The effects of Staphylococcus aureus on the β-lactamase enzymes and virulence factors of Pseudomonas aeruginosa in lung disease

Background the determine the statistically changes the adjusted p-values. All the presented were for multiple comparisons. Gene expression analysis was performed using the REST® (version The ΔΔCt method was used to determine the expression levels.


Background
Lung infections are usually colonized by several microorganisms that interact with each other constituting complex polymicrobial communities [1,2]. Two of the organisms that are frequently coisolated from this type of infection are Pseudomonas aeruginosa and Staphylococcus aureus [3]. In vitro studies suggest that P. aeruginosa prospers better than S. aureus by multiplying faster and acting as its antagonist [4]. This might be related to the fact that P. aeruginosa seems to induce the bio lm production of S. aureus [5,6]. Bio lms are structured communities that secrete and encase themselves within a protective matrix [1]. In addition, the presence of both pathogens is a sign of poor prognosis for the patients. Evidence shows that their virulence increases and antibiotic treatments become less e cient during co-infections [6,7].
As a matter of fact, P. aeruginosa inhibits the electron chain transport of respiration using 4-hydroxy-2heptylquinoline-N-oxide, kills S. aureus to acquire iron using the LasA protease, and produces rhamnolipids to disperse the S. aureus bio lm [8,9]. S. aureus, on the other hand, increases its resistance to the inhibition of respiration by forming small colony variants and ferments carbon sources to lactic acid as the end product [10]. P. aeruginosa establishes a bio lm within the lung of the susceptible patient coupled with the overproduction of the exopolysaccharide alginate [11,12]. Moreover, the pathogenicity and antibiotic resistance of P. aeruginosa increases in polymicrobial infections [13]. Two of the microorganisms that are frequently co-isolated from this type of lung infection are Staphylococcus aureus and P. aeruginosa although they have a competitive relationship [14]. Nevertheless, causative microbial agents vary in the different levels of co-existing with mild infections generally being considered as monomicrobial and moderate and severe infections being considered as polymicrobial. Various studies have demonstrated that the dual-microbial infection of P. aeruginosa and S. aureus is more virulent and resistant than those of single species.
Using cell culture as an in vivo condition is one of the best methods for assessing the performance of P. aeruginosa and S. aureus in co-culture [14]. Co-culture models are a step towards bridging the gap between in vitro models and in vivo systems [5,8]. Furthermore, accessory cells or support cells may aid in target cell growth, differentiation, and resistance to challenge. The cell line utilized in the present study is the widely used A549 type II alveolar epithelial cell model [1,15]. The A549 cell line is a transformed variant of type II alveolar epithelia commonly used as a model of toxicology and drug delivery in the pulmonary epithelium [15]. Moreover, resistance to beta-lactams increases in the co-infections of S. aureus and P. aeruginosa. Furthermore, some virulence factors of P. aeruginosa (such as bio lm) play a more bene cial role in enhancing beta-lactamase resistance [16].
In this study, the resistance to beta-lactam and the pathogenicity of P. aeruginosa were evaluated in an in vivo model using the A549 respiratory cell line in co-culture with S. aureus.

Long-term competition on the A549 cell line
The viability of P. aeruginosa in bio lm and planktonic co-culture with S. aureus was monitored by plate counts. At the beginning of the experiment, the relative abundance of the P. aeruginosa strains was slightly in favor of PAO1. In the bio lm state of co-culture, the survival of PA-2 and PA-3 was less than that of PA-4 ( Figure 1, A1 and A3). Moreover, in the planktonic form, the PA-4 strain survived longer (Figure 1, A2 and A4). In the bio lm state of co-culture, the PAO1 and PA-4 strains had a higher cell density than the PA-2 and PA-3 strains at 48 hours of growth ( Figure 1).
Antibiotic susceptibility pattern of the recovered P. aeruginosa strains Figure 2 illustrates the antibiotic patterns of the P. aeruginosa strains. Resistance to all antibiotics increased after co-culture in the bio lm and planktonic conditions. Antibiotic susceptibility decreased more e ciently in the planktonic form compared with the bio lm one. Antibiotic resistance to meropenem and imipenem was more effectively increased in the beta-lactamase-producing strain compared with the bio lm and toxin producers. After two days, resistance to imipenem increased in the toxin-producing strain (PA-3) so that the MIC of imipenem ranged from 8 μg/mL to 3 μg/mL. Besides, in the MDR strain, increased resistance to meropenem was more than the other antibiotics. In addition, a decrease of imipenem, meropenem, and doripenem zone inhibitions was more observed in the pathogenic strains compared with the MDR strain. (Figure 2).

