Amnion Epithelial Cells Promote Lung Repair via Lipoxin A4

Abstract Human amnion epithelial cells (hAECs) have been shown to possess potent immunomodulatory properties across a number of disease models. Recently, we reported that hAECs influence macrophage polarization and activity, and that this step was dependent on regulatory T cells. In this study, we aimed to assess the effects of hAEC‐derived proresolution lipoxin‐A4 (LXA4) on T‐cell, macrophage, and neutrophil phenotype and function during the acute phase of bleomycin‐induced lung injury. Using C57Bl6 mice, we administered 4 million hAECs intraperitoneally 24 hours after bleomycin challenge. Outcomes were measured at days 3, 5, and 7. hAEC administration resulted in significant changes to T‐cell, macrophage, dendritic cell, and monocyte/macrophage infiltration and phenotypes. Endogenous levels of lipoxygenases, LXA4, and the lipoxin receptor FPR2 were elevated in hAEC‐treated animals. Furthermore, we showed that the effects of hAECs on macrophage phagocytic activity and T‐cell suppression are LXA4 dependent, whereas the inhibition of neutrophil‐derived myleoperoxidase by hAECs is independent of LXA4. This study provides the first evidence that lipid‐based mediators contribute to the immunomodulatory effects of hAECs and further supports the growing body of evidence that LXA4 is proresolutionary in lung injury. This discovery of LXA4‐dependent communication between hAECs, macrophages, T cells, and neutrophils is important to the understanding of hAEC biodynamics and would be expected to inform future clinical applications. Stem Cells Translational Medicine 2017;6:1085–1095


INTRODUCTION
Idiopathic pulmonary fibrosis (IPF) is a debilitating chronic inflammatory disease that affects more than 300 million people worldwide. The disease is characterized by progressive chronic inflammation; patients usually have limited numbers of CD4 + / CD28 + /CD25 + /Foxp3 + regulatory T-regulatory cells and higher numbers of Th1 effector CD4 + T cells in the lung [1]. Patients with severe forms of IPF also have significantly higher numbers of neutrophils, macrophages, and dendritic cells present in their bronchoalveolar lavage fluid (BALF) [2]. Currently, there is no effective treatment; lung transplantation is the only therapeutic option for end-stage disease. However, cell-based therapies have been suggested as a promising treatment approach [3][4][5][6][7][8][9]. Studies with early administration of mesenchymal stromal cells (MSCs) after bleomycin challenge reported significant anti-inflammatory effects [10].
Preclinical studies have shown that stem cells reduce T-cell proliferation, myeloperoxidase production, and profibrotic chemokine CCL-2 and collagen deposition to repair lung injury [8,11,12]. Specifically, MSCs home to sites of injury and regulate inflammatory processes by reducing macrophage and neutrophil infiltration. These promising experimental studies led to early-phase clinical trials of MSCs in patients with IPF, and these confirmed the safety of MSCs [13][14][15]. A number of efficacy trials of MSCs as a treatment for IPF are now underway [14]. However, if successful, the future application of MSCs is likely to be limited by the low cell yields from MSC harvesting. In that regard, human amnion epithelial cells (hAECs), stem-like cells with potent immunomodulatory properties, are a promising alternative cell therapy. These cells appear equally effective in experimental models of lung injury [3,[5][6][7] and are more readily sourced in much greater cell numbers without the need for in vitro manipulation or expansion [4,16,17]. Like MSCs [18,19], it would appear that the principle mechanisms of action of hAECs are paracrine, modulating the host immune cell responses and supporting the recruitment and expansion of host niche lung progenitor cells [20][21][22].
Although it is not essential to clinical application of hAECs, a better understanding of the paracrine signaling involved in hAEC-modulated injury repair would be expected to improve the design of future clinical trials and, potentially, lead to novel non-cell-based therapeutics. In that regard, we wondered whether hAECs were exerting their reparative effects via lipoxins. Lipoxins are a subset of lipid mediators, inclusive of resolvins, protectins, and maresins, collectively called specialized proresolving mediators. Lipoxins are eicosanoids derived from arachidonic acid via enzymatic steps by lipoxygenases-5, -12, and -15. They are regulators of inflammation and act by limiting polymorphonuclear neutrophil (PMN) infiltration and by enhancing nonphlogistic phagocytosis of apoptotic PMNs. Preclinical studies have shown the potent anti-inflammatory properties of LXA4 in diverse disease models, including microbial infection, Alzheimer disease, liver transplantation, and myocardial ischemia reperfusion injury [23][24][25][26]. In the lung, LXA4 and aspirin-triggered lipoxin analogs have been shown to improve alveolarization in a neonatal model of hyperoxiainduced injury and to prevent inflammation and fibrosis in bleomycin-induced lung injury [27][28][29]. It has been shown that hAECs administered 24 hours after bleomycin challenge was beneficial [4,8,9]; therefore, we sought to assess the role of lipoxins in hAEC-mediated lung repair in the same mouse model of lung injury.

