Platelet-rich plasma affects bacterial growth in vitro

Cytotherapy, 2014; 16: 1294e1304

Platelet-rich plasma affects bacterial growth in vitro ERMINIA MARIANI1,2, GIUSEPPE FILARDO3, VALENTINA CANELLA1, ANDREA BERLINGERI4, ALESSANDRA BIELLI4, LUCA CATTINI1, MARIA PAOLA LANDINI4, ELIZAVETA KON3, MAURILIO MARCACCI3 & ANDREA FACCHINI1,2 1

Laboratory of Immunorheumatology and Tissue Regeneration/RAMSES, Rizzoli Orthopaedic Institute, Bologna, Italy, 2Department of Medical and Surgical Sciences, University of Bologna, Bologna, Italy, 3Laboratory of Biomechanics and Technology Innovation/NABI, 2nd Orthopaedic and Traumatologic Clinic, Rizzoli Orthopaedic Institute, Bologna, Italy, and 4Operative Unit of Clinical Microbiology, St. Orsola-Malpighi University Hospital, Regional Reference Centre for Microbiological Emergencies, Bologna, Italy Abstract Background aims. Platelet-rich plasma (PRP), a blood derivative rich in platelets, is a relatively new technique used in tissue regeneration and engineering. The increased quantity of platelets makes this formulation of considerable value for their role in tissue healing and microbicidal activity. This activity was investigated against five of the most important strains involved in nosocomial infections (Escherichia coli, Staphylococcus aureus, Pseudomonas aeruginosa, Klebsiella pneumoniae and Streptococcus faecalis) to understand the prophylactic role of pure (P)-PRP. Microbicidal proteins released from activated P-PRP platelets were also determined. Methods. The microbicidal activity of P-PRP and platelet-poor plasma (PPP) was evaluated on different concentrations of the five bacterial strains incubated for 1, 2, 4 and 18 h and plated on agar for 18e24 h. P-PRP and PPP-released microbicidal proteins were evaluated by means of multiplex beadebased immunoassays. Results. P-PRP and PPP inhibited bacterial growth for up to 2 h of incubation. The effect of P-PRP was significantly higher than that of PPP, mainly at the low seeding concentrations and/or shorter incubation times, depending on the bacterial strain. Chemokine (CC motif) ligand-3, chemokine (C-C motif) ligand-5 and chemokine (C-X-C motif) ligand-1 were the molecules mostly related to Pseudomonas aeruginosa, Staphylococcus aureus and Streptococcus faecalis inhibition. Escherichia coli and Klebsiella pneumoniae were less influenced. Conclusions. The present results show that P-PRP might supply an early protection against bacterial contaminations during surgical interventions because the inhibitory activity is already evident from the first hour of treatment, which suggests that physiological molecules supplied in loco might be important in the time frame needed for the activation of the innate immune response. Key Words: bacterial growth, kinocidins, microbicidal activity, microbicidal proteins, nosocomial infections, platelet-rich plasma

Introduction The use of autologous platelet-rich plasma (PRP), a blood derivative rich in platelets easily obtainable from a patient blood sample after centrifugation, is a relatively new technique and a new appealing method used in a variety of procedures, such as bone regeneration, soft tissue maturation (chronic or acute tendinopathies, ligament reconstruction, muscle lesions and cartilage treatment, etc) and in burn treatment and tissue engineering (1e6). The use of PRP enables the treatment of patients with their own derivatives and growth factors, thus preventing infection transmission and immunological reactions. The increased quantity of platelets, which

distinguishes pure PRP (P-PRP), makes this formulation of considerable value not only for the fundamental role of platelets in homeostasis regulation and tissue healing, but also for its microbicidal activity, a possibly important feature of P-PRP. Indeed, platelets are a rich reservoir of molecules that are able to interact and interfere with different microorganisms (7), thus acting as host defense effector cells. In addition to platelets, PRP preparations may also be enriched with leukocytes (L), giving rise to the L-PRP, which, thanks to this cellular component, contains additional leukocyte-derived proteins with microbicidal activity. Both the enriched preparations (P-PRP and L-PRP) can be variously activated to

Correspondence: Erminia Mariani, MD, Laboratory of Immunorheumatology and Tissue Regeneration, Rizzoli Orthopaedic Institute, Via di Barbiano 1/10, 40136 Bologna, Italy. E-mail: [email protected] (Received 23 October 2013; accepted 17 June 2014) ISSN 1465-3249 Copyright Ó 2014, International Society for Cellular Therapy. Published by Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.jcyt.2014.06.003

