Staphylococcus aureus aggregation in the plasma fraction of silkworm hemolymph

Staphylococcus aureus formed bacterial aggregates in the plasma fraction of the hemolymph of silkworm, the larva of Bombyx mori, in a growth-dependent manner. The addition of arabinose or galactose inhibited the formation of S. aureus aggregates in the silkworm plasma. Formation of the bacterial aggregates depended on S. aureus genes required for the synthesis of bacterial surface polysaccharides–ypfP and ltaA, which are involved in lipoteichoic acid synthesis, and the tagO gene, which is involved in wall teichoic acid synthesis. These findings suggest that S. aureus forms bacterial aggregates in the silkworm plasma via bacterial surface teichoic acids.


Introduction
Staphylococcus aureus is a human pathogenic bacterium that exists in the nares of 30% healthy persons and causes various diseases such as pneumonia, meningitis, and sepsis in immunocompromised patients. Since the 1960s, methicillin-resistant S. aureus (MRSA) has infected many people in hospitals and clinical care facilities, and a new type of MRSA, called community-acquired MRSA, has emerged in the last two decades, infecting healthy persons in the general population [1]. Overcoming S. aureus infectious diseases will require a comprehensive understanding of the detailed molecular mechanisms underlying S. aureus virulence.
S. aureus forms aggregates in mammalian blood plasma, which is called the "coagulase reaction" [2]. The coagulase reaction is an important clinical feature for identifying S. aureus [3][4][5]. S. aureus aggregation caused by the coagulase reaction facilitates bacterial escape from host immune responses and contributes to S. aureus virulence [6][7][8]. Thus, clarification of the molecular mechanism of the aggregation reaction of S. aureus in animal blood plasma is important for understanding S. aureus virulence. Although many molecular aspects of the coagulase reaction in mammalian plasma have been studied, it has remained unclear whether S. aureus forms aggregates in the plasma fraction from animals other than mammals.
Animal infection models are essential for investigating the virulence mechanism of human pathogenic bacteria, but the use of mammals such as mice is limited due to ethical and economical issues. On the other hand, large numbers of invertebrate animals such as silkworms, fruit flies, and nematodes, which share similar innate immune systems with mammals, can be a1111111111 a1111111111 a1111111111 a1111111111 a1111111111 used more easily, and have been used as animal models of human pathogenic bacterial infection [9,10]. We have used silkworms to investigate the virulence system of S. aureus and the host innate immune system against S. aureus [9,11]. S. aureus novel virulence factors identified by using a silkworm infection model contribute to S. aureus virulence in mice [12][13][14][15]. Lipid carrier proteins of both silkworms and mice suppress S. aureus virulence [16][17][18]. These findings suggest the usefulness of the silkworm infection model to investigate S. aureus-host interactions [19]. In the present study, we investigated whether S. aureus forms aggregates in silkworm plasma. We observed S. aureus aggregation in silkworm plasma and demonstrated the requirement of S. aureus genes involved in synthesis of teichoic acid, an S. aureus cell surface polysaccharide, for S. aureus aggregation.

Bacterial strains and culture conditions
S. aureus was aerobically cultured in tryptic soy broth (TSB) at 37˚C using an air shaker at a speed of 150 rpm (BR-3000LF, TAITEC, Tokyo, Japan). To culture S. aureus gene knockout strains, TSB was supplemented with 10 μg/ml erythromycin. The bacterial strains and plasmids used in this study are shown in Table 1.

