Background
The first use of human placental membranes for the treatment of diseases occurred with Chinese and Japanese traditional medicines centuries ago, and around the turn of the 20th century for Western cultures [
1,
2]. Since these early applications, investigators identified that there are antibacterial and other paracrine factors in amniotic membrane that modulate the wound healing process [
1,
3]. Like amniotic membrane, amniotic fluid (AF) plays important roles in the development and the protection of the fetus [
4]. One property of AF that helps to defend the fetus against pathogens is its antibacterial activity [
5‐
7]. AF contains numerous defense proteins and cytokines and antibacterial peptides, like cystatin C, lactoferrin and lysozyme [
8‐
11]. Other components that are present in AF that confer antibacterial activity include transferrin, β-lysin, peroxidases, immunoglobulins, and zinc-peptide complexes [
8,
12‐
14]. Also, chemokines such as CXCL1 and CXCL14 with known antibacterial activities are present in AF [
15].
Amniotic fluid is a complex matrix that can be viewed as containing multiple fractions: (1) an insoluble fraction that consists of components like lanugo, vernix and cellular elements; (2) a soluble fraction that is made up of carbohydrates, proteins, lipids, electrolytes and metabolites and; (3) a fraction that contains extracellular vesicles (ECVs). Based on the knowledge that cytokines, growth factors and chemokines play major roles in activating endogenous mechanisms to facilitate repair and regeneration, we developed a novel approach to produce a processed AF (pAF) that eliminates a majority of its insoluble components (i.e. lanugo, vernix and cells) [
16]. Interestingly, despite the removal of these insoluble components that include the stem cells, pAF supports the healing of chronic and acute wounds (unpublished observation). This observation suggests that active ingredients retained within the soluble fraction of pAF have properties capable of facilitating wound healing. From a previous study, we reported that 76% of a total of 400 proteins that we tested were present in pAF and that a majority of these proteins had roles in host defense [
16]. Among the host defense peptides are proteins known for their role in the inflammatory response, innate immunity, immune modulation, and/or as having antibacterial activity.
Chronic wound environments are often colonized with microbes [
17] and this colonization by microorganisms likely results in delayed healing due to infection [
18,
19]. Delays in the healing process for chronic wounds is also attributable to prolonged inflammation caused by a lack of proper growth factors and cytokines [
20]. Consequently, therapies that have antibacterial and anti-inflammatory activity would be a good choice for treating wounds. Recognizing that pAF facilitates the healing of chronic and acute wounds, the goals of this study were to assess the antibacterial activity of unprocessed AF relative to pAF, to evaluate the antibacterial activity of pAF from different donors, and to begin to characterize the role of specific antibacterial proteins in pAF against a panel of wound-associated pathogens.
Materials and methods
Collection and processing of amniotic fluid
Donor consent, screening and infectious disease testing of amniotic fluid (AF) were previously described [
16]. The collection and processing of AF were also previously described [
16]. Prior to processing AF, several aliquots were removed and stored frozen at − 80 °C (i.e. pre-processed AF). Upon the completion of the final processing step, aliquots of pAF were removed and the samples were maintained at − 80 °C until the time of use.
Bacterial culture and preparation of inoculums
Clinical isolates of ESKAPE bacterial strains were purchased from ATCC (Manassas, VA):
E. faecium ATCC
® 51559™,
S. aureus ATCC
® 25923™,
K. pneumoniae ATCC
® 700603,
A. baumannii ATCC
® 49466™,
P. aeruginosa ATCC
® 15692™, and
E. aerogenes ATCC
® 49469™. All bacterial strains were cultured and maintained as instructed by ATCC
®. Optimal culture media used are tryptic soy broth (TSB) for
S. aureus,
P. aeruginosa, E. aerogenes,
A. baumannii, brain heart infusion (BHI) medium for
E. faecium and nutrient broth for
K. pneumoniae. Preparation of bacterial inoculum was performed as described previously [
21]. Briefly, bacteria were cultured in their defined optimal culture medium at 37 °C with shaking until optical density reached 0.2 to 0.6 at 600 nm (OD
600). The number of colony forming units (CFUs) for each strain was estimated based on an OD
600 = 1.0, which corresponds to 10
9 CFU/mL. To prepare the inoculum for antibacterial assays, the bacterial stocks were serially diluted with culture medium or pAF to approximately 1 × 10
3 CFU/mL of bacteria. For each experiment, the actual CFU of each inoculum was determined by preparing serial dilutions and then plating onto TSB agar plates (BD, Franklin Lakes, NJ).