PRP and bacterial growth induce the release of a greater amount of soluble factors (8). Interaction of platelets with microbial pathogens may develop directly by contact (proceeding from platelet aggregation, shape transition and microtubule organization, to platelet degranulation and the secretion of an array of host defense molecules) and indirectly by bridging matrices (7). Depending on the diverse bacteria, these fundamental steps may differ, thus leading to various platelet antimicrobial response profiles, which might significantly influence host defense versus the virulence of a given pathogen. Activated platelets release a high number of platelet microbicidal proteins, most of which have a homologous structure with classic chemokines (containing a CC or CXC motif), thus called “kinocidins because of their dual role both as chemokines and microbicidal effectors (7,9). As far as we know, platelet factor (PF)-4 (CXCL-4), neutrophilactivating protein (NAP)-2 (CXCL-7), interleukin (IL)-8 (CXCL-8) and regulated upon activation, normal T cell expressed and secreted protein (RANTES) (CCL-5) are the best examples of these molecules, but other antimicrobial polypeptides contained in platelets appear to have a similar structure and microbicidal activity (7,9). As the possibility to develop infections after invasive procedures or with the use of medical devices (10) is an important parameter for the success of surgery, in the present study the potential role of P-PRP as microbicidal effector was investigated. The potential antibacterial activity of P-PRP was tested against five of the most important strains (Escherichia [E] coli, Staphylococcus [S] aureus, Pseudomonas [P] aeruginosa, Klebsiella [K] pneumoniae and Streptococcus [S] faecalis) potentially involved in bone, soft tissue and wound infections (10e12), to understand and possibly improve the role of P-PRP in preventing and controlling nosocomial infections.

Methods Donors Ten healthy men (mean age  standard deviation: 29.9  3.4 years), enrolled on a voluntary basis, signed a written informed consent to the study protocol that was approved by the institutional ethics committee. Subjects with systemic disorders and infections, presenting hemoglobin concentrations lower than 11 g/dL or platelet numbers of lower than 150  103/mL, were excluded from the study as were subjects with smoking habits and assuming non-steroidal antiinflammatory drugs in the 5 days before blood collection. Subject anonymity was ensured by code numbers assigned to the samples.

1295

Preparation of platelet concentrate Forty-five milliliters of venous blood was collected from each volunteer and divided into five tubes containing 1 mL of tri-sodium citrate solution (3.8%) as anticoagulant. The blood was centrifuged at 460g for 8 min (13,14) and platelet-poor plasma (PPP), P-PRP and erythrocyte concentrates were stratified separately. In sterile conditions, approximately 1 mL per tube of PPP, located at the top of the stratified solution, was aspirated. A similar volume of P-PRP, located on the red blood cell pellet, was then carefully harvested, avoiding leukocytes. The median platelet number was 290  103/mL (interquartile range, 182e385  103/mL) in P-PRP and 6  103/mL in PPP (interquartile range, 4e7  103/mL); white blood cell number was <200/mL in P-PRP and undetectable in PPP, as determined by an automated hematology analyzer (Coulter LH 750; Beckman Coulter Inc, Miami, FL, USA) (linearity was 5e1000  103/mL for platelet count and 0.1e100  103/mL for white blood cell count). Each sample was divided in two aliquots, one for the evaluation of antimicrobial activity and the second one for the quantification of microbicidal protein release.

Evaluation of antimicrobial activity Escherichia coli (American Type Culture Collection [ATCC] 25922; Manassas, VA, USA), S aureus (ATCC 29213), K pneumoniae (ATCC 700603), P aeruginosa (ATCC 27853) and S faecalis (ATCC 29212) were chosen as those bacterial species implicated in bone and soft-tissue infections (15). In addition, some of these pathogens have been selected as a reference for the grouping of microorganisms on the basis of their morphological characteristics of membrane and growth (for example, E coli and K pneumoniae as enterobacteriaceae, P aeuruginosa as Gram-negative non-fermenting bacteria) (16). Suspension of 108 colony-forming units (CFU)/ mL in tryptic soy broth (TSB) medium (MEUS S.R.L., Italy) were prepared for the culture of the selected bacterial species. Ten-fold serial dilutions (106, 105 and 104 CFU/mL) were prepared to obtain bacterial concentrations indicative of a localized infection in vivo (17,18). Ninety microliters of each P-PRP and PPP sample, activated with 10 mL of CaCl2 (calcium chloride, 22.8 mmol/L final concentration), were added to tubes containing 900 mL of each bacterial concentration. The tubes were then incubated at 37 C for 1, 2, 4 and 18 h; 10 mL of each sample as then plated on blood horse agar (MEUS S.R.L.) with the use of a seeding distributed in three quadrants with an appendix in the terminal part (Figure 1).

1296

E. Mariani et al.

A semi-quantitative evaluation of bacterial growth was performed after 18e24 h of incubation at 37 C by counting the colonies in each quadrant, thus allowing a rough estimate of the bacterial growth (CFU/mL), as indicated in Table I.

Quantification of microbicidal protein release P-PRP and PPP were immediately activated with 10% CaCl2 (22.8 mmol/L, final concentration) to improve soluble factor release and incubated at 37 C in 5% CO2 for 1 h and 18 h, corresponding to the first and last time point of bacterial incubation. After centrifugation (15 min at 2800 g at 20 C), the supernatants were collected and frozen at 30C until use. For the evaluation of microbicidal proteins, P-PRP and PPP samples were assayed in duplicate. Commercially available multiplex beadebased sandwich immunoassay kits were used to simultaneously evaluate the following soluble factors: macrophage inflammatory protein (MIP)-1a (CCL3), RANTES (CCL-5), growth-regulated protein homolog (GRO)-a (CXCL-1), IL-8 (CXCL-8), soluble CD40 ligand (sCD40L) (Bio-Rad Laboratories, Hercules, CA, USA); PF4 (CXCL-4), epithelial neutrophil-activating protein (ENA)-78 (CXCL-5) and NAP2 (CXCL-7), Fractalkine (CX3CL-1), epidermal growth factor (EGF), platelet-derived growth factor (PDGF)-AB/BB, (Milliplex MAP Kit; Millipore, Bedford, MA, USA), as previously described (19). Briefly, distinct sets of fluorescently dyed beads loaded with capture monoclonal antibodies, specific for each cytokine to be tested, were used. Plasma samples (50 mL/well, diluted according to the manufacturer’s recommendations) or standards (50 mL/well) were incubated with 50 mL of pre-mixed bead sets inside the wells of a 96-well microtiter plate. The formation of different sandwich immune complexes on distinct bead sets was measured and

