Syndecan-4 deficiency leads to high mortality of lipopolysaccharide-injected mice.

Syndecan-4 is a transmembrane heparan sulfate proteoglycan belonging to the syndecan family. Following intraperitoneal injection of lipopolysaccharide (LPS), syndecan-4-deficient mice exhibited high mortality compared with wild-type controls. Severe endotoxin shock was observed in the deficient mice: systolic blood pressure and left ventricular fractional shortening were lower in the deficient mice than in the wild-type controls 9 h after LPS injection. Although histological examinations revealed no apparent differences between two groups, the plasma level of interleukin (IL)-1beta was higher in the deficient mice than in the wild-type controls 9 h after LPS injection. Consistent with the regulatory roles of syndecan-4, its expression in monocytes and endothelial cells of microvasculature increased in the wild-type mice after LPS administration. Although IL-1beta was produced to the same extent by macrophages from syndecan-4-deficient and wild-type mice after LPS stimulation, inhibition of its production by transforming growth factor-beta1 was impaired in the syndecan-4-deficient macrophages. These results indicate that syndecan-4 could be involved in prevention of endotoxin shock, at least partly through the inhibitory action of transforming growth factor-beta1 on IL-1beta production.

Septic shock is defined as sepsis with hypotension or multiple systemic organ failure and is a common cause of death in patients in the intensive care unit. Lipopolysaccharide (LPS) 1 is one of the toxic principles of Gram-negative bacteria (1), and is a classic example of an initiator of septic shock. Shock triggered by LPS administration is often called endotoxin shock.
Cytokines, such as tumor necrosis factor (TNF)-␣ and interleukin (IL)-1␤, are produced in response to infection or LPS administration. Production of these cytokines has been shown to be important for the cascade leading to endotoxin shock (2)(3)(4).
Other classes of cytokines such as IL-10 and transforming growth factor (TGF)-␤1 play key roles in decreasing susceptibility to endotoxin shock, principally by suppressing the expression of proinflammatory cytokines (5,6). However, much remains to be clarified concerning the regulatory mechanisms for avoiding endotoxin shock.
Syndecan-4 (also called ryudocan) is a transmembrane heparan sulfate proteoglycan (7,8) belonging to the syndecan family, which consists of 4 members (9). Heparan sulfate chains of syndecans bind to growth/differentiation factors, anticoagulant factors, and cell adhesion molecules, and have been suggested to participate in various biological phenomena (9,10). Mice deficient in syndecan genes provide powerful means to analyze the functions of syndecan molecules in vivo (11)(12)(13)(14)(15). Thus, syndecan-1 has been shown to participate in Wnt-1 signaling (14) and enhancement of microbial virulence (15), while syndecan-4 in anticoagulation in fetal vessels of placental labyrinth (11) and prevention of -carrageenan deposition in the kidney (12). Here, we described an unexpected finding that syndecan-4 is involved in a mechanism avoiding endotoxin shock.
LPS Administration-LPS (Escherichia coli serotype O111:B4) was obtained from Sigma-Aldrich, and dissolved in autoclaved physiological saline at a concentration of 250 or 500 g/ml. LPS solution was injected intraperitoneally into Synd4(ϩ/ϩ) or Synd4(Ϫ/Ϫ) or commercially obtained C57BL/6J mice (Clea  These dimensions were determined  by left ventricular M-mode echocardiography with an Acuson system  equipped with a 13-MHz transducer (Sequia Ultrasound System,  Mountain View, CA). During the echocardiography, mice were anesthetized with pentobarbital sodium at the dose of 25 mg/kg, positioned prone on a warmed saline bag, and applied with the transducer from below to avoid preventing respiration.
Plasma Cytokine Assay-Blood was taken from the heart into heparinized syringes under mild anesthesia with diethyl ether at the indicated times after LPS injection. Plasma was separated by centrifugation at 5,000 ϫ g for 10 min. Concentrations of TNF-␣ and IL-1␤ in plasma were determined with immunoassay kits (TFB, Tokyo, Japan).
After incubation with the first antibody at room temperature for 1 h or at 4°C overnight, sections were incubated with anti-rabbit IgG antibody conjugated with biotin, followed by avidin-biotin-peroxidase complex (Vector Laboratories, Burlingame, CA). The staining was visualized with diaminobenzidine tetrahydrochloride-Ni 3ϩ , Co 2ϩ (Amersham Pharmacia Biotech, Tokyo, Japan).
