C-terminal mini-PEGylation of a marine peptide N6 had potent antibacterial and anti-inflammatory properties against Escherichia coli and Salmonella strains in vitro and in vivo

Enteropathogenic Escherichia coli and Salmonella pullorum are two important groups of zoonotic pathogens. At present, the treatment of intestinal pathogenic bacteria infection mainly relies on antibiotics, which directly inhibit or kill the pathogenic bacteria. However, due to long-term irrational, excessive use or abuse, bacteria have developed different degrees of drug resistance. N6, an arenicin-3 derivative isolated from the lugworm, has potent antibacterial activity and is poorly resistant to enzymatic hydrolysis and distribution in vivo. Polyethylene glycol (PEG) is an extensively studied polymer and commonly used in protein or peptide drugs to improve their therapeutic potential. Here, we modified the N-/C-terminal or Cys residue of N6 with liner PEGn of different lengths (n = 2, 6,12, and 24), and the effects of PEGylation of N6 on the stability, toxicity, bactericidal mechanism, distribution and efficacy were investigated in vitro and in vivo. The antimicrobial activity of the peptide showed that PEGylated N6 at the C-terminus (n = 2, N6-COOH-miniPEG) had potent activity against Gram-negative bacteria; PEGylated N6 at the N-terminus and Cys residues showed low or no activity with increasing lengths of PEG. N6-COOH-miniPEG has higher stability in trypsin than the parent peptide-N6. N6-COOH-miniPEG significantly regulated cytokine expression in lipopolysaccharides (LPS)-induced RAW 264.7 cells, and the levels of tumor necrosis factor-α (TNF-α), interleukin-6 (IL-6) and IL-1β were reduced by 31.21%, 65.62% and 44.12%, respectively, lower than those of N6 (-0.06%, -12.36% and -12.73%); N6-COOH-miniPEG increased the level of IL-10 (37.83%), higher than N6 (-10.21%). The data indicated that N6-COOH-miniPEG has more potent anti-inflammatory and immune-regulatory effect than N6 in LPS-stimulated RAW 264.7 cells. N6-COOH-miniPEG exhibited a much wider biodistribution in mice and prolonged in vivo half-time. FITC-labeled N6-COOH-miniPEG was distributed throughout the body of mice in the range of 0.75 – 2 h after injection, while FITC-labeled N6 only concentrated in the abdominal cavity of mice after injection, and the distribution range was narrow. N6-COOH-miniPEG improved the survival rates of mice challenged with E. coli or S. pullorum, downregulated the levels of TNF-α, IL-6, IL-1β and IL-10 in the serum of LPS-infected mice, and alleviated multiple-organ injuries (the liver, spleen, kidney, and lung), superior to antibiotics, but slightly inferior to N6. The antibacterial activity, bactericidal mechanism and cytotoxicity of N6-COOH-miniPEG and N6 were similar. N6-COOH-miniPEG has a higher resistance to trysin than N6. The distribution of N6-COOH-miniPEG in mice was superior to that of N6. In exploring the modulatory effects of antimicrobial peptides on cytokines, N6-COOH-miniPEG had stronger anti-inflammatory and immunomodulatory effects than N6. The results suggested that C-terminal PEGylated N6 may provide an opportunity for the development of effective anti-inflammatory and antibacterial peptides.


