Surface-Dried Viruses Can Resist Glucoprotamin-Based Disinfection

Abstract

Touching of contaminated objects and surfaces is a well-known method of virus transmission. Once they are attached to the hands, viruses can easily get adsorbed and initiate infection. Hence, disinfection of frequently touched surfaces is of major importance to prevent virus spreading. Here we studied the antiviral activity of a glucoprotamin-containing disinfectant against influenza A virus and the model virus vaccinia virus (VACV) dried on inanimate surfaces. The efficacy of the surface disinfectant on stainless steel, polyvinyl chloride, and glass coupons was investigated in a quantitative carrier test. Vacuum-dried viruses were exposed to 0.25%, 0.5%, and 1% disinfectant for 5 min, 15 min, and 30 min without agitation, and residual infectivity was determined by endpoint titration. Although glucoprotamin was highly active against both viruses in suspension, limited antiviral activity against the surface-dried viruses was detected. Even after 30 min of exposure to 1% disinfectant, VACV was not completely inactivated. Furthermore, influenza A virus inactivation was strongly affected by the surface composition during the 5-min and 15-min treatments with 0.25% and 0.5% disinfectant. The results presented in this study highlight the relevance of practical tests to assess the antiviral activity of surface disinfectants. High virucidal activity in solution is not necessarily indicative of high antiviral activity against surface-dried viruses. In addition, we want to emphasize that the mere exposure of surfaces to disinfectants might not be sufficient for virus inactivation and mechanical action should be applied to bring attached viruses into contact with virucidal compounds.

INTRODUCTION

In the middle of the 1990s, glucoprotamin was presented as a novel disinfectant with broad activity against bacteria, mycobacteria, fungi, and enveloped and even lipophilic nonenveloped viruses. The antimicrobial substance is nonodorous and nonvolatile, is easily dissolved in water, and has excellent ecotoxicological properties (1, 2). Glucoprotamin is less toxic than aldehyde-based disinfectants and, thus, contributes to the safety and health of staff working in the medical environment, where disinfection is an important part of the daily work. In addition to its antimicrobial activity, glucoprotamin has favorable cleaning properties and is compatible for use with metal and plastics. In contrast to aldehydes, blood is not fixed on surfaces after exposure to glucoprotamin, and no residues remain on treated surfaces, as was observed for quaternary ammonium compounds (3). Hence, glucoprotamin is a true alternative to aldehyde-based disinfectants or quaternary ammonium compounds and has successfully been used in the last decade. Like other amine derivatives, the mode of action of glucoprotamin is rather unspecific and takes place through the destruction and disorganization of biological membranes (4). Hence, the disinfectant exhibits antiviral activity mainly against lipid-enveloped viruses. Nonenveloped viruses might be affected only after prolonged exposure, and the lack of virucidal activity against poliovirus is known (3).

The antiviral activity of disinfectants can be assessed by different test methods, and according to CEN Technical Committee TC 216, the test program for chemical disinfectants can be divided into three phases. First, the basic microbicidal activity of a disinfectant is assessed qualitatively in a suspension test with different test organisms that considers the effective exposure times and effective concentrations (phase 1). Second, further information is gained in quantitative suspension tests by investigating the microbicidal properties at different concentrations and different exposure times and temperatures (phase 2, step 1). In those studies, the kinetics of virus inactivation in the absence or presence of a soil load should be determined. However, suspension tests poorly reflect real conditions because viruses are incubated in large volumes of disinfectant, which enables excessive contact between viruses and the virucidal active compounds. To draw real conclusions about the microbicidal activity of a disinfectant under practical conditions, test methods representative of conditions of practical use of the disinfectant, including quantitative surface tests or hand-rubbing and hand-washing tests, must be applied (phase 2, step 2). Third, the final proof of the disinfectant's activity requires field tests under practical conditions (phase 3) (5, 6).

