Monocyte chemoattractant protein-1 (MCP-1) directs migration of blood monocytes to inflamed tissues. Despite the central role of chemotaxis in immune responses, the regulation of chemotaxis by signal transduction pathways and their in vivo significance remain to be thoroughly deciphered. In this study, we examined the intracellular location and functions of two recently identified regulators of chemotaxis, Ca2+-independent phospholipase (iPLA2β) and cytosolic phospholipase (cPLA2α), and substantiate their in vivo importance. These enzymes are cytoplasmic in unstimulated monocytes. Upon MCP-1 stimulation, iPLA2β is recruited to the membrane-enriched pseudopod. In contrast, cPLA2α is recruited to the endoplasmic reticulum. Although iPLA2β or cPLA2α antisense oligodeoxyribonucleotide (ODN)–treated monocytes display reduced speed, iPLA2β also regulates directionality and actin polymerization. iPLA2β or cPLA2α antisense ODN–treated adoptively transferred mouse monocytes display a profound defect in migration to the peritoneum in vivo. These converging observations reveal that iPLA2β and cPLA2α regulate monocyte migration from different intracellular locations, with iPLA2β acting as a critical regulator of the cellular compass, and identify them as potential targets for antiinflammatory strategies.
Chemokine-induced recruitment of peripheral blood leukocytes to tissues is a critical step in development of inflammatory responses. Hence, specific inhibition of leukocyte migration is envisaged as a rational therapeutic approach for inflammatory diseases (1). Monocyte chemoattractant protein-1 (MCP-1), which is a critical ligand for monocyte chemotaxis, binds to its receptor CC chemokine receptor 2 (CCR2) and recruits monocytes to inflamed sites in a variety of chronic inflammatory diseases, such as atherosclerosis, multiple sclerosis, rheumatoid arthritis, and Alzheimer's disease (2). MCP-1 or CCR2 knockout mice exhibited a severe reduction in monocyte chemotaxis to thioglycolate-induced peritonitis and resistance to atherosclerosis (3–8), suggesting that MCP-1 and CCR2 are required for monocyte migration in both health and disease.
Despite the significance of MCP-1 and monocytes in the pathogenesis of inflammatory diseases, our understanding of how MCP-1 transforms random migration of monocytes to directed migration is limited to identification of a few regulatory signaling molecules. These include phospholipase C (9), Src, Syk, MAPKs p42/44ERK1/2, p38, JNK (10), Pyk2 (unpublished data), phosphatidylinositol 3-kinase (PI3K) (11), protein kinase Cβ (12), and association of Arp2/3 with Wiskott-Aldrich syndrome protein (13). Earlier, we discovered Ca2+-independent phospholipase (iPLA2β) and cytosolic phospholipase (cPLA2α) as critical regulators of monocyte chemotaxis to MCP-1. Monocytes rendered deficient in iPLA2β by treatment with antisense oligodeoxyribonucleotides (ODNs) displayed normal cPLA2α activity, and vice versa. Furthermore, monocytes rendered deficient in iPLA2β or cPLA2α by their antisense ODNs are restored for chemotaxis to MCP-1 if treated with lysophosphatidic acid (LPA) or arachidonic acid (AA), respectively, yet both enzymes are required for monocyte chemotaxis to MCP-1 (14). These observations led us to hypothesize that iPLA2β and cPLA2α might be recruited to different intracellular locations, and that they might regulate distinct properties of monocyte chemotaxis to MCP-1. In this study, we tested these possibilities by examining MCP-1–induced redistribution of these phospholipases and by evaluating their contributions to the characteristics of monocyte chemotaxis to MCP-1. We report that MCP-1 induces iPLA2β recruitment to the membrane-enriched pseudopod, whereas cPLA2α is recruited to the endoplasmic reticulum. Although both enzymes regulate speed and net migration of monocytes toward MCP-1, directionality is governed by iPLA2β, likely by regulating F-actin polymerization. Finally, we validate the requirement for both of these phospholipases for monocyte chemotaxis in vivo using a novel mouse model.
iPLA2β regulates monocyte chemotaxis to MCP-1
The antisense oligodeoxyribonucleotides (AS-ODN) used in our previous study to identify the involvement of iPLA2 was directed against the form of iPLA2 that was eventually classified as iPLA2β (14). iPLA2 isoforms (β/γ) display different sensitivities to R or S enantiomers of the pharmacological inhibitor bromoenol lactone (BEL), with iPLA2β being 10 times more sensitive to (S)-BEL than to (R)-BEL (15). As predicted, (S)-BEL caused stronger reduction of MCP-1–induced monocyte chemotaxis compared with (R)-BEL at all concentrations tested, providing additional evidence that monocyte chemotaxis to MCP-1 is indeed regulated by iPLA2β (Fig. 1 A).
iPLA2β and cPLA2α are recruited to different intracellular sites in response to MCP-1
iPLA2β is recruited to the plasma membrane and the pseudopod.