Virulence factors of the recovered P. aeruginosa strains
The co-culture of PAO1 and SA-1 showed no change in bio lm production after six days ( Figure 3A). The LasA and LasB elastases in the PA-2 and PA-3 strains were enhanced after two days. In the PA-4 strain, the amounts of LasA and LasB elastases decreased after three days ( Figure 3B). In the A549 cell line, pyocyanin production was increased in the co-cultures of PA-2 and SA-1 and PA-3 and SA-1 strains.
Moreover, the highest amount of pyocyanin was produced on the fth and sixth days. In the PAO1 strain, pyocyanin production decreased ( Figure 3D). The production of pyoverdine showed a signi cant change on day six in the A549 cell co-culture conditions in all the P. aeruginosa strains (except for PA-4). Nevertheless, after six days, the PA-4 strain showed a little change in the production of pyoverdine ( Figure  3E).
Gene expression of P. aeruginosa after co-culture in different states After the bio lm co-culture of P. aeruginosa and S. aureus, the activities of lasR and lasI were up-regulated in P. aeruginosa in response to the presence of S. aureus (Figure 4Ac, Bc). The expression level of the lasR gene was higher in the co-culture of the PA-2 and SA-1 strains than that of the PA-4 and SA-1 strains ( Figure 4Ab). However, the expression of the lasR gene in the planktonic state was lower than that of the bio lm form at different times ( Figure 4Aa). The up-regulations of lasI and algD were observed in the bio lm co-cultures of PA-2 and SA-1 and PA-3 and SA-1 strains. Hence, the SA-1 strain had the greatest effect on the expressions of lasI gene (Figure 4Ba and 4Ca). However, the activity levels of lasI were more remarkable in the virulent strains than the KPC-producing strain (Figure 4Bb). The results of this experiment revealed that the SA-1 strain induced more virulence genes in the PA-2, PA-3, and PA-4 strains than the PAO1 strain (Figure 4Cc-Dc). The down-regulation of the KPC gene was observed in the bio lm and planktonic co-culture. Furthermore, the activity of the mexR gene did not change signi cantly in the different strains of P. aeruginosa (Figure 4Da).

Analysis of the data
The chi-square test showed a statistical signi cance between the numbers of mono-cultured and cocultured bacteria (p < 0.001). According to the Student's t-test and χ2 test analysis (Figure 1), the bio lm co-culture containing SA-1+PAO1, SA-1+PA-2, and SA-1+PA-3 signi cantly decreased viability ((P:0.0004), (P0.0006), and (P:0.0009), respectively). However, after analyzing the t-test results, there was no signi cant difference between the viability results of the PA-4 mono-culture and the SA-1+PA-4 co-culture (P:0.059), indicating that no synergism occurred. All the other co-culture combinations had a strong and impressive effect on viability.
According to the Student's t-test and χ2 test analysis (Figure 2), the production of carbapenemase in the combinations containing SA-1+PA-2 and SA-1+PA-3 declined considerably. However, after analyzing the ttest results, there was a notable difference between the observed and expected results, indicating that SA-1 had blocked the antibiotic resistance activity in all the P. aeruginosa strains. Furthermore, all the other co-culture combinations had a remarkable effect on tobramycin resistance. The Tukey analysis found the same results. Figure 2 also shows signi cant t-test results for co-cultures containing SA-1+PA-4. However, these results demonstrated a signi cant positive difference, indicating that antagonism occurred while SA-1 was in the co-culture. The Tukey analysis found the same results.
As shown in Figure 4, the co-culture containing S. aureus+P. aeruginosa produced signi cant t-test results so that S. aureus remarkably suppressed resistance activity in the bio lm model (P<Q0.0001). According to the Tukey results, the S. aureus+P. aeruginosa combination signi cantly increased bio lm production and virulence activity in the bio lm co-culture of P. aeruginosa. However, S. aureus in the planktonic coculture had no signi cant effect on the virulence activity of P. aeruginosa. Figure 4 demonstrates that there was a signi cant positive difference between the observed and expected resistance to carbapenem antibiotics in the S. aureus+P. aeruginosa combination (P=0.0004). This indicates that the increase in the KPC expression was because of S. aureus. The one-away ANOVA analysis showed that each bio lm coculture had a signi cant effect on the antibiotic resistance and production of the virulence factors of P. aeruginosa. The results suggested that there was a strongly signi cant difference between the virulence gene expressions of S. aureus and P. aeruginosa (P< 0.0006).

