to account for antimicrobial effectiveness by disruption of the bacterial membrane.⁴⁵

A Model for WG-1 Surface Chemistry.

A surface-concentrated Silicone-Amine-Phase (SAP) is proposed for WG-1 silicone (Figure 7). This model is suggested by the similar solubility parameters of polydimethylsiloxane chains, 7.3 cal^1/2 cm^−3/2, and the CH₃ and 14 methylene moieties for C14 (8.0 cal^1/2 cm^−3/2) and the driving force of hydrogen bonding of water-protonated amine at the aqueous medium interface. The presence of bulk amine is shown in Figure 7 as uncharged molecules.

Figure 7.

Silicone-amine phase (SAP) model. Green for silicone and amine alkyl chains denotes miscibility; blue protonated amine surface concentration is driven by hydrogen bonding at the water interface. The protonated nitrogen is depicted as an atomic Janus-like moiety with opposing “faces” comprised of charge (N⁺-H = blue) and methyl groups (hydrocarbon entities, green). Bulk amine is green with blue N. Not shown is the presumed near-surface anion (H₂PO₄)⁻. Biophysical data support membrane disruption represented by uropathogenic E. coli with hair-like fimbriae (blue → red). Hydrophilicity (low θ_R) is thought to account for weak adsorption of cell remnants.

The outermost surface of the SAP model has a hydrophilic N⁺-H moiety depicted blue in Figure 7. This arises from protonation of the basic amine (pKa ~10). Strong hydrogen bonding of N⁺-H to water accounts for the hydrophilic character of the wetted surface. That is, WG-1 silicone has a 20° lower advancing contact angle (Figure 6-B-adv) than unmodified silicone (Figure 6-A-adv) and a receding contact angle of 30° that is 37° lower (Figure 6-B-rec) than unmodified silicone (Figure 6-A-rec). The decrease of 37° in receding contact angle for WG-1 compared to unmodified silicone is noteworthy. As shown by Francolini⁴⁶ a hydrophilic (wetted) surface resists bacterial adhesion by the physical characteristic of wettability. Strong steric repulsion by a dense hydration layer was shown for a combination of poly(ethylene glycol-dimethacrylate) (PEGDMA), and 2-methacryloyloxyethyl phosphorylcholine (MPC) that resulted contact angles of about 10°.⁴⁷ Steric repulsion by the hydration layer together with a sharp nanostructured topology resulted in synergistic dual functionalities that affected resistance to fouling and bactericidal activity.⁴⁷

While not reflective of a strong hydration layer, the contact angle change from hydrophobic (unmodified silicone) to hydrophilic for WG-1 silicone is attributed to SAP concentration at the outermost surface (Figure 7) and constitutes a component of steric repulsion that helps resistance to bacterial colonization and weak adsorption of bacterial membrane fragments by the physical characteristic of wettability.

In summary, the SAP model explains two important physical features of WG-1 silicone, namely hydrophilicity (especially receding contact angles) and charge (zeta potentials) that are synergistic in conferring resistance to biofilm formation and other favorable characteristics described herein.

Composition and SAP Model.

The average weight of an unmodified silicone half-pipe segment is 0.17 g. The surface area of a segment is approximately 2.4 cm² (Bard French 22, 1 cm × 1 cm × 0.1 cm). From 3 wt% feed (Figure 1), 1.1 wt% amine corresponds to 1.9 × 10¹⁸ C14/cm². The calculated surface C14 density for maximum packing in a perpendicular position from a model described in Figure S4 is 2 × 10¹⁶ C14/cm². This calculated C14 surface density is ~80 times lower than the calculated surface C14 density (1.9 × 10¹⁸ C14/cm²). Thus, about 99% C14 amine is sub-surface or bulk as depicted in Figure 7.

Antimicrobial tests

Minimum inhibitory concentration (MIC) and minimum biocidal concentration (MBC) for C14.