Background
Gram-negative pathogenic Escherichia coli and Salmonella are common pathogens of bacterial disease infections in livestock and poultry farming, which can cause diarrhea diseases and even sepsis in both animals and humans [1,2]. Over four billion episodes of diarrhea occur annually in developing countries with diarrheagenic E. coli outbreaks also being reported [3]. Salmonella has been reported to cause between 200 million and 1.3 billion cases of intestinal disease worldwide each year [4], causing hundreds of millions of dollars in losses to the livestock and poultry industry each year [5]. Currently, the prevention and treatment of bacterial diarrheal diseases on farms mainly rely on antibiotics, but due to long-term unreasonable use, has led to varying degrees of bacterial resistance [6,7]. Treatment with antibiotic therapy also induces the rapid release of high concentrations of LPS from the cell walls of gram-negative bacteria, and subsequently induces a variety of acute and chronic diseases [8]. Therefore, there is an urgent and growing need for the development of alternative antimicrobial agents with superior properties for the prevention and treatment of E. coli and Salmonella infections.
Antimicrobial peptides (AMPs) are promising therapeutic agents due to their rapid-killing and broad-spectrum antimicrobial properties [9]. AMPs may not cause widespread bacterial resistance due to their membranolytic mechanism that is distinct from conventional antibiotics, and thus they are a promising alternative to current antibiotics for disease treatment [10,11]. However, some side effects of AMPs such as sensitivity to enzymes, unknown toxicity towards host cells, rapid renal clearance, potent immunogenicity and short circulation half-life have limited their application in clinics [12,13]. Our preliminary study has demonstrated that N6 exhibited potent antimicrobial activity against gramnegative bacteria, especially Salmonellae and E. coli [14]. However, N6 showed poor trypsin resistance stability, and the inhibitory activity of N6 against Salmonella was reduced by 100% after 4 h of incubation with trypsin [14]. Therefore, improving the anti-trypsin capacity and circulating life of N6 is the main problem to solve the future clinical application of N6.
PEG, also known as a macrogols, is a polyether composed of repeated ethylene glycol units [-(CH 2 CH 2 O) n ] with molecular weights (MWs) from 0.4 to 150 kDa; PEG has been used in pharmaceutical fields, consumer products, and cosmetics for decades [15]. PEGylation is a process through which PEG chains are conjugated to drug molecules to improve their in vivo efficacy and tissue distribution; PEGylation is one of the widely used technologies for proteins, peptides and small-molecule drugs due to its non-immunogenicity, degradability and excellent biocompatibility [16,17]. PEG modification can provide more advantages in enhancing resistance to proteolytic degradation, prolonging blood circulation, improving water solubility, and enhancing pharmacokinetics of drugs [18,19]. PEG molecules include a wide variety of linear, branched, Y-shaped, or multiarm geometries forms; among them, linear PEGs are the simplest and most commonly used in bioconjugation and crosslinking of peptides or proteins [20]. Until now, a few AMPs have been covalently conjugated with PEG and they are currently under preclinical trials. Imura et al. [21] investigated the effects of PEGylation of tachyplesin I from Tachypleus tridentatus on the mechanism and found that the attachment of the large PEG moiety (5 kDa) did not change the mode of action of tachyplesin I, but its binding affinity to lipid bilayers was reduced by 2 -3 fold. Morris et al. [22] synthesized PEG-CaLL at N-terminus, which is derived from LL-37 and cecropin A, and found that PEGylated CaLL derivatives had higher activity against B. anthracis (including vegetative and spore forms), Staphylococcus aureus and Escherichia coli than the parental LL-37 peptide. Benincasa et al. [23] modified Bac7 (1e35) by using cleavable ester bonds or non-hydrolyzable amide bonds and found that PEGylated Bac7 (1e35) improved the bioavailability and led to the wide distribution in mice and slow renal clearance of Bac7 (1e35). Singh et al. [24] found that PEG-binding to KYE28 effectively reduced the toxicity and serum protein clearance, increased selectivity, and retained antiinflammatory effects. PEGylation at the C-terminus of the three-Lys-branching core of M33 led to an increase in stability to Pseudomonas aeruginosa elastase [25]. Meanwhile, other PEGylated AMPs such as M33 [25], KYE28 [24], CaLL [22], Nisin A [26], MA [27], tachyplesin I [21], magainin 2 [28], 73c [21] and Bac7   [23] improved solubility [29,30], proteolytic stability [31,32], and circulating half-life in the blood, respectively [33,34]; they also reduced toxicity [35]. Noticeably, the conjugation site, distribution over the peptides and structure (linear, branched or dendritic) of PEG may influence antimicrobial effects and toxicity of drugs [24]. However, little is still known about how linear PEG length and localization affect the antibacterial and anti-inflammatory activity of AMPs in vitro and in vivo, respectively.
Considering this, in this study, a series of PEGylated N6 analogues were obtained by adding different lengths of linear PEGn (n = 2, 6, 12, and 24) with MWs from 145 to 1,127 Da to the N-terminus, C-terminus and Cys residues at positions 7 and 16 of a marine AMP-N6 (GFAWN-VCVYRNGVRVCHRRAN), an AMP with potent bactericidal activity against Gram-negative bacteria in our previous study, but issues with protease sensitivity [14]. The effects of PEGn lengths (n = 2, 6, 12, and 24) at different sites on the bactericidal activity of N6 against major enteric pathogens-E. coli and Salmonella pullorum were firstly investigated, followed by their toxicity, stability, and mechanism of action. Furthermore, the in vivo therapeutic effect of PEGylated N6 was evaluated in ICR mice infected with E. coli and S. pullorum, respectively.

N6-COOH-miniPEG potently permeabilize bacterial cell membranes and disrupt membrane potentials
The ability of peptides to permeabilize the outer membranes was measured with the N-phenyl-1-naphthylamine (NPN) fluorescent dye. As shown in Fig. S10A and B, both N6 and N6-COOH-miniPEG caused a time-dependent and concentration-dependent increase in fluorescence in E. coli CVCC195 and S. pullorum CVCC533 cells. As shown in Fig. S10A, both N6 and N6-COOH-miniPEG had disruptive effect on the outer membrane of E. coli CVCC195. The fluorescence intensity of 4 × MIC N6 (t = 1 min, fluorescence intensity = 17,884) and N6-COOH-miniPEG (t = 1 min, fluorescence intensity = 29,939) treated E. coli CVCC195 increased rapidly within 1 min, indicating that 4 × MIC N6 and N6-COOH-miniPEG could rapidly penetrate the outer membrane of E. coli CVCC195 within 1 min. As shown in Fig The ability of peptides to permeabilize the inner membranes was measured with the propidium iodide (PI) fluorescence. As shown in Fig. S11A-B, the percentages of E. coli CVCC195 or S. pullorum CVCC533 stained with PI were 0.36-0.797% and 0.217 -0.64% in the absence of N6 or N6-COOH-miniPEG, indicating intact bacterial inner membranes. After treatment with N6-COOH-min-iPEG for 5-120 min, the percentages of PI-permeable E. coli CVCC195 cells were 4.89-14.9%, higher than those of N6 (3.03 -15.5%), indicating a higher level of internalization of PEGylated N6. As shown in Fig. S11B, after treatment with 4 × MIC N6, the membrane permeabilizing ratio of bacteria increased to 26.1% (0.5 h) and 45.8% (2 h). After treatment with 4 × MIC N6-COOH-miniPEG for 0.5 -2 h, the penetration ratio increased up to 29.2% and 35.9%, respectively. Therefore, the inner membrane disruption ability of N6-COOH-miniPEG is slightly lower than that of N6 against S. pullorum CVCC533.
To further examine effects of peptides on bacterial cell membranes, the membrane potentials were monitored with the voltage-sensitive anionic lipophilic dye 3,3′-dipropylthiadicarbocyanine iodide (DiSC 3 (5)) [36]. After treatment with 0.25-8 × MIC N6 and N6-COOH-miniPEG, the fluorescence values of E. coli CVCC195 and S. pullorum CVCC533 suspensions were enhanced rapidly and in a concentration-dependent manner (Fig.

