An orally-active adiponectin receptor agonist mitigates cutaneous fibrosis, inflammation and microvascular pathology in a murine model of systemic sclerosis

The hallmarks of systemic sclerosis (SSc) are autoimmunity, microangiopathy and fibrosis. Skin fibrosis is accompanied by attrition of the dermal white adipose tissue layer, and alterations in the levels and function of adiponectin. Since these findings potentially implicate adiponectin in the pathogenesis of SSc, we employed a novel pharmacological approach to augment adiponectin signaling using AdipoRon, an orally active adiponectin receptor agonist. Chronic treatment with AdipoRon significantly ameliorated bleomycin-induced dermal fibrosis in mice. AdipoRon attenuated fibroblast activation, adipocyte-to-myofibroblast transdifferentiation, Th2/Th17-skewed polarization of the immune response, vascular injury and endothelial-to-mesenchymal transition within the lesional skin. In vitro, AdipoRon abrogated profibrotic responses elicited by TGF-β in normal fibroblasts, and reversed the inherently-activated profibrotic phenotype of SSc fibroblasts. In view of these broadly beneficial effects on all three cardinal pathomechanisms underlying the clinical manifestations of SSc, pharmacological augmentation of adiponectin signaling might represent a novel strategy for the treatment of SSc.

Systemic sclerosis (SSc) is a chronic multi-factorial disease characterized by early inflammation, followed by microangiopathy and fibrosis in the skin and internal organs 1 . While fibroblast activation elicited by transforming growth factor (TGF)-β, chemokines and other extracellular cues is recognized as an essential early step in the development of fibrosis, the factors driving persistent vasculopathy, loss of intradermal adipose tissue and chronic unresolving fibrosis in SSc are not well understood 2,3 . White adipose depots, widely distributed throughout the body, are comprised of hormonally active adipocytes producing a range of adipokines with pleiotropic functions. Secreted adipokines exert potent paracrine and humoral effects on a variety of cells, and regulate inflammatory cell accumulation and endothelial activation in diverse pathological processes [4][5][6][7] . Recent studies have generated substantial evidence implicating deregulated adipocyte function and adipokine secretion as pathogenic factors in fibrosis and SSc.
A pathologic hallmark of SSc is accumulation in the lesional tissue of activated myofibroblasts that are responsible for excessive extracellular matrix (ECM) deposition and tissue remodeling. Profibrotic myofibroblasts might originate from resident fibroblasts, or via in situ transdifferentiation of other cell lineages resident within the fibrotic microenvironment 8 . Recent studies identified adipocytic cells within intradermal adipose tissue as another potential source of activated myofibroblasts during tissue injury 6,9 . Since differentiated adipocytes produce adipokines that help maintain tissue homeostasis through autocrine, paracrine and endocrine mechanisms 10 , altered adipokine balance resulting from adipocyte loss or dysfunction might contribute to

AdipoRon prevents dermal fibrosis induced by bleomycin.
To investigate the effect of AdipoRon in a murine model of SSc, mice were randomly assigned to treatment with vehicle, bleomycin, or those in combination with AdipoRon. Prolonged administration of AdipoRon (50 mg/kg) for up to 28 days was well tolerated and no signs of toxicity were observed. AdipoRon treatment significantly attenuated the increase in dermal thickness induced by bleomycin, while augmenting the dermal white adipose tissue layer (Fig. 1a). Collagen content and the number of myofibroblasts within the lesional skin were significantly reduced in mice treated with AdipoRon (Fig. 1b,c). Additionally, transdifferentiation of intradermal adipocytes to myofibroblasts, detected by double immunofluorescence for α-SMA and perilipin, was reduced ( Supplementary Fig. 1). Levels of Col1a1 and Col1a2 mRNA in the skin were reduced in AdipoRon-treated mice, while Mmp13 mRNA levels were significantly elevated (Fig. 1d). Since TGF-β1 and connective tissue growth factor (CTGF) are recognized as critical mediators of fibrosis 33 , we examined the expression of these growth factors in the lesional skin. The results showed a significant decrease in Tgfb1 and Ctgf gene expression in mice treated with AdipoRon ( Fig. 1d), which was also confirmed by immunohistochemistry ( Fig. 1e and Supplementary Table 1). Taken together, these findings indicate that chronic AdipoRon treatment exerts a potent anti-fibrotic effect in mice by attenuating myofibroblast transition, adipocyte-to-myofibroblast transdifferentiation, collagen accumulation and expression of key pro-fibrotic growth factors within the lesional skin. Importantly, bleomycin increased the expression of adiponectin receptors, AdipoR1 and AdipoR2, in dermal fibroblasts, endothelial cells, inflammatory cells, and adipocytes irrespective of AdipoRon treatment (Supplementary Fig. 2 and Supplementary Table 2), and AdipoRon treatment increased the phosphorylation of adenosine monophosphate-activated protein kinase (AMPK), a major intracellular mediator of adiponectin signaling, in skeletal muscles from both control (PBS-treated) and bleomycin-treated mice ( Supplementary Fig. 3), suggesting that AdipoRon at least in part directly exerts its effect on bleomycin-induced skin fibrosis.
