EphA2 Transmembrane Domain Is Uniquely Required for Keratinocyte Migration by Regulating Ephrin-A1 Levels

Accepted Manuscript EphA2 transmembrane domain is uniquely required for keratinocyte migration by regulating ephrin-A1 levels Rosa Ventrella, Nihal Kaplan, Paul Hoover, Bethany E. Perez White, Robert M. Lavker, Spiro Getsios PII:

S0022-202X(18)31890-6

DOI:

10.1016/j.jid.2018.04.011

Reference:

JID 1400

To appear in:

The Journal of Investigative Dermatology

Received Date: 2 December 2017 Revised Date:

11 April 2018

Accepted Date: 12 April 2018

Please cite this article as: Ventrella R, Kaplan N, Hoover P, Perez White BE, Lavker RM, Getsios S, EphA2 transmembrane domain is uniquely required for keratinocyte migration by regulating ephrin-A1 levels, The Journal of Investigative Dermatology (2018), doi: 10.1016/j.jid.2018.04.011. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT

EphA2 transmembrane domain is uniquely required for keratinocyte migration by

RI PT

regulating ephrin-A1 levels

AUTHORS: Rosa Ventrella1, Nihal Kaplan1, Paul Hoover1, Bethany E. Perez White1, Robert M. Lavker1, and Spiro Getsios1,*

1

SC

AFFILIATIONS:

Department of Dermatology, 303 E. Chicago Ave, Ward 9, Northwestern University, Chicago,

M AN U

IL 60611

*Corresponding author’s current address: GlaxoSmithKline, 1250 S. Collegeville Rd.,

TE D

Collegeville, PA 19426. Email address: [email protected]. Phone: (773) 715-5310.

AC C

EP

SHORT TITLE: EphA2 transmembrane domain role in keratinocyte migration

Abbreviations: Eph receptors-Erythropoietin-producing human hepatocellular receptors; Knockdown-KD; Normal human epidermal keratinocytes-NHEKs; Transmembrane domain-TMD; Receptor tyrosine kinase-RTK; Reconstituted human epidermis-RHE

ACCEPTED MANUSCRIPT

ABSTRACT EphA2 receptor tyrosine kinase (RTK) is activated by ephrin-A1 ligand, which harbors a GPI-

RI PT

anchor that enhances lipid raft localization. While EphA2 and ephrin-A1 modulate keratinocyte migration and differentiation, the ability of this cell-cell communication complex to localize to different membrane regions in keratinocytes remains unknown. Using a combination of

biochemical and imaging approaches, we provide evidence that ephrin-A1 and a ligand-activated

SC

form of EphA2 partition outside of lipid raft domains in response to calcium-mediated cell-cell

M AN U

contact stabilization in normal human epidermal keratinocytes (NHEKs). EphA2 transmembrane domain (TMD) swapping with a shorter and molecularly distinct TMD of EphA1 resulted in decreased localization of this RTK at cell-cell junctions and increased expression of ephrin-A1, which is a negative regulator of keratinocyte migration. Accordingly, altered EphA2 membrane distribution at cell-cell contacts limited the ability of keratinocytes to seal linear scratch wounds

TE D

in vitro in an ephrin-A1-dependent manner. Collectively, these studies highlight a key role for the EphA2 TMD in receptor-ligand membrane distribution at cell-cell contacts that modulates

wound healing.

EP

ephrin-A1 levels to allow for efficient keratinocyte migration with relevance for cutaneous

AC C

INTRODUCTION

Eph receptors represent the largest family of receptor tyrosine kinases (RTKs) in

mammals and are activated by ephrin ligands at cell-cell contacts. This results in bidirectional signaling with forward signaling through the receptor expressing cell and reverse signaling through the ligand expressing cell (Nikolov et al., 2013, Noberini et al., 2012, Pasquale, 2005). Specifically, the ephrin-A family of ligands activate EphA receptors and contain a glycosylphosphatidylinositol (GPI)-linked tail allowing for their partitioning into lipid raft

2

ACCEPTED MANUSCRIPT

domains (Campbell et al., 2008, Davy et al., 1999, Gauthier and Robbins, 2003). Lipid rafts are highly ordered membrane microdomains that are rich in a variety of sterols, most notably cholesterol, and organize protein complexes at the membrane to allow for efficient signal

RI PT

transduction (Allen et al., 2007, Munro, 2003). In keratinocytes, lipid rafts impact proliferation, migration, apoptosis, and differentiation through mechanisms that remain relatively unclear (Calay et al., 2010, Giltaire et al., 2011, Jans et al., 2004, Lambert et al., 2008, Mathay et al.,

SC

2008).

M AN U

Following ligand binding, Eph receptors form a signaling cluster allowing for activation of downstream signaling pathways, which can be terminated by endocytosis of the ligand/receptor complex (Pitulescu and Adams, 2010). Alternatively, when the receptor and ligand are co-expressed in the same cell, ligand-mediated cis-inhibition results in attenuation of receptor activation (Egea and Klein, 2007, Falivelli et al., 2013, Kao and Kania, 2011). This

TE D

inhibition can occur by ephrins interacting with the Eph ligand binding domain or by preventing formation of Eph signaling clusters (Carvalho et al., 2006, Yin et al., 2004). Ephrin binding induces the transmembrane domain (TMD) of the receptor to undergo a

EP

conformational change thereby promoting reorganization of neighboring unliganded receptors to

AC C

form active Eph signaling clusters (Bocharov et al., 2010, Bocharov et al., 2017, Sharonov et al., 2014). Also, the TMD can affect the affinity that a transmembrane receptor has for lipid raft membrane domains (Lorent et al., 2017, Yuan et al., 2006). Interestingly, the EphA2 receptor is highly homologous to the other EphA family member EphA1, however they differ in the sequence and properties of their TMD (Bocharov et al., 2017, Sharonov et al., 2014). This led to the hypothesis that the distinct TMD of EphA2 regulates its plasma membrane localization relative to ephrin-A1 ligand thus modulating its downstream signaling.

3

ACCEPTED MANUSCRIPT

Recent studies have illustrated that ephrin-A1-induced forward signaling through EphA2 inhibits keratinocyte migration (Kaplan et al., 2012, Walsh and Blumenberg, 2011). Although there is evidence for ephrin distribution in lipid rafts, the ability of Eph receptors to associate

RI PT

with these membrane regions in keratinocytes is unknown. Since the balance of Eph/ephrin

trans-activation and cis-inhibition impacts downstream signaling, the membrane distribution of receptor and ligand likely plays an important role in modulation of keratinocyte behavior. Hence,

SC

we sought to define the localization of EphA2 and ephrin-A1 in the plasma membrane of

keratinocytes to better understand the dynamics of this cell-cell communication complex as

M AN U

keratinocytes stabilize their cellular junctions. We found that ligand-activated EphA2, along with ephrin-A1, concentrate outside of lipid raft domains during calcium-induced cell-cell contact stabilization. Moreover, the EphA2 TMD governs receptor localization to cell-cell contacts where ephrin-A1 ligand is present, thereby regulating ephrin-A1 protein levels and keratinocyte

TE D

migration. These results provide insight into the requirement of the EphA2 TMD for controlling receptor localization and downstream signaling events and will be particularly useful in

RESULTS

EP

delineating how Eph/ephrins regulate epithelial cell migration and tissue repair.