As noted earlier, an objective for characterization of WG-1 is differentiating between pathogen kill due to leaching and kill due to membrane disruption on contact with the surface. To understand the significance of “normal growth” in medium in which WG-1 silicone had been immersed, the MIC and MBC for C14 were determined for uropathogenic E. coli (UPEC, ATCC# 700928) by adapting a reported method.⁴⁸ A decrease in optical density (OD) at 10 μg/mL signaled the MIC for C14 against UPEC while colonies were not observed for C14 concentrations ≥ 20 μg/mL which is the MBC. Experimental data are shown in Figure S5.

To obtain a broader knowledge of the biocidal character of C14, the MIC and MBC were determined by the same method for several Gm+ and Gm− bacteria (Table 3). The contrast in MIC for Pseudomonas aeruginosa and susceptibility to WG-1 silicone in an ASTM E2149 “shake flask” test is noted below.

Table 3.

MIC and MBC for C14 against Common Pathogenic Bacteria

pathogen/Gram (±)	MIC (μg/mL)	MBC (μg/mL)
uropathogenic E. coli (Gm−)	10	20
E. faecalis (Gm+)	6.25	12.5
P. aeruginosa (Gm−)	>100	>100
K. pneumoniae (Gm−)	25	50
S. aureus (Gm+)	25	50
S. epidermidis (Gm+)	12.5	25

ASTM E2149, the shake-flask test for biocidal effectiveness.

Antimicrobial testing was conducted according to a standard ASTM “shake flask” test, E2149–01.^{25, 49} The pathogens in Table 4 are listed in order of estimated prevalence for Catheter Associated Urinary Tract Infections (CAUTIs).^{29, 50} However, these pathogens are associated with differing prevalence for other silicone devices such as endotracheal tubes.⁵¹ The acronym ESKAPE includes six nosocomial pathogens that exhibit multidrug resistance and virulence: E. faecium, S. aureus, K. pneumoniae, A. baumannii, P. aeruginosa, and Enterobacter spp.⁵² These pathogens are common isolates from patients with multi-drug resistant hospital acquired infections.⁵³ Results from ASTM E2149 tests for four ESKAPE pathogens are designated by “E” in the pathogen designation column of Table 4.

Table 4.

Antimicrobial Activity of WG-1 against Pathogen Challenges Determined by Shake Flask and Biofilm Tests

		shake flask test		biofilm test
pathogen/Gram (±)e	percent infectionsb (%)	log reductiona,c	% kill	log reductionc	% reduction
E. coli (G−) K-12	23.90	4.3B	>99.99	d
E. coli (G−) uropathogenic	23.90	>5B	>99.999	3.2	99.93
C. albicans	17.80	>4B	>99.99	3.4	99.9
E. faecalis (G+, E)	11	>5B	>99.999	d
P. aeruginosa (G–, E)	10.3	0.8B	~80	1.4	95.9
K. pneumoniae (G–, E)	10.1	5B	99.999	2.9	99.9
S. aureus (G+, E)	1.6	0.9M9	~90	2.3	99.5

^aSuperscripts indicate the test medium, both at pH ~7; B = PB; M9 = M9 medium.

^bPercent pathogens for UTIs vary in reviews. Those listed are from Infect. Control Hosp. Epidemiology, 2016, 37, 1288–130 and Nat Rev Microbiol. 2015, 13, 269–284.

^cLog reduction for C14–3 is based on growth for untreated Bardex all-silicone.

^dNot determined.

^e“ESKAPE” pathogens are designated “E” (see text).

Gm− bacteria:

Uropathogenic Escherichia coli (UPEC) is commonly associated with CAUTIs and comprises ~80% of urinary tract infections.^54–55 WG-1 silicone achieved a log 5 (99.999%) reduction via a 24 h ASTM E2149 shake flask test. Separately, after WG-1 was immersed for 28 d in pH 6.5 M3 medium an ASTM E2149 shake flask test confirmed antimicrobial effectiveness against UPEC. No surviving bacteria were observed suggesting one measure of durability (> 99.99% or 4 log kill).