AdipoRon attenuates fibroblast activation. The molecular mechanism underlying the anti-fibrotic properties of AdipoRon were further investigated in vitro with normal human dermal fibroblasts. Initial studies established the absence of cytotoxicity of AdipoRon at concentrations up to 50 μM (Fig. 2a); therefore, all subsequent experiments were performed with AdipoRon at concentrations ≤50 μM. Incubation of confluent fibroblasts with AdipoRon for 24 h resulted in dose-dependent down-regulation of type I collagen, fibronectin and α-SMA, while expression of MMP-1 was increased, as expected 34 (Fig. 2b-d, full-length blots of Fig. 2d in Supplementary  Fig. 4). Significantly, AdipoRon abrogated TGF-β1-induced stimulation of type I collagen and α-SMA gene expression in a dose-dependent manner (Fig. 2e). Furthermore, in vitro wound-healing assay demonstrated that treatment with AdipoRon abrogated the increase in fibroblast migration elicited by TGF-β1 (Fig. 2f).
To further explore the anti-fibrotic mechanism of AdipoRon, we next focused on downstream signaling pathways implicated in adiponectin receptor signaling. AMPK and peroxisome proliferator-activated receptor-α (PPARα) are cardinal mediators for adiponectin signal transduction in fibroblasts 5,35 . AdipoRon treatment of normal fibroblasts resulted in a rapid increase in AMPK phosphorylation, and was relatively short-lived (Fig. 3a). Intriguingly, the AdipoRon effect on AMPK was more rapid than that elicited by adiponectin, possibly due to slight differences in strength/affinity ligand binding to receptors. Immunocytochemistry further demonstrated that in normal fibroblasts, AdipoRon treatment caused a reduction in cellular type I collagen, while increasing p-AMPK; fibroblasts showing the highest levels of p-AMPK exhibited signficant reductions in collagen levels (Fig. 3b). The anti-fibrotic effects of AdipoRon were abrogated by pretreatment of the cultures with Compound C, a selective AMPK inhibitor, while GW6417, a selective inhbitor of PPARα, showed no effect (Fig. 3c). These observations support the importance of AMPK activation as a possible mechanism to account for the anti-fibrotic effect of AdipoRon.
AdipoRon inhibits Th2/Th17-skewed immune polarization induced by bleomycin. Since fibroblast activation by inflammatory cells and their secreted product plays an important pathogenic role in bleomycin-induced  (f) In vitro wound healing assays. Cultures were incubated with TGF-β1 and 20 μM of AdipoRon. Gap widths were determined at 24 h using ImageJ. For A, C, and E, bars represent mean ± SEM of triplicate determinations. For B and D, images representative from three independent experiments were shown. For F, bars represent mean ± SEM of six determinations from three randomly selected fields. *p < 0.05 versus cells without any treatment. **p < 0.01 versus cells without any treatment for C and F, and cells treated with TGF-β1 alone for E. Ctl, control; AU, arbitrary unit. skin fibrosis 8 , we next focused on the effect of AdipoRon on inflammatory responses. To this end, we evaluated in vivo changes in cutaneous cytokines and chemokines at an early time point (7 days) when skin fibrosis is still relatively modest. A significant reduction in Il4 and Il17a mRNA levels was seen in AdipoRon-treated mice, indicating suppression of Th2 and Th17-dominant immune responses, while levels of the Ifng and Il10 mRNA were modestly elevated (Fig. 4a). Flow cytometry of cells isolated from skin-draining lymph nodes showed a decrease in the total numbers of lymphocytes and decreased proportions of IL-4-and IL-17A-producing CD4 + T cells in mice treated with AdipoRon, while the proportions of IFN-γ-producing and Foxp3 + cells were unaltered (Fig. 4b). In view of the critical role of M2 macrophages and mast cells in SSc fibrosis [36][37][38] , we assessed the impact of AdipoRon on these cell populations. The results showed that expression of the M2 macrophage markers, such as Ang1, Fizz1 and Ym1, and the numbers of toluidine blue-positive cells in the lesional skin were unchanged upon treatment with AdipoRon ( Fig. 4c,d).