Unstimulated monocytes showed different morphologies that varied from spherical to polar with diffuse, cytoplasmic iPLA2β staining (Fig. 1, C, E, and G, left). Upon MCP-1 stimulation, migrating cells displayed polar morphology, and iPLA2β showed increased distribution at the cell periphery, particularly in the membrane-enriched pseudopod (Fig. 1, C, E, G, right). Quantification of mean fluorescent intensity of iPLA2β in the tail, midbody, and pseudopod of migrating monocytes suggests that, in the absence of MCP-1, iPLA2β was localized predominantly in the tail of monocytes; however, in response to MCP-1, the distribution was reversed and the majority of iPLA2β was detected in the pseudopod of moving cells (Fig. 1 B). These observations demonstrate that MCP-1 induces preferential recruitment of iPLA2β from the tail to the pseudopod of migrating monocytes.
Because MCP-1–induced chemotaxis of iPLA2β-deficient cells could be rescued with LPA (14), which is known to induce focal adhesion assembly (16), it is likely that iPLA2β affects monocyte migration by interacting with an essential component of the pseudopod. Therefore, we evaluated the colocalization of iPLA2β with F-actin and Cdc42, which are two known markers of the pseudopod (Fig. 1, C and E). In the presence of MCP-1, F-actin was enriched in the pseudopod of migrating monocytes and showed increased colocalization with iPLA2β (Fig. 1 C, right, merged image). Quantitative evaluation showed that the MCP-1–induced increased colocalization of F-actin with iPLA2β was not dependent on the polar morphology of monocytes (Fig. 1 D, filled versus black-hatched bar), and that it correlated with significantly enhanced colocalization of iPLA2β with Cdc42 (Fig. 1 E and 1F). Together, these results show that MCP-1 induces redistribution of iPLA2β in nonpolar monocytes that eventually culminates in the increased presence of iPLA2β in the pseudopod of polar monocytes.
Association of iPLA2β with the cell membrane was evaluated by visualizing the colocalization of iPLA2β with Na-K-ATPase (Fig. 1 G). Quantitative evaluation suggested that MCP-1–treated polar monocytes displayed significantly higher colocalization of iPLA2β with Na-K-ATPase compared with untreated monocytes (Fig. 1 H). To rule out the possibility that this was an artifact caused by the small size of the monocytes, the percentage distribution of iPLA2β associated with the cell membrane was also determined by quantifying pixel intensity of iPLA2β on the peripheral cell membrane and in the whole cell (17). As shown in Fig. S1 , MCP-1 induced approximately twofold induction of iPLA2β recruitment to the cell membrane in both polar or nonpolar monocytes, suggesting that recruitment of iPLA2β to the cell membrane is specifically induced by MCP-1, regardless of cell polarity.
cPLA2α is recruited to the endoplasmic reticulum.
Because of the stimulus and cell type–dependent variability in intracellular locations of cPLA2α, we initially explored the recruitment of cPLA2α to early endosomes, Golgi, and endoplasmic reticulum by costaining with Rab 5, lectin GS-II, and protein disulfide isomerase (PDI), respectively. cPLA2α showed minimal or no localization with the Golgi or early endosomes (unpublished data). In the absence of MCP-1, cPLA2α was distributed throughout the cytoplasm (Fig. 2 A, left) and did not reveal specific colocalization with PDI (Fig. 2 A, arrowhead). In MCP-1–stimulated cells, cPLA2α was concentrated in the posterior region of the cell and colocalized with PDI (Fig. 2 A, right). MCP-1 induced a significant increase in recruitment of cPLA2α to the endoplasmic reticulum, irrespective of cell polarity (Fig. 2 B, filled versus black-hatched bar).
iPLA2β and cPLA2α do not colocalize.
To further substantiate that MCP-1 induces localization of iPLA2β and cPLA2α to different intracellular sites, both enzymes were visualized together, and the percentage of colocalization of the phospholipases was determined. In MCP-1–stimulated monocytes, cPLA2α was concentrated in the posterior main body of the cell; in contrast, iPLA2β was concentrated in the pseudopod (Fig. 2 C, right, merged image). Furthermore, the cPLA2α-rich posterior region was relatively devoid of iPLA2β staining, whereas the iPLA2β-enriched pseudopod showed minimal cPLA2α staining (Fig. 2 C, right, merged image, arrow versus arrowhead). The colocalization of iPLA2β with cPLA2α (28 ± 6%), or vice versa (32 ± 6%), remained statistically unaltered because of MCP-1 treatment (Fig. 2 D).
