| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
-derived phosphoinositide metabolite
Department of Cell Biology and Oncology, Consorzio Mario Negri Sud, Santa Maria Imbaro, Italy;
* Santa Lucia Foundation/CNR, Rome, Italy; and
Dipartimento di Biologia e Patologia Cellulare e Molecolare, Università Federico II di Napoli, c/o Istituto di Endocrinologia e Oncologia Sperimentale del CNR, Naples, Italy
1Correspondence: Department of Cell Biology and Oncology Consorzio Mario Negri Sud, Via Nazionale 8, 66030 Santa Maria Imbaro, Chieti, Italy. E-mail: mariggio@negrisud.it or corda{at}negrisud.it
ABSTRACT
The phosphoinositides have well-defined roles in the control of cellular functions, including cytoskeleton dynamics, membrane trafficking, and cell signaling. However, the interplay among the phosphoinositides and their diffusible derivatives that originate through phospholipase A2 action (the lysophosphoinositides and glycerophosphoinositols) remains to be fully elucidated. Here we demonstrate that in PCCl3 rat thyroid cells, the intracellular levels of glycerophosphoinositol are finely modulated by ATP and norepinephrine through the P2Y metabotropic and 
-adrenergic receptors, respectively. The enzyme involved here is phospholipase A2 IV (PLA2IV), which in these cells specifically hydrolyzes phosphatidylinositol, forming lysophosphatidylinositol, glycerophosphoinositol, and arachidonic acid. This receptor-mediated activation of PLA2IV leads to stimulation of PCCl3 cell growth. The involvement of a PLA2IV-mediated pathway is demonstrated by inhibition of the increase in intracellular glycerophosphoinositol levels and cell proliferation by specific inhibitors, RNA interference, and overexpression of the dominant-negative PLA2IV1–522. Modulation of PCCl3 cell growth is not seen with inhibitors of arachidonic acid metabolism. In conclusion, these data characterize glycerophosphoinositol as a mediator of the purinergic and adrenergic regulation of PCCl3 cell proliferation, defining a novel regulatory cascade specifically involving this soluble phosphoinositide derivative and widening the involvement of the phosphoinositides in the regulation of cell function.—Mariggiò, S., Sebastià, J., Filippi, B. M., Iurisci, C., Volonté, C., Amadio, S., De Falco, V., Santoro, M., Corda, D. A Novel pathway of cell growth regulation mediated by a PLA2-derived phosphoinositide metabolite.
Key Words: glycerophosphoinositol • phosphatidylinositol • thyroid cells • phospholipases • cell signaling
THE PHOSPHOINOSITIDES (PIS) are membrane bilayer components with important roles in membrane structure and the control of cellular functions (1)
. This control is exerted either directly by the differently phosphorylated PIs or after hydrolysis of the PIs, which is mediated by phospholipases and leads to the formation of active compounds such as diacylglycerol (DAG), the inositol phosphates (InsPs), lysophosphoinositides, phosphatidic acid, and the glycerophosphoinositols (2
, 3)
. Virtually all cell functions are directly or indirectly under the control of these inositol-containing molecules: the PIs are part of the mechanisms for recruitment, recognition, and docking of protein complexes in different cell compartments, and the PI-derived diffusible messengers (such as the InsPs and the glycerophosphoinositols) arise from activation of various specific phospholipases by signaling cascades initiated at the plasma membrane by hormonal stimulation. Due to this relatively large number of biologically active, inositol-containing molecules, definition of their metabolism, function, and localization in different cell compartments continues to provide information toward elucidating the interplay between their formation pathways and their activities.
The glycerophosphoinositols are water-soluble compounds that result from enzymatic deacylation of the membrane PIs (reviewed in ref. 3
), and although they were originally identified and studied in the context of Ras-transformed cells (4
, 5)
, they can be detected in virtually all cell types. In intact cells, the intracellular levels of the glycerophosphoinositols can be modulated by calcium ionophore via the activation of a phospholipase A2 (PLA2) and by receptor activation (3
, 6)
, whereas a phosphodiesterase (GDE1/MIR16) regulated by G-protein-coupled receptors has been shown to preferentially hydrolyze glycerophosphoinositol (GroPIns) to inositol and glycerol phosphate (7)
.