AdipoRon treatment mitigates ICAM-1 expression in microvascular endothelial cells.
We had previously demonstrated that Th2/Th17 cell infiltration in the dermis is governed by the balance between intercellular adhesion molecule 1 (ICAM-1) and glycosylation-dependent cell adhesion molecule 1 (GlyCAM-1) acting as positive regulators, and P-selectin and E-selectin acting as negative regulators 39 . We therefore next investigated the impact of AdipoRon on the expression of these cell adhesion molecules. Initial experiments confirmed expression of AdipoR1 and AdipoR2 in endothelial cells ( Supplementary Fig. 5). The increase in Icam1 mRNA levels in bleomycin-treated mice was mitigated by AdipoRon, while those of the Glycam1, Selp, and Sele genes were not altered (Fig. 5a). Immunohistochemistry confirmed that in bleomycin-treated mice AdipoRon markedly attenuated ICAM-1 levels in small vessels of the lesional skin (Fig. 5b). Consistent with these in vivo observations, we found that treatment of human dermal microvascular endothelial cells (HDMECs) with AdipoRon ex vivo abrogated the stimulation of ICAM1 mRNA induced by TNF-α (Fig. 5c). These results suggest that the inhibitory effects of AdipoRon on Th2/Th17 cell accumulation within the lesional skin in bleomycin-treated mice could be mediated, at least in part, by preventing ICAM-1 expression of endothelial cells.
AdipoRon treatment attenuates endothelial-to-mesenchymal transition in the lesional skin of bleomycin-treated mice. Vascular injury in the bleomycin model of SSc is thought to contribute to the accumulation of interstitial myofibroblasts through the process of EndoMT 40 . Therefore, we evaluated the impact of AdipoRon treatment on this cellular transition by double immunofluorescence for VE-cadherin and fibroblast-specific protein 1 (FSP1), a validated indicator of EndoMT in vivo [41][42][43][44] . As previously reported, increased number of spindle-shaped double-positive cells was readily observed in the lesional skin from bleomycin-treated mice 15, 40,45,46 . AdipoRon treatment markedly attenuated the accumulation of double-positive cells (Fig. 6a). Furthermore, mRNA expression of Snail1, encoding a master regulator of EndoMT, was significantly attenuated (Fig. 6b). Parallel in vitro experiments with HDMECs showed that AdipoRon effectively suppressed enhanced SNAIL1 mRNA expression induced by TNF-α (Fig. 6c). Since EndoMT is closely associated with vascular destabilization 46 , we next evaluated stability of dermal microvessels by assessing phenotypical alteration of pericytes using immunostaining for α-SMA, a marker of pericytes with angiostatic phenotype, and platelet-derived growth factor receptor β (PDGFRβ), a marker of pericytes with pro-angiogenic phenotype. As previously reported, α-SMA expression was decreased 15,45 , while PDGFRβ expression was increased in dermal vessels of the lesional skin, indicating that bleomycin promotes a pro-angiogenic state (left panels of Fig. 6d and Supplementary Fig. 6). Of note, AdipoRon treatment reversed the expression profile of α-SMA and PDGFRβ in dermal small vessels of bleomycin-treated mice (right panels of Fig. 6d and supplementary Fig. 6). Relevant to these findings, bleomycin induced the formation of immature vascular networks consisting of tiny vessels, which was also attenuated by AdipoRon treatment (Fig. 6a). Viewed together, these results indicate that AdipoRon has beneficial effects on bleomycin-induced dermal microvascular damage by preventing EndoMT and promoting vessel stabilization.