N6-COOH-miniPEG enhances the ability to bind to Lipopolysaccharide (LPS)
The ability of peptides to bind to LPS was determined using the fluorescent probe BODIPY-TR cadaverine (BC) displacement assay [28]. In the concentration range of 0.1 -3 μM, both N6 and N6-COOH-miniPEG could displace the BC probe and bind to LPS in a dose-dependent manner (Fig. S13). N6-COOH-miniPEG has a better ability to bind LPS than N6 in the concentration range of 0.1 -3 μM, except for 2 μM; at a concentration of 2 μM, the ability of N6-COOH-miniPEG to bind LPS with N6 was comparable (Additional File Sheet 28).

N6-COOH-miniPEG more strongly affected the morphology and integrity of the bacteria
The morphology and integrity of E. coli CVCC195 and S. pullorum CVCC533 treated with the peptides were observed by scanning electron microscopy (SEM). In the untreated group, it was observed a normal cell morphology and smooth intact cell surface of E. coli and S. pullorum (Fig. 3). After treatment with 4 × MIC N6 and N6-COOH-miniPEG for 2 h, the cellular content leakage of bacterial cells was 55% and 63%, respectively. And cell shrinkage was also found in E. coli CVCC195. For S. pullorum CVCC533, after treatment with N6 and N6-COOH-miniPEG, the bacterial cells were deformed or collapsed, and filiferous substances were observed outside the cells. The percentages of S. pullorum CVCC533 showing leakage after the treatment of 4 × MIC N6 and 4 × MIC N6-COOH-miniPEG were 62% and 70%, repectively. After treatment with N6-COOH-miniPEG, the number of bacteria in the visual field was significantly lower than that in N6 treatment group (Red arrows: typical morphological changes such as crumpling or cytoplasmic leakage of the bacterium).
As shown in Fig. S14, in the absence of N6 and N6-COOH-miniPEG, smooth intact cell surfaces were observed in E. coli CVCC195 and S. pullorum CVCC533 cells. After exposure to N6 or N6-COOH-miniPEG for 2 h, the disruption of cell membranes and uneven electron cloud density and disruption of the internal structure were observed in both E. coli CVCC195 and S. pullorum CVCC533 cells. And the percentages of E. coli showing morphological changes after treatment with 4 × MIC N6 and 4 × MIC N6-COOH-miniPEG were 49% and 63%, respectively. The percentages of S. pullorum CVCC533 showing morphological changes after treatment with 4 × MIC N6 and 4 × MICN6-COOH-miniPEG were 61% and 65%, respectively.

N6-COOH-miniPEG has a much wider biodistribution in mice
As shown in Fig. 5, free fluorescein isothiocyanate (FITC) in the control mice was rapidly distributed within 10 min and showed a slower renal clearance at 6 h post injection. The weak fluorescence of FITC-N6 was accumulated in the abdomen of mice after injection and the fluorescence did not distribute throughout the body. Comparably, however, strong fluorescence of FITC-N6-COOH-min-iPEG was distributed throughout the body at 45 min post injection; it decreased after 3 h and disappeared after 24 h, indicating a more extensive distribution of N6 after PEGylation at the C-terminus and a slower renal clearance compared to the unmodified peptide. Moreover, it was observed a significant concentration of the PEGylated N6 in kidney even at 24 h after injection (Fig. 5), indicating a prolonged in vivo half-time for N6 after its conjugation to miniPEG.