In summary, microscopic observations reveal that MCP-1 induces redistribution of both enzymes; cPLA2α is recruited to the endoplasmic reticulum, whereas iPLA2β is recruited to the membrane-enriched pseudopod of monocytes, and these phospholipases do not colocalize.
iPLA2β and cPLA2α regulate MCP-1–induced net distance of monocyte migration
To gather insight as to how cPLA2α and iPLA2β modify chemotaxis of monocytes to MCP-1, we evaluated their effects on the net distance traveled by monocytes in a gradient of MCP-1 using an under-agarose assay. For these studies, we used previously characterized antisense ODNs to assess the effect of iPLA2β and cPLA2α deficiency on MCP-1–induced chemotaxis of monocytes. It is important to note that these antisense ODNs have been shown to specifically target these enzymes and do not modulate the expression of CCR2 (14, 18, 19). Monocytes were allowed to respond to MCP-1 for 1 h, fixed, and analyzed for net distance. Net distance was determined by measuring the distance of all migrated cells from the periphery of the well to their final position, and it reflects the cumulative effect of change in speed and direction. As shown (Fig. 3), MCP-1 increased the chemotaxis of monocytes and the migrated cells traveled to 1.82-fold greater net distance (3,082.2 μm) than unstimulated cells (1,722 μm). When treated with iPLA2β antisense ODN, 98.7% reduction was observed in MCP-1–induced chemotaxis (Fig. 3 A). These monocytes displayed reduced net distance (88%) in response to MCP-1 (1,631.4 μm) compared with MCP-1–stimulated cells that were not treated with ODN (Fig. 3 B). Similarly, inhibition of cPLA2α expression with antisense ODN caused 94% reduction in MCP-1–induced chemotaxis (Fig. 3 A) and 84% inhibition of net distance traveled compared with ODN-untreated, MCP-1–unstimulated cells (Fig. 3 B). These observations suggest that, upon MCP-1 stimulation, more monocytes migrate toward MCP-1, and that migrating monocytes travel twice the distance compared with unstimulated cells.
iPLA2β regulates both speed and directionalityof monocyte chemotaxis
Reduced net distance of migration of monocytes rendered deficient in phospholipase A2 by antisense ODN, as described in the previous section, in response to MCP-1 (Fig. 3 B) can be caused by a change in the total distance, speed, or net migration of monocytes toward MCP-1 (net-X distance). Therefore, the relative contribution of iPLA2β and cPLA2α to these parameters was delineated by quantitative analysis of time-lapse photomicrographs.
As shown in wind-rose plots (Fig. 4 A), in the absence of MCP-1, monocytes showed random migration (top left). Upon MCP-1 stimulation, cells migrated toward the source of MCP-1 and traveled longer distances (bottom left). Cells treated with sense ODN to iPLA2β migrated similarly to untreated cells (Fig. 4 A, middle). Interestingly, monocytes treated with iPLA2β antisense ODN showed a more clustered migratory pattern both in the absence or presence of MCP-1 (Fig. 4 A, right top and bottom). Quantitative analysis of all cells that could be tracked for 60 min (n) revealed that in response to MCP-1, monocytes migrated significantly faster than unstimulated monocytes and equally to sense ODN–treated cells (Fig. 4 A). In contrast, treatment of monocytes with antisense ODN to iPLA2β significantly decreased the speed of migrating monocytes from 7.58 ± 0.62 to 5.93 ± 0.49 m/min (Fig. 4 A). Furthermore, iPLA2β antisense ODN–treated cells showed a significant reduction in net distance, whereas the net distance traveled by sense ODN-treated monocytes remained unaltered (Fig. 4 C). These findings confirm the results presented in Fig. 3.
We reasoned that speed or net distance merely reflect migration of monocytes, irrespective of their direction. In contrast, the ability of monocytes to reach inflamed sites depends on their ability to navigate toward the source of MCP-1. Hence, the contribution of iPLA2β in directed migration of monocytes toward MCP-1 was determined by quantification of net-X distance that denotes the net value of the X coordinate toward MCP-1. As shown in Fig. 4 D, MCP-1 induced a fivefold induction in net-X. Treatment of monocytes with sense ODN to iPLA2β had no effect; in contrast, monocytes treated with antisense ODN to iPLA2β displayed significant reduction in net-X.