Although the establishment of standard test methods to assess the virucidal activity of disinfectants under conditions of practical use lags behind tests for bactericidal activity, several international and national standards focusing on this topic are available. Among them are an ASTM standard method for assessing the virucidal activity of a disinfectant on nonporous surfaces and a quantitative disc carrier test method that can be applied to bacteria, fungi, and viruses (7, 8). In 2013, the OECD published a guidance document on quantitative methods for evaluating the activity of microbicides used on hard nonporous surfaces, where the test procedure based on a ring trail with adenovirus was described (9). Until now, in Europe EN 14476:2013 has been the standard for evaluating the virucidal activity of disinfectants, but it comprises solely a quantitative suspension test (phase 2, step 1) (10). A quantitative standard test method for evaluation of the virucidal activity of chemical disinfectants on nonporous surfaces, prEN 16777:2014, is currently under development, but a period for public comments has only recently been opened (11). Also, the DVV (German Association for the Control of Virus Diseases) provides guideline for the quantitative determination of the virucidal activity of disinfectants on nonporous surfaces (12). The DVV guideline differs from the above-mentioned standards as it discriminates between tests for limited virucidal activity, low-level virucidal activity, and high-level virucidal activity, depending on the spectrum of activity of the investigated disinfectant against different test viruses with different levels of physicochemical resistance.

The DVV guideline was applied in the present study to analyze the antiviral activity of a glucoprotamin-containing surface disinfectant (26 g glucoprotamin in 100 g concentrate) against influenza A virus and vaccinia virus (VACV) dried on inanimate surfaces. Influenza A virus was chosen as the relevant test virus because the disinfectant should be used in facilities where infectious influenza viruses are handled. In this context, influenza A virus can be considered a model for enveloped respiratory viruses and represents viruses that are highly contagious and can be spread through droplets that might be present on environmental surfaces, like doorknobs or handrails. To comply with the DVV guideline and to extend the spectrum of investigated viruses, the established model virus VACV was included in this study to represent a further surrogate for lipid-enveloped viruses, like hepatitis B and C viruses, human immunodeficiency virus (HIV), and herpesviruses (13). In addition, VACV is also transmitted through droplet secretions or by inhalation of dust particles containing the infectious agent (14). Both test viruses are mentioned in international standards and national guidelines for use in the testing of biocides for their antiviral activity and represent enveloped viruses with low to medium levels of resistance against physicochemical treatments (7, 15). In contrast to suspension assays, which can be performed easily and relatively quickly, in the present study the antiviral activity was determined on carriers to reflect more practical conditions of use of the disinfectant. The surfaces stainless steel, glass, and polyvinyl chloride (PVC; floor material) were chosen to represent relevant surfaces that are typically found in hospitals or biopharmaceutical facilities and could be contaminated with infectious agents. Viruses were exposed to the disinfectant at different concentrations recommended by the manufacturer for different times, and the log₁₀ reduction factors (LRFs) achieved on the different surfaces were determined. The influence of the surface composition on virus inactivation was examined, and the antiviral activity of the glucoprotamin-containing disinfectant against viruses dried on inanimate surfaces was compared to the virucidal efficacy of the disinfectant in suspension tests.

MATERIALS AND METHODS

Viruses and cells.

Vaccinia virus strain Elstree (VR-1549; ATCC) was propagated and titrated in Vero cells (84113001; ECACC). Vero cells were grown at 37°C with 5% CO₂ in Dulbecco's modified Eagle's medium (DMEM; Sigma-Aldrich) supplemented with 10% fetal bovine serum (FBS) and 1% l-glutamine. For maintenance of the cells in culture and for VACV production, the FBS content in the medium was reduced to 2%. The cell culture-adapted influenza virus A/Singapore/6/86 (H1N1) (kindly provided by J. P. Gregersen, Novartis Vaccines and Diagnostics GmbH, Marburg, Germany) was grown and titrated in MDCK cells (kindly provided by the Institute of Virology, University of Ulm, Ulm, Germany). MDCK cells were propagated at 37°C with 5% CO₂ in DMEM-GlutaMAX (Life Technologies) supplemented with 3% FBS. For influenza A virus production and titration, cells were incubated at 34°C with 5% CO₂ in medium without FBS but complemented with 0.25 μg/ml trypsin.