N6-COOH-miniPEG protects mice from lethal challenge with E. coli CVCC195 or S. pullorum CVCC533
To evaluate efficacy of N6 and N6-COOH-miniPEG in a peritonitis model, the mice were inoculated intraperitoneally with E. coli CVCC195 and S. pullorum CVCC533, respectively and treated with peptides. The uninoculated mice of the control group survived throughout the experimental period (Fig. 6A-B). As shown in Fig. 6A, all mice infected with E. coli CVCC195 in the control group died within 2 d. After treatment with 2 and 4 μmol/kg N6, the survival rates of mice were 50% and 90%, respectively, higher than those of N6-COOH-miniPEG (40% and 70%) and 0.125 μmol/kg polymyxin B (PMB) (30%). The survival rate of mice treated with 0.25 μmol/kg PMB was 100% (Additional File Sheet 17). The mice infected with S. pullorum CVCC533 in the control group developed the symptoms of diarrhea and weight loss, and all died within 6 d. After treatment with 8 and 10 μmol/kg N6, the survival rates of mice were 50% and 83.33%, respectively, higher than those of N6-COOH-miniPEG (33.33% and 66.67%) (Additional File Sheet 18). The survival rate of mice treated with 1 μmol/kg CIP was 100% (Fig. 6B). The results implied that PEGylation of N6 at the C-terminus slightly reduced the efficacy of the N6 in vivo.
To investigate the effects of peptides on translocation of E. coli CVCC195 or S. pullorum CVCC533 from the peritoneal cavity to other organs, livers, spleens, kidneys and lungs were collected and analyzed by the plate count method. As shown in Fig. 6C (Fig. 6D). The results suggested that N6-COOH-miniPEG is less able to inhibit bacterial translocation in mice than N6 and antibiotics.
To further explore the effects of N6 and N6-COOH-miniPEG on inflammatory cytokines, an endotoxemia The concentrations of cytokines were measured by using an ELISA kit. The analyses were measured by one-way ANOVA, with Duncan's multiple comparisons test. A probability value of < 0.05 was considered significant. (*) Indicates the significance between control and treatment groups. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0. 0001. (*) Indicates the significance between control and treatment groups; (**) Values represent as the mean ± SE of three independent experiments performed in triplicate mouse model was constructed by intraperitoneal injection with 10 mg/kg LPS, followed by treatment with peptides or antibiotics. As shown in Fig. 7A-D Sheets 21,22,23,24). These results suggested that both N6 and N6-COOH-miniPEG can regulate LPS-induced cytokines, and both are far superior to antibiotics.
To investigate whether N6 and N6-COOH-mini-PEG alleviate organ injury in mice challenged with E. coli CVCC195 and S. pullorum CVCC533, the livers, spleens, kidneys and lungs were dissected from mice at different times after treatment with peptides or antibiotics. As shown in Fig. 8 and Fig. S15, in the uninfected group, no pathological symptoms were observed in the organs of the mice. However, the untreated mice infected with E. coli CVCC195 developed severe organ damage, especially in the lungs where tissue destruction was evident, with massive inflammatory exudate in the alveolar cavity. In contrast, after treatment with N6, N6-COOH-miniPEG and PMB, there was apparently less damage of liver, spleen, kidney and lung at 24 h and 5 d post-treatment. As shown in Fig. S16   Fig. 5 Biodistribution of N6 and N6-COOH-miniPEG in the healthy nude mice. The nude mice were injected intraperitoneally with 10 mg/kg FITC-labeled N6, N6-COOH-miniPEG or free FITC, and fluorescence (ventral) was observed at 10, 30, 45, 60, 90, 120, 180, 240, 360, and 1440 min, respectively. The mice from left to right were free FITC, FITC-labeled N6, FITC-labeled N6-COOH-miniPEG, and the blank control, respectively and 17, the untreated mice infected with S. pullorum CVCC533 developed acute injury, characterized by a large number of inflammatory cells, degeneration, cord atrophy or necrosis in the tissues. In contrast, after treatment with N6, N6-COOH-miniPEG and CIP, there was apparently less damage of liver, spleen, kidney and lung at 48 h and 4 d post-treatment. The efficacy of N6-COOH-miniPEG was slightly lower than that of N6 and antibiotics. The results suggested that N6-COOH-miniPEG improves the histopathology of mouse tissue damage induced by E. coli or S. pullorum, superior to antibiotics, but inferior to N6.