To further probe whether the loss of directionality contributed to the reduced net-X distance (Fig. 4 D) and clustered migratory pattern of iPLA2β antisense ODN-treated monocytes (Fig. 4 A, bottom right), turn analysis was performed. As shown in Fig. 4 E, MCP-1–treated monocytes make fewer wide turns (≥30 degrees, 9 ± 0.81), which is similar to monocytes treated with sense ODN to iPLA2β (9.66 ± 0.76). In contrast, iPLA2β antisense ODN–treated monocytes make more frequent wide turns (>30 degrees, 14.2 ± 1.39). The MCP-1–stimulated monocytes without the ODN treatment showed a significant reduction in average angle per turn from 71.08 ± 7.37 to 35.54 ± 4.18 degrees compared with monocytes without MCP-1 (Fig. 4 F). This effect was totally abrogated in iPLA2β antisense ODN–treated monocytes (Fig. 4 F), suggesting that iPLA2β is a contributor to the cellular compass.
cPLA2α regulates only the speed of monocyte chemotaxisto MCP-1
As shown in Fig. 5 A, wind-rose plots of untreated or monocytes treated with sense ODN to cPLA2α showed random migration in the absence of MCP-1 and directed migration in the presence of MCP-1. Migratory patterns of monocytes treated with antisense ODN to cPLA2α were more clustered than the other groups. Quantitative analysis of the migratory patterns of the maximum number of monocytes (n) that could be tracked for 60 min revealed that antisense ODN to cPLA2α significantly reduced the speed of monocytes (3.96 ± 0.28 μm/min) compared with ODN-untreated, MCP-1–stimulated cells (7.58 ± 0.62 μm/min; Fig. 5 B). cPLA2α antisense ODN–treated cells also displayed significantly reduced net distance (Fig. 5 C) and net migration toward MCP-1 (net–X distance; Fig. 5 D) compared with either no ODN or sense ODN–treated monocytes. Notably, cPLA2α antisense ODN-treated cells displayed no significant reduction in the number of wide (≥30 degrees) turns or the average angle per turn compared with controls (Fig. 5, E and F). These observations suggest that cPLA2α regulates speed, but not directionality, of monocyte chemotaxis to MCP-1.
iPLA2β regulates F-actin polymerization
MCP-1–induced recruitment of iPLA2β to the F-actin–rich pseudopod and the significant loss of directionality of iPLA2β antisense ODN–treated monocytes suggested that iPLA2β might be involved in actin polymerization. We tested this possibility by evaluating the effect of iPLA2β inhibition on MCP-1–induced actin polymerization. MCP-1 induced a significant increase in the level of F-actin. Interestingly, inhibition of iPLA2β by BEL caused significant reduction in F-actin levels (Fig. 6 A). Furthermore, iPLA2β antisense ODN–treated monocytes displayed a significant reduction in actin polymerization compared with ODN-untreated or iPLA2β sense ODN–treated monocytes (Fig. 6 B). These observations indicate that iPLA2β regulates MCP-1–induced actin polymerization.
In vivo validation of the roles of iPLA2β and cPLA2αin monocyte chemotaxis
We developed a novel adoptive transfer mouse model to validate the regulatory functions of iPLA2β and cPLA2α in monocyte chemotaxis to MCP-1 in vivo. It is based on the following: (a) migration of monocytes to the peritoneum in thioglycolate-induced peritonitis in wild-type mice is dependent on MCP-1(7); (b) mouse peripheral blood mononuclear cell preparations do not contain neutrophils, and lymphocytes do not respond to MCP-1; (c) monocytes, by virtue of being fluorescently labeled with a stable PKH26, can be distinguished from unlabeled endogenous cells of the recipient mice; (d) we hypothesize that treatment of cells with phospholipase A2–specific inhibitors or antisense ODNs before adoptive transfer will cause migration of fewer adoptively transferred monocytes; and (e) because adoptively transferred cells constitute only a small fraction of total blood cells (11–14%), migration of endogenous leukocytes of the recipient mouse to the peritoneum will not be affected (Fig. 7 A).
Pharmacological inhibition of cPLA2α/iPLA2β suppresses monocyte migration in vivo.
Earlier, we reported that the iPLA2 inhibitor racemic-BEL and the cPLA2/iPLA2 inhibitor arachidonyl trifluoromethyl ketone (AACOCF3) both suppressed MCP-1–induced chemotaxis of primary monocytes (14). As predicted from these in vitro data, significantly fewer AACOCF3-treated monocytes migrated to the peritoneum in the adoptive transfer model (Fig. 7 B). Surprisingly, no significant difference was observed in the migration of monocytes treated with racemic-BEL compared with untreated monocytes. The inefficiency of racemic-BEL to inhibit monocyte migration in vivo was found to be caused by time-dependent recovery of BEL-treated monocytes at 37°C in vitro (Fig. S2). To determine whether the apparent efficacy of AACOCF3 in vivo was caused by toxicity, we examined monocyte survival and chemotaxis to MCP-1 24 h after treatment with AACOCF3 in vitro. As shown in Fig. S3, AACOCF3 treatment did not affect monocyte survival, and the monocytes still displayed significant reduction in chemotaxis to MCP-1 compared with untreated monocytes. These observations suggest that effectiveness of AACOCF3 was caused by the stable inhibitory effect of AACOCF3 on cPLA2α and iPLA2β. As expected, no change was observed in either total monocytes or total leukocytes in the peritoneal lavage (Fig. 7 B).