Infected cell cultures were subjected to two freeze-thaw cycles to release the viruses. Cell debris was removed by centrifugation at 800 × g for 10 min, and the virus-containing supernatants were stored in aliquots at −70°C.

Disinfectant.

The investigated disinfectant is commercially available as a concentrated solution composed of 25 to 35% glucoprotamin, 10 to 20% butyl diglycol, 10 to 20% phenoxyethanol, and 1 to 5% fatty alcohol ethoxylates. The indicated active ingredient in 100 g of the concentrate is 26 g glucoprotamin.

Quantitative suspension test.

Quantitative suspension tests were performed under clean conditions as described by Blümel and colleagues, with slight modifications (16). Differing from the DVV guideline for quantitative suspension tests, the inoculum was prepared with bovine serum albumin (BSA; final concentration, 0.3 mg/ml), thereby adopting the method for inoculum preparation for tests under clean conditions of the DVV guideline for quantitative carrier tests (12). One hundred microliters of virus suspension in 0.3 mg/ml BSA (Cohn fraction V; Serva Electrophoresis, Germany) was mixed with 100 μl of water of standardized hardness (WSH) and 800 μl of 1.25-fold-concentrated disinfectant (prepared in WSH). The final concentrations of the disinfectant in the reaction mixture were 0.25%, 0.5%, and 1%, corresponding to glucoprotamin contents of 0.065 g/100 ml, 0.13 g/100 ml, and 0.26 g/100 ml, respectively. The mixtures were incubated at room temperature for 5 min, 15 min, and 30 min. At the end of the exposure times, the virucidal activity was stopped by immediate preparation of serial 10-fold dilutions with ice-cold cell culture medium, and residual infectivity was determined by titration in susceptible cells. Control reaction mixtures were prepared in exactly the same way but received WSH instead of disinfectant. Experiments were performed in three independent biological replicates.

Quantitative carrier test.

Stainless steel carriers (material 1.4435; surface roughness [Ra], <0.8) were obtained from Kasag Langnau AG (Langnau, Switzerland), PVC carriers (Forbo Colorex) were obtained from Intertapis AG Bodenbeläge & Parkett (Liebefeld/Köniz, Switzerland), and glass carriers were obtained from Walther+Müller Glasbau AG (Bern, Switzerland). The dimensions of the stainless steel and PVC carriers were 30 mm by 20 mm by 2 mm, and the dimensions of the glass carriers were 30 mm by 20 mm by 1.5 mm. Preparation of the carriers and the experimental procedures were in accordance with the DVV guideline for the evaluation of the virucidal activity of chemical disinfectants on nonporous surfaces (12). Briefly, 50 μl virus suspension in 0.3 mg/ml BSA was applied in the center of each carrier and dried under vacuum in a desiccator for about 45 min. Afterwards, the carriers were transferred into adequate plastic containers prefilled with 0.5 g of glass beads (bead size, 0.25 to 0.5 mm; Carl Roth), which served as an abrasive to increase virus recovery (17). Without mechanical disruption, the dried inoculum was overlaid with 100 μl of 0.25%, 0.5%, or 1% disinfectant and incubated at room temperature for 5 min, 15 min, and 30 min. The disinfectant was applied in a way to ensure complete coverage of the inoculum and to remain as a thin layer on the carrier, as illustrated in OECD guidance document ENV/JM/MONO(2013)11 (9) and in ASTM standard E2197-11 (8). At the end of the exposure times, the virucidal activity was stopped by addition of 900 μl of ice-cold cell culture medium and the container was intensely vortexed to recover residual virus. Subsequently, susceptible cells were inoculated with 10-fold serial dilutions of the eluate to determine residual infectivity. Control reactions were treated in exactly the same way but received WSH instead of disinfectant. For each combination of carrier, disinfectant concentration, and exposure time, three independent biological experiments were performed.

Cytotoxicity determination.