Discussion
The scientists who study the virulence of P. aeruginosa usually focus on its virulence factors to describe its pathogenic potential. However, the synthesis of these virulence factors requires a functional metabolism that furnishes the energy necessary for the lifestyle and virulence of this microorganism. This study showed that the virulent strains (PA-2 and PA-3) co-cultured with SA-1 in planktonic and bio lm conditions survived less than the PA-4+SA-1 strains (MDR). However, the PAO1 strain exhibited better survival than the PA-4 strain. Meanwhile, the P. aeruginosa populations in the bio lm co-culture were more than those of the planktonic form. Thus, SA-1 has a much more notable inhibitory effect on the PA-2 and PA-3 strains in the A549 cell culture. Our ndings are in agreement with the studies of Alves et al. [16] and Hotterbeekx et al. [13] who suggested that S. aureus had a more inhibitory effect on the pathogenic strains of P. aeruginosa. Because of the depletion of oxygen and the suppression of growth in polymicrobial populations, P. aeruginosa changes its metabolic pathway for its survival and pathogenicity. Frapwell et al. [10] and Ali Mirani et al. [11] demonstrated that in the bio lm co-culture of P. aeruginosa with S. aureus, the viability of P. aeruginosa was reduced after three days which was the most important reason for the type of metabolic pathway and the occurrence of genetic mutations in these bacteria. According to Orazi and O'Toole [17], the interaction between P. aeruginosa and S. aureus in coculture conditions alters the metabolic pathway of P. aeruginosa and the bacterium shifts to fermentative growth and reduced antibiotic resistance. Although the bio lm-producing strain (PA-2) was more inhibited by S. aureus than most other P. aeruginosa strains, it remains to be seen how its pathogenicity and resistance changed.
Our observations indicated that the SA-1 strain had an inhibitory effect on the antibiotic resistance of PA-2 and PA-3 strains, while such effect was observed slightly less in the MDR strain of P. aeruginosa (PA-4). Increased resistance to tobramycin in the bio lm-producing strains of P. aeruginosa in co-culture with S. aureus was reported in the study of Beaudoin et al. [3]. Unlike the other strains, the PA-2 strain demonstrated an increase in tobramycin resistance. Furthermore, the antibiotic inhibition of P. aeruginosa strains in the planktonic state was more than that of the bio lm co-culture, although this sensitivity was lower in the MDR strains as expected based on the studies of Chan et al. [18] and DeLeon et al. [14] on the co-culture of P. aeruginosa with S. aureus. Yang et al. [5] also suggested that bio lm formation may be a bene cial survival characteristic in the co-culture. The bio lm can physically protect bacteria from antimicrobial agents particularly those with large polar molecules. This study demonstrated that, after six days, resistance to imipenem decreased in the PA-2 and PA-3 strains so that the MIC of meropenem ranged from 8 μg/mL to 4 μg/mL. Besides, increased resistance to meropenem in the MDR strain was more than those of the other antibiotics. Moody [19], Frapwell et al. [10], and Tognon et al. [9] showed that S. aureus reduced the antibiotic resistance of P. aeruginosa in co-culture conditions. Therefore, there was a signi cant correlation between SA-1 and the viability of P. aeruginosa strains (p:0.0001, p:0.0002, and p:0.0001 for PA-2, PA-3 and PA-4, respectively) so that the S. aureus co-culture had the lowest effect on the virulence factors and the highest impact on the antibiotic resistance of P. aeruginosa. Our ndings are in agreement with those of Frapwell et al. [10]. P. aeruginosa is a denitrifying bacterium and thus possesses the ability to convert nitrate (NO3-) into nitrogen gas (N2). In response to S. aureus, P. aeruginosa differentially expresses nitrogen metabolism enzymes. Increasing nitrogen has an inducible effect on the pathogenesis of P. aeruginosa and augments the production of such