AC C

Lipid raft domain association of EphA2 and ephrin-A1 in human keratinocytes Lipid rafts contribute to keratinocyte differentiation and migration through mechanisms that include regulation of desmosome dynamics and EGFR signaling, respectively (Overmiller et al., 2016, Stahley et al., 2014). Since ephrin-A1 is a negative regulator of keratinocyte migration largely through action on EphA2, we characterized the lipid raft localization of this ligand/receptor combination (Kaplan et al., 2012).

4

ACCEPTED MANUSCRIPT

In order to assess Eph/ephrin dynamics during cell-cell contact stabilization, confluent keratinocytes were cultured in low calcium (0.03 mM) medium to limit terminal differentiation and then switched to high calcium (1.2 mM) culture medium (Hennings et al., 1980). Under

RI PT

confluent culture conditions, basal activation of EphA2 was evident as detected by its

phosphorylation at residue Y772 (pY772 EphA2; Sup. Fig. 1A). Ligand-activated EphA2 was further elevated after a 60-minute calcium switch (Sup. Fig. 1A-B). When keratinocytes were in

SC

low calcium medium, EphA2 and pY772 EphA2 were present in lipid raft domains along with ephrin-A1 (Fig. 1A, 0 minutes). After a calcium switch, EphA2 distribution was increased in

M AN U

lipid raft domains whereas pY772 EphA2 was more prevalent outside of lipid raft fractions (Fig. 1A-C). The abundance of ephrin-A1 outside of lipid raft domains increased as cell-cell contacts were stabilized following the pattern of ligand-activated EphA2 (Fig. 1A, E). EphA1, which is highly homologous to EphA2 and also expressed in keratinocytes, was localized to lipid raft

TE D

domains but remained steady in these membrane regions during cell-cell contact stabilization (Fig. 1A, D). Changes in protein lipid raft concentration were likely not dependent on how lipid raft fractions were being molecularly defined since concentrations of caveolin-1 and flotillin-1 in

EP

lipid raft defined membrane fractions remained unchanged among samples (Fig. 1A, F-G). Prior to cell-cell contact stabilization in high calcium, EphA2 and basally active pY772

AC C

EphA2 were concentrated in lipid raft domains (Fig. 1A-C) and present at cell borders (Fig. 2A). As cell-cell contacts mature in response to a calcium switch, the colocalization of pY772 EphA2 in caveolin-positive lipid raft domains decreased (Fig. 2A-B) matching the pattern of lipid raft depletion observed biochemically (Fig. 1A, C). However, colocalization between total EphA2 and caveolin-1 remained unchanged (Fig. 2A, C). Collectively, these data suggest that a portion of EphA2 is present in lipid rafts along with ephrin-A1 in keratinocytes and an activated form of

5

ACCEPTED MANUSCRIPT

this receptor is either preferentially depleted out of lipid rafts or accumulates in these distinct membrane domains from a separate pool along with ligand upon cell-cell contact stabilization.

RI PT

EphA2 forward signaling can be studied by pharmacological delivery of a soluble ephrinA1-Fc recombinant protein (Barquilla and Pasquale, 2015). We used this agonist-based approach to further assess how activation of EphA2 causes membrane redistribution of this receptor. As expected, ephrin-A1-Fc treatment resulted in increased pY772 activation of EphA2 15 and 60

SC

minutes after treatment (Sup. Fig. 2A-B) and correlated with a partitioning of pY772 EphA2, but

M AN U

not total EphA2, out of lipid raft domains (Sup. Fig. 2F-G). Again, molecularly defined fractions containing caveolin-1 and flotillin-1 remained unchanged between samples (Sup. Fig. 2H-I). These observations indicate that ephrin-A1-induced activation of EphA2 triggers an accumulation of pY772 EphA2 outside of lipid raft domains.

TE D

The EphA2 TMD is required for receptor localization at cell-cell contacts During cell-cell contact stabilization, EphA2 is localized to borders where it is likely activated by ephrin ligands (Lin et al., 2010). To minimize the impact of differentiation, an intermediate

EP

(0.2 mM) calcium concentration was used to assess receptor distribution in keratinocytes with relatively stable cell-cell contacts (O'Keefe et al., 1987, Yin et al., 2005). Interestingly, EphA2

AC C

and EphA1 exhibited distinct localization patterns in keratinocytes after cell-cell contacts were stabilized in 0.2 mM calcium culture medium for 24 hours (Fig. 3A). Specifically, EphA2 was concentrated at sites of cell-cell contact whereas EphA1 had a diffuse localization pattern. Although these two receptors have high amino acid sequence homology (approximately 65%), they contain several structural differences including EphA1 having a TMD domain that is predicted to be shorter than EphA2 with decreased sequence homology compared to the entirety of the receptor (approximately 36%) (Fig. 3B) (Bocharov et al., 2017, Sharonov et al., 2014).

6

ACCEPTED MANUSCRIPT

Since TMD length and amino acid sequence impact membrane localization, we hypothesized that the unique properties of the EphA2 TMD may play a role in this RTKs border localization

RI PT

following cell-cell contact stabilization (Diaz-Rohrer et al., 2014). To assess the functional significance of the EphA2 TMD in cell-cell junction localization, this region of EphA2 was replaced with the corresponding amino acid sequence of EphA1 to generate a chimera with the EphA1 TMD inserted between the EphA2 extracellular and

SC

cytoplasmic domains (Chimera 212) (Fig. 3B). In order to compare the localization of Chimera

M AN U

212 to that of EphA2, these ectopic proteins were expressed in keratinocytes where endogenous levels of EphA2 were knocked down (shEphA2.pLKO) (Fig. 3A, C). Unlike wild-type EphA2, Chimera 212 failed to localize to cell-cell contacts (Fig. 3C). Interestingly, the localization pattern of Chimera 212 more closely resembled that of EphA1 than EphA2 suggesting that the TMD plays an important role in these distinct localization patterns. Although Chimera 212 did

TE D

not concentrate at borders, a detectable portion of Chimera 212 was present at cell-cell contacts and biotinylated at the cell surface as junctions were being established in high calcium for 15 minutes (Sup. Fig. 3). Taken together, these findings suggest that Chimera 212 retains the ability

EP

to be trafficked to the cell surface but is poorly stabilized at regions of cell-cell contact.