To estimate the cytotoxicity of the disinfectant, the 0.25%, 0.5%, and 1% disinfectant solutions were serially diluted 10-fold with cell culture medium and 100 μl of Vero or MDCK cell suspensions was inoculated with 100 μl of the disinfectant dilution series. In that way, cells were exposed to the same final disinfectant concentrations that were present during the virus inactivation experiments. Cells were observed for cytotoxic effects for the same incubation period that was afterwards applied in the virus infectivity assays. The level of cytotoxicity was determined according to the method of Spearman (18) and Kärber (19).

Interference determination.

To analyze whether glucoprotamin affects the susceptibility of the cells to virus infection, Vero and MDCK cells were pretreated with the nontoxic dilutions of the 0.25%, 0.5%, and 1% disinfectant. After a contact time of 1 h, the disinfectant was replaced by cell culture medium and the cells were inoculated with dilution series of the viruses in cell culture medium. Control cells were treated in exactly the same way but received phosphate-buffered saline instead of disinfectant. Finally, the virus titers on disinfectant-treated and mock-treated cells were determined according to the method of Spearman (18) and Kärber (19). The absence of interference with virus infectivity would be indicated by a ≤0.5-log₁₀-unit difference in virus titers (12).

Neutralization validation.

Cells were inoculated with the dilution series of the viruses prepared in the 0.25%, 0.5%, and 1% dilutions of the disinfectant previously determined to be noncytotoxic and noninterfering. In that way, cells were exposed to the same final disinfectant concentrations that were present during the virus inactivation experiments. Comparative titrations were prepared in cell culture medium, and virus titers were determined according to the method of Spearman (18) and Kärber (19). Successful quenching of virucidal activity would be indicated by a ≤0.5-log₁₀-unit difference in virus titers (12, 16).

Determination of infectivity and statistical analysis.

Viral infectivity was determined by endpoint dilution titration in 96-well microtiter plates. Samples drawn after the exposure times indicated below were immediately serially diluted 10-fold with cell culture medium, and 100 μl was transferred into 100 μl cell suspension, with seven replicates performed per dilution. After 5 to 7 days, plates were analyzed for virus-specific cytopathic effects (CPEs), and virus titers, expressed as the log₁₀ number of 50% tissue culture infective doses (TCID₅₀)/ml, were calculated by the method of Spearman (18) and Kärber (19). The reduction in titer was calculated as the difference between the virus titer after exposure to disinfectant and the control virus titer at the same time. This difference displays the LRF, and the corresponding 95% confidence limits (C_95%) were calculated for each replicate. Finally, the mean LRF of three replicates and its standard deviation (SD) were calculated. For statistical analysis, an unpaired two-tailed t test (assuming previously determined variances) was performed to determine whether the differences observed between two comparison groups were statistically significant.

RESULTS

The cytotoxicity of the glucoprotamin-containing disinfectant was determined by exposure of the detection cell lines to dilution series of 0.25%, 0.5%, and 1% disinfectant. The 1% disinfectant was cytotoxic for both Vero and MDCK cells at dilutions up to and including 1:1,000, and hence, the limit of detection (LOD) was 4.5 log₁₀ TCID₅₀/ml for both VACV and influenza A virus. For the less concentrated disinfectant solutions, cytotoxicity was detected at dilutions up to and including 1:100, resulting in a lower LOD of 3.5 log₁₀ TCID₅₀/ml for both viruses. The dilution with no visible signs of cytotoxicity was considered the noncytotoxic dilution. To exclude the possibility of interference with virus detection, Vero and MDCK cells were pretreated with the noncytotoxic disinfectant dilutions determined as described above, and comparative virus titrations in disinfectant-treated and mock-treated cells were performed. Differences in virus titers determined on mock-treated and disinfectant-treated cells of ≤0.43 log₁₀ TCID₅₀/ml for VACV and ≤0.21 log₁₀ TCID₅₀/ml for influenza A virus proved that the noncytotoxic disinfectant solutions did not affect the susceptibility of the detection cell lines to virus infection. To confirm that neutralization by dilution in cell culture medium was sufficient to render the disinfectant inactive against the test viruses, comparative virus titrations in cell culture medium and in the noncytotoxic and noninterfering disinfectant dilutions determined as described above were performed. The absence of residual virucidal activity was indicated by differences in virus titers determined on mock-treated and disinfectant-treated cells of ≤0.36 log₁₀ TCID₅₀/ml for VACV and ≤0.14 log₁₀ TCID₅₀/ml for influenza A virus. Thus, all controls complied with the general requirement that the differences between the titers of the mock treatment and the disinfectant treatment should not exceed 0.5 log₁₀ TCID₅₀/ml, implicating that the virucidal activity of the disinfectant under the given conditions could be reliably determined (8, 9, 12).