Construction of S. aureus gene knockout strains
S. aureus gene-knockout mutant was constructed as we previously reported with minor modifications [12]. An internal region of the ypfP or ltaA gene was amplified by polymerase chain reaction (PCR) from the RN4220 genome as a template using oligonucleotide primers ( Table 2). The amplified DNA fragment was inserted into pMutinT3, resulting in pT0875R or pT0874. S. aureus RN4220 strain was electroporated with pT0875R or pT0874 by using an pltaA pHY300 with intact ltaA from RN4220 This study ptagO pHY300 with intact tagO from RN4220 [42] electroporator (Gene Pulser Xcell, BIO-RAD) in 0.2 cm cuvette at 2.3 kV, 25 microfarads, and 100 O, of which condition was slightly modified from our previous method [20]. The electroporated cells was immediately suspended in B2 broth (Casamino acids, 10g/l; Yeast extract, 25g/l; K 2 HPO 4 , 1g/l; Glucose, 5g/l; NaCl, 25g/l) [21] and incubated for 1 h at 37˚C. The cells were spread on TSB agar plates containing 10 μg/ml erythromycin and the plates were incubated at 37˚C for more than 2 days and colonies were obtained. The single colony was cultured overnight in TSB medium supplemented with 10 μg/ml erythromycin at 37˚C and the proliferated bacterial cells were treated with 10 μg/ml lysostaphin. The genomic DNA was extracted as reported previously [22]. The gene knockout was confirmed by Southern blot analysis.

Construction of a plasmid carrying the ltaA gene
To construct pltaA containing the ltaA gene with the native promoter, DNA fragment containing the native promoter and the ypfP-ltaA operon was amplified by PCR from the RN4220 genome using oligonucleotide primer pairs (ypfPltaAcomp-F and ypfPltaAcomp-R) and KOD plus DNA polymerase (TOYOBO co. ltd., Osaka, Japan) ( Table 2). The amplified DNA fragment was inserted into pHY300, resulting in pypfP-ltaA. To remove the ypfP region from pypfP-ltaA, DNA fragment without the ypfP region was amplified by PCR from pypfP-ltaA using ologonucleotide primer pairs (ypfPdel-F and ypfPdel-R) ( Table 2) and self-ligated, resulting in pltaA.

Preparation of silkworm plasma
Fertilized eggs of silkworms (Hu/Yo × Tsukuba/Ne) were purchased from Ehime Sansyu (Ehime, Japan). Hatched silkworms were raised in a laboratory as previously described [15,23]. Prolegs of fifth instar silkworms at day 2 after molting were cut and the dropped hemolymph was collected into ice-cold 50 mL conical tubes (Falcon) using a funnel. The collected hemolymph was supplemented with 0.05 mM phenylthiourea and centrifuged at 8,600 g at 4˚C for 10 min. The centrifuged supernatant was frozen in liquid nitrogen and stored at -80˚C. The frozen supernatant was melted and used as silkworm plasma.

S. aureus aggregation in silkworm plasma
A single colony of S. aureus was aerobically cultured in TSB at 37˚C overnight. The overnight culture (50 μl) was inoculated into silkworm plasma (2 ml) in 50 mL conical tubes (Falcon), and aerobically cultured at 37˚C at a speed of 150 rpm. After 7 h culturing, aggregates were collected by centrifugation at 70 g at room temperature for 5 min (RL-131, TOMY SEIKO CO., Tokyo, Japan). The aggregates were suspended with 2 ml of phosphate buffered saline (PBS) and centrifuged at 70 g at room temperature for 5 min. The washing was repeated one more  time. The aggregates were suspended in 1 ml PBS and poured into 10 x 35 mm dishes (Corning). The dishes were photographed by using a scanner (LiDE 210, CANON Inc., Tokyo, Japan). To measure the amount of the aggregates, the aggregates were dried in 1.5 ml tubes using a centrifuge evaporator (CC-105, TOMY SEIKO CO., Tokyo, Japan) and the dry weight was measured. We used 7 h cultured sample for measurement of bacterial aggregates, because enough bacterial aggregates were observed at 7 h culturing.

Microscopy of S. aureus aggregates
S. aureus aggregates were suspended in PBS, placed on slide glass, and Gram-stained (Merck). The slide glass was observed under a light microscope (DM4000B, Leica).

Requirement of bacterial growth for S. aureus aggregation in silkworm plasma
A single S. aureus colony was aerobically cultured in TSB at 37˚C overnight. The S. aureus overnight culture was autoclaved at 121˚C for 15 min. The autoclaved or non-autoclaved culture (2 ml) was centrifuged at 21,400 g at 4˚C for 5 min, and the precipitated S. aureus cells were suspended in 1 ml of PBS. The cells were collected by centrifugation at 21,400 g at 4˚C for 5 min, and were used as heat-killed cells or viable cells, respectively. The heat-killed or viable S. aureus cells were suspended in 2 ml of silkworm plasma, and were aerobically incubated at 37˚C using an air shaker at a speed of 150 rpm. After 7 h incubation, the aggregates were analyzed as described above.