Discussion
AMPs are a class of small cationic polypeptides that are important components of biological innate immunity [37]; AMPs are thought to be effectively conserved immune molecules in ancient evolutionary mammals [38,39] and thus have become the most promising new generation of drugs to replace traditional antibiotics due to their broad antibacterial spectrum [37], high antibacterial activity [40,41], and low resistance [42,43]. To date, however, few AMPs has been approved by the US Food and Drug Administration (FDA) due to their toxicity, protease cleavage, and short half-life [11, Mice were intraperitoneally injected with E. coli (1 × 10 9 CFU/mL) followed by N6, N6-COOH-miniPEG or PMB at 0.5 and 8 h; Mice were intraperitoneally injected with S. pullorum (5 × 10 7 CFU/mL) followed by N6, N6-COOH-miniPEG or CIP at 0.5 and 8 h. Survival rates were recorded for 7 days. Effects on the bacterial translocation. Bacteria were counted in the livers, spleens, kidneys and lungs of mice infected with (C) E. coli or (D) S. pullorum after treatment with N6 or N6-COOH-miniPEG. The untreated mice were used as the negative control. Data were expressed as mean ± SE (n = 3) 34, 43,44]. Our preliminary study has demonstrated that N6 exhibits low toxicity to eukaryotic cells, low hemolysis, and potent antimicrobial activity against Gram-negative bacteria, especially for Salmonellae and E. coli [14], however, N6 is completely degraded by trypsin [41]. Therefore, in this study, PEGylated N6 was generated by adding the different lengths of linear PEGn (n = 2, 6, 12, and 24) to N-/C-terminus or Cys residue, and their antibacterial/anti-inflammatory activity, stability, mechanism and efficacy were systematically evaluated in vitro and in vivo.
PEGylation defines the modification of protein, peptide or non-peptide molecules by the linking of one or more PEG molecules [19]. The PEG polymer is non-toxic, non-immunogenic, non-antigenic and highly soluble in water and has been approved by FDA to modify many drugs [45]. In this study, we firstly attached linear PEGn (n = 2, 6, 12, and 24) with different MWs of 145-to 1,127-Da to N-/C-terminus and Cys residues at positions 7 and 16 of a marine peptide N6 to improve stability. The antibacterial activity of PEGylated N6 showed that N6 modified by linear miniPEG (n = 2, MW of 145 Da) at the C-terminus (N6-COOH-miniPEG) had the highest antibacterial activity against Gram-negative bacteria (MICs of 1.53 -6.1 μM) ( Table 2), which may be related to PEGylation of peptides [46][47][48]. Conversely, all PEGylation of N6 at the N-terminus had low/no antimicrobial activity (Table 2), which is agreement with the previous study in which N-terminal modification Fig. 7 Effects on cytokines in mice. Mice were challenged with LPS (10 mg/kg, 200 μL), followed by injection with N6 or N6-COOH-miniPEG (0.25 μmol/kg). Sera were collected and the levels of (A) TNF-α, (B) IL-6, (C) IL-1β, and (D) IL-10 were detected by using an ELISA kit after 2 and 8 h after treatment. The analyses were measured by one-way ANOVA, with Duncan's multiple comparisons test. A probability value of < 0.05 was considered significant. (*) Indicates the significance between control and treatment groups. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0. 0001. (*) Indicates the significance between control and treatment groups; (**) of glucose-dependent insulinotropic polypeptide (GIP) (1-30) with 40 kDa PEG lost functional activity; C-terminal PEGylation of GIP (1-30) retained agonism and reasonable potency at the GIP receptor and conferred a high level of DPPIV resistance [49]. Additionally, antibacterial activity of PEGylated N6 reduced with the increase in PEG length (MW of 145 -1,127 Da) at the N-terminus (Table 2), which was agreement with PEGylated KYE28 [24]. Both N6-Cys7-miniPEG and N6-Cys16-miniPEG with the PEG modification at Cys residues (Cys7 and Cys16) lost the antibacterial properties against gram-negative and gram-positive bacteria (with MICs > 128 μg/mL) ( Table 2), indicating that it may be related to PEGylation that change structural characteristics of peptides at the specific site ( Fig. S7  and 8). Some previous studies showed that PEGylation of AMPs such as CaLL and aurein 2.2 could inhibit α-helix formation [22,50]. In this study, PEGylation of N6 at the C-terminus (N6-COOH-miniPEG) did not have a detrimental impact upon α-helical content in ddH 2 O (with a decrease of 0.1%), 20 mM SDS (with an increase of 1.8%), and 50% TFE (with an increase of 2.2%); however, the antiparallel of N6-COOH-miniPEG was increased by 5.9% (ddH 2 O), 13% (SDS), and 10.3% (TFE), respectively when compared to N6 (Table S1). The result indicated that PEGylation of N6 may alter the secondary structure of the peptide and thereby it may affect antibacterial activity.
The mechanisms of action of AMPs against bacteria include both membrane-breaking and non-membrane breaking structures [51]. In a previous study, Yang et al. [14] have demonstrated that marine peptide N6 has a significant disruptive effect on the inner and outer membranes of E. coli CVCC195. In this study, by further investigating the effect of miniPEG modification on the bactericidal mechanism of N6, it was found that N6-COOH-miniPEG disrupted the internal and external membrane of bacteria to similar extent as N6 (Fig.  S10 and 11). The result was also observed in SEM and transmission electron microscopy (TEM) (Fig. 3 and   Fig. 8 Effects of N6 and N6-COOH-miniPEG on organ injury in mice. Mice were infected intraperitoneally with E. coli CVCC195 (1 × 10 9 CFU/mL, 200 μL) and treated with N6 (4 μmol/kg) or N6-COOH-miniPEG (4 μmol/kg). Livers, spleens, kidneys and lungs were harvested from sacrificed mice at 24 h after infection. CK group: The livers, spleens, kidneys, and lungs were normal; Infected E. coli CVCC195 group: Inflammatory lesions such as foci of hepatocellular necrosis, reactive enlargement of small splenic nodules, localized foci of interstitial inflammation in the kidney, and diffuse inflammatory lesions in the interstitial lung (arrow) (× 200, scale bar = 100 μm); N6 treatment group: There is a reduced density of lymphocytes in the local splenic nodules and a small infiltration of inflammatory cells in the interstitial lung (arrow) (× 200, scale bar = 100 μm); N6-COOH-miniPEG treatment group: The splenic nodules are mildly enlarged, the renal tubules are partially atrophied and a small amount of inflammatory cell infiltration is present in the interstitial lung (arrow) (× 200, scale bar = 100 μm); PMB treatment group: Reactive enlargement of splenic nodules, atrophy of the renal tubules at the site of the lesion and infiltration of inflammatory cells in the interstitial lung (arrow) (× 200, scale bar = 100 μm) S14), respectively. However, we were surprised to find that high concentrations of N6-COOH-miniPEG had a stronger effect on bacterial inner membrane permeability (Fig. S11) and inner membrane depolarization (Fig.  S12) than N6. It may be related to miniPEG attachment, which was more favorable for N6 to shift to a α-helical conformation, which in turn may disrupt the bacterial cell membrane and cause bacterial death (Fig. S9 and Table S1) [14,24,28]. Meanwhile, miniPEG linkage may stabilize the ordered conformation of N6, which was consistent with the findings of previous reports [22]. Additionally, N6-COOH-miniPEG more strongly bound to bacterial LPS than N6 (Fig. S13) and thereby it may prevent inflammatory responses.
PEGylation via amine modification is known to be a reliable method for effectively shielding proteolysis of peptides [52,53]. Theoretically, there should be a difference in the amount of degradation between the native and mini-PEGylated peptide because of PEG interference of the trypsin cleavage process [54]. In our study, we found that miniPEG conjugation prolonged the activity of N6 in trypsin buffer and improved the thermal stability of N6 ( Fig. 2C and Table S2). This result was consistent with the finding that PEGylation improved the anti-trypsin stability of the recombinant interferon (rhIFN) α-2a [55]. Furthermore, this result was consistent with the finding that PEGylation improved the thermal stability of the model protein CT-322 and PEGylated CT-322 could refold correctly after thermal denaturation [56].
The in vivo distribution of peptides showed that mini-PEGylation of N6 had a wider distribution and a slower renal clearance than unmodified N6 (Fig. 5), indicating a prolonged in vivo half-life of N6 after binding to linear PEG [34,57]. This is mainly because PEGylation offers the possibility to specifically protect endangered termini and furthermore increases molecular mass, leading to slow absorption after injection, reduced glomerular filtration and lower renal excretion, thus indirectly increasing the in vivo half-life of the short peptide [58,59]. Furthermore, it has already been demonstrated that the effect of protecting PEGylated proteins from proteolysis is especially strong when high-molecular-weight, branched PEGs are used [60]. Moreover, N6-COOH-miniPEG was found to have a stronger anti-inflammatory function than N6 in RAW 264.7 cells (Fig. 4). However, N6-COOH-miniPEG showed a slightly lower efficacy in mice infected with E. coli CVCC195 or S. pullorum CVCC533 than N6; this result was consistent with the finding that PEGylated peptides such as tachyplesin I substantially inhibited antibacterial effects against S. epidermidis, E. coli and other strains [21]. The reduced antimicrobial activity of PEGylation of peptides may be due to a decrease in the number of protonated amino groups and shielding of the positive charges by PEG chains, leading to decreasing in electrostatic interactions of peptides with the negatively charged bacterial surface [61].