The exogenous administration of the glycerophosphoinositols to cells affects specific functions in different cell lines. In particular, we have shown that glycerophosphoinositol 4-phosphate (GroPIns4P) inhibits cyclic AMP (cAMP) production; as a consequence, it can affect cAMP-dependent functions such as cell proliferation in epithelial thyroid cells and fibroblasts (3
, 6
, 8
, 9)
. GroPIns4P can also activate small GTPases of the Rho family and induce the formation of actin ruffles and stress fibers (10)
. Both GroPIns and GroPIns4P can inhibit invasion of the extracellular matrix (ECM) by tumor cell lines (epithelium-derived MDA-MB-231 breast carcinoma and A375MM melanoma cells) by decreasing the ability of these cells to degrade the matrix components (11)
.
On the basis of these effects of exogenously added glycerophosphoinositols, we investigated whether GroPIns levels can be modulated physiologically, and thus lead to the regulation of specific cellular pathways. To this end, we initially screened selected cell lines for those where increased levels of the glycerophosphoinositols are seen after receptor activation; these were then investigated for the pathway of formation and the intracellular actions of GroPIns.
We focused this study on a continuous line of follicular thyroid cells (PCCl3) particularly suitable for studying the control of cell proliferation and differentiation in an epithelial cell setting (12)
. These cells express a set of thyroid-specific genes [thyroglobulin, thyroperoxidase (TPO), and the sodium/iodide symporter (NIS); ref. 12
, 13
], expression of which is under the control of thyroid-specific transcription factors (TTF-1, a homeodomain-containing protein, and PAX8, a paired-domain transcription factor) (13)
. In addition, PCCl3 cells require a mixture of six hormones (6H, including TSH) for cell proliferation (12)
, and maintenance of the thyroid differentiated phenotype is strictly connected to the control of proliferation (12
, 14)
. In most thyroid model systems, cell growth is mainly controlled by the TSH-dependent cAMP pathway and growth factor-dependent tyrosine kinase receptors, as well as other factors like ATP, norepinephrine, and the bioactive lipid sphingosine 1-phosphate that activate separate signaling cascades, including a Ca2+ response (14
15
16
17)
.
Our results define a novel regulatory cascade involving PLA2IV, which, once activated by purinergic and adrenergic receptors, controls epithelial cell growth through the hydrolysis of phosphatidylinositol (PtdIns) to GroPIns.
MATERIALS AND METHODS
Reagents and antibodies
Dulbecco’s modified Eagle medium (DMEM), HBSS, and calf serum (CS) were from Life Technologies, Inc., Brl (Grand Island, NY, USA). FBS was from Biochrom KG (Berlin, Germany), insulin from Elanco Products (Indianapolis, IN, USA), L-noradrenaline hydrochloride from Fluka (Buchs-SG, Switzerland), PD98059, SB203580, and sPLA2IIA inhibitor-I from Calbiochem (San Diego, CA, USA), U0126 from Promega (Madison, WI, USA), proline from ACROS (Morris Plains, NJ, USA), [6-3H]-thymidine (18.4 Ci/mmol) from Perkin Elmer Life Science (Boston, MA, USA), [3H]-myo-inositol (16 Ci/mmol), and [5,6,8,9,11,12,14,15-3H]-arachidonic acid (210 Ci/mmol) from Amersham Pharmacia (Piscataway, NJ, USA). Pyrrophenone was provided by K. Seno (Shionogi Research Laboratories, Osaka, Japan) (18)
. The cPLA2(N-216), ERK1(K-23), p38(H-147), P2X5(V-19) antibodies were from Santa Cruz (San Diego, CA, USA), the 12D4 antiphospho-MAP kinases (ERK1/2) clone from Upstate Technology (Lake Placid, NY, USA), antiphospho-p38(Thr180/Tyr182) and anti-rabbit IgG from Cell Signaling Technology (Beverly, MA, USA), and anti-goat IgG from Chemicon International Inc. (Temecula, CA, USA). The antiglyceraldehyde phosphate dehydrogenase antibody (Ab) (anti-GAPDH) was from Biogenesis Ltd. (Sandown, NH, USA). The affinity-purified antibodies for P2X1,2,3,4,6,7 and P2Y1,2,4,6,12 (Alomone, Jerusalem, Israel) were raised against purified peptides corresponding to epitopes not present in other known proteins (P2X2:457–472; P2X4:370–388; P2X6: 363–379; P2X7:576–595; P2Y1:242–258; P2Y2:227–244; P2Y4:337–350; P2Y6:311–328). All other reagents were at the highest purities available from Sigma-Aldrich (Milan, Italy).