Examination of the inhibitory effect of monosaccharides on S. aureus aggregation in silkworm plasma
S. aureus overnight culture (50 μl) was inoculated into 2 ml of silkworm plasma supplemented with 200 mM of monosaccharide (glucose, galactose, mannose, maltose, arabinose, or N-acetyl-glucosamine) and cultured at 37˚C using an air shaker at a speed of 150 rpm. After 7 h culturing, the aggregates were analyzed as described above.

Silkworm infection experiment
A single colony of S. aureus was aerobically cultured in TSB at 37˚C overnight. The S. aureus overnight culture was serially diluted with PBS. To concentrate the bacterial solution, the overnight culture was centrifuged at 8,600 g at 4˚C for 5 min, and bacterial cell pellet was suspended in an appropriate amount of PBS. Fifth instar silkworms were fed an antibiotic-free artificial diet (Katakura Industries, Tokyo, Japan) for 1 day, and were injected with bacterial solution via the intra-hemolymph route. Silkworm survival was examined at 24 h after the bacterial injection. Because the S. aureus gene knockout strains used in this study produced fewer colony forming units than the parent strain, we performed an ATP-based assay to measure the viable bacterial cell number, as described previously [24]. Briefly, the ATP concentration in the S. aureus overnight culture was measured using firefly luciferase and the substrate (Kikkoman Corp., Tokyo, Japan), and S. aureus viable cell number per culture volume was calculated.
To determine the half-maximal lethal dose (LD 50 ), the survival curve was determined by logistic regression based on the dose-response survival plots [25].

S. aureus forms bacterial aggregates in silkworm plasma
Overnight culture of S. aureus RN4220 strain was inoculated into silkworm plasma and aerobically cultured at 37˚C. After 7 h incubation, precipitates were observed at the bottom of the culture tube (Fig 1A). The precipitates were not observed when S. aureus was cultured in TSB or S. aureus was not inoculated into silkworm plasma (Fig 1A). The precipitates were collected by low centrifugation (70 g for 5 min), whereas the planktonic S. aureus cells in TSB was not collected by this centrifugation condition (Fig 1B). The precipitates were Gram-stained and analyzed under a microscope. The precipitates contained many numbers of S. aureus cells ( Fig  1C), suggesting that S. aureus forms bacterial aggregates in silkworm plasma. In the beginning of incubation, the bacterial growth curves were not much different between plasma and TSB, but in the later stage of incubation, S. aureus growth in silkworm plasma was decreased compared to that in TSB (Fig 1D), when the bacterial aggregates were observed. The weight of bacterial aggregates was 5% of the total bacterial weight (Fig 1E). We examined whether the S. aureus aggregation activity was observed in other S. aureus strains. S. aureus NCTC8325-4 strain formed a large amount of aggregates, whereas S. aureus MSSA1 and COL strains formed few aggregates (Fig 1F). The result indicates that the aggregation activities are different among S. aureus strains. We used S. aureus RN4220 strain for the subsequent analysis of this study.

S. aureus aggregation in silkworm plasma requires bacterial growth
To understand the molecular mechanism for S. aureus aggregation in silkworm plasma, we examined whether S. aureus growth is required for the formation of S. aureus aggregates in the silkworm plasma. S. aureus cells were inoculated into silkworm plasma with or without antibiotics (chloramphenicol or erythromycin) and incubated for 7 h. The addition of antibiotics completely abolished the S. aureus aggregation in silkworm plasma (Fig 2A). To further address whether bacterial growth or presence of enough number of viable cells are required for the S. aureus aggregation, we examined the aggregation capacities of heat-killed S. aureus cells or S. aureus viable cells in silkworm plasma, in which cell concentration is corresponding to that at full growth phase. Viable S. aureus cells from 2 mL overnight culture or heat-killed S. aureus cells from 2 mL overnight culture was inoculated into 2 mL of silkworm plasma, and aerobically cultured at 37˚C. After 7 h incubation, both the viable cells from 2 mL culture and the heat-killed cells severely decreased the formation of the bacterial aggregates ( Fig 2B). On the other hand, viable S. aureus cells from 0.05 mL overnight culture did produce bacterial aggregates (Fig 2B). These findings suggest that S. aureus growth in silkworm plasma is required for S. aureus aggregation in silkworm plasma.

