1

SERRATIA MARCESCENS SUPPRESSES HOST CELLULAR IMMUNITY VIA THE PRODUCTION OF AN ADHESION-INHIBITORY FACTOR AGAINST

IMMUNOSURVEILLANCE CELLS Kenichi Ishii, Tatsuo Adachi, Hiroshi Hamamoto, and Kazuhisa Sekimizu

From the Laboratory of Microbiology, Graduate School of Pharmaceutical Sciences, The University of Tokyo, Tokyo 113-0033, Japan.

Running title: S. marcescens inhibits immune cell adhesion Address correspondence to: Kazuhisa Sekimizu, 7-3-1 Hongo, Bunkyo-ku, Tokyo 113-0033, Japan.

Tel.: 81-3-5841-4820; Fax: 81-3-5684-2973; E-mail: sekimizu@mol.f.u-tokyo.ac.jp Background: Serratia marcescens induced an acute increase in immune cell numbers in silkworms. Results: Serralysin metalloprotease was identified as a factor that inhibited immune cell adhesion and bacterial clearance. Conclusion: S. marcescens suppresses cellular immunity by degrading adhesion molecules. Significance: A novel mode of immune evasion by S. marcescens is proposed. SUMMARY Injection of a culture supernatant of Serratia marcescens into the bloodstream of the silkworm Bombyx mori increased the number of freely circulating immunosurveillance cells (hemocytes). Using a bioassay with live silkworms, serralysin metalloprotease was purified from the culture supernatant and identified as the factor responsible for this activity. Serralysin inhibited the in vitro attachment of both silkworm hemocytes and murine peritoneal macrophages. Incubation of silkworm hemocytes or murine macrophages with serralysin resulted in degradation of the cellular immune factor BmSPH-1 or calreticulin, respectively. Furthermore, serralysin suppressed in vitro phagocytosis of bacteria by hemocytes and in vivo bacterial clearance in silkworms. Disruption of the serralysin gene in S. marcescens attenuated its host killing ability in silkworms and mice. These findings suggest that serralysin metalloprotease secreted by S. marcescens suppresses cellular immunity by decreasing the adhesive properties of immunosurveillance cells, thereby contributing to bacterial pathogenesis. Living organisms are constitutively exposed to

the danger of infection by microorganisms present in the surrounding environment, such as in the air, water, food, and soil (1). S. marcescens is an environmental bacteria that causes sepsis in immunocompromised people (2). S. marcescens strains resistant to multiple antibiotics were recently reported (3). Therefore, overcoming S. marcescens infection is an important clinical issue. Recent studies revealed the biosynthetic mechanisms of virulence components such as the flagella (4,5) and outer membrane vesicles (6) of S. marcescens. On the other hand, how S. marcescens interacts with the host immune system and escapes attack by immunosurveillance cells remains unclear. To elucidate bacterial pathogenesis, it is crucial to understand the interactions between bacteria and host immune systems. Especially in the battle against bacteria at the early stage of infection, the innate immune system acting independently of antibodies plays a critical role. Invertebrates and mammals share a common basis of innate immunity (7). Our laboratory studies issues regarding bacterial infection through analyses of the innate immune system using the silkworm Bombyx mori. Silkworms are highly suited for biochemical analysis of hemocytes, immunosurveillance cells circulating within the insect bloodstream, because of their large body size and the technical simplicity of blood (hemolymph) collection (8). We recently demonstrated that bacterial and fungal cell wall components induce the activation of an insect cytokine named paralytic peptide (PP) in the silkworm hemolymph (9), and that active PP promotes the expression of adhesion molecules in hemocytes (10). Our group has established silkworm infection models to analyze the virulence mechanisms of various pathogens (11-14). In the course of our

http://www.jbc.org/cgi/doi/10.1074/jbc.M113.544536The latest version is at JBC Papers in Press. Published on January 7, 2014 as Manuscript M113.544536

Copyright 2014 by The American Society for Biochemistry and Molecular Biology, Inc.

by guest on July 18, 2020http://w

ww

.jbc.org/D

ownloaded from

2

studies of S. marcescens pathogenesis using the silkworm infection model (15), we found that the number of freely circulating hemocytes in silkworms increased immediately after infection. We speculated that the bacteria inhibited the adhesion of hemocytes to the body cavity and tissues. In general, the adhesive properties of immune cells are important for cellular immune responses in host animals. For example, adhesion molecules are involved in bacterial phagocytosis by hemocytes in the tobacco hornworm Manduca sexta (16). Moreover, in the fruit fly Drosophila melanogaster, mutation of the gene encoding a transmembrane protein mediating bacterial phagocytosis (eater) leads to an increase in the number of circulating hemocytes (17). Based on these reports, we hypothesized that S. marcescens decreases the adhesive abilities of immune cells and exerts its virulence via the suppression of host cellular immunity. In the present study, we purified factors that increase the cell density of silkworm hemocytes from the culture supernatant of S. marcescens. EXPERIMENTAL PROCEDURES Animals, bacteria, and reagents—Silkworm (Hu·Yo × Tukuba·Ne) eggs were purchased from Ehime Sansyu (Ehime, Japan). Silkworm larvae were reared on an artificial diet (Silkmate 2S, Nihon Nosan Kogyo) at 27°C. Mice (C57BL/6JJc1) were purchased from CLEA Japan, Inc. A red-color pigmented Serratia marcescens 2170 strain was harvested in BHI medium (Becton Dickinson and Co.) at 30°C. E. coli W3110 and S. aureus Newman were harvested in tryptic soy broth (Becton Dickinson Co.) at 37°C. The active form of the insect cytokine PP and its truncated form lacking the N-terminus ENF residues were chemically synthesized (18,19). Anti-PP rabbit antiserum was prepared as described previously (18). Measurement of hemocyte numbers in silkworm hemolymph—Silkworm larvae (day 2 of 5th instar, 2 g/larva) were injured or injected with liquid samples into the back between the 8th and the 9th segment using 27 G needles and 1-ml syringes (Thermo). Silkworms were incubated at 27°C, and the legs were cut with scissors to collect the hemolymph. In some experiments, equal volumes of hemolymph obtained from several larvae were

pooled. To prevent melanization and other serine-protease mediated reactions, 5 µl of hemolymph was immediately mixed with 15 µl of 10 mM benzamidine chloride dissolved in insect physiologic saline (IPS) (150 mM NaCl, 5 mM KCl, 1 mM CaCl2). Samples were loaded on a cytometer and hemocyte numbers were counted under a microscope. Purification of the hemocyte-increasing factor from S. marcescens culture supernatant—S. marcescens was inoculated in 400 ml BHI medium (40 tubes; Conical centrifuge tubes, 50 ml polypropylene, Becton Dickinson Co. containing 10 ml each of the inoculated medium) and shaken overnight at 30°C. The collected culture containing approximately 1010 cells/ml was centrifuged and the cell pellet was re-suspended in 40 ml IPS (40 tubes containing 1 ml each of the suspended culture). The IPS culture was statically incubated at 30°C overnight. The collected culture was centrifuged and the supernatant was filtered through a Millipore filter (Millex-GV, 0.22 µm, PVDF) to obtain the “IPS culture supernatant” fraction (Fr. I). Phenyl-Toyopearl resin (12 ml; Phenyl-650M, TOSOH) was washed with reverse-osmosis water (ROW) and equilibrated with 1 M ammonium sulfate. Fraction (Fr.) I (39 ml) was mixed with 13 ml of 4 M ammonium sulfate and applied to the column. The column was sequentially washed with 3-column volumes (36 ml) of 1 M and 0.25 M ammonium sulfate. The column was then loaded with 36 ml of IPS, and the eluted sample was collected as Fr. II. Hydroxyapatite resin (4 ml; Seikagaku-kogyo, Japan) was swelled in 1 mM phosphate buffer (pH 6.8) for at least 1 day. The column was washed with 400 mM phosphate buffer (pH 6.8) and equilibrated with 10 mM phosphate buffer (pH 6.8). Fr. II was dialyzed in the 10 mM phosphate buffer prior to loading. After applying 36 ml of the dialyzed sample, the column was washed with 3 column volumes (12 ml) of 10 mM phosphate buffer. The column was then loaded with 12 ml of 50 mM phosphate buffer (pH 6.8), and the eluted sample was collected as Fr. III. DEAE-Toyopearl resin (4 ml; DEAE-650M, TOSOH) was washed with ROW and 1 M NaCl in 50 mM Tris-HCl (pH 8.0), and then equilibrated with 50 mM Tris-HCl (pH 8.0). Fr. III was