factors as elastase, pyocyanin, bio lm, and toxins. Moreover, P. aeruginosa has many quorum-regulated virulence factors including elastase, pyocyanin, pyoverdine, and elastase. The studies by Kim et al. [20], Radlinski et al. [21], and Yang et al. [5] con rmed that the virulence of P. aeruginosa was altered by interaction with S. aureus. Based on the present study, the co-culture of PAO1+SA-1 showed no change in bio lm production after six days. However, SA-1 augmented LasA and LasB elastases in the PA-2 and PA-3 strains after two days. An enhanced production of pyocyanin in the PA-2+SA-1 and PA-3+SA-1 co-cultures was con rmed. Furthermore, S. aureus reduced the production of pyoverdine in all strains of P. aeruginosa after three days except for PA-4. Abisado et al. [22] and Yang et al. [5] demonstrated that S. aureus had a more signi cant inhibitory effect on the virulence and bio lm production of P. aeruginosa in an in vivo co-culture. Koley et al. [23] recently showed that pyocyanin created a redox potential gradient in the bio lm called 'electro line' which increased iron availability which is essential for the development of bio lm. In the interaction of bacteria, S. aureus acts as a potent iron supplier for P. aeruginosa. Iron plays a crucial role in enhancing the Pseudomonas quinolone signal (PQS) activity of P. aeruginosa. Moreover, it induces bio lm production and quorum sensing (QS) [24]. This change provides the basis for the increased pathogenicity of P. aeruginosa in fermentative conditions. Another important issue addressed in the current study (also con rmed by the studies of Armbruster et al. [7], and Alves et al. [16] is that the production of pyoverdine and pyocyanin increased after the co-culture of the PA-3 and SA-1 strains. According to the study of Hotterbeekx et al. [13], there was a signi cant relationship between the increased pathogenicity of P. aeruginosa and the effects of S. aureus. They found that in polymicrobial infections, S. aureus increased the virulence factors of P. aeruginosa, which con rms our results.
In the current study, some differences were observed in the virulence factors and the KPC expression was signi cant among PAO1, PA-2, PA-3, and PA-4. In the bio lm state of co-culture, the expression of lasR was escalated in the PA-2+SA-1 strains. However, the expression of this gene decreased in PA-4+SA-1. The expression of the lasR gene in the planktonic state was lower than that of the bio lm form at different times. The increased expressions of the lasI gene in the co-cultures of PA-2+SA-1, PA-3+SA-1, and PA-4+SA-1 on the fourth and fth days showed the most prominent effect on the ΔCTlasI gene. Woods et al. [8] observed that the expressions of the lasR and lasI genes of P. aeruginosa in the co-culture with S. aureus increased. These genes play a signi cant role in controlling QS in bacteria. With the increase of QS, the expression of antibiotic resistance genes decreases. In this study, it was demonstrated that the levels of lasI gene in bio lm-and toxin-producing strains were more pronounced than that of the carbapenem-resistant strain. The algD gene was highly expressed in the co-cultures of the PA-2+SA-1 and PA-3+SA-1 strains, however, no increase was observed in the PA-4+SA-1 co-culture. The PA-3+SA-1 strain was also not expressed on the sixth day. Similar results were reported by Limoli et al. 25] and Tognon et al. [9]. At rst, the KPC gene expression in the bio lm and planktonic co-culture was reported. The fold change of KPC was increased in all the strains of P. aeruginosa and the highest activity of this gene was for the PA-4+SA-1 strain. In other words, all the strains showed the increased activity of carbapenemase from the third day onward. DeLeon et al. [14] used the term 'mutual bene t' for P. aeruginosa and S. aureus in cell culture conditions. Although these bacteria inhibit each other's viability, pathogenicity, and antibiotic resistance, with some genetic changes, the surviving population becomes more infectious and more resistant.