Conclusions
In this study, we designed PEGylated peptides based on a marine peptide-N6 and examined the effects of different lengths of PEGn (n = 2, 6, 12, and 24) modification at different sites on the toxicity, stability, mechanism, distribution, and in vitro and in vivo antibacterial/antiinflammatory activities of N6. The results showed that PEGylated N6 at the C-terminus (n = 2, N6-COOH-miniPEG) significantly improved proteolytic stability. N6-COOH-miniPEG more potently bound to LPS and prolonged in vivo retention than N6; N6-COOH-mini-PEG retained the anti-inflammatory activity of the parent peptide N6. N6-COOH-miniPEG showed a slightly lower efficacy in mice than N6. The results suggested that PEGylation of N6 contributes to the design of effective anti-inflammatory and antibacterial PEGylated peptides for clinical applications.

Strains, peptides, and animals
The strains of E. coli CVCC195, E. coli CVCC1515, E. coli CVCC25922, E. coli ATCCO157, S. enteritidis CVCC3377, S. pullorum CVCC533, S. pullorum CVCC1789 and S. pullorum CVCC1802 were purchased from China Veterinary Culture Collection Center (CVCC). S. typhimurium ATCC14028, S. aureus ATCC546 and S. aureus ATCC25923 were purchased from American Type Culture Collection (ATCC). P. aeruginosa CICC21630 were purchased from China center of industrial culture collection (CICC). Candida albicans CMCC98001 was purchased from National Center for Medical Culture Collections. One clinical strain of S. hyicus 437-2 was obtained from Animal Husbandry and Veterinary Research Institute (Tianjin, China). N6 and its PEGylated analogues with different lengths of PEGn (n = 2, 6, 12, and 24) were synthesized by Mimotopes (Wuxi, China) and WuXi App Tec (Shanghai, China), respectively. The purity of all peptides was greater than 90%. Six-week-old specific-pathogen-free (SPF) female ICR and nude mice (approximately 20 ± 2 g/mouse) were obtained from the Vital River Laboratories (VRL, Beijing, China). All other chemical reagents were of analytical grade.

Antibacterial activities and bactericidal curves of N6 and PEGylated N6 analogues
The MIC was determined according to the method developed by Clinical and Laboratory Standards Institute (CLSI) [62]. The strains were cultured in Mueller-Hinton Broth medium (MHB) at the appropriate temperature until mid-log phase. A series of two-fold dilutions of N6 and PEGylated N6 (10 μL) and bacterial suspension (1 × 10 5 CFU/mL, 90 μL) were added into 96-well plates and incubated at 37 °C for 18 -24 h. PBS and MHB were used as the negative and blank control, respectively. All assays were conducted in triplicate.