S. aureus aggregation in silkworm plasma is inhibited by the addition of monosaccharides
Lectin binds to polysaccharides of foreign substances and functions in host defense. Monosaccharides inhibit lectin binding [26]. We examined whether the addition of monosaccharides inhibits S. aureus aggregation in silkworm plasma. S. aureus was aerobically cultured in silkworm plasma supplemented with or without several monosaccharides (arabinose, galactose, glucose, maltose, mannose, and N-acetyl-glucosamine) at 37˚C. Addition of arabinose and galactose decreased the amount of S. aureus aggregates in silkworm plasma, whereas glucose, maltose, mannose, and N-acetyl-glucosamine did not decrease the amount of S. aureus aggregates (Fig 3A and 3B). We measured the S. aureus growth curves in silkworm plasma supplemented with these monosaccharides to see whether the effect of monosaccharide is due to bacterial growth inhibition. The addition of all of these monosaccharides did not inhibit bacterial growth during 0-3 h (Fig 3C). There were differences of growth curves during 4-7 h, in which bacterial growth in the presence of arabinose and galactose were higher than that without addition of monosaccharide, whereas bacterial growth in the presence of N-acetyl-glucosamine was lower than that without addition of monosaccharide ( Fig 3C). These results suggest that arabinose and galactose inhibit S. aureus aggregation without inhibiting bacterial growth.

Fig 1. S. aureus forms bacterial aggregates in silkworm plasma.
A) S. aureus overnight culture (0.05 ml, 1.5 x 10 8 CFU) was inoculated into 2 ml of silkworm plasma (left) or TSB (right) and was incubated at 37˚C for 7 h. Silkworm plasma without bacterial inoculation was incubated in the same condition (center). Precipitate in silkworm plasma is indicated by a black arrow. B) S. aureus was cultured in silkworm plasma or TSB at 37˚C for 7 h. The culture was centrifuged at 70 g for 5 min to collect precipitate and the precipitate was spread in a Petri dish. C) S. aureus was cultured in silkworm plasma at 37˚C for 7 h. The formed precipitate were Gram-stained and examined under a light microscope. D) S. aureus growth curves in silkworm plasma or TSB at 37˚C were measured. A fluid part of bacterial culture was sampled during the time course of incubation and the OD 600 value was measured. E) Relative weights of the bacterial aggregates and the planktonic bacteria to the total bacterial weight were measured. S. aureus was cultured in silkworm plasma or TSB at 37˚C for 7h. The bacterial aggregates were collected by low centrifugation and the dry weight of the aggregates was measured. The planktonic bacterial cells in the centrifuged supernatant were collected by further centrifugation at 21,400 g for 5 min and the dry weight was measured. Vertical axis represents the relative weight to the total bacterial weight in plasma or TSB. ND means not detected. F) S. aureus strains including RN4220, MSSA1, NCTC8325-4, and COL were cultured in silkworm plasma at 37˚C for 7 h. Bacterial aggregates were spread in a Petri dish.
https://doi.org/10.1371/journal.pone.0217517. g001   Fig 2. S. aureus growth is required for S. aureus aggregation in silkworm plasma. A) S. aureus was cultured in silkworm plasma or in silkworm plasma supplemented with 50 μg/ml chloramphenicol or 10 μg/ml erythromycin at 37˚C for 7 h. The aggregates were collected and spread in a Petri dish. B) S. aureus viable cells from 0.05 mL overnight culture (left), S. aureus viable cells from 2 mL overnight culture (center), or S. aureus heat-killed cells from 2 mL overnight culture (right) was inoculated into 2 ml of silkworm plasma, and cultured at 37˚C for 7 h. The aggregates were collected and spread in a Petri dish.