Cell lines and generation of transfectants
PCCl3 rat thyroid cells were grown in Coon’s modified Ham’s F12 medium, 5% CS, and a six-hormone mix (6H) that included thyrotropin (TSH) (12
, 14)
. In some procedures, either TSH (5H medium) or TSH and somatostatin (4H medium) were withdrawn. Chinese hamster ovary (CHO) cells were grown in DMEM, 10% FBS, 34 µg/ml proline, 2 mM L-glutamine. Transfections were performed with Lipofectamine-plus (Invitrogen, Burlington, Ontario, Canada) according to the manufacturer’s instructions. PLA2IVwt(pBK--CMV) and PLA2IV1–522(pCDNA3.1/Zeo) plasmids were provided by I. Kudo (Showa University, Tokyo, Japan) (19)
.
PCCl3 cells were transfected by electroporation (20)
. Antibiotic-resistant clones were isolated by limiting dilution. Transfection efficiency and levels of ATP-activated PLA2 were verified by Western blot and arachidonic acid release assays. A single representative clone was chosen from each transfection from among the five to eight obtained: the Vclone was transfected with an empty vector, and WTclone and DNclone with PLA2IVwt and PLA2IV1–522, respectively. PLA2IV sequence-specific silencing was performed as described (21)
, using an electroporated small interfering RNA (siRNA) (20)
and an siCONTROL nontargeting duplex (Dharmacon Inc., Lafayette, CO, USA) as control. After electroporation, cells were seeded in serum-free growth medium for 4 h, then 10% CS was added for 72 h.
Labeling and analysis of [3H]-inositol-containing compounds
PCCl3 cells were labeled for 48 h in Medium 199, 1% CS, 6H, 5 µCi/ml [3H]-myo-inositol. Transfected CHO cells were labeled for 24 h in Medium 199 containing 5 µCi/ml [3H]-myo-inositol. Cell stimulation and extraction, HPLC separation of [3H]-inositol-labeled water-soluble metabolites, and thin-layer chromatography of organic phases were as described previously (22)
. In the figures, the HPLC and TLC cpm were expressed as percentages of their individual controls for calculations of means. For completeness, in unstimulated cells the total [3H]-inositol-containing aqueous compounds were 429,932 ± 30,881 cpm/well, of which 4476 ± 965 cpm/well were in [3H]-GroPIns. Similarly, the total [3H]-inositol-containing lipid levels were 48,765 ± 18,810 cpm/well, of which 484 ± 125 cpm/well were in[3H]-lysophospatidylinositol ([3H]-LysoPtdIns).
Labeling and analysis of [3H]-arachidonic acid-containing compounds
For arachidonic acid release assays, PCCl3 cells were labeled for 18 h in growth medium with 0.1 µCi/ml [3H]-arachidonic acid. [3H]-Arachidonic acid released in the medium was quantified in triplicate, as described previously (23)
; radioactivity released from cells is expressed as percentages of total incorporated radioactivity. As above, in unstimulated cells across the full experimental data, the total [3H]-arachidonic acid-containing lipid levels were 504,106 ± 36,975 cpm/well, of which 120,930 ± 28,211 cpm/well were in [3H]-PtdIns.
For analysis of [3H]-arachidonic acid-containing lipids, PCCl3 cells were labeled for 18 h in growth medium with 0.8 µCi/ml [3H]-arachidonic acid. Cell stimulation, extraction, and HPLC separation of [3H]-arachidonic-labeled lipids were as described previously (24)
. This labeling was adopted to improve the detection of PtdIns hydrolysis after activation of PLA2.
Western blot
Cells were lysed in 10 mM Tris-HCl, pH 7.5, 5 mM EDTA, 7 mM MgCl2, and protease inhibitors, then homogenized and centrifuged (700 g, 10 min, 4°C). Fifty micrograms of supernatant were analyzed by 10% SDS-PAGE and transferred to nitrocellulose Protran (PerkinElmer Life Science, Boston, MA, USA). Activities of the extracellular signal-regulated kinases 1/2 (ERK1/2) were determined as described (25)
.
In vitro PLA2 activity assay
Sf9 cells (American Type Culture Collection-LGC Promochem s.r.l., Milan, Italy) were transfected with the pAcHLT-PLA2IV vector (kindly provided by C. Leslie, National Jewish Medical and Research Center, Denver, CO, USA), using the BaculoGoldTM transfection kit from PharMingen (San Diego, CA, USA). Recombinant His-tagged PLA2IV was then purified using nickel-nitrilotriacetic acid metal affinity chromatography (Quiagen, Milan, Italy), following the manufacturer’s instructions.