AC C

Epidermal junctions are arguably more reflective of the in vivo state when studied in 3-D reconstituted human epidermis (RHE) models (Simpson et al., 2010). In this culture model, EphA2 exhibited a junctional localization pattern whereas Chimera 212 remained mostly intracellular emulating its localization pattern during 2-D cell-cell contact stabilization (Fig. 3D). Collectively, these data add weight to the idea that the EphA2 TMD contributes to the localization of this RTK at cell borders. Ephrin-A1 levels are regulated by EphA2 in a manner that depends on its TMD

7

ACCEPTED MANUSCRIPT

Differences in receptor localization at cell-cell contacts between EphA2 and Chimera 212 led us to investigate the ability of Chimera 212 to partition in lipid raft domains. Sucrose gradients were performed 24 hours after cell-cell contacts were stabilized where endogenous EphA2 is normally

RI PT

concentrated at cell borders. When EphA2 or Chimera 212 was expressed in keratinocytes with endogenous EphA2, Chimera 212 decreased total receptor, including both endogenous EphA2 and Chimera 212, distribution in lipid rafts (Fig. 4A, B). Ectopic expression of wild type EphA2

SC

also trended toward a reduction in total receptor lipid raft distribution suggesting that this

outcome may be a consequence of exogenous EphA2 receptor overexpression in keratinocytes

M AN U

(Fig. 4A, B). Consequently, we examined the membrane distribution of Chimera 212 and wild type EphA2 in keratinocytes lacking endogenous EphA2 (shEphA2.pLKO). Both EphA2 and Chimera 212 were similarly concentrated in lipid raft domain fractions, despite markedly distinct localization patterns, suggesting that the EphA2 TMD is not uniquely required for receptor lipid

TE D

raft domain association (Fig. 4C). Exogenous receptor expression did not alter the relative abundance of caveolin-1 or flotillin-1 in lipid raft fractions in control (pLKO) or EphA2 knockdown (shEphA2.pLKO) keratinocytes (Sup. Fig. 4).

EP

Although, EphA2 and Chimera 212 concentrated in lipid raft domains to a similar extent, the membrane distribution of pY772 EphA2 and ephrin-A1 was distinct from wild-type receptor

AC C

conditions. In particular, Chimera 212 led to an accumulation of pY772 EphA2 outside of lipid raft domains despite having increased ephrin-A1 (Fig. 4D, E). These observations suggest that Chimera 212 likely concentrates in lipid raft domains void of ephrin-A1 whereas endogenously activated EphA2 co-distributes with ephrin-A1 at cell-cell contacts during calcium-induced differentiation (Figure 1). These results suggest that the EphA2 TMD does not affect the ability

8

ACCEPTED MANUSCRIPT

of this receptor to enter in lipid raft domains but rather impairs receptor localization to membrane regions at intercellular junctions where ephrin-A1 ligand is present.

RI PT

Following ephrin-induced activation of EphA2 at cell-cell contacts, Eph/ephrin complexes are internalized and degraded or recycled back to the cell surface (Pitulescu and

Adams, 2010). Ultimately, this results in a down-modulation of ephrin levels when EphA2 is prominently expressed (Sabet et al., 2015). This clearing of ephrins from the plasma membrane

SC

likely relies on the ability of Eph receptors to engage ligand at cell-cell contacts. Therefore, we

M AN U

hypothesized that lack of Chimera 212 localization at cell-cell contacts and depletion of endogenous EphA2 from lipid rafts would impact the ability of this receptor to regulate ephrinA1 levels in keratinocytes. Interestingly, total ephrin-A1 protein levels were markedly increased when Chimera 212 was expressed in keratinocytes under conditions where cell-cell contacts were stabilized for 48 hours (Fig. 5A, B). We failed to detect major changes in ephrin-A1

TE D

transcript levels suggesting that this change in expression did not reflect altered transcriptional regulation (Fig. 5C). Also, expression of Chimera 212 resulted in smaller protein species likely representing ephrin-A1 fragments generated by protease cleavage as previously described in

EP

human glioblastoma and mammary adenocarcinoma cells (Beauchamp et al., 2012, Janes et al., 2005, Wykosky et al., 2008). These findings indicate that Chimera 212 impacted ephrin-A1

AC C

homeostasis in keratinocytes. To examine if ephrin-A1 internalization at cell-cell contacts was altered by Chimera 212,

a silicone chamber co-culture system was used (Kaplan et al., 2018). The silicone chamber allowed for separation of two different keratinocyte populations; one population overexpressing ephrin-A1 and the other population either overexpressing wild type EphA2 or expressing Chimera 212. Removal of the silicon chamber allowed for formation of a heterotypic interface

9

ACCEPTED MANUSCRIPT

between these two cell populations. When ephrin-A1 confronted EphA2 expressing keratinocytes, both the receptor and ligand colocalized in discrete puncta in the EphA2 overexpressing cells likely due to trans-endocytosis of ephrin-A1 at cell-cell contacts (Fig. 5D).

RI PT

However, when ephrin-A1 overexpressing cells encountered Chimera 212 expressing cells

ephrin-A1 puncta were largely absent in the Chimera 212 cells (Fig. 5E). Collectively, these data suggest that Chimera 212-induced increases in ephrin-A1 are possibly due to inability of this

SC

mutant receptor to efficiently internalize ephrin-A1 at cell-cell contacts.

M AN U

EphA2 TMD is required for efficient keratinocyte migration

Due to increased ephrin-A1 in keratinocytes expressing Chimera 212 and the ability of ephrinA1 to negatively regulate cell migration, we investigated how expression of Chimera 212 would impact keratinocyte migration (Kaplan et al., 2012, Miao et al., 2009). We anticipated that the migratory response of Chimera 212 would be altered due to enhanced availability of ephrin-A1

TE D

in the cellular microenvironment. To test this possibility, scratch-wounds were made in confluent keratinocyte cultures with cell-cell contacts stabilized for 24 hours.

EP

Consistent with previous studies, treatment with ephrin-A1-Fc recombinant protein (1 µg/mL) at the time of wounding inhibited keratinocyte migration (Kaplan et al., 2012, Miao et

AC C

al., 2009) (Fig. 6A). Furthermore, keratinocytes overexpressing EphA1 or EphA2 showed reduced migration when treated with ephrin-A1-Fc while overexpression of the respective receptor alone had minimal impact (Fig. 6A). Strikingly, expression of Chimera 212 significantly inhibited keratinocyte migration to a similar extent as ephrin-A1-Fc treatment (Fig. 6A). Since Chimera 212 increased ephrin-A1 levels in keratinocytes, we hypothesized that inhibition of migration would be normalized by preventing the induction in ephrin-A1, which we achieved via siRNA-mediated knockdown (Fig. 6B). Consistent with its role in restricting 10

ACCEPTED MANUSCRIPT

migration, reducing ephrin-A1 levels by itself led to more efficient closure of linear scratch wounds (Fig. 6C). Moreover, when ephrin-A1 was knocked down in keratinocytes expressing Chimera 212, migration was normalized to control levels, consistent with the notion that this

RI PT

EphA2 mutant deficient in lipid raft association inhibited migration largely as a consequence of increased ephrin-A1 levels (Fig. 6C). Collectively, our data suggest that the EphA2 TMD has important functions in regulating ephrin-A1 levels in order to allow for efficient keratinocyte

SC

migration, at least in a simplified cell culture model of re-epithelialization.