by guest on July 18, 2020http://w

ww

.jbc.org/D

ownloaded from

3

dialyzed in 50 mM Tris-HCl (pH 8.0) before loading. After applying 11 ml of the dialyzed sample, the column was washed with 50 mM Tris buffer, and then a linear-gradient elution from 0 to 0.3 M NaCl using 60 ml of 50 mM Tris-HCl (pH 8.0) and 60 ml of 0.3 M NaCl in 50 mM Tris buffer (30-column volumes in total) was performed with a gradient maker. The elution speed was approximately 1 ml/min, and 30 fractions (4 ml/Fr.) were collected in 15-ml tubes. As indicated in Fig. 2, Fr. #14 and 15 were pooled as peak 2 or Fr. IV. For further chromatography of the peak 1 fraction eluted from the DEAE column, we performed gel-filtration column chromatography with Superose 12. The peak 1 fraction was concentrated by Amicon Ultra Centrifugal Filters with Ultracel-3K membranes (Millipore) to reduce the volume approximately 40-fold. The sample was applied to the Superose 12 10/300 GL gel filtration column (GE Healthcare) and eluted by the AKTApurifier system (GE Healthcare). Elution was performed in 100 mM NaCl in 50 mM Tris-HCl (pH 8.0) at a flow rate of 1 ml/min, and 50 fractions (0.5 ml/Fr.) were collected in 2-ml tubes. All fractions except for Fr. I were dialyzed in IPS prior to assessing their hemocyte-increasing activities. Protein concentrations were measured by using Coomassie PlusTM protein assay reagent (Thermo) with bovine serum albumin as the standard. The purification procedure using independently prepared bacterial cultures was repeated three times with similar results. Amino acid sequence analysis was performed by in-gel trypsin digestion followed by the Edman degradation method. Fragments separated with HPLC were analyzed by mass spectrometry. Production of recombinant serralysin protein—Genome DNA of S. marcescens was prepared by using a QIAamp DNA Blood Mini kit (Qiagen). The coding region of the serralysin gene was amplified by polymerase chain reaction (PCR) using forward (5’-AAGAATTCTAACCGTGGCTTAC-3’) and reverse (5’-AACTCGAGCGCGTTAAAGTACCT-3’) primers and the S. marcescens genome DNA as a template. The amplified 1.6-kb DNA fragment and pET28a expression plasmid were

double-digested with restriction enzymes EcoR I and Xho I, and the purified DNA fragments were ligated using a TAKARA ligation kit. The constructed plasmid was transformed into E. coli BL21(DE3)pLysS. Cells were harvested in 5 ml LB10 medium containing 50 µg/ml kanamycin, and 1 ml of the overnight culture was transferred into 1 l of LB10 medium containing 50 µg/ml kanamycin. The culture was shaken at 37°C for 2 h, and then 1 ml of 0.5 M isopropyl-β-D-thigalactopyranoside was added. After further incubation at 37°C for 2 h, the culture was centrifuged. Cells were suspended in 10 ml PBS and rapidly frozen in liquid nitrogen. Samples were thawed and supplied with 100 µl of 100 mg/ml lysozyme, and cells were disrupted by sonicating twice using a Sonifier 450 (Branson) with output control of 2 for 30 s. Samples were frozen in liquid nitrogen again, followed by thawing and sonicating four times for 30 s with 30 s intervals. After centrifugation, the precipitate was solubilized in 20 ml PBS containing 8 M urea. Samples were then centrifuged and the supernatant was collected as the “urea soluble fraction”. To purify the His-tagged serralysin protein from the soluble fraction, Ni-affinity column chromatography was performed under a denaturing condition. Probond resin (3-5 ml' Invitrogen) was sequentially washed with 0.5 N NaOH, ROW, and 20 mM phosphate buffer (pH 7.8), and then equilibrated with 5 to 10 column volumes of buffer A (20 mM phosphate buffer [pH 7.8] containing 0.5 M NaCl and 8 M urea). A 20-ml aliquot of the urea soluble fraction mixed with 2.4 ml of 200 mM phosphate buffer (pH 7.8) and 2.4 ml of 5 M NaCl (final concentrations: 20 mM for phosphate and 0.5 M for NaCl) was applied to the column, and the resin was washed with 3-column volumes of buffer B (20 mM phosphate buffer [pH 7.8] containing 10 mM imidazole, 0.5 M NaCl, and 8 M urea). Ten-column volumes of buffer C (20 mM NaH2PO4 [pH 4.0], 0.5 M NaCl, and 8 M urea) were then applied to the column, and 2-column volumes per fraction of the eluate were collected (E1-E5). The eluted fractions were dialyzed in IPS before assessing their biologic activities. Attachment assay of silkworm hemocytes—Hemolymph (5 ml) was collected from 40 larvae (day 2 of 5th instar) in an ice-cold

by guest on July 18, 2020http://w

ww

.jbc.org/D

ownloaded from

4

tube containing 10 ml of 10 mM benzamidine chloride. The hemolymph was centrifuged at 1000x g for 5 min at 4°C, and the precipitated cells were suspended in 2 ml IPS. Cells were transferred to tissue-culture treated 24-well polystyrene plates and incubated in the presence of recombinant serralysin at 27°C for 3 h. The plate was gently shaken by hand 10 times and the aliquots were collected. Cell numbers were counted under a microscope. Two-dimensional electrophoresis analysis—For the analysis of silkworm hemocytes, membrane fractions were prepared according to a previous report (20) with modifications. Approximately 30 ml of hemolymph was collected from 240 larvae (day 2 of 5th instar) in an ice-cold tube containing 10 ml of 10 mM benzamidine chloride. The hemolymph was centrifuged and the collected hemocytes (approximately 108 cells) were washed with IPS. Cells suspended in 0.4 ml of IPS were divided into 2 groups. An equal volume of either IPS or 1.6 µM recombinant serralysin was added to the hemocyte suspension. After incubation at 27°C for 1 h, the samples were centrifuged and the cells were suspended in 1 ml homogenization buffer (IPS containing protease inhibitor cocktail). Cells were homogenized and sonicated at Output control of 1 for 10 s, and then the samples were centrifuged at 100x g for 1 min to remove the debris. The supernatant was further centrifuged at 12,000 x g at 4°C for 30 min. Precipitates were suspended in 400 µl of 2% CHAPS (3-[(3-cholamidopropyl)dimethyl-ammonio]propanesulfonate) and sonicated at an Output control of 1 for 5 s. Trichloroacetic acid (44 µl) was added to each sample and placed at 4°C overnight. Samples were centrifuged and the precipitates were washed two times with ethanol. The precipitated samples were dissolved in a solubilization buffer (8 M urea, 50 mM dithiothreitol, 2% CHAPS, 0.001% bromophenol blue, and 0.2% carrier ampholytes [Bio-Rad]) by sonication. Isoelectric focusing on ReadyStripTM IPG strips pH 3-10, 11 cm (Bio-Rad, #163-2014) was performed with PROTEAN IEF Cell (Bio-Rad). Strips were sequentially treated with equilibration solution I (6 M urea, 2% SDS, 375 mM Tris-HCl [pH 8.8], 20% glycerol, and freshly supplied 20 mg/ml dithiothreitol) for 20 to 30 min and equilibration solution II (6 M urea, 2% SDS,

375 mM Tris [pH 8.8], 20% glycerol, and freshly supplied 25 mg/ml iodoacetamide) for 10 min at room temperature. Strips were then subjected to SDS-polyacrylamide gel electrophoresis using 12.5% gels (12 cm x 14 cm), and the gels were stained with Coomassie Brilliant Blue. For analysis of mouse peritoneal macrophages, cells were obtained from C57BL/6JJc1 female mice as described previously (15). Approximately 0.5 x 108 cells obtained from 15 mice were treated with serralysin, and the membrane fraction was prepared and analyzed as described above. Phagocytosis assay of silkworm hemocytes—Phagocytosis of S. aureus Newman strain by hemocytes collected from silkworm larvae (day 3 of 5th instar) was assessed as described previously (10). Bacterial counts per hemocyte measured in each group treated with or without recombinant serralysin after 1 h at 27°C were normalized with those of the control group incubated at 4°C to calculate the phagocytic index. Construction of a S. marcescens-disrupted mutant of the serralysin gene—Gene disruption of S. marcescens was performed according to previous reports (15,21,22). Briefly, the 145-309th nucleotides of the internal region within the serralysin open reading frame (1536 bp) was amplified by PCR using forward (5’-AAGAATTCGACGACCTGCTGCATTATCA-3’) and reverse (5’-AAGAATTCATCCGGGAAGGAGAAGGTTA-3’) primers and S. marcescens 2170 genome DNA as a template. The amplified DNA fragments were treated with EcoR I and ligated with EcoR I-digested pir-dependent plasmid pFS100. E. coli S17-1 λpir strain transformed with the constructed plasmid was further conjugated with S. marcescens. Disruption of the targeted region in the S. marcescens genome was confirmed by PCR. Western blot analysis—To test the degradation of PP by serralysin in vitro, 1 µM of chemically synthesized PP was incubated with 800 nM of recombinant serralysin at 27°C for 1 h. Western blot analysis of PP was performed as described previously (9). To observe the serralysin-dependent degradation of calreticulin in mouse macrophages, peritoneal macrophages were collected from 10

