Peptide 19-2.5 Inhibits Heparan Sulfate-Triggered Inflammation in Murine Cardiomyocytes Stimulated with Human Sepsis Serum

Myocardial dysfunction in sepsis has been linked to inflammation caused by pathogen-associated molecular patterns (PAMPs) as well as by host danger-associated molecular patterns (DAMPs). These include soluble heparan sulfate (HS), which triggers the devastating consequences of the pro-inflammatory cascades in severe sepsis and septic shock. Thus, there is increasing interest in the development of anti-infective agents, with effectiveness against both PAMPs and DAMPs. We hypothesized that a synthetic antimicrobial peptide (peptide 19-2.5) inhibits inflammatory response in murine cardiomyocytes (HL-1 cells) stimulated with PAMPs, DAMPs or serum from patients with septic shock by reduction and/or neutralization of soluble HS. In the current study, our data indicate that the treatment with peptide 19-2.5 decreases the inflammatory response in HL-1 cells stimulated with either PAMPs or DAMPs. Furthermore, our work shows that soluble HS in serum from patients with Gram-negative or Gram-positive septic shock induces a strong pro-inflammatory response in HL-1 cells, which can be effectively blocked by peptide 19-2.5. Based on these findings, peptide 19-2.5 is a novel anti-inflammatory agent interacting with both PAMPs and DAMPs, suggesting peptide 19-2.5 may have the potential for further development as a broad-spectrum anti-inflammatory agent in sepsis-induced myocardial inflammation and dysfunction.


Introduction
Sepsis remains one of the most common cause of death in intensive care units worldwide [1]. Thereby septic cardiomyopathy is recognized in at least 50% of patients with septic shock and Trypsin activity was blocked by the trypsin inhibitor from glycine max (soya bean; Sigma-Aldrich, St. Louis, MO, USA) at a ratio of 10 μl per 1 cm 2 of cells.

Transfections, stimulation and luciferase assays
HL-1 cells were plated on 6-well plates 48-72 h before transfection. The transfection complex contained per plate 635 μl of supplemented Claycomb medium, 49.1 μl of FuGENE HD transfection reagent, and 16.3 μg of DNA. The cells were transfected using the firefly pGL4.32 [luc2P/ NFκB-RE/Hygro] vector and the Renilla pGL4.74 [hRluc/TK] vector (all Promega, WI, USA). A total of 15.6 μg of luc2P/NFκB pGL4.32 was used in conjunction with 0.66 μg of Renilla pGL4.74 per plate. The transfection complex was added to the cell culture and incubated for 24 h, followed by additional 24 h incubation in supplemented Claycomb medium. LPS from the rough mutant Ra from Salmonella enterica serovar Minnesota (R60) was extracted as described in [15]. LPS, fibroblast stimulating lipopeptide-1 (FSL-1) (EMC Microcollections, Tübingen, Germany), HS (AMS Biotechnology, Oxon, United Kingdom), serum from patients with Gramnegative or Gram-positive septic shock or HS-free serum was added in different concentrations to the supplemented Claycomb medium 4 h prior to luciferase measurement, in the presence or absence of peptide 19-2.5 (20 μg/ml, sequence GCKKYRRFRWKFKGKFWFWG, molecular weight 2711). In preceding experiments the dose of 20 μg/ml peptide 19-2.5 combined the highest efficiency with the lowest toxicity [15]. NFκB-Luciferase activity was assayed with the Dual-Glo Luciferase system (Promega, Madison, WI, USA) as per the manufacturer's instructions.
Firefly luciferase values were normalized to Renilla luciferase values for each set of readings as per the manufacturer's instructions.
cultures with either Gram-negative or Gram-positive strains. Furthermore, we collected plasma from healthy human donors (n = 10). To stimulate HL-1 cells we used the sera from 6 of the 18 septic shock patients with either Gram-negative (n = 3) or Gram-positive (n = 3) strain of infection (Table 1).

