5-(N-Ethyl-N-isopropyl)-Amiloride

Sodium hydrogen exchanger and phospholipase D are required for a1-adrenergic receptor stimulation of metalloproteinase-9 and cellular invasion in CCL39 ﬁbroblasts

Abstract

Matrix metalloproteinase 9 (MMP-9) plays a critical role in digesting the extracellular matrix and has a vital function in tumor metastasis and invasion; this protease activity is signiﬁcantly increased in non- small cell lung cancers. The sodium hydrogen exchanger isoform 1 (NHE1) functions as a focal point for signal coordination and cytoskeletal reorganization. NHE1 is thought to play a central role in estab- lishing signaling components at the leading edge of a migrating cell. Therefore, we studied the relation- ship between NHE1 and MMP-9 activity in Chinese hamster lung ﬁbroblasts (CCL39) stimulated with phenylephrine (PE). We show that PE increases MMP-9 gelatinolytic activity in CCL39 cells. The inhibition of phospholipase D (PLD) signaling abrogated PE-induced MMP-9 activity. The role of PLD as an essential signaling intermediate was conﬁrmed when the addition of permeable phosphatidic acid increased MMP-9 activity in the same cells. PE-induced invasion was increased 1.9-fold over controls and the PE response was lost when 1-butanol was used to block PLD signaling. Cells pre-treated with the NHE1 inhibitor, 5-(N-ethyl-N-isopropyl) amiloride (EIPA) prior to PE addition resulted in a notable decrease in MMP-9 activation and cell invasion as compared to untreated PE-stimulated cells. CCL39 NHE1 null cells demonstrated no increase in MMP-9 protease activity or cell invasion in response to PE treatment. Reconstitution of NHE1 expression recovered the PE-induced activation of protease activity and cell inva- sion. MMP-9 processing was altered in cells expressing a proton transport defective NHE1 but retained the ability to respond to PE. Conversely, cells expressing an ezrin, radixin, moesin (ERM)-binding deﬁ- cient NHE1 had a lower MMP-9 activity and the protease did not respond to PE addition. Parallel studies on NCI-H358 non-small cell lung cancer (NSCL) cells showed that PE stimulated both MMP-9 activity and cell invasion in an NHE1 dependent manner. This work describes for the ﬁrst time a PE-induced relation- ship between NHE1 and MMP-9 and a new potential mechanism by which NHE1 could promote tumor formation and metastasis.

The coordination of tumor invasion and migration has recently been associated with the ion transport activity of the sodium hydrogen exchanger isoform 1 (NHE1) [11–17]. In addition to reg- ulation of intracellular and extracellular pH, NHE1 is found at the leading edge of migrating cells where NHE1 acts as a plasma mem- brane anchor point for cytoskeleton proteins and recruits proteins necessary for cell invasion [18]. Cells lacking NHE1 display poorly coordinated migration and commonly move indiscriminately [12]. NHE1 activation in tumor cells commonly leads to higher intracel- lular pH levels and substantially lower extracellular pH levels. In human breast carcinoma, serum deprivation decreases extracellu- lar pH and increased cell invasion [19,20]. This decreased extracel- lular pH is thought to be critical to the activation of extracellular proteases and the degradation of the extracellular matrix within tumors [16,21]. In the early stages of cell migration, invasion and metastasis this degradation is primarily due to the actions of ma- trix metalloproteinases 2 and 9 (MMP-2 and 9).1 MMP-2 and MMP-9 are responsible for hydrolysis of both extracellular matrix and basement membrane proteins (see reviews [22–24]. MMP-9 is synthesized as 92 kDa pro-enzymes and is secreted before conver- sion to the fully active 82 kDa form. Increased MMP-9 expression and activity have been correlated with increased tumor aggressive- ness and lower survival rates in several cancer types [23,25,26]. MMP-9 levels were found to be signiﬁcantly higher in patients with NSLC than the control subjects [23,25,26]. Consistent with these results, antisense treatment of MMP-9 in nude mice de- creased the metastatic capacity of NSLC tumors [27]. While the activation of MMP-9 is complex and not well deﬁned, a potential role for NHE1 in MMP-9 activation has been identiﬁed. NHE1 and MMP-9 activities are co-localized in invadopodia in breast cancer cells [14,28]. Additionally, in Ltk-mouse muscle ﬁbroblasts expressing proton transport deﬁcient NHE1 cells were nearly de- void of MMP-9 expression or activity [29].