For analysis by confocal laser scanning microscopy (Bio-Rad), antirabbit IgG antibody conjugated with fluorescein isothiocyanate (Sigma-Aldrich) and streptoavidin-cytochrome 5 (Pharmingen) were used as the second antibodies.
Syndecan-4 Assay in the Liver after LPS Injection-Livers were obtained from 8-week-old Synd4(ϩ/ϩ) mice before, 1, 3, and 9 h after intraperitoneal injection of LPS at a dose of 5 mg/kg. Livers were also obtained from 8-week-old Synd4(Ϫ/Ϫ) mice before and 3 h after LPS injection. The samples were homogenized in Dulbecco's phosphatebuffered saline without calcium and magnesium (PBS(Ϫ)) containing 5 mM EDTA and 1 mM phenylmethylsulfonyl fluoride, and then the homogenate was centrifuged at 100,000 ϫ g for 1 h at 4°C. The precipitate was homogenized in PBS(Ϫ) containing 5 mM EDTA, 1 mM phenylmethylsulfonyl fluoride, and 1% Nonidet P-40, and then the homogenate was centrifuged at 100,000 ϫ g for 1 h at 4°C. Total protein concentration of the supernatant was determined with BCA protein assay kit (Pierce), adjusted to 1 mg/ml in PBS(Ϫ) containing 0.5 mM EDTA, 0.1 mM phenylmethylsulfonyl fluoride, and 0.1% Nonidet P-40, and then the supernatant was used as the membrane fraction of the liver. Onehundred l of the membrane fraction derived from each sample was applied in a well of a 96-well plate (Becton Dickinson, Lincoln Park, NJ), and the plates were incubated at 4°C overnight. After washing four times with PBS(Ϫ) containing 0.1% Tween 20, the plates were incubated with 1% bovine serum albumin (BSA) in PBS(Ϫ) at room temperature for 1 h for blocking, and then with anti-syndecan-4 antibody, diluted with PBS(Ϫ) containing 1% BSA (1:200), or with PBS(Ϫ) containing 1% BSA alone as a control. After washing, 96-well plates were incubated with anti-rabbit IgG antibody conjugated with horseradish peroxidase. Visualization was performed with tetramethylbenzidine at room temperature for 30 min. The optical density (OD) at 450 nm was determined with ImmunoMINI NJ-2300 (InterMed, Tokyo, Japan). From the observed OD values, OD values of the controls, which were around 0.070, were subtracted to calculate syndecan-4 immunoreactivity.
Flow Cytometry-To isolate peripheral white blood cells, ammonium chloride was used for lysis of red blood cells. Cells were incubated at 4°C for 15 min in 2 g/ml of Fc blocker (anti-CD32/16 antibody, Pharmingen), stained at 4°C for 30 min with anti-syndecan-4 antibody and biotylated anti-Mac-1 antibody or anti-Gr-1 antibody in PBS(Ϫ) containing 2% fetal calf serum, and then with anti-rabbit IgG antibody conjugated with fluorescein isothiocyanate and streptoavidin-cytochrome 5. A FACS Calibur flow cytometer and Cell Quest software (Becton Dickinson) were used to analyze the stained cells.
Binding Assay of 125 I-Labeled TGF-␤1 to Peritoneal Macrophages-Recombinant human TGF-␤1 without carrier protein was purchased from Genzyme-Techne (Cambridge, MA) and labeled with Na 125 I by the chloramine-T method. The labeled TGF-␤1 was separated from Na 125 I by gel filtration through a PD-10 desalting column (Amersham Pharmacia Biotech). Approximately 1.5 ϫ 10 8 cpm/g of 125 I-labeled TGF-␤1 was recovered. Autoradiography of the labeled material after 10% SDSpolyacrylamide gel electrophoresis showed an expected band with a molecular mass of 25 kDa (data not shown).
Resident peritoneal exudative cells were collected by peritoneal lavage with RPMI-HEPES containing 1% fetal calf serum in 10-cm culture dishes (Becton Dickinson) and incubated for 2 h at 37°C in 5% CO 2 . After washing twice with RPMI-HEPES, adherent cells were further incubated in RPMI-HEPES containing 1% autologous serum and 100 ng/ml LPS for 3 h. After washing twice with PBS(Ϫ), the cells were incubated in 1 ml of PBS(Ϫ) containing 0.2% EDTA and 5% fetal calf serum for 10 min at 4°C and then harvested by mild scraping. After washing the cells with PBS(Ϫ) containing 1% BSA, viable cells were counted with the aid of trypan blue staining, adjusted to 1 ϫ 10 5 in 500 l of PBS(Ϫ) containing 1% BSA, and incubated with 125 I-labeled TGF-␤1 (1 ϫ 10 5 cpm) for 30 min at 4°C in the absence or presence of heparin (10 g/ml). After washing three times with PBS containing 1% BSA, the amount of 125 I-labeled TGF-␤1 bound to the cells was quantified with an Aloka ␥-counter (Aloka, Tokyo, Japan).
