An Organoselenium Compound Inhibits Staphylococcus aureus Biofilms on Hemodialysis Catheters In Vivo

Abstract

Colonization of central venous catheters (CVCs) by pathogenic bacteria leads to catheter-related bloodstream infections (CRBSIs). These colonizing bacteria form highly antibiotic-resistant biofilms. Staphylococcus aureus is one of the most frequently isolated pathogens in CRBSIs. Impregnating CVC surfaces with antimicrobial agents has various degrees of effectiveness in reducing the incidence of CRBSIs. We recently showed that organoselenium covalently attached to disks as an antibiofilm agent inhibited the development of S. aureus biofilms. In this study, we investigated the ability of an organoselenium coating on hemodialysis catheters (HDCs) to inhibit S. aureus biofilms in vitro and in vivo. S. aureus failed to develop biofilms on HDCs coated with selenocyanatodiacetic acid (SCAA) in either static or flowthrough continuous-culture systems. The SCAA coating also inhibited the development of S. aureus biofilms on HDCs in vivo for 3 days. The SCAA coating was stable and nontoxic to cell culture or animals. This new method for coating the internal and external surfaces of HDCs with SCAA has the potential to prevent catheter-related infections due to S. aureus.

INTRODUCTION

Intravascular catheters, such as central venous catheters (CVCs) and hemodialysis catheters (HDCs), are essential components of treatment within the health care system. The main complication that stems from utilizing these catheters is the development of bacterial infections (9, 24). Such infections may be either localized to the catheter and surrounding skin or systemic—spread via the bloodstream to deep internal organs. In the United States, the rate of nosocomial catheter-related bloodstream infections (CRBSIs) in intensive care units ranges from 3 to 8% (10). The mortality rate associated with these infections may reach as high as 36% (1). It is estimated that 80,000 cases of CRBSI occur annually, with an estimated cost of $300 million to $5 billion (21, 24).

At the time of insertion, skin surface microorganisms may contact and colonize the external surface of the catheter, leading to CRBSIs (24, 35). Alternatively, microorganisms from the patient's skin or health care worker's hands may be transferred to the catheter hub, which leads to the colonization of all surfaces of the catheter and, subsequently, to CRBSIs (23, 24, 35). Coagulase-negative Staphylococcus species, Staphylococcus aureus, and other bacteria commonly found on the skin are the organisms most frequently isolated from the blood of patients with CRBSIs (3, 24, 35), although CRBSIs are also associated with other pathogens, including Pseudomonas aeruginosa, Enterococcus species, and Acinetobacter baumannii (26, 30, 32). In patients on dialysis, the mortality rate of S. aureus-related CRBSIs may reach as high as 20% (3).

After it colonizes either the inner or outer surface of the catheter, S. aureus produces a biofilm. Biofilms are defined as microbially derived, sessile communities within which bacterial cells are attached to a substratum, an interface, or to one another (7). During the formation of these biofilms, bacteria proliferate and produce exopolymers (extracellular matrix) (12, 16, 20). Within the biofilm, bacteria are highly resistant to antimicrobial agents, including antibiotics, germicides, and disinfectants (6, 7, 22). The development of these biofilms on the surfaces of CVCs within the first 2 weeks after catheter placement, followed by the migration of bacteria into the bloodstream, are important steps in the occurrence of CRBSIs (3).

Thus, to decrease the incidence of CRBSIs, it is essential to prevent the initial colonization and the subsequent development of bacterial biofilms on the surfaces of CVCs. As a result, several strategies to prevent biofilm formation on these catheters have been developed and clinically evaluated. Among these strategies is coating the catheters with antimicrobial agents, including chlorhexidine-silver sulfadiazine (CSS) and minocycline-rifampin (3, 14, 27, 35).