M AN U

DISCUSSION

An extensive body of work has contributed to our understanding of Eph/ephrin signaling, specifically in tissue patterning and carcinogenesis. However, a knowledge gap remains in how these receptor/ligand pairs localize to specific membrane microdomains to allow for pathway activation to occur (Barquilla and Pasquale, 2015). We now demonstrate that EphA2 and ephrin-

TE D

A1 exhibit dynamic movements between membrane microdomains, including lipid rafts, during keratinocyte cell-cell contact stabilization. Additionally, the EphA2 TMD plays an important role in receptor localization at cell-cell contacts, which impacts ephrin-A1 levels and

EP

keratinocyte migration.

AC C

A single-pass transmembrane receptors affinity for lipid rafts is dependent on its TMD length, the surface area of the amino acid side chains in its TMD, and palmitoylation proximal to the cytoplasmic side of the TMD (Lorent et al., 2017). Both EphA1 and EphA2 have similar amino acid side chain surface areas compared to that of the lipid raft preferring Linker for Activation of T-cells (LAT) (Lorent et al., 2017, Yuan et al., 2006). However, unlike LAT, EphA1 and EphA2 lack cysteine residues within 15 residues on the cytoplasmic side of their TMD suggesting that they lack palmitoylation sites that would influence lipid raft localization.

11

ACCEPTED MANUSCRIPT

EphA2, but not EphA1, contains four cysteine residues (C9, C247, C376, and C612) in its extracellular domain that are predicted as potential palmitoylation sites. However, we were unable to detect EphA2 palmitoylation either in NHEKs under a variety of experimental

RI PT

conditions (data not shown).

The main TMD differences between EphA1 and EphA2 are in TMD length and amino acid sequence compositions. Here we show that the EphA2 TMD plays an important role for its

SC

localization to cell-cell contacts and impacts on ephrin-A1 levels as well as keratinocyte

M AN U

migration. Similar to the TMD of VE-Cadherin binding to VEGFR2/3 in endothelial, EphA2 TMD may also interact with other membrane RTKs or adhesion receptors in order to prevent ephrin-A1 accumulation and allow for efficient keratinocyte migration to occur (Coon et al., 2015).

Ephrin-A ligands have the ability to localize to lipid raft domains and are negative

TE D

regulators of keratinocyte migration (Gauthier and Robbins, 2003, Kaplan et al., 2012, Walsh and Blumenberg, 2011). Although, the ability of Eph receptors to localize to lipid raft domains was previously unknown, caveolin-1 was identified as part of an EphA2 interactome study in

EP

RHE cultures (Perez White et al., 2017). In epidermis, ephrin ligands and EphA2 are both

AC C

present in the basal layer where lipid rafts are enriched (Gniadecki and Bang, 2003, Gordon et al., 2013, McGuinn and Mahoney, 2014). Therefore, lipid rafts could play a vital role in organizing Eph/ephrin complexes in the progenitor layer of the epidermis. One possible outcome of this overlapping expression pattern is continual ligand-induced activation of EphA2 in basal keratinocytes that leads to EphA2/ephrin-A1 removal from lipid raft domains. Similar dynamics have been seen where ephrin-A1/EphA2 undergo movements that are inversely correlated with the lipid raft protein, CD44 (Murai, 2015, Salaita et al., 2010). This clearing of the ligand-

12

ACCEPTED MANUSCRIPT

activated EphA2 from lipid raft domains in basal cells may limit keratinocyte migration under homeostatic conditions.

RI PT

Our findings reveal that the TMD of EphA2 plays a role in the localization of this receptor to cell-cell contacts without impacting lipid raft domain association. Lack of Chimera 212 localization at cell-cell contacts suggest that this chimeric receptor may be concentrating in lipid rafts at regions other than those present at intercellular junctions such as endosomes, Golgi

SC

apparati, or other plasma membrane regions (Rajendran and Simons, 2005). Lack of Chimera

M AN U

212 junctional localization, most likely impairs the ability of this mutant receptor to engage with ephrin-A1 at cell-cell contacts resulting in altered ligand internalization. Under homeostatic conditions, ephrin-A1 levels are maintained at a steady-state with a balance of transcriptional upregulation and ligand turnover that is governed by the presence of receptors like EphA2 (Pitulescu and Adams, 2010). Following ligand binding, ephrins can be cleaved by proteases

TE D

allowing for ephrin transendocytosis or proteolysis (Atapattu et al., 2012, Beauchamp et al., 2012, Janes et al., 2005). Since Chimera 212 decreases endogenous EphA2 concentration in lipid rafts, we speculate that this altered localization impedes ephrin-A1 transendocytosis and

EP

turnover. Interestingly, processed ephrin-A1 fragments were seen with Chimera 212 expression indicative of altered ephrin-A1 degradation under these conditions. Our data suggest that this

AC C

increase in ephrin-A1 is likely responsible for Chimera 212-induced inhibition of keratinocyte migration.

Although cell migration is required for normal tissue patterning and wound healing,

cancer cells utilize similar mechanisms to promote tumor invasion. Redundant molecular mechanisms control cell migration, including those engaged by Eph/ephrins (Ventrella et al., 2017). Tumors that express high levels of EphA2 are often accompanied by loss of ephrin-A1.

13

ACCEPTED MANUSCRIPT

This expression pattern results in EphA2 ligand-independent cell migration, which can be inhibited by restoring ligand in the microenvironment with recombinant ephrin-A1-Fc (Miao et al., 2009). Conversely, when wound healing is delayed, like in the diabetic corneal epithelium,

RI PT

ephrin-A1 expression is increased (Kaplan et al., 2012, Xu and Yu, 2011). Our studies provide additional support for the notion that ephrin-A1 abundance is a key determinant of cell

migration. We now extend these findings to include the importance of the EphA2 TMD for

SC

receptor cell-cell contact localization as a mechanism in regulating ephrin-A1 levels in epithelial cells. The ability of ephrin-A1 to decrease keratinocyte migration provides more weight to the

M AN U

argument that recombinant ephrin-A1-Fc treatment or other EphA2 targeting approaches that disrupt the receptor’s lipid raft localization can be used to reduce tumor cell migration. Additionally, studies to better understand EphA2/ephrin-A1 endocytic trafficking and degradation pathways following removal from lipid raft domains may prove useful in identifying

TE D

potential therapeutic targets to decrease ephrin-A1 levels and promote keratinocyte migration during cutaneous wound healing. MATERIALS & METHODS

EP

Cell culture and generation of RHE

AC C

Primary normal human epidermal keratinocytes (NHEKs) were isolated from neonatal foreskins obtained from the Skin Disease Research Center (SDRC) Tissue Engineering Core as previously described (Simpson et al., 2010). Experimental replicates used clones isolated from at least three different donors and pooled in order to account for possible clonal variation. RHE cultures were generated as described earlier (Simpson et al., 2010). Patient consent for experiments was not required as neonatal foreskin tissue is de-identified and considered discarded material by Northwestern University Institutional Review Board policy.