While the signaling pathways involved in the combined regulation of NHE1 and MMP-9 have not been fully characterized, MMP-9 activity has been found downstream of phospholipase D (PLD) [7,30–32]. We have recently found that stimulation of the a1AR in CCL39 lung ﬁbroblasts requires both PLD and ERK activation for stimulation of NHE1 proton transport and cell migration [7,33]. In this study we describe a novel role for a1AR in cell inva- sion in a pathway involving PLD, cumulating in NHE1 and MMP-9 activation. Evidence is presented which demonstrates that the expression and proton transport activity of NHE1 is critical for MMP-9 function and cellular invasion in CCL39 ﬁbroblasts. Finally, we transition this research from CCL39 cells to a human lung can- cer cell line. Using the human NSCLC cell line NCI-H358 we demon- strate that PE stimulation increases MMP-9 activation and cancer cell invasion in an NHE1 dependent manner.

Experimental procedures

Materials

Reagents: 5-N(Ethyl)-isoproplyamiloride (EIPA), phenylephrine (PE), 4-amin- ophenylmercuric acetate (APMA), Dulbecco’s modiﬁed Eagle’s medium with high glucose were purchased from Sigma Chemical Co, St. Louis, MO. Short chain (dihexanoyl)-phosphatidic acid (sc-PA) was from Avanti Polar Lipids, Alabaster, AL. The MMP-2/MMP-9 inhibitor SB-3CT 3-(4-phenoxyphenylsulfonyl)-propylthi- rane, anti-MMP-9 mouse monoclonal antibody and recombinant MMP-9 were pur- chased from Chemicon. Centricon YM10 ﬁltration devices were from Millipore, Billerica, MA. The BD BioCoat Matrigel invasion chamber was purchased from BD Biosciences, Bedford, MA.

Cell culture

CCL39 (American Type Culture Collection, Manassas VA) and PS120 cells (NHE11-null cells derived from CCL39 ﬁbroblasts—a gift from J. Pouyssegur, Univer- sity of Nice) were maintained in high glucose DMEM plus 10% heat-inactivated fetal bovine serum with 100 units/ml penicillin, 100 lg/ml streptomycin. PS120 cells stably expressing wild-type NHE1 (PSN), mutated NHE1 with impaired ERM bind- ing (KR/A) or proton transport defective NHE1 (E266I) cells, both kind gifts from Diane Barber University of California, San Francisco, CA, were maintained in high glucose DMEM medium containing 10% heat-inactivated fetal bovine serum, 100 units/ml penicillin, 100 lg/ml streptomycin, and 0.4 mg/ml G418 (Calbiochem San Diego, CA). NCI-H358 human non-small cell lung carcinoma cells (ATCC) were maintained in RPMI 1640 containing 10% heat-inactivated fetal bovine serum, 100 units/ml penicillin, 100 lg/ml streptomycin. All cells were maintained at 37 °C in a 5% CO2 incubator. Where indicated, short chain cell-permeable phospha acid, (sc-PA) was prepared by drying a chloroform sc-PA mixture and resus- pending the lipid in phosphate buffered saline by vortexing and sonication.

Gelatin zymography

One milliliter of conditioned medium from CCL39 ﬁbroblasts was analyzed for MMP-9 activity by zymography after 20-fold concentration by lyophilization. The lyophilized samples were re-suspended to 50 ll in loading buffer (0.125 M Tris– Cl pH 7.2, 5% SDS, 40% sucrose, and 0.1% bromophenol blue). Lyopholized samples from E266I cell culture were re-suspended to 25 ll of loading buffer. MMP-9 activ-
ity from H358 cells were also subjected to zymography. These samples were ﬁrst concentrated 80-fold by ultraﬁltration using a YM-10 centricon membrane (Milli- pore). Following concentration, protein samples were resolved by 10% SDS–PAGE in the presence of 1 mg/ml gelatin (Biorad, Hercules, CA). The resulting gels were washed in 2% Triton-X100 for 60 min and then incubated for 24 h at 37 °C in zymo- gen gel buffer (0.1 M Tris–Cl pH 8.0, 1 mM CaCl2, and 2% Triton-X100). The gels were then stained using Coomassie brilliant blue R-250, gelatinase activity was identiﬁed as clear bands. Band intensities were analyzed using Kodak Image Station EDAS 290 and the Kodak Image Station 1.62 software.