Organoselenium that is covalently bound to a solid matrix retains its ability to catalyze the formation of superoxide radicals that prevent bacterial colonization of solid surfaces (19). We recently showed that organoselenium-methacrylate polymer-coated cellulose disks inhibit attachment and biofilm formation by both S. aureus and P. aeruginosa (34). In this study, we investigated the efficacy of the organoselenium antimicrobial agent selenocyanatodiacetic acid (SCAA) in preventing the development of S. aureus biofilms, both in vitro and in vivo, when applied as a coating to the inner and outer surfaces of hemodialysis catheters.

MATERIALS AND METHODS

Bacterial strains, media, and growth conditions.

Staphylococcus aureus strain AH133, which constitutively expresses green fluorescent protein from plasmid pCM11, was used for all studies (18). The strain was routinely grown in Luria-Bertani (LB) broth at 37°C with shaking (250 rpm). Biofilm formation was examined using Trypticase soy broth (TSB) (MP Biomedicals, Solon, OH) as the growth medium. To maintain pCM11 in AH133, both LB broth and TSB were supplemented with 3 μg/ml erythromycin.

Catheters.

Uncoated polyurethane Decathlon high-flow, long-term catheters were provided by Spire Biomedical (Bedford, MA). Each catheter was cut into several 1-cm segments. The lumen, as well as the outer surface, of each segment was pretreated with 100% ethylene oxide under 100% humidity at 55°C for 30 min. To dissipate the ethylene oxide, the segments were kept in the treatment chamber at room temperature for 5 to 6 h and then allowed to degas in a chemical fume hood for an additional 24 h at room temperature. The segments were then coated externally and internally by covalent attachment of SCAA as follows. Each segment was completely submerged for 5 h under shaking conditions in 1.5 ml of dry acetone containing 4 mg/ml SCAA (Eburon Organics, Lubbock, TX), 3.3 mg/ml N-hydroxysulfosuccinimide (Pierce Chemical, Rockford, IL), and 3.3 mg/ml 1-ethyl-3-(3-dimethyl aminopropyl)carbodiimide hydrochloride (Pierce Chemical). The coating solution was then decanted, and the segments were rinsed several times for 1 h in sterile distilled water (dH₂O). The segments were then soaked overnight in sterile dH₂O and left to dry for 4 to 6 h at room temperature.

In vitro biofilm systems. (i) Static biofilm.

The static biofilm assay was done as previously described, using the microtiter plate assay with some modifications (11, 25). Biofilms were quantified by crystal violet assay as previously described (11) or by determining the CFU per catheter segment. To determine the CFU per segment, each segment was carefully removed from the well, rinsed gently with sterile distilled H₂O, and placed into a microcentrifuge tube containing 1 ml of phosphate-buffered saline (PBS). The tubes were placed in a water bath sonicator for a total of 10 min to loosen the cells within the biofilm and then vigorously vortexed three times for 1 min to detach the cells. Suspended cells were serially diluted 10-fold in PBS, and 10-μl aliquots of each dilution were spotted onto LB agar plates. The plates were incubated at 37°C for 24 h, and the CFU were counted. The CFU per segment were determined using the following formula: CFU × dilution factor × 100. To confirm the efficacy of our protocol for recovery of biofilm-associated bacteria from each HDC catheter segment, the segments were placed into fresh tubes containing PBS and the sonication and vortexing process was repeated. We recovered no CFU from plating of the PBS (data not shown); nor were residual bacteria visualized on the vortexed segments.

(ii) Flowthrough continuous-culture system.

Development of S. aureus biofilm on 1-cm catheter segments was achieved using a multicell flowthrough continuous-culture system (5, 31). Catheter segments were secured in the flow cell with stainless steel brackets, the flow cells were attached to silicone tubing, and these units were sterilized by autoclaving. Sterile TSB plus 3 μg of erythromycin per ml was maintained in a 10-liter reservoir and pumped to the flow cell through 1/8-inch (internal diameter) silicone tubing at a flow rate of 1 ml per min using a six-roller-head peristaltic pump. Biofilms were allowed to develop for 5 days. Catheter segments were then extracted from the flow cells, gently rinsed in PBS, and examined by confocal laser scanning microscopy (CLSM) using an Olympus Fluoview FV300 (Olympus America, Center Valley, PA).

