Surface texture limits transfer of S. aureus, T4 Bacteriophage, Influenza B virus and Human coronavirus

The bead transfer method and Staphylococcus aureus testing

To develop a method for evaluating transfer of microorganism onto surfaces, we applied the following design criteria for the method: 1) is applicable for different types of microorganisms including bacteria, fungi, and viruses; 2) is applicable for a wide variety of surfaces, including three-dimensional surfaces that are commonly found in daily uses; 3) shows acceptable levels of attributes that are critical for standardized disinfectant tests, including reasonableness, relevancy, validity, ruggedness, resemblance, responsiveness, repeatability, and reproducibility (29). It was reasoned that the bead transfer method (Figure 1C) could meet all these requirements and therefore was selected for further development.

For initial testing, the bead transfer method was used to evaluate attachment of Staphylococcus aureus on control (smooth) silicone. Briefly, glass beads were coated with S. aureus inoculum then applied to sterilized silicone surfaces (Figure 1C). After inoculation, tiny droplets are evenly distributed on the surface. The surface was sampled by a Replicate Organism Detection and Counting (RODAC) plate after it is visibly dry, and the amount of bacteria transferred onto the surface was determined by colony forming units (CFUs). To assess resemblance of the method, i.e. the repeatability in control samples, this assay was repeated on two different days by three different operators. A linear mixed-effects model was used to fit the log₁₀ transformed CFUs (logCFU) from all independent assays, and the resemblance repeatability standard deviation (or control repeatability standard deviation, denoted by CSr) was estimated to be 0.283 (Table 2). This value meets the historical acceptance criteria of CS_r ≤ 0.5 (30), suggesting that the bead transfer method has reasonable resemblance for testing S. aureus attachment. Among the factors contributing to CS_r, it was estimated that 46.1% of the variance stems from between-date variation, and 32.1% results from between-operator variation.

Previously, we have demonstrated that an engineered Sharklet® micropattern, SK2×2, effectively reduces bacterial attachment to surfaces (20, 23, 24). To test whether this micropattern is responsive to the bead transfer method, attachment of S. aureus on SK2×2 patterned silicone surface (Figure 1A, Table 1) was measured as described above. Compared to the smooth control, SK2×2 micropattern significantly reduces the amount of S. aureus transferred by beads, as shown by log1oCFU (Figure 2, unpaired two-tailed t-test, p ≤ 0.0001). For each independent assay, log reduction (LR) was calculated by subtracting log10CFU for patterned surface from log10CFU for control surface. Similarly, to account for between-date and between-operator variations, a linear mixed-effects model was used to fit LRs from all assays, in which LR was estimated to be 1.66 ± 0.24 (Table 3). In other words, SK2×2 micropattern results in an average reduction of S. aureus attachment on silicone surfaces by 97.8%. From the model, the repeatability standard deviation (Sr) was estimated to be 0.438 and meets the historical acceptance criteria of S_r ≤ 1.0 (30), suggesting that the bead transfer method has reasonable repeatability for measuring the efficacy of antimicrobial surfaces against S. aureus attachment.

• Download figure
• Open in new tab

Figure 2. SK2×2 micropattern limits transfer of Staphylococcus aureus on silicone surfaces

Smooth or SK2×2 micropatterned silicone samples were tested against Staphylococcus aureus using bead transfer method. The log transformed colony forming units (log1oCFUs) recovered per sample is indicated for each sample. Horizontal lines indicate sample mean. Percent reduction compared to smooth is indicated above the data set. Statistical significance was determined by unpaired two-tailed t test (****, p ≤ 0.0001).

Transfer of Bacteriophage T4

To investigate whether the bead transfer method can be applied for other types of microorganisms, we used this method to measure transfer of bacteriophage T4 on silicone surfaces. Inoculation of the surface was essentially the same as described above. To prevent desiccation which can be detrimental to phage survival, phage was recovered from the surface immediately after inoculation through extensive rinsing, and the amount of phage transferred on to the surface was quantified by plaque assay. Similarly, resemblance of the method for bacteriophage T4 was evaluated by performing the assay on three different days by three different operators. By fitting log transformed plaque forming units (log10PFU) to a linear mixed-effects model, CS_r was estimated to be 0.239, suggesting that the bead transfer method has reasonable resemblance for testing bacteriophage T4 transfer. Interestingly, the fitted model was (near) singular, where both between-date and between-operator variance was (nearly) zero (Table 2). This could be due to less biological variation between experiments, since phages are obligate parasites and do not metabolize by themselves.