In another investigation of durability, after an initial shake flask test one set of samples was rinsed with D.I. water, subjected to a second UPEC challenge identical to the first, and then rinsed and tested a third time. The results for this sequence of three tests were: (1) 5.5 log reduction, >99.99% kill; (2) 3.6 log reduction, 99.97% kill, and (3) 4.2 log reduction, >99.99 % kill. These sequential tests on the same samples attest to sustainability of antimicrobial effectiveness for WG-1.

P. aeruginosa has an environmental origin and causes a wide array of life-threatening infections, particularly for patients with compromised immunity.^56–57 P. aeruginosa is the main cause of morbidity and mortality in cystic fibrosis (CF) patients.⁵⁸ As noted by Ciofu, the multiple mechanisms employed by Gm− P. aeruginosa to survive antibiotic treatment while growing in a biofilm represent an important therapeutic challenge.⁵⁹

A 24 h shake flask test resulted in an 80% reduction of P. aeruginosa by WG-1 (Table 4). In contrast, the MBC for P. aeruginosa is greater than 100 μg/mL, which was the highest concentration of C14 tested. The solubility limit of C14 is about 75 μg/mL which means that at 100 μg/mL C14 is partly soluble and partly in emulsion. Considering the resistance of P. aeruginosa to antibiotics, the antimicrobial effectiveness of WG-1 against this pathogen is worthy of further study.

Klebsiella pneumoniae has recently gained notice as an infectious agent due to a rise in the number of severe infections and strains that have acquired additional genetic traits and become either hypervirulent or antibiotic resistant. K. pneumoniae is the ‘K’ in the ESKAPE pathogens, the group of opportunistic pathogens that together account for the majority of clinically significant multi-drug resistant (MDR) hospital acquired infections (HAIs).⁶⁰ K. pneumoniae is often prevalent in CAUTIs and cases of Ventilator Associated Pneumonia (VAP).^61–62 WG-1 is highly effective against K. pneumoniae with a log 5 reduction for the 24 h shake flask test.

Gm+ bacteria:

As noted above, Gm+ S. aureus is one of the six ESKAPE pathogens having multidrug resistance and virulence. For S. aureus, a 0.9 log (~90%) reduction was obtained for the shake flask test. This can be compared with a MIC of 25 μg/mL and an MBC of 50 μg/mL in aqueous medium (Table 3). As with P. aeruginosa, performance is better for WG-1 silicone compared to C14 in aqueous medium.

Gm+ Enterococcus faecalis has a dual nature with some strains recognized as probiotics but others are ESKAPE pathogens with multidrug resistance and virulence.^63–64 For E. faecalis, a 5 log (~99.999%) reduction was obtained for the shake flask test. This can be compared with a MIC of 6.25 μg/mL and an MBC of 12.5 μg/mL in growth medium (Table 3). The sensitivity of E. faecalis and the moderate reduction for S. aureus are significant results that attest to WG-1 effectiveness against both Gm+ and Gm− bacteria.

Candida albicans: a mycological pathogen.

The virulence of Candida spp. remains controversial.^{19, 65} In any event, WG-1 has strong growth inhibition for the yeast C. albicans with a > log 4 reduction (> 99.99%) for the shake flask test. The effectiveness of WG-1 silicone against C. albicans may prove to be of importance to extending the life of voice prostheses. As noted above, fouling by C. albicans is the primary cause of interference with the valve operation that shortens the lifetime of the device.