In vivo biofilm system. (i) Animal model.

In vivo inhibition of S. aureus biofilm by SCAA-coated HDCs was examined in a murine model based on the previously described murine and rabbit models of bacterial colonization of subcutaneously inserted devices (4, 15, 29). Adult female BALB/c mice weighing 20 to 24 g were anesthetized using a mixture of isoflurane and oxygen, and their backs were shaved. The shaved areas were completely cleansed with 95% ethanol, and a 1-cm incision was made centrally in the shaved area. Subcutaneous pockets (about 2 cm by 1.5 cm) were generated on the back of each mouse (one pocket on each side) by dissecting the skin from the underlying muscles through this incision. Either uncoated or SCAA-coated catheter segments (1 cm) were inserted in these pockets, one segment per pocket. Aliquots containing 10⁴ CFU AH133 in 100 μl were injected in the area of the inserted catheter segments. The incisions were closed by suturing, and the mice were returned to their cages with free access to food and water. The mice were monitored twice a day for signs of infection or distress. After 3 days of observation, the mice were euthanized, the connective tissue around the catheter segments was dissected, and the segments were carefully removed to avoid disrupting the biofilm. Extracted catheter segments were gently rinsed in PBS, and the biofilms were analyzed by CLSM as described above. Animals were treated in accordance with the protocol approved by the Institutional Animal Care and Use Committee at Texas Tech University Health Sciences Center in Lubbock, TX.

(ii) Quantification of in vivo biofilms.

Biofilms were quantified using the COMSTAT program (13). We obtained 20-image stacks of each biofilm by CLSM and analyzed the images for the following parameters (Table 1): biomass, or volume of the biofilm (μm³/μm²); average thickness (μm); surface area (μm²), a reflection of the efficiency with which the strain colonizes the surface; and surface area to biomass ratio, an estimate of the portion of the biofilm exposed to nutrients.

Table 1.

Quantitative analysis of 3-day biofilms formed on uncoated and SCAA-coated hemodialysis catheters^a

Location and treatment	Biomass (μm3/μm2)	Avg thickness (μm)	Surface area (μm2)b	Surface area/biomass ratioc
Outer surface
Uncoated	0.4770	0.5937	347,748	1.5189
SCAA	0	0	0	0
Inner surface
Uncoated	0.1643	0.2789	101,385	1.2856
SCAA	0	0	0	0

^aTwenty-image stacks were acquired from each surface of the catheter; the images were analyzed using the COMSTAT program (13).

^bReflects the efficiency with which AH133 colonized the surface.

^cEstimates the portion of the biofilm exposed to nutrients.

Analysis of the SCAA coating. (i) Long-term stability.

One-centimeter segments of the SCAA-coated HDCs were completely immersed in 10 ml of PBS in sterile glass tubes and incubated at room temperature for 4, 6, or 8 weeks. At the end of the incubation period, the segments were removed, dried, and utilized in the in vitro static biofilm model system as described above to determine the durability of the coating.

(ii) Release of selenium from the SCAA-coated HDCs.

Three 1-cm catheter segments were placed into a sterile 15-ml plastic tube to which 3 ml of sterile 0.9% NaCl was added, an amount sufficient to completely cover the segments. The tubes were incubated at 37°C for 3 days. The amount of free selenium in the solution was determined at a commercial facility (TraceAnalysis, Inc., Lubbock, TX).

(iii) In vitro toxicity.

