Vesicular nucleotide transporter mediates ATP release and migration in neutrophils

Neutrophils migrate to sites infected by pathogenic microorganisms. This migration is regulated by neutrophil-secreted ATP, which stimulates neutrophils in an autocrine manner through purinergic receptors on the plasma membrane. Although previous studies have shown that ATP is released through channels at the plasma membrane of the neutrophil, it remains unknown whether it is also released through alternate secretory systems involving vesicular mechanisms. In this study, we investigated the possible involvement of vesicular nucleotide transporter (VNUT), a key molecule for vesicular storage and nucleotide release, in ATP secretion from neutrophils. RT-PCR and Western blotting analysis indicated that VNUT is expressed in mouse neutrophils. Immunohistochemical analysis indicated that VNUT mainly colocalized with matrix metalloproteinase-9 (MMP-9), a marker of tertiary granules, which are secretory organelles. In mouse neutrophils, ATP release was inhibited by clodronate, which is a potent VNUT inhibitor. Furthermore, neutrophils from VNUT−/− mice did not release ATP and exhibited significantly reduced migration in vitro and in vivo. These findings suggest that tertiary granule-localized VNUT is responsible for vesicular ATP release and subsequent neutrophil migration. Thus, these findings suggest an additional mechanism through which ATP is released by neutrophils.

Purinergic chemical transmission is involved in regulating the function of many types of blood cells (1). Polymorphonuclear neutrophils (PMNs) 3 are primary phagocytic cells that play a crucial role in defense against invading microorganisms such as bacteria, fungi, and protozoa. When these organisms invade, PMNs sense the infection and secrete nucleotides such as ATP, which act as autocrine or paracrine signals to initiate a series of responses including granular release, chemotaxis, the production of superoxide, and the regulation of apoptosis upon binding to purinoceptors in neutrophils (2)(3)(4)(5)(6)(7). For example, during chemotaxis, ATP released from chemoattractant-stimulated neutrophils is involved in signal amplification, controlling gradient sensing, and controlling migration speed via the purinergic P2Y2 receptor and A 3 receptor in an autocrine manner (3). Thus, purinergic chemical transmission plays a crucial role in neutrophil function.
Even though ATP signal reception by neutrophils is well understood, the mechanism through which neutrophils secrete ATP is relatively unknown. At least two major mechanisms have been postulated for the release of ATP: one is that ATP release is mediated by ATP channels at the plasma membrane, and another is exocytosis of ATP (8,9). Connexin-43 and pannexin-1 channels are involved in ATP release from neutrophils, leading to migration (10 -12). However, the mechanism of ATP release is controversial; specifically, there are multiple ATP release pathways in neutrophils, including the secretion of granules that store ATP, because the amount of ATP released is dependent on the concentration of the chemoattractant (13). These studies suggest the existence of granule-mediated ATP exocytosis.
Vesicular nucleotide transporter (VNUT) is responsible for the vesicular storage of ATP and plays an essential role in the exocytosis of ATP upon stimulation (14). This protein is a member of the SLC17 anion transporter family and transports nucleotides such as ATP and ADP into secretory vesicles using the membrane potential (⌬; inside positive) across membranes, which is established by vacuolar H ϩ -ATPase (14,15). Recent studies indicated that VNUT is expressed and functions in neuroendocrine and immune cells, which have been reported to be associated with purinergic tissues or cells (14 -22). VNUT Ϫ/Ϫ mice were shown to lose vesicular ATP contents and vesicular ATP release, resulting in loss of purinergic chemical transmission in vivo (19). VNUT is primarily responsible for vesicular ATP storage and release and plays an essential role in purinergic chemical transmission.
In this study, we determined whether VNUT is expressed and functions in neutrophils. We showed that VNUT is cro ARTICLE expressed and localized in the tertiary granules of neutrophils and that neutrophils release ATP in a VNUT-dependent manner. In addition, VNUT-mediated ATP release was shown to be involved in neutrophil migration, which was blocked by the clinically available VNUT inhibitor clodronate.