Ethics statement
This study and the collection of serum and plasma were approved by the local ethics committee of the University Hospital Aachen (EK_206_09). All patients or their legal representative gave written informed consent before sampling.

Heparan sulfate ELISA
The amount of HS in serum and plasma was determined using ELISA (AMS Biotechnology, Oxon, United Kingdom) according to the manufacturer's instructions. The absorbance was measured at 450 nm on a microplate reader (Sunrise Tecan, Crailsheim, Germany).
Elimination and reconstitution of HS in serum of septic shock patients in vitro HS was eliminated from serum (n = 6) of septic shock patients (SsP) in vitro using a biotinconjugated polyclonal antibody against HS (host: rabbit, clone: PAA565Hu71, USCN Life Science Ltd Co., Wuhan, China) and affinity chromatography (Pierce Streptavidin Agarose Columns, Thermo Scientific Inc., Worcester, MA, USA) according to the manufacturers' instructions. SsP was incubated with anti-HS antibody (1:200 dilution) for 10min at room temperature and added to the column. The column was placed in a collection tube and centrifuged at 500 x g for 1 minute. A specific ELISA (AMS Biotechnology, Oxon, United Kingdom) was used to test the absence of HS in SsP according to the manufacturer's instructions. To exclude that other factors are co-eliminated we reconstituted the detected amount of HS with artificial HS (AMS Biotechnology, Oxon, United Kingdom) to each sample (reconstituted serum) and re-performed the measurements.

Statistical analyses
The PCR-derived data were derived using a relative expression software tool (REST, (http:// www.gene-quantification.de/rest.html, rest-mcs-beta-9august 2006) [22]. The expression ratios are calculated on the basis of the mean crossing point (CP) values for reference and target genes. All data are given as mean ± standard deviation (SD). We used a multiple t-test with Holm-Šídák correction when comparing differences between experimental (peptide treatment) and control (untreated cells) groups. We used a 1-way-ANOVA and Tukey´s-Test for multiple comparisons when comparing differences in heparan sulfate levels between healthy volunteers and septic shock patients with Gram-negative or Gram-positive strain of infection. We performed all calculation and figures using GraphPad Prism 6 (GraphPad, San Diego, CA, USA). A p-value of p < 0.05 was considered significant.

Results
PAMPs-mediated inflammatory response in cell culture

HS-mediated inflammatory response in cell culture
Stimulation of HL-1 cells with HS resulted in a significant dose-dependent increase in NFκBluciferase reporter activity compared to non-stimulated cells ( Fig 3A). Furthermore, the mRNA levels and secreted protein concentrations of TNF-α and IL-6 were upregulated in HL-1 cells stimulated with HS in a dose-dependent manner relative to non-stimulated cells (Fig 2B-2E). Treatment with peptide 19-2.5 significantly lowered NFκB-luciferase reporter activity levels (Fig 2A) and significantly decreased TNF-α and IL-6 mRNA expression and secreted protein concentrations in HL-1 cells stimulated with HS compared to untreated cells ( Fig 2B-2E).

Inflammatory response in cells stimulated with human septic shock serum
To investigate whether peptide 19-2.5 reduces inflammation induced by human sepsis serum we tested the inflammatory response of HL-1 cells stimulated with serum from patients with Gram-negative or Gram-positive septic shock at different concentrations. NFκB-luciferase reporter activity significantly increased after stimulation with serum from patients with Gramnegative or Gram-positive septic shock compared to non-stimulated cells (Fig 4A and 4B). The mRNA levels and secreted protein concentrations of TNF-α and IL-6 were significantly upregulated relative to non-stimulated cells (stimulation with 5% serum see left part of Fig 5A-5H, other concentrations see Tables 2 and 3). Treatment with peptide 19-2.5 significantly lowered NFκB-luciferase reporter activity levels (Fig 4A and 4B) and significantly decreased mRNA expression and secreted protein concentrations of TNF-α and IL-6 in HL-1 cells stimulated with serum from septic shock patients compared to untreated cells (Fig 5A-5H, left part and Tables  2 and 3).