14

ACCEPTED MANUSCRIPT

Antibodies Antibody information is available in Supplementary Table 1.

RI PT

Gene silencing and re-expression Gene silencing of EphA2 was performed using a PLKO-based lentiviral vector (gift from

Bingcheng Wang; Case Western Reserve University, Cleveland, OH) and packaged in HEK293T

SC

cells with help from the SDRC DNA/Gene Delivery Core (Miao et al., 2015). siRNA

oligonucleotide duplexes (20 nM) were used to silence ephrin-A1 (Santa Cruz Biotechnology,

M AN U

Dallas, TX, sc-39426) and GC-matched siRNA was used as a negative control (ThermoFisher Scientific, Walthan, MA) as described before (Lin et al., 2010).

Full-length human EphA1 (Open Biosystems; Dharmacon Inc., Lafayette, CO) and EphA2 (gift from Bingcheng Wang) were subcloned into the pLZRS-Linker retroviral vector and

TE D

packaged in Phoenix cells as previously described (Denning et al., 2002, Miao et al., 2009). Generation of Chimera 212

Chimera 212 contains the EphA2 extracellular domain (amino acids 1-522), EphA1 TMD

EP

(amino acids 544-572), and EphA2 cytoplasmic domain (amino acids 564-976). In-Fusion cloning (Takara Bio Inc., Mountain View, CA) was used to subclone Chimera 212 into the

AC C

pLZRS-Linker vector (Denning et al., 2002, Miao et al., 2009). Western blot analysis

NHEKs were lysed in urea sample buffer (USB) and protein was separated using SDS-PAGE as previously described (Lin et al., 2010). FIJI imaging software was used to quantify protein band intensities (Schindelin et al., 2012). When two groups were compared a paired t-test was used. When more than two groups were compared, a paired one-way ANOVA was used followed by a Tukey post hoc test between all of the groups.

15

ACCEPTED MANUSCRIPT

Sucrose density centrifugation gradients NHEK pellets were resuspended in 0.5 M Sodium-carbonate buffer (0.5 M Na-carbonate pH 11,

RI PT

1 mM sodium orthovanadate, protease/phosphatase inhibitors) (Song et al., 1996). Lysates were homogenized by passing through needles and a Dounce homogenizer. A 1:1 dilution of sample and 90% sucrose in MESNA buffer (25 mM Mes pH6.5, 0.15 M NaCl) was added to the bottom of the ultracentrifuge tube to create a 45% sucrose mixture. Then, 35% and 5% sucrose buffers

SC

were layered for a discontinuous sucrose gradient. Samples were centrifuged at 44,000 rpm for

M AN U

18 hours in an ultracentrifuge (Sorvall WX Ultra 80, ThermoFisher Scientific) followed by collection of 12 equal-volume fractions into Laemmli buffer. Equal volumes from each fraction were subjected to immunoblotting.

Densitometry was done on all 12 fractions and positive lipid raft fractions were defined as having greater than 10% of the total amount of either caveolin-1 or flotillin-1. Remaining

TE D

fractions were then identified as low density or high density relative to the lipid raft fractions (Averaimo et al., 2016). Two-way ANOVA was used to test for significance between different fractions. All experiments were compared to their control group: 0 minutes, Fc treated, or LZRS.

EP

Since significance was compared to a single control Dunnett’s post hoc test was used.

AC C

Cell surface biotinylation

Cultures were labeled on ice with 0.5 mg/mL EZ Link Sulfo-NHS-SS-biotin (ThermoFisher Scientific) for 20 minutes (Chen et al., 2012). Excess biotin was quenched with 100 mM glycine and lysates were collected in USB lacking BME. Samples were immunoprecipitated using streptavidin (SA) conjugated beads (ThermoFisher Scientific) overnight at 4°C and eluted using Laemmli buffer. Samples were separated by SDS-PAGE.

16

ACCEPTED MANUSCRIPT

Scratch wound healing assay Confluent NHEKs were kept in 0.2 mM Ca2+ containing complete NHEK medium for 24 hours

RI PT

prior to scratch wound assays and the percentage of wound closure was calculated for each condition 24 hours after wounding as previously described (Kaplan et al., 2012). Following wounding, human Fc protein (Jackson ImmunoResearch, West Grove, PA) or ephrin-A1-Fc (R&D Systems Inc., Minneapolis, MN) were added to the 0.2 mM Ca2+ medium. No differences

SC

were seen in migration rates with 0.4 µg/mL mitomycin C (Millipore Sigma, Billerica, MA)

M AN U

pretreatment (data not shown). To test for significance, paired two-way ANOVA was used followed by a Bonferroni post hoc test when two groups were compared or a Tukey post hoc test when more than two groups were compared. Microscopy and image processing

NHEKs or optimal cutting temperature embedded RHE sections were fixed and permeabilized in

TE D

methanol. For the co-culture confrontation assay, silicone chambers with a 500 µm separation (Ibidi, Fitchburg, WI) were used to prevent intermixing of cell populations as previously

EP

described (Kaplan et al., 2018). On one side of the silicone chamber, ephrin-A1 overexpressing NHEKs were plated and on the other side wild type EphA2 overexpressing or Chimera 212

AC C

expressing NHEKs were plated in NHEK growth medium. After 24 hours, the silicone chamber was removed and medium was changed to 0.2 mM Ca2+ medium. After 48 hours, NHEKs were fixed and permeabilized in methanol. Images were captured using a Zeiss AxioImager Z.1 microscope with ApoTome (Carl

Zeiss AG, Oberkochen, Germany). Colocalization was analyzed using the Coloc2 plug-in on FIJI (Schindelin et al., 2012). Significance between multiple groups were compared using a paired one-way ANOVA was used followed by a Tukey post hoc test between all groups.

17

ACCEPTED MANUSCRIPT

RNA isolation and analysis Total RNA was collected from cells using a purification kit (RNeasy; Qiagen, Germantown, MD) and cDNA was generated using a reverse transcription kit (Suprscript III;

RI PT

ThermoFisher Scientific). Real-time qPCR was performed (LightCycler 96 System; Roche, Indianapolis, IN) using a quantitative FastStart SYBR green PCR kit (Roche). Primer sequence information is available in Supplementary Table 2. Significance was determined by a paired one-

SC

way ANOVA using a Tukey post hoc test between all groups.