Biofilm Tests

In essence, the biofilm test described in the Experimental Section assesses viability of pathogens that have a range of adherence to the surface. To quantify the biofilm growth, trypsin was used to degrade the biofilm and detach the bacterial cells from the material surface. Zhou, et al., investigated the degradation effect of bovine trypsin on multispecies biofilm. It was reported that bovine trypsin can overcome the biofilm structure, disperse biofilm and bacteria flora, and reduce the exopolysaccharide (EPS) and bacterial biomass without significant impact on the ratio of live/dead bacteria after the trypsin treatment.⁶⁶

Table 4 shows results for WG-1 biofilm resistance against UPEC, three ESKAPE pathogens (P. aeruginosa, K. pneumoniae, S. aureus) and C. albicans. Among these pathogens, S. aureus and Candida albicans are common isolates from patients with MDR HAIs.^{53, 67–68} WG-1 is effective in reducing the biofilm formation of UPEC (3.2 log reduction, > 99.9% kill), Klebsiella pneumoniae (2.9 log reduction, 99.9% kill) and S. aureus (2.3 log reduction, 99.5% kill) compared to untreated controls. These results are interesting because WG-1 exhibited high antimicrobial effectiveness against these pathogens in shake flask tests with higher than 4 log reduction (Table 4). Thus, there are minimal surviving bacteria in the biofilm that was removed by trypsin and in the medium sampled the shake flask test.

Although a shake flask test showed only modest antimicrobial performance for C14 against P. aeruginosa (0.8 log reduction, ~80% kill) and S. aureus (0.9 log reduction, ~90% kill), the biofilm test gave a 1.4 log reduction (95.9% kill) for P. aeruginosa and 2.3 log reduction (99.5% kill) for S. aureus. These results suggest that a longer residence time on the surface for strains such as P. aeruginosa and S. aureus results in more effective membrane disruption compared to transient contact during the shake flask test. The ability of WG-1 to reduce biofilm formation of these pathogens is significant as biofilm formation is a crucial factor in the persistence and recurrence of VAP.

Leaching.

A recent review analyzed studies of silver coated catheters that included a total of 36,783 patients. Overall, there was no significant difference in the CAUTI rate between silver coated and non-coated catheters. However, the CAUTI risk was lower in the coated group among patients requiring catheterization for longer than 14 days.⁶⁹ A coating comprised of a palladium, gold, and silver alloy layer and a hydrogel outer layer significantly decreases the incidence of CAUTI.⁷⁰ An interesting study reported leaching of metals into urine was minimal or undetectable.⁷¹ In one urine sample 0.2 μg/L silver was detected which is 4% of the permitted daily exposure for drugs.⁷⁰

In response to pathologies that include acute hydrocephalus and traumatic brain injury, external ventricular drains (EVDs) are put in place for patients with elevated intracranial pressure. The rifampin-clindamycin antibiotic combination incorporated into Bactiseal EVD is aimed at inhibiting Gm+ bacterial colonization.^72–73 However, the rifampin-clindamycin antibiotic combination is ineffective against multidrug resistant (MDR) Gm− bacteria such as E. coli that can cause ventriculitis.⁷⁴ While uncommon, ventriculitis was reported for a rifampicin resistant S. epidermidis strain.⁷⁵

The prior section on antimicrobial tests for WG-1 demonstrated broad activity against Gm+, Gm− bacterial strains and the yeast C. albicans. Experimental studies in this section are aimed at ascertaining whether bacterial kill occurs due to C14 leaching or cell contact with the surface. As noted earlier, avoiding a drug-device classification is important to facilitate bringing a product to market.⁷⁶

Zone of Inhibition (ZOI) test.

This is a qualitative test described by Bauer.⁷⁷ Figure 8-A shows a ZOI from an instructional publication by the American Society of Microbiology (Figure 10 in this ASM publication).⁷⁸ The ZOI from diffusion of an antibiotic from the white disk creates a clear zone of inhibited bacterial growth that is indicated by the white arrow. The ruler in the image is employed to measure the width of the zone.

Figure 8.

A. Example of measurement of a ZOI from disk-diffusion of an antibiotic. B. Undetectable ZOI for WG-1 silicone tubing segment on a light tan “lawn” of E. coli after 24 h incubation.