by guest on July 18, 2020http://w

ww

.jbc.org/D

ownloaded from

5

mice and the membrane fraction was prepared as described above for the two-dimensional electrophoresis analysis. Protein samples were separated by SDS-polyacrylamide gel electrophoresis using 12.5% gels, and transferred to Immobilon-P PVDF membranes (Millipore). Membranes were soaked in 5% skim milk dissolved in tris-buffered saline supplied with 0.1% Tween20 (TBST) at room temperature for 2 to 3 h, and washed twice with TBST for 5 min. The membranes were then reacted overnight at 4°C with either primary antibody against calreticulin (Cell Signaling Technology, #2891) or beta-actin (Cell Signaling Technology, #4967) diluted with blocking buffer (5% bovine serum albumin in TBST) by 1/1000 or 1/5000-fold, respectively. After washing with TBST twice, membranes were further reacted with 1/5000-fold diluted horseradish peroxidase-linked anti-rabbit Ig (GE Healthcare, #NA934) in the blocking buffer at room temperature for 3 to 4 h. Bands were detected using Western Lightning Plus-ECL (Perkin Elmer, #NEL104001EA). Infection experiments—In the silkworm infection experiments, serial 2-fold dilutions of overnight cultures were injected into 5 larvae (day 2 of 5th instar) per group, and the survival rates were measured. LD50 values (at 16 h post infection of S. marcescens or 24 h post infection of S. aureus) were determined as previously described (15). In the mouse infection experiments, C57BL/6JJc1 female mice (7-8 weeks old) were intravenously injected with 200 µl S. marcescens (3 x 108 cells/ml) suspended in PBS. The survival of 6 mice per group was monitored, and the statistical difference between the survival curves was analyzed by log-rank test. RESULTS Decrease in silkworm hemocyte number by injury stimulation, and suppression of this phenomenon by the S. marcescens culture supernatant—When silkworms were injured by needles, the number of hemocytes freely circulating in the bloodstream decreased after 0.5 h (Fig. 1A). This effect was more prominent when water, insect physiologic saline (IPS), or bacterial culture medium was injected into the silkworm hemolymph (Fig. 1A). This phenomenon is considered an “injury effect”, i.e., an acute

immune response in insects (23). Similar observations are reported in mammals at early stages of infection (24). In contrast, when silkworms were injected with a live suspension of S. marcescens, the decrease in the free hemocyte number was suppressed (Fig. 1B). Injection of E. coli or S. aureus, on the other hand, did not induce the suppression (Fig. 1B). We and others have demonstrated that fewer than 10 colony forming units of S. marcescens kills insects, including silkworms (15,25). We therefore hypothesized that the above phenomenon, suppression of the injury-dependent reduction in hemocyte numbers by S. marcescens, is responsible for the high virulence of S. marcescens against silkworms. The suppressive effect on the injury-dependent hemocyte decrease was also observed by injecting S. marcescens culture supernatant. Injection of brain heart infusion (BHI) medium into the silkworm hemolymph induced a transient decrease in the number of free hemocytes within 0.5 h, followed by a slow recovery of the cell number (Fig. 1C). Injection of the S. marcescens culture supernatant led to a much higher increase in the number of hemocytes compared with control groups after 0.5 to 24 h (Fig. 1C). We previously demonstrated that live S. marcescens induces apoptotic cell death in silkworm hemocytes and suppresses innate immune responses (15). We also reported that the S. marcescens culture supernatant does not induce a rapid killing of hemocytes (15). In contrast, under conditions in which S. marcescens culture supernatant suppressed the decrease in hemocyte numbers, cell viabilities were not affected, based on by trypan blue staining (data not shown). Therefore, we considered that S. marcescens culture supernatant contains factors that promote the release of free hemocytes into the bloodstream without killing them. Purification of the responsible factors is described in the following section. Purification of hemocyte number-increasing factors from S. marcescens culture supernatant—S. marcescens culture supernatant suppressed the injury effect of IPS injection, which caused a rapid decrease in the number of free hemocytes. In other words, injection of S. marcescens culture supernatant into silkworms induced an apparent increase in the hemocyte cell

by guest on July 18, 2020http://w

ww

.jbc.org/D

ownloaded from

6

density compared with the IPS-injected group. Hereafter, we refer to this activity “hemocyte number-increasing activity”. For the purification of active factors, activity that induced a 1.5-fold increase in the hemocyte number compared with IPS-injected group was defined as 1 unit. Heat treatment of the S. marcescens BHI-culture supernatant at 100ºC for 30 min significantly reduced its hemocyte-increasing activity (Fig. 1D). An increased number of free hemocytes was also observed following the injection of a filtered supernatant of a static S. marcescens culture in IPS (instead of BHI medium; Fig. 2A). The activity of S. marcescens IPS-culture supernatant was also abolished by heat treatment at 100ºC for 30 min (Fig. 2B). We used this IPS culture supernatant as a starting material for the purification. Purification of the silkworm hemocyte number-increasing factor in the S. marcescens culture supernatant was performed by sequential column chromatography using Phenyl-Toyopearl, hydroxyapatite, and DEAE-Toyopearl (Table 1). Two active fractions eluted at a low and a high salt concentration, designated peaks 1 and 2, respectively, were obtained by DEAE-Toyopearl column chromatography (Fig. 2C). Activities of both fractions were sensitive to heat treatment (Fig. 2D, E). Regarding the peak 2 fraction, co-elution of the protein concentration with hemocyte-increasing activity was observed (Fig. 2C). Sodium dodecyl sulfate (SDS)-polyacrylamide gel electrophoresis analysis of the peak 2 fraction revealed a Coomassie Brilliant Blue (CBB)-stained single band with a molecular mass of 50 kDa (Fig. 2F). We further applied the peak 2 fraction to Superose 12 gel-filtration column chromatography and observed a single peak of activity (Fig. 3A), consistent with the presence of the 50-kDa band (Fig. 3B). In addition, specific activity of the eluted fraction was identical to that of the applied sample, indicating that the active substance in peak 2 was purified. The protein was extracted from the gel and treated with trypsin, and then subjected to amino acid sequence analysis. Two peptide sequences were obtained, which matched the internal sequence of S. marcescens serralysin metalloprotease (data not shown). We then cloned the serralysin gene from S.

marcescens genome DNA and introduced it into E. coli BL21(DE3)pLysS. E. coli cells expressing the S. marcescens serralysin gene were lysed and subjected to purification using nickel-affinity column chromatography to obtain His-tagged recombinant serralysin protein (Fig. 4A, B). The eluted fraction containing the recombinant serralysin protein increased the hemocyte number when injected into the silkworm hemolymph in a dose-dependent manner (Fig. 4C). The specific activity of recombinant serralysin prepared from E. coli cells was 1.3 units/µg protein, which was 1/9 that of peak 2 fraction purified from S. marcescens IPS culture supernatant. The activity of the recombinant protein was abolished by heat treatment at 100ºC for 30 min (Fig. 4D). The protein degradation activities of metalloproteases including serralysin are inhibited by EDTA, a chelating agent of divalent metal ions (26). In the above experiment using the recombinant serralysin fraction, the addition of 10 mM EDTA (final concentration in silkworm hemolymph) to the samples completely inhibited the increase in the hemocyte density (Fig. 4E), suggesting that divalent metal ions are necessary for the hemocyte number-increasing activity of serralysin. This means that the protein degradation activity of serralysin is required for the hemocyte number-increasing activity. Suppression of host cellular immune responses by serralysin—S. marcescens culture supernatant induced a rapid increase in the hemocyte density in the silkworm hemolymph within 30 min (Fig. 1C). Based on this observation, we speculated that serralysin blocked the adhesion of hemocytes to body tissues without promoting cell growth in hematopoietic organs. To test this hypothesis, we performed the following experiments. Hemocytes were pre-cultured on polystyrene dishes. The number of cells that were detached from the bottom surface by gentle shaking was higher in cultures supplied with recombinant serralysin than in control groups (Fig. 5A). Because serralysin is a protease, we assumed that serralysin degraded adhesion molecules present on the hemocyte membrane and thus attenuated the cell adhesiveness. We therefore attempted to identify the membrane proteins degraded by serralysin. Silkworm hemocytes were incubated in the presence of recombinant

by guest on July 18, 2020http://w

ww

.jbc.org/D

ownloaded from

7

serralysin, and membrane fractions were subjected to two-dimensional electrophoresis analysis. We found two protein spots that disappeared in response to serralysin treatment (Fig. 5B). We further cut out gel pieces that contained these proteins and performed amino acid sequence analysis. As a result, one protein contained clip domain serine protease 11 (BmSPH-1) precursor and the other protein contained a serine proteinase-like protein precursor (Fig. 5B). BmSPH-1 is involved in a cellular immune response called nodule formation, a process characterized as an aggregation of hemocytes that trap foreign substances and attach to internal tissues (27). Adhesion molecules of immunosurveillance cells play critical roles in bacterial engulfment. Based on our results suggesting that serralysin reduced cell adhesiveness, we considered that serralysin might inhibit bacterial phagocytosis by hemocytes. To test this hypothesis, we investigated the effect of serralysin on host cellular immune responses by in vitro phagocytosis assay against S. aureus as previously described (10). The numbers of live S. aureus cells incorporated by hemocytes were decreased in the presence of serralysin, and this effect was dependent on the serralysin concentration (Fig. 5C). Furthermore, we injected silkworms with a mixture of serralysin and S. aureus to examine the effect of serralysin on the transition of live bacterial counts in the bloodstream. When serralysin was co-injected with S. aureus, higher numbers of live bacteria were collected from the hemolymph compared to the control group injected with S. aureus alone (Fig. 5D). Insect cytokine PP induces morphologic alterations accompanied with cellular elongation in silkworm hemocytes (28). We previously reported that active PP also promotes the expression of phagocytosis-related genes in hemocytes (10). We therefore verified our assumption that PP increased the cell adhesiveness leading to a reduction of free hemocytes and that serralysin suppressed this process. The numbers of free hemocytes in IPS-injected larvae were further decreased by injection of active PP (Fig. 6A). In contrast, an inactive form of PP with a truncated N-terminus lacking ENF-residues (19) failed to show this