Phospholipase and lysophospholipase activities were measured using 700 µM PtdIns, plus 50,000 cpm of [3H]-PtdIns as a tracer, in an assay mixture (40 µl) containing 80 mM HEPES (pH 7.4), 150 mM NaCl, 10 mM CaCl2, 4 mM Triton X-100, 30% glycerol, 1 mg/ml BSA. The reaction was initiated by adding 3 µg purified PLA2IV or 20 ng bee venom PLA2 (Cayman Chemical Company, Ann Arbor, MI, USA), and terminated after 3 h at 25°C by addition of 750 µl chloroform/methanol/HCl 1 M (1:1:1, v/v/v). [3H]-GroPIns in the aqueous phase was analyzed by HPLC, and LysoPtdIns in the organic phase by thin-layer chromatography, as described (22)
.
[3H]-Thymidine incorporation
Cells were starved for 48 h in Coon’s modified Ham’s F12 medium with 0.3% BSA, 2 mM L-glutamine, 100 U/ml penicillin, 0.1 mg/ml streptomycin. Stimuli (48 h in 4H medium; quadruplicates) and subsequent [3H]-thymidine incorporation were as described previously (16)
. The stable transfectants were starved as above before addition of 5H medium (replaced every 48 h). Blinded cell counting was done in a Neubauer chamber.
Intracellular cAMP assay
Cells were stimulated in HBSS with added Mg2+ and Ca2+ (HBSS++), 10 mM HEPES, pH 7.4, 0.4% BSA, 0.5 mM 3-isobutyl-1-methylxanthine for 30 min at 37°C. cAMP content was measured in triplicate using a commercial "Cyclic AMP [3H]-BiotrakTM Assays System" (Amersham Pharmacia).
Determination of [Ca2+]i
Cells were loaded with 2 µM Fluo-3 acetoxymethyl ester (AM) (Molecular Probes, Eugene, OR, USA) in 20 mM HEPES, pH 7.2, 118 mM NaCl, 4.6 mM KCl, 10 mM glucose, 0.4 mM CaCl2 for 30 min at 37°C, as described previously (26)
. Fluorescence was measured (27)
in single cells with a IX70 microscope (Olympus, Hamburg, Germany) equipped with a TILL Photonics imaging system (Gräfelfing, Germany).
RNA extraction and RT-polymerase chain reaction (RT-PCR)
Total RNA was extracted from PCCl3 cells and TaqMan real-time polymerase chain reaction (PCR) was performed (28)
using the following primers: thyroglobulin forward, 5'-GAGTGATGCTCCCAGCTTCT-3'; thyroglobulin reverse, 5'-AGTTCCTGGTGGCTGAAATG-3'; TTF1 forward, 5'-CTACTGCAACGGCAACCTG-3'; TTF1 reverse, 5'-CTCATATTCATGCCGCTCG-3', Na/I symporter forward, 5'-CAGCACTGCATCCACCAG-3'; Na/I symporter reverse, 5'-CGTGAAGGCGCCTAGTAGAG-3'; PAX8 forward, 5'-AGCAGCAGTAGTGGTCCTCG-3'; PAX8 reverse, 5'-CCGTCATCCAGGGTACTGTT-3'.
Statistical analysis
Data are presented as mean ± SE of results from at least three independent experiments. Statistical analysis was by Student’s t test.
RESULTS
Receptor-mediated hydrolysis of the membrane phosphoinositides
Hydrolysis of the membrane PIs by PLA2 activities has been shown to lead to production of the glycerophosphoinositols and lysophosphoinositides in some cell lines (3)
. In PCCl3 cells, levels of these compounds were quantified after equilibrium labeling with [3H]-myo-inositol (see Materials and Methods; ref. 22
). Addition of the ionophore A23187 (1 µM) increased GroPIns levels in these cells to 2.5-fold that of control, whereas 10 µM ATP and 20 µM L-noradrenaline hydrochloride (NE) resulted in an 
2-fold increase of the control (Fig. 1
A). These effects were concentration (1–100 µM) and time dependent (see Fig, 1B
and legend to Table 1
). For NE, the effect was observed after 5 min, with maximum stimulation after 120 min; a level higher than the control was maintained for at least 8 h (Fig. 1B
). Other agonists, including TSH (10–7 M) and carbachol (1 µM), did not modulate GroPIns levels in these cells, indicating that ATP and NE activate specific signaling cascades that lead to PtdIns hydrolysis (see below).