The high antiviral activity of the disinfectant against enveloped viruses was confirmed in quantitative suspension tests. After 5 min of exposure to 0.25% disinfectant, both viruses were already inactivated to levels below the limit of detection. Log₁₀ reduction factors (LRFs) of ≥2.76 and ≥4.33 were achieved for VACV and influenza A virus, respectively (see Table S1 in the supplemental material). Most likely, higher virucidal activity could be shown, but the high LODs of 3.5 and 4.5 log₁₀ TCID₅₀/ml for VACV and influenza A virus, respectively, and the restricted availability of high-titer VACV limited the virus reduction that was demonstrable.

For determination of antiviral activity against viruses attached to surfaces, the test viruses were dried on stainless steel, glass, and PVC carriers. Drying was performed under vacuum for 45 min to ensure uniform test conditions. To estimate the loss of infectivity after drying, the virus spikes were wetted with WSH for 5 min and recovered by intense vortexing in cell culture medium. Virus recovery was evaluated by comparison to the virus load initially applied. On the basis of the titer of the VACV stock suspension, 6.7 log₁₀ TCID₅₀/ml should be recovered after drying. Actual titers ranged from 5.98 to 6.17 log₁₀ TCID₅₀/ml, and hence, drying significantly reduced the infectivity by 0.53 to 0.72 log₁₀ TCID₅₀/ml (Fig. 1A). Any influence of the surface composition on virus recovery was not detected. Influenza A virus was even more susceptible to the drying procedure, and the significant decrease in infectivity ranged from 1.47 to 1.99 log₁₀ TCID₅₀/ml in comparison to the applied viral load. It might be assumed that virus recovery from PVC carriers was slightly higher than that from stainless steel and glass carriers, but the observed differences were not statistically significant (Fig. 1B).

FIG 1.

Recovery of surface-dried viruses from stainless steel, PVC, and glass carriers. VACV (A) and influenza A virus (B) suspensions in 0.3 mg/ml BSA were dried on the nonporous surfaces under vacuum. Spiked carriers were placed in a container with some glass beads, which served as an abrasive, and spotted areas were overlaid with WSH. After 5 min incubation at room temperature, cell culture medium was added and the carriers were intensely vortexed to recover infectious virus. The means and SDs of three independent biological experiments are shown. Black line, expected virus titers. *, significant difference (P = 0.1) at the 90% confidence level; **, significant difference (P = 0.05) at the 95% confidence level.

Next we investigated the virucidal activity of 0.25%, 0.5%, and 1% disinfectant against surface-dried viruses after exposure times of 5 min, 15 min, and 30 min. In contrast to the observations in the suspension assay, 0.25% disinfectant was not able to inactivate VACV completely. After 5 min and 15 min, LRFs were still below 1.5, regardless of the surface. Only after 30 min, virus inactivation exceeded 2 log₁₀ units on the steel and PVC carriers but remained at 1.67 log₁₀ units on the glass carrier (Fig. 2A). Similar findings were obtained with 0.5% disinfectant. No distinct differences in virus inactivation were observed after 5 min and 15 min exposure. After 30 min, the decrease in infectivity was in the range of 2.14 to 2.43 log₁₀ units on all carriers (Fig. 2B). The 1% disinfectant solution was already able to reduce VACV infectivity by approximately 2 log₁₀ units after 5 min exposure. Virus inactivation increased over the exposure time, and after 30 min, LRFs of 2.33 on the stainless steel and glass carriers and 2.71 on the PVC carriers were measured (Fig. 2C).

