Published online before print April 12, 2007
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Cell Biology and Physiology, Wright State University, School of Medicine, Dayton, Ohio, USA
1 Correspondence: Cell Biology and Physiology, Wright State University School of Medicine, 3640 Colonel Glenn Highway, Dayton, OH 45435, USA. E-mail: julian.cambronero{at}wright.edu
ABSTRACT
Phospholipase D (PLD) is an enzyme that catalyzes the conversion of membrane phosphatidylcholine to choline and phosphatidic acid (PA; a second messenger). PLD is expressed in nearly all types of leukocytes and has been associated with phagocytosis, degranulation, microbial killing, and leukocyte maturation. With the application of recently developed molecular tools (i.e., expression vectors, silencing RNA, and specific antibodies), the demonstration of a key role for PLD in those and related cellular actions has contributed to a better awareness of its importance. A case in point is the recent findings that RNA interference-mediated depletion of PLD results in impaired leukocyte adhesion and chemotaxis toward a gradient of chemokines, implying that PLD is necessary for leukocyte movement. We forecast that based on results such as those, leukocytes may prove to be useful tools to unravel still-unresolved mechanistic issues in the complex biology of PLD. Three such issues are considered here: first, whether the cellular actions of PLD are mediated entirely by PA (the product of its enzymatic reaction) or whether PLD by itself interacts with other protein signaling molecules; second, the current difficulty of defining a "PA consensus site" in the various intracellular protein targets of PA; and third, the resolution of specific PLD location (upstream or downstream) in a particular effector signaling cascade. There are reasons to expect that leukocytes and their leukemic cell line counterparts will continue yielding invaluable information to cell biologists to resolve standing molecular and functional issues concerning PLD.
Key Words: signal transduction • differentiation
INTRODUCTION
Phospholipase D (PLD): the genes and the proteins
PLD is an enzyme that hydrolyzes the phosphodiester bond in phosphatidylcholine (PC), yielding choline and phosphatidic acid (PA) [1
2
3
]. PLD is involved in physiological and cellular signaling pathways, primarily through the production of second messengers such as PA and indirectly, diacylglycerol (DAG). Some of these physiological processes include cytoskeletal rearrangement, vesicle trafficking, exocytosis, phagocytosis, oncogenesis, and neuronal and cardiac stimulation [4
].
Mammalian PLD is encoded by two genes: PLD1 and PLD2, which give rise to proteins PLD1 and PLD2 and four other slightly shorter, spliced forms (Fig. 1 ). The PLD1 gene has been localized in the long arm (q) of chromosome 3 (3q26) [5 ] and is defined by 31 exons, 27 of which yield the expression of four splice variants. PLD1 splice variants include PLD1a, PLD1a2, PLD1b, and PLD1b2 [6 , 7 ]. The longest form, PLD1a, is a 1072-amino acid protein of 120 kDa MW, which is mostly membrane-associated [8 ]. A mammalian PLD2 isoform was cloned from mouse embryos [9 ] and later from human B lymphocytes [10 ]. The PLD2 gene is located on the short arm (p) of chromosome 17 (17p13) [11 ] and encodes a 933-amino acid protein with an apparent MW of 106 kDa [9 ]. PLD2 has two splice variants, PLD2a and PLD2b (Fig. 1) , which are functionally indistinguishable [12 ].