To determine whether SK2×2 micropattern is effective against bacteriophage T4, the bead transfer method was repeated on patterned silicone surfaces. Compared to the smooth control, log10PFU were significantly reduced by the SK2×2 micropattern (Figure 3, unpaired two-tailed t-test, p ≤ 0.0001), indicating an inhibitory effect on bacteriophage T4 transfer. Account for between-date and between-operator variations, SK2×2 micropatterns resulted in an LR of 1.15 ± 0.25, or an average reduction of 93.0% in bacteriophage T4 transfer compared to the smooth control. Sr for LRs was estimated to be 0.477 (Table 3), suggesting a reasonable level of repeatability for the assays.

• Download figure
• Open in new tab

Figure 3. SK2×2 micropattern limits transfer of Bacteriophage T4 on silicone surfaces

Smooth or SK2×2 micropatterned silicone samples were tested against Bacteriophage T4 using beads transfer method. The log transformed plaque forming units (Log10PFUs) recovered per sample is indicated for each sample. Horizontal lines indicate sample mean. Percent reduction compared to smooth is indicated above the data set. Statistical significance was determined by unpaired two-tailed t test (****, p ≤ 0.0001).

Transfer of influenza B virus on silicone surface

Virus transfer has been shown to vary among viral species (31-33), largely influenced by surface properties of viral particles (34). In contrast to bacteriophage T4, where capsid proteins directly interact with inanimate surfaces, enveloped viruses have a membrane structure that shields viral particles from the outer environment. To determine whether the bead transfer method can reliably measure transfer of enveloped viruses to silicone surfaces, we applied this method to Influenzavirus B/Taiwan/2/62, an enveloped RNA virus. Independent assays were performed on two different days by three different operators using smooth silicone, and the amount of virus transferred was quantified by 50% tissue culture infective dose (TCID50) method in MDCK cells. Fitting log transformed TCID50 (log10TCID50) to a linear mixed-effects model estimated CS_r to be 0.169, suggesting that the bead transfer method has reasonable resemblance for influenza B virus (IBV). Between-operator variance contributes to 49.9% of CS_r, whereas between-date variance was (nearly) zero (Table 2).

Next, we examined the efficacy of SK micropatterns against transfer of IBV. In addition to SK2×2, we included a new generation SK topography, SK5×3, in our assays. In the SK5×3 micropattern, features are designed to be 4μm high, 5μm wide, spaced 3μm apart, and with length ranging from 10μm to 40μm (Figure 1A, Table 1). Previously, this new micropattern has been shown to limit bacterial touch transfer while exhibiting improved mechanical properties (internal proprietary data). When patterned silicone surfaces were subjected to the bead transfer method, both types of surfaces showed significant decrease in the amount of IBV transfer, as quantified by log10TCID50s (Figure 4A, ordinary one-way ANOVA followed by Tukey’s multiple comparison test, p ≤ 0.001). SK2×2 and SK5×3 micropatterns resulted in LRs of 0.46 ± 0.06 and 0.51 ± 0.11, or average reduction of 65.2% and 69.3% in IBV transfer compared to the smooth surface, respectively (Table 3). There was no significant difference between LRs from SK2×2 and SK5×3 (unpaired two-tailed t-test, p = 0.159). Sr for SK2×2 and SK5×3 was estimated to be 0.161 and 0.300, respectively, suggesting reasonable repeatability for both micropatterns.

• Download figure
• Open in new tab

Figure 4. SK micropattern limits transfer of human viruses on silicone surfaces

Smooth, SK2×2, or SK5×3 micropatterned silicone samples were tested against Influenzavirus B/Taiwan/2/62 (A) or Coronavirus 229E (B) using beads transfer method. The log transformed TCID50 (Log10TCID50) recovered per sample is indicated for each sample. Horizontal lines indicate sample mean. Percent reduction compared to smooth is indicated above the data set. Statistical significance was determined by ordinary one-way ANOVA followed by Tukey’s multiple comparison test (****, p ≤ 0.0001; ***, p ≤ 0.001; ns, not significant, p ≥ 0.05).