FIG 2.

Time- and concentration-dependent virucidal activity of glucoprotamin-containing disinfectant against surface-dried viruses. VACV (A to C) and influenza A virus (D to F) suspensions in 0.3 mg/ml BSA were dried on stainless steel, PVC, and glass carriers and exposed to 0.25%, 0.5%, and 1% disinfectant for the indicated exposure periods. Afterwards, cell culture medium was added to stop the inactivation reaction and the carriers were intensely vortexed to recover infectious virus. Control reaction mixtures were treated in exactly the same way but received WSH instead of disinfectant. LRFs were calculated in relation to the results for the corresponding controls. The means and SDs of three independent biological experiments are shown. *, significant difference (P = 0.1) at the 90% confidence level; **, significant difference (P = 0.05) at the 95% confidence level.

Contrary to above-described findings, the surface composition of the test carriers had a strong influence on the antiviral activity of glucoprotamin against influenza A virus. After 5 min exposure to 0.25% disinfectant, more than 2 log₁₀ units of influenza A virus was inactivated on the stainless steel and glass carriers, whereas just a 1-log₁₀-unit decrease in infectivity was detected on the PVC carrier. Even more clearly, complete influenza A virus inactivation was achieved on the glass and steel carriers after 15 min, but inactivation on PVC carriers increased only slightly to 1.19 log₁₀ units; therefore, compared to virus inactivation on the glass and steel carriers, influenza A virus inactivation on the PVC carrier was significantly lower, by 1.1 log₁₀ and 2 log₁₀ units, respectively. Not even after 30 min was complete inactivation achieved on the PVC surface (Fig. 2D). A similar behavior was detected with 0.5% disinfectant. Again, after 5 min and 15 min, high LRFs on the stainless steel and glass carriers were determined, and viruses were completely inactivated after the latter period. However, infectivity on the PVC carriers was decreased by only 1.33 and 1.76 log₁₀ units after 5 min and 15 min, respectively. No remaining infectious virus was recovered from the PVC carriers only after 30 min exposure to 0.5% disinfectant (Fig. 2E). When 1% disinfectant was investigated, the protecting effect of the PVC surface was markedly reduced, and residual virus infectivity was recovered only after the 5-min exposure period. However, compared to virus inactivation on the glass carrier, virus inactivation on PVC was significantly lower by more than 1.2 log₁₀ units. After 15 min, influenza A viruses were completely inactivated on all surfaces (Fig. 2F).

The C_95% of all individual virus titers as well as of all LRFs were ≤0.5 log₁₀ unit. The significance of the differences between the applied and the recovered viral loads and between the LRFs detected on the different surfaces was calculated to prove that the differences that were found were true and not simply due to chance. Statistically significant differences are indicated by asterisks in Fig. 1 and 2.

DISCUSSION

The aim of this study was to test the antiviral activity of a disinfectant that should be used to disinfect areas where mainly influenza viruses are handled. Hence, a disinfectant with proven limited virucidal activity (according to the DVV/RKI [German Federal Authority] recommendations), like the glucoprotamin-based disinfectant investigated in the present study, should be sufficient for this purpose (12). Influenza A virus was chosen as the test virus because it should be inactivated in the practical application of glucoprotamin and therefore represents a relevant test virus. Whenever possible, the relevant virus should be used in virus inactivation studies (20). To comply with the DVV guideline, the well-established model virus VACV was also included in this study. Numerous data on the resistance of VACV to chemical disinfectants are available, and hence, these data should facilitate the comparison of the results obtained in the present with already published findings.