VNUT is localized in tertiary granules of neutrophils
As the first step of the study, we isolated neutrophils from the bone marrow of wildtype and VNUT Ϫ/Ϫ mice and examined the status of VNUT. Regarding the form of neutrophils, there was no morphological difference between those from wildtype and VNUT Ϫ/Ϫ mice, based on Giemsa staining (Fig. 1A). RT-PCR analysis indicated that a 523-bp VNUT-specific transcript was amplified from mouse neutrophils, whereas no such transcript was detected in wildtype neutrophils without the RT reaction or in neutrophils from VNUT Ϫ/Ϫ mice (Fig. 1B). Immunoblotting with specific antibodies against VNUT indicated that the membrane fraction of wildtype mouse neutrophils contained an immunoreactive polypeptide with an apparent molecular mass of 70 kDa (Fig. 1C). This was absent in neutrophils from VNUT Ϫ/Ϫ mice. The protein levels of other neutrophil granules or vesicle proteins, including V-ATPase, formyl peptide receptor-like 1 (FPRL1), and VAMP2 (vesicleassociated membrane protein 2), were unaffected by the absence of VNUT (Fig. 1C). Furthermore, indirect immunofluorescence microscopy indicated that VNUT was present in mouse neutrophils and exhibited a punctate distribution, whereas control reactions with preabsorbed antibodies resulted in only background levels of staining (Fig. 1D). In VNUT Ϫ/Ϫ mice, no VNUT immunoreactivity was observed, similar to that observed in the negative control (Fig. 1D).

VNUT-dependent vesicular ATP release from mouse neutrophils
Further, we examined whether the vesicular release of ATP from neutrophils occurs in a VNUT-dependent manner. ATP release was observed in isolated mouse neutrophils upon stimulation with 5 M A23187, a Ca 2ϩ ionophore, or 100 nM W-peptide, a formyl peptide receptor ligand; however, this was almost completely abolished in cells from VNUT Ϫ/Ϫ mice ( Fig.   Figure 1. Expression of VNUT in mouse neutrophils. A, neutrophils from the bone marrow of WT and VNUT Ϫ/Ϫ (KO) mice were stained with a Giemsa stain. Segmented nuclei were stained purple. Bar, 10 m. B, RT-PCR analysis was performed using total RNA isolated from bone marrow-derived neutrophils of WT and KO mice (523 bp) after RT reaction (ϩRT) and without RT reaction (ϪRT). The PCR product from VNUT cDNA was shown as a positive control. The expression of mouse glyceraldehyde-3-phosphate dehydrogenase (mG3PDH) was also shown for internal RNA quality control (150 bp). C, Western blot of bone marrow-derived neutrophil membrane vesicles (100 g) prepared from wildtype and VNUT Ϫ/Ϫ mice and probed using anti-VNUT antibody. The position of VNUT (70 kDa) is marked with an arrowhead. The expression of V-ATPase subunit A, FPRL1, and VAMP2 is also shown. D, indirect immunofluorescence microscopy revealed that VNUT is expressed in wildtype mouse bone marrow-derived neutrophils. No VNUT immunoreactivity was observed in neutrophils from VNUT Ϫ/Ϫ mice. Inset, background signal with preabsorbed anti-mouse VNUT antibody. Bars, 10 m. 3, A and B, left panels). On the other hand, no differences were observed in the release of MPO after stimulation with 5 M A23187 and 100 nM W-peptide or in the release of MMP-9 upon stimulation with 5 M A23187 in cells from wildtype and VNUT Ϫ/Ϫ mice; however, W-peptide-induced MMP-9 release was reduced in VNUT Ϫ/Ϫ cells (Fig. 3, A and B, middle and right panels). ATP release was also blocked by 0.1 or 1 M clodronate, a potent and selective VNUT inhibitor (24) (Fig. 3C). After adding 0.1 or 1 M clodronate, neutrophil viability was 97.6 and 96.9%, respectively. Previous studies reported that vesicular ATP release is strongly inhibited by low temperatures or Ca 2ϩ chelators (16,24,25). In this study, we showed that A23187stimulated ATP release is significantly reduced when neutrophils are incubated at 20°C or 4°C (Fig. 3D). In the presence of an extracellular and intracellular Ca 2ϩ chelator EGTA and EGTA-AM, the release of ATP stimulated by 5 M A23187 was also reduced (Fig. 3E). These results strongly suggested that ATP release from mouse neutrophils can occur via a VNUTmediated exocytotic mechanism.

