Enzymatically-crosslinked gelatin hydrogels containing paenipeptin and clarithromycin against carbapenem-resistant pathogen in murine skin wound infection

Background The recent rise and spread of carbapenem-resistant pathogens pose an urgent threat to public health and has fueled the search for new therapies. Localized delivery of topical antibiotics is an alternative for the treatment of infected wounds caused by drug-resistant pathogens. In this study, we aimed to develop antimicrobial-loaded hydrogels for topical treatment of wound infections in a murine skin wound infection. Results Paenipeptin analogue 1, a linear lipopeptide, potentiated clarithromycin against multidrug-resistant Acinetobacter baumannii, Enterobacter cloacae, Escherichia coli, and Klebsiella pneumoniae. Enzymatically-crosslinked gelatin hydrogels were developed to encapsulate paenipeptin analogue 1 and clarithromycin. The encapsulated antimicrobials were gradually released from hydrogels during incubation, reaching 75.43 and 53.66% for paenipeptin and clarithromycin, respectively, at 24 h. The antimicrobial-loaded hydrogels containing paenipeptin and clarithromycin synergistically resulted in 5-log reduction in carbapenem-resistant A. baumannii within 6 h in vitro. Moreover, the antimicrobial-loaded hydrogels reduced 3.6- and 2.5-log of carbapenem-resistant A. baumannii when treated at 4 or 20 h post infection, respectively, in a murine skin wound infection. Conclusions Enzymatically-crosslinked gelatin hydrogels loaded with paenipeptin analogue 1 and clarithromycin exhibited potent therapeutic efficacy against carbapenem-resistant A. baumannii in murine skin wound infection. Supplementary Information The online version contains supplementary material available at 10.1186/s12866-021-02383-z.


Background
Bacterial infections caused by carbapenem-resistant pathogens are difficult to treat and have been recognized as an urgent threat [1,2]. It was estimated that carbapenem-resistant Acinetobacter caused 8500 cases of infections and 700 deaths in the United States in 2017 [3]. Drug-resistant Acinetobacter is a challenging threat to patients in healthcare facilities. Acinetobacter can cause lung, wound, urinary tract, and bloodstream infections [3]. Wound infections caused by multidrugresistant pathogens, including Acinetobacter baumannii, are serious problems with regard to morbidity and mortality [4][5][6].
Localized delivery of topical antibiotics is an alternative to systemic antibiotics for the treatment of infected wounds [7]. The use of topical medications allows for the delivery of high concentrations of antimicrobials at the site of infection. Antimicrobial-loaded hydrogels can serve as topical antimicrobial carriers for the treatment Open Access *Correspondence: ehuang@uams.edu 1 Department of Environmental and Occupational Health, University of Arkansas for Medical Sciences, 4301 West Markham Street, Little Rock, AR 72205, USA Full list of author information is available at the end of the article of wound infections [8,9]. For example, the lipopeptide antibiotic colistin was encapsulated in glycol chitosanbased hydrogels for the treatment of infection in mice. The localized release of colistin from hydrogels exhibited potent activity against Pseudomonas aeruginosa in the in vivo animal "burn" infection model [8]. Similarly, topically delivered moxifloxacin reduced P. aeruginosa and Staphylococcus aureus wound infections and prompted wound healing [9].
Gelatin is a mixture of water-soluble proteins obtained by collagen hydrolysis. Gelatin has great biocompatibility and biodegradability and has been extensively used as a biomaterial for tissue engineering and drug delivery [10,11]. Gelatin hydrogels can be made using chemical and enzymatic crosslinking to increase mechanical and thermal strength. However, the chemical crosslinking agents, such as glutaraldehyde, may be potentially cytotoxic [12], which limits its biomedical applications. Biochemical crosslinking approaches are more biocompatible when compared to chemical crosslinking. Microbial transglutaminase has been used to enzymatically form thermally stable gelatin hydrogels to be used as a tissue engineering scaffold [12] and a biomimetic tissue sealant [13,14]. The favorable biocompatibility of transglutaminase crosslinked gelatin hydrogels makes it a promising drug carrier for the delivery of topical antimicrobial agents.
Paenipeptins are synthetic linear lipopeptides that sensitize Gram-negative pathogens to antibiotics that have little effect when used alone [15]. We previously reported that paenipeptin analogues potentiated clarithromycin and rifampicin against 10 carbapenem-resistant pathogens, including five isolates of A. baumannii and five isolates of Klebsiella pneumoniae [16]. Moreover, systemic administration of paenipeptin analogues in combination with clarithromycin or rifampicin was effective against mcr-1-mediated polymyxin-resistant Escherichia coli in a neutropenic murine thigh infection model [17]. In this study, we aimed to develop antimicrobial-loaded hydrogels for localized delivery of paenipeptin-clarithromycin mixture for topical treatment of wound infections in a murine skin wound infection.