Limitation
A limitation of the current study is that many variables can affect the growth of the bacterium in the A549 cell line. Therefore, it is suggested that in future studies, the nutritional status of the cell be taken into account when evaluating QS-based gene expression. QS-based gene regulation models based on the studies of planktonic cells must be modi ed to explain the behavior of bio lm gene expression since gene expression in bio lms is dynamic. In addition, determining the physiological differences between bio lm and planktonic cultures is critical for understanding P. aeruginosa infections (such as those found in the cystic brosis lung) or for removing problematic bio lms from tissue infections.

Conclusion
Our ndings demonstrated that an initial foundation is needed to explain how factors other than cell density can control the expressions of quorum sensing-regulated genes and carbapenemase genes. Since bio lm formation, toxicity, and carbapenem resistance cause the up-or down-regulation of quorum sensing regulated genes (lasR/lasI), it is not inconceivable that globally regulated genes can be controlled by more than one factor. Even though this conclusion is novel, it is not surprising. Furthermore, in the coculture in the A549 cell line, a signi cant relationship was observed among the viability of P. aeruginosa, the activity of pathogenic enzymes, incubation time, resistance to carbapenem, and the expression of virulence genes. Carbapenemase enzymes played a more critical role than pathogenic enzymes in maintaining bacterial growth. Hence, in respiratory infections, resistance to carbapenem antibiotics in P. aeruginosa can provide a basis for the development and spread of co-infection with S. aureus. Besides, the production of pathogenic enzymes and bio lms by P. aeruginosa changes the metabolic pathways of the bacteria and causes the emergence of pathogenic strains.

Materials And Methods
Preparation of the standard strains Some standard strains including P. aeruginosa PAO1, S. aureus ATCC25923, P. aeruginosa NCTC13359 (a strong bio lm-former and KPC-producing strain), P. aeruginosa NCTC13618 (a toxin and KPC-producing strain), and P. aeruginosa NCTC 13620 (a KPC-producing strain) were used in this study. All the strains were derived from clinical isolates and incubated at 37 °C unless described otherwise (200 rpm). Trypticase Soy Broth (supplemented by 1% agar when needed) was applied to culture the bacterial strains. Mannitol salt agar and cetrimide agar were used to recover S. aureus and P. aeruginosa, respectively.

DNA extraction and Sanger sequencing
The bacterial DNA was extracted using the QIAamp® DNA Mini Kit (QiaGene Inc., Chatsworth, Calif., USA), following the manufacturer's instructions. To determine the sensitivity of the reaction, 100 μL of the serially diluted S. aureus reference strain was used during DNA extraction. The ampli cation of 16S rDNA was performed as follows. The bacterial DNA was ampli ed using three sets of primers including mutL, aroC, and rpoD, [26,27]synthesized by Pishgam Research, Pte. Ltd., Iran (Table 1).
All the reactions were performed in a 25 µl volume containing 1 µl of 10 pmol of each primer, 12.5 µl 2X master mix (Ampliqon, Denmark), 2 µl DNA template, and 8.5 µl of deionized water (Sigma-Aldrich, USA). The ampli cation of the genes was performed using a C1001 Touch Thermal Cycler (BioRad, USA) with the following thermocycling program: initial denaturation was done at 95 °C for 5 minutes followed by 35 cycles of denaturation at 95 °C for 1 minute, annealing at 57 °C for 1 minute, extension at 72 °C for 1 minute, and a nal extension at 72 °C for 5 minutes [10]. This study was approved by the Ethics Committee of Hamadan University of Medical Sciences (No: IRUMSHA. REC. 1396.694). The PCR products were sequenced using the service of Sequiserve company (Pishgam Co., Tehran, Iran). Using the Chromas sequence analysis software (version 2.6.5, Technelysium Pty Ltd., South Brisbane, AU), the sequencing results were analyzed by employing the NCBI (https://blast.ncbi.nlm.nih.gov/Blast.cgi) database (BLAST).
Co-culture of P. aeruginosa and S. aureus on the A549 cell line The co-culture assays were done as described by the study of Anderson et al. [28]. Brie y, the bacterial strains were cultured on a TSB medium at 37 °C overnight and then centrifuged at 10000g for ve minutes. The pellets were suspended in a minimal essential medium (MEM) supplemented by Lglutamine and the OD 600 was adjusted to 0.1. 250 µl of each bacterial suspension was added to the A549 monolayer and incubated at 37 °C and 5% CO 2 . In one-and ve-hour time intervals after incubation, the supernatants were removed and replaced by fresh MEM+L-Glu. The plates were incubated for an additional six days and then the supernatants were collected, serially diluted in PBS, and plated on MSA and CA for eighteen hours to count CFU/mL. The established bio lms were treated with 0.1% Triton X-100 in PBS and shaken vigorously for thirty minutes. All the tests were done in triplicate.
Bio lm assay of the recovered P. aeruginosa strains A 500-µl sample of an overnight stationary phase broth culture was diluted 1:100 in a fresh and sterile broth media which was then grown to mid-exponential phase (00 600 0.7-1.0) at 37 ºC. 200 µl of this culture was pipetted into each well of a 96-well microtiter plate which was then incubated for four hours at 37 ºC. After incubation, the contents of the wells were gently aspirated. Each well was washed three times with 200 µl of sterile phosphate-buffered saline. 200 µl of safranin-O dye was then pipetted into each well to stain any resultant bio lm and the wells were then rinsed out with tap water. The plate was then dried in an incubator. Next, 200 µl of 70% ethanol was pipetted into each well and the plate was placed on a shaker at 100 revolutions per minute for fteen minutes. The resultant solution in the microtiter plate wells was then read using a plate reader and the results were recorded [5]. All tests were done in triplicate.
Determination of the minimal inhibitory concentration of the recovered P. aeruginosa strains Antibiotic disks (MAST, UK) and E-tests (Lio lchem, Italy) were used to examine the susceptibility of the bacteria to several antibiotics before and after their co-culture growth. The recovered bacteria from planktonic and bio lm conditions of co-culture were employed for antibiotic susceptibility testing using disk diffusion and MIC based on CLSI 2018. Antibiotic susceptibility was performed for imipenem, meropenem, doripenem, and tobramycin. P. aeruginosa ATCC 27853 was used as the reference strain. All tests were done in triplicate.