 View larger version (29K): [in a new window] |
Figure 1. Genomic and protein map of PLD isoforms. Representation of the PLD1 (upper half of the figure) and PLD2 (lower half of the figure) cDNA homologies in relation to the genomic organization of each PLD gene. Data used to construct the figure were derived from analysis of the human genome (www.ncbi.nlm.nih.gov). The specific PLD1- or PLD2-spliced exons are depicted in red boxes. Exon numerals are indicated beneath the exon boxes (blue for PLD1 and gray for PLD2 genes); nucleotide positions are in black and on top of each cDNA representation. Human PLD1 and PLD2 genes share 50% identity and are located in the long arm of chromosome 3 (reverse strand locus 3q26) and in the short arm of chromosome 17 (locus 17p13.1), respectively. The human PLD1 gene covers  20.8 kb genomic DNA defined by 31 exons and directs the expression of at least four alternative spliced transcripts: PLD1a, PLD1b, PLD1a2, and PLD1b2. PLD1b, an evolutionary conserved splice variant of PLD1a, arises from splicing of a 38-amino acid, codified by the alternate exon 19 (yellow box), whereas PLD1a2 is the result of exon 29 splicing (green box). In the case of PLD2, 16.3 kb genomic DNA, located in 25 exons, defines the gene. The PLD2 gene directs the generation of at least two alternatively spliced isoforms: PLD2a and PLD2b. PX, phox consensus sequence; PH, plekstrin homology; HKD, HxxxxKxD; PIP, phosphatidylinositol phosphate; 3'-UTR, 3' untranslated region.
|
50% homology. However, all members of the PLD superfamily contain two highly conserved, catalytic domain HKD motifs. Other conserved sequences within the PLD structure include the PX, the PH domain, and the phosphatidylinositol 4,5-bisphosphate [PI(4,5)P2]-binding site [13
14
15
] (Fig. 1)
. Deletion of the PH and/or the PX region leads to a pronounced mislocalization of PLD, which alters the catalytic activity of the isoenzymes. Several reports have indicated that the PX domain is involved in binding regulatory factors (such as PIP) and proteins [13
, 15
, 16
].
Regulation of PLD function
PLD1 and PLD2 are expressed as phosphotyrosine proteins in vivo [17
]. However, the role of tyrosine phosphorylation in the regulation of PLD functions is not well understood. Although evidence implicates epidermal growth factor (EGF), platelet-derived growth factor, insulin, and other mitogens in the phosphorylation of PLD2 [17
], no direct proof of PLD2 as a direct substrate of tyrosine kinase action is available. Several tyrosine residues in PLD2 are potential targets for phosphorylation, and it has been found that Y169 is necessary for high basal PLD2 activity [18
]. However, the nature of the kinase(s) responsible for its phosphorylation as well as the mechanisms implicated in the regulation of Y169 phosphorylation are still elusive.
PLD1 enzymatic activity is regulated through phosphoinositides, protein kinase C (PKC), ADP ribosylation factor (ARF), Rho family GTPases, fatty acids, Ca2+, and protein phosphorylation [19
, 20
]. Other PLD1 regulators are Rac1 and Cdc42. In the presence of GTP
S, the Rho family proteins stimulate PLD1 activity directly in several cell types [21
, 22
]. These small GTPases are often associated with PLD1, as mutation of the Rho-binding site on PLD1 prevents PLD1/ARF interaction [15
, 23
]. The Rho family of GTPases is also involved in indirect regulation of PLD1 enzymatic activity through stimulation of PI(4,5)P2kinase, Rho kinase [24
], or by intracellular translocation of the PLD isoforms [25
26
27
]. Conversely, Rac1 does not regulate PLD2 activity [9
, 10
]. This enzyme is obviously tightly regulated, as the reaction product on PC, PA, once produced, has multiple targets inside the cell and regulates a plethora of cellular functions [28
29
30
31
32
].
THE IMPORTANCE OF PLD ACTIVITY IN LEUKOCYTE FUNCTION
PLD1 and PLD2 are expressed in a wide variety of mammalian tissues, and as such, they are present in primary leukocytes [12 , 33 ]. Human neutrophils express PLD1 and PLD2 at the protein level [34 35 36 37 38 39 40 ], and detailed molecular analyses revealed that human neutrophils express all known variants of PLD1 and PLD2 [41 ]. PLD enzymes are expressed in monocytes [42 43 44 45 ], macrophages [46 , 47 ], basophils [46 , 48 ], eosinophils [49 , 50 ], dendritic cells [45 , 51 , 52 ], lymphocytes [53 , 54 ] and NK cells [55 , 56 ]. In addition to being present in primary cells, PLD1 and PLD2 have been found in a variety of leukemic cell lines, such as U937, THP-1, HL-60, and PLD-985 [57 58 59 60 61 ]. These cell lines have also served the purpose of investigating PLD expression during maturation and differentiation of myeloid lineages. In a series of studies, which have used a combination of primary cells and leukemic cell lines, PLD has been implicated in at least two key leukocyte functionalities: phagocytosis and chemotaxis. We shall discuss each in detail.