Transfer of Coronavirus 229E on Silicone Surface

In light of the emerging pandemic of coronavirus disease 2019 (COVID-19), we were interested in whether SK micropatterns can be used for preventing the spread of coronaviruses. We reason that the bead transfer method can be readily applied for evaluating transfer of coronaviruses, which are enveloped RNA viruses like influenza, on surfaces. As a proof-of-concept study, we used human coronavirus (hCoV) 229E, a surrogate strain for SARS-COV-2 as suggested by ASTM E35.15 (35). Similarly, control (smooth) silicone was independently tested on two different days by two different operators, and surface contamination was quantified by the TCID50 method in MRC-5 cells. The resulting log10TCID50 was fitted into a linear mixed-effects model, which estimates CS_r to be 0.102, confirming that the bead transfer method has reasonable resemblance for hCoV 229E. Between-date variance contributes to 45.7% of CS_r, whereas between-operator variance was (nearly) zero (Table 2).

To determine the efficacy of SK micropatterns against hCoV 229E, SK2×2 and SK5×3 patterned silicone was subject to the bead transfer method alongside with smooth control. As seen with influenza, both micropatterns significantly reduce hCoV transfer to silicone surfaces (Figure 4B ordinary one-way ANOVA followed by Tukey’s multiple comparison test, SK2×2, p ≤ 0.0001; SK5×3, p ≤ 0.001). Accounting for between-date and between-operator variation, SK2×2 micropattern led to an LR of 0.49 ± 0.11, or an average reduction of 67.6% in hCoV transfer; SK5×3 micropattern led to an LR of 0.44 ± 0.11, or an average reduction of 63.5% in hCoV transfer (Table 3). There was no significant difference between LRs from SK2×2 and SK5×3 (unpaired two-tailed t-test, p = 0.960). From the linear mixed models used to fit the LRs, Sr for SK2×2 and SK5×3 was estimated to be 0.195 and 0.092, respectively, indicating reasonable repeatability for both micropatterns. These results were highly reminiscent of those seen with IBV, suggesting that SK micropatterns have similar inhibitory effects towards the transfer of enveloped RNA viruses on surfaces.

Sharklet® polypropylene film

Having established that the bead transfer method is suitable for evaluating surface contamination by human viruses, we applied this method for a commercially available product, Sharklet® polypropylene (PP) film, to assess its efficacy in limiting viral transfer. The PP film was produced by a partner manufacturer with SK2×2 micropattern on the front side, and smooth on the back side (Table 1, Figure 5A). Therefore, the back side of the film was used as the control surface. Six independent assays were performed by two different operators on two different days for both Influenza virus B/Taiwan/2/62 and hCoV 229E. When PP film was challenged with IBV, SK2×2 micropattern resulted in a significant reduction in viral transfer (Figure 5B, unpaired two-tailed t-test, p ≤ 0.001), with an LR of 0.76 ± 0.11, or an average reduction of 82.8%. Similarly, when tested with hCoV, SK2×2 micropatterned PP film showed a significant reduction in viral transfer (Figure 5C, unpaired two-tailed t-test, p ≤ 0.0001) compared to the smooth control, with an LR of 0.83 ± 0.08, or an average reduction of 85.1%.

• Download figure
• Open in new tab

Figure 5. SK micropattern reduces touch transfer of human viruses on Sharklet Shieldpolypropylene film

(A) Representative confocal images of smooth control and SK2×2 polypropylene surfaces. Scale bar, 20μm.

(B-C) Smooth or SK2×2 micropatterned polypropylene film was tested against Influenzavirus B/Taiwan/2/62 (B) or Coronavirus 229E (C) using beads transfer method. The log transformed TCID50 (Log10TCID50) recovered per sample is indicated for each sample. Horizontal lines indicate sample mean. Percent reduction compared to smooth is indicated above the data set. Statistical significance was determined by unpaired two-tailed t test (****, p ≤ 0.0001; ***, p ≤ 0.001).

SK micropattern alters the interaction of surfaces with water

To our surprise, for both IBV and hCoV, SK2×2 micropattern led to a significantly greater reduction in viral transfer on PP film compared to silicone (unpaired two-tailed t-test, IBV, p = 0.0271; hCoV, p = 0.0018). One mechanism of action for Sk micropatterns is through lowering surface energy and therefore wettability of the surface. To investigate the difference in SK micropattern efficacy between different material, we measured water contact angle for smooth and SK2×2 micropatterned silicone and PP film (Figure 6). Interestingly, smooth PP film is significantly more wettable than smooth silicone (ordinary one-way ANOVA followed by Holm-Sidak’s multiple comparison test, p ≤ 0.0001), whereas SK2×2 micropatterned PP film showed similar wettability as SK2×2 micropatterned silicone (p = 0.8751). We reasoned that the greater difference in wettability between smooth and micropatterned PP film may explain the elevated effects of SK2×2 micropattern on PP film compared to silicone.