quantified by use of the Bio-Plex Protein Array System (Bio-Rad Laboratories). A 50-mL volume was sampled by each well and the fluorescent signal of a minimum of 50 beads per region (chemokine/ cytokine) was evaluated and recorded. Values with a coefficient of variation above the 10% were discarded before the final data analysis. Data were analyzed by the Bio-Plex Manager software version 6.0 (Bio-Rad Laboratories). Standard levels between 70% and 130% of the expected values were considered to be accurate and were used. In general, at least six standards were accepted and used to establish standard curves, following a FiveParameter Logistic (5PL) regression model. Sample concentrations were immediately interpolated from the standard curves. Statistical analysis Values are presented as medians, interquartile ranges and extreme values, as appropriate. Analysis was performed with the use of the Friedman analysis of variance (ANOVA) test for comparisons among bacterial growth in the different culture conditions (PRP, PPP and control medium). Differences between culture conditions were analyzed by means of the Wilcoxon matched pair test or the MannWhitney test, as appropriate. Correlations between concentrations of released soluble factors and the growth of each bacterium at different time points were analyzed by means of the Kendall-Tau rank correlation test. The level of statistical significance was set at P < 0.05. Data were analyzed with the use of Statistica 6 software (StatSoft Inc, Tulsa, OK, USA). Results Antimicrobial activity of P-PRP and PPP The effects of P-PRP and PPP on bacterial growth inhibition were evaluated in E coli, S aureus, K pneumoniae, P aeruginosa and S faecalis at different bacterial concentrations (104, 105, 106 CFU/mL) compared with spontaneous bacterial growth in Table I. Correlation between the number of colonies detected on each quadrant and approximate growth (CFU/mL). 1st 2nd 3rd Approximate Quadrant Quadrant Quadrant Appendix growth (CFU/mL)

Figure 1. Plating pattern. (I) First quadrant; (II) second quadrant; (III) third quadrant. App., appendix.

0 1e4 5 5 5 5

<5 5 5 5

<5 5 5

0 1

<103 103 104 105 106 >106

PRP and bacterial growth

Figure 2. Escherichia coli growth. Comparison between P-PRP (dark gray boxes) and PPP (light gray boxes). Standard control growth (white boxes) is also indicated. Results are reported as medians, interquartile ranges (boxes) and extreme values (whiskers). Friedman ANOVA: (A) P < 0.0001; (B, C) P < 0.001; (D) P < 0.005; (E) P < 0.02; (F) not significant.

control medium without blood-derived products and at different periods of incubation (1, 2, 4 and 18 h, respectively). Overall, both preparations (P-PRP and PPP) produced a fair growth inhibition for at least up to 2 h of incubation whatever the strain examined, although not for all the seeding conditions. Because none of the strains were entirely inhibited, the number of bacteria increased with the time of incubation (4 h), reaching comparable growth among P-PRP, PPP and control conditions after 18 h of

1297

Figure 3. Pseudomonas aeruginosa growth. Comparison between P-PRP (dark gray boxes) and PPP (light gray boxes). Standard control growth (white boxes) is also indicated. Results are reported as medians, interquartile ranges (boxes) and extreme values (whiskers). Friedman ANOVA: (A) P < 0.0001; (B, C) P < 0.001; (D) P < 0.005; (E) P < 0.0005; (F) not significant.

incubation, independently of the starting bacterial concentration and whatever the strain examined. Escherichia coli (Figure 2), P aeruginosa (Figure 3) and S faecalis (Figure 4) growth was different, depending on P-PRP, PPP and control culture

1298

E. Mariani et al.

Figure 5. Staphylococcus aureus growth. Comparison between P-PRP (dark gray boxes) and PPP (light gray boxes). Standard control growth (white boxes) is also indicated. Results are reported as medians, interquartile ranges (boxes) and extreme values (whiskers). Friedman ANOVA: (A, C) P < 0.0001; (B, G, H) P < 0.001; (D, E, I) P < 0.0005; (F) P < 0.005.