S. aureus teichoic acids are required for S. aureus aggregation in silkworm plasma
Based on the results that arabinose and galactose inhibited S. aureus aggregation, we hypothesized that bacterial surface polysaccharides are involved in S. aureus aggregation in silkworm plasma and examined the effect of knockout of S. aureus genes involved in teichoic acid synthesis: ypfP, which encodes synthetase for diglucosyl-diacylglycerol, the membrane anchor of lipoteichoic acids [27]; ltaA, which encodes translocase of diglucosyl-diacylglycerol from the inner to the outer leaflet of the membrane [28]; and tagO, which encodes synthetase for wall teichoic acids [29]. The ypfP, ltaA, and tagO knockout strains exhibited severely decreased aggregation in silkworm plasma compared with the parent strain (Fig 4A and 4B). The decreased aggregations of the ypfP, ltaA, and tagO knockout strains were respectively restored by introducing wildtype ypfP, ltaA, tagO genes (Fig 4A and 4B). These findings suggest that bacterial cell surface teichoic acids are required for S. aureus aggregation in silkworm plasma.

S. aureus ypfP and ltaA genes are required for S. aureus virulence in silkworms
To determine whether the ability of S. aureus to form aggregates in silkworm plasma is involved in S. aureus virulence against silkworms, we examined the virulence activity of ypfP, ltaA, and tagO knockout strains by measuring the LD 50 value, the bacterial dose that kills half of the silkworms. The LD 50 values of the ypfP and ltaA knockout strains against silkworms were higher than the parent strain, whereas the LD 50 value of the tagO knockout strain was not (Fig 5). The LD 50 values of the ypfP and ltaA knockout strains were restored to the level of the parent strain by introducing intact ypfP and ltaA genes, respectively (Fig 5). These findings suggest the ypfP and ltaA genes are required for S. aureus killing activity against silkworms.