Phagocytosis and the oxidative burst
The direct role of PLD activity in phagocytosis, degranulation, and microbial killing by neutrophils and macrophages has been subject to intense investigation in the last 15 years or so [62
63
64
65
66
67
68
]. Resting macrophages and neutrophils exhibit little PLD activity. In contrast, when these cells are challenged in vitro with virulent or killed microorganisms, PLD activity increases more than 10 times [63
, 67
, 69
]. Although the involvement of surface receptors in phagocytosis is relatively well established [64
], the early demonstration that PLD activation correlates closely with phagocytosis is yet to be understood completely at the mechanistic level. PLD concentrates at forming phagosomes [47
, 70
], and PA is concomitantly produced [71
], demonstrating that the localized PLD is active. It is known that phagocytosis requires the presence of specific protein tyrosine kinases, GTP-binding proteins, PKC, NADPH activity, actin polymerization, and membrane movement. All these players are downstream of PLD activity [30
].
Neutrophil degranulation and microbial killing are accompanied by increased PA formation, activation of tyrosine kinases, and stimulation of the NADPH oxidase system [72 73 74 ]. The respiratory burst is PA-dependent [49 , 74 75 76 77 78 79 80 ], and PLD inhibitors affect IgG-dependent phagocytosis in human monocytes [43 , 47 , 63 ]. Moreover, the absence of PLD-mediated generation of PA correlates with a severely impaired oxidative burst or microbial killing in neutrophils and macrophages [47 , 68 ]. However, new evidence suggests that at least partially, PLD-generated PA may not be related directly to phagocytosis. RAW cells, expressing a catalytically inactive version of PLD2 (K758R), did not result in PLD-mediated inhibition of phagocytosis [70 ]. Although the reasons for these apparent discrepancies are not known, it may be related to the cells (RAW vs. monocytes) and/or the tools used to study the role of PA in phagocytosis: pharmacological inhibition of PLD versus molecular depletion of PLD activity by point mutants. Recently, it has been demonstrated that the PH domains of PLD1 and PLD2 are strong modulators of the membrane recycling machinery, resulting in regulated growth factor receptor endocytosis [81 ].
PLD and cell adhesion and chemotaxis
In addition to the established role of PLD mediated by PA in leukocyte function as discussed (albeit still lacking full understanding of intracellular mechanisms), a recurrent theme in the scientific literature has been cell chemotaxis. Leukocyte motility is a complex process, requiring such integrated pathways as actin polymerization, cytoskeletal reorganization, morphological polarization, specific adhesiveness, and cell-substratum detachment [82
]. Information about the molecular bases underlying those has been derived often from pharmacological manipulation of a wide variety of signaling cascades (i.e., PKC, PLD, MAPK, PI-3K, PLA2, and tyrosine phosphorylation) [83
, 84
]. It is the precise contribution of PLD to cellular movement that we shall discuss next.
PLD activation is an important event in neutrophil signal transduction following exposure of adherent cells to the cytokine GM-CSF [85 ]. Chemotactic migration of neutrophils through endothelial monolayers in response to fMLP results in endothelial cell PLD activation [86 ]. PLD1 is known to be activated by small GTPases, and Rho GTPase-controlled cytoskeletal dynamics are needed for cell migration and other neutrophil functions [87 ]. Rac GTPase controls chemotaxis in live neutrophils [88 ], and suppression of Rac2 activation concomitantly with reactive oxygen species generation is mediated by the Rho family exchange factor (guanine nucleotide exchange factor), Vav-1 [89 ]. In neutrophils, Rac1 plays an important role during gradient detection and actin assembly via PI-3K and Akt [90 ]. The ribosomal S6 protein kinase, p70S6K, is also involved in leukocyte chemotaxis, in response to GM-CSF or fMLP [91 ] (Fig. 2 ).