Experimental Animals
Animal experiments were carried out with approval by Monash University Ethics Committee and conducted in accordance to the Australian Code of Practice for the Care and Use of Animals for Scientific Purposes. A total of 38 female mice, aged 8-12 weeks and weighing 18-21 g were administered 15 IU of bleomycin intranasally (Willow Pharmaceuticals, New South Wales, Australia http://willowpharma.com.au). Each mouse was randomly allocated to receive either 4 million hAECs in 0.2 ml of saline or saline alone via intraperitoneal injection 24 hours after receiving bleomycin. Mice were culled 1, 3, 5, or 7 days after bleomycin challenge. Bronchoalveolar lavage fluid and lung tissues were collected and processed as previously described [4].

Human Amnion Epithelial Cell Isolation
Amnion cells were collected from placentae of consenting women undergoing an elective cesarean section delivery at term. Isolation of hAECs was performed according to previously described studies [16] in accordance with guidelines and approval from Monash Health Human Research Ethics Committee. Mean gestational age at collection was 38 weeks and 4 days. A total of 10 amnions were used in these in vitro and 1 of the 10 was randomly selected and used for all in vivo experiments.

Tissue Collection
Animals were culled at days 1, 3, 5, and 7 by CO 2 asphyxiation, and 2 ml of bronchoalveolar lavage fluid was collected with saline. The right lung was ligated at the right mainstem bronchus, then the trachea was exposed and the left lung was instilled with 4% paraformaldehyde. The right lung was excised for RNA analysis and the left lung for histological analysis. A separate cohort of animals was used for flow cytometric analysis, for which the entire lung (left and right) was processed (bleomycin and saline, n = 5, bleomycin and hAECs, n = 6).

Histological and Immunohistochemical Analysis Immunofluorescent Staining for Lipoxin Receptor and Macrophages
To measure the effect of hAECs on endogenous lipoxin A 4 receptor expression and macrophage number, we performed immunohistochemistry for the lipoxin receptor N-formyl peptide receptor 2 (FPR2) and for the macrophage marker F4/80 on lung slices. Briefly, paraffin-embedded slices (0.5-mm thick) were dewaxed and rehydrated in water. Antigen retrieval was performed with 10 mM citrate buffer, pH 6.0, in a microwave oven for 20 minutes. Blocking was performed with a universal protein blocking solution before immunostaining with anti-FPR2 antibody at 1:100 (NSL1878; Novus Biologicals, Littleton, CO, https://www.novusbio.com) and anti-F4/80 antibody at 1:200 (MCA497; Bio-Rad Laboratories, Oxford, U.K., https://www.bio-rad-antibodies.com) with an overnight incubation at 4°C. Secondary antibody incubation was performed at room temperature for 1 hour, followed by nuclear stain with 49,6-diamidino-2-phenylindole (DAPI) for 10 minutes at room temperature. For each section, five fields of view were taken using the Nikon C1 confocal microscope running the NIS Elements Software (Nikon, Tokyo, Japan, http://www.nikon.com), where dual positive staining was manually quantified and analyzed with FIJI ImageJ analysis software (version 1.480; http://imagej.net/).