Phospholipase A2 antisense ODN inhibits MCP-1–mediated chemotaxis of monocytes.
Because AACOCF3 inhibits both iPLA2 and cPLA2 and racemic-BEL proved ineffective in vivo, we tested the effect of specific antisense ODNs. iPLA2β antisense ODN used in this study has previously been demonstrated to reduce iPLA2β protein expression in both murine and human monocytes/macrophages (14, 18, 19). It significantly reduced the expression of iPLA2β without affecting cPLA2α protein expression or enzyme activity (14) and was effective in reducing monocyte chemotaxis to MCP-1 in vitro even 24 h later (unpublished data). Similarly, cPLA2α antisense ODN reduced monocyte chemotaxis in vitro without affecting iPLA2β protein expression or enzyme activity (14). It reduced the expression of cPLA2α both in human and murine monocytes (Fig. S4) (14).
As shown (Fig. 8 A), adoptively transferred monocytes treated with iPLA2β antisense ODN displayed statistically significant reduction in migration to the peritoneum compared with untreated or sense ODN–treated monocytes (top). Migration of endogenous monocytes and total leukocytes remained statistically similar. These data suggest that unlike racemic-BEL, iPLA2β antisense ODN–treated monocytes displayed sustained reduced migration in vivo.
To evaluate the in vivo requirement for cPLA2α in MCP-1 chemotaxis, adoptively transferred cells were pretreated with antisense ODN to cPLA2α and monitored for their response to thioglycolate. As shown in Fig. 8 B (top), significantly fewer cPLA2α antisense ODN–treated cells migrated to the peritoneum compared with untreated or cPLA2α sense ODN–treated cells. As expected, infiltration of monocytes (Fig. 8 B, middle) or total leukocytes (Fig. 8 B, bottom) remained unaffected in animals that received either untreated or cPLA2α sense ODN–treated cells. These results demonstrate that iPLA2β and cPLA2α are required for MCP-1–induced chemotaxis of monocytes in vivo, thus validating our previous in vitro findings.
Our studies characterize the important relative contributions of cPLA2α and iPLA2β for regulating monocyte chemotaxis to MCP-1. Reduced migration of adoptively transferred, phospholipase-antisense ODN–treated mouse monocytes to the peritoneum in response to thioglycolate indicates the important signaling roles of iPLA2β and cPLA2α in monocyte chemotaxis to MCP-1 in vivo.
The intracellular locations of these phospholipases vary with the type of cell and nature of stimulus. For example, in INS-1 cells, iPLA2β translocates to the perinuclear region during endoplasmic reticulum stress or stimulation with glucose (20, 21), whereas in smooth muscles cells, iPLA2β is associated with the plasma membrane (22). Similarly, cPLA2α has been reported to be associated with the nuclear membrane, Golgi complex, endoplasmic reticulum, plasma membrane, and an undefined intracellular compartment (23–25). Our observations of the significant increase in colocalization of iPLA2β with membrane and pseudopod markers and of cPLA2α with the ER marker in nonpolar but MCP-1–stimulated monocytes suggest that their recruitment to distinct intracellular locations is induced by MCP-1. This conclusion is further supported by the insignificant colocalization of iPLA2β with cPLA2α, and vice versa.
Earlier, we showed the restoration of chemotaxis of iPLA2β-deficient monocytes to MCP-1 with LPA (14). LPA has been reported to induce phospholipase D (PLD) activity, cytoskeleton rearrangements, and assembly of focal adhesions (26). PLD can generate phosphatidic acid (PA) from membrane phospholipids (27). Interestingly, PA is a preferred substrate of iPLA2β, yielding LPA (28).We have observed that PLD activity is induced by MCP-1 and required for monocyte chemotaxis (unpublished data). Furthermore, PLD is required for LPA-induced cytoskeletal rearrangements, likely through LPA formation (27). As schematically represented in Fig. 9, these observations tempt us to speculate that in response to MCP-1, PLD acts on membrane phospholipids to generate PA, and then iPLA2β acts on PA to liberate LPA. LPA possibly interacts with components of pseudopod assembly and regulates cytoskeletal reorganization; hence, speed and directionality of monocytes. This conjecture is reinforced by observations that actin polymerization is spatially localized to the membrane-enriched pseudopod, where iPLA2β is recruited in response to MCP-1 (Fig. 1, B and C), MCP-1 induces association of iPLA2β with F-actin, irrespective of cell polarity (Fig. 1 D), and MCP-1–induced actin polymerization is significantly reduced in response to iPLA2β inhibition or reduced expression (Fig. 6).