 View larger version (66K): [in a new window] |
Figure 2. Molecular "crosstalk" between two related signaling pathways, MAPK and PLD/PA, enhance of the ribosomal S6 kinase (p70S6K), while the immunosuppressant rapamycin inhibits it. Chemotaxis starts with an agonist, GM-CSF (from the host’s compromised tissue) or fMLP (from bacteria), and the activation of p70S6 kinase plays a key role [91
, 92
].
|
In addition to stimulating chemotaxis, IL-8 (a ELR+ CXC chemokine) is involved in neutrophil lipid metabolism. Sozzani et al. [96 ] demonstrated that IL-8 activates PLD and the respiratory burst, thus involving PA, the product of the PLD reaction, in neutrophil signaling. Furthermore, the formation of IL-8-induced PA is inhibited by pertussis toxin and tyrosine kinase inhibitors [97 ]. During inflammation and phagocytosis, PA promotes degranulation [39 ]. However, new molecular tools (expression vectors, silencing RNA, and specific antibodies) have propelled these studies to new levels of understanding. Two recent papers have linked PLD to cell adhesion and chemotaxis, one from Iyer et al. [98 ] and the other from our own laboratory [99 ]. As known, leukocyte adhesion and migration are essential steps for the antimicrobial and cytotoxic functions of leukocytes. These studies, along with other studies in nonleukocytic cells, which we shall also discuss, are opening a new paradigm for the role of PLD in cell physiology.
In the first study, Iyer et al. [98 ] demonstrated that adherence of human (PLB, TH1-P) and murine (RAW) myeloid-macrophage cell lines to fibronectin, fibrinogen, collagen, and plastic results in activation of PLD. PLD1 activity is enhanced rapidly upon cell adhesion, and this regulates the initial stages of neutrophil and macrophage adhesion. Conversely, inhibition of PLD activity leads to reduction of cell adhesion. To associate cell adhesion of macrophages and neutrophils with the extracellular membrane, this study also showed that integrins are involved with PLD activity. The second study by Lehman et al. [99 ] linked PLD1 and PLD2 isoforms to chemotaxis. PLD1 was found to be activated by CXCR1, and PLD2 was activated by CXCR2 or CXCR1 in HL-60 cells induced to express the neutrophilic phenotype. During transient expression of PLD1, chemokinesis and chemotaxis toward IL-8 and fMLP (but not toward ENA-78) were increased, suggesting that PLD1 may be regulated differentially by chemoattractants. Regardless, PLD1 and PLD2 appear to be required for leukocyte chemotaxis, as the RNA interference-mediated depletion of either isoform eliminates the potential of phagocytes to adhere and then migrate along a gradient of chemokines.
Although the consequences of PLD action in downstream signaling mechanisms are still being investigated, adhesion and chemotaxis (both required for the inflammatory actions of leukocytes) are clearly modulated by PLD. Here, we propose a simple model, whereby PLD1 is involved in rolling/adhesion/diapedesis as well as in chemotaxis, and PLD2 seems more specialized for the latter (Fig. 3 ).