Neutralization of Lipoxygenases With Nordihydroguaiaretic Acid in hAECs
Human amnion epithelial cells were cultured in a T75 flask at a density of 5 3 10 6 cells. Neutralization of lipoxygenases was performed by adding 2.5 mM or 10 mM of nordihydroguaiaretic acid (NDGA) and incubated at 37°C for 24 hours. After the 24-hour incubation, hAECs were collected for quantitative polymerase chain reaction analysis of lipoxygenase-5, -12, and -15 expression and supernatant was collected for enzyme-linked immunosorbent assay (ELISA) of LXA4 before further coculture studies. Conditioned medium was obtained according to previous protocols after neutralization of NDGA [18].

Macrophage Phagocytic Assay
Macrophage phagocytosis was determined as previously described [18]. Briefly, macrophages were plated in 6-well flat-bottom culture plates at a density of 5 3 10 5 cells per well for 48 stimulated lipopolysaccharide (LPS; 10 ng/ml) with or without NDGA preprimed hAECs and with or without primary hAECs (1:1 ratio). Staphylococcus aureus particles labeled with pHrodo (Thermo Fisher) were added to each well (10 mg/ml) and incubated for 30 minutes. Incubation on ice inhibits membrane movement and was used as a negative control. Only cells that phagocytosed pHrodo-labeled S. aureus were fluorescent and stained positive on fluorescenceactivated cell sorting (FACS).

Neutrophil Myeloperoxidase Activity Assay
Neutrophils were isolated from mouse femoral bone-marrow exudates using a CD11b magnetic-bead isolation kit (Miltenyi Biotec). Enriched neutrophils (0.5 3 10 6 ) were cultured in complete RPMI with or without NDGA-preprimed hAECs and with or without primary hAECs (0.5 3 10 6 ), and activated with granulocyte-colony stimulating factor (100 ng/ml; Amgen, Sydney, Australia, http:// www.amgen.com) for 12 days. The medium was changed every other day. After 12 days, cells were lysed with cetyltrimethylammonium bromide at 500 ml per 1 3 10 6 cells. Then, 50 ml was added to 950 ml of 50 mM potassium phosphate buffer (pH 6.0) that contained 0.167 mg/ml o-dianisidine (D9143; Sigma-Aldrich) and 0.0005% H 2 O 2 (Sigma-Aldrich), and the absorbance at 460 nm (A460) was measured. One unit of myeloperoxidase (MPO) activity was defined as a change in A460 of 1.0 after 5 minutes; results were expressed as mU of MPO activity per milligram of lysate.

LXA4 ELISA or Th1/Th2/Th17 Cytokine Bead Array
The concentration of lipoxin A 4 in BALF and Th1/Th2/Th17 cytokines in culture supernatants were measured by ELISA and bead array according to manufacturer's guidelines. Data were collected using the FACS LSRII analyzer and analyzed with BD FRAP Array software (BD Biosciences).

Statistical Analysis
Data were expressed for each experimental group as mean 6 SEM. Differences between two experimental groups were determined using an unpaired, one-tailed t test. Differences across three or more experimental groups were determined using one-way analysis of variance with the Bonferroni post hoc test. Confidence intervals of 95% were deemed significant. All analyses were performed using GraphPad Prism (GraphPad Software, San Diego, CA, http://www.graphpad.com). Next, we assessed changes to the immune-cell population in the lung based on surface markers listed in Table 1 and gated according to the gating strategies seen in Figure 1D.

hAECs' Effect on Lipoxin A 4 and its Receptor, FPR2
Given earlier reports that lipoxin A 4 is highly expressed in placental tissues [20,22] and plays a valuable role in resolving inflammation and fibrosis, we evaluated the role of lipoxin A 4 in hAEC-mediated lung repair. Protein analysis of bronchoalveolar lavage fluids revealed that, compared with the control group of animals treated with bleomycin plus saline, hAEC administration was associated with significantly higher lipoxin A 4 levels at day 7 (0. 13 Fig.  2A). Compared with the control group, hAEC treatment was associated with significantly elevated expression of F4/80 and FPR2 in the lungs of day 7 animals (Fig. 2C). Representative immunofluorescent staining showing colocalization positive for the macrophage marker F4/80, FPR2, and nuclear stain DAPI of hAEC-treated lung is shown in Figure 2D.