|
|
In parallel and under identical conditions, we examined whether other PIs metabolites were formed upon ATP and NE stimulation of PCCl3 cells (10 and 20 µM, respectively, concentrations used from now on unless otherwise specified). As expected, there was an increase to 2-fold basal levels of LysoPtdIns, the first product of PtdIns deacylation by PLA2 activity (Fig. 1C
). The formation of GroPIns and LysoPtdIns was paralleled by a decrease in PtdIns levels (Fig. 1D
), as detected by labeling of the arachidonic acid moiety (see Materials and Methods). NE also induced a rapid increase (maximum within 2 min) in inositol 1,4,5-trisphosphate formation (reaching 2-fold the basal levels), paralleled by a decrease in PtdIns4,5P2 (40%). ATP behaved in a similar manner (data not shown).
ATP and NE stimuli also produced increases in GroPIns4P (Fig. 1E
), which can arise from hydrolysis of phosphatidylinositol 4-phosphate or phosphorylation of GroPIns, although under normal growth conditions detection of GroPIns4P production in cells generally is hampered by its dephosphorylation to GroPIns as it is formed (3
, 6
, 29)
. GroPIns4P levels were thus evaluated in the presence of a generic phosphatase inhibitor, sodium orthovanadate, whereby ATP, NE, and A23187 increased GroPIns4P levels to the same extent as GroPIns and LysoPIns (Fig. 1E
). Under these experimental conditions, glycerophosphocholine and lysophosphatidylcholine, the two products of PLA2 action on phosphatidylcholine (30)
, were not detected (data not shown).
These data thus indicate that in PCCl3 cells, ATP and NE stimulate signaling pathways that lead to hydrolysis of the PIs, with the production of LysoPtdIns, GroPIns, and GroPIns4P.
Activation of PLA2IV is required for GroPIns formation
We next sought to identify the specific PLA2 involved in this receptor-mediated GroPIns and LysoPtdIns formation in PCCl3 cells. The first criterion to discriminate among different classes of PLA2 was to evaluate their dependence on Ca2+. The ATP- and NE-induced increases in GroPIns were completely inhibited by preincubation of PCCl3 cells with the intracellular Ca2+ chelator BAPTA-AM (Table 1)
, thus indicating the Ca2+ dependency of this PLA2 activation. Since under physiological conditions the increase in intracellular Ca2+ concentration ([Ca2+]i) could derive from a parallel activation of phospholipase C (PLC), the specific PLC inhibitor U-73122 was used; here, the ATP-induced inositol 1,4,5-trisphosphate formation and the consequent rise in [Ca2+]i were completely blocked (see supplemental material), as were GroPIns and LysoPtdIns production, further underlining the Ca2+ requirement for their formation. The inactive analog U-73343 was ineffective (Table 1)
.
In addition, specific inhibitors of Ca2+-independent PLA2 (iPLA2; 10 µM palmitoyl trifluoromethyl ketone, 1 µM bromo-enol lactone [BEL]) and secretory PLA2 (sPLA2; 50 µM sPLA2IIA inhibitor-I) did not inhibit agonist-stimulated GroPIns production (Table 1)
. However, at a concentration known to be nonspecific for the different PLA2 isoforms (50 µM), BEL inhibited ATP and NE stimulation of GroPIns production by >65% (Table 1)
. Finally, pyrrophenone, a specific PLA2IV inhibitor (18)
, completely blocked ATP- and NE-induced GroPIns and LysoPtdIns production at all time points examined (while inhibiting A23187 stimulation by 75%), indicating that the PLA2 involved in this stimulation of the production of GroPIns and LysoPtdIns is the cytosolic, calcium-dependent PLA2IV (Fig. 2
A–D, Table 1
).
|
Receptor-mediated activation of PLA2IV leads to the parallel formation of arachidonic acid and GroPIns
In intact cells, PLA2IV has been shown to be selective for phospholipids substituted in the sn-2 position with arachidonic acid (31)
. A release of arachidonic acid thus should accompany LysoPtdIns and GroPIns formation in our system. Indeed, under the experimental conditions above, both ATP and NE induced arachidonic acid release in a time-dependent manner: with ATP, 70% of the maximal effect was seen after 2 min of stimulation; a plateau was reached within 10 min that lasted at least 20 min. The maximal effects of ATP and NE produced increases to 7- and 2-fold the control (Fig. 2E
) whereas A23187 (1 µM) produced an 30-fold stimulation (Table 2
).
|
The inhibitors used to analyze GroPIns formation were also tested on arachidonic acid release, and, not surprisingly, showed the same pattern of response. Thus, both the Ca2+ chelator BAPTA-AM and the PLC inhibitor U-73122 prevented agonist-induced arachidonic acid release, but the inactive analog U-73343 was ineffective (Table 2)
. Pyrrophenone (0.1–0.5 µM) inhibited ATP- and NE-stimulated arachidonic acid release by up to 90% (Fig. 2F
).