activity (Fig. 6A). On the other hand, recombinant serralysin inhibited the hemocyte reduction caused by both IPS and active PP injection (Fig. 6A). Furthermore, in vitro incubation of active PP in the presence of recombinant serralysin did not reduce the protein level of PP itself (Fig. 6B). Thus, we concluded that serralysin suppressed the PP-dependent hemocyte decrease without directly degrading PP in the silkworm hemolymph. Involvement of serralysin in silkworm killing by pathogenic bacteria—In general, adhesion of blood-circulating immune cells is involved in a variety of host immune responses and contributes to bacterial clearance and cellular migration to other tissues. Our results suggesting that serralysin decreased the adhesive properties of the silkworm hemocytes and suppressed cellular immune responses led us to assume that serralysin contributed to virulence against host animals. To verify the necessity of serralysin for the pathogenesis of S. marcescens, we constructed a disruption mutant of the serralysin gene and examined its host-killing effect. IPS culture supernatant of the mutated strain Δser had lower hemocyte number-increasing activity than the wild-type strain (Fig. 7A). When this Δser IPS culture supernatant was subjected to the fractionation procedure for serralysin purification as shown in the Table 1, the activity corresponding to peak 2 in the DEAE-Toyopearl column chromatography was abolished (Table 2, Fig. 7B). In addition, the protein with a molecular mass of 50 kDa was not detected in eluted fractions corresponding to peak 2. On the other hand, the activity corresponding to peak 1 was not significantly affected (Table 2); the total activity in DEAE-Toyopearl column chromatography from 3 independent cultures of Δser mutant was 530 ± 140 units (mean ± SD). We then evaluated the LD50 value of the Δser mutant against silkworm larvae. The LD50 value of Δser was approximately 5 times higher than that of the parent S. marcescens strain (Fig. 7C, D). On the other hand, co-injection of the recombinant serralysin protein with Δser S. marcescens suppressed the increase in the LD50 value (Fig. 7D). Moreover, in an S. aureus infection model, injection of the recombinant serralysin protein reduced the LD50 value of S. aureus against silkworms (Fig. 7E). Neither disruption of the

by guest on July 18, 2020http://w

ww

.jbc.org/D

ownloaded from

8

serralysin gene in S. marcescens nor the addition of serralysin protein to the S. aureus culture medium significantly affected the generation time of each bacterium (data not shown). Involvement of serralysin in S. marcescens virulence against mice—The results of the experiments described above demonstrated that serralysin attenuated the adhesive properties of silkworm hemocytes and that serralysin was required for silkworm killing by S. marcescens. We further examined whether serralysin was involved in S. marcescens virulence against mammals. Serralysin inhibited the adhesion of mouse peritoneal macrophages to polystyrene dishes (Fig. 8A). Moreover, two-dimensional electrophoresis revealed a protein spot in the membrane fraction of mouse peritoneal macrophages that disappeared in response to serralysin treatment (Fig. 8B). Amino acid sequence analysis revealed that this protein spot contained calreticulin (Fig. 8B), a multifunctional Ca2+-binding protein located on membrane surfaces of immune cells that is involved in cellular adhesion (29-32). The serralysin-dependent degradation of calreticulin in mouse macrophages was confirmed by Western blot analysis (Fig. 8C). Furthermore, the killing of mice infected intravenously with the serralysin gene-disrupted S. marcescens mutant was delayed compared with that of mice infected with wild-type S. marcescens (Fig. 8D), suggesting that serralysin is necessary for the pathogenicity of S. marcescens in mice. DISCUSSION

In both vertebrates and invertebrates, pathogenic microorganisms invading the host bloodstream are initially recognized by circulating immunosurveillance cells and then rapidly eliminated through early defense responses such as phagocytosis. Immunosurveillance cell dysfunction causes severe problems in infected hosts. In the present study, we found that the opportunistic pathogen S. marcescens reduced the adhesiveness of freely circulating immune cells in silkworms, leading to the suppression of host cellular immunity. We identified serralysin metalloprotease as a factor responsible for this activity. Moreover, our findings suggested that serralysin attenuated adhesion in murine

peritoneal macrophages and is involved in host killing by S. marcescens against silkworms and mice. In mammals, including humans, white blood cell (WBC) density is an indicator of inflammation. When inflammation occurs in mammals, granulocyte colony stimulation factor and other cytokines that promote cell growth of neutrophils and macrophages are produced, then WBC values increase within several days post-infection. On the other hand, in the early stage of infection, WBC adhere to the blood epithelia and migrate to damaged tissue, which transiently decreases WBC values (24). In silkworms, the injection of salt solutions or culture medium rapidly reduced the number of circulating hemocytes. Moreover, the insect cytokine PP dissolved in IPS induced a further decrease in hemocyte counts compared to IPS alone (Fig. 6A). These phenomena observed in silkworm hemocytes seem to correspond to the acute responses of WBC in mammals infected by pathogens. Serralysin is a metalloprotease that is conserved among various bacterial species. Proteus mirabilis, a pathogen that infects the human urinary tract, and an entomopathogenic bacteria, Photorhabdus luminiscens, possess serralysin homologs named ZapA (33) and PrtA (34), respectively. These serralysin family metalloproteases degrade humoral immune factors. For example, in vitro studies revealed that P. mirabilis ZapA degrades immunoglobulins IgA and IgG, and antimicrobial peptides such as human β-defensin 1 and LL-37 (35). To our knowledge, however, there are no reports demonstrating that serralysin affects “cellular” immune systems of host animals. The present study demonstrated that S. marcescens serralysin inhibits the adhesion of immune cells and suppresses bacterial clearance. Moreover, our results suggested that serralysin acts as a metalloprotease to degrade BmSPH-1, a factor involved in cellular immune responses, by mediating the adhesion of hemocytes to the tissue surface. Sakamoto et al. recently reported that activated BmSPH-1 localized within hemocyte aggregates when nodule formation is induced in silkworms infected by bacteria (27). That is, while BmSPH-1 is normally present in the hemolymph of non-infected silkworms, it is recruited to hemocyte surfaces after immunologic stress and

by guest on July 18, 2020http://w

ww

.jbc.org/D

ownloaded from

9

contributes to cellular adhesion and melanization processes within the aggregates (27). Based on our in vitro results, we considered that BmSPH-1 was activated through the hemocyte isolation process and recruited to cell surfaces, and then degraded by recombinant serralysin, leading to the suppression of cellular immune responses. Under our experimental conditions, it is possible that the degradation of other factors caused a serralysin-dependent inhibition of BmSPH-1 localization to the cell surface and subsequent immune reactions. Factors other than BmSPH-1 that are reported to be involved in the regulation of nodule formation are Noduler (36) and BmLBP C-type lectin (37), which directly bind to bacteria. These factors are also reported to be necessary for bacterial clearance via the formation of nodules at an early stage of infection, and thus they are potential serralysin target candidates. Moreover, the serine proteinase-like protein precursor identified through the analysis of protein spot #2 that disappeared following serralysin treatment (Fig. 5B) could be an uncharacterized factor involved in acute immune responses and targeted by serralysin. We previously demonstrated that the insect cytokine PP is converted into the active form via reactive oxygen species production by hemocytes stimulated with bacterial cell wall components (9). Moreover, we reported a genome-wide analysis of gene expression in the immune organs of silkworms injected with active PP (10). This microarray analysis revealed factors involved in cell adhesion among upregulated genes in the hemocytes of PP-injected silkworms (10). In the present study, active PP decreased the number of freely circulating hemocytes, whereas serralysin inhibited this process. Because serralysin did not directly degrade PP, we considered that serralysin degraded adhesion molecules on hemocytes induced by active PP, thereby leading to decreased adhesiveness of the immune cells to internal tissues. The relationship between PP-dependent innate immune responses and the nodule formation mentioned above is yet to be clarified. Because both reactions are triggered by bacterial cell wall components via host serine protease cascades within a relatively short time post-infection, at least some parts of these activation pathways may share common

mechanisms. One possible mechanism is that PP activated by bacteria rapidly induces cellular immune-factors (such as BmSPH-1) that promote hemocyte adhesion, followed by a decrease in the number of freely circulating cells in silkworm hemolymph. Further attempts to elucidate the relationship between PP and nodule formation is necessary to test this hypothesis. Artificial suppression of PP activation and cellular immune responses such as nodule formation by neutralizing antibodies or gene knockdown inhibits pathogen clearance in infected hosts (9,27,36). Studies of virulence factors that target host immunity would also shed light on the mechanisms underlying the immune system. Our findings indicate that serralysin was responsible, at least in part, for the hemocyte number-increasing activity of an IPS culture supernatant of S. marcescens. To our knowledge, this is the first study to identify a virulence factor that affects immune cell density in the bloodstream using a bioassay with whole animals. Cellular factors in the innate immune system have been intensively studied using invertebrates such as Drosophila melanogaster (17,38-41) and Caenorhabditis elegans (42,43). Because of its larger body size compared to these animals, however, we propose that the silkworm is a useful model for bioassays to identify virulence factors that suppress host cellular immunity. Because a band corresponding to serralysin was not observed in the peak 1 fraction and disruption of the serralysin gene did not significantly affect the activity of the peak 1 fraction obtained by DEAE-Toyopearl column chromatography (Table 2, Fig. 7B), factors other than serralysin may also contribute to this effect. We found two genes having high homology with the serralysin gene in the genome database of S. marcescens, and studies of their involvement in virulence are ongoing. Further studies to identify the remaining hemocyte-increasing factors would clarify the picture of cellular immune suppression by S. marcescens. Acknowledgements—This work was supported by a Grant-in-aid for Japan Society of the Promotion of Science Fellows 21-10519 from the Japan Society for the Promotion of Science (JSPS), and in part by JSPS Grant-in-Aid for Young Scientists (A) (24689008).

by guest on July 18, 2020http://w

ww

.jbc.org/D

ownloaded from

10

REFERENCES 1. Hejazi, A., and Falkiner, F. R. (1997) Serratia marcescens. J. Med. Microbiol. 46, 903-912 2. Villari, P., Crispino, M., Salvadori, A., and Scarcella, A. (2001) Molecular epidemiology of an

outbreak of Serratia marcescens in a neonatal intensive care unit. Infect. Control Hosp. Epidemiol. 22, 630-634

3. Maragakis, L. L., Winkler, A., Tucker, M. G., Cosgrove, S. E., Ross, T., Lawson, E., Carroll, K. C., and Perl, T. M. (2008) Outbreak of multidrug-resistant Serratia marcescens infection in a neonatal intensive care unit. Infect. Control Hosp. Epidemiol. 29, 418-423

4. Castelli, M. E., Fedrigo, G. V., Clementin, A. L., Ielmini, M. V., Feldman, M. F., and Garcia Vescovi, E. (2008) Enterobacterial common antigen integrity is a checkpoint for flagellar biogenesis in Serratia marcescens. J. Bacteriol. 190, 213-220

5. Castelli, M. E., and Vescovi, E. G. (2011) The Rcs signal transduction pathway is triggered by enterobacterial common antigen structure alterations in Serratia marcescens. J. Bacteriol. 193, 63-74

6. McMahon, K. J., Castelli, M. E., Garcia Vescovi, E., and Feldman, M. F. (2012) Biogenesis of outer membrane vesicles in Serratia marcescens is thermoregulated and can be induced by activation of the Rcs phosphorelay system. J. Bacteriol. 194, 3241-3249

7. Hoffmann, J. A., and Reichhart, J. M. (2002) Drosophila innate immunity: an evolutionary perspective. Nat. Immunol. 3, 121-126

8. Imamura, K., Ishii, K., Hamamoto, H., and Sekimizu, K. (2011) Hemocytes and humoral factors in silkworm blood are cooperatively involved in sheep erythrocyte aggregation. Drug Discov. Ther. 5, 125-129

9. Ishii, K., Hamamoto, H., Kamimura, M., and Sekimizu, K. (2008) Activation of the silkworm cytokine by bacterial and fungal cell wall components via a reactive oxygen species-triggered mechanism. J. Biol. Chem. 283, 2185-2191

10. Ishii, K., Hamamoto, H., Kamimura, M., Nakamura, Y., Noda, H., Imamura, K., Mita, K., and Sekimizu, K. (2010) Insect cytokine paralytic peptide (PP) induces cellular and humoral immune responses in the silkworm Bombyx mori. J. Biol. Chem. 285, 28635-28642

11. Kaito, C., Akimitsu, N., Watanabe, H., and Sekimizu, K. (2002) Silkworm larvae as an animal model of bacterial infection pathogenic to humans. Microb. Pathog. 32, 183-190

12. Kaito, C., Kurokawa, K., Matsumoto, Y., Terao, Y., Kawabata, S., Hamada, S., and Sekimizu, K. (2005) Silkworm pathogenic bacteria infection model for identification of novel virulence genes. Mol. Microbiol. 56, 934-944

13. Usui, K., Miyazaki, S., Kaito, C., and Sekimizu, K. (2009) Purification of a soil bacteria exotoxin

by guest on July 18, 2020http://w

ww

.jbc.org/D

ownloaded from

11

using silkworm toxicity to measure specific activity. Microb. Pathog. 46, 59-62 14. Ishii, K., Hamamoto, H., Imamura, K., Adachi, T., Shoji, M., Nakayama, K., and Sekimizu, K.

(2010) Porphyromonas gingivalis peptidoglycans induce excessive activation of the innate immune system in silkworm larvae. J. Biol. Chem. 285, 33338-33347

15. Ishii, K., Adachi, T., Imamura, K., Takano, S., Usui, K., Suzuki, K., Hamamoto, H., Watanabe, T., and Sekimizu, K. (2012) Serratia marcescens induces apoptotic cell death in host immune cells via a lipopolysaccharide- and flagella-dependent mechanism. J. Biol. Chem.

16. Zhuang, S., Kelo, L., Nardi, J. B., and Kanost, M. R. (2007) An integrin-tetraspanin interaction required for cellular innate immune responses of an insect, Manduca sexta. J. Biol. Chem. 282, 22563-22572

17. Kocks, C., Cho, J. H., Nehme, N., Ulvila, J., Pearson, A. M., Meister, M., Strom, C., Conto, S. L., Hetru, C., Stuart, L. M., Stehle, T., Hoffmann, J. A., Reichhart, J. M., Ferrandon, D., Ramet, M., and Ezekowitz, R. A. (2005) Eater, a transmembrane protein mediating phagocytosis of bacterial pathogens in Drosophila. Cell 123, 335-346

18. Kamimura, M., Nakahara, Y., Kanamori, Y., Tsuzuki, S., Hayakawa, Y., and Kiuchi, M. (2001) Molecular cloning of silkworm paralytic peptide and its developmental regulation. Biochem. Biophys. Res. Commun. 286, 67-73

19. Ishii, K., Adachi, T., Hamamoto, H., Oonishi, T., Kamimura, M., Imamura, K., and Sekimizu, K. (2013) Insect cytokine paralytic peptide activates innate immunity via nitric oxide production in the silkworm Bombyx mori. Dev. Comp. Immunol. 39, 147-153

20. Kobayashi, H. K., S.; Natori, S. (1991) Purification of the 200 kDa hemocyte membrane protein of Sarcophaga peregrina and its specific interaction with fat body. Insect Biochem. 21, 517-522

21. Toratani, T., Shoji, T., Ikehara, T., Suzuki, K., and Watanabe, T. (2008) The importance of chitobiase and N-acetylglucosamine (GlcNAc) uptake in N,N'-diacetylchitobiose [(GlcNAc)2] utilization by Serratia marcescens 2,170. Microbiology 154, 1326-1332

22. Watanabe, T., Kimura, K., Sumiya, T., Nikaidou, N., Suzuki, K., Suzuki, M., Taiyoji, M., Ferrer, S., and Regue, M. (1997) Genetic analysis of the chitinase system of Serratia marcescens 2170. J. Bacteriol. 179, 7111-7117

23. Hou, R. F. C., J. (1985) Cellular defense response to Beauveria bassiana in the silkworm, Bombyx mori. Appl. Ent. Zool. 20, 118-125

24. Qi, W., Ebbert, K. V., Craig, A. W., Greer, P. A., and McCafferty, D. M. (2005) Absence of Fer protein tyrosine kinase exacerbates endotoxin induced intestinal epithelial barrier dysfunction in vivo. Gut 54, 1091-1097

25. Chadwick, J. S., Caldwell, S. S., and Chadwick, P. (1990) Adherence patterns and virulence for Galleria mellonella larvae of isolates of Serratia marcescens. J. Invertebr. Pathol. 55, 133-134

by guest on July 18, 2020http://w

ww

.jbc.org/D

ownloaded from

12

26. Kida, Y., Inoue, H., Shimizu, T., and Kuwano, K. (2007) Serratia marcescens serralysin induces inflammatory responses through protease-activated receptor 2. Infect. Immun. 75, 164-174

27. Sakamoto, M., Ohta, M., Suzuki, A., Takase, H., Yoshizawa, Y., Kitami, M., and Sato, R. (2011) Localization of the serine protease homolog BmSPH-1 in nodules of E. coli-injected Bombyx mori larvae and functional analysis of its role in nodule melanization. Dev. Comp. Immunol. 35, 611-619

28. Nakahara, Y., Kanamori, Y., Kiuchi, M., and Kamimura, M. (2003) Effects of silkworm paralytic peptide on in vitro hematopoiesis and plasmatocyte spreading. Arch. Insect Biochem. Physiol. 52, 163-174