All chemicals, unless otherwise noted, were obtained from Fisher Scientific (Fair Lawn, NJ). Dulbecco's modified Eagle's medium (DMEM) supplemented with sodium pyruvate (0.11 mg/ml), penicillin (100 U/ml), streptomycin (0.1 mg/ml), and 10% (vol/vol) fetal bovine serum was used for cell culture. COS-1 cells (ATCC, Manassas, VA) were plated into 12-well dishes (Nunc; Nalge Nunc International, Rochester, NY) for 48 h. Wells were then washed twice with 2 ml of 0.9% NaCl that had been incubated with control- or SCAA-coated HDC segments for 3 days (as described above; ca. 0.07 mM selenium), or with excess SCAA (1 mM selenium). Each solution was diluted 1:10 in DMEM prior to use. A third 2-ml aliquot of the appropriate solution was then added to the cells and incubated for 48 h. Images were obtained using a Zeiss microscope equipped with a 10× objective and a Canon EOS 1/Ds camera and optimized using Adobe Photoshop software. Following imaging, the cells were removed from the surfaces of the wells using 0.125% (wt/vol) pancreatin in a basic salt solution containing EDTA. Enzyme activity was quenched with 0.4 ml of DMEM, and the cells were counted using a hemocytometer. Duplicate wells for each condition were analyzed.

(iv) In vivo toxicity.

SCAA-coated HDC segments containing 4 mg of selenium per segment were examined for acute systemic toxicity in mice at a commercial medical device testing laboratory in an ACUC-approved procedure (NAMSA, Northwood, OH). The analysis was based on the International Organization for Standardization's standard for tests for systemic toxicity (12a). Selenium was extracted from the SCAA-coated HDC segments using either 0.9% NaCl or sesame oil and injected into adult male H1a CVF mice (5 mice/extract) at a dose of 50 ml/kg of body weight (approximately 58.4 μg selenium or 18.1 μg selenium per mouse, respectively). Control groups were injected with either 0.9% NaCl or sesame oil. The mice were monitored for mortality, distress, and clinical signs of infection. Total body weight was recorded prior to injection and every day for 3 days postinjection.

Statistical analyses.

The results of the crystal violet assays and the CFU assays were analyzed using Prism version 4.03 (GraphPad Software, San Diego, CA). Comparison of the biofilms formed on uncoated and SCAA-coated HDCs was done using unpaired, two-tailed t tests. The results of the selenium release assay were analyzed by one-way analysis of variance (ANOVA) with Dunnett's multiple comparison posttest to determine significant differences.

RESULTS

Selenocyanatodiacetic acid coating inhibits the development of S. aureus biofilm on hemodialysis catheters in vitro.

We examined the ability of SCAA covalently attached to the inner and outer surfaces of the HDCs to inhibit AH133 biofilms using the microtiter plate assay (11, 25). The SCAA coating significantly reduced biofilm biomass (Fig. 1a) when measured by crystal violet assay. Additionally, CFU of AH133 were reduced by the presence of SCAA on the HDCs (Fig. 1b). To confirm these results, biofilms were allowed to develop on uncoated and SCAA-coated HDC segments for 48 h and were visualized by CLSM. Characteristic AH133 biofilm was seen on the inner and outer surfaces of the uncoated segments (Fig. 2a and b). In contrast, very few microorganisms were detected on either surface of the SCAA-coated segments (Fig. 2c and d). Thus, SCAA coating inhibited the development of AH133 biofilms on both surfaces of the HDCs.

Fig 1.

Selenocyanatodiacetic acid (SCAA) coating on hemodialysis catheters (HDCs) inhibits development of S. aureus biofilms. S. aureus strain AH133 was inoculated into microtiter wells containing TSB plus 3 μg erythromycin/ml. The microtiter plate assay was done as described in Materials and Methods. (a) Biofilm biomass was quantified by the crystal violet assay (absorbance at 595 nm). (b) Numbers of viable organisms (CFU) present in the biofilms. To determine stability of the SCAA coating on exposure to aqueous environment, coated and uncoated HDC segments were soaked in PBS for indicated times and used in the microtiter biofilm assay. Values represent the means of 3 or more independent experiments ± standard deviations.