AdipoRon mitigates the intrinsic fibrotic phenotype of SSc fibroblasts. All four known adiponectin receptors (AdipoR1, AdipoR2, AdipoR3, and T-cadherin) were found to be expressed in SSc fibroblasts at levels comparable to those in normal fibroblasts (Fig. 7a). Treatment with AdipoRon ex vivo resulted in variable suppression of fibrotic gene expression in SSc and normal fibroblasts, but the effects were considerably stronger in SSc fibroblasts (Fig. 7b), especially for α-SMA (the fold induction, 0.61 ± 0.13 versus 1.04 ± 0.09 [p < 0.05] for α-SMA and 0.32 ± 0.11 versus 0.55 ± 0.17 [p = 0.30] for type I collagen). In foreskin fibroblasts, treatment with TGF-β1 or ectopic expression of Egr-1 had no effect on expression of adiponectin receptors (Supplementary Table 3). These results indicate that AdipoRon exerts its anti-fibrotic effects on dermal fibroblasts irrespective of their activation status.

Discussion
In view of the emerging importance of adiponectin in modulating a broad range of cellular activities implicated in SSc pathogenesis, we sought to investigate the impact of AdipoRon, the first orally-active AdipoR agonist, in a preclinical model of SSc. We found that AdipoRon treatment prevented fibrosis by attenuating fibroblast activation, adipocyte and endothelial cell transition to myofibroblasts, Th2/Th17-skewed immune polarization, and vascular activation. Importantly, the expression of AdipoR1 and AdipoR2 was increased by bleomycin treatment in multiple cell types in the skin, while adiponectin expression is generally suppressed in bleomycin-treated skin 6 . Therefore, activation of adiponectin receptor signaling might represent a plausible strategy to attenuate SSc-related disease processes, a notion supported by a series of experiments with AdipoRon.
Increased number of myofibroblasts within the lesional tissue is a histopathological hallmark of SSc shared with other fibrotic disorders 8 . Myofibroblasts originate from resident fibroblasts, as well as via transition of epithelial cells, endothelial cells, pericytes and adipocytic progenitor cells 6,40,45,46 . In bleomycin-treated mice, transition of endothelial cells and adipocytes to myofibroblasts were both attenuated by AdipoRon. Since myofibroblast conversion of adipocytes and/or endothelial cells has been implicated in tissue fibrosis of the skin, lung, heart, kidney, and pulmonary artery 40,45-51 , inhibition of myofibroblast accumulation is likely to represent a key mechanism underlying the potent anti-fibrotic effects of adiponectin. were injected with bleomycin (BLM) or PBS for one week alone or together with AdipoRon. Tnfa, Il1b, Ccl2, Ifng, Il4, Il6, Il10, Il13, and Il17a mRNA levels in the lesional skin were measured by qRT-PCR (n = 4-6 mice/ group, a). The total number of lymphocytes and the proportions of IFN-γ-, IL-4-, and IL-17A-producing cells and Foxp3 + cells in CD4 + T cells were determined by flow cytometry with cells from skin-draining lymph nodes (n = 5-6 mice/group, b). mRNA expression of the Arg1, Fizz1, and Ym1 genes was evaluated in the lesional skin by qRT-PCR (n = 4-6 mice/group, c). The number of toluidine blue-positive cells was determined by immunostaining (n = 4-6 mice/group, d). Each graph represents mean ± SEM of the indicated parameter. Representative images are shown. A scale bar is 20 μm. *p < 0.01 compared with mice treated with PBS and vehicle. **p < 0.05 compared with mice treated with PBS and vehicle. # p < 0.01 compared with mice treated with BLM and vehicle. AU, arbitrary unit; HPF, high power field; ARon, AdipoRon; VE, vehicle.
Adiponectin modulation of the Th cytokine balance has been implicated in autoimmune and inflammatory conditions. For instance, in collagen-induced murine arthritis adiponectin augments articular bone destruction by enhancing Th17 response 52 , while in asthma it exacerbates respiratory inflammation via its anti-Th1 effect 53 .
Taken together with the current finding that AdipoRon suppresses Th2/Th17 responses in bleomycin-treated  mice, adiponectin receptor signaling seems to modulate the Th balance in a context-dependent manner. Importantly, the regulatory effect of adiponectin receptor signaling on Th2/Th17 responses is consistent with a datum regarding Th cell subsets and adiponectin levels in SSc. In early dcSSc pro-fibrotic Th2/Th17 responses are predominant 54,55 , while the Th balance shifts towards an anti-fibrotic Th1 response along with the resolution of skin sclerosis 56,57 . Serum adiponectin levels are decreased in early, but not in late-stage dcSSc and increased in patients treated with anti-fibrotic therapy 4,11,12 . Therefore, decreased adiponectin might contribute to SSc fibrosis via pro-fibrotic inflammation as well as fibroblast activation.