Correlation between PLA2IV expression and GroPIns formation
To further substantiate the involvement of PLA2IV in GroPIns formation, its wild-type (WT) (PLA2IVwt) and dominant-negative (PLA2IV1–522; able to compete with PLA2IVwt for membrane binding; see ref. 19
) forms were overexpressed in CHO cells, a cellular model selected for its high transfection efficiency. After confirmation of the overexpression of PLA2IVwt and the PLA2IV1–522 mutant by Western blot (Fig. 3
A), ATP stimulation (after subtraction of their respective basal levels) showed a 66% increase for PLA2IVwt and a 44% decrease for PLA2IV1–522 over the empty vector response (Fig. 3B
). The parallel ATP stimulation of GroPIns production showed a 207% increase and a 63% decrease, respectively (Fig. 3C
). No significant changes in LysoPtdIns levels were seen under these conditions, although changes were evident in permanently transfected clones (see below).
|
The same analysis was performed in a series of permanently transfected PCCl3 cell clones expressing either the PLA2IV constructs or the empty vector (see Materials and Methods). In the clone overexpressing PLA2IVwt (WTclone; representative of 5 isolated), ATP-induced GroPIns and LysoPtdIns increased to 363% and 238%, respectively, compared with the ATP-stimulated empty vector clone (Vclone, representative of 8 isolated); this stimulation was completely blocked by pyrrophenone (Fig. 3D, E
). The clone overexpressing PLA2IV1–522 (DNclone, representative of 5 isolated) showed instead an 43% inhibition of ATP-induced GroPIns synthesis and had no effect on LysoPtdIns production with respect to the Vclone ATP stimulation (Fig. 3D, E
). Arachidonic acid levels monitored under the same conditions consistently showed an increase in the ATP effect in Vclone, which was inhibited by pyrrophenone, and a decrease in the DNclone (Fig. 3F
). Compared with the control (see below) and mock-transfected (data not shown) PCCl3 cells, activation rates of ERK1/2 and p38 in these clones after ATP addition were not different, indicating that these clones maintain normal control of ERK1/2 and p38 activation (data not shown).
Finally, we used an siRNA specific for PLA2IV; this decreased the PLA2IV levels by 40% (compared with siCONTROL nontargeting-duplex-treated cells) without affecting levels of the highly homologous PLA2IVß, monitored as a control (Fig. 4
A). This loss of PLA2IV was reflected in an 26% decrease in ATP-induced arachidonic acid release (Fig. 4B
). At the same time, basal GroPIns levels were reduced by 25%, and ATP and NE stimulation of GroPIns production was decreased by 28% and 33%, respectively (Fig. 4C
).
|
In line with previous studies (ref. 31
and references therein), in vitro experiments performed using the recombinant protein and purified lipids (see Materials and Methods) showed that PLA2IV possesses phospholipase and lysolipase activities that hydrolyze PtdIns to LysoPtdIns and LysoPtdIns to GroPIns, respectively (Fig. 4D-F
; bee venom PLA2 is shown as control).
Collectively, the different approaches above indicate that PLA2IV is the enzyme involved in the formation and modulation of GroPIns levels in PCCl3 cells. This PLA2 isoform can perform both deacylation steps, as shown by in vitro assays using PtdIns and LysoPtdIns as substrates.
The PLA2IV leading to GroPIns formation is under MAP kinase control
ERK1/2 MAP kinases and p38 stress-activated protein (SAP) kinase are well-known regulators of PLA2IV (31)
. Indeed, ATP-, NE-, and A23187-induced GroPIns production was partially blocked (50–90%) by inhibitors of ERK1/2 and p38, including PD-98059, U0126, and SB203580 (Table 1)
. In parallel, ATP- and NE-induced arachidonic acid release was inhibited by 67% and 35% and by 90% and 70% by U0126 and SB203580, respectively (Table 2)
, in line with their effects on agonist-induced formation of GroPIns.