Figure 4. Streptococcus faecalis growth. Comparison between PRP (dark gray boxes) and PPP (light gray boxes). Standard control growth (white boxes) is also indicated. Results are reported as medians, interquartile ranges (boxes) and extreme values (whiskers). Friedman ANOVA: (A, C, E) P < 0.0001; (B) P < 0.002; (D) P < 0.005; (F) P < 0.05.

conditions for up to 2 h of incubation, when 104 and 105 CFU/mL were seeded (Friedman ANOVA: P < 0.005 at least). The difference was evident only after 1 h of incubation when increasing the number of E coli and P aeruginosa seeded bacteria to 106 CFU/mL (Friedman ANOVA: P < 0.02) (Figures 2 and 3), whereas S faecalis (Figure 4) showed a significant

difference for up to 2 h of incubation (Friedman ANOVA: P < 0.01). Concerning S aureus (Figure 5) and K pneumoniae (Figure 6), their growth varied according to P-PRP, PPP and control conditions for each bacterial concentration tested for up to 4 h of incubation (Friedman ANOVA: P < 0.05 at least). In particular, further analysis showed that E coli growth (Figure 2) in the presence of P-PRP or PPP was similar and always lower than that of the control condition after 1 and 2 h of incubation when 104 CFU/mL bacteria was seeded (P < 0.01, at least) (Figure 2A,B) and after 1 h with the inoculum of 105 and 106 bacteria (P < 0.02 at least) (Figure 2C,E). The growth of 105 CFU/mL seeded bacteria was significantly inhibited after 2 h of incubation only by P-PRP and not by PPP (Wilcoxon matched pairs test: P < 0.05) (Figure 2D).

PRP and bacterial growth Staphylococcus aureus growth (Figure 5), incubated both with P-PRP and PPP, was significantly lower than that of control medium conditions at every time stage up to 4 h and for all the bacterial concentrations tested (P < 0.01, at least) (Figure 5AeI). P-PRP was more effective than PPP after 1 h of incubation when 105 and 106 CFU/mL bacteria were seeded (Wilcoxon matched pairs test: P < 0.05 and P < 0.01, respectively) (Figure 5D,G). Klebsiella pneumoniae growth (Figure 6) in P-PRP and PPP was always significantly lower than that of control conditions (P < 0.05 at least) (Figure 6AeI). Bacterial growth was more inhibited by P-PRP compared with PPP after 1 and 2 h of incubation for 104 and 105 CFU/mL bacterial seeding (Wilcoxon matched pairs test: P < 0.05, at least) (Figure 6A,B,D,E) and for 106 CFU/mL seeding only after 2 h of incubation (Wilcoxon matched pairs test: P < 0.02) (Figure 6H). Concerning P aeruginosa (Figure 3), both P-PRP and PPP induced a significant growth inhibition at 1 and 2 h with 104 and 105 CFU/mL (Figure 3AeD) and only after 1 h for 106 CFU/mL concentrations (P < 0.001 at least) (Figure 3E). P-PRP induced a significant reduction of bacterial growth compared with PPP only after 1 h of incubation at 104 CFU/mL bacterial concentration (Wilcoxon matched pairs test: P < 0.05) (Figure 3A). Streptococcus faecalis growth (Figure 4) was inhibited after treatment with P-PRP and PPP compared with that of control growth conditions at 1 and 2 h for 104, 105 and 106 CFU/mL bacterial concentrations (P < 0.05, at least) (Figure 4AeF). Statistically significant differences between P-PRP and PPP were found after 1 h of incubation at 105 CFU/mL bacterial concentration (Figure 4C) (Wilcoxon pairs matched test: P < 0.05). Microbicidal protein evaluation Concentrations of the main soluble factors described as microbicidal proteins are reported in Table II. Correlations between the growth of the bacterial strains analyzed and the concentrations of specific soluble factors, which have been reported to play a role in bacterial growth inhibition, were performed, comparing bacterial growth for up to 4 h of incubation and the concentrations of microbicidal proteins released by P-PRP after 1 h of activation (Table III). After 18 h of P-PRP activation, all microbicidal protein concentrations were increased, excluding PF-4, ENA-78, NAP-2 and Fractalkine (data not shown). However, it was not possible to analyze the corresponding correlations with bacterial inhibition

1299

Figure 6. Klebsiella pneumoniae growth. Comparison between P-PRP (dark gray boxes) and PPP (light gray boxes). Standard control growth (white boxes) is also indicated. Results are reported as medians, interquartile ranges (boxes) and extreme values (whiskers). Friedman ANOVA: (A, B) P < 0.0001; (C, F) P < 0.01; (D, G) P < 0.0005; (E) P < 0.0002; (H) P < 0.005; (I) P < 0.02.

because the bacteria were in an overgrowth phase at 18 hours. MIP-1a, RANTES and GRO-a were the molecules mostly related to P aeruginosa, S aureus and S faecalis inhibition. In particular, the concentrations of these chemokines correlated with growth inhibition after 2 and/or 4 h of incubation, independent of the seeding bacterial concentrations (Kendall-Tau rank correlation test: P < 0.05 at least). In addition, S aureus and S faecalis strains showed a significant susceptibility to IL-8 at different incubation times (Kendall-Tau rank correlation test: P < 0.05 at least) and an isolated correlation at 4 h with sCD40L (Kendall-Tau rank correlation test: P < 0.05). Staphylococcus aureus growth was also correlated with EGF and Fractalkine, starting from 2 h of incubation (Kendall-Tau rank correlation test: P < 0.05 at least).