MCP-1 chemotaxis can be restored in cPLA2α-deficient monocytes by AA (14). Thus, cPLA2α may contribute to monocyte migration by generating AA in the endoplasmic reticulum (Fig. 9). AA can be converted into various metabolites, such as prostaglandins, lipoxins, LTB4, and H(P)ETEs via COX-2, cytochrome P450, or lipoxygenase pathways (29). Lipoxins induce RhoA- and Rac-mediated actin reorganization in monocytes/macrophages (30), whereas prostaglandin 2 increases migration of dendritic cells (31). Our observations, in view of these reports, call for identification of key metabolites of AA that regulate monocyte migration to MCP-1.
Directed cell migration is initiated by sensing of chemotactic stimuli, and culminates in the directed migration of cells. During this process, an external shallow gradient of chemokine is perceived, translated, and amplified into a stable intracellular gradient of signaling molecules that results in the formation of pseudopod at the front and uropod in the rear of a migrating cell. The pseudopod is formed by the activation of the heterotrimeric GTP binding protein Gi, whereas the uropod is formed by the activation of heterotrimeric GTP-binding protein G12/13. Phosphatidylinositol (3,4,5) triphosphate (PtdIns[3,4,5]P3), a key lipid messenger, is synthesized in the leading edge by PI(3)Ks and is degraded by PTEN in the rear edge of migrating cells (32). This vectorial metabolism gives rise to an intracellular gradient of PtdIns(3,4,5)P3. Pseudopod is organized by Rac/PtdIns(3,4,5)P3/actin and is stabilized by the scaffold Hem-1 (33, 34). In contrast, the uropod is organized by Rho-mediated, myosin-based contractions (35). This distinct intracellular hierarchy is thought to commence polarity and ensure directed migration of cells in a chemokine gradient. Multiple signaling molecules are believed to orchestrate these distinct events at the opposing edges of a migrating cell. For example, Cdc42 regulates stability and location of the leading edge, whereas Rac, which transforms random migration of fibroblasts and epithelial cells to directed migration (36), only controls morphological and invasion characteristics of macrophages (37). Vav1 is supposed to control the speed of migrating macrophages (38). In neutrophils, deficiency of Rho GEF Lsc causes faster migration and loss of directionality (39), whereas deficiency of PI3Kγ causes only loss of directionality (40). In Dictyostelium discoideum, PTEN regulates levels of PtdIns(3,4,5)P3; in contrast, in neutrophils, it is metabolized by the PtdIns(3,4,5)P3 phosphatase SHIP1 rather than PTEN. These observations not only point toward evolution of cell-based variability in key signaling molecules, they also provide an impetus to search for unknown members of these cascades (41, 42).
Once considered a probable cellular compass, recent reports suggest that PtdIns(3,4,5)P3 is primarily a component of positive-feedback loop controlling persistent actin dynamics (43). Emerging concepts argue for the existence of parallel signaling pathways for chemotaxis and indicate that an unknown cellular compass exists upstream of PI3Kγ (42, 44). Interestingly, a calcium-independent phospholipase (PLA2A) that liberates AA from phosphatidylcholine and acts in parallel with PI3K/PTEN has recently been implicated in chemotaxis of D. discoideum (45). Because iPLA2 does not show a preference for the fatty acyl moiety at the sn-2 position of phospholipids, and D. discoideum does not contain cPLA2α, it is possible that PLA2A merely performs twin functions of cPLA2α and iPLA2β by liberating LPA and AA. Our observations on loss of directionality of iPLA2β-deficient monocytes raise the possibility that iPLA2β is either a cellular compass or a component of a cellular compass. Our observations, together with recent reports (44, 45), suggest that phospholipases are critical components of parallel signaling pathways regulating monocyte chemotaxis as schematically represented in Fig. 9. How iPLA2β directs monocytes toward MCP-1, as well as whether it is a universal regulator of chemotaxis, remains to be deciphered. One key function ascribed to iPLA2β is membrane remodeling and membrane trafficking that brings about changes in shape and fluidity of the plasma membrane (46). It will be interesting to explore whether iPLA2β regulates monocyte chemotaxis by changing lipid composition of membrane rafts, thereby modifying assembly of signaling molecules in the pseudopod of migrating cells (47).
The impaired in vivo migration of adoptively transferred mouse monocytes that contain either inactive PLA2 (by AACOCF3) or were rendered deficient in PLA2 by AS-ODN validates our in vitro observations (14) in a dynamic in vivo milieu. We now have adapted this in vivo assay to monitor human monocyte chemotaxis (48). Our reports show the importance of in vitro studies for identification of potential signaling components. Our in vivo MCP-1 chemotaxis assay provides a straightforward approach to verify the relevance of signaling pathways that appear promising for regulating MCP-1–dependent chemotaxis in vitro.