Binding of Syndecan-4 to TGF-␤1-Livers were obtained from 8-week-old Synd4(ϩ/ϩ) or Synd4(Ϫ/Ϫ) mice 3 h after intraperitoneal injection of LPS at a dose of 5 mg/kg. The proteoglycan fraction was prepared from the livers as described previously (17): the final step was absorption to a column of Q-Sepharose and elution with a buffer containing 1 M NaCl after washing with a buffer containing 0.4 M NaCl. The proteoglycan fraction was adjusted to the concentration of 6 g (as core proteins)/ml in PBS(Ϫ) containing 1% BSA. In some cases, the proteoglycan fraction was pretreated with heparitinase (Seikagaku Co., Tokyo, Japan) (0.5 unit/ml) at 37°C for 3 h.
Each well of a 96-well plate was incubated at 4°C overnight with BSA, basic fibroblast growth factor (Roche Molecular Biochemicals), or TGF-␤1 at the concentration of 10 g/ml in 50 l of distilled water or 4 mM HCl (in case of TGF-␤1), which was necessary to change the latent form of TGF-␤1 to the active form. After washing four times with PBS(Ϫ) containing 0.1% Tween 20, the plates were incubated with 1% BSA in distilled water at room temperature for 1 h. Then 50 l of the proteoglycan fraction in the absence or presence of heparin (10 g/ml) was added. After incubation at room temperature for 1 h, anti-syndecan-4 antibody and anti-rabbit IgG antibody conjugated with horseradish peroxidase were applied successively. Visualization was performed with tetramethylbenzidine at room temperature for 30 min. The OD at 450 nm was determined with ImmunoMINI NJ-2300.
Statistical Analysis-Stat View 4.5 (SAS Institute Inc., Carry, NC) was used for statistical analyses, and p values were calculated with Log-rank test or Student's t test where appropriate. A p value of 0.05 was considered to be statistically significant.
We investigated possible differences in proinflammatory cytokine levels after LPS administration between Synd4(ϩ/ϩ) and Synd4(Ϫ/Ϫ) mice. The profile of TNF-␣ concentration in plasma was not different between the two genotypes, which reached the maximal level 1 h after, and declined nearly to the basal level 3 h after intraperitoneal injection of LPS at a dose of 10 mg/kg ( Fig. 2A). On the other hand, the plasma level of IL-1␤ was significantly higher in Synd4(Ϫ/Ϫ) mice than in Synd4(ϩ/ϩ) controls 9 h after LPS injection, while it was not different between both genotypes until 3 h after LPS injection (Fig. 2B).
Syndecan-4 Expression Was Increased 3 and 9 h after LPS Injection-We investigated whether the expression of syndecan-4 was affected by LPS administration. To avoid a loss of mice, the dose of LPS was reduced to 5 mg/kg. Even at the dose of 5 mg/kg, systolic blood pressure 9 h after LPS administration and survival rate were significantly lower in Synd4(Ϫ/Ϫ) mice than in Synd4(ϩ/ϩ) controls (data not shown). The specificity of the anti-mouse syndecan-4 antibody was confirmed by Western blot analysis: the antibody reacted only to the GST-syndecan-4 fusion protein among GST fusion proteins carrying the ectodomains of mouse syndecan-1, -2, -3, and -4 (data not shown). Although syndecan-4 expression was not changed 1 h after intraperitoneal injection of LPS (data not shown), it was increased 3 and 9 h after LPS injection in the microvasculature of the lung (Fig. 3B), intestine (data not shown), kidney (data not shown), and liver (Fig. 3, D and E). Strong expression of syndecan-4 was also detected in Mac-1-positive cells 3 and 9 h after LPS injection (Fig. 3, E-G). The assay to quantify syndecan-4 expression in the liver confirmed that its expression was increased 3 and 9 h after LPS injection (Fig. 4). Flow cytometry revealed that syndecan-4 expression was increased in Mac-1positive cells 3 h (Fig. 5) and 9 h (data not shown) after LPS injection. Syndecan-4 expression in these cells declined 1 day after, and returned to the basal level 4 days after LPS injection (data not shown). Little expression of syndecan-4 was observed in Gr-1-positive cells by immunohistochemical staining and flow cytometry (data not shown). Thus, Mac-1-positive cells expressing syndecan-4 strongly were considered to be monocytes and Kupffer cells in the sinusoids of the liver. When Mac-1-positive cells from Synd4(Ϫ/Ϫ) mice were analyzed, a small number of cells were stained with anti-syndecan-4 antibody and the second antibody, namely anti-rabbit IgG antibody conjugated with fluorescein isothiocyanate (Fig. 5). However, the positive cells were detected similarly even in the absence of anti-syndecan-4 antibody (data not shown), indicating that the second antibody reacted nonspecifically with this population of cells. Although a small number of cells stained with the second antibody were also present in Mac-1-positive cells from Synd4(ϩ/ϩ) mice, the number was not elevated by LPS injection (data not shown). Therefore, the increase of Mac-1-positive cells with syndecan-4 expression after LPS injection (Fig. 5) was not due to increase of the nonspecifically stained cells.