Effect of hAECs on Macrophage and Neutrophil Function
Pulmonary inflammation induced by bleomycin is managed through a balance of neutrophil infiltration, apoptosis, and subsequent macrophage efferocytosis of apoptotic neutrophils. Lipid mediators, in particular, LXA4, are potent regulators of this process and, accordingly, we evaluated the direct effects of hAECderived LXA4 on macrophage and neutrophil function in vitro. We first showed that we could induce and inhibit hAEC production of LXA4 and its precursor enzymes ALOX-5, -12, -15. Proinflammatory cytokines interferon (IFN)-g and tumor necrosis factor (TNF)-a significantly increased LXA4 production in hAECs (0.5309 6 0.0359 ng/ml vs. 1.0450 6 0.1393 ng/ml, p = .012; Fig. 3A).
Next we showed that in the presence of nordihydroguaruretic acid (NDGA), gene expression of lipoxygenase-5 and -12 was significantly reduced as well (ALOX-5: 0.  Fig. 3B),suggestingthat biosynthesis of LXA4 may occur in hAECs in an inflammatory environment.
Next, we assessed the effects of hAECs on myeloperoxidase (MPO) production by primary neutrophils. Whereas coculture of CD11b + neutrophils with hAECs significantly reduced their production of MPO, the addition of 2.5 mM (data not shown) and 10 mM NDGA did not affect MPO production, as determined by absorbance measurements (0.154 6 0.0323 A460/minute vs. 0.038 6 0.005 A460/minute vs. 0.039 6 0.004 A460/minute, p = .001; Fig. 3E). Successful inhibition of hAEC-derived lipoxin A 4 by NDGA is shown in Figure 3F

hAEC-Mediated Change in T-Cell Phenotype Is Not Mediated via Lipoxin A 4
We evaluated the effect of hAECs and hAEC-derived LXA4 on T-cell phenotype, function, and behavior. In comparison with the control treatment (i.e., medium alone), hAECs significantly decreased IL-2 and increased IL-17A levels in vitro (IL-2: 37.20 6 0.00 vs. 21.60 6 0.81, p = .031; IL-17A: 6.670 6 0.40 vs. 26.960 6 8.76, p = .005; Fig. 4A). We showed that T-cell chemotaxis toward CCL17 was unaltered by hAECs (Fig. 4B). However, coculture with hAECs significantly suppressed T-cell proliferation, which diminished in the presence of 2.5 mM NDGA (data not shown for 10 mM; 1.     [27]. We thus focused our attention on lipoxin A 4 a key regulator of macrophage infiltration and function in hAECmediated repair of lung fibrosis. We observed that hAEC treatment was associated with increased IL-4 cytokine expression at day 3 after bleomycin challenge, accompanied by a reduction in IL-10 at day 5, and IL-6 and IL-2 at day 7. Levels of inflammatory cytokines may affect immune cell infiltration into the lung, as we observed increased inflammatory monocyte numbers at day 3 followed by fewer macrophages and dendritic cells by day 5. These changes progressed, and on day 7, CD4 + -naïve T-cell and macrophage numbers were reduced in animals treated with hAECs. Next, we measured the effect of hAEC administration on endogenous production of lipoxin A 4 and its receptor FPR2, and showed significant increase in LXA4 levels in BALF and increased macrophage FPR2 expression on lung tissue at day 7. We further characterized lipoxygenase-mediated effects of hAECs on neutrophil, macrophage, and T-cell behavior, using a lipoxygenase-neutralizing compound, NDGA. We showed that hAECs increased macrophage phagocytosis and suppressed T-cell proliferation in a lipoxygenase-mediated fashion. However, hAECs' ability to reduce primary neutrophil MPO activity appeared to be independent of LXA4. Polarity of Th1 and Th2 cells is a stochastic process usually directed by proinflammatory (i.e., IL-6, IL-2, IL-12) and anti-inflammatory (i.e., IL-10, IL-4) cytokines, which provides positive feedback signaling for downstream inflammatory response. Bronchoepithelial cells express IL-4 receptors, which, upon activation by IL-4, results in the production of cytokines (i.e., monocyte-chemotactic protein -1, IL-8, and IL-1Ra) and increased migration of human airway epithelial cells [31,32]. IL-4is a known profibroticcytokine in preclinical studies using bleomycin. Huaux and colleagues showed that IL-4 2/2 mice have lower profibrotic factors such as OH-proline, soluble collagen, fibronectin, and tumor growth factor (TGF)-b1 from day 14 onward [33]. However, as our current study investigates changes during acute inflammation, we draw our attention toward the role of IL-4 in inflammation. Interestingly, IL-4 is a potent inducer of lipoxygenase -12 and -15. Chaitidis et al. showed that human peripheral blood monocytes strongly upregulated LOX-15 and conversely downregulated classical proinflammatory cytokines TNFa, IL-6, IL-1, and IL-8 when cultured with IL-4 [34]. In our study, we observed that administration of hAECs modestly increased levels of IL-4 at day 3. This suggests that application of hAECs can be detrimental in a fibrotic environment but beneficial in an inflammatory environment, and considerations toward clinical applications should be made with this in mind.
Conversely, we found significant reduction in Th1 drivers IL-6 and IL-2 by day 7, suggesting a repression of inflammatory responses. This is supported by an earlier report that showed blocking IL-6 at days 8, 9, and 10 after bleomycin instillation can ameliorate lung fibrosis [35]. Surprisingly, we observed a transient reduction in IL-10 levels on day 5, which contradicts a previous finding by Manuelpillai and colleagues that described an increase in IL-10 levels following hAEC treatment in a mouse model of liver fibrosis [36]. CD11c dendritic cells [37,38] and macrophages are potent producers of IL-10. Sakamoto and colleagues showed that after bleomycin challenge, the levels of IL-10 in bronchoalveolar mononuclear cells began increasing at day 3, then peaked at day 7 and persisted until day 14. It is, therefore, likely that the reduced levels of IL-10 on day 5 are reflective of the reduced infiltration of dendritic cells and macrophages on day 3 [39].
Lipoxin A 4 is a lipid mediator capable of anti-inflammatory signaling by regulating neutrophil infiltration and proresolving signaling through macrophage polarization and nonphlogistic uptake of apoptotic PMNs. Their innate ability to perpetuate and limit inflammation is of particular importance in our study to define probable immune regulatory roles of hAECs. As our first step in identifying the importance of LXA4 in hAEC-mediated repair, we evaluated endogenous levels of LXA4 and gene expression of its precursor enzymes ALOX-5, -12, and -15 in vivo.
Here, hAECs significantly increased levels of LXA4 in BALF at day 7 after bleomycin challenge. Gene expression of the enzymes ALOX-5, -12, and -15 responsible for the biosynthesis of lipoxin was significantly increased in hAEC-treated animals at day 7 (ALOX-5) and day 5 (ALOX-12 and -15), suggesting that hAEC treatment had an effect on endogenous LXA4 and lipoxygenase synthesis in bleomycin-induced lung injury. However, changes to lipoxygenase expression varied between days 5 and 7 and may be a consequence of the behavior of lipoxygenaseexpressing immune cells [40]. Treatment with hAECs could indirectly contribute to lipoxygenase expression of endogenous immune cells.
In our study, we saw no change in neutrophil numbers across all days between injured and treated animals, suggesting that, in our study at least, hAECs did not alter neutrophil infiltration to the site of injury and may, instead, mediate activity of neutrophils to resolve injury. Neutrophils are cells recruited as the first line of inflammatory defense, hence behavioral changes to these cells will inevitably influence downstream inflammatory processes. Neutrophil release of MPOscontributes to alveolar damage and, subsequently,fibrosisin patients with IPF, and their activity can be regulated by aspirin-triggered 15-epi-lipoxin A 4 in vitro [41][42][43]. Our previous studies showed that hAECs significantly reduced proinflammatory MPO-positive cells in vivo [4,8]. In this study, we showed that hAECs reduced neutrophil MPO activity in vitro, but this was unlikely to be mediated through LXA4. Although these results are contrary to our initial hypothesis, it was unsurprising, because other lipid-based mediators are involved in this process. Resolvin E1 administration inhibited PMN infiltration and reduced MPO levels in aspiration pneumonia in mice [44,45]. Taken together, these studies suggest that hAEC inhibition of neutrophil MPO activity is driven by alternatives to the lipoxygenase pathway.
Next, we assessed the changes to pulmonary monocyte subpopulations where two distinct subtypes can be found: Ly6C hi monocytes, which secrete inflammatory cytokines TNF-a and IL-1b; and Ly6C lo monocytes, which patrol the endothelium. The percentage of CD11b + Ly6C hi monocytes was significantly higher in hAEC-treated animals at days 3 and 5 of this study. This was unexpected because there was less inflammation and honeycomb injury in those mice. Although denoted as inflammatory monocytes, CD11b + Ly6C hi cells have the ability to differentiate into mature alternatively activated macrophages [46]. Gibbons and colleagues reported that the adoptive transfer of bonemarrow-derived Ly6C hi inflammatory monocytes during the progressive fibrotic phase led to increased numbers of M2 macrophages and exacerbation of fibrosis [47][48][49]. In a similar vein, M2 polarization during the peak of lung inflammation has been reported to have a protective effect in the bleomycin model [18], an outcome that was not achieved with later administration of hAECs, coinciding with the progressive fibrotic phase [45,50]. Collectively, this suggests that the spatiotemporal infiltration of inflammatory monocytes could be supporting an M2-driven proreparative response at day 7. Additionally, increase in LXA4 levels at day 7 in animals treated with hAECs could lead to the increase in inflammatory monocytes seen in our study. Studies have shown that LXA4 stimulates monocyte migration, adherence, and Ca 2+ motility [51,52]. This finding coincided with increased numbers of FPR2+ macrophages in lung tissues at day 7 in mice treated with hAECs; however, FPR expression was not wholly restricted to macrophages. This is unsurprising because FPR2 expression is not restricted to macrophages and is expressed by other phagocytic leukocytes such as neutrophils. Given that annexin A1 has been shown to bind strongly to FPR2 on the surface of activated PMNs [53] and is critical for leukocyte adhesion and rolling, as well as the attenuation of MPO activity in neutrophils [54], FPR2-annexin A1 interaction could explain the nonmacrophage effects observed in our current study.
Our observations suggest that hAECs exert their effects, at least in part, through LXA4, which is a potent stimulus for macrophage phagocytosis [55,56]. Macrophage phagocytosis of apoptotic PMNs is a rate-limiting step in inflammation, which could contribute to the pathogenesis of IPF. Alveolar macrophages of patients with IPF have reportedly impaired efferocytosis activity compared with alveolar macrophages from patients affected by other forms of interstitial lung disease [57,58]. It is plausible, therefore, that hAECs' ability to promote phagocytosis may contribute toward lung repair. However, to determine if hAEC-mediated increase in phagocytosis was driven by LXA4, we showed that neutralization of lipoxygenases negated the ability of hAECs to promote macrophage phagocytosis. Data reported here extend previous observations that hAECs polarized macrophages from an M1 to an M2 phenotype by increasing their mannose receptor CD206 expression [18]. As shown by Shirey and colleagues, neutralization of lipoxygenase pathways inhibits M2 macrophage differentiation in mice infected with respiratory syncytial virus [50,59]. Moreover, Maderna and colleagues showed that bone-marrow-derived macrophages from fpr2 2/2 have defective phagocytic ability for apoptotic PMN [25,60], further supporting the importance of LXA4-FPR2 binding in regulating macrophage phagocytosis.
LXA4 is a potent regulator of inflammation and resolution; however, other specialized lipid mediators released by hAECs remain to be investigated [61][62][63]. Our study showed an increase in LXA4 levels in vivo and in vitro, albeit at low levels (,0.05 ng/ml). Notably, lipoxins are formed in picogram and nanogram amounts in humans, with their potency maintained. Guilherme and colleagues showed that treatment with relatively low concentrations of 1 mg and 0.1 mg of aspirin-triggered lipoxin A 4 per mouse at days 7 and 10 significantly reduced inflammatory cell infiltration and reversed established fibrosis [27]. Furthermore, it should be noted that although NDGA has been shown to preferentially inhibit lipoxygenase-5 [64][65][66], its activity is not exclusive to lipoxygenases. Indeed, NDGA has also been shown to also block leukotriene B4 and it has potent antioxidative properties [67].
Dendritic cells (DCs) are antigen-presenting cells known to influence T-cell proliferation and plasticity. In this study, we showed that hAEC treatment significantly reduced DC infiltration (day 5) to the site of injury and this might subsequently have led to changes in T-cell numbers that we saw on day 7. This study identified possible mechanisms of hAEC-mediated regulation of T-cell behavior and phenotype. Our coculture migration and proliferation assays found that T-cell migration toward CCL17 was unaffected by the hAEC-conditioned medium but was able to suppress T-cell proliferation. To elucidate the consequence of hAEC-derived LXA4 on T-cell proliferation, we found that addition of NDGA mitigated hAEC suppression of T-cell proliferation. This may be due to the inhibition of extracellular signal-regulated kinase signaling, which is known to regulate T-cell proliferation [59].
The bleomycin mouse model of pulmonary fibrosis has come under recent scrutiny [10]. IPF is a disease with a complex pathology, and simple anti-inflammatory and antifibrotic approaches are not curative. Although the bleomycin model does not capture the full complexity of the disease, it does replicate the acute exacerbations of the disease [10] where proinflammatory cytokines predominate (e.g., TNF-a, IL-1b, IL-6, TGF-b, fibronectin, and procollagen-1) [68]. Furthermore, at the early stages of inflammation, when PGE2 and LTB4 provoke neutrophil extravasation, LXA4 and its epimer 15-epi-LXA4 was shown to have a profound inhibitory effect on PMN infiltration in vivo [69]. Our focus on acute inflammatory events provided an opportunity to investigate the immune regulatory mechanism of hAECs during acute exacerbations. Immunological changes exerted by hAECs beyond this period were not examined and should be the subject of future research. Previous work showed 4 million hAECs given 24 hours after bleomycin challenge via intraperitoneal administration ameliorated lung fibrosis and restored lung function in bleomycinchallenged mice irrespective of hAEC engraftment [4,45]; thus, this model was selected for our study. Whether lower doses, repeated dosing, or varying the route of administration will improve acute inflammatory exacerbations and/or long-term outcomes on fibrosis remains to be investigated. Assessing the efficacy of hAECs in a model of repeated bleomycin challenge will also likely be useful in terms of evaluating their therapeutic utility in intractable lung fibrosis.

CONCLUSION
Our current study provides evidence that hAECs can alter neutrophil, macrophage, and T-cell migration, phenotype, and behavior in resolving bleomycin-induced lung injury, in part through LXA4. In the process of lung repair, hAECs promote resolution by upregulating macrophage phagocytosis, partly through LXA. It was highlighted at the 2012 National Institutes of Health, National Heart, Lung, and Blood Institute workshop that understanding the mechanism of action of stem cells is paramount for the advancement of cell-based therapies [70]. Acumen of the biodynamics of cell therapy candidates in various disease or injury settings, therefore, will expedite the development of cell-based therapies for lung disease. Additional studies on how these exogenously delivered stem cells and stem-like cells aid other endogenous repair processes, such as their effects on stem/progenitor cell populations and extracellular matrices, during different phases of injury will be needed to best exploit their regenerative properties.