SDS–PAGE and Western blot analysis

MMP-9 expression was determined by Immunoblot. Cells were cultured and prepared as above. Conditioned medium was collected and concentrated by ultra- ﬁltration (Millipore YM-10 Centriprep) and 10 lg of concentrated medium was re- solved by 10% SDS–PAGE. Proteins were transferred to polyvinylidene diﬂuoride (PVDF) membrane using a semi-wet electroblotting apparatus. MMP-9 was de- tected by immunoblotting the membrane using a 1:2000 dilution of rabbit anti- rat MMP-9 polyclonal antibody (Chemicon International) in TTBS (10 lM Tris and 150 mM NaCl, pH 8.0 with 0.05% Tween 20) with 5% dry milk overnight at 4 °C.

Bound antibodies were detected using a secondary antibody, anti-rabbit horserad- ish peroxidase-linked IgG (Cell Signaling Technologies), diluted at 1:10,000 in TTBS with 5% dry milk. MMP-9 protein was detected using an enhanced chemilumines- cence detection reagent (Santa Cruz Biotechnology).

PLD assay

In vivo PLD activity assay was performed using the transphosphatidylation as- say as described previously by Park et. al. [34]. Brieﬂy, cells were serum deprived and labeled with 2 lCi/ml [9,10-3H] myristic acid for 16–18 h prior to agonist stim- ulation. Unincorporated [3H] myristic acid was removed by washing with PBS and the cells were incubated in serum free high glucose DMEM medium for 1 h. For the
ﬁnal 15 min. of pre-incubation, 0.3% 1-butanol was included. After agonist activa- tion, the cells were scraped, lipids extracted using an acidiﬁed chloroform methanol extraction and resolved by TLC [34]. Standards for phosphatidic acid (PA) and phos- phatidylbutanol (PtdBut) were visualized with iodine and bands co-migrating with the authentic standards were scraped into scintillation vials. PLD activity was deter- mined as the incorporation of radioactivity into PtdBut [34].

Transwell invasion assay

Invasion was quantiﬁed using the BD BioCoat Matrigel invasion chamber (BD Bioscience, Bedford, MA) according to the manufacturer’s instructions. Serum- starved CCL39 or H358 cells were incubated at 37 °C for 30 min in 10 mM Cell Tracker Green Dye (Chemicon). A total of 1 × 106 cells in serum free medium were seeded onto Matrigel-coated ﬁlters. Agonists or inhibitors were added to both the upper and lower chamber as indicated. Cells were allowed to migrate/invade for 18 h. Cells that had invaded through the ﬁlter were counted under magniﬁcation (randomly selected high-power ﬁelds). The counting was performed in ﬁve random ﬁelds for each sample, and mean values from three or more independent experi- ments were used.

Results

Mature active form of MMP-9 is expressed in phenylephrine stimulated cells

In CCL39 cells, a1AR activation increased cell motility as mea- sured by decrease wound closure times in a standard wounding as- say. This led us to investigate whether the speciﬁc agonist PE increased MMP-9 activity in these cells. We measured MMP-9 activity using gelatin zymography in medium conditioned by CCL39 cells in the presence of 50 lM PE (Fig. 1). Addition of 50 lM PE increased the amount of MMP-9 activity 2.3-fold compared to unstimulated cells at 24 h. These bands co-migrated with the mature 82 kDa active MMP-9 isoform of partially digested recombinant MMP-9 (Fig. 1A). Activity from the 92 kDa pro-form of MMP-9 was virtually undetectable in the conditioned medium from unstimulated or PE treated cells (Fig. 1A). Concentration of the medium or longer gel incubation reveled only small amounts of the 92 kDa band (data not shown). To further identify the cleared band on the zymogen gel as MMP-9 medium and lysates were subjected to Western blot analysis. MMP-9 immunoanalysis of PE conditioned medium using MMP-9 antibodies, identiﬁed a major band at 82 kDa. This correlates to the smaller molecular weight generated when recombinant pre-MMP-9 was incubated with APMA and is consistent with the fully active form of MMP-9 (Fig. 1C). The concentration of MMP-9 in control conditioned medium was too low to be detected by the MMP-9 Ab (data not shown). Due to the lack of pro-MMP-9 in the conditioned medium, we also examined the lysates of control and stimulated cells for expression of MMP-9 (Fig. 1D). In both stimulated and control ly- sates, the Western blot showed that the major form of MMP-9 is the 92 kDa, pro-MMP-9 and that the expression level did not change with addition of PE (Fig. 1D). These data suggest that CCL39 ﬁbroblasts primarily secrete the mature, 82 kDa MMP-9 iso- form and that PE can induce the activation of the mature form of MMP-9 (Fig. 2).