Soluble HS in human septic shock serum
To determine the effects of soluble HS in serum from patients with septic shock, we first measured levels of soluble HS in the applied serum (Table 1). Next, we eliminated soluble HS from      the serum and stimulated HL-1 cells with HS-free serum from patients with Gram-negative or Gram-positive septic shock. We found a significant lower NFκB-luciferase reporter activity compared to cells stimulated with the primary serum (2,5%, 5%, and 10%, respectively, all p < 0.05) (HS-free serum see Fig 4C and 4D, primary serum see Fig 4A and 4B). Treatment with peptide 19-2.5 did not significantly alter NFκB-luciferase reporter activity in HL-1 cells stimulated with HS-free serum compared to untreated cells (2,5%, 5%, and 10%, respectively, all n.s.) (Fig 4C and 4D). To exclude that other relevant factors had been co-eliminated during HS elimination we reconstituted the detected amount of HS using artificial HS to each serum sample and re-performed the measurements. In cells stimulated with reconstituted serum we obtained nearly the same elevated NFκB-luciferase reporter activity (Fig 4C and 4D), than in cells stimulated with primary serum (Fig 4A and 4B). Concomitantly to the assessment of NFκB-luciferase reporter activity, we measured TNF-α and IL-6 mRNA and secreted protein concentrations of cells stimulated with HS-free serum (stimulation with 5% HS-free serum see center part of Fig 5A-5H, other concentrations see Tables 2 and 3). Application of HS-free serum induced significant lower levels of both TNF-α and IL-6 mRNA and secreted protein concentrations compared to primary serum (2,5%, 5%, and 10%, respectively, all p < 0.05). Treatment with peptide 19-2.5 did not significantly alter TNF-α and IL-6 mRNA expression and secreted protein concentrations in HL-1 cells stimulated with HS-free serum compared to untreated cells (Fig 5A-5H, middle part and Tables 2 and 3). Application of reconstituted serum to HL-1 cells resulted in increased levels of both TNF-α and IL-6 mRNA and secreted protein concentrations compared to cells stimulated with HS-free serum (2,5%, 5%, and 10%, respectively, all p < 0.05), which were comparable to primary serum (Fig 5A-5H, right part and Tables 2 and 3).

Study population characteristics
Included patients with septic shock had a mean age of 70 ± 15 years (78% male). The healthy volunteers had a mean age of 67 ± 19 years (50% male). HS level were significantly higher in patients with septic shock compared to healthy volunteers (p < 0.0001). There was a significant difference of HS levels between the patients with Gram-negative to those with Gram-positive septic shock (Fig 6). Additional characteristics of patients' sera used for cell stimulation (n = 6) are shown in Table 1.

Discussion
This study demonstrates that treatment with peptide 19-2.5 decreases inflammatory response in HL-1 cells stimulated with both PAMPs and DAMPs. Furthermore our work shows that soluble HS in serum from patients with Gram-negative or Gram-positive septic shock induces a strong pro-inflammatory response in HL-1 cells, which can be effectively blocked by peptide 19-2.5.
response in HL-1 cells stimulated with several concentrations of HS in the presence or absence of peptide 19-2.5. We measured a dose-dependent increase in NFκB-luciferase reporter activity as well as elevated cytokine mRNA levels and secreted protein concentrations. These results are in line with findings from Johnson et al. who stimulated different cell lines with HS concentrations from 0.3 to 10 μg/ml and detected a dose-dependent NFκB activation, notably after 30-min stimulation [10], compared to 4h stimulation in our experiments. As part of the innate immune system, Toll-like receptors rapidly react on a pathogen challenge without prior exposure. HS are known as endogenous TLR-4 ligands [9,10,[24][25][26][27][28][29][30][31], which induce the release of cytokines [32] and trigger the pro-inflammatory cascades in severe sepsis and septic shock [9,10]. Our work is the first that identifies peptide 19-2.5 as a potential therapeutic option of blocking the HS-associated inflammatory response. Recently, an investigation showed that peptide 19-2.5 binds to HS moieties attached to their proteoglycan on cells, thereby inhibiting the entry of enveloped viruses [18]. Similarly to the described changes of the LPS aggregate structure by peptide 19-2.5 [15] the decreased pro-inflammatory response by peptide 19-2.5 in cells stimulated with HS could be explained by neutralization of the HS charge. It was described before that peptides interfering with protein-protein or viral protein-host membrane interfaces may have the potential to serve as novel antiviral drugs [33]. Krepstakies et al. investigated the binding of the positively charged peptide 19-2.5 to the negatively charged HS by biophysical analysis. They detected an alteration of the peptide's secondary structure and a characteristic change in the hydration and sulfation status of the HS moieties due to a pronounced interaction of peptide 19-2.5 and HS [18]. Recently, it was shown that high sulfation in O-position of HS is required for their immunomodulatory activities [34]. Thus, reduction of pro-inflammatory response in HL-1 stimulated with soluble HS (Fig 3) may be explained by neutralization of the HS structure by peptide 19-2.5, impeding binding of HS to Toll like receptor 4.