Expression of VNUT in human neutrophils
To examine the function of VNUT in greater detail, we prepared human neutrophils from blood samples obtained from volunteers and determined whether VNUT was expressed in human neutrophils (Fig. 4A). Through RT-PCR analysis, a 115-bp VNUT-specific transcript was amplified in human neutrophils (Fig. 4B). Indirect immunofluorescence microscopy further indicated that VNUT immunoreactivity occurred in human neutrophils and exhibited a punctate distribution, whereas control staining with preabsorbed antibodies exhibited only background levels of staining (Fig. 4C). Furthermore, in agreement with our results in mouse neutrophils, double-labeling immunofluorescence microscopy and digital image analysis (colocalization coefficients) indicated that VNUT immunoreactivity was colocalized with MMP-9 (M1: 0.796, M2: 0.639) (Fig. 4D). In contrast, this immunoreactivity was not colocalized with MPO (M1: 0.191, M2: 0.420), lactoferrin (M1: 0.341, M2: 0.354), or CD35 (M1: 0.167, M2: 0.407). Taken together, it can be suggested that VNUT is also localized in tertiary granules in human neutrophils.

VNUT ؊/؊ mice and VNUT inhibitors result in impaired neutrophil migration
Finally, we studied the role of VNUT-independent ATP release on neutrophil function. To assess neutrophil migration in vitro using cells isolated from the bone marrow of wildtype and VNUT Ϫ/Ϫ mice, we performed Transwell assays consisting of upper wells with neutrophils and lower wells with 100 nM W-peptide separated by a filter with 3-m pores. Consistent with the results of ATP release, we found that the W-peptideinduced migration of neutrophils from VNUT Ϫ/Ϫ mice was decreased by 54% compared with that of cells from wildtype mice (Fig. 5A). No effect on exogenous ATP was observed in the W-peptide-induced migration of neutrophils from VNUT Ϫ/Ϫ mice, whereas the addition of adenosine or IB-MECA, an A 3 receptor agonist, increased the migration of neutrophils from VNUT Ϫ/Ϫ mice (Fig. 5, A and B). The addition of 1 M clodronate also inhibited W-peptide-induced neutrophil migration by 41% (Fig. 5C). To confirm the importance of VNUT for neutrophil migration in vivo, we assessed cell recruitment to the

VNUT-mediated vesicular ATP release in neutrophils
hind paw of wildtype control and VNUT Ϫ/Ϫ mice after a subcutaneous injection of 20 l of 1 mg/ml complete Freund's adjuvant (CFA). As shown in Fig. 5 (D and E), CFA-induced inflammatory and neutrophil recruitment was observed in the hind paw of wildtype mice, but this was reduced in VNUT Ϫ/Ϫ mice. Compared with neutrophil numbers in wildtype mice, those in VNUT Ϫ/Ϫ mice were decreased by 73, 46, 38, and 45% at 6, 12, 24, and 48 h, respectively, after injecting CFA (Fig. 5E).
These results indicated that VNUT is important for neutrophil migration both in vitro and in vivo.

Discussion
How and where ATP is stored in neutrophil granules and how vesicular ATP release is regulated during neutrophil migration are unresolved questions. In the present study, we found that VNUT is involved in the vesicular release of ATP

VNUT-mediated vesicular ATP release in neutrophils
from neutrophils. VNUT is expressed and associated with granules in neutrophils. In vitro, neutrophils were found to release ATP and migrate upon W-peptide stimulation, which is abolished by a VNUT inhibitor and in VNUT Ϫ/Ϫ mice. In vivo, CFA induced neutrophil migration into the footpad of wildtype mice, whereas this migration was reduced in VNUT Ϫ/Ϫ mice. These observations indicated that VNUT is involved in ATP release and migration in neutrophils.
Neutrophils possess several distinct granule subsets, namely azurophil granules, specific granules, and tertiary granules, and secretory vesicles (23). These granules and vesicles are secretory organelles that are classified based on their size, contents, and other parameters. It remains unknown whether ATP is stored in neutrophil granules and subsequently released through granule exocytosis. Our immunohistochemical analyses indicated that VNUT is colocalized with MMP-9, indicating that VNUT-containing granules are tertiary granules that contain V-ATPase as the driving force of ATP transport (⌬) (26). These granules also contain matrix-degrading enzymes (such as MMPs) and membrane receptors (23). Exocytosis of these enzymes from tertiary granules is essential for interstitial structures during neutrophil migration. Upon stimulation, the formyl-Met-Leu-Phe receptor, which is localized to tertiary granules, translocates to the plasma membrane