Phenylephrine increases phosphatidylbutanol formation in CCL39 cells

A number of potential upstream signaling regulators of MMP-9 have been linked to PLD activity in different cell types using vari- ous agonists [7,32]. In our laboratory, we have recently shown that incubating cells with PE (15 min.) stimulates PLD activity and that inhibition of PLD signiﬁcantly reduces the cell motility of CCL39 cells [7]. To identify PLD as an important signaling component of cell migration and invasion, we determined if chronic a1AR stimu- lation resulted in the activation of PLD in a time-dependent man- ner consistent with the MMP-9 activation observed previously. The time course of PLD activation showed a signiﬁcant increase at 10 min with maximal activation being achived by 30 min. This maximal level of PLD hydrolysis continued throughout the 24 h
PE stimulation (9237 ± 639 DPM; p < 0.001 compared to zero time point). This data establishes that chronic stimulation of the a1AR allows for a long-term stimulation of PLD in contrast to other sig- naling intermediates such as MAPK, where the maximal stimula- tion by PE occurs at 10 min and then sharply declines by 30 min of agonist stimulation [33]. Phenylephrine mediated MMP-9 activation and cellular invasion require PLD activity Because the PKC-dependent activation of MMP-9 in ﬁbrosar- coma cells requires PLD activity [32], we wished to establish a rela- tionship between PLD and MMP-9 activity in a1AR stimulated cells. To do this, we ﬁrst used a primary alcohol, 1-butanol to block the formation of phosphatidic acid. As shown in Fig. 3, treatment with 0.5% 1-butanol prior to PE addition signiﬁcantly decreased PE-in- duced MMP-9 activity (p < 0.05) (Fig. 3). This indicated that either the production of phosphatidic acid catalyzed by PLD or one of the lipid’s metabolic products is required to stimulate MMP-9 activity. To further establish a role for PLD in cell migration, we tested the ability of CCL39 ﬁbroblasts to migrate through a collagen matrix using a BD BioCoat Matrigel Invasion Chamber. Both 1-butanol and sc-PA were used to examine if PLD was involved in inducing the invasiveness of these cells. We have found that cell permeable sc-PA can be used to mimic exogenously added PA, the product of PLD activity. As shown in Fig. 4, 50 lM PE stimulated the number of cells invading through the collagen 1.9-fold (p > 0.001) com- pared to unstimulated cells. Addition of 40 lM sc-PA increased the number of invading cells 1.8-fold over controls (p > 0.01). This concentration of sc-PA is similar to the effective concentration determined in earlier studies in our laboratory [7]. Conversely, when PLD was inhibited by the addition of 1-butanol, the differ- ence in the ability of PE to induce invasion was found to be signif- icantly decreased (p > 0.001) when compared to PE stimulation alone. Our results suggest that a1ARs stimulation by PE can acti- vate cell invasion. These results indicate that PLD activity is re- quired for MMP-9 stimulation by PE and that a1AR activation leads to an increased invasiveness of CCL39 ﬁbroblasts. The hall- mark of MMP-9 activity is the degradation of the extracellular ma- trix, thus these results suggest that PE addition stimulates a PLD activity crucial in mediating the effects of MMP-9 enhanced cell invasion.