To test WG-1 silicone, a tubing segment was placed on a nutrient agar plate lightly inoculated with E. coli (K-12). After 24 h incubation to form a lawn of bacteria, the agar plate was removed from the incubator and imaged. As shown by Figure 8B, a zone of inhibition was not observed around the WG-1 silicone tubing segment. Thus, detectable leaching of an antimicrobial substance did not occur. As a control, a paper disc with ampicillin showed the expected ZOI on the lawn of bacteria (not shown).

Immersion of WG-1 segments and bacterial growth (immersion/growth or IG test).

This is a test we devised for bacterial growth in medium after removal of WG-1 silicone that had been immersed therein for up to 28 d at pH of 6 or greater. The purpose of carrying out this test was to discern if there was leaching of an antimicrobial constituent from WG-1 silicone into growth medium that, after removal, would subsequently inhibit normal bacterial growth.

Growth rates for bacteria that are often exponential in media such as M9 are fundamental in microbiological studies.⁷⁹ This rapid bacterial growth can be inhibited as was found for a secondary amine that reduced the growth of Escherichia coli and Acinetobacter baumannii but not the growth of the Gm+ Bacillus subtilis.⁸⁰ Given the sensitivity of growth rates to environmental and chemical factors, the IG test was devised to assess whether leaching of moieties from WG-1 silicone would affect bacterial growth.

The IG test protocol is depicted in Figure 10: (1) WG-1 and untreated silicone segments were added to separate 50 mL flasks containing Luria-Bertani (LB) medium as an extraction medium; (2) The flasks were shaken for 1, 3, 5, 7, 14 or 28 d at 37 °C; (3) The segments were removed from their respective flasks; (4) The media were inoculated with uropathogenic E. coli (UPEC, ATCC 700928, ~1 × 10⁵ CFU / mL) followed by 24 h incubation; (5) 100 μL aliquots from the respective flasks were removed, serially diluted, spread on agar, incubated for 24 h at 37 °C, and colonies counted (Figures 10–5A and 10–5B).

Figure 10.

Depiction of steps 1–4 for the immersion-growth (IG) test to detect leaching. A. UPEC growth after immersion of unmodified silicone segments for 28 d. and B. Indistinguishable growth in medium after immersion of WG-1 silicone segments. Similar plate images were obtained after other immersion times (1, 3, 5, 7, and 14 days (Step 1).

Unexpectedly, in an initial test the UPEC concentration was consistently higher for WG-1 silicone compared to untreated silicone. However, a test after untreated silicone was immersed in ethanol alone at 37 °C showed growth to 2 × 10⁹ CFU / mL which matched that of WG-1. Apparently, some processing aid or other constituent inhibited growth for as-received silicone segments. Therefore, the term “untreated” means immersed in ethanol alone for 1 h and not treated with C14.

IG test combined with MIC. The IG test showed normal UPEC growth after prior immersion of WG-1 in growth medium for up to 28 d. For normal UPEC growth, the concentration of C14 must be well below the MIC of 10 μg/mL for 24 h. However, from ASTM E2149, “the shake flask test” a 5-log reduction (99.999% kill) was found in 24 h (Table 3). If a leach rate of 10 μg/mL is ascribed to the first day, it is obviously too high to account for the result of the IG test. Rather, this result and other studies described below support contact kill and minimal levels of leaching.

Dialysis bag growth tests for Gm− UPEC (ATCC# 700928) and Gm+ Enterococcus faecalis (ATCC#29212).

These tests were designed to augment IG leaching studies for WG-1 and to compare antimicrobial effectiveness for WG-1 silicone to Bard Lubri-Sil. Adapting a method described by Mohorcic, a test using dialysis tubing was developed.²⁸ The Experimental section describes tests for Gm− uropathogenic Escherichia coli (UPEC, ATCC# 700928) and Gm+ Enterococcus faecalis (ATCC#29212).²⁹

Figure 4 shows dialysis bags after UPEC incubation. The results are summarized according to designations in Figure 4.