Soluble HS in serum from septic shock patients
Although several studies have evaluated circulating levels of glycosaminoglycans in plasma of critically ill patients [35][36][37], our work is the first to identify a difference in HS levels according to the type of bacterial infection (Fig 6). In addition to the pro-inflammatory response in HL-1 cells stimulated with HS, incubation with serum from septic shock patients also induced an inflammatory response (Figs 4 and 5). Our measurements are consistent with another study using sera (2.5-10%) collected from mice 4 h after sepsis induced by cecal ligation and puncture (CLP) [38]. Data from this model suggest the time-dependent generation of inflammatory cell injury in primary cultures of mouse cortical tubular epithelial cells [38]. Johnson et al. administered HS by intraperitoneal injection to mice. Eighty percent mice injected with HS died, however 5 mg of HS for intraperitoneal injection was used [9]. To determine the relevance of soluble HS in human serum for an inflammatory response, we eliminated HS from serum and found significantly attenuated inflammatory response relative to that observed after exposure to primary serum from patients with septic shock (Figs 4 and 5). Notably, addition of peptide 19-2.5 to the HS-free serum did not alter the inflammatory response, suggesting an HS-dependent mechanism of peptide 19-2.5. It has been well documented that HS binds an array of growth factors, chemokines and cytokines [39]. Indeed, there have been many cases in which factors were studied using elimination experiments, which was later found to be not reproducible due to co-elimination of other factors [40]. To exclude other than HS effects after elimination, we reconstituted the detected amount of HS to each serum sample and re-performed the measurements using artificial HS. Stimulation with reconstituted serum reproduced the increase in NFκB-luciferase reporter activity, cytokine mRNA levels and secreted protein concentrations as detected after stimulation with primary serum (Figs 4 and 5 and Tables 2 and 3).
Yet, there are some limitations of our study. First, we investigated only the early phase of sepsis in humans. The results may differ in later stages of sepsis after initial improvement by adequate therapy. Second, the use of a cell culture model to study peptide treatment limits the transferability to human sepsis. Third, although we showed that HS induces inflammatory responses in murine cardiomyocytes, our findings are limited to in vitro measurements. Thus we will further investigate the role of HS in triggering cardiac inflammation and dysfunction during sepsis in vivo.
In summary, our data indicate for the first time that the treatment with peptide 19-2.5 decreases the inflammatory response in HL-1 cells stimulated with either PAMPs or DAMPs. Moreover, we demonstrated for the first time that soluble HS in serum from patients with Gram-negative or Gram-positive septic shock induces a strong pro-inflammatory response in HL-1 cells, which can be effectively blocked by peptide 19-2.5. Thus, to our knowledge peptide 19-2.5 is the only anti-infective agent interacting with both PAMPs and DAMPS, suggesting peptide 19-2.5 may have the potential for further development as a broad-spectrum anti-inflammatory agent in sepsis-induced myocardial inflammation and dysfunction.