M AN U

Statistics

Data are expressed as the mean ± Standard Deviation (SD) and significance was determined when p≤0.05. For all experiments n≥3 replicates. Statistics were performed with GraphPad Prism 7 (La Jolla, CA). For statistical analysis, t-test or ANOVA with a post-hoc test was used where appropriate.

TE D

CONFLICT OF INTEREST

No competing interests declared. SG is an employee and shareholder of GlaxoSmithKline, llc.

EP

(GSK). This article does not relate to work being done at GSK, nor does his status as an employee and shareholder unduly influence its contents.

AC C

ACKNOWLEDGEMENTS

This work was supported by the National Institutes of Health [AR062110 (SG); EY06769, EY017539, EY019463 (RML)], Cellular and molecular basis of disease training fellowship [T32-GM08061 (RV)], Cancer Smashers Foundation Award [RV], Research Career Development Award from the Dermatology Foundation [BPW], and Skin Disease Research Center Grant [P30-AR057216-01].

18

ACCEPTED MANUSCRIPT

We thank Ariel Finkielsztein, Ph.D., Andrea Brown, and Cynthia Ventrella for their contributions.

RI PT

REFERENCES

AC C

EP

TE D

M AN U

SC

Allen JA, Halverson-Tamboli RA, Rasenick MM. Lipid raft microdomains and neurotransmitter signalling. Nat Rev Neurosci 2007;8(2):128-40. Atapattu L, Saha N, Llerena C, Vail ME, Scott AM, Nikolov DB, et al. Antibodies binding the ADAM10 substrate recognition domain inhibit Eph function. J Cell Sci 2012;125(Pt 24):6084-93. Averaimo S, Assali A, Ros O, Couvet S, Zagar Y, Genescu I, et al. A plasma membrane microdomain compartmentalizes ephrin-generated cAMP signals to prune developing retinal axon arbors. Nat Commun 2016;7:12896. Barquilla A, Pasquale EB. Eph receptors and ephrins: therapeutic opportunities. Annu Rev Pharmacol Toxicol 2015;55:465-87. Beauchamp A, Lively MO, Mintz A, Gibo D, Wykosky J, Debinski W. EphrinA1 is released in three forms from cancer cells by matrix metalloproteases. Mol Cell Biol 2012;32(16):3253-64. Bocharov EV, Mayzel ML, Volynsky PE, Mineev KS, Tkach EN, Ermolyuk YS, et al. Lefthanded dimer of EphA2 transmembrane domain: Helix packing diversity among receptor tyrosine kinases. Biophys J 2010;98(5):881-9. Bocharov EV, Sharonov GV, Bocharova OV, Pavlov KV. Conformational transitions and interactions underlying the function of membrane embedded receptor protein kinases. Biochim Biophys Acta 2017. Calay D, Vind-Kezunovic D, Frankart A, Lambert S, Poumay Y, Gniadecki R. Inhibition of Akt signaling by exclusion from lipid rafts in normal and transformed epidermal keratinocytes. J Invest Dermatol 2010;130(4):1136-45. Campbell TN, Davy A, Liu Y, Arcellana-Panlilio M, Robbins SM. Distinct membrane compartmentalization and signaling of ephrin-A5 and ephrin-B1. Biochem Biophys Res Commun 2008;375(3):362-6. Carvalho RF, Beutler M, Marler KJ, Knoll B, Becker-Barroso E, Heintzmann R, et al. Silencing of EphA3 through a cis interaction with ephrinA5. Nat Neurosci 2006;9(3):322-30. Chen J, Nekrasova OE, Patel DM, Klessner JL, Godsel LM, Koetsier JL, et al. The C-terminal unique region of desmoglein 2 inhibits its internalization via tail-tail interactions. J Cell Biol 2012;199(4):699-711. Coon BG, Baeyens N, Han J, Budatha M, Ross TD, Fang JS, et al. Intramembrane binding of VE-cadherin to VEGFR2 and VEGFR3 assembles the endothelial mechanosensory complex. J Cell Biol 2015;208(7):975-86. Davy A, Gale NW, Murray EW, Klinghoffer RA, Soriano P, Feuerstein C, et al. Compartmentalized signaling by GPI-anchored ephrin-A5 requires the Fyn tyrosine kinase to regulate cellular adhesion. Genes Dev 1999;13(23):3125-35. Denning MF, Wang Y, Tibudan S, Alkan S, Nickoloff BJ, Qin JZ. Caspase activation and disruption of mitochondrial membrane potential during UV radiation-induced apoptosis of human keratinocytes requires activation of protein kinase C. Cell Death Differ 2002;9(1):40-52.

19

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

Diaz-Rohrer BB, Levental KR, Simons K, Levental I. Membrane raft association is a determinant of plasma membrane localization. Proc Natl Acad Sci U S A 2014;111(23):8500-5. Egea J, Klein R. Bidirectional Eph-ephrin signaling during axon guidance. Trends Cell Biol 2007;17(5):230-8. Falivelli G, Lisabeth EM, Rubio de la Torre E, Perez-Tenorio G, Tosato G, Salvucci O, et al. Attenuation of eph receptor kinase activation in cancer cells by coexpressed ephrin ligands. PLoS One 2013;8(11):e81445. Gauthier LR, Robbins SM. Ephrin signaling: One raft to rule them all? One raft to sort them? One raft to spread their call and in signaling bind them? Life Sci 2003;74(2-3):207-16. Giltaire S, Lambert S, Poumay Y. HB-EGF synthesis and release induced by cholesterol depletion of human epidermal keratinocytes is controlled by extracellular ATP and involves both p38 and ERK1/2 signaling pathways. J Cell Physiol 2011;226(6):1651-9. Gniadecki R, Bang B. Flotillas of lipid rafts in transit amplifying cell-like keratinocytes. J Invest Dermatol 2003;121(3):522-8. Gordon K, Kochkodan JJ, Blatt H, Lin SY, Kaplan N, Johnston A, et al. Alteration of the EphA2/Ephrin-A signaling axis in psoriatic epidermis. J Invest Dermatol 2013;133(3):712-22. Hennings H, Michael D, Cheng C, Steinert P, Holbrook K, Yuspa SH. Calcium regulation of growth and differentiation of mouse epidermal cells in culture. Cell 1980;19(1):245-54. Janes PW, Saha N, Barton WA, Kolev MV, Wimmer-Kleikamp SH, Nievergall E, et al. Adam meets Eph: an ADAM substrate recognition module acts as a molecular switch for ephrin cleavage in trans. Cell 2005;123(2):291-304. Jans R, Atanasova G, Jadot M, Poumay Y. Cholesterol depletion upregulates involucrin expression in epidermal keratinocytes through activation of p38. J Invest Dermatol 2004;123(3):564-73. Kao TJ, Kania A. Ephrin-mediated cis-attenuation of Eph receptor signaling is essential for spinal motor axon guidance. Neuron 2011;71(1):76-91. Kaplan N, Fatima A, Peng H, Bryar PJ, Lavker RM, Getsios S. EphA2/Ephrin-A1 signaling complexes restrict corneal epithelial cell migration. Invest Ophthalmol Vis Sci 2012;53(2):936-45. Kaplan N, Ventrella R, Peng H, Pal-Ghosh S, Arvanitis C, Rappoport JZ, et al. EphA2/EphrinA1 Mediate Corneal Epithelial Cell Compartmentalization via ADAM10 Regulation of EGFR Signaling. Invest Ophthalmol Vis Sci 2018;59(1):393-406. Lambert S, Ameels H, Gniadecki R, Herin M, Poumay Y. Internalization of EGF receptor following lipid rafts disruption in keratinocytes is delayed and dependent on p38 MAPK activation. J Cell Physiol 2008;217(3):834-45. Lin S, Gordon K, Kaplan N, Getsios S. Ligand targeting of EphA2 enhances keratinocyte adhesion and differentiation via desmoglein 1. Mol Biol Cell 2010;21(22):3902-14. Lorent JH, Diaz-Rohrer B, Lin X, Spring K, Gorfe AA, Levental KR, et al. Structural determinants and functional consequences of protein affinity for membrane rafts. Nat Commun 2017;8(1):1219. Mathay C, Giltaire S, Minner F, Bera E, Herin M, Poumay Y. Heparin-binding EGF-like growth factor is induced by disruption of lipid rafts and oxidative stress in keratinocytes and participates in the epidermal response to cutaneous wounds. J Invest Dermatol 2008;128(3):717-27.