VNUT-mediated vesicular ATP release in neutrophils
through exocytosis (27).receptors are involved in signal amplification These results suggested that the translocation of the formyl-Met-Leu-Phe receptor to the leading edge of the plasma membrane coincides with vesicular ATP release and promotes more efficient migration in neutrophils. Identification of VNUT-containing granules will be helpful for understanding the mechanisms that regulate ATP release and migration in neutrophils.
We found that W-peptide-stimulated ATP release is abolished in neutrophils derived from VNUT Ϫ/Ϫ mice, whereas nonstimulated cells constantly release ATP (Fig. 3B). Previous inhibitor-based studies have shown that ATP release is also dependent on pannexin 1 hemichannels and connexin 43 from leukotriene B4 (LTB4)-stimulated neutrophils, as well as vesicular release (8 -12). However, we previously found that the disruption of ATP release using inhibitors of hemichannels also inhibits VNUT (28). In addition, the amount of ATP released varies according to the stimulation intensity (13), suggesting that neutrophils have multiple ATP release pathways. Neutrophils might use multiple ATP release pathways depending on the type and concentration of chemoattractant. Therefore, the characterization of ATP release pathways might help to elucidate how VNUT-mediated and channel-mediated ATP release is involved in the pathogenic mechanisms of inflammatory diseases.
In the present study, it was determined that A23187-stimulated MMP-9 release from VNUT Ϫ/Ϫ mice was the same as that from wildtype mice (Fig. 3A). This suggests that VNUT is involved in the vesicular storage and release of ATP but not granular accumulation or the release of other contents. On the other hand, W-peptide-stimulated MMP-9 release was reduced in neutrophils derived from VNUT Ϫ/Ϫ mice (Fig. 3B). Extracellular ATP caused an increase in intracellular Ca 2ϩ through the P2Y2 receptor, leading to degranulation (11). These results suggested that MMP-9 release is dependent on physiological stimuli and is positively controlled by VNUT-mediated ATP release. A similar mechanism of autocrine regulation of ATP has been reported in connection with the release of adrenaline from chromaffin cells (19). VNUT might be a key molecule for autocrine regulation of ATP in neutrophils.

VNUT-mediated vesicular ATP release in neutrophils
Another possibility of this observation is that VNUT regulates tertiary granules secretion independent of ATP secretion under physiological stimulation. We also found that neutrophil migration was not rescued by the addition of ATP, but migration was rescued by adding adenosine and IB-MECA, an A 3 agonist, in VNUT Ϫ/Ϫ cells in vitro. (Fig. 5, A and B). Previous reports showed that A 3 receptors promote neutrophil migration and P2Y2 receptors are involved in signal amplification, gradient sensing and cell motility (29). Our results suggest that exogenous ATP access P2Y2 receptors omnidirectionaly, resulting in a loss of gradient sensing and polarity in neutrophils.
Recently, we identified a potent allosteric inhibitor of VNUT, clodronate (24). VNUT mediated-ATP transport activity is allosterically regulated by Cl Ϫ and inhibited competitively and reversibly by keto acids such as acetoacetate and glyoxylate (22,30). A previous study indicated that clodronate inhibits vesicular ATP release and decreases the release of inflammatory mediators, thereby attenuating chronic inflammation (24). In the present study, we found that clodronate also inhibits vesicular ATP release from neutrophils and reduces neutrophil migration, indicating a novel mechanism through which VNUT mediates its anti-inflammatory effects (Fig. 5C).
Neutrophils migrate to the sites of pathogenic invasion or inflammation and release cytokines in response to a concentration gradient of chemoattractants from microbial pathogens or injured cells. Interestingly, purinergic chemical transmission is involved not only in migration but also in phagocytosis and the production and release of inflammatory mediators (31,32). It has been reported that adenosine, which is a product of ATP degradation, affects phagocytosis via the A 1 or A 2A receptor and modulates inflammatory mediators such as tumor necrosis factor via the A 2A receptor (31,32). Neutrophils generate reactive oxygen species to kill pathogens targeted for phagocytosis; however, the release of such compounds from overactivated neutrophils can also damage host tissues in inflammatory diseases. Purinergic chemical transmission also affects the generation of reactive oxygen species in neutrophils (31,33). VNUT can be a major source of extracellular ATP and is involved in neutrophil activation; it might also be the cause of some inflammatory diseases.
It is noteworthy that VNUT could be a good target for the treatment of chronic inflammatory diseases such as chronic obstructive pulmonary disease and cystic fibrosis. In these diseases, controlling of neutrophil function including recruitment would be an efficient therapy to treat inflammation (34). Clodronate is a specific inhibitor of VNUT and is currently used as an anti-osteoporotic drug (24,35). Our previous study on the biochemistry and pharmacology of clodronate showed that the compound has an anti-inflammatory effect. Therefore, our findings support the clinical significance of clodronate for the treatment of acute sepsis and chronic inflammatory diseases. Further pathological studies regarding the inhibitory effects of clodronate on neutrophil migration are currently in progress in our laboratories.
In conclusion, we demonstrated that VNUT is localized to tertiary granules and is responsible for vesicular ATP release from neutrophils, which is blocked by a clinically available VNUT inhibitor, clodronate. These results strongly suggest that VNUT represents the missing link in the ATP release pathway of neutrophils.