Toxicity of N6 and N6-COOH-miniPEG
Erythrocytes from SPF mice were centrifuged at 1,500 rpm for 5 min and washed with 0.9% NaCl three times. 75 μL of 8% (v/v) erythrocyte solution was mixed with 75 μL of N6 or N6-COOH-miniPEG (the final concentrations of 0.5 -128 μg/mL); the mixture was added into 96-well plates, incubated at 37 °C for 1 h, and centrifuged at 1,500 rpm for 5 min. The supernatants were measured at 540 nm. 0.9% NaCl and 0.1% Triton X-100 were used as the blank (A 0 ) and positive (A 100 ) control, respectively [14,65]. The hemolysis percentages of peptides were calculated by the following equation: Hemoly- The cytotoxicity of N6 and N6-COOH-miniPEG was evaluated using the 3-(4, 5-dimethylthiazolyl-2)-2, 5-diphenyltetrazolium bromide (MTT) method [66]. RAW 264.7 cells (2.5 × 10 5 cells/mL) were seeded into a 96-well plate (100 μL/well) and cultured at 37 °C (5% CO 2 , 95% saturated humidity) for 24 h. The medium was then removed and the cells were washed twice with PBS. The peptide solutions (with the concentrations from 1 to 256 μg/mL) were then added to the plates (100 μL/well), incubated for 24 h, and washed twice with PBS. Next, 5 mg/mL MTT (100 μL/well) was added and incubated for 4 h, followed by an addition of dimethyl sulfoxide (DMSO) (150 μL/well). After shaking in a shaker, the crystals in the bottom of wells were completely dissolved, and the absorbance of each well was measured at 570 nm (A peptide ). PBS served as the control (A control ). The cell survival rate was calculated by the equation: Survival rate (%) = (A peptide / A control ) × 100.

Stability of N6 and N6-COOH-miniPEG
The thermal stability of peptides was determined after treatment for 1 h at different temperatures (4,20,40,60,80, and 100 °C). The antibacterial activity of peptides against E. coli CVCC195 was determined by the MIC assay [67]. To evaluate pH stability, the peptides were dissolved in glycine-HCl buffer (pH 2.0), sodium acetate buffer (pH 4.0), sodium phosphate buffer (pH 6.0), tris-HCl buffer (pH 8.0), or glycine-NaOH buffer (pH 10.0) and incubated for 3 h. The MIC values of the peptides against E. coli CVCC195 were measured as described above [40]. Similarly, salt stability of peptides was determined after a 3 h-incubation at different concentrations of NaCl solution (50,100,200, 300, 400, and 500 mM). Additionally, to evaluate the protease stability, N6 and N6-COOH-miniPEG (200 μg/mL) were incubated with trypsin or papain (250 U/mg, pH 8.0) for different times, an aliquot of 20 μL N6 or N6-COOH-miniPEG mixture was taken and tested against E. coli CVCC195 by the inhibition zone assay [68,69]. The untreated peptides were used as the positive control and buffers alone were used as the negative control. All assays were conducted in triplicate.
Stability of N6 or N6-COOH-miniPEG in SGF and SIF was carried out as previously described [68,70]. Briefly, N6 or N6-COOH-miniPEG (200 μg/mL) was dissolved in SGF or SIF incubated for 5 -60 min at 37 °C. At different time intervals, an aliquot of 20 μL N6 or N6-COOH-miniPEG was taken and tested against E. coli CVCC195 by the inhibition zone assay. N6 or N6-COOH-miniPEG (200 μg/mL) dissolved in PBS was used as a positive control; SGF and SIF were used as the negative controls. To determine the serum of the peptides, N6 or N6-COOH-miniPEG were dissolved in PBS or mice serum at 37 °C to detect their serum stability and samples were removed at different time points to determine their residual activity by the inhibition zone or reverse-phase high-performance liquid chromatography (RP-HPLC) methods [41].
All assays were conducted in triplicate. PC-N6, equivalent N6 prepared in PBS was used as positive control; PC-N6-COOH-miniPEG, equivalent N6-COOH-mini-PEG prepared in PBS was used as positive control.

CD analysis of N6 and N6-COOH-miniPEG
The secondary structure of the peptides was determined by CD spectroscopy in ddH 2 O, SDS, and TFE, respectively [22,69]. N6 and N6-COOH-miniPEG were dissolved in ddH 2 O, 20 mM SDS or 50% TFE and their CD spectra were measured via a MOS-450 spectropolarimeter (Bio-Logic, Grenoble, France) using a 1.0 mm pathlength cuvette. The spectra of peptides were recorded from 180 to 260 nm at 25 °C at a scanning speed of 100 nm/min with a step resolution of 2.0 nm and an integration time of 2 s. Data were analyzed using CDNN software.

Effects of N6 and N6-COOH-miniPEG on the bacterial cell membrane and membrane potential
The outer membrane permeabilization abilities of peptides were determined using the fluorescent NPN assay. Mid-log phase E. coli CVCC195 and S. pullorum CVCC533 cells were collected by centrifugation, washed twice, and suspended in n-2-hydroxyethylpiperazinen-2-ethane sulfonic acid (HEPES) buffer (pH 7.4) to an OD 600nm of 0.4. Cell suspension and NPN solutions (10 μM) were added into 96-well black plates, followed by an addition of peptide solutions (1 × , 2 × , and 4 × MIC). Fluorescence intensity was recorded until no further increase with a microplate reader (excitation/emission, 328/438 nm). The cells treated with PBS were used as the negative control; the untreated cells were used as the blank control [71].
The DiSC 3 (5) assay was used to investigate cytoplasmic membrane depolarization [36,73]. Mid-log phase E. coli CVCC195 and S. pullorum CVCC533 cells were washed in 5 mM HEPES buffer (pH 7.2) containing 20 mM glucose and resuspended in buffer (5 mM HEPES buffer, 20 mM glucose, and 100 mM KCl, pH 7.2) to an OD 600 of 0.1. The fluorescence intensity of bacterial cells was monitored for 13 min. N6 and N6-COOH-miniPEG were added and the fluorescence intensity (excitation/emission, 620/670 nm) was measured using a Tecan Infinite M200 PRO microplate reader. PBS was used as the blank control. The experiment was repeated three times.