20

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

McGuinn KP, Mahoney MG. Lipid rafts and detergent-resistant membranes in epithelial keratinocytes. Methods Mol Biol 2014;1195:133-44. Miao H, Gale NW, Guo H, Qian J, Petty A, Kaspar J, et al. EphA2 promotes infiltrative invasion of glioma stem cells in vivo through cross-talk with Akt and regulates stem cell properties. Oncogene 2015;34(5):558-67. Miao H, Li DQ, Mukherjee A, Guo H, Petty A, Cutter J, et al. EphA2 mediates ligand-dependent inhibition and ligand-independent promotion of cell migration and invasion via a reciprocal regulatory loop with Akt. Cancer Cell 2009;16(1):9-20. Munro S. Lipid rafts: elusive or illusive? Cell 2003;115(4):377-88. Murai T. Lipid Raft-Mediated Regulation of Hyaluronan-CD44 Interactions in Inflammation and Cancer. Front Immunol 2015;6:420. Nikolov DB, Xu K, Himanen JP. Eph/ephrin recognition and the role of Eph/ephrin clusters in signaling initiation. Biochim Biophys Acta 2013;1834(10):2160-5. Noberini R, Rubio de la Torre E, Pasquale EB. Profiling Eph receptor expression in cells and tissues: a targeted mass spectrometry approach. Cell Adh Migr 2012;6(2):102-12. O'Keefe EJ, Briggaman RA, Herman B. Calcium-induced assembly of adherens junctions in keratinocytes. J Cell Biol 1987;105(2):807-17. Overmiller AM, McGuinn KP, Roberts BJ, Cooper F, Brennan-Crispi DM, Deguchi T, et al. cSrc/Cav1-dependent activation of the EGFR by Dsg2. Oncotarget 2016;7(25):37536-55. Pasquale EB. Eph receptor signalling casts a wide net on cell behaviour. Nat Rev Mol Cell Biol 2005;6(6):462-75. Perez White BE, Ventrella R, Kaplan N, Cable CJ, Thomas PM, Getsios S. EphA2 proteomics in human keratinocytes reveals a novel association with afadin and epidermal tight junctions. J Cell Sci 2017;130(1):111-8. Pitulescu ME, Adams RH. Eph/ephrin molecules--a hub for signaling and endocytosis. Genes Dev 2010;24(22):2480-92. Rajendran L, Simons K. Lipid rafts and membrane dynamics. J Cell Sci 2005;118(Pt 6):1099102. Sabet O, Stockert R, Xouri G, Bruggemann Y, Stanoev A, Bastiaens PI. Ubiquitination switches EphA2 vesicular traffic from a continuous safeguard to a finite signalling mode. Nat Commun 2015;6:8047. Salaita K, Nair PM, Petit RS, Neve RM, Das D, Gray JW, et al. Restriction of receptor movement alters cellular response: physical force sensing by EphA2. Science 2010;327(5971):1380-5. Schindelin J, Arganda-Carreras I, Frise E, Kaynig V, Longair M, Pietzsch T, et al. Fiji: an opensource platform for biological-image analysis. Nat Methods 2012;9(7):676-82. Sharonov GV, Bocharov EV, Kolosov PM, Astapova MV, Arseniev AS, Feofanov AV. Point mutations in dimerization motifs of the transmembrane domain stabilize active or inactive state of the EphA2 receptor tyrosine kinase. J Biol Chem 2014;289(21):1495564. Simpson CL, Kojima S, Getsios S. RNA interference in keratinocytes and an organotypic model of human epidermis. Methods Mol Biol 2010;585:127-46. Song KS, Li S, Okamoto T, Quilliam LA, Sargiacomo M, Lisanti MP. Co-purification and direct interaction of Ras with caveolin, an integral membrane protein of caveolae microdomains. Detergent-free purification of caveolae microdomains. J Biol Chem 1996;271(16):9690-7.

21

ACCEPTED MANUSCRIPT

M AN U

SC

RI PT

Stahley SN, Saito M, Faundez V, Koval M, Mattheyses AL, Kowalczyk AP. Desmosome assembly and disassembly are membrane raft-dependent. PLoS One 2014;9(1):e87809. Ventrella R, Kaplan N, Getsios S. Asymmetry at cell-cell interfaces direct cell sorting, boundary formation, and tissue morphogenesis. Exp Cell Res 2017;358(1):58-64. Walsh R, Blumenberg M. Specific and shared targets of ephrin A signaling in epidermal keratinocytes. J Biol Chem 2011;286(11):9419-28. Wykosky J, Palma E, Gibo DM, Ringler S, Turner CP, Debinski W. Soluble monomeric EphrinA1 is released from tumor cells and is a functional ligand for the EphA2 receptor. Oncogene 2008;27(58):7260-73. Xu K, Yu FS. Impaired epithelial wound healing and EGFR signaling pathways in the corneas of diabetic rats. Invest Ophthalmol Vis Sci 2011;52(6):3301-8. Yin T, Getsios S, Caldelari R, Kowalczyk AP, Muller EJ, Jones JC, et al. Plakoglobin suppresses keratinocyte motility through both cell-cell adhesion-dependent and -independent mechanisms. Proc Natl Acad Sci U S A 2005;102(15):5420-5. Yin Y, Yamashita Y, Noda H, Okafuji T, Go MJ, Tanaka H. EphA receptor tyrosine kinases interact with co-expressed ephrin-A ligands in cis. Neurosci Res 2004;48(3):285-96. Yuan Z, Zhang F, Davis MJ, Boden M, Teasdale RD. Predicting the solvent accessibility of transmembrane residues from protein sequence. J Proteome Res 2006;5(5):1063-70.