RNA extraction and gene expression of the recovered P. aeruginosa strains
The total RNA was isolated during the log phase of the mono-cultures and over six periods during the log phase of the co-cultures. The strains were inoculated into LB broth (Merck, Germany) and then incubated at 37 °C. The RNA was extracted and cDNA synthesis was performed using the GeneAll RNA extraction kit and the GeneAll cDNA synthesis kit (GeneAll, Korea) according to the manufacturer's instructions. Quantitative real-time PCR was used to determine the expressions of lasR, lasI, algD, mexR, and KPC genes using the SYBR Green method and rpoD was employed as the reference gene. The primers used from different studies [2,6,26,27,33,34] and are listed in Table 1. Each reaction contained 3 µL molecular grade water, 2 µL primers with a nal concentration of 0.5 µM, and 10 µL SYBR Green master mix (Takara Bio, Inc., Otsu, Japan). The ABI StepOne-Plus LightCycler 96 (Applied Biosystems, Foster City, USA) was used. The cycling parameters included one denaturing cycle at 94 o C/15 minutes, followed by 40 three-step cycles of ampli cation (95 o C/30 seconds, 59 o C/30 seconds, and 72 o C/30 seconds). A melting curve was also drawn on the rst run for each sample. The melting curve analysis was performed using a temperature range of 65 o C to 90 o C with a three-second interval. P. aeruginosa ATCC 27853 was used as the negative control. All tests were done in triplicate.

Statistical analysis
All the data were presented as mean ± SEM. For all the data collected, a two-way analysis of variance (ANOVA) was performed using GraphPad Prism 6.0 (Graph Pad Software, USA). When necessary, Tukey's test, the chi-square test, and the Student's t-test were applied to the data to determine the statistically signi cant changes by providing the adjusted p-values. All the presented p-values were adjusted for multiple comparisons. Gene expression analysis was performed using the REST® software (version 2009, Qiagen, Germany). The ΔΔCt method was used to determine the expression levels.  The bio lm (A) and planktonic (B) states of the co-culture of P. aeruginosa -S. aureus on the A549 cell line. 108 CFU/mL of PAO1 was obtained from corneas with and without the S. aureus co-culture. The viability of P. aeruginosa was measured as log10 (CFU/well) in the co-culture with S. aureus in a six-day period. PAO1: P. aeruginosa PAO1; SA-1: S. aureus ATCC 25923; PA-2: P. aeruginosa NCTC13359; PA-3: P. aeruginosa NCTC13618; PA-4: P. aeruginosa NCTC12903. The means of the colony counts in six different times (day 1 to day 6) are shown in the A549 cell line; star: sixth day; polygone: fth day; diamond: fourth day; three-pointed star: third day; right arrowhead: second day; circle: rst day. The error bars indicate the standard errors of the means in a representative triplicate time. The Student's t-test and χ2 test were used to test the differences between the groups. *: p < 0.05, **: p < 0.001, ***: p < 0.0001.