Animals
C57BL/6 mice were obtained from Japan SLC (Shizuoka, Japan). VNUT Ϫ/Ϫ mice were generated as previously described (19). All animal procedures and care were approved by the Institutional Animal Care and Use Committee and were performed according to the guidelines of the Okayama University. For the in vivo experiment, 20 l of 1 mg/ml CFA was injected subcutaneously into the plantar region of the left hind paw of mice (male, 9 -13 weeks old) using a 100-l Hamilton microsyringe with a 27-gauge needle.

Isolation of mouse PMNs from bone marrow and of human PMNs from peripheral blood
Neutrophils were isolated from the bone marrow of mice as previously described (36). Briefly, tibias and femurs were removed and stripped of their muscles. The bone marrow was flushed using neutrophil isolation buffer containing 0.4% sodium citrate in Hanks' balanced salt solution (HBSS). Neutrophils were separated by density centrifugation using a Percoll (62% (v/v) in neutrophil isolation buffer; GE Healthcare) gradient at 1,000 ϫ g for 30 min at 4°C. PMNs were recovered as a pellet at the bottom of the 62% Percoll gradient and washed with PBS. Erythrocytes were removed by hypotonic lysis.
Human PMNs were isolated from peripheral blood samples using Histopaque-1119 and Histopaque-1077 (density: 1.119 g/ml and 1.077 g/ml, respectively; Sigma-Aldrich) double-density gradient centrifugation. Peripheral blood was obtained from healthy volunteer donors in compliance with the guidelines of the ethics committee of Okayama University (permit number 1388). Blood samples (36 ml) were mixed with 4 ml of 3.8% sodium citrate immediately after drawing, layered on a Histopaque gradient, and centrifuged at 700 ϫ g for 30 min at room temperature. The PMN fraction was collected and washed with PBS. The purity of mouse and human neutrophil populations was Ͼ90 and Ͼ93%, respectively, and cell viability was Ͼ97 and Ͼ93%, respectively, as determined by Giemsa and Trypan blue staining. The specimens were observed using an Olympus IX83 microscope and Olympus DP80 camera.

VNUT-mediated vesicular ATP release in neutrophils RT-PCR analysis
An RNeasy mini kit (Qiagen) was used for RNA extraction from neutrophils, according to the manufacturer's instructions. cDNA was generated using PrimeScript RT Master Mix (Takara Bio, Shiga, Japan) with total RNA as the template. RT-PCR was performed as previously described (37). For PCR amplification, cDNA was added to the reaction buffer containing 0.2 mM total dNTP (50.0 M each dNTP), 10 pmol primers, and 1.0 unit of Ex Taq (Takara Bio). The reaction was conducted over 35 cycles as follows: for human neutrophils, denaturation at 95°C for 30 s, annealing at 62°C for 35 s, and extension at 72°C for 20 s; and for mouse neutrophils, denaturation at 95°C for 30 s, annealing at 54°C for 30 s, and extension at 72°C for 1 min. The amplification products were analyzed using a 10% polyacrylamide gel electrophoresis.