Paenipeptin analogue 1 potentiated clarithromycin against multidrug-resistant pathogens
The MIC range, MIC 50 , and MIC 90 of paenipeptin analogue 1, clarithromycin, and their combination are summarized in Table 1. Paenipeptin at sub-inhibitory concentration (4 μg/ml) enhanced the activity of clarithromycin against A. baumannii, Enterobacter cloacae, E. coli, and K. pneumoniae. The MIC 90 of clarithromycin when used alone was > 32 μg/ml for all tested isolates in these four genera, whereas the MIC 90 of clarithromycin in the presence of paenipeptin analogue 1 (4 μg/ml) ranged from 0.0625 to 8 μg/ml. Paenipeptin promoted the uptake of clarithromycin by disrupting the outer membrane of the susceptible Gram-negative pathogens [17]. However, paenipeptin failed to increase the activity of clarithromycin against P. aeruginosa. The results were consistent with the previous reports. For example, NAB741, a polymyxin B derivative, potentiated clarithromycin and other antibiotics against E. coli, K. pneumoniae and A. baumannii [18], but not P. aeruginosa [19]. Similarly, a linear cationic peptide, unacylated tridecaptin A1 (H-TriA1), substantially lowered the MIC of rifampicin against E. coli and K. pneumoniae strains but not P. aeruginosa [20]. However, the mechanism of lacking synergistic effect against P. aeruginosa remains unknown.

Release kinetics and time-kill kinetics of antimicrobial-loaded gelatin hydrogels in vitro
The encapsulated antimicrobials were gradually released from hydrogels during incubation, reaching 75.43 and 53.66% for paenipeptin and clarithromycin, respectively, at 24 h ( Table 2). In the in vitro time-kill kinetics tests in tryptic soy broth, single treatments by hydrogels loaded with paenipeptin analogue 1 or Similarly, paenipeptin analogue 1 slowed the growth of A. baumannii AR 0063 by 6 h, but the final population of the bacterium reached a similar level as the PBS control at 24 h. Clarithromycin slightly reduced the growth of A. baumannii AR 0063 but the final population of the bacterium also exceeded the inoculum level and reached 8.0 log CFU/ml at 24 h. In contrast, the combined treatments with both compounds synergistically resulted in 5-log reduction in both strains within 6 h (Fig. 1).

In vivo therapeutic efficacy of antimicrobial-loaded gelatin hydrogels
In the murine skin infection model, single treatments started at 4 h post infection by gelatin hydrogels containing paenipeptin and clarithromycin resulted in 1.6-and 2.5-log reduction of A. baumannii ATCC 19606, respectively. The combined treatment started at 4 h post infection led to 4.2 log-reduction whereas the combined treatment started at 20 h post infection resulted in 2.8log reduction (Fig. 2). The results indicated that bacterial pathogens in the established wounds (20 h post infection) became less susceptible to antimicrobial treatments than that in the freshly infected wounds (4 h post infection). Therefore, the initiating time for treatment after infection affected the therapeutic efficacy. It was reported that the common topical antimicrobial agents, including mupirocin and bacitracin, had reduced efficacy against methicillin-resistant S. aureus (MRSA) when treatment was initialed at 24 h post infection compared to 4 h after infection in a superficial murine wound model [21]. We further studied the therapeutic efficacy against a carbapenem-resistant strain, A. baumannii AR0063. Similarly, when the treatments started at 4 h post infection, single treatments by hydrogels with paenipeptin and clarithromycin resulted in 2.0-and 2.6-log reduction of A. baumannii AR0063, respectively. The combined treatment started at 4 h post infection led to 3.6-log reduction whereas the combined treatment started at 20 h post infection resulted 2.5-log reduction (Fig. 3). Other antimicrobial peptides have been studied as topical   [22].

Skin irritation tests of antimicrobial-loaded gelatin hydrogels
Gelatin hydrogels containing both paenipeptin and clarithromycin at 0.1-0.4 mg/ml did not cause any signs of irritation after placing the hydrogels on the mice skin for 24 h (Table 3). However, at the end of 48 h extended incubation with a new hydrogel at 24 h, skin rash was observed in four out of six mice treated with hydrogels ( Fig. S1) containing both paenipeptin and clarithromycin at the highest tested concentration (0.4 mg/ml), which was four times of the effective therapeutic concentration tested in the wound infection model.