 View larger version (18K): [in a new window] |
Figure 3. Cell biology scheme for PLD in cell adhesion and chemotaxis. Early reports [85
, 86
] described PLD participation in cell adhesion, although without conferring specificity of isoforms. The particular implication of the two different PLD isoforms (PLD1 or PLD2) has been clarified only recently, thanks to new molecular tools. PLD1 would play a major role in the initial step of phagocyte-endothelium adhesion [98
], whereas PLD1 and PLD2 are needed for the ensuing chemotaxis in tissues [99
].
|
A recent paper reporting about the role of PLD in actin localization and actin-based motility in Dictyostelium [103 ] is also relevant, as Dictyostelium and leukocytes are functionally and biochemically similar. Zouwail et al. [103] have shown that PLD activity is essential for normal cell activity. Inhibition of PLD with 1-butanol causes aberrant F-actin distribution. With the alcohol, F-actin, which normally accumulates in macropinocytic crowns and at the leading edge, changes to an aggregation-staining pattern throughout the cells. Moreover, disruption of PLD signaling causes reduced phagocytosis as well as fluid-phase endocytosis. The authors have also identified the signaling pathway involved in these effects, showing that the loss of PLD activity leads to inhibition of PI(4,5)P2 synthesis, which serves to emphasize the role of PA as a key second messenger in cell function.
Kim et al. [104 ] have also addressed the role of PLD in adhesion and migration in nonimmune cells. Endogenous PLD1 is present during the organization of the actin-based cytoskeleton in association with cell adhesion and migration. By overexpressing a plasmid that encodes an enzyme-inactive version of PLD1, the authors have found that some of the changes in cellular morphology (accumulation of stress fibers, cell elongation, and cell flattening) are independent of PLD activity. Lehman et al. [99 ] have also indicated that chemokinesis (or random migration) in neutrophils and differentiated HL-60 cells is, likewise, unrelated to PLD activity. However, the similarities with nonimmune cells end there, as chemotaxis in response to IL-8 and other chemokines is dependent on PLD activity. Possibly, the reason for these differences between cell types may be related to the specific nature of inflammatory signals, which are sensed only by leukocytes [105 ] in injured tissues [97 , 106 , 107 ].
QUESTIONS THAT REMAIN UNANSWERED IN PLD BIOLOGY
As PLD is beginning to emerge as an important signaling molecule in key leukocyte functions (phagocytosis and chemotaxis), the mechanism(s) underlying these functionalities will undoubtedly be investigated in more detail. Perhaps, leukocytes could prove to be adequate "tools" to unravel still-unresolved mechanistic issues in the general biology of PLD. In this section, we will consider three issues in detail: whether the actions of PLD are mediated by its enzymatic activity or whether there are additional possible mechanisms, namely interaction with other protein signaling molecules; although many PA substrates have been identified, the specific consensus site of PA binding remains to be delineated; and finally, the realization that PLD could be localized at different places in complex signaling pathways.
The dichotomy of enzyme activity versus protein–protein interaction
As PLD is an enzyme, it is logical to conclude that its actions in the cell are derived from its enzymatic activity, i.e., the synthesis of PA from membrane phospholipids. Substantial research in the field of phospholipases has focused on the PLD-mediated generation of PA by using pharmacological approaches and in a somewhat lesser degree, lipase-inactive mutants of PLD. It has been assumed that overexpression of PLD translates into overproduction of PA. This seems to be the case for PLD2 but not necessarily for PLD1. Even with PLD2, there is only a narrow concentration of PA, which stimulates chemotaxis effectively [99
]. Thus, it is still not clear if spatial-temporal production of PA and/or PLD per se is the modulator of cellular functions. We comment about the idea that PLD can regulate target substrates independently of its catalytic activity, via protein–protein interaction between PLD and its target.