NHE1 mediates phenylephrine activation of MMP-9 and cell invasion

Recently we have shown the role of PLD in the regulation of cell migration through the ERK signaling pathway in CCL39 cells. Like- wise ERK activation by PE regulates the activation of NHE1 in the same cell line [7,33,35]. A link between NHE1 and cancer metasta- sis has been seen in breast cancer cells where NHE1 acidiﬁes the extracellular microenvironment leading to an increase in breast tu- mor invasion [19]. Finally, the activity of MMP-9 in mouse muscle ﬁbroblasts lacking NHE1 expression is signiﬁcantly diminished [29]. Thus, we next investigated a potential role for NHE1 in regu- lating both MMP-9 activity and cell invasion through matrigel. In an experiment similar to (Fig. 4), conditioned medium was collected from cells stimulated with PE in the presence of 50 lM EIPA (a speciﬁc amiloride analogue which acts by inhibiting H+ trans- port) and analyzed by gelatin zymography. Exposure of cells to 50 lM EIPA abrogated the ability of PE to stimulate MMP-9 activity (Fig. 5A). To ensure that MMP-9 activity wasn’t inhibited by the NHE1 inhibitor, 50 lM EIPA was incubated with recombinant MMP-9 before or during gel zymography. In these studies no effect on MMP-9 gelatinolysis was observed (data not shown). To further investigate the speciﬁc nature of NHE1 on MMP-9 regulation, PS120 cells were used. PS120 cells (NHE1 deﬁcient cells derived from CCL38 ﬁbroblasts [36] were found to have a signiﬁcantly re- duced basal level of MMP-9 activity in conditioned medium and did not show a response to addition of 50 lM PE (Fig. 5B). Longer incubation of the zymogen gel or a 150-fold concentration of the conditioned PE-stimulated medium did not result in a detectable level of MMP-9 activity or show MMP-9 by immunoanalysis (data not shown). Reconstitution of wild-type NHE1 expression in PS120 cells (PSN cells) resulted in a recovery of basal level MMP-9 activity (Fig. 6). The MMP-9 activity in these cells was enhanced 2.1 ± 0.02 fold (p < 0.01) upon PE stimulation when compared to the unstim- ulated control PSN cells. The importance of proton transport and extracellular pH on MMP-9 activation was examined by measuring MMP-9 activity in cells expressing a proton defective NHE1. We found that, like PSN cells, expression of the proton deﬁcient NHE1 in PS120 cells (E266I) reconstituted background MMP-9 activity. This further demonstrated that PS120 cells are capable of MMP-9 expression when NHE1 activity is reconstituted and that the loss of MMP-9 activity in PS120 cells was not a non-speciﬁc re- sult of the generation of the NHE1 null cell line. More importantly, the loss of proton transport by NHE1 in E266I cells, resulted in the presence of a second, higher molecular weight band identiﬁed after zymography corresponding as pro-MMP-9. This indicates a role for extracellular acidiﬁcation in the complete processing of MMP-9. a1-adrenergic receptor stimulation of E266I cells increased MMP-9 activity 1.8 ± 0.05 (p < 0.05) fold over control. This increase in activity was accompanied by a signiﬁcant reduction in the amount of pro-MMP-9 found in the PE-stimulated conditioned medium (Fig. 6). We then examined the effect of ERM binding to NHE1 on MMP-9 activation using KR/A cells (PS120 cells expressing NHE1 which does not bind ERM). In serum-starved KR/A cells, addition of PE did not signiﬁcantly change MMP-9 activation (0.9 ± 0.12 fold of control, p = 0.21). However, the overall level of MMP-9 activity in these cells was lower than the activity found in either PSN or E266I cells. This was demonstrated by the fact that the cultured medium from E266I cells had to be concentrated twice that of the other cells to detect signiﬁcant protease activity (Fig. 6). To establish the interplay between NHE1 and cellular invasion, CCL39 cells were cultured on matrigel-coated invasion plates and treated with either the MMP-2/MMP-9 inhibitor SB-3CT or EIPA prior to stimulation with 50 lM PE (Fig. 7). As expected, the addition of PE increased the number of cells invading through matrigel 1.9 ± 0.21 fold over control (p < 0.01). Treatment of cells with either SB-3CT or EIPA signiﬁcantly inhibited PE-stimulated cellular invasion as compared to control levels (3.5 ± 0.16 and 2.1 ± 0.20 fold lower than PE treated cells, p < 0.01). These observations strongly support the conclusion that NHE1 expression, cytoskeletal binding via ERM proteins and proton transport of NHE1 are each crucial for the activation and processing of MMP-9. This work also shows that the proper regulation