Vasculopathy in SSc encompasses vascular activation and structural abnormalities, both of which contribute to tissue fibrosis 58 . AdipoRon modulates both of these two aspects of SSc vasculopathy. A hallmark of SSc vascular activation is up-regulation of ICAM-1 in endothelial cells 40,59 , and ICAM-1 deficiency significantly attenuates bleomycin-induced dermal fibrosis by decreasing inflammatory Th2/Th17 cell accumulation 39 . Since AdipoRon reduced bleomycin-induced dermal fibrosis in parallel with decreased Th2/Th17 cell infiltration and endothelial ICAM-1 down-regulation, adiponectin receptor signaling likely inhibits a pro-fibrotic inflammatory response through vascular inactivation. With respect to structural abnormalities, capillary dilation and arteriolar stenosis are evident in SSc due to vascular destabilization and intimal fibro-proliferation, respectively 58 . Since AdipoRon attenuated bleomycin-induced vascular leakage and loss of vascular integrity, decreased adiponectin expression may underlie vascular destabilization in SSc. Adiponectin selectively accumulates in injured vessel walls by binding specifically to collagen types I, III, V and VIII of the vascular intima, and may contribute to vascular repair 60 . Considering the regulatory role of adiponectin in tissue fibrosis, decreased adiponectin expression may promote the fibro-proliferative changes of arterioles in SSc. Thus, adiponectin receptor agonists might exert a protective effect on SSc vasculopathy, possibly leading to the attenuation of tissue fibrosis.
AdipoRon is thought to bind exclusively to AdopoR1 and AdopoR2 30 , while adiponectin also activates AdipoR3 61 . Both AdipoR1 and AdipoR2 contain seven transmembrane domains but, in contrast to other G-protein-coupled receptors, they present an extracellular C and a cytosolic N terminus 62 . Recent resolution of the crystal structure of AdipoR1 and AdipoR2 is expected to lead to increasing insights into the regulation of adiponectin signaling 63 . Pharmacological approaches to augment adiponectin signaling have focused on boosting adiponectin levels or augmenting activity with receptor agonists. Adiponectin levels can be increased by treatment with thiazolidinediones, as well as dietary compounds such as curcumin, gingerol, capsaicin, and omega-3 fatty acids 64,65 . Naturally occurring adiponectin receptor agonists include osmontin, a plant-derived protein resembling the β barrel of adiponectin domain I 66 . Screening of small molecular plant libraries have also identified matairesinol, arctiin, (-)-arctigenin and gramine as AdipoR1 agonists, while parthenolide, taxifoliol, deoxyschizandrin, and syringin are AdipoR2 agonists 67 . Undesirable potential effects of adiponectin such as reduced bone density, left ventricular hypertrophy, infertility and growth of established tumors due to enhanced angiogenesis 68 , are concerns for all AdipoR agonists.
The present studies are significant since they represent the first to demonstration for the pleiotropic anti-fibrotic effects of AdipoRon in vitro and in vivo. In view of its potent beneficial impact on all three cardinal pathomechanisms underlying the clinical manifestations of SSc, adiponectin receptor agonists represent promising novel therapeutic approaches for SSc.

Materials and Methods
Ethics Statement. The  Animal experiments. Bleomycin (Nippon Kayaku, Tokyo, Japan) was dissolved in PBS at the concentration of 1 mg/ml and sterilized by filtration, and 200 μg was injected subcutaneously into the shaved backs of C57BL/6 wild-type female mice (Japan SLC Inc., Tokyo, Japan). AdipoRon (Arking Pharma Scientific Inc., ON, Canada) was suspended in aqueous solution of 0.5% carboxy methyl cellulose. AdipoRon (50 mg/kg) or vehicle was administered orally together with s.c. bleomycin. Mice were harvested at indicated time points. All animal protocols were approved by the Animal Care and Use Committee of University of Tokyo (approval ID: P14-064), and all methods were performed in accordance with the relevant guidelines and regulations.