1. Figure 4-A. A positive control with 20 μL C14 (N,N-dimethyltetradecylamine) inside the dialysis bag filled with 10 mL M3 broth inoculated with UPEC. The dialysis bag was placed in a 50 mL centrifuge tube containing 20 mL of M3 broth that was inoculated with UPEC. The 3.5 kDa cutoff for dialysis is 10 times greater than the molecular weight of C14 and the concentration of C14 was more than 10 times higher than the MBC. As a result, C14 diffused from inside to outside the bag with 100% kill inside and outside.

2. Figure 4-B. Fifteen untreated Bard silicone catheter segments (negative control) were placed inside the bag. After 24 h, the UPEC concentration reached 1.1 × 10⁹ CFU/mL inside and outside the bag. Similarly, Bard silicone catheter segments had negligible effect on E. faecalis growth, which reached about 6 × 10⁶ CFU/mL inside and outside the bag (Figures S6 and S7, Supporting Information).

3. Figure 4-C. Fifteen Bard Lubri-Sil segments were placed inside the bag. Compared with unmodified silicone, Bard Lubri-Sil is ineffective against UPEC in this test with comparable growth inside and outside the bag. Similarly, Bard Lubri-Sil had negligible effectiveness against E. faecalis inside and outside the bag. In other tests and trials Bard Lubri-Sil has shown variable antimicrobial effectiveness.⁸¹

4. Figure 4-D. Fifteen WG-1 silicone segments were placed inside the bag. After 24 h incubation a black arrow points to the clear medium inside the bag that reflects negligible bacterial concentration but the white arrow points to the cloudy medium outside the bag where normal bacterial growth occurred. Table 1 shows WG-1 effected a 9-log kill of Gm− UPEC inside the bag (clear) but UPEC grew normally outside the bag (Figure 4D, cloudy). The image in Figure 4-D and results summarized in Table 1 show that at pH 7 UPEC is killed primarily by contact with undetectable leaching.

Figure S6 in Supporting Information shows images for the dialysis test for Gm+ E. faecalis (ATCC#29212). Under our conditions, 24 h growth of E. faecalis for the control (silicone segments) was about 10⁷ CFU/mL, which is less vigorous than E. coli (10⁹ CFU/mL). Table 1 shows WG-1 effects a 6-log kill inside the bag but there is no kill of E. faecalis outside the bag with normal growth to 1 × 10⁷ CFU/mL. Figure S7 shows one set of four plates after incubation. In one test for WG-1 segments, growth of E. faecalis could not be detected inside the tube; in a duplicate test, one surviving colony was detected giving an average log 6 reduction. In contrast, results for tests of Bard Lubri-Sil showed negligible E. faecalis kill inside and outside the tube.

GC-MS.

The semi-qualitative results from ZOI, IG, and dialysis bag tests showed undetectable C14 leaching. Additional quantitative data were obtained by gas chromatography coupled with mass spectrometry (GC/MS). The experimental design was adapted from that used by Li.⁸² while the analysis of the mass spectrum followed that reported for C14 by Abu-Serag and confirmed a molecular ion for C14 at 239.4 AMU.⁸³

GC on control samples showed that C14 detection at low concentrations was at the instrumental limit of ~1–5 mg/L or 1–5 ppm (Figure 3-Left). The GC data for the octane extract after WG-1 immersion in PB for 1, 2, and 7 d at 37 °C is shown in Figure 3 (Right). No trend is apparent. Rather, each data set is characterized by a slight rise at ~9.72 min. Thus, it appears that the C14 concentration is no greater than ~5 mg/L after 24 h and there is no detectable increase with time.

The surface area to volume ratio for the IG test was 1.5/1 and 1:1 for the GC test. Considering experimental limitations, the GC finding of about 5 μg/mL for 24 h is in fair agreement with an upper limit of less than 10 μg/mL for the IG test described above.