PLD interacts and regulates a wide variety of proteins. The N-terminal region of PLD2, containing the PX (amino acid, 64–192) and PH (amino acid, 210–313) domains, is the most representative region involved in PLD2 protein–protein interactions [18 , 108 109 110 111 112 ]. The interaction of PLD with its cognate partners appears not to be modulated by PLD-mediated generation of PA. Two illustrative examples can be found in two recent works [18 , 81 ]. The PH domain of PLD is sufficient to interact with and regulate the GTPase cycle of dynamin, acting as a guanine-exchange activator protein (GAP) [81 ], and two residues in the PX domain (Y169 and Y179) are involved in the recruitment of the growth factor receptor-binding protein 2 (Grb2)/son of sevenless protein (SOS) complex. In the absence of functional Y179, residue Y169 commands the recruitment of the Grb2/SOS and further activation of the Ras/MAPK signaling cascade, resulting in increased de novo DNA synthesis [18 ]. Dynamin is a known GTPase involved in vesicle-mediated recycling of the EGF receptor (EGFR) [113 ], and Grb2 can be associated with the EGFR and modulate its response to the growth factor [114 115 116 ]. Furthermore, EGF stimulates the endocytosis of its own receptor, which leads to the activation of the Ras/MAPK signaling cascade [115 ]. Four conserved R residues (R128, R145, R165, and R197 in PLD1 and R101, R118, R138, and R170 in PLD2) were identified in the C-terminal regions of the PLD1 and PLD2–PX domains as necessary and sufficient for the GAP activity [81 ], whereas two residues in the same domain of PLD2 were required for PLD2-mediated Grb2/SOS recruitment in vivo [18 ].
It is interesting to note that PLD2 residues R170 (involved in GAP activity) and Y179 (involved in Grb2/SOS recruitment) are part of a perfect match (179YRNY182) for Src homology 2 (SH2)-mediated recruitment of the Grb2/SOS complex [18 ]. This raises the possibility that the GAP activity of PLD2 may be modulated by a Y179-dependent Grb2 recruitment (or vice versa). PX-related functions of PLD2, GAP activity and Grb2 recruitment, are independent of PLD2 catalysis, as mutation of Y179 (to render PLD2-Y179F) or deletion of the last 600 amino acids of PLD2 does not eliminate PLD2 catalysis [18 , 81 ]. There could also be complex, inter-regulatory "loops." For example, PLD2-Y179F (via Y169-mediated recruitment of Grb2) could potentially activate AKT by phosphorylating residues T308 and S473 in a PA-PI-3K-dependent manner (M. Di Fulvio, Kathleen Frondorf, Karen Henkels, J. Gomez-Cambronero, unpublished observations). In turn, PLD2-Y179F-activated AKT could phosphorylate PLD2 on T175 (as this residue is in close proximity of the Grb2-binding site) and inhibit the Ras/MAPK signaling pathway.
At any rate, new results will surely appear broadening the PLD-Grb2 protein–protein regulation example, validating it across other cell types and signaling pathways. It will then be possible to ascertain whether PLD-Grb2 (or similar protein–protein interactions) plays a role in PLD biology, in addition to the well-known actions mediated by its enzymatic activity.
In search of a "PA-binding domain"
Once synthesized by PLD action, PA acts as a lipidic second messenger, and a plethora of target substrates have been defined [77
]. Some substrates are well-known kinases and phosphatases related to the most diverse cellular functions. They include PIP, Raf-1, SH2-containing tyrosine phosphatase 1, mammalian target of rapamycin (mTOR), protein phosphatase 1 c, and PKC. Recent experimental advances within this framework allow us to cite mTOR as a prime example. PA binds to mTOR at a site, which is also targeted by the inhibitor rapamycin [117
]. The premise that PA binds to and activates mTOR is defined currently in the literature as being under control of PLD and PI-3K. Still, other targets are being discovered at a steady pace. Recently, our laboratory has described that PLD2-derived or exogenously administrated PA binds and activates p70S6K independently of mTOR in vivo and in vitro [92].
Nevertheless, the question remains: Where does PA bind to its targets? In spite of the reports mentioned previously, it is still not possible to draw a general picture regarding PA binding to its substrates. This is in stark contrast to, for example, the PH domains, which binds specifically to PI(3,4)P2 and phosphatidylinositol 3,4,5-triphosphate, products of PI-3K action, which serve to dock and activate PDK and AKT. No consensus or binding motif for PA has been described as yet, leaving an important gap in knowledge. In some cases, the PA-binding site overlaps with other recognition sites, lending to the possibility that PA may modulate the cognate function of such proteins. For example, the PA-binding site of Raf-1 overlaps with the ATP-binding site of the kinase, and PA modulates Raf-1 translocation to the plasma membrane, assisting in its activation [118 , 119 ]. Moreover, PA binding to PKC depends on the simultaneous binding to DAG [120 , 121 ]. However, it is not yet known whether PA displaces ATP from a kinase, allowing it to be recruited to the cognate place of activity and once in there, whether ATP displaces PA, resulting in a fully active kinase.