In summary, our studies provide insight into how two seemingly similar phospholipases regulate directed migration of monocytes from different intracellular sites. We report that MCP-1 induces chemotaxis of monocytes by increasing speed, as well as introducing directionality. By translocating to the pseudopod and regulating directionality and speed, iPLA2β acts front and center, whereas in contrast, cPLA2α is recruited to the endoplasmic reticulum and regulates only speed. The profound defects in migration of iPLA2β- or cPLA2α-deficient mouse monocytes in vivo validate the important roles for these enzymes in regulating extravasation of blood monocytes to sites of inflammation.
Materials And Methods
Antibodies and reagents
BEL and AACOCF3 were obtained from Biomol. Bel enantiomers (R and S) were provided by R. Gross (Washington University School of Medicine, St. Louis, MO). Human recombinant MCP-1 (BD Biosciences) was used at a concentration of 5.75 nM. Anti-iPLA2β (1:50; goat polyclonal), anti–human Cdc-42 (1:50; rabbit polyclonal), and anti–human PDI (1:100; rabbit polyclonal) were purchased from Santa Cruz Biotechnology. Anti–human cPLA2α (rabbit polyclonal) and a mouse monoclonal antibody against human Na-K ATPase (1:200) were obtained from Cell Signaling Technology. Rhodamine phalloidin (1:40), Rab5, lectin GSII-Alexa Fluor 594, and Alexa Fluor 594– or 488–tagged species-specific secondary antibody raised in donkey (1:200) were obtained from Invitrogen.
Isolation of human monocytes and cell culture
Human monocytes were isolated from freshly drawn human whole blood, as described by McNally et al. (49), and maintained in DME (10% BCS) in polypropylene tubes at 2 × 106 cells/ml. Studies on human monocytes were approved by the Institutional Review Board of the Cleveland Clinic.
Microchamber chemotaxis assay.
This assay was performed as previously described (14) and was used to quantify monocyte migration to MCP-1.
Under-agarose cell migration assay.
The under-agarose cell migration assay was performed as described by Heit and Kubes (50) using 10% BCS containing DME. Net distance in this assay denotes the distance between the edge of the well and the final position of migrating cells. It was determined from the composite photomicrograph of the entire field of migration.
Treatment of monocytes with pharmacological inhibitorsand antisense ODNs
Monocytes were pretreated with Racemic-BEL or its R or S enantiomers (0.1–5 μM) or AACOCF3 (50 μM) at 37°C in 10% CO2 for 1 h, and washed in serum-free DME before being used for chemotaxis. The sense and antisense ODN sequences for phospholipases used in this study were phosphorothioate modified and purified by HPLC (Sigma-Aldrich). These ODNs were of proven efficacy and specificity, and were delivered under standardized experimental conditions, as described (14). These specific antisense ODNs were effective in reducing the expression of iPLA2β or cPLA2α, as described in our previous study (14).
Immunostaining and confocal microscopy
Monocytes were allowed to migrate using the under-agarose cell migration assay (1 h), fixed, and permeabilized with Triton X-100 (0.1% in PBS for 4 min). Nonspecific sites were blocked with 3% BSA in PBS (37°C for 1 h). Cells were incubated with primary antibody (for 1 h at 37°C), washed, and visualized with fluorescently labeled secondary antibody (for 1 h at 37°C). Stained cells were mounted in Vectashield (Vector Laboratories) and observed using a laser-scanning confocal microscope (DMR XE attached to TCS SP2; Leica). Horizontal sections were imaged using a 63× objective (magnification 5×, pinhole 1, and format of 1,024 × 1,024). Images were analyzed by Image-Pro Plus (Media Cybernetics) and processed using Photoshop CS (Adobe).
For this purpose, under-agarose cell migration was performed in serum-coated polystyrene six-well culture dishes, essentially as described in the previous section. The plate was placed on a preheated stage at 37°C under 10% CO2. Images were acquired with an inverted microscope (DMIRB; Leica) using a 10× phase objective. Images were acquired at 3-min intervals using 30-ms exposure times and recorded with a cooled charge-coupled device camera (Cool SNAP HQ; Roper Scientific) using Metamorph software (Universal Image).
Image analysis was conducted with Image-Pro Plus. In the absence of MCP-1, randomly migrating cells often return to the well where they originated. Because of the large number of cells in the well itself, returning cells could not be followed after reentering the well. Therefore, cells that can be tracked for the 60-min duration were used for migration analysis. Total distance depicts total length of a migratory path, speed denotes total distance/60 min, and net distance depicts the linear distance between the first and the last position of a migrating cell. Net-X is the net value of the X coordinate after 60 min. The turn angle is the absolute value of the angle between successive positions of migrating cell measured from the direction of the preceding position. Average angle/turn depicts the average angles of the turns taken by a migrating cell during 60 min of migration at 3-min intervals.