FIGURE LEGENDS:

Figure 1. Ephrin-A1 and ligand activated EphA2 are concentrated outside of lipid raft

TE D

domains during calcium-induced cell-cell contact stabilization. (a) Sucrose density gradients of EphA2, ligand-induced active pY772 EphA2, EphA1, ephrin-A1, flotillin-1, and caveolin-1 in keratinocytes maintained in low Ca2+ (0.03 mM) or switched to high

EP

Ca2+ (1.2 mM) for 15 and 60 minutes. Densitometry is quantified for EphA2 (b), pY772

AC C

EphA2 (c), EphA1 (d), ephrin-A1 (e), flotillin-1 (f), and caveolin-1 (g) as low density, lipid raft, and high density fractions. Location of these fractions were determined by flotillin-1 and caveolin-1 for each time point. (WC=Whole cell lysate; Error bars represent SD; *p≤0.05, **p≤0.01; paired two-way ANOVA with Dunnett post hoc test

compared to 0 minutes, n=4). Figure 2. Decreased colocalization of caveolin-1 with ligand activated EphA2 during calcium-induced cell-cell contact stabilization. (a) Representative pY772 EphA2,

22

ACCEPTED MANUSCRIPT

EphA2, and caveolin-1 immunofluorescent images of NHEKs maintained in low Ca2+ (0.03 mM) and then switched into high Ca2+ (1.2 mM) for the indicated time points. DAPI was used to stain nuclei. (Scale bar=20 µm). Colocalization analysis of pY772

RI PT

EphA2 in caveolin-1 and EphA2 in caveolin-1 is quantified as Mander’s coefficients in (b) and (c), respectively. (Error bars represent SD; ***p≤0.001; paired one-way ANOVA

SC

with Tukey post hoc test, n=4).

Figure 3. The transmembrane domain of EphA2 is required for cell-cell contact

M AN U

localization. (a) Immunofluorescent images of EphA1 and EphA2 in control (pLKO) and EphA2 knockdown (KD; shEphA2.pLKO) NHEKs with stable cell-cell contacts formed in 0.2 mM Ca2+ for 24 hours. DAPI was used to stain nuclei. (Scale bar=50 µm). (b) Diagram depicting Chimera 212 that contains the extracellular domain from EphA2 (amino acids 1-522), transmembrane domain from EphA1 (amino acids 544-572), and the

TE D

cytoplasmic domain from EphA2 (amino acids 564-976). The differences in the predicted amino acid sequence of the transmembrane domain between EphA1 and EphA2 are shown below. (c) Immunofluorescent images of wild type EphA2 and Chimera 212

EP

expressed in EphA2 KD NHEKs after maintained in 0.2 mM Ca2+ for 24 hours. DAPI

AC C

was used to stain nuclei. (Scale bar=50 µm). (d) Representative EphA2 and E-cadherin immunofluorescent images of reconstituted human epidermis (RHE) overexpressing EphA2 or Chimera 212 in EphA2 KD NHEKs 9 days after being lifted to an air-liquid interface. DAPI was used to stain nuclei. (Scale bar=100 µm).

Figure 4. Chimera 212 alters pY772 EphA2 and ephrin-A1 lipid raft distribution. (a) Sucrose density gradients for EphA2 in NHEKs expressing control vector (LZRS), EphA1, EphA2, or Chimera 212 after being maintained in 0.2 mM Ca2+ medium for 24

23

ACCEPTED MANUSCRIPT

hours. The percentage of EphA2 in each fraction, as defined by flotillin-1 and caveolin-1, is quantified in (b). (c-e) Quantification of sucrose density gradients for EphA2, pY772 EphA2, and ephrin-A1 in keratinocytes lacking endogenous EphA2 (shEphA2) but

RI PT

expressing ectopic wild type EphA2 or Chimera 212. NHEKs were maintained in 0.2 mM Ca2+ medium for 24 hours and each fraction was defined by flotillin-1 or caveolin-1 (Error bars represent SD; *p≤0.05, **p≤0.01; (b, c-e) paired two-way ANOVA with

SC

Dunnett post hoc test compared to LZRS (b) or Bonferroni post hoc test (c-e), n=3).

M AN U

Figure 5. Loss of the EphA2 TMD leads to an accumulation of ephrin-A1. (a) Western blot analysis of EphA1, EphA2, and ephrin-A1 in NHEKs that have been maintained in 0.2 mM Ca2+ for 48 hours and express LZRS, EphA1, EphA2, or Chimera 212 vectors. Processed ephrin-A1 fragments are indicated by arrowheads. GAPDH was used as a loading control. The expression level of ephrin-A1 is quantified in (b). (c) q-RT-PCR

TE D

analysis of ephrin-A1 mRNA levels in NHEKs expressing LZRS, EphA1, EphA2, or Chimera 212 after 48 hours in 0.2 mM Ca2+ medium. (d, e) Immunofluorescent images of the confrontation area between ephrin-A1 overexpressing NHEKs and EphA2

EP

overexpressing (d) or Chimera 212 expressing NHEKs (e) 48 hours after silicone

AC C

chamber was removed and medium was changed to in 0.2 mM Ca2+. DAPI was used to stain nuclei and white dashed line indicate confrontation area of the two cell populations. (Scale bar=20 µm). (Error bars represent SD; *p≤0.05, **p≤0.01; paired one-way

ANOVA with Tukey post hoc test, n=3-4).

Figure 6. The EphA2 TMD is required for efficient keratinocyte migration through regulation of ephrin-A1. (a) Scratch wound assays were performed in keratinocytes expressing LZRS, EphA1, EphA2, or Chimera 212 in 0.2 mM Ca2+ for 24 hours.

24

ACCEPTED MANUSCRIPT

Keratinocytes were then treated with Fc or ephrin-A1-Fc (EfnA1-Fc) recombinant protein (1.0 µg/mL) in 0.2 mM Ca2+ and allowed to migrate into the wound area for 24 hours. Percent wound closure was then calculated. (b) Western blot analysis of EphA2 and

RI PT

ephrin-A1 in keratinocytes that have been treated with control or ephrin-A1 siRNA. GAPDH was used as a loading control. (c) Keratinocytes expressing LZRS or Chimera 212 were treated with control or ephrin-A1 siRNA for knockdown and subsequently

SC

allowed to migrate for 24 hours after generation of a linear scratch wound and percent wound closure was then calculated. (Error bars represent SD; ns=not significant,

M AN U

*p≤0.05, **p≤0.01, ****p≤0.0001; paired two-way ANOVA with Bonferroni or Tukey post hoc test for comparison between two groups or more than two groups, respectively,

AC C

EP

TE D

n=4).

25

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT