Lipoxin A4 Counter-regulates Histamine-stimulated Glycoconjugate Secretion in Conjunctival Goblet Cells

Conjunctival goblet cells synthesize and secrete mucins which play an important role in protecting the ocular surface. Pro-resolution mediators, such as lipoxin A4 (LXA4), are produced during inflammation returning the tissue to homeostasis and are also produced in non-inflamed tissues. The purpose of this study was to determine the actions of LXA4 on cultured human conjunctival goblet cell mucin secretion and increase in intracellular [Ca2+] ([Ca2+]i) and on histamine-stimulated responses. LXA4 increased mucin secretion and [Ca2+]i, and activated ERK1/2 in human goblet cells. Addition of LXA4 before resolvin D1 (RvD1) decreased RvD1 responses though RvD1 did not block LXA4 responses. LXA4 inhibited histamine-stimulated increases in mucin secretion, [Ca2+]i, and ERK1/2 activation through activation of β-adrenergic receptor kinase 1. We conclude that conjunctival goblet cells respond to LXA4 through the ALX/FPR2 receptor to maintain homeostasis of the ocular surface and regulate histamine responses and could provide a new therapeutic approach for allergic conjunctivitis and dry eye diseases.

Conjunctival goblet cells are specialized cells that span the thickness of the conjunctiva from the ocular surface to the stroma. These cells synthesize and secrete mucins which include the gel forming mucin MUC5AC in humans and in rats 1,2 . These mucins are responsible for maintenance of ocular surface hydration, lubrication, and prevention of destructive interaction of foreign bodies and pathogens with the conjunctiva. Goblet cells also play a role in the innate immune response of the conjunctiva and can be activated by cytokines produced during inflammation 3,4 . In the context of the ocular surface, the types of inflammation observed include seasonal allergic conjunctivitis, and dry eye syndrome 5,6 . Allergic conjunctivitis alone affects 15-25% of Americans 6 . Dry eye disease is a chronic, multifactorial disease and can be a result of graft vs host disease, autoimmune diseases, normal aging or refractive and cataract surgeries [7][8][9][10][11] . It has been estimated that the overall cost of dry eye disease treatment in the United States is more than $3.8 billion though that number is likely underestimated 12 . Uncontrolled inflammation is a hallmark of these ocular surface diseases causing redness, itching, and discomfort and creating a significant impact on quality of life. There are few effective treatments for these diseases and most are only palliative.
During the allergic response, mast cells are recruited to the ocular surface, degranulate, and release histamine and leukotrienes (LT) 13,14 . We previously showed that goblet cells of the conjunctiva play an active role in the response of the ocular surface to histamine and leukotriene challenge [15][16][17] . All four receptors (H1-H4) for histamine as well as cysteinyl leukotriene receptors, CysLT 1 and CysLT 2 , are expressed in goblet cells 17 . Activation of each of the these receptor subtypes caused an increase in intracellular [Ca 2+ ] ([Ca 2+ ] i ) and high molecular weight glycoconjugate secretion including MUC5AC 15,17 .
Termination of inflammation occurs with the biosynthesis of the specialized pro-resolution mediators (SPMs) resolvins (Rvs), lipoxins (LX), maresins, and protectins from omega-3 and -6 essential fatty acids 18 . These resolution-phase mediators alter the magnitude and the duration of the inflammatory response through mechanisms involving counter regulation of inflammatory mediators as well as phagocytosis of microbes and cell . The increase in [Ca 2+ ] i stimulated by LXA 4 was directly linked to mucin secretion as chelation of intracellular Ca 2+ blocked secretion 19 . In the present study, we investigated the actions of LXA 4 with cultured human conjunctival goblet cells, as well as the impact of LXA 4 on histamine-stimulated increase in [Ca 2+ ] i , mucin secretion, and ERK 1/2 activation in rat and human goblet cells. In human goblet cells, LXA 4 binds to the ALX/FPR2 and GPR32 receptors while RvD1 binds to GPR32 receptors. In rat, LXA 4 and RvD1 preferentially bind to the ALX/FPR2 receptor. In addition, we report that LXA 4 utilizes β -adrenergic receptor kinase (β ARK) 1 to counter-regulate the H1 histamine receptor.
Next, the actions of LXA 4 on glycoprotein secretion from cultured human goblet cells was determined. Goblet cells were serum-starved for 2 h and LXA 4 was added (10 −10 -10 −8 M) for 2 h and glycoconjugate secretion measured. LXA 4 (10 −9 M) increased secretion 2.6 ± 0.1 fold above basal (p = 0.01, Fig. 2). In cells from the same individuals, histamine, as a positive control, increased glycoconjugate secretion 2.5 ± 0.3 fold above basal (p = 0.005, data not shown). These data show that in human goblet cells, similar to rat goblet cells, LXA 4 activates the ALX/ FPR2 receptor to increase [Ca 2+ ] i and stimulate glycoconjugate secretion.

Presence and Localization of ALX/FPR2 Receptors in Human Conjunctival Goblet Cells. As
LXA 4 stimulates an increase in [Ca 2+ ] i and glycoconjugate secretion and the ALX/FPR2 receptor inhibitor, BOC-2, blocks LXA 4 -stimulated increase in [Ca 2+ ] i , we confirmed that ALX/FPR2 is expressed in human goblet cells. RT-PCR, using primers specific for this receptor, was performed. As shown in Fig. 3A, one band at the expected size was detected. The receptor is also expressed at the protein level as detected by western blot analysis from cells grown from 3 individuals (Fig. 3B). The ALX/FPR2 receptor is known to be glycosylated which could account for the multiple bands observed 24 . In cultured human goblet cells, ALX/FPR2 (shown in red) was present throughout the cytosol of the cells (Fig. 3C). UEA-1, shown in green was used to confirm the identity of cultured goblet cells (Fig. 3C). There was substantial overlap in the localization of ALX/FPR2 and UEA-1. These data confirm that ALX/FPR2 is present in human goblet cells.
Interaction of LXA 4 and RvD1 via ALX/FPR2 and GPR32. The ALX/FPR2 receptor has multiple agonists which can bind to it. These agonists include RvD1, the protein annexin A1, as well as LXA 4 25 . In addition to ALX/FPR2, RvD1 and LXA 4 also bind to the receptor GPR32 26 , which we previously demonstrated to be present in cultured human goblet cells 20 . We explored the interaction between LXA 4 and RvD1 with ALX/FPR2 and GPR32 in human goblet cells. In a first set of experiments, we determined the extent to which RvD1 binds to ALX/FPR2 in human goblet cells. Goblet cells were preincubated with ALX/FPR2 inhibitor BOC-2 (10 −4 M) prior to stimulation with RvD1 (10 −8 M) and [Ca 2+ ] i was measured. In the absence of inhibitor, RvD1 increased [Ca 2+ ] i by 242.8 ± 70.8 nM (Fig. 4A). Preincubation with BOC-2 had no effect on the RvD1 response (p = 0.14).
In a second set of experiments, the following experimental paradigm was used: addition of first agonist was followed 5 minutes later by addition of second agonist and the increase in peak [Ca 2+ ] i . was measured approximately 30 seconds after addition of agonist. In cultured human goblet cells, the addition of RvD1 (10 −8 M) first caused a peak increase in [Ca 2+ ] i of 256.2 ± 62.5 nM (Fig. 4B,C). A second addition of RvD1, 5 minutes after the first, resulted in a peak increase of [Ca 2+ ] i of 17.9 ± 13.8 nM (Fig. 4B,C). This was a decrease from the response obtained when RvD1 was added first (p = 0.04). If LXA 4 is added first and RvD1 is added 5 min later, the RvD1 response was 11.3 ± 11.3 nM. This is also a decrease from the response obtained if RvD1 is added first (p = 0.03).
We previously established that histamine increases [Ca 2+ ] i in a concentration-dependent manner which was blocked by RvD1 and AT-RvD1 20 . To determine if LXA 4 also blocks histamine responses, cultured goblet cells were preincubated with LXA 4 prior to stimulation with histamine. In goblet cells cultured from rat, histamine (   To ensure that the actions of LXA 4 on histamine response are mediated by the ALX/FPR2 receptor, rat goblet cells were pretreated with BOC-2 (10 −4 M) for 15 min prior to addition of LXA 4 (10 −9 M) for 30 min. The [Ca 2+ ] i was then measured in response to histamine (10 −5 M). In the absence of BOC-2 and LXA 4 , the change in peak [Ca 2+ ] i in response to histamine was 586.8 ± 173.1 nM (p = 0.01, Fig. 6E). LXA 4 added first reduced the histamine response by 90.3 ± 1.9% to 55.5 ± 8.5 nM (p = 0.03). Preincubation with BOC-2 reversed the inhibition by LXA 4 on the histamine response and increased [Ca 2+ ] i by 495.0 ± 76.6 nM (Fig. 6E).

ALX/FPR2 Uses βARK1, but Not Protein Kinase C, to Block the H1 Histamine Receptor
Simulated Increase in [Ca 2+ ] i . Examination of the H1 histamine receptor for phosphorylation sites using Scan Site (http://scansite.mit.edu/), showed that this receptor has consensus sequences for β -adrenergic receptor kinase 1 (β ARK1), also known as G-protein coupled receptor kinase (GRK)-2 and protein kinase C (PKC). We previously demonstrated that RvD1 binding to GPR32 activates both these kinases to counter-regulate the H1 histamine receptor to block the increase in [Ca 2+ ] i 20 . To determine if ALX/FPR2 and LXA 4 also use β ARK1 and/ or PKC to counter regulate histamine H1 receptor, rat goblet cells were pretreated with either LXA 4 or LXA 4 plus inhibitors to β ARK1 and PKC. The increase in [Ca 2+ ] i in response to the specific H1 receptor agonist, histamine dimaleate was measured in cultured rat goblet cells. Pretreatment with LXA 4 decreased the histamine dimaleate stimulated increase in [Ca 2+ ] i from 823.7 ± 154.1 nM above basal in the absence of LXA 4 to 140.9 ± 68.7 nM (p = 0.02, Fig. 8A,B). β ARK1 inhibitor peptide (10 −6 M) alone did not have an effect of the histamine dimaleate (p = 0.16, Fig. 8A,B). When cells were pretreated with β ARK1 inhibitor peptide followed by LXA 4 , blockage of the histamine dimaleate response by LXA 4 was completely reversed (Fig. 8A,B).
These data indicate that the activation of the ALX/FPR2 uses β ARK1 but not PKC to counter regulate the H1 histamine receptor in rat goblet cells. This is in contrast to RvD1 that uses both β ARK1 and PKC to counter regulate the H1 receptor.

Discussion
Our results demonstrate that LXA 4 plays a role in goblet cell function in both normal non-inflamed conditions and acute inflammatory conditions. Our hypothesis is that the ALX/FPR2 receptor is present in human conjunctival goblet cells and activation of the receptor by LXA 4 stimulates an increase in [Ca 2+ ] i and mucin secretion, which in rat goblet cells is protective in the eye and involves activation of phospholipase (PL) C, PLD, and PLA2 signaling pathways (Fig. 9A) 19 . In circumstances such as inflammation or pharmacological addition, LXA 4 inhibits histamine-stimulated increase in [Ca 2+ ] i , ERK 1/2 activation, and mucin secretion through the counter-regulation of histamine receptor by β ARK1 (Fig. 9B). This may be relevant in controlling excessive histamine release into the conjunctiva.
It is currently not known if any cells in the conjunctiva, including goblet cells, produce and secrete LXA 4 . Along these lines, Gronert et al. have demonstrated that the epithelial cells of the cornea endogenously express LXA 4 and the amount is increased upon wounding 22 . This LXA 4 could then diffuse via the tears to the goblet cells to stimulate mucin secretion. While LXA 4 is an appreciated pro-resolution mediator, the results from several studies indicate that LXA 4 and other pro-resolution mediators can also play a role within other organs in, physiological conditions that maybe organ specific. For example, LXA 4 is endogenously produced in the cornea and lacrimal gland under non-inflamed conditions 17 . RvD1 and AT-RvD1, similar to LXA 4 , alone stimulate conjunctival goblet cell functions 20 . These results imply that these mediators could assist in the maintenance of the normal homeostasis of the ocular surface by regulating goblet cell mucin secretion that is linked to ocular surface health. Allergic conjunctivitis is the most common type of inflammation of the ocular surface. In this condition, histamine interacts with H1-H4 histamine receptors, all of which are expressed in rat and human conjunctival goblet cells 15 . Histamine also increases [Ca 2+ ] i and mucin secretion in a concentration dependent manner 15 . Pre-incubation with LXA 4 blocked histamine-stimulated increase in [Ca 2+ ] i , mucin secretion and ERK 1/2. Thus, LXA 4 likely acts as a pro-resolution mediator acting on goblet cells of the conjunctiva to return mucin levels to normal. LXA 4 is likely to have similar effects on histamine-stimulated responses in other tissues. For example, LXA 4 inhibits histamine release from human lung mast cells 27 and histamine-stimulated paw edema in mice 28 .
This study examined the actions of LXA 4 on conjunctival goblet cells only. The ocular surface consists of multiple cell types and is covered by tears, which are a complex film that overspreads the ocular surface 29 . The actions of LXA 4 have not been tested on other types of cells on the ocular surface nor in the presence of tears.
Cultured human goblet cells often react similarly to LXA 4 as cultured rat goblet cells. In goblet cells from both species, LXA 4 stimulated an increase [Ca 2+ ] i , and mucin secretion to the same extent (current study and 20 ). Mucin secretion stimulated by cysteinyl leukotrienes in human goblet cells was also similar to that obtained with rat goblet cells 17 . There does appear to be several differences between rat and human goblet cells. In human goblet cells, the concentration LXA 4 required to maximally inhibit histamine-stimulated increase in [Ca 2+ ] i was 10 fold less that than that required in rat goblet cells. An additional difference was demonstrated by experiments involving interactions of LXA 4 and RvD1 with their receptors. In rat goblet cells, initial addition of either LXA 4 or RvD1 blocked the increase in [Ca 2+ ] i stimulated by a second addition of either LXA 4 or RvD1 indicating that these two SPMs bind to the same receptor 19 . However in human goblet cells, while an initial addition of LXA 4 blocks the RvD1 response, an initial addition of RvD1 does not block the LXA 4 response. In addition, BOC-2 does not alter RvD1-stimulated increase in [Ca 2+ ] i . These results support the notion that in human cells RvD1 preferentially activates GPR32 while LXA 4 activates both receptors (Fig. 9). LXA 4 is an established agonist of ALX/FPR2 and has been shown to bind to GPR32 in human phagocytes 26 . It is also known that RvD1 binds to both ALX/FPR2 and GPR32 26 . At this point it is not known if a rat homolog of GPR32 is present and functional in rat goblet cells. Since GPR32 has not yet been identified in rat, it is possible that RvD1 only binds to ALX/FPR2 in these rat cells. There are many other situations in which rat differs from human including regulatory T cell phenotypes 30 , wound healing in skin 31 , and glomerulonephritis 32 .
We previously showed the mechanism by which RvD1 prevents the actions of histamine in rat goblet cells 20 . We found that RvD1 counter-regulates the H1 histamine receptor by activation of both β ARK1 and PKC to prevent the H1 specific agonist-stimulated increase in [Ca 2+ ] i 20 . In contrast to RvD1, only an inhibitor of β ARK1 reversed the LXA 4 inhibition of H1 histamine receptor. Cooray et al. have demonstrated that ALX/FPR2 receptor can form hetero-and homodimers depending on the agonist bound 24 . Thus RvD1 and LXA 4 could form different dimer formations in rat goblet cells.
The signaling pathways activated by LXA 4 after binding to ALX/FPR2 are dependent on the cell type (Table 1). LXA 4 acting through ALX/FPR2 stimulated an increase in [Ca 2+ ] i , chemotaxis and adherence in human monocytes 33,34 . In human neutrophils, ALX/FPR2 activation leads to lipid remodeling, arachidonic acid release, and In conclusion, we demonstrate that ALX/FPR2 receptors are present on cultured human goblet cells, and that LXA 4 alone increases [Ca 2+ ] i , mucin secretion and ERK 1/2 activation. In addition, LXA 4 counter-regulates the H1 histamine receptor to block its activation thereby returning the ocular surface to homeostasis. LXA 4 thus plays a critical role in ocular surface health and maintenance in physiological conditions. In addition, LXA 4 protects the ocular surface from challenges of the external environment that induce ocular surface inflammatory and allergic diseases. Thus LXA 4 and this receptor axis may provide the basis for new therapeutic treatments for these diseases.

Materials and Methods
Synthetic LXA 4 was purchased from EMD Millipore (Billerica, MA) and RvD1 was purchased from Cayman Chemical, Ann Arbor, MI). Both compounds were dissolved in ethanol as supplied by the manufacturer and were stored at − 80 °C with minimal exposure to light. Immediately prior to use, the SPMs were diluted in with Krebs-Ringer bicarbonate buffer with HEPES (KRB-HEPES, 119 mM NaCl   In rat goblet cells, histamine via the H1 histamine receptor subtype activates phospholipase (PL) -C to stimulate extracellularregulated kinase 1/2 (ERK 1/2) which leads to mucin secretion. In addition, inositol trisphosphate (IP 3 ) is produced which also leads to release of Ca 2+ i and activation of Ca 2+ channels leading to mucin secretion Also in rat goblet cells activation of the ALX/FPR2 receptor stimulates PLC, -D, and A2. These phospholipases activate ERK 1/2 through phosphorylation (pERK 1/2), and protein kinase C (PKC). IP 3 is produced which leads to release of Ca 2+ i and activation of Ca 2+ channels leading to mucin secretion. (A) Activation of ALX/ FPR2 by either LXA 4 or RvD1 activates β -adrenergic receptor kinase 1 (β ARK1) to counter-regulate the H1 histamine receptor to prevent histamine-stimulated mucin secretion (B). In human goblet cells, RvD1 binds to GPR32 receptor and regulates goblet cells (A,B). The rat homolog of the human GPR32 if present remains to be identified.