Binding of TGF-␤1 to Syndecan-4 -We investigated whether TGF-␤1 binds to syndecan-4 to exert inhibitory effects on production of IL-1␤. First, we compared binding of 125 I-labeled TGF-␤1 to Synd4(Ϫ/Ϫ) peritoneal macrophages with that to Synd4(ϩ/ϩ) cells. The amount of 125 I-labeled TGF-␤1 bound to Synd4(Ϫ/Ϫ) macrophages was significantly (p Ͻ 0.01) less than that to Synd4(ϩ/ϩ) controls (Fig. 7). Heparin (10 g/ml) reduced the binding of 125 I-labeled TGF-␤1, and abolished the difference between the binding to Synd4(Ϫ/Ϫ) cells and that to Synd4(ϩ/ϩ) cells (Fig. 7). The results suggest that syndecan-4 is one of the TGF-␤1-binding molecules in macrophages. To obtain more data on binding between TGF-␤1 and syndecan-4, proteoglycan fraction was isolated from livers obtained 3 h after LPS injection. Syndecan-4 in the proteoglycan fraction was found to bind to TGF-␤1; the level of the binding was more than that to BSA, but was less than that to basic fibroblast growth factor, which is known to bind to syndecan-4 (20) (Fig.  8). The binding of syndecan-4 to TGF-␤1 was almost completely abolished by heparin or digestion with heparitinase (Fig. 8), indicating that syndecan-4 binds to TGF-␤1 via its heparan sulfate chains. DISCUSSION To evaluate the roles of syndecan-4 in the events after LPS injection, Synd4(Ϫ/Ϫ) and Synd4(ϩ/ϩ) mice were injected intraperitoneally with LPS. Syndecan-4 deficiency resulted in increased mortality after LPS injection. This finding suggests that syndecan-4 promotes survival of LPS-injected mice. Systolic blood pressure and left ventricular fractional shortening were lower in Synd4(Ϫ/Ϫ) mice than in Synd4(ϩ/ϩ) controls 9 h after LPS administration, although histological examina- tions revealed no apparent difference between Synd4(ϩ/ϩ) and Synd4(Ϫ/Ϫ) mice 3 or 9 h after LPS injection. These observations indicate that syndecan-4 deficiency could increase susceptibility to endotoxin shock.