Accordingly, these receptor agonists also induced an increase in ERK1/2 and p38 phosphorylation and consequent phosphorylation of PLA2IV (Fig. 5
). ERK1/2 phosphorylation was time dependent: it was evident after 4 to 5 min of stimulation with either agonist (Fig. 5A
) and was maintained maximally for up to 10 to 15 min for ATP (Fig. 5C
); it was mostly switched off after 10 min with NE (data not shown) and remained detectable for up to 3 h with ATP (Fig. 5C
) and NE (data not shown).
|
Under normal growth conditions (and under the starving conditions used for inositol labeling), 50% of PLA2IV was in the phosphorylated form (31)
and became fully phosphorylated within 5 min of stimulation with either ATP or NE (Fig. 5A
). Activation of ERK1/2 is required here, since treatment with the inhibitor U0126 prevented both ATP- and NE-induced ERK activation and PLA2IV phosphorylation (Fig. 5B
). Activation of PLA2IV phosphorylation was rapid and was sustained for at least 3 h (Fig. 5C
). Under these conditions, carbachol and TSH, which did not induce GroPIns formation (see above), did not activate either ERK1/2 or PLA2IV phosphorylation, indicating that their specific receptors are uncoupled from this signaling pathway (data not shown).
An NE- and ATP-stimulated, inositol-1,4,5-trisphosphate-dependent increase in [Ca2+]i (see below and Table S1 in supplemental material) is required for ERK1/2 activation and consequent PLA2IV phosphorylation, as confirmed by treatment with BAPTA-AM and PLC inhibitors, which prevented ERK1/2 and PLA2IV phosphorylation (data not shown).
Taken together, these data indicate that the PLA2IV that mediates this agonist-stimulated GroPIns and LysoPtdIns formation is regulated by the MAP and SAP kinases (31)
.
PLA2IV involvement in the regulation of cell differentiation
We investigated whether this receptor activation of PLA2IV and the subsequent formation of these PIs metabolites affect the expression of differentiated gene markers in PCCl3 cells. The effects of exogenous addition of GroPIns and receptor-induced intracellular GroPIns production on mRNA levels of the thyroglobulin, PAX8, TTF1, TPO, and NIS genes were evaluated by real-time RT-PCR. Exogenous addition of GroPIns (300 µM) (Table 3
) had no significant effects on the above mRNA levels in PCCl3 cells; NE alone exerted only minimal down-modulatory effects on differentiation marker mRNA levels (semiquantitative PCR; data not shown). The combined treatment with NE and pyrrophenone further reduced mRNA levels of thyroglobulin, PAX8, and TTF1; concomitant addition of exogenous GroPIns (100 µM) partially reversed this effect on TTF1 mRNA (Table 3)
. Semiquantitative analysis performed at later time points confirmed these findings. Thus, although NE can modestly affect thyroid differentiation, modulation of the PLA2IV pathway did not impair the expression of thyroid differentiation markers. Rather, GroPIns can at least partially antagonize this effect.
|
PLA2IV in the regulation of cell proliferation
The receptor-stimulated modulation of cell proliferation was then examined on the basis that PLA2IV and the PIs metabolites that form through its activation could intervene in this process in epithelial cells, as proposed (32
, 33)
. With [3H]-thymidine incorporation in PCCl3 cells in 4H medium (see Materials and Methods), addition of ATP (100 µM) and NE (20 µM) resulted in 2-fold the control [3H]-thymidine incorporation (Fig. 6
A). This effect was prevented by pyrrophenone pretreatment, suggesting that a PLA2IV product has a role in this modulation. Pyrrophenone did not affect [3H]-thymidine incorporation per se (Fig. 6A
).
|
The involvement of prostaglandin E2 (PGE2), an arachidonic acid metabolite in FRTL5 cells (33)
, in ATP- and NE-induced cell proliferation was excluded. This was seen through lack of activity of the general cyclooxygenase inhibitor lysinate acetylsalicylic acid (300 µM), which did not influence basal and hormone-induced arachidonic acid production but was active in preventing arachidonic acid metabolism (data not shown). Moreover, the direct addition of PGE1, PGE2, and tromboxane B2 (TXB2) in a range of concentrations from 10 to 300 nM did not induce any significant effects on PCCl3 cell growth (with concentrations 
300 nM being toxic for the cells). Instead, exogenous addition of arachidonic acid (100 nM and 1 µM) induced an increase in [3H]-thymidine incorporation of 1.5-fold that of control (Fig. 6A
). Another potentially active PLA2 product is platelet-activating factor (PAF), which can be formed and secreted after PLA2 activation and acts in an autocrine/paracrine manner (34)
. However, under the above assay conditions, 0.1 to 10 µM PAF were ineffective (data not shown). Instead, the exogenous addition of GroPIns (100–400 µM) and LysoPtdIns (10 µM) induced an increase in [3H]-thymidine incorporation of from 1.5- to 2-fold that of control (Fig. 6A
).