PLD signaling: who regulates what?
As is often the case in cell signaling, it is difficult to ascertain which protein(s) are located upstream or downstream in a particular signaling cascade. In the case of kinases and phospholipases, both important to myeloid biology, there are examples of a kinase phosphorylating PLD and also examples of PLD, via the synthesis of PA, affecting a kinase. The case of MAPK highlights this point (Fig. 4
). PLD and derived lipid products in turn activate MAPK-p42 [118
], MAPK-p38 [122
], and NF-
B signaling [124
]. PLD2 recruits the Grb2/Sos complex, resulting in the activation of the MAPK signaling cascade [18
]. Nevertheless, the converse reaction, i.e., a kinase affecting a phosphatase, also holds true. In cell-free systems, ERKs increase ATP-dependent PLD activity, and direct phosphorylation of PLD2 has been demonstrated [84
]. PLD2, but not PLD1, contributed to fMLP-mediated PLD activity. We suspect that what matters in the cell is that a particular protein is activated at the proper time.
 View larger version (18K): [in a new window] |
Figure 4. Cross-signaling of PLD with other major cellular pathways. PLD can act on other intracellular signaling pathways such as ERK (MAPK-p42/44 and p38) [18
, 118
, 122
]; however, MAPK can also phosphorylate PLD2 [84
]. In some cases, the enzymatic activity of PLD might not even be necessary, as just a protein–protein interaction between Grb2 and the PX domain in PLD2 suffices during crosstalk with the MAPK pathway, leading to an increase in overall cell proliferation [18
]. The enzymatic activity of PLD [i.e., the synthesis of PA (blue arrow)] is conversely needed for other cellular actions, such as leukocyte chemotaxis. In addition to PA binding to the mTOR kinase [117
], it binds to and activates S6K, independently of mTOR [92]. Activation of S6K leads to an increase in chemotaxis [123
]. PCNA, Proliferating cell nuclear antigen.
|
POSSIBLE ADVANTAGES OF LEUKOCYTES IN PLD INVESTIGATION
In addition to normal, primary cell phagocytes, several leukemic or lymphocytic cell lines have proven to be immensely useful tools for dissecting signaling mechanisms, enzymatic activity, and isoform makeup, as they can be conveniently cultured and are amenable to be manipulated with molecular biology techniques. Overexpression of PLD2 results in increased PLD activity in rat basophilic leukemia 2H3 mast cells [126 ] and concomitant degranulation [127 , 128 ]. H2O2 increases PLD activity in mouse lymphocytic leukemic L1210 cells, which express only PLD2 [129 ]. PLD is involved in ATP-induced apoptosis of chronic lymphocytic leukemic cells [130 ], but it does not induce any proliferative effect in immature myeloid cells [124 ]. El Marjou et al. [57 ] have studied PLD during maturation and differentiation of myeloid cells in three different systems: leukemic myeloblasts, peripheral blood neutrophils, and four different human myeloid cell lines. They have concluded that PLD activity increases with maturation and differentiation, regardless of the chemical inducer used, and the terminal induction involved the granulocytic or the monocytic lineage. Our laboratory has found that the increase in basal activity could be ascribed mainly to PLD1 in immature HL-60 cells [41 ].