The high levels of antimicrobial and antiviral activity of glucoprotamin were demonstrated in several previous studies and could be proven in a suspension assay (2, 21). To achieve sufficient virus inactivation, the manufacturer suggests that a concentration of 0.25% disinfectant be used for 30 min or a concentration of 0.5% disinfectant be used for 15 min. We confirmed this recommendation and even found complete inactivation of VACV and influenza A virus after 5 min exposure to 0.25% disinfectant. Due to the high levels of cytotoxicity, which were also reported by Dvorakova and colleagues, LRFs higher than 2.76 for VACV and 4.33 for influenza A virus could not be demonstrated (22). However, the main objective of the present study was to determine the antiviral activity of the glucoprotamin-based disinfectant against viruses dried on the inanimate surfaces stainless steel, glass, and PVC. In recovery studies, we demonstrated that viruses remained infectious after drying, although a significant loss of infectivity of up to 0.73 and 1.99 log₁₀ units was detected for VACV and influenza A virus, respectively. The high loss of infectivity upon drying revealed that influenza A virus, in general, is an inappropriate test virus for determination of antiviral activity against viruses dried on surfaces. However, due to its high relevance, we could not exclude influenza A virus from this study.

Terpstra and coworkers performed similar studies and investigated the virucidal activity of several disinfectants against viruses dried on stainless steel surfaces (23). They also found decreased levels of infectivity of the enveloped viruses bovine diarrheal disease virus, HIV, and pseudorabies virus ranging from 0.5 to 2 log₁₀ units after the viruses were dried in a cell culture medium or plasma matrix (23). They also reported that viruses in a dried state were less susceptible to disinfectants. This is in accordance with our findings, and especially, VACV infectivity was only slightly affected by exposure to 0.25% and 0.5% glucoprotamin-based disinfectant.

Two infectious forms of VACV exist, and it remains to be elucidated to what extent the different types of virions contributed to the drastically enhanced resistance of VACV in the dried state against disinfectant concentrated to low to medium levels. The most abundant form of VACV is the intracellular mature virus (IMV), which is retained as a robust, stable virion in so-called virus factories inside the cytoplasm until cell lysis. During VACV morphogenesis, IMV is wrapped by an additional lipid envelope and the extracellular enveloped VACV (EEV) is released from the cell (24, 25). Obviously, IMV and EEV differ in the number and composition of lipid envelopes, suggesting that the two forms differ in their physicochemical properties, like resistance to chemical disinfectants or at least wetting properties. However, drying onto surfaces does not always increase the resistance of viruses to disinfectants. Thevenin and colleagues showed in an interesting study that the nonenveloped feline calicivirus was inactivated up to 4 log₁₀ units in 10 min in the dried state, but inactivation was only moderate when the virus was present in solution (26).

Most surprising was the strong influence of the surface composition of the carriers on the virucidal activity of glucoprotamin against influenza A virus. Apparently, PVC protected surface-dried viruses, and this effect was most evident for disinfectant concentrated to a low level and short incubation times. We found significantly lower LRFs on PVC than on stainless steel and glass carriers after 5 min exposure to 0.25% disinfectant. After 15 min and 30 min exposure to 0.25% disinfectant, the LRFs on PVC were also significantly lower than those on stainless steel. Similar findings were obtained with 0.5% disinfectant, but significant differences compared to the results obtained on stainless steel and glass carriers were observed after only 5 min and 15 min exposure. As expected, increased virucidal activity was observed with increasing disinfectant concentration. Exposure to 1% disinfectant suppressed the protecting effect of PVC, and significant differences between LRFs obtained on PVC and those obtained on the other two carriers, stainless steel and glass, were revealed only for the 5-min exposure. That surface material itself can affect influenza A virus infectivity was reported by Noyce and colleagues (27). They inoculated influenza A virus onto copper and stainless steel coupons and determined the viral infectivity remaining after incubation at room temperature. They found decreased infectivity of nearly 4 log₁₀ units after 6 h of incubation on copper carriers, whereas high infectivity was recovered after incubation on the stainless steel coupons. Inhibition of viral replication by degradation of genomic material or the interaction of copper ions with nucleic acid strands was discussed.