• Download figure
• Open in new tab

Figure 6. Effects of SK micropatterns on surface wettability

(A) Representative images of water drops captured by a goniometer on smooth and SK2×2 silicone or polypropylene surfaces. Water contact angles are reported belowimagesby mean ± SEM (silicone, n = 8; smooth polypropylene n = 16; SK2×2 polypropylene, n = 22).

(B) Comparison of water contact angle on different surfaces, error bars represent SEM. Statistical significance was determined by ordinary one-way ANOVA followed by Holm-Sidak’s multiple comparison test (****, p ≤ 0.0001; ns, not significant, p ≥ 0.05, p > 0.05).

Silicone sample generation

Silicone samples used for testing were casted inside a clean room using Elastosil M 4641 (Wacker). Elastosil part A was mixed with Elastosil part B at a 10:1 ratio by weight in a SpeedMixer DAC600.2 VAC-LR (FlackTek, Inc.) for one minute under pressure following the manufacturer’s recommendation. Mixed silicone was poured onto a nickel master with desired topography (smooth or SK patterned) and covered with a glass plate to allow even spreading across the master. The thickness of silicone samples was controlled by 1mm-thick spacers taped around the nickel master. The casting apparatus was incubated at 65°C with 20lb weights on top for 2 hours for silicone to cure. Silicone samples were removed from the master and cut into 50mm x 50mm squares using a razor blade. The fidelity of surface topography on silicone samples was verified by confocal microscopy (LEXT 0LS4000, Olympus, Figure 1B). A summary of pattern dimensions is included in Table 1. The silicone samples were submerged in 70% ethanol for 10 seconds and placed pattern side up onto a 100mm x 15mm petri dish. Samples were UV sterilized for 15 minutes before used in experiments.

Contact Angle Measurement

Three-phase water contact angle for each silicone surface was measured using a Rame-hart contact angle goniometer (Model 250). Surfaces were rinsed with 95% ethanol and air dried completely before measurement. A 10uL water droplet was deposited onto the surface using a micropipette, and the image of the drop was recorded and analyzed using the built-in DROPimage Advanced software. For micropatterned surfaces, the length of SK features was placed along the light path. Measurement was repeated 4-11 times with each type of surface, and each measurement gives two readings for contact angle, from the left and right sides of the water drop. A total of 8-22 contact angle readings were collected for each type of surface and used for data analysis in Figure 6.

Strains and media

Staphylococcus aureus (ATCC 6538) was grown in tryptic soy broth (TSB, Criterion C7141) at 37°C and 280 rpm for 16 hours. Before testing, bacteria were pelleted by centrifugation at 250 x g for 5 minutes. Supernatant was discarded and S. aureus was resuspended in phosphate buffer saline (PBS, HyClone SH30258.02). OD600 of bacterial suspension was measured using a BioWave CO8000 (BioChrom), and theoretical CFU per mL was calculated using empirically determined OD600-CFU relationship under identical growth condition (1 OD600 = 5.065 x 108 CFU/mL). S. aureus was diluted to desired concentration (5 x 103 CFU/mL) using PBS and was used within 30 minutes of preparation. For each independent assay, the inoculum was serially diluted and plated on tryptic soy agar to confirm its concentration falls within acceptable range (2.5 x 103 - 1 x104 CFU/mL).

Escherichia coli B strain (ATCC 11303) was grown in Luria-Bertani (LB) broth (Lennox, Sigma Aldrich L3022) at 37°C and 280 rpm for 16 hours before phage propagation and plaque assays. MDCK cells (ATCC CCL-34) and MRC-5 cells (ATCC cCl-171) were maintained in cell culture media [Dulbecco’s Modified Eagle Medium (DMEM, Gibco 11995) supplemented with 100 U/mL penicillin, 100 μg/mL streptomycin (Gibco 15140-122), and 10% fetal bovine serum (FBS, Gibco 10082-147)] at 37°C under 5% CO₂.