Fig 2.

SCAA visibly reduces S. aureus biofilm formation on HDCs. Outer and inner surfaces of uncoated (a and b, respectively) and SCAA-coated (c and d, respectively) HDC segments were examined by confocal laser scanning microscopy (CLSM) at ×20 magnification after 48 h of incubation.

SCAA coating inhibits the development of AH133 biofilms on both surfaces of the HDCs under continuous-flow conditions.

S. aureus biofilms were initiated as described in Materials and Methods and allowed to mature over a 5-day period. At the end of the fifth day, catheter segments were removed, gently rinsed in PBS to eliminate loosely attached planktonic cells, and examined by CLSM. S. aureus AH133 produced a biofilm on the outer surface of the uncoated HDC segments (Fig. 3a). Additionally, the bacteria colonized the entire inner surface of the uncoated HDC segments and were in early stages of biofilm development (Fig. 3b). Differences in the stages of biofilm development between the inner and outer surfaces are possibly due to the increased accessibility of the outer surface to the bacteria. Clinically, biofilm generally develops on the outer surface of central venous catheters, followed by extension to the inner surface. Compared with the uncoated segments, very few individual cells were found attached to either surface of the SCAA-coated HDC segments (Fig. 3c and d). These results indicate that the SCAA coating is effective in inhibiting S. aureus biofilm formation under conditions resembling host blood flow for an extended period.

Fig 3.

SCAA coating inhibits the development of S. aureus biofilm on HDCs in the flowthrough continuous-culture biofilm system. Biofilms were developed for 5 days, and the outer and inner surfaces of uncoated (a and b, respectively) and SCAA-coated (c and d, respectively) HDC segments examined by CLSM at ×20 magnification.

SCAA coating on HDCs inhibits the development of S. aureus biofilms in vivo.

These experiments were done using the modified murine model of chronic biofilm infection (4, 15, 29). The catheter segments were then removed and examined by CLSM. AH133 produced a biofilm on both sides of the uncoated catheter segments (data not shown). In contrast, no biofilm development was detected on either side of the SCAA-coated catheter segments (data not shown). The 3-day in vivo biofilms were quantified using the previously described COMSTAT program (13). AH133 produced a biofilm with clusters of differentiated microcolonies on the outer surface of the catheter segments, with an average thickness of 0.5937 μm and surface area of 347,748 μm² (Table 1). The biofilm formed on the inner surface of the uncoated catheter segments was less vigorous, with an average thickness of 0.2789 μm and surface area of 101,385 μm² (Table 1). In contrast, no biofilm structure was detected on either side of the SCAA-coated catheter segments (Table 1). These results demonstrate that the SCAA coating also inhibits the development of catheter-related biofilm in vivo.

The inhibitory effect of the SCAA coating is long-lasting.

We determined whether the antibiofilm activity of SCAA-coated catheters is stable over a prolonged period of exposure to aqueous solution. Compared to uncoated catheters, a significant reduction in CFU was detected on SCAA-coated HDC segments after up to 8 weeks of aqueous exposure (Fig. 1b). However, there was no statistical difference between SCAA-coated segments, regardless of exposure time (Fig. 1b), showing that the SCAA coating on the catheter is stable and that the effect on biofilm is durable.

The SCAA coating is nontoxic in vitro or in vivo.

Uncoated or SCAA-coated HDC segments were incubated in 0.9% NaCl at 37°C for 3 days. The solution was analyzed by a commercial laboratory (TraceAnalysis) and found to contain 58.4 mg selenium/liter, indicating the release of approximately 58.4 μg selenium per 1-cm segment. To determine if the level of released selenium is toxic to tissue cells, 1-cm uncoated or SCAA-coated HDC segments were incubated in 0.9% NaCl at 37°C for 3 days and the eluates used in a cytotoxicity assay as described in Materials and Methods. The released selenium had no effect on the viability of COS-1 cells, while the presence of 1 mM selenium dramatically reduced cell viability and cell count (Fig. 4). Similar to cells incubated with the 0.9% NaCl or uncoated eluate, cells incubated with the SCAA eluate showed no changes in morphology, attachment, or density (Fig. 4).