LPS elicits the production of host mediators, causing hyper-thermia, shock, and intravascular coagulation (21,22). Among these mediators, TNF-␣ and IL-1␤ are critical in the pathogenesis of endotoxin shock and organ failure, and they act synergistically in local or systemic inflammation (23). Previous studies have demonstrated that mice deficient in IL-1␤-converting enzyme (2) or caspase-11 (3), which are required for the processing of IL-1␤, and mice injected intravenously with anti-IL-1 neutralizing antibodies or antagonists of IL-1␤ (24) become resistant to high-dose LPS-induced shock. Analyses of cytokines in plasma revealed that concentration of IL-1␤ was higher in Synd4(Ϫ/Ϫ) mice than in Synd4(ϩ/ϩ) mice 9 h after LPS injection, while the level of TNF-␣ was not different between Synd4(Ϫ/Ϫ) and Synd4(ϩ/ϩ) mice. These observations suggest that excessive elevation of the IL-1␤ level could be one of the causes of high susceptibility of Synd4(Ϫ/Ϫ) mice to endotoxin shock. We then investigated the mechanism leading to the excessive elevation of IL-1␤. As the first step, we investigated whether syndecan-4 expression was altered by LPS administration. Previous studies have shown that syndecan-4 expression in the microvasculature is limited in the fetal vessels of the placental labyrinth and the capillaries between renal tubules (11,12). The present study demonstrated that syndecan-4 expression in the microvasculature became evident in the lung, intestine, kidney, and liver 3 and 9 h after LPS injection. Strong expression of syndecan-4 in Mac-1-positive cells was observed at the same time by immunohistochemical staining and flow cytometry, while Gr-1-positive cells expressed little syndecan-4. Mac-1 is expressed not only in macrophages, including monocytes and Kupffer cells, but also in neutrophils. Gr-1 is expressed in neutrophils, but not in macrophages. Thus, Mac-1positive cells expressing syndecan-4 strongly were considered to be monocytes and Kupffer cells in the sinusoids of the liver. Zhang et al. (25) have found that TNF-␣ induces syndecan-4 expression in endothelial cells by both increasing syndecan-4 gene expression in an NF-B-dependent manner and by extending the half-life of syndecan-4 mRNA. TNF-␣ concentration in plasma was elevated 1 h after LPS injection, which was prior to syndecan-4 induction. Thus, TNF-␣ could be one of the molecules involved in induction of syndecan-4 expression after LPS injection.
It is noteworthy that syndecan-4 expression was increased after LPS administration in monocytes/Kupffer cells and endothelial cells, which produce proinflammatory cytokines mediating endotoxin shock (26 -28). However, peritoneal macrophages of both genotypes produced IL-1␤ at the same levels on stimulation with LPS at 1, 100, and 10,000 ng/ml. This finding suggests that syndecan-4 deficiency does not alter the functions of cell-surface receptors involved in production of IL-1␤, including LPS receptor, Toll-like receptor-4 (29), and recently identified triggering receptor expressed on myeloid cells-1 (30). Therefore, we examined the hypothesis that increased syndecan-4 expression served to down-regulate the synthesis of IL-1␤. Indeed, LPS-induced production of IL-1␤ was inhibited to a lesser extent by TGF-␤1 or TGF-␤1 and IL-10 in combination in Synd4(Ϫ/Ϫ) macrophages than in Synd4(ϩ/ϩ) controls. Previous studies have demonstrated that TGF-␤1 binds to heparin/heparan sulfate (31,32). Our experiments showed that syndecan-4 deficiency reduced the amount of 125 I-labeled TGF-␤1 bound to macrophages and that syndecan-4 in the proteoglycan fraction from livers obtained 3 h after LPS injection bound to TGF-␤1 via its heparan sulfate chains. Heparan sulfate is known to modulate the function of growth factors by promoting their accumulation to specific sites and also serving as their co-receptors (10, 33-35). Thus, the interaction between TGF-␤1 and syndecan-4 on macrophages, and probably on en-  8. Binding of syndecan-4 to TGF-␤1. The proteoglycan fraction was isolated from livers obtained 3 h after LPS injection as described under "Experimental Procedures," and applied to wells of 96well plates coated with BSA, basic fibroblast growth factor, or TGF-␤1. In some cases, the proteoglycan fraction was applied with heparin, or pretreated with heparitinase. Bound syndecan-4 was detected with anti-syndecan-4 antibody, and estimated as A 450 , as described under "Experimental Procedures." The experiments with duplicate samples were performed independently twice, showing the same results. Open columns, the proteoglycan fraction from Synd4(ϩ/ϩ) mice; dotted column, the proteoglycan fraction was applied with heparin; shaded column, the proteoglycan fraction was pretreated with heparitinase; closed column, the proteoglycan fraction from Synd4(Ϫ/Ϫ) mice. dothelial cells, may play a role in regulation of cytokine synthesis to avoid endotoxin shock.
Syndecan-4 is a member of the syndecan family, which consists of 4 closely related molecules (10). The plasma membrane also contains the glypican family, which are heparan sulfate proteoglycans linked to the membrane via a glycosylphosphatidylinositol (10). Therefore, it was remarkable that deficiency of only one heparan sulfate proteoglycan significantly alters susceptibility to endotoxin shock. This finding will be important in investigating genetic factors affecting susceptibility to septic shock in patients.