Phage and virus propagation

Bacteriophage T4 (ATCC 11303-B4) was prepared as described before (38). E. coli B strain was subcultured in LB broth supplemented with 1mM CaCl2 and MgCl2 at 37°C and 280 rpm until mid-log phase (OD600 = 0.4-0.8). Bacteriophage T4 was added to the E. coli culture at an MOI of 0.003, and incubated with agitation for about five hours until the lysate is visibly clear. Phage lysate was clarified by centrifugation at 250 x g for 30 minutes followed by passing through a 0.22μm syringe filter (Millipore SLGV033RS). The resulting phage stock was quantified by plaque assay (see below) and stored at 4°C for up to two months before use. Phage was diluted to desired concentration (2.0 x 106 PFU/mL) using SM buffer (50mM Tris-HCl, pH 7.5, 100mM NaCl, 8mM MgSO4) before testing and was used within 30 minutes of preparation. For each independent assay, the inoculum was serially diluted and quantified by plaque assay to confirm its concentration falls within acceptable range (1.0 - 5.0 x 106 PFU/mL).

Influenza B virus (IBV, ATCC VR-295) was propagated essentially as described (39). MDCK cells were grown in a T175 until 80-90% confluence. Growth media was aspirated off and cells were washed three times with PBS before infected with 2mL IBV in infectious PBS [iPBS, PBS supplemented with 0.3% bovine serum albumin (BSA, Fisher BP1600), 100 U/mL penicillin, 100 μg/mL streptomycin, 2 mM MgCl2, and 1 mM CaCl2] at an MOI of 0.01. Virus was absorbed for 60 minutes at 37°C, shaking every 15 minutes. At the end of adsorption, viruses were aspirated off, and cells were washed three times with influenza infection media [IM-flu, DMEM supplemented with 100 U/mL penicillin, 100 μg/mL streptomycin, 0.3% BSA, 0.1% FBS, 20mM N-2-hydroxyethylpiperazine-N-2-ethane sulfonic acid (HEPES, Gibco 15630-080), and 1 μg/mL TPCK-treated trypsin (Sigma Aldrich T1426)]. IBV infected cells were incubated at 37°C under 5% CO2 for 48-72 hours. When 90% of cells were detached from the flask, the supernatant containing IBV was harvested and clarified at 500 x g for 5 minutes. The supernatant was aliquoted and stored at −80°C, and the titer of IBV was determined by TCID50 assay in MDCK cells (see below). Virus stocks were diluted to approximately 1.0 x 106 PFU/mL. IBV was thawed on ice and used undiluted for testing.

Human coronavirus 229E (ATCC VR-740) was propagated in MRC-5 cells similarly as for IBV. Cells were infected with 2mL coronavirus 229E in infectious DMEM (iDMEM, DMEM supplemented with 100 U/mL penicillin, 100 μg/mL streptomycin, 0.2% BSA) at an MOI of 0.01. After an hour of adsorption, hCoV was removed and CoV infection media (IM-CoV, DMEM supplemented with 100 U/mL penicillin, 100 μg/mL streptomycin, and 2% FBS) was added.

Once cytopathic effects (CPE) reach 50% or more, cell culture flask was frozen at −80°C for an hour and partially thawed in a 37°C water bath. Freeze thaw was repeated two more times to facilitate cell lysis and virus release. After the last thawing step, cell lysate was clarified at 3,000 x g for 10 minutes at 4°C. The supernatant was aliquoted and stored at −80°C, and the titer of coronavirus 229E was determined by TCID50 assay in MRC-5 cells (see below). Virus stocks were diluted to approximately 1.0 x 105 PFU/mL. Coronavirus was thawed on ice and used undiluted for testing.

Bead transfer method and microbial recovery from testing surfaces

For each testing sample, 15g of 3mm soda-lime glass beads (Walter Stern 100C) were dispensed into a sterile sample cup (Thermo Scientific 021038). Immediately before inoculation, an appropriate amount of inoculum (1mL for bacteria or phage, 500μL for human viruses) was added to the beads (Figure 1A). The sample cup was capped with a lid and shaken by hand vigorously in a circular motion for 10 seconds for even dispersal of inoculum on glass beads.

The lid was removed, and a testing sample was placed face down over the sample cup opening. The testing sample was held firmly against the sample cup by hand while inverted three times to allow inoculation of the surface by glass beads.