It seems clear that PLD expression and activity change upon induced maturation of leukemic cells. However, it is not known whether PLD plays a passive role in maturation or whether maturation itself is the consequence of increased PLD expression or activity. It is our opinion that PLD may act as a product of maturation rather than being involved in maturation. This is not to say that PLD is not important in growth and maturation of leukemic cells. It might very well be that PLD is one of the several genes up-regulated by chemically induced maturation of leukemic cells. The mature white blood cell benefits from such regulated gene expression as PLD can carry out specialized functions, such as cell migration and the oxidative burst.
Further research is likely to address whether PLD regulates programmed cell differentiation related to the specialized functions of mature leukocytes. Overall, it is expected that the use of cell lines such as HL-60, RAW-264, THP-1, U-937, and others will continue to be useful in outlining PLD-unresolved, mechanistic issues.
FUTURE DIRECTIONS
In our opinion, the future areas for investigators working on PLD will most likely be centered around the unresolved issues in PLD biology, which we have cited in Questions That Remain Unanswered in PLD Biology—whether enzymatic activity is enough to explain all PLD function in the cell (or whether protein–protein interaction with other signaling elements are also required)—along with the still-needed definition of a PA consensus site in target molecules and also, the fine-tuning of PLD localization in signaling in ever-more complex networks. The recent discovery that PLD plays a role in adhesion and chemotaxis in leukocytes [98 , 99 ] and in other mobile cells [103 , 104 ] will undoubtedly lead to the realization that leukocytes can be an ideal tool for unraveling the biological role of PLD. A detailed knowledge of the leukocyte migratory machinery may lead to the discovery of cell-specific modulators of inflammation. Further conceptual advances in designing efficient methods of differentiating PLD1 and PLD2 in specific enzyme assays are currently needed, as is the identification of isoform-specific inhibitors. The differential role of PLD1 versus PLD2 in cell physiology has only now begun to be clarified in leukocytes, as PLD1 seems more relevant for the initial stages of adhesion and PLD2 for chemotaxis. Although it is clear that PLD enzymes participate in the inflammatory functions of leukocytes, the molecular details and mechanisms remain, understandably. Even so, normal leukocytes and their leukemic cell line counterparts have a proven record of being invaluable and powerful tools for cell biologists to gain an understanding of these issues.
ACKNOWLEDGEMENTS
This work has been supported by a grant from the National Institutes of Health (HL056653) to J. G-C. We thank Kathleen Frondorf and Karen Henkels for their help in editing the manuscript.
Received January 17, 2007; revised February 20, 2007; accepted March 20, 2007.
REFERENCES
J. Biol. Chem. 272,3860-3868 in RBL-2H3 cells Mol. Biol. Cell 13,1252-1262 receptor ligation Blood 98,1448-1455 S-stimulated polymorphonuclear leukocytes Blood 103,2363-23681 and phospholipase D during the differentiation of human monocytes to dendritic cells Int. Immunopharmacol. 4,911-920[CrossRef][Medline]RIIIA lymphocyte-mediated cytotoxicity Blood 106,577-583
during Fc receptor-dependent phagocytosis requires the C1B domain and phospholipase C-1 Mol. Biol. Cell 17,799-813RI to nicotinamide adenine dinucleotide phosphate (reduced form) oxidative burst and immune complex trafficking requires the activation of phospholipase D1 Blood 98,3421-3428, stimulates phospholipase D activity Blood 85,522-531-Synuclein interacts with phospholipase D isozymes and inhibits pervanadate-induced phospholipase D activation in human embryonic kidney-293 cells J. Biol. Chem. 277,12334-12342 1 with phospholipase D2 is important for epidermal growth factor signaling J. Biol. Chem. 278,18184-18190-actinin in an ADP-ribosylation factor-reversible manner J. Biol. Chem. 275,21295-21301 Biochemistry 42,11661-11668[CrossRef][Medline] to the plasma membrane in RBL-2H3 cells Mol. Biol. Cell 14,4885-4895B activation in tumor-necrosis-factor--treated immature acute-myeloid-leukaemia cells Biochem. J. 351,459-467[CrossRef][Medline] Mol. Pharmacol. 66,293-301Related Article
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