Biocompatibility.

In vitro biocompatibility.

As a modification designed to minimize microbial colonization on a medical device, biocompatibility with mammalian cells is of importance for clinical applications and FDA regulations.^{25, 84} Tests for in vitro cytotoxicity are an essential part of establishing biocompatibility for WG-1 silicone. A test was conducted by following the ISO 10993–5 standard as it has been found reproducible, rapid, and sensitive for evaluation of in vitro biocompatibility for medical devices.

Figure 9 shows HDF cell morphology after 48-h incubation for cell growth medium as a blank control (Figure 9-A), untreated silicone as a negative control (Figure 9-B), WG-1 silicone (Figure 9-C) and latex as a positive control (Figure 9-D). The morphology and proliferation of HDFs incubated with the extract of WG-1 silicone was identical to that for cell growth medium and untreated silicone. In contrast, significant cell death and poor cell proliferation was observed for latex positive control. The results for alamarBlue^® assays for HDFs after 40-h incubation are shown in Figure 11. Cell viability was found to be 100 ± 3% for cell growth medium, 101 ± 3% for untreated silicone, 102 ± 2% for WG-1 silicone, and 0% for latex. These results indicated that the WG-1 silicone does not release toxic agents in the presence of cell growth medium. The 100% viability on alamarBlue assay shows WG-1 did not cause any toxicity effects to HDFs and should be a safe candidate for medical devices and clinical applications.

Figure 9.

Cytotoxicity study following ISO 10993–5 standard. Images of HDF elongated live cell morphology were captured after 48-h incubation with A, medium control; B, untreated silicone control; C, WG-1 silicone; D, latex positive control with rounded, dead cells.

Figure 11.

AlamarBlue^® assays for Human Dermal Fibroblasts (HDFs) after 40 h incubation.

To confirm in-lab studies, North American Science Associates (NAMSA) carried out cytotoxicity assays on WG-1 according to ISO 10993–5 employing L-929 mouse fibroblast cells. WG-1 Silicone tubing segments were extracted in growth medium for three days at 37 °C. The undiluted growth medium extract was added to a 70% confluent layer of L-929 mouse fibroblast cells and incubated for 3 d (37 °C, 5% CO₂). WG-1 Silicone gave 100% cell viability (non-cytotoxic) that was indistinguishable from untreated silicone segments.

Cell adhesion.

Coatings that can prevent cell or tissue adhesion are useful in a wide range of applications. One example noted in the introduction is ventricular shunts that remove excess cerebrospinal fluid for patients with hydrocephalus. The ingrowth of human tissue into the lumen of ventricular shunts often causes shunt obstruction, which is a life-threatening failure for patients.⁸⁵

A conventional approach of achieving resistance to cell adhesion is improving coating hydrophilicity. Manfredini, et al., reported a coating method based on the adsorption of poly(styrene-co-3-sulfopropyl methacrylate) to prevent nonspecific adhesion of cells on tissue culture polystyrene (TCPS) surfaces. A non-fouling efficacy of 97% was achieved with a 16 ± 5° static contact angle.⁸⁶ Almousa, et al., modified polyvinylchloride surfaces with polyethylene glycol (PEG) and polyvinylpyrrolidone (PVP) polymers end-capped with succinimidyl groups to improve hydrophilicity. The modified surface exhibited significantly reduced 3T3 cell adhesion with a 50%–69% decrease for PEG and a 64%–81% decrease for PVP, as compared to unmodified polyvinylchloride.⁸⁷

A cell adhesion study for WG-1 with HDFs employed glass disks coated with Silastic^® MDX4–4210 biomedical grade silicone, which is often used for facial rehabilitation.⁸⁸ After cure, the coated disks were modified by the WG-1 process (Figure 1). Untreated and WG-1 samples were placed in wells of a 6-well plate. HDFs in growth medium were added to the wells with samples. After 48-h incubation, HDFs attached and grew on the untreated Silastic silicone but there was no attachment or growth on WG-1 (Figure 12). The 20° lower advancing CA and 37° lower receding CA for WG-1 (Figure 6-B) compared to untreated silicone is thought to confer WG-1 silicone with resistance to cell adhesion. This assumption is supported by Gao and McCarthy’s study that showed a low receding contact angle signals hydrophilic character and is due to hydrophilic groups that emerge rapidly at the molecular or nanoscale on the wetted surface.⁸⁹

Figure 12.