1300

E. Mariani et al.

Table II. Microbicidal protein concentrations in the different plasma preparations after 1 hour of incubation. Microbicidal proteins CCL-3/MIP-1a CCL-5/RANTES CXCL-1/GRO-a CXCL-4/PF4 CXCL-5/ENA-78 CXCL-7/NAP-2 CXCL-8/IL-8 CX3CL-1/Fractalkine sCD40L PDGF-AB/BB EGF

PRP 1.01 3443 88.5 688 1182.55 441,718 5.73 13.98 506 8705 89.27

[0e2.45] [2735e4754] [0e126.7] [638e978] [687.79e1343.27] [363,407e779,653] [4.40e9.92] [0e43.60] [450e864] [6342e12139] [34.66e171.33]

PPP 0 1867 0 131 96.48 188,030 3.20 0 0.70 291.7 9.45

[0e1.50] [990e3043] [0e10.42] [104e165] [27.23e187.31] [121,815e239,831] [1.81e0.09] [0e19.45] [0e18.95] [0e613.59] [4.23e15.98]

Wilcoxon matched pair test NS P < 0.05 NS P < 0.001 P < 0.0001 P < 0.05 NS NS P < 0.005 P < 0.0001 NS

Results are expressed as medians and [interquartile ranges] of 10 donors. NS, not significant.

Among the bacterial strains considered, E coli and K pneumoniae appear to be the least influenced by the analyzed molecules showing only few correlations. In particular, sCD40L concentration appeared to play an important role only in E coli growth inhibition; indeed, a significant correlation was evident from the 1st up to the 4th hour of incubation (Kendall-Tau rank correlation test: P < 0.05 at least). PDGF-AB/BB concentration correlated at 2 and 4 h, whereas Fractalkine showed only an isolated correlation at 4 h (Kendall-Tau rank correlation test: P < 0.05 at least). Concerning K pneumoniae, growth inhibition correlated with EGF, Fractalkine and ENA-78 concentrations only at 2 h (Kendall-Tau rank correlation test: P < 0.05). Finally, for PF-4 and NAP2, no correlation was observed with any of the bacterial strains examined. Discussion Over the past two decades, the regenerative potential of platelet concentrates during wound-healing and repair processes has been explored thoroughly (1,2,5,6), and these preparations have been used in different clinical areas (3,4,20), including chronic wounds, bone regeneration and orthopedic surgery, oral and maxillofacial surgery, plastic and cardiovascular surgery and sports medicine. The intrinsic microbicidal potential of platelets has been described (7), thus marking the concept of platelet concentrates as a rich reservoir of molecules acting also in host defense. Nonetheless, the antimicrobial activity mechanisms of platelet concentrates remain partially unknown. Several antimicrobial factors have been proposed, including platelet antimicrobial proteins and peptides of the innate immune response, whereas the respective impact of the plasma and cellular components has yet to be clarified; thus, opinions are contradictory and a matter of intense debate for the research community (21). For this study in particular, we used an autologous preparation of P-PRP characterized by a high

number of platelets and a negligible number of leukocytes, and we showed that the bacterial growth of five strains (E coli, S aureus, K pneumoniae, P aeruginosa and S faecalis) can be inhibited by the presence of PRP during in vitro culture. The five bacterial strains studied were chosen because they represent common causes of nosocomial infections of surgical origins (10e12). In general, it was observed that bacterial growth inhibition was transient and lasting up to 2 or 4 h of treatment, depending on the bacterial strain, thus suggesting a differential consumption of/susceptibility to bactericidal factors contained in P-PRP suspension. In E coli, the bacterial growth inhibition determined by P-PRP was early, within the first 2 h of incubation. Inhibition was also described after a relatively long experimental time (16e18 h) by Bielecki et al. (22), but the results are difficult to interpret. In fact, we cannot exclude that methodological differences and the presence of leukocytes might account for the prolonged microbicidal activity, which, however, was restricted to two of the strains used (E coli and S aureus). On the other strains examined (K pneumoniae, P aeruginosa), no effects of PRP were observed despite the presence of leukocytes. Conversely, in our experimental conditions, the last three strains were susceptible to the bacterial activity of a P-PRP preparation containing only platelets. We did not find a similar long-lasting effect; starting from the 4th hour, we observed a progressive increase in bacteria growth reaching the concentrations found in control medium after 18 h. It is conceivable that incubation times that are too long (16e18 h) might result in the loss of the PRP peak of antimicrobial activity, thus allowing an overgrowth of bacteria (22). A recent study (21) described a reduction in S aureus growth within 4 h of treatment, either after treating cultures with a preparation containing only platelets (similar to our P-PRP) or with a formulation containing more platelets and/or leukocytes. Even if

PRP and bacterial growth

1301

Table III. Correlations between bacterial strain growth and specific soluble factor concentrations.