of MMP-9 via NHE1 is critical for the invasion capacity of CCL39 ﬁbroblasts. NHE1 activity is important for the function of MMP-9 and invasion in human non-small lung cancer cells The human NSCLC cell line, H358, was isolated from a primary bronchoalveolar carcinoma. This cell line shows strong colony forming efﬁciency in soft agarose, aggressively forms tumors in athymic nude mice, and is a potent xenobiotic metabolizer. MMP-9 activity has been implicated in the ability of H358 cells to respond to growth factor signaling that enhances cell invasion through a collagen matrix. Several proteases including MMP-9 have been suggested to be good targets for therapy of this chemo- therapy resistant cancer [37,38]. For these reasons we examined the effect of EIPA on PE-stimulated MMP-9 activity and cellular invasion in H358 carcinoma cells. Activity of both the latent MMP-9 and the 92 kDa form of MMP-9 were identiﬁed in resting H358 cells (Fig. 8A). The activity of both forms of MMP-9 increased after 24 h incubation with 50 lM PE (2.2 ± 0.14 fold over control, p < 0.01). Treatment with 10 lM EIPA suppressed the ability of PE to increase MMP-9 activity. After a 12 h incubation with 50 lM PE, we found a 1.8 ± 0.31 fold increase in cells invading through the matrigel layer as compared to control, p < 0.01. As seen with the CCL39 ﬁbroblasts, treatment with either the MMP-2/ MMP-9 inhibitor SB-3CT or EIPA, blocked the PE-stimulated invasion (3.38 ± 0.24 and 1.95 ± 0.42 fold decrease from PE alone for SB-3CT + PE and EIPA + PE treated cells, respectively p < 0.01). Discussion Cellular invasion and migration is a complex process involving the coordination of intra-and extracellular events. The extracellu- lar pH of tumor cells is often lower than that of normal tissues, indicating a potential and critical role for NHE1 in neoplastic pro- gression. Activation by serum depletion results in increased proton exchange mediated by NHE1 and enhanced motility and invasion of human breast cancer cells [19–21]. However, coordination of cellular motility is more complex than simple extracellular acidiﬁ- cation. NHE1 plays a role in regulating the intracellular dynamics of the cytoskeleton as well as the formation of invasive structures called invadopodia. In these invadopodia, an increase in extracellu- lar conditions leads to enhanced cell motility and proteolytic deg- radation of the extracellular matrix [14]. Increase extracellular protease activity has been associated with NHE1 and cell migration in several cell lines [19,39]. Earlier work published by Barber’s lab- oratory showed that loss of NHE1 expression resulted in a signiﬁ- cant decrease in MMP-9 activity [29]. Our work in CCL39 ﬁbroblasts has implicated NHE1 in a1AR mediated cell signaling and migration involving PLD [7]. Our ﬁndings here, establish a role for phenylephrine stimulation in cellular invasion via PLD, NHE1 and MMP-9. There are a number of studies that have examined the involve- ment of cellular migration and invasion by MMP-9, however, its regulation is diverse and may involve extracellular acidiﬁcation [22,23]. Our earlier work has shown that activation of the a1AR by PE produces an increase in intracellular pH through NHE1, acti- vates the ERK growth factor pathway, stimulates PLD activity, in- duces the formation of stress ﬁbers and enhances the rate of cellular migration in a wound assay [7,33]. Because each of these signaling factors have been implicated in the control of MMP-9 activity, we ﬁrst determined that MMP-9 was present in CCL39 ﬁbroblasts and if its activity was stimulated be PE. Although the activity found in the medium was increased by a1AR stimulation, no detectable activity was found in cell lysates or membranes (data not shown). Unexpectedly only the lower molecular weight 82 kDa form of MMP-9 was observed in the zymogen gels (Fig. 1). Immunoanalysis of the conditioned medium conﬁrmed that only the ma- ture active form of MMP-9 was present. However, only the pre–pro MMP-9 intermediate isoform was identiﬁed in cell lysates. Fur- thermore, we found that the expression proﬁle of each MMP-9 form was unchanged in lysates with or without addition of PE (Fig. 1D). This indicates that changes in MMP-9 are likely to be due to processing of small amounts of the cell-bound precursor MMP-9. MMP-9 processing includes C-terminal hydrolysis and interactions with regulators including tissue inhibitor of metallo- proteinases (TIMP). The lack of immature, higher molecular weight MMP-9 often found in conditioned medium may indicate a deﬁ- ciency