Scientific RepoRts | 6:36124 | DOI: 10.1038/srep36124 Cell Culture. Goblet cells from human and rat conjunctiva were grown in organ culture as described and extensively characterized previously 4,16,17,[40][41][42] . The tissue plug was removed after nodules of cells were observed. First passage goblet cells were used in all experiments. The identity of cultured cells was periodically checked by evaluating staining with antibody to cytokeratin 7 (detects goblet cell bodies) and the lectin Ulex europaeus agglutinin (UEA)-1 (detects goblet cell secretory product) to ensure that goblet cells predominated. Measurement of Glycoconjugate Secretion. Cultured goblet cells were serum starved for 2 h before use and then stimulated with either LXA 4 or histamine in serum-free RPMI 1640 supplemented with 0.5% BSA for 2 h. Inhibitors were added 30 min prior to stimulation. Goblet cell secretion was measured using an enzyme-linked lectin assay (ELLA) with the lectin UEA-I. UEA-1 detects high molecular weight glycoconjugates containing L-fucose including mucin MUC5AC produced by goblet cells 43 . The media were collected and analyzed for the amount of lectin-detectable glycoconjugates, which quantifies the amount of goblet cell secretion as described earlier 17 . Glycoconjugate secretion was expressed as fold increase over basal that was set to 1.

Measurement of [Ca
Reverse Transcriptase (RT)-PCR. Cultured human goblet cells were homogenized in TRIzol and total RNA was isolated. One microgram of purified total RNA was used for complementary DNA (cDNA) synthesis using the Superscript First-Strand Synthesis system for RT-PCR (Invitrogen, Carlsbad,CA). The cDNA was amplified by the polymerase chain reaction (PCR) using primers specific to human ALX/FPR2 receptor using the Jumpstart REDTaq Readymix Reaction Mix (Sigma-Aldrich, St. Louis, MO) in a thermal cycler (Master Cycler, Eppendorf, Hauppauge, NY). The primers were from published sequences 44 . The forward primer sequence was GGA TTT GCA CCC ACT GCA TTT and reverse primer was ATC CAA GGT CCG AGA TCA C. These primers generated a product of 528 base pairs. β − Actin served as the positive control. The primers were from published sequences 45 . The conditions were as follows: 5 min at 95 °C followed by 35 cycles of 1 min at 94 °C, 3 s at annealing temperature for 1 min at 72 °C with a final hold at 72 °C for 10 min. Samples with no cDNA served as the negative control. Amplification products were separated by electrophoresis on a 1.5% agarose gel and visualized by ethidium bromide staining.
Immunofluorescence Microscopy. First passage cells were grown on glass cover slips and were fixed in 4% formaldehyde diluted in phosphate buffered saline (PBS, 145 mM NaCl, 7.3 mM Na 2 HPO 4 , and 2.7 mM NaH 2 PO 4 (pH 7.2)) for 4 hours at 4 °C. The coverslips were rinsed for 5 minutes in PBS, and nonspecific sites were blocked by incubation with 1% bovine serum albumin, and 0.2% Triton X-100 in PBS for 45 minutes at room temperature. ALX/FPR2 receptor antibody (Novus Biologics) was used at 1:100 dilution overnight at 4 °C. UEA-1 directly conjugated to FITC (Sigma-Aldrich, St. Louis, MO) was used at a dilution of 1:300 to identify goblet cells. Secondary antibodies were conjugated to Cy 3 (Jackson ImmunoResearch Laboratories, West Grove, PA) was used at a dilution of 1:150 for 1 h at room temperature. Negative control experiments included incubation with the isotype control antibody. The cells were viewed by fluorescence microscopy (Eclipse E80i; Nikon, Tokyo, Japan) and micrographs were taken with a digital camera (Spot; Diagnostic Instruments, Inc, Sterling Heights, MI).

Statistical analysis.
Results were expressed as the fold-increase above basal. Results are presented as mean ± SD. Data were analyzed by ANOVA followed by post-hoc Tukey or Student's t-test. P < 0.05 was considered statistically significant.