29. Ghiran, I., Klickstein, L. B., and Nicholson-Weller, A. (2003) Calreticulin is at the surface of circulating neutrophils and uses CD59 as an adaptor molecule. J. Biol. Chem. 278, 21024-21031

30. Rauch, F., Prud'homme, J., Arabian, A., Dedhar, S., and St-Arnaud, R. (2000) Heart, brain, and body wall defects in mice lacking calreticulin. Exp. Cell Res. 256, 105-111

31. Fadel, M. P., Dziak, E., Lo, C. M., Ferrier, J., Mesaeli, N., Michalak, M., and Opas, M. (1999) Calreticulin affects focal contact-dependent but not close contact-dependent cell-substratum adhesion. J. Biol. Chem. 274, 15085-15094

32. Cho, J. H., Homma, K., Kanegasaki, S., and Natori, S. (1999) Activation of human neutrophils by a synthetic anti-microbial peptide, KLKLLLLLKLK-NH2, via cell surface calreticulin. Eur. J. Biochem. 266, 878-885

33. Walker, K. E., Moghaddame-Jafari, S., Lockatell, C. V., Johnson, D., and Belas, R. (1999) ZapA, the IgA-degrading metalloprotease of Proteus mirabilis, is a virulence factor expressed specifically in swarmer cells. Mol. Microbiol. 32, 825-836

34. Bowen, D. J., Rocheleau, T. A., Grutzmacher, C. K., Meslet, L., Valens, M., Marble, D., Dowling, A., Ffrench-Constant, R., and Blight, M. A. (2003) Genetic and biochemical characterization of PrtA, an RTX-like metalloprotease from Photorhabdus. Microbiology 149, 1581-1591

35. Belas, R., Manos, J., and Suvanasuthi, R. (2004) Proteus mirabilis ZapA metalloprotease degrades a broad spectrum of substrates, including antimicrobial peptides. Infect. Immun. 72, 5159-5167

36. Gandhe, A. S., John, S. H., and Nagaraju, J. (2007) Noduler, a novel immune up-regulated protein mediates nodulation response in insects. J. Immunol. 179, 6943-6951

37. Watanabe, A., Miyazawa, S., Kitami, M., Tabunoki, H., Ueda, K., and Sato, R. (2006) Characterization of a novel C-type lectin, Bombyx mori multibinding protein, from the B. mori hemolymph: mechanism of wide-range microorganism recognition and role in immunity. J. Immunol. 177, 4594-4604

38. Cuttell, L., Vaughan, A., Silva, E., Escaron, C. J., Lavine, M., Van Goethem, E., Eid, J. P., Quirin,

by guest on July 18, 2020http://w

ww

.jbc.org/D

ownloaded from

13

M., and Franc, N. C. (2008) Undertaker, a Drosophila Junctophilin, links Draper-mediated phagocytosis and calcium homeostasis. Cell 135, 524-534

39. Ramet, M., Manfruelli, P., Pearson, A., Mathey-Prevot, B., and Ezekowitz, R. A. (2002) Functional genomic analysis of phagocytosis and identification of a Drosophila receptor for E. coli. Nature 416, 644-648

40. Stroschein-Stevenson, S. L., Foley, E., O'Farrell, P. H., and Johnson, A. D. (2006) Identification of Drosophila gene products required for phagocytosis of Candida albicans. PLoS Biol. 4, e4

41. Stuart, L. M., and Ezekowitz, R. A. (2008) Phagocytosis and comparative innate immunity: learning on the fly. Nat. Rev. Immunol. 8, 131-141

42. Reddien, P. W., and Horvitz, H. R. (2000) CED-2/CrkII and CED-10/Rac control phagocytosis and cell migration in Caenorhabditis elegans. Nat. Cell Biol. 2, 131-136

43. Means, T. K. (2010) Fungal pathogen recognition by scavenger receptors in nematodes and mammals. Virulence 1, 37-41

FIGURE LEGENDS TABLE 1. Purification of the hemocyte-increasing factor secreted by S. marcescens. TABLE 2. Purification of the hemocyte-increasing factor secreted by S. marcescens Δser mutants. A representative result of 3 independent experiments is shown. FIGURE 1. Decrease in free hemocytes by injection stimulation, and suppression of this response by S. marcescens. A, Alteration of hemocyte density after wounding or liquid injection. Silkworm larvae (day 2 of 5th instar, 2 g/larva) were either injured by needles or injected with 50 µl of liquid samples (Milli-Q water, saline, LPS-free saline, insect physiologic saline [IPS], 50 mM Tris/HCl [pH 8.0], 10 mg/ml bovine serum albumin, LB10 culture medium, or BHI culture medium). After 0.5 h, hemolymph was collected, and the number of hemocytes was counted under a microscope. Data represent mean ± SDs of 4-6 larvae. Statistical analysis was performed using one-way analysis of variance (ANOVA) with Tukey’s multiple comparison tests. Statistically significant differences were observed between columns with different letters (a-d; p<0.05). B, Ten silkworms per group were injected with 100 µl of either BHI culture medium, live bacterial suspensions (1010 cells/ml) of S. aureus, E. coli, or S. marcescens. Hemolymph collected from 10 silkworms after 3 h were pooled, and hemocyte density was determined under a microscope. Data represent mean ± SDs of 3-4 experiments. Statistical analysis was performed using a one-way ANOVA, and statistical differences compared with the control “BHI medium” group were analyzed by Dunnett’s multiple comparison test (*, p<0.001). C, Silkworms were injected with 100 µl of either culture medium (open circle) or a filtered fraction of an overnight BHI-culture supernatant of S. marcescens (closed circle), and the number of free hemocytes was counted at the indicated time-points. Data represent mean ± SDs of 5 larvae. Statistical analysis was performed using Student’s t-test between two groups at each time-point (#, p<0.005; *, p<0.001). D, Effect of heat treatment on hemocyte-increasing activity of the BHI-culture supernatant of S. marcescens. Silkworms were injected with 100 µl of either a filtered fraction of an overnight BHI-culture

by guest on July 18, 2020http://w

ww

.jbc.org/D

ownloaded from

14

supernatant of S. marcescens (open circle, “Control”) or the supernatant heated at 100°C for 30 min (closed circle, “Heat-treated”). After 0.5 h, the hemolymph was collected, and hemocyte numbers were counted. Data represent mean ± SDs of 4 larvae. Statistical analysis was performed using Student’s t-test between two groups at injection volumes of 13, 25, 50, and 100 µl (#, p<0.01; *, p<0.001). FIGURE 2. Hemocyte number-increasing activities of eluted fractions of culture supernatant of S. marcescens from DEAE-Toyopearl column. A, Silkworm hemocyte-increasing activity of S. marcescens IPS-culture supernatant. Overnight-cultured S. marcescens cells were suspended in IPS (1011 cells/ml) and statically incubated at 30°C for 1 d. The suspension was then centrifuged and the supernatant was filtered through a 0.22-µM filter. Three larvae per group (day 2 of 5th instar) were injected with 100 µl of each diluted sample, and the hemolymph was collected after 0.5 h. Hemolymph from 3 larvae was pooled and hemocyte numbers were counted under a microscope. Activity with a 1.5-fold increase in the number of hemocytes relative to that obtained from IPS-injected silkworms was defined as 1 unit. B, Effect of heat treatment on hemocyte-increasing activity of the IPS-culture supernatant of S. marcescens. Silkworms were injected with 100 µl of either a filtered fraction of an IPS-culture supernatant of S. marcescens or the supernatant heated at 100°C for 30 min. After 0.5 h, the hemolymph was collected, and hemocyte numbers were counted. Data represent mean ± SDs of 3 larvae. Statistical analysis was performed by one-way ANOVA with Tukey’s multiple comparison tests (*, p<0.01). C, Hemocyte number-increasing activities (filled squares) and amounts of protein (open circles) in eluted fractions of culture supernatant of S. marcescens from the DEAE-Toyopearl column. Dashed lines indicate the NaCl concentration. D-E, Effect of heat treatment on hemocyte-increasing activity of eluted fractions of S. marcescens culture supernatant from the DEAE-Toyopearl column. Peak 1 or peak 2 fractions obtained by DEAE-Toyopearl column chromatography were heated at 100°C for 30 min. Silkworms were injected with 100 µl of heat-treated peak 1 (D) or peak 2 (E) fraction. After 0.5 h, the hemolymph was collected, and hemocyte numbers were counted. Data represent mean ± SDs of 3 larvae. Statistical analysis was performed using Student’s t-test (D) or one-way ANOVA with Tukey’s multiple comparison test (E) (*, p<0.01; #, p<0.001). F, Fraction Nos. 12-19 of the above DEAE-Toyopearl column chromatography were subjected to SDS-polyacrylamide gel electrophoresis, and the gel was stained with CBB. M, molecular weight markers. FIGURE 3. Hemocyte number-increasing activities of eluted fractions of culture supernatant of S. marcescens from the Superose 12 gel-filtration column. A, Peak 2 fraction obtained from DEAE-Toyopearl column chromatography of S. marcescens culture supernatant was further applied to a Superose 12 gel-filtration column. Hemocyte number-increasing activities (filled squares) and amounts of protein (open circles) in eluted fractions are shown. B, Fraction Nos. 19-30 of the above Superose 12 column chromatography were subjected to SDS-polyacrylamide gel electrophoresis, and the gel was stained with CBB. M, molecular weight markers. FIGURE 4. Expression of recombinant serralysin protein in E. coli and its hemocyte number-increasing activity. A, Expression of recombinant serralysin protein in transformed E. coli cells. E. coli harboring the expression plasmid were incubated in the presence of isopropyl-β-D-thigalactopyranoside, and insoluble fractions were prepared from cell lysates. Samples were then solubilized in 8 M urea and subjected to SDS-polyacrylamide gel electrophoresis. “pET28a”, E. coli cells transformed with an empty vector; “pET28a/ser”, E. coli cells transformed with a serralysin ORF-containing vector; “P”, insoluble fractions