Inhibition of bacterial growth by pAF
An inoculum of P. aeruginosa or S. aureus (1 × 103 CFU/mL) was added to pAF that had been prepared by diluting pAF in TSB. The cultures were incubated at 37 °C for 24 h. The growth of bacteria was monitored using an alamarBlue assay (ThermoFisher Scientifics; Waltham, MA) by following the manufacture’s protocol. The fluorescent intensity was measured using a TECAN Spark 10 M plate reader (TECAN, Morrisville NC) at Ex560 nm/Em590 nm. The growth of bacteria in the presence of pAF was normalized to the growth in the absence of pAF.
Quantification of bacterial growth in the presence of pAF
Bacterial growth was assessed as described with modifications [
22]. Each strain of bacteria (1 × 10
3 CFU/mL) was added to 1 mL of AF or 1 mL optimal culture medium. The cultures were incubated at 37 °C with shaking for 24 h. Serial dilutions were then prepared for each culture and plated onto TSB agar plates. CFUs were counted after overnight incubation at 37 °C. The antibacterial activity of pAF was expressed in log reductions, which was calculated as the log
10CFU (Control) − log
10CFU (pAF). Data from two independent experiments (n = 3 for each experiment) were pooled together to calculate the mean and standard deviation.
Protein array
Quantitative Protein arrays were performed as previously described [
16]. Briefly, pAF from three maternal collections were sent to RayBiotech to simultaneously and quantitatively measure the concentration of 400 human cytokines using the Quantibody
® Human Cytokine Antibody Array 9000 (RayBiotech, In., Norcross, GA). Controls and serial dilutions of cytokine standards were prepared according to the manufacturer’s instructions and were added to chip wells. After processing the chips according to the manufacturer’s instructions, the chips were analyzed using the Quantibody
® Q-Analyzer software (RayBiotech, Inc.). Proteins were classified according to their biological function by surveying the Human Protein Reference Database (
http://www.hprd.org/index_html), Cytokines & Cells Online Pathfinder Encyclopedia (COPE)
http://www.copewithcytokines.de/), GeneCards
® (
http://www.genecards.org/), and the biomedical literature in PubMed (
http://www.ncbi.nlm.nih.gov/pubmed).
Detection of human lysozyme, cystatin C and lactoferrin in pAF using ELISA
The presence of lysozyme, cystatin C and lactoferrin from different lots of pAF was quantified using quantitative sandwich ELISA assays according to the manufacturer’s instructions (Abcam, Cambridge, MA): Lysozyme (Human Lysozyme ELISA kit ab108880), Lactoferrin (Human Lysozyme ELISA kit ab200015), and cystatin C (Human Lysozyme ELISA kit ab179883).
Immunoprecipitation (IP)
Selective depletion of lysozyme and lactoferrin from AF was accomplished by immunoprecipitation (IP) as described [
22]. Briefly, rabbit polyclonal anti-lysozyme antibody (ab2408) and anti-lactoferrin antibody (ab15811) were purchased from Abcam. Anti-lysozyme antibodies (15 μg/mL) and anti-lactoferrin antibodies (4 μg/mL) were added individually (IP) or together (Co-IP) to 1 mL of AF. As a control, the same volume of PBS was added to 1 mL of AF. The samples were mixed at 4 °C for overnight. A 50 µL slurry of Protein A/G agarose plus resin: sc-2003 (Santa Cruz Biotechnology; Dallas, TX) was washed twice with 1 mL of PBS and mixed with each pAF with or without antibodies for 4 h at 4 °C. Mixtures were then centrifuged at 2000 rpm for 3 min and resulting supernatants were transferred to individual bacterial culture tubes. Inoculums of 1 × 10
3 CFU of
P. aeruginosa or
S. aureus were added to each culture tube and incubated with shaking at 37 °C. After 24 h, CFUs for each culture were quantified by serial dilution plating as previously described [
22].
Statistical analysis
Each independent experiment contained 3 or more biological repeat samples (n ≥ 3), and data is presented as the mean ± standard deviation. One-way ANOVA with a Tukey’s multiple comparisons test was performed to determine statistical significance. Differences were considered significant at a p value of < 0.05.
Discussion
Reports of stem cells in amniotic fluid (AF) has attracted the efforts of a number of groups for developing commercial products of AF for the treatment of wounds and other therapeutic applications [
25,
26]. Based on our observation that there is a paucity of stem cells in AF and that expansion of the stem cells would be required to obtain therapeutic doses. We developed a processing method that removes a majority of the insoluble components (i.e. cells, lanugo and vernix) while retaining soluble components (i.e. cytokines, growth factors, chemokines, antibacterial peptides) and extracellular vesicles [
16]. Our formulation of pAF is based on the hypothesis that the soluble components and extracellular vesicles found in AF will activate endogenous mechanisms to facilitate immune responses, tissue repair and regenerative events.