Interaction between N6 or N6-COOH-miniPEG and bacterial LPS
The affinity between the peptides and LPS was detected by the fluorescent probe BC [74]. Equal volumes of LPS (40 μg/mL) and BC probe (10 μM) in 50 mM Tris buffer (pH 7.4) were mixed and added into a black 96-well microtiter plate (180 μL/well), followed by an addition of different concentrations of N6 or N6-COOH-miniPEG (20 μL/well). Fluorescence was measured by a Tecan Infinite M200 PRO microplate reader at room temperature (excitation/emission, 580/620 nm) and the BC displacement rate was calculated as previously described [41].
For TEM observations, the bacterial cells were treated with peptides and fixed as described above. After fixation, bacterial cells were washed with 0.01 M PBS for five times and post-fixed in 1% osmium tetroxide (OsO 4 ) for 1 h. The cells were dehydrated with a series of graded ethanol solutions (50%, 70%, 85%, 95%, and 100%, × 2 times) for 15 min/time, transferred to epoxy resins, and stained with 1% uranium acetate; the cells were observed using a JEM1400 TEM (JEDL, Japan) [14].

Effects of N6 and N6-COOH-miniPEG on cytokines in macrophages challenged with LPS
RAW 264.7 cells (2.5 × 10 5 cells/mL) were cultured in 12-well plates. N6 or N6-COOH-miniPEG (100 μg/mL) was mixed with LPS (100 ng/mL) for 30 min and added into the cells. After treatment for 4 h at 37 °C, the supernatant from each well was collected, and the levels of TNF-α, IL-6, IL-1β, and IL-10 were determined by the ELISA [76]. The cells challenged with LPS (100 ng/mL) were used as the negative control (LPS), and the cells without challenge and treatment were used as the blank control (CK).
The mice (10 mice/group) were challenged with E. coli CVCC195 (1 × 10 9 CFU/mL, 0.2 mL) or S. pullorum CVCC533 (5 × 10 7 CFU/mL, 0.2 mL) by intraperitoneal injection and treated with N6 (4 and 10 μmol/kg) or N6-COOH-miniPEG (4 and 10 μmol/kg). To evaluate bacterial loads in organs, the livers, spleens, kidneys and lungs were removed from the sacrificed mice at 24 h post treatment and homogenized in sterile PBS for E. coli CVCC195 colony counting; similarly, the livers, spleens, kidneys and lungs were removed from the sacrificed mice at 48 h after treatment and homogenized in sterile PBS for S. pullorum CVCC533 colony counting. The uninfected mice served as the blank control (CK); the infected mice treated with PBS were used as the negative control. The mice were sacrificed by neck dislocation under inhaled isoflurane anesthesia (RWD Life Science Co., LTD. Shenzhen, China).
The mice (10 mice/group) were injected intraperitoneally with N6 and N6-COOH-miniPEG (0.25 μmol/ kg), PMB (10 μmol/kg) or CIP (5 μmol/kg) 0.5 h after challenge with 10 mg/kg E. coli 0111:B4 LPS. Blood was collected in a heparin sodium tube at 2 or 8 h after LPS injection and incubated for 30 min at 37 °C and then overnight at 4 °C. After centrifugation at 3,000 rpm for 10 min at 4 °C, serum was collected from supernatant. The cytokine levels of TNF-α, IL-6, IL-1β and IL-10 in serum were detected by the Jiaxuan Biotech. Co. Ltd. (Beijing, China) using an ELISA. RAW 264.7 cells untreated with LPS and antimicrobial drugs were used as a blank control (CK).
The mice were treated with N6 (4 and 10 μmol/kg) and N6-COOH-miniPEG (4 and 10 μmol/kg) after intraperitoneal injection of E. coli CVCC195 for 0.5 and 8 h or S. pullorum CVCC533 for 0.5 h as described above. Livers, spleens, kidneys and lungs were removed from the mice at 24 h and 5 d post-treatment after E. coli CVCC195 injection and at 48 h and 4 d post-treatment after S. pullorum CVCC533 injection, respectively, washed in PBS, fixed in 4% paraformaldehyde at 4 °C for 24 h, and stained with hematoxylin and eosin (HE) for histopathology analysis. Finally, the tissue samples were fixed and observed by the light microscope. The uninfected mice served as the blank control (CK); the infected mice being treated with PBS were used as the negative control.

Statistics
All data were analyzed by Origin 2018pro or GraphPad Prism 7. The results were given as means ± standard error (SE). Statistical analyses between treatments or groups were determined using one-way analysis of variance (ANOVA) models in SAS 9.2 (SAS Institute Inc., Cary, NC, USA), followed by Dunnett's multiple comparisons test. A p-value of < 0.05 was considered statistically significant.