by guest on July 18, 2020http://w

ww

.jbc.org/D

ownloaded from

15

of cell lysates; “S”, urea-soluble fractions. B, Fractionation of recombinant serralysin protein by Ni-affinity column chromatography. The above urea-soluble fractions were applied to Probond resin. The column was washed and eluted as described in the Methods section. FT, flow-through fraction; W, washed fraction; E1-5, eluted fractions. C, Hemocyte-increasing activity of recombinant serralysin protein. The E1 fraction of the above Ni-affinity column chromatography was dialyzed in IPS and injected into silkworm hemolymph. Three larvae (day 2 of 5th instar) were injected with 100 µl of each sample, and after 0.5 h hemolymph were collected and pooled. Hemocyte numbers were counted under a microscope. Cell counts relative to that of IPS-injected group are shown. D, Effect of heat treatment on hemocyte-increasing activity of the recombinant serralysin protein. Silkworms were injected with 100 µl of either purified recombinant serralysin (2.6 µM) or the protein sample heated at 100°C for 30 min. After 0.5 h, the hemolymph was collected, and hemocyte numbers were counted. Data represent mean ± SDs of 3 larvae. Statistical analysis was performed by one-way ANOVA with Tukey’s multiple comparison tests (*, p<0.001). E, Effect of EDTA on the hemocyte-increasing activity of the recombinant serralysin protein. Recombinant serralysin was incubated in the presence of 50 mM EDTA at 4°C for 3 h according to a previous report (26), and then the sample was injected into silkworm larvae (100 µl/larva). After 0.5 h, hemocyte numbers were counted as above. Data represent mean ± SDs of 4-5 larvae. Statistical differences between EDTA(-) and EDTA(+) at each serralysin concentration were analyzed using Student’s t-test (*, p<0.005). FIGURE 5. Suppression of hemocyte-dependent cellular immune responses by serralysin. A, Inhibition of hemocyte attachment to solid surfaces by serralysin. Silkworm hemocytes suspended in IPS were incubated in cell culture plates with or without recombinant serralysin protein at 27°C for 3 h. Plates were mildly shaken, and hemocytes collected in aliquots were counted as detached cells. Data represent mean ± SDs of 4 experiments (*, p<0.05). B, Alterations in protein patterns of hemocyte membrane fractions by serralysin. Silkworm hemocytes were suspended in IPS and incubated in the presence of recombinant serralysin (700 nM) at 27°C for 1 h. Membrane fractions of hemocytes were subjected to two-dimensional electrophoresis. CBB-stained gels of either the control or serralysin-treated group are shown in the upper or middle panels, respectively. Areas surrounding spots altered by serralysin treatment are magnified in the black frames. These pattern alterations were reproducibly observed in 3 experiments. Lower panels show results of MASCOT analysis of amino acid sequences obtained from protein extraction of gels indicated as protein spots #1 and #2. Gene products showing the highest identity with each sequence in the database are indicated, and sequences matching those obtained from each spot are written in red. C, Effect of serralysin on phagocytosis of S. aureus by silkworm hemocytes in vitro. Hemocytes were suspended in PBS and incubated with S. aureus in the presence of recombinant serralysin (rSer) at 27°C for 1 h. Hemocytes were then lysed, and numbers of incorporated S. aureus were determined. Bacterial numbers per hemocyte relative to those of non-reacted control groups incubated without rSer at 4°C are indicated as the phagocytic index. Data represent mean ± SDs of 4 experiments (*, p<0.01). D, Effect of serralysin on changes in live cell density of S. aureus in silkworm hemolymph in vivo. S. aureus cells suspended with or without recombinant serralysin protein were injected into silkworm larvae, and changes in the live bacterial cell density in the hemolymph were monitored at the indicated time-points. Data represent mean ± SDs of 6 larvae. Statistical differences were analyzed between two groups at each time point using Student’s t-test (*, p<0.01). FIGURE 6. Decrease in the density of silkworm hemocytes by the insect cytokine PP, and suppression of this effect by serralysin. A, Alterations of the hemocyte density in silkworm hemolymph induced by active PP. Silkworms were injected with 50 µl of either IPS, active PP (1 µM), a truncated form of PP lacking the N-terminal ENF

by guest on July 18, 2020http://w

ww

.jbc.org/D

ownloaded from

16

residues (PPΔENF, 1 µM), or a mixture of active PP (1 µM) and recombinant serralysin (800 nM), and after 0.5 h the density of hemocytes was determined. Data represent mean ± SDs of 4-5 silkworms. Statistical analysis was performed using a one-way ANOVA with Tukey’s multiple comparison tests. Statistically significant differences were observed between columns with different letters (p<0.05). B, Effect of serralysin on the protein level of active PP. Active PP (1 µM) was incubated in IPS for 1 h with or without recombinant serralysin (800 nM). A 5-fold serial dilution of each sample was applied to electrophoresis, and PP was detected by Western blot analysis. Ref, 10 ng of chemically synthesized active PP. FIGURE 7. Effect of serralysin on silkworm killing by bacteria. A, Requirement of the serralysin gene for the increase in the cell density of hemocytes induced by S. marcescens. IPS supernatant of either S. marcescens wild-type (WT) or the disrupted mutant of the serralysin gene (Δser) was injected into silkworms (100 µl/larva), and after 0.5 h the cell density of hemocytes in the hemolymph collected from 3 larvae per group was determined. Data represent hemocyte counts relative to that of the IPS-injected group. B, Hemocyte number-increasing activities of eluted fractions of culture supernatant of the serralysin gene-disrupted mutant from the DEAE column. Hemocyte number-increasing activities (filled squares) and amounts of protein (open circles) in eluted fractions from the DEAE-Toyopearl column using culture supernatant of the Δser mutant as the starting material. Dashed lines indicate the NaCl concentration. C, Requirement of the serralysin gene in silkworm killing by S. marcescens. Serial diluted suspension of either S. marcescens wild-type (WT) or the disrupted mutant of the serralysin gene (Δser) was injected into silkworms in each group, and survival rates after 16 h were determined. Each point represents the survival rate of 5 larvae infected with the indicated number of S. marcescens. D, Effect of co-injection of serralysin on silkworm killing by S. marcescens disruption mutant of the serralysin gene. Silkworms were injected with either S. marcescens wild-type cells suspended in IPS (WT, IPS), the disrupted mutant of the serralysin gene (Δser), or Δser mutant suspension supplemented with recombinant serralysin protein (Δser, rSer). LD50 values were determined from survival rates monitored after 16 h of infection. Data represent mean ± SDs of 4-5 experiments. One-way ANOVA with Tukey’s multiple comparison test was performed, and statistically significant differences were observed between columns with different letters (a or b; p<0.05). E, Effect of serralysin on silkworm killing by S. aureus. Silkworm larvae (day 2 of 5th instar) were injected with S. aureus suspension supplemented with recombinant serralysin protein. LD50 values were determined from survival rates monitored after 24 h of infection. Data represent mean ± SDs of 4 experiments. Statistical analysis was performed using Student’s t-test (*, p<0.05). FIGURE 8. Involvement of serralysin in S. marcescens virulence against mice. A, Inhibition of murine peritoneal macrophage attachment to solid surfaces by serralysin. Macrophages suspended in PBS were pre-incubated with or without recombinant serralysin (800 µM) at 37°C for 1 h. Cell suspensions were diluted and placed into cell culture dishes, and aliquots were removed after incubation at 37°C for 2 h. Cells detached from the plate surfaces by cell scrapers were counted. Data represent mean ± SDs of triplicates (*, p<0.05). B, Alterations in protein patterns of macrophage membrane fractions by serralysin. Mouse peritoneal macrophages were suspended in IPS and incubated in the presence of recombinant serralysin (700 nM) at 37°C for 1 h. Membrane fractions of macrophages were subjected to two-dimensional electrophoresis. CBB-stained gels of either the control or serralysin-treated groups are shown in the upper or middle panels, respectively. The spot altered by serralysin treatment is magnified in the black frames. Lower panel shows the result of MASCOT analysis of amino acid sequences obtained from protein extraction of the gel. The gene product showing the highest identity with the obtained sequence is indicated, and the matched sequences are written in red.