For bacteria recovery, the inoculated surface was air dried completely at room temperature for 5 minutes. RODAC plate was stamped onto the surface for 5 seconds, applying about 400g of pressure ensuring there are no bubbles trapped between the two surfaces. RODAC plates were incubated at 37°C overnight. Plates that contain between 30-300 individual colonies were counted to determine CFU per sample.

For virus recovery, 5mL SM buffer (phage) or 2mL Infection Media (IM, human viruses) were immediately pipetted onto the inoculated sample and over the surface three times bottom to top in each of the four directions. Viruses recovered from each testing surface were quantified by plaque assay (phage) or TCID50 assay (human viruses).

Phage plaque assay

Phage recovered from testing samples were diluted in LB broth for countable plaques. 1mL recovered phage at proper dilution (neat, 10x, or 100x), 1mL E.coli B strain overnight culture, and 3mL molten 0.7% LB Agar were combined in a 15 mL sterile conical tube, and the tube was inverted once slowly to mix without generating any bubbles. The mixture was carefully poured onto an 1% LB agar plate and evenly distributed by tilting. The plate was allowed to sit undisturbed at room temperature until the agar has solidified (about five minutes) and incubated overnight at 37°C to allow plaque formation. Plates that contain 30-300 individual plaques were counted to determine PFU/mL for each sample:

TCID50 assay

Cells compatible with viruses to be quantified were grown to ~90% confluency in a T175. On the day of the experiment, cells were washed once with PBS and detached from the flask by incubating with 2mL Trypsin-EDTA (HyClone, SH3023602) at 37°C under 5% CO2. Cells were diluted to 50,000 cells/mL in cell culture media and aliquoted into 96-well plates at 100μL per well. Cells were incubated for 1-2 hours at 37°C or until cells were visually adhered to the bottom before use. Virus samples were serially diluted in cold IM in 24-well plates. IBV was diluted 10, 30, 90, 270, 810, 2430, 7290, and 21870-fold; CoV 229E was diluted 3, 9, 27, 81, 243, 729, and 2187-fold. Cell culture media was removed from the 96-well plates and cells were washed once with 100μL IM before 100μL diluted viruses were added. 8-12 wells were used for each virus dilution. For each 96-well plate, at least one well was used as uninfected control, where 100μL IM was added instead of viruses. Cells were incubated at 37°C under 5% CO2 for 4-5 days or until CPE became stagnant. At the end of incubation, cells were fixed with 4% paraformaldehyde in PBS (Thermo Scientific J19943K2) and stained with 0.2% (w/v) crystal violet (Sigma-Aldrich V5265) in 20% methanol. CPE was recorded for each well, and TCID50 was calculated using Reed-Muench method (40).

Statistical Analysis

A linear mixed-effects model was used to fit the log transformed counts (log₁₀CFUs, log₁₀PFUs, or log10TCID50) or log reductions (LRs) from all independent assays performed on each species. Random effects included in the model were date of experiment and operator performing the assay. Therefore, the mixed-effects model estimated the between-date variance (V_D), between-operator variance (V_O), and within-date and operator variance (residual error, V_R). The model can be expressed as, where y_ijk represents the experimental readouts (log₁₀CFUs, log₁₀PFUs, log₁₀TCID50, or LRs) from the ith date, the j_th operator, and the k_th assay, μ represents the mean readout over all experiments and operators, Di represents the between-date variance, Oj represents the between-operator variance, and ε_ijk represents the within-experiment and operator variance (residual error). All linear mixed models were fit in RStudio v.1.2.5042 using R v.3.6.3 (41) and package lme4 (42). Residual, scale-location, and quantile-quantile plots were used to check linearity, homoscedasticity, and normality assumptions of the linear mixed-effects model. The repeatability standard deviation (RSD) for each species tested was calculated by:

Comparison between different surfaces (Figure 2-6) were performed in Prism 8 (GraphPad) using the following workflow: 1) outliers were identified using ROUT method (Q = 1%) and excluded from data analysis; 2) normality of the sample distribution was tested at significance level of 0.05 using Anderson-Darling test, D’Agostino-Pearson omnibus normality test, Shapiro-Wilk normality test, and Kolmogorov-Smirnov normality test with Dallal-Wikinson-Lillie for P value; 3) since all sample distributions reported in this study passed all normality tests, t test (two samples) or ordinary one-way ANOVA (three or more samples) and multiple comparison test recommended by the software was used for comparison.