Conclusions
Paenipeptin analogue 1 potentiated clarithromycin against multidrug-resistant Gram-negative pathogens, including A. baumannii, E. cloacae, E. coli, and K. pneumoniae but not P. aeruginosa. Moreover, we developed enzymatically-crosslinked gelatin hydrogels loaded with paenipeptin analogue 1 and clarithromycin. Gelation without using transglutaminase resulted in hydrogels that were not heat stable at 37 °C [12]. The enzymatic gelation method increased the thermal stability of hydrogels without using toxic crosslinking agents, such as glutaraldehyde. To the best of our knowledge, this is the first report using transglutaminase crosslinked gelatin hydrogels as a carrier for the delivery of topical antimicrobial agents. The antimicrobial-loaded hydrogels gradually released antibiotics and exhibited potent therapeutic efficacy against carbapenem-resistant A. baumannii in murine skin wound infection.
(Piscataway, NJ). The purity was determined by HPLC, and the peptide structure was verified by high resolution mass spectrometry (HRMS) and nuclear magnetic resonance (NMR) as described previously [15].

Minimum inhibitory concentration determination
The minimum inhibitory concentration (MIC) of paenipeptin analogue 1, and clarithromycin (Sigma, St. Louis, MO) with or without paenipeptin analogue 1 was determined as described previously [15]. In this current study, we expanded the susceptibility tests against 70 isolates in the Gram-Negative Carbapenemase Detection panel obtained from the FDA-CDC Antibiotic-Resistance (AR) Bank. The majority of these isolates in the panel are resistant to carbapenem antibiotics [23].

Release kinetics of antimicrobials from hydrogels
To determine the release kinetics of antimicrobials, antimicrobial-loaded gelatin hydrogels were incubated at 37 °C for 24 h. Paenipeptin analogue 1 and clarithromycin released from hydrogels in aqueous solution were collected at 3, 4, 6, 8, 10, and 24 h during incubation, followed by quantification using HPLC as described previously [15]. Briefly, the gelatin hydrogels (0.9 ml) contained paenipeptin analogue 1 and clarithromycin at an equal concentration of 0.1 mg/ml. Release experiments (three independent replicates) were carried out in vitro during incubation at 37 °C. Quantification was performed using Waters Acquity UPLC with Waters BEH C 18 Column (130 Å, 1.7 μm, 2.1 mm × 50 mm).

Time-kill kinetics in vitro
The activity of antimicrobial-loaded gelatin hydrogels containing paenipeptin analogue 1 (0.1 mg/ml), clarithromycin (0.1 mg/ml), or both compounds (0.1 mg/ml for each) was evaluated in vitro using time-kill kinetics assay. Briefly, antimicrobial-loaded hydrogels (180 μl) was placed in 2 ml tryptic soy broth, which was inoculated with A. baumannii ATCC 19606 or the carbapenem-resistant strain, A. baumannii AR 0063, followed by incubation at 37 °C with agitation at 80 rpm for 24 h. Bacterial population was determined at 0, 1, 2, 4, 6 and 24 h. Data were analyzed using one-way analysis of variance (ANOVA) (GraphPad Prism, version 9.0; GraphPad Software Inc., San Diego, CA) and considered significant at p < 0.05. After hair removal, a partial-thickness skin wound was generated under isoflurane anesthetization by repeated brief touching the back skin of CD-1 mouse using a rotary tool (Dremel 8050) with a sterile sanding attachment until skin tissues were red and glistening [21]. After disinfection with alcohol wipes, an area of ~ 1 cm 2 was inoculated with A. baumannii ATCC 19606 or A. baumannii AR 0063 (5.5-6.0 log CFU in 5 μl saline). At 4 or 20 h post infection, an antimicrobial-loaded gelatin hydrogel (0.9 ml), which was loaded on a bandage, was applied and secured to the infected wound. After antimicrobial treatment overnight for 18 h, mice were euthanized and wound tissues were collected for bacterial enumeration. All animals were euthanized by CO 2 followed by cervical dislocation. Data were analyzed using one-way analysis of variance (ANOVA) (GraphPad Prism, version 9.0; GraphPad Software Inc., San Diego, CA) and considered significant at p < 0.05.

Skin irritation tests
To test the potential skin irritation on mice (female and male CD-1, n = 6), gelatin hydrogels containing paenipeptin analogue 1 and clarithromycin at various concentrations (0.1, 0.2, or 0.4 mg/ml for both compounds) were placed and secured on mice skin for 24 h. After 24 h, the hydrogel was removed and replaced by a new hydrogel at the same concentration for another 24 h. All animals were euthanized by CO 2 followed by cervical dislocation.
Signs of irritation such as redness or rash were observed at 24 and 48 h.
Pictures were taken at 48 h after 2 repeated treatments at every 24 h. Black inks were used to mark the skin to guide the placement of the hydrogels. Red arrow at 0.4 mg/ml indicated the rash on the skin.