A. Elongated HDF cells attached and grew on untreated biomedical silicone; B. No attachment or growth on WG-1 silicone.

Conclusion

Evidence is presented for a surface-concentrated Silicone-Amine-Phase (SAP) for WG-1 silicone, which is prepared by a one-step method using C14 amine (Figures 7 and 1). Physical characterization reveals increased wettability and, by analogy with a zeta potential study on a model system, increased surface charge that is assumed to originate from protonation of the outermost amine layer. The SAP model explains hydrophilicity and charge that are synergistic in conferring resistance to biofilm formation.

MICs and MBCs for aqueous C14 in growth medium were determined to assess effects of possible leaching (Table 3). The highest sensitivity is Gm+ Enterococcus faecalis with a MIC of 6.25 μg/mL and MBC 12.5 μg/mL. Gm− uropathogenic E. coli (UPEC) is also sensitive with a MIC of 10 μg/mL and MBC 20 μg/mL. The least sensitive to C14 in growth medium is Gm− P. aeruginosa with MIC and MBC greater than 100 μg/mL, the highest concentration employed. Interestingly, a 24 h shake flask test resulted in an 80% reduction of P. aeruginosa by WG-1 (Table 4).

Other than UPEC and E. faecalis, most pathogens have relatively high MBCs (Table 3), but greater than 99.9% kill in the shake flask test. This may be understood in terms of concentration of charge at the SAP surface that operates over the nano-to-microscale and facilitates bacterial membrane disruption illustrated in Figure 7. A similar membrane disruption mechanism is the basis for effectiveness of multi-charged aggregates,⁹⁰ peptide aggregates,⁹¹ and polycations.^{25, 92–93} This mechanism may provide a pathway for antimicrobial effectiveness without resistance buildup.⁹⁴

An immersion-growth (IG) test was developed to discern if there was detectable leaching of an antimicrobial constituent from WG-1 silicone into growth medium. After prior immersion of WG-1 silicone in growth medium from 1 to 28 d the IG test showed normal growth for UPEC. In particular, after 24 h, the concentration of C14 must be less than 10 μg/mL, the MIC for UPEC. The 5 log reduction (99.999% kill) in 24 h using the shake flask test (ASTM E2149, Table 3) cannot be due to a leaching concentration less than the MIC (10 μg/mL). The results of the shake flask test thus support contact kill.

In vitro ISO 10993–5 tests in our laboratory and third-party tests for cell compatibility provide strong evidence for biocompatibility of WG-1 silicone. Interestingly, human cells (HDF) attached and grew on a model untreated silicone (a Silastic coating) but there was no attachment or growth on WG-1 (Figure 12). Increased hydrophilicity compared to untreated silicone and charge are thought to confer resistance to cell adhesion.

Undetectable leaching places WG-1 at an advantage over many anti-infective devices currently available commercially as WG-1 has comparable or in some cases better performance over a broad range of pathogens while not relying on leaching. The ease of preparation and properties advances WG-1 silicone toward the goal of device designation from FDA. All tests reported herein were at pH 6.5 or 7 which is near many bodily fluids such as CSF and saliva. However, human urine can have a broad pH range depending on sex⁹⁵ and disease state.^96–97 In future work tests for effectiveness of WG-1 silicone will be extended to a broader pH range, tests for hemocompatibility will be carried out and control of attachment of human cells will be explored to widen the range of potential applications for WG-1 silicone.