the platelet number was different among the preparations, they displayed a comparable microbicidal activity. Similar results were reported by Moojen et al. (17), who detected a bacterial growth inhibition delayed up to 8 h of treatment with the use of a PRP not activated by calcium chloride, thus suggesting that the components responsible for antibacterial effects are probably activated after interaction with bacteria or the spontaneous degeneration of the platelets because after this time, the PRP effect was similar to the thrombin-activated preparations. The P-PRP activation by calcium chloride, which was performed in the present study, might then accelerate the release of microbicidal molecules from platelets and thus the process of bacterial inhibition, allowing maximum activity of P-PRP during the first hours after application. Furthermore, it was found that the addition of thrombin to the preparation did not activate contained leukocytes (17); therefore, the microbicidal activity that they found after 1 and 4 h may be attributable to the activity of platelets alone, being therefore comparable with our experimental conditions. Because S aureus infections are strictly related to surgical implants and devices, activated PRP might be useful in surgical prophylaxis for preventing infections. Similar data were also obtained by Chen et al. (23), who activated PRP with thrombin and described a

long-lasting PRP-inhibiting activity, which was already evident in the first 4 h of treatment and continued for up to 24 h. In their experimental conditions, the presence of leukocytes was uninfluential as demonstrated by the similar inhibition obtained by blocking or not the leukocytes. Other authors (24) attributed the antimicrobial effects of the different preparations to plasma rather than cellular components on the basis of the evidence that the activation with CaCl2 did not improve PRPinhibiting effects on bacterial growth compared with inactivated conditions or PPP. In the present study, an inhibitory component was also obviously attributable to plasma at least in some experimental conditions, but the presence of an elevated number of activated platelets induced more evident effects. The early inhibition, found in the present experimental conditions, was particularly relevant for K pneumoniae because of its well-known ability to resist antibiotic treatment (12). Indeed, for this bacterium, few not-univocal experimental evidences are available (22,24) on the susceptibility to the PRP microbicidal effect. Concerning P aeruginosa, a lack of inhibition (25) was observed at bacterial concentrations 10-fold lower than the lowest used in the present experimental conditions. Moreover, a P aeruginosa flare-up induction was shown at long incubation times, even in the presence of leukocytes in PRP preparation (22).

1302

E. Mariani et al.

High levels of antibiotic resistance contribute to the pathogenicity of S faecalis, which is also frequently found to persist in root canaletreated teeth (26). Our data showed a P-PRP-inhibiting activity within the first 2 h of treatment. Recently, Drago et al. (25) found a PRP-antimicrobial activity against S faecalis after an incubation time of 24 h, with the use of platelet concentration ranges lower than blood circulating levels and roughly similar to those we obtained in PPP, whose effect, however, was consumed within 2 h. Despite some overlapping results, the few data available are difficult to compare. Several factors, including protocols used for platelet concentrate production (with or without leukocytes), modalities of PRP activation (CaCl2 and/or thrombin, or no activation), the non-univocal terminology used by different authors to identify substantially identical preparations, different bacterial susceptibility to the antimicrobial components within platelets, different time points of the evaluation and finally, different assays for testing PRP bacterial inhibition, might account for these differences. Moreover, as observed for all the bacterial strains, the inhibition of growth was transient, suggesting a potential consumption of bactericidal factors within the PRP mixture. These results may find practical and clinical applications in the way that blood-derived materials are used. It is therefore reasonable to hypothesize that the correct concentration of platelets in PRP, supported by appropriate protocols for PRP activation and estimation of the inhibition of bacterial strain, might lead to the formulation of standardized protocols, thus allowing comparable results between different studies. To get a better understanding of the PRP mechanism of action, this study analyzed the correlation between the concentrations of several soluble factors released by platelets after activation. These factors have been previously described to display a microbicidal activity and bacterial growth inhibition. For the first time, Tang et al. (27) described “kinocidins” and showed the direct role of chemokines as antimicrobial effector molecules. Yount et al. (28) then identified the presence of a highly conserved Y-core motif able to mediate the antimicrobial activity in a broad spectrum of classic antimicrobial peptides, including several kinocidins. By the use of recombinant peptides, they found that RANTES, IL-8 and, partially, GRO-a, were able to inhibit S aureus growth (29), which is supported by the results of the present study. A strong correlation between S aureus inhibition and GRO-a concentration was also found by Yang et al. (30), who surprisingly failed to observe any correlation between bacterial growth inhibition and IL-8, RANTES, MIP-1a and Fractalkine

concentrations, despite the use of higher amounts of recombinant chemokines than in our physiological levels. In addition, we observed two more correlations with sCD40L and EGF levels. To our knowledge, Fractalkine has been described as being secreted by oral fibroblasts, playing an important role in the oral immune response to Candida albicans infection (31), and the present data appear to be the first to suggest the possible role of Fractalkine in reducing other pathogen infections, thus prompting us to suppose a possible role of this molecule as an antimicrobial effector involved in host defense. Concerning E coli, our data are in agreement with those of Yang (30), similarly not highlighting correlations of bacterial inhibition with MIP-1a, RANTES and IL-8 concentrations. Conversely, a correlation was found with Fractalkine, PDGF AB/ BB, and sCD40L. To our knowledge, no evidence of a role of these three factors on E coli inhibition has been described until now. Surprisingly, E coli and K pneumoniae are bacterial strains that show the weakest correlations with few of the analyzed soluble factors with antimicrobial activity (EGF, Fractalkine and ENA-78). This is clearly in contrast with the inhibition of growth obtained after PRP treatment, which leads us to suppose that other molecules/mechanisms of action, not considered in this study, might be responsible for the PRP-antimicrobial effect observed. Until now, no previously published data have shown a PRP-inhibiting activity against P aeruginosa (22,23,25). Conversely, we found strong PRP-inhibitory activity, and some interesting correlations were found between P aeruginosa bacterial growth inhibition and MIP-1a, RANTES and GRO-a concentrations; the majority of these factors clearly display a direct antimicrobial role against the different analyzed pathogens. To our knowledge, these are the only available data presented in the literature concerning the involvement of selected platelet-derived kinocidins in the inhibition of P aeruginosa. Finally, we found correlations between S faecalis and MIP-1a, RANTES, GRO-a and IL-8, thus confirming the concept that these are the most common kinocidins involved in bacterial growth inhibition, as previously highlighted for the other considered bacterial strains. No correlation was observed between each bacterial strain, PF-4 and NAP-2. Noteworthy, the present data originate from the use of PRP compositions, characterized by the presence of soluble factors naturally deriving from blood-derivative components, rather than the use of recombinant proteins, as occurred in several previously published works and therefore more representative of physiological in vivo conditions.