In the present case, virus recovery from stainless steel, glass, and PVC carriers was within a narrow range. Therefore, the possibility that one of the materials exhibited direct antiviral activity can be excluded and a different mechanism must account for the differences in virus inactivation. The microstructure of PVC probably favors influenza A virus stability during drying and, in addition, inhibits penetration of the disinfectant into the deeper layers of the dried virus spike. Furthermore, the dried inoculum must first be dissolved by the applied disinfectant before virucidal compounds can inactivate the virus. The equal activity of 0.25% and 0.5% disinfectant against VACV after 5 min and 15 min was possibly a consequence of the initial dissolving of the dried viruses. Once they were dissolved, the viruses were rapidly inactivated by the disinfectant, as displayed by the similar increases in LRFs after the 30-min exposure period. Enveloped viruses like VACV and influenza A virus can be considered lipophilic due to their additional lipid shell. The lipid envelope makes them easier to inactivate by chemical disinfectants than nonenveloped hydrophilic viruses or intermediate viruses like adenovirus or reovirus, which also lack an envelope but which are more soluble in lipids than other nonenveloped viruses (28). However, enveloped viruses differ in their lipophilic properties, depending on the extent to which hydrophilic proteins are incorporated in their envelope. Differences in the envelope compositions of VACV and influenza A virus possibly accounted for the different inactivating properties of the investigated disinfectant in the low concentration range. Similar to quaternary ammonium compounds, glucoprotamin mainly targets biological membranes and exhibits its antiviral activity by the destruction of membrane integrity (4, 29). To unfold its virucidal activity, the disinfectant must be in contact with fluid lipid membranes, and the preceding drying of the virus might hamper or at least delay the development of antiviral effects. Hence, the wetting and dissolving of contaminants must be taken into account for surface disinfection and disinfection procedures for sensitive instruments. Widmer and Frei analyzed the process of disinfection of surgical instruments by another glucoprotamin-containing disinfectant (30). The devices were first immersed in saline solution, followed by soaking in a glucoprotamin solution for 60 min. Even without mechanical action, they found that it had excellent antimicrobial activity (30). However, our data implicate that dried viruses might resist such a treatment and that disinfection procedures must be carefully validated.

Mostly, more attention is given to the matrix material in which the viruses are present. For antiviral activity testing, animal serum, BSA-erythrocyte mixtures, or a soil load composed of BSA, tryptone, and mucin is usually used to mimic blood or bodily fluids (8, 12). Viruses in protein-containing solutions dried on surfaces are often less susceptible to disinfectants due to the quenching effects of the organic compounds (13, 23). We also found this in other carrier tests with influenza A virus, adenovirus, and hepatitis A virus, in which the protective effect of the soil load was the most evident with oxidizing agents (unpublished data). However, the distinct influence of the surface on virus inactivation could be detected only with the glucoprotamin-based disinfectant. Protection by the surface probably comes into effect only after virus exposure to disinfectants with limited virucidal activity.

Nevertheless, the importance of environmentally compatible disinfectants may not be neglected, considering that high-level virucidal disinfectants are active chemicals and always represent a risk and the potential for harm, especially when applied thoughtlessly or incorrectly (31). Particularly in sensitive areas like hospitals, where only mild disinfection can be applied, special attention should be paid to disinfectants with high wetting capacities and efficient cleaning programs to prevent the survival and transmission of viruses from contaminated surfaces (32). Cleaning practices should also take into account the facts that surface composition may facilitate virus resistance and different surfaces may present different pathogen reservoirs (33). This is of particular importance when medical devices or sensitive equipment for surgical applications is cleaned after use solely by immersion in disinfectant solution rather than by application of mechanical action. Whenever possible, mechanical action should be applied during disinfection to force dissolving of adhered viruses and to enable the access of antiviral active compounds. Finally, not only the matrix where the viruses are present during drying can affect virus stability, but also the surface composition can substantially contribute to virus survival and might protect viruses from inactivation by surface disinfectants.