of inhibitors of MMP-9 processing. While not examined, one possible reason is a lack of TIMP binding partners or other, similar MMP-9 regulators. A high TIMP to MMP ratio results in a delay of activation of the protease and decreases the processing of pro-MMP-9 into mature active form of MMP-9 [40]. Kinetic examination of MMP-9 activation in the absence of TIMP has been suggested to occur immediately instead of the latent activation ob- served in the presence of TIMP [40,41]. The importance of MMP-9 in cell motility has been demon- strated in many cell types by numerous agonists, however there are only a few examples of a1AR mediated MMP-9 activity. One example is the chronic activation of a1AR in cardiac ﬁbroblasts which results in tissue remodeling and MMP-9 expression [42]. Another is a Gq coupled activation of MMP-9 activity in pituitary gonadotrope cells [43]. Thus the data presented here showing acti- vation of MMP-9 by PE suggest a novel role for PE stimulation of a1AR in ﬁbroblasts. For some time PLD has been believed to be involved in cell migration and invasion [44,45]. The lipase was found to mediate MMP-9 secretion in both aggressive HT 1080 ﬁbrosarcoma cells [32,46] and in colon cancer cells [47]. Manipulation of extracellular acidic conditions in metastatic melanoma cells, resulted in en- hanced PLD catalysis, while inhibition of PLD activity in the same cells blocked the secretion of MMP-9 [48]. In CCL39 cells inhibition of PLD signaling resulted in a decrease in motility of the ﬁbroblasts moving into a vacated area [7]. Therefore, we next investigated whether PLD was responsible for MMP-9 activation and cell invasion in CCL39 cells. Our studies demonstrated that the a1AR-mediated activation of PLD was sustainable over the same time course found for MMP-9 activation (Fig. 1) and was consistent with the motility studies already published [7]. Either addition of exoge- nous PA or blocking PLD formation of PA with 1-butanol further demonstrates that PLD is involved in mediating PE activation of MMP-9 and cell invasion. NHE1 has two critical functions in cell motility, to act as a membrane anchor for cytoskeletal and signaling proteins and extracellular acidiﬁcation by proton transport. In migrating cells, NHE1 is localized to the leading edge of the cell where it facili- tates directed cell movement [49]. Inhibition of NHE1 proton transport by EIPA resulted in the abrogation of PE-stimulated MMP-9 activity (Fig. 5). A similar loss of MMP-9 activation by a1AR signaling was identiﬁed in parallel studies with NHE1 null PS120 cells. These results are in agreement with work conducted by the Barber laboratory, where analysis of NHE1 nulls muscle cells displayed a near total loss of MMP-9 activity [29]. In our experiments, we found that reconstitution of NHE1 expression recovered both basal and PE-stimulated MMP-9 activity. Inter- estingly, mutation of either the proton transport function or the ability of NHE1 to dock with ERM proteins resulted in slightly different MMP-9 activation. While retaining the ability to increase the MMP-9 activity, E2661 expressing cells (transport deﬁcient) did not fully process the 92 kDa pro-MMP-9 form, yet expression of a functional transporter with diminished ability to bind to the cytoskeleton (KR/A cells) resulted in fully processed MMP-9 (Fig. 6). However the MMP-9 activity in these cells did not respond to a1AR stimulation. Therefore, the critical processing and activation of MMP-9 must rely on both functions of NHE1. We have also shown that both MMP-9 activation and NHE1 pro- ton transport are critical for PE-mediated cellular invasion. Inhibi- tion of either MMP-9 or NHE1 blocked PE-mediated invasion to levels near control cells (Fig. 7). This led us to ask if the model of NHE1 as a crucial target against cell invasion would transfer to non-small cell lung cancer cells. Non-small cell lung cancer cells are highly chemoresistant and form very aggressive tumors. Thus NHE1 would provide a novel modality for the treatment of these lung cancer cells. As found with most cancer cells, the MMP-9 basal levels were elevated and addition of PE-stimulated the activity of pro-MMP-9. The MMP-9 activity and the ability of PE to induce cell invasion was signiﬁcantly inhibited when NHE1 was inhibited by addition of EIPA (Fig. 8). Our results demonstrate that the proper processing and activation of MMP-9 by NHE1 is important in the invasive potential of a lung cancer cell type that is refractory 5-(N-Ethyl-N-isopropyl)-Amiloride to chemotherapy-resistant tumors.