PRP and bacterial growth Nonetheless, further investigations are still required for a more complete comprehension of PRP antibacterial activity. The inclusion of leukocytes in different PRP preparations is currently the most popular topic of discussion (1,8); moreover, many characteristic features can influence antimicrobial chemokine activity (tridimensional structures, charged electrostatic patch on the surface of the protein, pH variations, chemokine concentration, in vitro/in vivo conditions and the sites of wound onset) (20,30,32). All these aspects still must be clarified to optimize the protocols for PRP preparation and application and to make PRP a routinely useful mixture to fight bacterial strain infections. Furthermore, the standardized experimental approach used does not reflect the in vivo settings, in which bacterial and antibacterial concentrations in tissues and body fluids may fluctuate widely, possibly influenced also by local concomitant inflammatory conditions. In conclusion, these results shed new light on the effects of PRP on bacterial growth inhibition against different pathogens, and it should be pointed out that because the inhibitory activity is already evident from the first hour of treatment, PRP supplies an early protection against possible bacterial contaminations during minor or radical surgical interventions. Indeed, inadequate penetration into the infection site is one of the main causes of failure of antibacterial therapy; therefore, the immediate availability of PRP functional molecules shifts the importance from its real bactericidal ability to a more valuable preventative/restrictive one. The prophylactic function displayed by physiological molecules supplied in loco by PRP might be important in the time frame needed to activate the innate immune response against nosocomial infection agents. Acknowledgments This work was supported by grants from Rizzoli Orthopaedic Institute (Ricerca Corrente), University of Bologna (RFO funds to EM), Italian Ministry of Health (Project RF-2009egrant no. 1498841) and PRRU (Emilia-Romagna Region/University of Bologna Project) 2010e2012; “5  1000” funds. The authors thank Mr Keith Smith for English editing, Mrs Patrizia Rappini and Graziella Salmi for manuscript typing, Berardo Di Matteo, MD, for blood sampling and Alice Roffi, BScD, for P-PRP preparation. Disclosure of interests: The authors have no commercial, proprietary, or financial interest in the products or companies described in this article.

1303

References 1. Anitua E, Sanchez M, Orive G. Potential of endogenous regenerative technology for in situ regenerative medicine. Adv Drug Deliv Rev. 2010;62:741e52. 2. Cole BJ, Seroyer ST, Filardo G, Bajaj S, Fortier LA. Plateletrich plasma: where are we now and where are we going? Sports Health. 2010;2:203e10. 3. Filardo G, Presti ML, Kon E, Marcacci M. Nonoperative biological treatment approach for partial Achilles tendon lesion. Orthopedics. 2010;33:120e3. 4. Torricelli P, Fini M, Filardo G, Tschon M, Pischedda M, Pacorini A, et al. Regenerative medicine for the treatment of musculoskeletal overuse injuries in competition horses. Int Orthop. 2011;35:1569e76. 5. Tschon M, Fini M, Giardino R, Filardo G, Dallari D, Torricelli P, et al. Lights and shadows concerning platelet products for musculoskeletal regeneration. Front Biosci. 2011;3:96e107. 6. Sanchez-Ilarduya MB, Trouche E, Tejero R, Orive G, Reviakine I, Anitua E. Time-dependent release of growth factors from implant surfaces treated with plasma rich in growth factors. J Biomed Mater Res A. 2013;101:1478e88. 7. Yeaman MR. Platelets in defense against bacterial pathogens. Cell Mol Life Sci. 2010;67:525e44. 8. Dohan Ehrenfest DM, Sammartino G, Shibli JA, Wang HL, Zou DR, Bernard JP. Guidelines for the publication of articles related to platelet concentrates (platelet-rich plasma-PRP, or platelet-rich fibrin-PRF): the international classification of the POSEIDO. POSEIDO J. 2013;1:17e27. 9. Wiesner J, Vilcinskas A. The ancient arm of the human immune system. Virulence. 2010;1(5):440e64. 10. Zalavras CG, Patzakis MJ, Holtom P. Local antibiotic therapy in the treatment of open fractures and osteomyelitis. Clin Orthop Relat Res. 2004;427:86e93. 11. Thanni LO, Osinupebi OA, ji-Agboola M. Prevalence of bacterial pathogens in infected wounds in a tertiary hospital, 1995e2001: any change in trend? J Natl Med Assoc. 2003;95: 1189e95. 12. Lautenbach D, Patel JB, Bilker WB, Edelstein PH, Fishman NO. Extended-spectrum B-lactamase-producing Escherichia coli and Klebsiella pneumoniae risk factors for infection and impact of resistance on outcomes. J Assoc Physicians India. 2013;12(58):1162e71. 13. Anitua E, Orive G, Aguirre JJ, Andia I. Five-year clinical evaluation of short dental implants placed in posterior areas: a retrospective study. J Periodontol. 2008;79:42e8. 14. Perut F, Filardo G, Mariani E, Cenacchi A, Pratelli L, Devescovi V, et al. Preparation method and growth factor content of platelet concentrate influence the osteogenic differentiation of bone marrow stromal cells. Cytotherapy. 2013; 15:830e9. 15. Zuluaga AF, Galvis W, Jaimes F, Vesga O. Lack of microbiological concordance between bone and non-bone specimens in chronic osteomyelitis: an observational study. BMC Infect Dis. 2002;2:8. 16. Mader JT, Calhoum J. Osteomyelitis. In: Mandell GL, Bennet JE, Dolin R. Principles and Practice of Infectious Diseases. Philadelphia: Churchill Livingston; 2000. p. 1182e96. 17. Moojen DJ, Everts PA, Schure RM, Overdevest EP, van ZA, Knape JT, et al. Antimicrobial activity of platelet-leukocyte gel against Staphylococcus aureus. J Orthop Res. 2008;26: 404e10. 18. Prabhakara R, Harro JM, Leid JG, Harris M, Shirtliff ME. Murine immune response to a chronic Staphylococcus aureus biofilm infection. Infect Immun. 2011;79:1789e96.