Quantification and colocalization of phospholipases
Mean fluorescent intensity of iPLA2β in tail, midbody, and pseudopod of polar cells was determined by measuring pixel intensity of iPLA2β and area clearly demarcated based on cellular morphology shown by differential interference contrast (DIC) images using Image-Pro Plus. Colocalization (percentage) of enzymes with marker proteins were determined as previously described (51).
Determination of F-actin
This assay was used to determine the effect of iPLA2β inhibition on F-actin polymerization, as described by Finkel et al. (52) using 2 million cells.
Isolation of mouse peripheral blood mononuclear cells
Animals (BALB/CJ; Jackson ImmunoResearch Laboratory) were used according to the protocol approved by the Cleveland Clinic Foundation Institutional Animal Care and Use Committee. BALB/CJ female mice (7–8 mo, two donors per recipient) were anesthetized with avertin (0.4 ml/animal). Blood was collected by cardiac puncture in a 1-ml syringe containing 0.05 ml of EDTA (0.5 M). Blood (10 ml) was diluted with 10 ml of Tricine-buffered saline (TBS; 10 mM, pH 7.4) and mixed with 5 ml of OptiPrep (Sigma-Aldrich). Blood was transferred to a suitable tube, overlayed with 0.5 ml of TBS and centrifuged at 1,000 g for 30 min at 20°C without the brake. The supernatant, containing mononuclear cells, was collected, diluted with 2 volumes of TBS, and centrifuged at 400 g for 10 min. The pellet was reconstituted in 0.5 ml of TBS, layered on dialyzed calf serum (2 ml), and centrifuged at 200 g for 15 min to remove platelets. Cells were suspended in TBS and counted. The recovery of mononuclear cells by this method was 7.45 × 104/g mouse.
Fluorescent labeling of mononuclear cells with PKH26
Cells were labeled per the manufacturer's instructions. Cells were suspended in DMEM and treated with inhibitors as previously described (14).
Adoptive transfer of mononuclear cells and initiation of peritonitis
Recipient mice (3–4 per group) were lightly anesthetized with avertin. Peritoneal inflammation was induced by thioglycolate injection (1 ml, 4% in physiological saline) into the peritoneal cavity. Tail veins were dilated with limonene and cleaned with 95% ethanol. The labeled mononuclear cells (1.4–1.8 million per animal) were injected into the tail vein, and the time of injection was recorded. Injected cells represent only 11–14% of total peripheral blood mononuclear cells of similar weight recipient animals. After 24 h, peritoneal cells were harvested, washed, resuspended in PBS (1 ml), and counted. Cells were fixed and used to prepare 10 cytospins for each sample. Half of them were stained with HEMA per the manufacturer's (Thermo Fisher Scientific) instruction, to determine the number of monocytes. The remainder was mounted in Vectashield. PKH26-positive cells were counted on an upright microscope (model DMR; Leica) using a Texas red filter.
Statistical significance of observations was calculated using the Student's t tests. P < 0.05 was considered significant.
Online supplemental material
Fig. S1 presents data indicating that MCP-1 induces membrane localization of iPLA2β in both polar and nonpolar monocytes. Fig. S2 demonstrates the transient nature of BEL inhibition of monocyte chemotaxis to MCP-1 in a time-dependent manner. Fig. S3 shows the stable inhibitory effect of AACOCF3 on monocyte chemotaxis to MCP-1, and that AACOCF3 does not affect monocyte viability under these conditions. Fig. S4 shows inhibition of cPLA2α protein expression in murine monocytes/macrophages by treatment with antisense ODN. Supplemental materials and methods are also provided.
The authors thank Dr. Richard Gross for the gift of BEL enantiomers; Amit Vasanji for help with image analysis; and Drs. Josephine Adams and Ashish Bhattacharjee for their helpful comments.
This study is supported by National Institutes of Health grants (HL 510681 and HL 61971) to M.K. Cathcart and by the General Clinical Research Center Grant M01-RR-018390.
The authors have no competing financial interests.
Abbreviations used: AA, arachidonic acid; AACOCF3, arachidonyl trifluoromethyl ketone; BEL, bromoenol lactone; CCR2, CC chemokine receptor 2; cPLA2, cytosolic phospholipase; DIC, differential interference contrast; iPLA2, Ca2+-independent phospholipase; LPA, lysophosphatidic acid; MCP-1, monocyte chemoattractant protein-1; ODN, oligodeoxyribonucleotide; PA, phosphatidic acid; PDI, protein disulfide isomerase; PI3K, phosphatidylinositol 3-kinase; PLD, phospholipase D; PtdIns(3,4,5)P3, phosphatidylinositol (3,4,5) triphosphate; TBS, Tricine-buffered saline.