Recognizing that delayed healing can occur when wounds are colonized with microbes and that a therapy with antibacterial activity is advantageous towards healing infected wounds, this study determined whether AF depleted of its insoluble components retained its antibacterial activity. We found that pAF maintains its antibacterial activity relative to unprocessed AF and that the soluble components in pAF are key contributors to its antibacterial activity. This study also shows that differences in the antibacterial activity of different lots of pAF exist against different pathogens (Table
1). Given that a standardized method is used to process AF, it is more likely that the differences in the antibacterial activity of different lots is due to inherent differences among donors rather than to the manufacturing strategy.
Of the 31 known anti-microbial proteins that we tested, 26 of them are present in pAF. Cystatin C and lactoferrin were amongst the highest expressed antibacterial proteins that we tested. Cystatin C not only had the highest quantitative levels among antibacterial proteins, but was also among the top 5 most highly expressed proteins of ~ 300 proteins that we identified and quantified in pAF [
16]. Cystatin C is a cysteine protease inhibitor [
27] that is present in almost all tissue and body fluids and is reported to be a potent regulator of the inflammatory response. Cystatin C plays important roles in innate immunity by binding to components of the classical complement pathway and by modulating the actions of neutrophils via superoxide inhibition and chemotaxis [
28‐
30]. Lactoferrin, is the most abundant protein in the whey fraction of human milk [
31]. Lactoferrin plays a critical role in protecting the newborn infant from infection via its iron binding function that inhibits bacteria, fungus, viral and parasitic infections and by its anti-inflammatory and immunomodulatory activities [
32‐
34].
Seventeen of the antibacterial proteins that we identified in pAF also possess chemokine activity, a property that involves directing the immune response to sites of injury and infection. For example, PARC, the most highly expressed chemokine in pAF triggers lymphocyte responses, but not neutrophils [
35]. While GRO and GROa attract neutrophils [
36]. Others of the identified chemokines present in pAF are proteins that can attract monocytes, eosinophils, and even basophils. Moreover, some chemokines have also been shown to have antibacterial activity that directly interfere with infectious agents [
15]. The significance of this finding is that different chemokines in pAF may recruit different types of white cells to sites of infection, inflammation and/or injury to promote repair of the site.
To determine whether a relationship could be established between the levels of antibacterial proteins and their activity against specific pathogens, we focused our effort on 3 proteins found in AF, lactoferrin, cystatin C and lysozyme. Knowing that lactoferrin has antibacterial activity against
S. aureus [
37] and lysozyme has lytic activity against
P. aeruginosa over
S. aureus [
24,
38], we were particularly interested in whether the levels of these proteins correlated with the bactericidal activity of pAF. We found that of six pAF lots tested, all lots except for lot 2 showed bactericidal activity against
S. aureus (Fig.
2b) and lot 2 also had the lowest levels of lactoferrin (Fig.
4c). Likewise, lower levels of lysozyme in lots 3 and 5 (Fig.
4a) corresponded with lower anti-
P. aeruginosa activity (Table
1). However, lower levels of lysozyme activity for lot 6 did not correspond to lower anti-
P. aeruginosa activity. Also, the levels of cystatin C, lysozyme and lactoferrin in lot 6 were not higher than that of lot 5, but lot 6 had better overall antibacterial activity than lot 5. These results suggest that lactoferrin and lysozyme levels may contribute to differential antibacterial activities of pAF. However, we did not identify one single protein of the 3 antibacterial proteins we tested whose level in AF determines the antibacterial activity of AF. This suggests that other proteins/peptides with antibacterial proteins in pAF are likely to contribute to pAF’s antibacterial activity. Evidence that other antibacterial proteins in pAF are involved is supported by the results of our immunoprecipitation study. By specifically depleting lactoferrin and lysozyme from pAF, we showed that there was a direct involvement of lysozyme and lactoferrin in the antibacterial activity of pAF. However, the depletion of these proteins without abolishing the antibacterial activity of pAF indicates that additional antibacterial factors are involved in the antibacterial activity of pAF.
Conclusion
The results of this study indicate that even after the removal of the insoluble components of human AF that the antibacterial activity of pAF is preserved and that a majority of the antibacterial activity of AF is contained within the soluble fraction. As we move towards controlled clinical studies to investigate the efficacy of using pAF in patients with burns and wounds, the antibacterial proteins identified under this study and their expression levels will be valuable in helping to assess their contributions in clinical outcome. In conclusion, pAF is rich in host defense proteins in which some are known for their antibacterial, antifungal, antiviral, anti-parasitic, anti-inflammatory and immunomodulatory activities.
Authors’ contributions
All authors were provided copies of the manuscript for their review. YM: study design, performed assays, analyzed and interpreted data, and helped to write the manuscript. JP: collection of birth tissue. AS-V: performed assays. MB: reviewed donor acceptance criteria. JK: study design. JAR: study design, analyzed and interpreted data and helped to write the manuscript. All authors read and approved the final manuscript.