Discussion
The present study demonstrated that S. aureus forms bacterial aggregates in silkworm plasma in vitro, which is the first indication that S. aureus bacterial aggregates in animal plasma is not specific to mammals, but is conserved in insects.
This study revealed that S. aureus growth in silkworm plasma is required for the formation of aggregates and that the aggregation process requires the ypfP, ltaA, and tagO gene-encoding enzymes involved in teichoic acid synthesis. Thus, the interaction between S. aureus cell surface teichoic acids and host factor(s) in the silkworm plasma during S. aureus growth causes bacterial aggregation. Because S. aureus growth was required for the reaction, S. aureus molecule(s) specific to the logarithmic growth phase is required for the aggregation. In the coagulase reaction, which is evaluated by culturing S. aureus cells in rabbit plasma for 2-4 h, coagulase secreted from S. aureus cells activates coagulase reacting factor (CRF) in the rabbit plasma, and the activated CRF converts fibrinogen to fibrin, resulting in plasma clotting [30, Fig 3. S. aureus aggregation in silkworm plasma is inhibited by the addition of monosaccharides. A) S. aureus was cultured in silkworm plasma supplemented with or without monosaccharides [arabinose (Ara), galactose (Gal), glucose (Glu), maltose (Mal), mannose (Man), or N-acetyl-glucosamine (GlcNAc)] at 37˚C for 7 h. Bacterial aggregates were spread in a Petri dish. B) The dry weight of the aggregates observed in A) was measured. Means±standard errors from three independent experiments are presented. Asterisks indicate Student's t-test p-value less than 0.05 between the absence and presence of monosaccharide. C) S. aureus growth curves in silkworm plasma supplemented with or without monosaccharides [arabinose (Ara), galactose (Gal), glucose (Glu), maltose (Mal), mannose (Man), or Nacetyl-glucosamine (GlcNAc)] were measured at 37˚C. This experiment was performed in parallel with the experiment of Fig 1D and the growth curve in silkworm plasma without monosaccharide is same with the data in Fig 1D (Plasma). https://doi.org/10.1371/journal.pone.0217517.g003 In vitro bacterial aggregation in insect plasma 31]. Identification of such coagulase-like molecule is important to understand the molecular mechanism of S. aureus aggregation in silkworm plasma.
The coagulase reaction utilizes plasma clotting, a host immune reaction, to enable bacterial escape from humoral and cellular bactericidal reactions in the host and contributes to bacterial virulence [7,8]. Single knockout of the coagulase gene dose not decrease S. aureus virulence in mice or rats [32][33][34], although double knockout of the coagulase gene and von Willebrand factor binding protein gene, another coagulation factor, decreases S. aureus virulence in mice [35]. The present study revealed that S. aureus gene knockout mutants of the ypfP and ltaA genes with decreased aggregation activity in silkworm plasma exhibit attenuated virulence in A) The S. aureus parent strain (WT) transformed with an empty vector (pHY300) and knockout strains of the ypfP, ltaA, and tagO genes (ΔypfP, ΔltaA, ΔtagO), which were transformed with an empty vector or a plasmid carrying respective wildtype gene (pypfP, pltaA, and ptagO), were cultured in silkworm plasma at 37˚C for 7 h and aggregates were observed. B) The dry weights of the aggregates observed in C) were measured. Means±standard errors from three independent experiments are presented. Student's t-test p-value less than 0.05 are shown ( � , the gene-knockout strain vs. the parent strain; �� , the complemented strain vs. the gene-knockout strain). https://doi.org/10.1371/journal.pone.0217517.g004 In vitro bacterial aggregation in insect plasma silkworms, whereas the tagO-knockout mutant with decreased aggregation activity does not attenuate virulence. Because the decreased aggregation activities were indistinguishable between the three mutants, it cannot conclude the involvement of the aggregation activity in S. aureus virulence in silkworms. To clarify the physiological role of S. aureus aggregation in silkworm plasma, further studies are needed such as constructing a gene-knockout mutant with severely decreased aggregation, or evaluating detailed infection processes in the silkworm model.
Nodule formation is an aggregate of insect hemocytes encapsulating invading microorganism, a cellular immune reaction in insects [36]. Carbohydrate binding proteins in the insect hemolymph called lectins, such as hemocytin and BmMBP, are involved in nodule formation [37,38]. Because S. aureus aggregation in silkworm plasma was inhibited by the addition of monosaccharides, lectins in the silkworm plasma are presumed to be involved in the S. aureus aggregation formation. Thus, nodule formation and S. aureus aggregation both involve lectins, raising the possibility that the S. aureus aggregation in silkworm plasma is one of the aspects of nodule formation. However, there are several different features between nodule formation and the S. aureus aggregation in silkworm plasma. First, nodule formation requires insect hemocytes, whereas the S. aureus aggregation occurs in silkworm plasma lacking hemocytes. Second, S. aureus aggregation requires bacterial growth, but nodule formation does not require bacterial growth, because nodule formation occurs by formaldehyde-treated E. coli or Micrococcus luteus [37,39]. Third, nodule formation is a rapid reaction that occurs within minutes after bacterial invasion [40]. When Saccharomyces cerevisiae or E. coli was injected into the silkworm haemocoel, transparent nodules formed within 1 min [37], and black aggregates Silkworms (n = 5/ dose) were injected with serial dilutions of S. aureus overnight cultures of the parent strain (WT/pHY300), the knockout strains of the ypfP, ltaA, and tagO genes (ΔypfP/pHY300, ΔltaA/pHY300, ΔtagO/pHY300), and the complemented strains (ΔypfP/pypfP, ΔltaA/pltaA, ΔtagO/ptagO). The silkworm survival was evaluated at 24 h after the injection and the LD 50 values were determined by logistic regression. The mean LD50 values with standard errors from three independent experiments are presented. Student's t-test p-value less than 0.05 are shown ( � , the gene-knockout strain vs. the parent strain; �� , the complemented strain vs. the gene-knockout strain). NS means not significant. https://doi.org/10.1371/journal.pone.0217517.g005 In vitro bacterial aggregation in insect plasma were observed after 30 min [39]. On the other hand, S. aureus aggregation in silkworm plasma was not observed for at least 5 h. Based on these differences between nodule formation and S. aureus aggregation in silkworm plasma, we assume that S. aureus aggregates are formed via different molecular mechanisms than nodules.
This study unveiled a new ability of S. aureus to form bacterial aggregates in silkworm plasma and identified teichoic acids as a key factor. This finding will facilitate our understanding of S. aureus versatile ability to act not only on mammalian system but also on insect system, and enables future studies to investigate the conserved molecular mechanism of bacterial aggregation between mammals and insects.