Fig 4.

Selenium eluted from SCAA-coated HDC segments is not cytotoxic. COS-1 cells were plated and incubated for 48 h with DMEM containing (a) 0.9% NaCl, (b) 0.9% NaCl incubated with uncoated HDC segments, (c) 0.9% NaCl incubated with SCAA-coated HDC segments (0.074 mM selenium), or (d) 1 mM selenium (from SCAA). Magnification is ×53. Following imaging, cells were removed from the surfaces of the wells and duplicate wells were counted using a hemocytometer. Cell counts are indicated for each condition.

Potential in vivo toxicity of SCAA-coated HDCs was evaluated by a commercial medical device testing laboratory (NAMSA). Uncoated or SCAA-coated HDC segments were incubated in 0.9% NaCl or sesame oil for elution of selenium to occur, and the extracts were injected into adult male H1a CVF mice (5 mice/extract). Over 3 days, all mice survived, were clinically normal, and their body weights increased by 9 to 11% (data not shown).

DISCUSSION

The results of this study point to the effectiveness of the covalent attachment of the organoselenium compound selenocyanatodiacetic acid to HDCs in preventing the development of S. aureus biofilms. The SCAA-coated hemodialysis catheters are highly effective in inhibiting the development of S. aureus biofilms. Image analysis of the static biofilm clearly showed only a few AH133 cells attached to the inner and outer surfaces of SCAA-coated HDCs (Fig. 2c and d), a finding strongly supported by quantification of the biofilm by crystal violet assay and CFU enumeration (Fig. 1). Similarly, AH133 failed to colonize and initiate biofilm on SCAA-coated HDCs in the 5-day flowthrough continuous-culture system (Fig. 3c and d). Organoselenium appears to be more effective in inhibiting S. aureus biofilms than other previously proposed or commercially available antimicrobial agents. Darouiche et al. (4) showed that coating an intravenous catheter with triclosan (antiseptic) and dispersin B (antibiofilm enzyme) was more efficient in inhibiting the colonization of catheters by different pathogens than coating with chlorhexidine-silver sulfadiazine. However, compared with uncoated catheters, the greatest reduction in S. aureus colonization by either triclosan-dispersin B or CSS-coated catheters, as determined by CFU, reached only 2.5 to 3 logs (4). In comparison, SCAA-coated HDCs reduced S. aureus colonization by 4 to 5 logs (Fig. 1b). This effect is not limited to strain AH133. SCAA-coated HDCs completely inhibited biofilm development by other S. aureus strains, including clinical isolates obtained from burn patients (data not shown).

In vivo, the effect of SCAA-coated HDCs on S. aureus colonization appears to be superior to those of triclosan-dispersin B or CSS-coated catheters. Experiments using the murine model of biofilm development revealed that, while AH133 produced biofilms on both sides of the control HDCs, it failed to colonize either side of the SCAA-coated HDCs (data not shown). COMSTAT quantitative analysis of the biofilm supported these findings (Table 1). Previous experiments using the rabbit model of biofilm development revealed that S. aureus colonized 3.3% of triclosan-dispersin B-coated catheters, 13.3% of CSS externally coated catheters, and 3.3% of CSS externally and internally coated catheters (4). Although the two studies used different species of animals (mouse versus rabbit), the experimental design in both studies is basically the same (4, 15, 29).