Western blotting analysis
For immunoblotting, mouse neutrophils were suspended in 25 ml of 20 mM MOPS-Tris (pH 7.0) containing 0.3 M sucrose, 5 mM EDTA, 10 g/ml pepstatin A, and 10 g/ml leupeptin and then pressurized with N 2 for 20 min at 350 psi with constant stirring at 4°C (38). The cavitate was centrifuged at 500 ϫ g (mouse neutrophils) for 10 min. The resulting supernatant was centrifuged at 250,000 ϫ g for 1 h. The pellet (membrane fraction) was suspended in the same buffer and then denatured with sample buffer containing 1% SDS and 10% ␤-mercaptoethanol. Polyacrylamide gel electrophoresis and Western blotting were performed as described previously (39). Immunoreactivity was visualized by ECL amplification according to the manufacturer (GE Healthcare). Protein concentrations were determined using a protein assay kit (Bio-Rad) with BSA as a standard.

ATP release assay
Mouse neutrophils (1.0 ϫ 10 6 cells) were preincubated in Krebs-Ringer solution consisting of 128 mM NaCl, 1.9 mM KCl, 1.2 mM KH 2 PO 4 , 1.3 mM MgSO 4 , 26 mM NaHCO 3 , 2.4 mM CaCl 2 , 10 mM D-glucose, 10 mM HEPES-Tris (pH 7.4), and 0.2% BSA in the absence or presence of inhibitors for 30 min at 37°C. The cells were stimulated through the addition of 5 M A23187 (Ca 2ϩ ionophore) or 100 nM W-peptide at 37°C. After aliquots were taken at the indicated times, the amount of ATP was measured using an ATP bioluminescent assay kit (Sigma-Aldrich) based on a luciferin-luciferase reaction and with Varioskan Flash Microplate Readers (Thermo Scientific).

MPO assay and MMP-9 assay
MPO and MMP-9 supernatants were analyzed by an ELISA using the mouse MPO ELISA kit (Abcam) and the mouse total MMP-9 quantikine ELISA kit (R&D Systems), according to the manufacturer's instructions.

Transwell migration assay
Migration assays using a Transwell system were performed with a 96-well multiscreen MIC plate (Millipore, Billerica, MA) with a pore size of 3.0 m, as described previously (3). A 100-l suspension of 1 ϫ 10 6 cells in HBSS ϩ10% FBS, with or without inhibitors, with nucleotides or agonists at the indicated concentrations, was added to each well of the upper filter plate. Chemoattractants in HBSS ϩ10% FBS (150 l) were added to each of the lower wells. After incubation at 37°C for 50 min, the upper plate was removed, and cells in the lower plate were counted.

VNUT-mediated vesicular ATP release in neutrophils Assessment of neutrophil migration into footpads
CFA-treated footpads were decalcified with 19% EDTA, fixed in 4% paraformaldehyde in PBS, and cut into 10-m-thick sections. Immunohistochemical analysis was performed using the horseradish peroxidase-DAB method, as described previously (37). In brief, samples were quenched of endogenous peroxidase activity with 0.3% H 2 O 2 and incubated in PBS containing 1.5% goat serum for 30 min. The samples were then incubated with anti-mouse Gr-1 antibodies (1:500) in PBS containing 0.1% BSA and 0.05% Tween 20 for 1 h at room temperature. The samples were washed four times with PBS containing 0.05% Tween 20, and then incubated with biotinylated-labeled anti-rat IgG (1:200; Vector Laboratories, Burlingame, CA) as the secondary antibody for 30 min at room temperature. Samples were washed four times with PBS containing 0.05% Tween 20, incubated with Vectastain ABC reagent (Vector Laboratories) for 30 min, and then incubated with a peroxidase substrate solution consisting of 0.02% DAB and 0.005% H 2 O 2 . Finally, samples were mounted with Mount-Quick (Daido Sangyo, Tokyo, Japan) and observed using a BZ-X700 microscope (Keyence, Osaka, Japan). Stained and migrated neutrophils in the footpads were quantified using BZ-X Analyzer software (Keyence).

Data analysis
All numerical values are shown as the means Ϯ S.E. unless otherwise specified. Statistical significance was determined by performing a two-tailed paired Student's t test or Dunnett's test for multiple comparisons after analysis of variance. These tests were performed using GraphPad Prism version 6 software (GraphPad Software, La Jolla, CA). Differences were considered significant at p Ͻ 0.05.