1304

E. Mariani et al.

19. Mariani E, Cattini L, Neri S, Malavolta M, Mocchegiani E, Ravaglia G, et al. Simultaneous evaluation of circulating chemokine and cytokine profiles in elderly subjects by multiplex technology: relationship with zinc status. Biogerontology. 2006;7:449e59. 20. Anitua E, Andia I, Ardanza B, Nurden P, Nurden AT. Autologous platelets as a source of proteins for healing and tissue regeneration. Thromb Haemost. 2004;91:4e15. 21. Anitua E, Alonso R, Girbau C, Aguirre JJ, Muruzabal F, Orive G. Antibacterial effect of plasma rich in growth factors [PRGF(R)-Endoret(R)] against Staphylococcus aureus and Staphylococcus epidermidis strains. Clin Exp Dermatol. 2012;37:652e7. 22. Bielecki TM, Gazdzik TS, Arendt J, Szczepanski T, Krol W, Wielkoszynski T. Antibacterial effect of autologous platelet gel enriched with growth factors and other active substances: an in vitro study. J Bone Joint Surg Br. 2007;89:417e20. 23. Chen L, Wang C, Liu H, Liu G, Ran X. Antibacterial effect of autologous platelet-rich gel derived from subjects with diabetic dermal ulcers in vitro. J Diabetes Res. 2013;269527. 24. Burnouf T, Chou ML, Wu YW, Su CY, Lee LW. Antimicrobial activity of platelet (PLT)-poor plasma, PLT-rich plasma, PLT gel, and solvent/detergent-treated PLT lysate biomaterials against wound bacteria. Transfusion. 2013;53:138e46. 25. Drago L, Bortolin M, Vassena C, Taschieri S, Del FM. Antimicrobial activity of pure platelet-rich plasma against

26.

27. 28.

29.

30.

31.

32.

microorganisms isolated from oral cavity. BMC Microbiol. 2013;13:47. Stuart CH, Schwartz SA, Beeson TJ, Owatz CB. Enterococcus faecalis: its role in root canal treatment failure and current concepts in retreatment. J Endod. 2006;32:93e8. Tang YQ, Yeaman MR, Selsted ME. Antimicrobial peptides from human platelets. Infect Immun. 2002;70:6524e33. Yount NY, Yeaman MR. Multidimensional signatures in antimicrobial peptides. Proc Natl Acad Sci U S A. 2004;101: 7363e8. Yount NY, Waring AJ, Gank KD, Welch WH, Kupferwasser D, Yeaman MR. Structural correlates of antimicrobial efficacy in IL-8 and related human kinocidins. Biochim Biophys Acta. 2007;1768:598e608. Yang D, Chen Q, Hoover DM, Staley P, Tucker KD, Lubkowski J, et al. Many chemokines including CCL20/MIP3alpha display antimicrobial activity. J Leukoc Biol. 2003;74: 448e55. Ohta K, Nishi H, Fukui A, Shigeishi H, Takechi M, Kamata N. CX3CL1 expression induced by Candida albicans in oral fibroblasts. FEMS Immunol Med Microbiol. 2010;60: 179e85. Cole AM, Ganz T, Liese AM, Burdick MD, Liu L, Strieter RM. Cutting edge: IFN-inducible ELR- CXC chemokines display defensin-like antimicrobial activity. J Immunol. 2001;167:623e7.