by guest on July 18, 2020http://w

ww

.jbc.org/D

ownloaded from

17

C, Degradation of calreticulin in mouse macrophages by serralysin. Western blot analysis for calreticulin was performed using membrane fractions of mouse peritoneal macrophages treated with recombinant serralysin as above. Beta-actin was detected as a loading control. D, Requirement of the serralysin gene of S. marcescens for mouse killing. Live bacteria suspension (200 µl; 3 x 108 CFU/ml) of either wild-type (WT) or the disrupted serralysin gene mutant (Δser) was intravenously injected into mice (N=6), and survival rates were monitored. A log-rank test revealed a significant difference between the survival curves of “WT” and “Δser” (p<0.05).

by guest on July 18, 2020http://w

ww

.jbc.org/D

ownloaded from

Purification steps Total activity(units)

Protein(μg)

Specific activity(units/μg)

Yield(%)

Purification(fold)

I. IPS supernatant 25703400 0.76 100 1

II. Phenyl-Toyopearl 2000960 2.1 78 3

III. Hydroxyapatite 92060 15 36 20

IV. DEAE-Toyopearl peak 1 (fr. 6-9) 7605 160 30 210

peak 2 (fr. 14-15) 27023 12 11 16

Table 1

18

by guest on July 18, 2020http://w

ww

.jbc.org/D

ownloaded from

Purification steps Total activity(units)

Protein(μg)

Specific activity(units/μg)

Yield(%)

Purification(fold)

I. IPS supernatant 13003240 0.40 100 1

II. Phenyl-Toyopearl 440720 0.61 34 2

IV. DEAE-Toyopearl

4008 53 31 130

fr. 14-15 (corresponding topeak 2 in the parent strain) <80<3 <6

Table 2

19

fr. 4-7 (corresponding topeak 1 in the parent strain)

III. Hydroxyapatite 360120 3.0 28 8

by guest on July 18, 2020http://w

ww

.jbc.org/D

ownloaded from

Den

sity

of h

emoc

ytes

(x 1

06 cel

ls/m

l)

A

C D

Figure 1

0

2

4

6

8

10

BHImedium

S. aureus E. coliS. mar-cescens

*B

WaterSalin

e

LPS-free sa

line

Insect

physi

ologic

al salin

e

Tris buffe

r

Bovine

serum

album

in

LB10

med

ium

BHI med

ium0

2

4

6

Den

sity

of h

emoc

ytes

(x 1

06 cel

ls/m

l)

No treatm

ent

Needle injury

c

a

b

d

cc

cc c

d

20

Den

sity

of h

emoc

ytes

(x 1

06 cel

ls/m

l)

Time post injection (h)

0

2

4

6

8

10

0 2 4 6 8 10 12 14 16 18 20 22 24

*

S. mar BHI-sup BHI medium

**

#

**

0 20 40 60 80 100S. marcescens BHI-culture supernatant (μl)

Control Heat-treated

Den

sity

of h

emoc

ytes

(x 1

06 cel

ls/m

l)

0

2

4

6

8

10

*#*

#

by guest on July 18, 2020http://w

ww

.jbc.org/D

ownloaded from

0.0

1.0

2.0

3.0

4.0

0

10

20

30

40

50

1 5 10 15 20 25 30 Fraction number (4 ml/fr.)

Act

ivity

(uni

ts/m

l) Protein (μg/m

l)

NaC

l (M)

0.0

0.1

0.2

0.3

C

FM 12 13 14 15 16 17 18 19 M

Fraction number

(kDa)

6645

26

14

Figure 2

peak 1 peak 2

21

Protein (μg)

Den

sity

of h

emoc

ytes

(rel

ativ

e to

IPS

inje

ctio

n)

1.0

1.5

2.0

2.5

0 2 4 6 8 10 12 14 16

A

0

2

4

6

8

Den

sity

of h

emoc

ytes

(x 1

06 cel

ls/m

l) * *B

0

2

4

6

8

0

1

2

3

4

5

Den

sity

of h

emoc

ytes

(x 1

06 cel

ls/m

l)

Den

sity

of h

emoc

ytes

(x 1

06 cel

ls/m

l)

IPSNon-treated

Heat-treated

*# #

Peak 1 fr.

Non-treated

Heat-treated

Peak 2 fr.

D E

IPS Non-treated

Heat-treated

IPS-culture sup

by guest on July 18, 2020http://w

ww

.jbc.org/D

ownloaded from

10 15 20 25 30 35 40 45 47Fraction number (0.5 ml/fr.)

1000

800

600

400

200

0

Act

ivity

(uni

ts/m

l) Protein (μ g/m

l)400

300

200

100

0

A

B19 20 21 22 23 24 25 26 27 28 3029 MM

4836

28

64

(kDa)

Fraction number

Figure 3

22

by guest on July 18, 2020http://w

ww

.jbc.org/D

ownloaded from

SSSSP P PPM

AFT W E1 E2 E3 E5E4 M

B

C

Recombinant Serralysin (μg)

Den

sity

of h

emoc

ytes

(r

elat

ive

to IP

S in

ject

ion)

IPTG+- +-pET28a/serpET28a

Figure 4

1.0

1.5

2.0

2.5

3.0

3.5

4.0

0 2 4 6 8 10 12 14

0

1

2

3

4

5

6

0 0.5 1 2 4

Den

sity

of h

emoc

ytes

(x 1

06 cel

ls/m

l)

Recombinant Serralysin (μg)

EDTA (-) EDTA (+)

D

**

*

624835

28

21

(kDa)

624835

28

21

(kDa)

23

0

2

4

6

8

10

IPS rSer rSer

* *

(heat-treated)

Den

sity

of h

emoc

ytes

(x 1

06 cel

ls/m

l)

E

by guest on July 18, 2020http://w

ww

.jbc.org/D

ownloaded from

Num

ber o

f det

ache

dhe

moc

ytes

(x 1

05 cel

ls/m

l)

Den

sity

of l

ive

bact

eria

in

silk

wor

m h

emol

ymph

(x

105 C

FU/m

l)

A

C D

Figure 5

0 7 70 700Recombinant Serralysin (nM)

0

5

10

15

20

25 **

B Recombinant Serralysin (-)

1 2

Spot #1; clip domain serine protease 11 (BmSPH-1)precursor (gi|112983100)

Spot #2; serine proteinase-like protein precursor (gi|114052256)

Recombinant Serralysin (+)

24

0 7 70 7000

1

2

3

4

Pha

gocy

totic

inde

x

Recombinant Serralysin (nM)

* *

1 2

1

10

100

1000

0 2 4 6 8 10 12 14 16TIme post injection (h)

rSer (-)rSer (+)

* * *

*

by guest on July 18, 2020http://w

ww

.jbc.org/D

ownloaded from

No inje

ction IPS PP

PPΔENF

rSer

PP + rS

er0

1

2

3

4

5

Den

sity

of h

emoc

ytes

(x 1

06 cel

ls/m

l)

a

c

aa

b b

A

Ref

B PP PP + rSer

Figure 6

25

by guest on July 18, 2020http://w

ww

.jbc.org/D

ownloaded from

0

20

40

60

80

100

1 10 100 1000

WT

Number of injected bacteria (CFU/larva)

Sur

viva

l rat

e of

silk

wor

ms

(%)

Δser

0.0

1.0

2.0

3.0

4.0

1 10 100

Den

sity

of h

emoc

ytes

(rel

ativ

e to

IPS

inje

ctio

n)

Volume of IPS supernatant (μl)

WT Δser

0

A

D E

rSer (-) rSer (+) LD50

of S

. aur

eus

(x 1

07 CFU

/larv

a)

Figure 7

*

0

1

2

3

4

5

0

20

40

60

80

100

WT, IPS Δser, IPS Δser, rSer

LD50

of S

. mar

cesc

ens

(CFU

/larv

a)

B

a

a

b

C

50

40

30

20

10

0A

ctiv

ity (u

nits

/ml) P

rotein (μ g/ml)

4.0

3.0

2.0

1.0

0.01 15 20 25 305 10

Fraction number (4 ml/fr.)

NaC

l (M)

0.0

0.1

0.2

0.3

by guest on July 18, 2020http://w

ww

.jbc.org/D

ownloaded from

0

20

40

60

80

100

0 5 10 15 20 25

Sur

viva

l rat

e of

mic

e (%

)

Time post infection (h)

WT ΔserPBS

Figure 8

0

1

2

3

4

5

6

rSer (-) rSer (+)

Num

ber o

f atta

ched

mac

roph

ages

(x 1

06 cel

ls/m

l)

*

A D

B Recombinant Serralysin (-)

Spot #1; calreticulin precursor (gi|6680836)

Recombinant Serralysin (+)

1

27

Calreticulin

Beta-actin

+-rSerC

by guest on July 18, 2020http://w

ww

.jbc.org/D

ownloaded from

Kenichi Ishii, Tatsuo Adachi, Hiroshi Hamamoto and Kazuhisa SekimizuAdhesion-Inhibitory Factor Against Immunosurveillance Cells

Serratia marcescens Suppresses Host Cellular Immunity via the Production of an

published online January 7, 2014J. Biol. Chem. 

  10.1074/jbc.M113.544536Access the most updated version of this article at doi:

 Alerts:

  When a correction for this article is posted• 

When this article is cited• 

to choose from all of JBC's e-mail alertsClick here

by guest on July 18, 2020http://w

ww

.jbc.org/D

ownloaded from