Histological assessment and immunostaining. All skin sections were taken from the paramidline, lower back region. Sections were stained with hematoxylin and eosin (H&E) and toluidine blue. We examined dermal thickness, which was defined as the thickness of skin from the top of the granular layer to the junction between the dermis and subcutaneous fat. Immunohistochemistry and immunofluorescence were performed with antibodies against α-SMA (Sigma-Aldrich, St. Louis, MO, USA), ICAM-1 (BD Pharmingen, San Diego, CA, USA), VE-cadherin (Santa Cruz Biotechnology, Santa Cruz, CA, USA), FSP-1 (Abcam, Cambridge, MA, USA), perilipin (Santa Cruz Biotechnology), AdipoR1 (Abcam), AdipoR2 (Santa Cruz Biotechnology) and PDGFR-β (R&D Systems, Minneapolis, MN, USA) as described previously 45 . All sections were examined independently by two investigators in a blinded manner. To quantify signal intensity (ICAM-1 and α-SMA), color images were converted to grayscale, and brightness of vessels was measured in 5 different randomly selected vessels per specimen using ImageJ.

Determination of hydroxyproline content in skin. Following instructions of the QuickZyme Total
Collagen Assay kit (QuickZyme Biosciences, Leiden, Netherlands), 6-mm punch biopsy skin samples were hydrolyzed with 6 N HCl and collagen content was quantified.
Quantitative real-time reverse transcription PCR. Gene expression levels were determined by quantitative real-time reverse transcription PCR (qRT-PCR), as described previously 69 . mRNA levels of target genes were normalized to those of the GAPDH gene or the Gapdh gene. The sequences of primers were summarized in Supplementary Table 4. Cell cultures. Primary fibroblasts (passage 3-6) established with explantation from neonatal foreskin fibroblasts or skin biopsies from forearm of healthy adults and SSc patients were used. Cells were cultured in DMEM supplemented with 1% glutamine, 50 µg/ml penicillin/streptomycin, 10% fetal bovine serum, 1% vitamin solutions and 2 mM L-glutamine in humidified atmosphere of 5% CO 2 at 37 °C until confluence. All cell culture reagents were from Lonza (Basel, Switzerland). For experiments, fibroblasts were placed in serum-free media containing 0.1% bovine serum albumin (BSA) for 24 h prior to addition of AdipoRon (Sigma-Aldrich) dissolved in DMSO. Cultures were preincubated with AdipoRon and or compound C (Sigma-Aldrich) for 60 min prior to TGF-β1 (PeproTech, Rocky Hill, NJ). Toxicity was determined using lactate dehydrogenase assay according to manufacturer instructions (Biovisison, Milpitas, CA, USA).
Human dermal microvascular endothelial cells. HDMECs were purchased from Lonza Ltd. (Basel, Switzerland). Confluent cultures of HDMECs were treated with AdipoRon at the indicated concentrations for 60 min prior to 0.5 ng/ml of TNF-α. Total RNA was isolated 24 h later from cell lysates as described above.
Immunoblotting. Fibroblasts were harvested, whole-cell lysates were prepared with RIPA buffer and proteinase/phosphatase inhibitor cocktail and equal amounts of proteins (20-50 µg/lane) were subjected to immunoblotting as described previously 5 . Antibodies specific for human type I collagen (Southern Biotechnology, Birmingham, AL, USA), α-SMA (Sigma -Aldrich), tubulin (Sigma-Aldrich), GAPDH (Zymed, San Francisco, CA, USA), fibronectin (Santa Cruz Biotechnology), p-AMPK, AMPK (Cell Signaling, Danvers, MA. USA), AdipoR1 and AdipoR2 were used as primary antibodies with overnight incubation. After washing membranes were incubated with appropriate secondary antibodies and subjected to enhanced chemiluminescence detection using ECL Reagent (Pierce, Rockford, IL, USA). Scratch assays. Fibroblast function was observed using scratch assay. Confluent foreskin fibroblast monolayers were mechanically wounded using p1000 pipette tips. Following 24-hour incubation of the cultures wound gap widths were determined at six randomly selected sites/high power field.

Statistical analysis.
Statistical analysis was done with one-way ANOVA followed by Turkey post hoc test for multiple comparisons and with Mann-Whitney U-test to compare the distributions of two unmatched groups. Statistical significance was defined as a P value of < 0.05.