The extreme effectiveness of organoselenium is probably due to its ability to catalyze the formation of superoxide radicals that inhibit bacterial attachment on the surface to which organoselenium is attached. We previously showed that cellulose disks coated with 0.2% selenium effectively inhibited bacterial attachment and biofilm formation by both S. aureus and Pseudomonas aeruginosa (34). Despite the effectiveness of SCAA-coated HDCs in inhibiting AH133 biofilm in vivo and in vitro, it will be essential to examine such effectiveness in clinical trials. Several clinical studies showed that different coating formulations vary in their effectiveness in inhibiting the colonization of CVCs by different bacterial pathogens, including S. aureus (2, 8, 17, 33).

One of the major challenges of the antimicrobial coating of medical devices, including CVCs, is the durability of the antimicrobial coating. An antimicrobial coating may be highly effective initially but gradually lose its effectiveness. As we showed in this study, covalently attached SCAA on HDCs is very durable in its anti-S. aureus effectiveness. SCAA-coated catheter segments reduced AH133 colonization and prevented the development of AH133 biofilms even after 8 weeks of soaking in aqueous solution following coating (Fig. 1b). This stability of the in vitro inhibitory effect of SCAA-coated catheters is superior to that obtained from catheters coated with other antimicrobial agents. After 4 days of soaking in growth medium, the CFU of S. aureus recovered on CSS-coated CVC increased by 3 logs compared with the CFU recovered on those soaked for only 24 h only (4). Similarly, Bach et al. (2) showed that after 15 days of aqueous exposure, the anti-S. aureus effectiveness of CSS-impregnated catheters decreased, as the zone of inhibition around the catheter segments was reduced from 12 to 7 mm.

Besides the in vitro durability, in vivo durability of the antimicrobial coating on the catheter is also essential. The risk of endoluminal colonization of catheters by bacteria is increased 7 to 10 days postinsertion (8). Clinical studies designed to compare the colonization rate of uncoated and antiseptic-coated catheters have produced differing results (8, 17, 33). Dünser et al. (8) showed that, in critically ill patients, whereas standard uncoated CVCs were first colonized 2 to 3 days postinsertion, those with CSS coating were first colonized 7 days postinsertion. In our animal model, SCAA-coated HDCs completely inhibited S. aureus colonization for 3 days after insertion (Table 1). We obtained similar results when the catheter segments remained in place for 5 days (data not shown). We attempted to increase the duration postinsertion; however, the fluorescent signal produced by AH133 was drastically reduced by 6 to 7 days postinsertion. We plan to conduct the in vivo experiment for 10 to 15 days and analyze the AH133 biofilm by determining the CFU within the biofilms on uncoated and SCAA-coated HDC segments.

Another major problem associated with antimicrobial-coated catheters is that systemic toxicity may result from the release of antimicrobial agents. However, the toxicity of organoselenium is extremely low (19, 28). Analysis showed that the amount of selenium released from the SCAA-coated catheters over several days was very low, 0.074 mM or less. In vitro, selenium extracted from catheter segments neither changed epithelial cell morphology nor reduced cell count (Fig. 4). Also, using a commercially available hemolysis assay, we ruled out the possibility that SCAA-coated HDCs cause a significant lysis of red blood cells (data not shown). Finally, the extracted selenium was not lethal when injected into experimental animals. Throughout the 3-day experiment, neither control nor experimental animals lost weight or showed signs of sickness. In contrast, although no signs of silver toxicity were demonstrated by Roe et al. (29) when they examined the toxicity of plastic catheters coated with bioactive silver nanoparticles, the animals implanted with silver nanoparticle-coated catheter segments lost about 8% of their body weight.

In conclusion, we developed a new method for coating the internal and external surfaces of HDCs with the organoselenium compound selenocyanatodiacetic acid. The SCAA coating on the HDCs was nontoxic, durable, and inhibited the development of S. aureus biofilms in vitro and in vivo. The use of SCAA-coated HDCs has the potential to prevent catheter-related infections due to S. aureus.