The role of PLA2IV in PCCl3 cell growth was further confirmed using the V, WT, and DN PCCl3 clones described above (Fig. 6B
). In 5H medium (lacking TSH, the main growth factor for PCCl3 cells; see Materials and Methods), the WTclone had a higher growth rate than the Vclone; this difference was abolished when the WTclone was incubated with pyrrophenone (Fig. 6C
), indicating that this greater growth rate is a consequence of PLA2IV activity. In complete 6H medium, TSH stimulation overcame this PLA2IV effect, and both the Vclone and the WTclone had similar growth rates. The DNclone did not grow at all in the absence of TSH (5H medium; Fig. 6B
). This indicates that by blocking endogenous PLA2IV, the deleted mutant inhibits PCCl3 cell growth in the DNclone, thus confirming the relevance of the PLA2IV pathway in the modulation of cell proliferation.
Purinergic and adrenergic receptor subtypes coupled to PLA2IV activation
A series of experiments was undertaken to elucidate the subtypes of purinergic and adrenergic receptors involved in the activation of PLA2IV in PCCl3 cells. The results of the pharmacological and signaling analyses converged on the role of the P2Y and -adrenergic receptors in the formation of GroPIns (see supplemental material).
DISCUSSION
We have defined a novel regulatory cascade that specifically links receptor activation of PLA2IV to regulation of cell proliferation through formation of the PIs metabolite GroPIns, thus adding to the number of regulatory pathways controlled by inositol-containing molecules.
PLA2IV is widely expressed in cells, and its function has been related to the metabolism of arachidonic acid and prostaglandin formation (31)
. In both in vitro assays and intact cells, PLA2IV has been shown to hydrolyze phosphatidylcholine (30)
. Our data show that in epithelial cells, PLA2IV can specifically act on the membrane PIs, as indicated by the parallel formation of LysoPtdIns, GroPIns, and arachidonic acid and by the decrease in PtdIns levels; in contrast, it does not lead to the formation of glycerophosphocholine, suggesting either that phosphatidylcholine is not the preferred substrate for PLA2IV in this cellular system or that a rapid reacylation of lysophosphatidylcholine to phosphatidylcholine takes place (3
, 35)
.
In this PCCl3 thyroid cell line, PLA2IV activity is modulated by specific receptor agonists. NE and ATP, acting through 1- and 2-adrenergic and P2Y receptors, respectively (see supplemental material and Fig. 7
), activate PLA2IV in these cells, unlike other known ligands such as TSH and carbachol, which do stimulate other thyroid cells (see Results and ref. 23
, 36
).
|
The criteria we applied to investigate coupling of the activated receptors to PLA2IV activity and to subsequent hydrolysis of the PIs included the requirement for Ca2+, the use of inhibitors, RNA interference, and overexpression and down-regulation of PLA2IV. Through these we have been able to define the signaling cascade involved in the formation of GroPIns (Fig. 7)
. Thus, by stimulating PLC, the activated purinergic and adrenergic receptors produce the increase in [Ca2+]i required for activation of MAPK and for membrane translocation of PLA2IV (31
, 37)
. This requirement was further demonstrated by use of the ionophore A23187, which mimics receptor activation. The specific inhibition of this pathway by the PLC inhibitor U-73122 and BAPTA-AM indicates that ERK1/2 is activated downstream of this [Ca2+]i increase and that ERK1/2 in turn phosphorylates, and thus activates, PLA2IV (Fig. 7)
. As well as being in agreement with the activation of PLA2IV reported in different cell systems (31)
, these data delineate the specific pathways leading to PLA2-mediated PIs hydrolysis in these cells. Indeed, a similar activation of a PLA activity that specifically forms LysoPtdIns and GroPIns was also reported upon overexpression of PI-transfer protein-, which also correlated with an increased growth rate of NIH3T3 fibroblasts (38)
.
The possibility that the recently cloned murine PLA2 isoforms cPLA2
and cPLA2
abundantly expressed in the thyroid (39)
are involved in this PIs hydrolysis was excluded. This was based on uncoupling of these enzymes from ERK regulation (they lack the consensus sequence for ERK1/2 phosphorylation), which is instead part of this cascade of GroPIns formation. Similarly, the contribution of an sPLA2 (sPLA2IIA) that acts on the PIs in vitro (31)
was excluded based on the lack of effect of the specific inhibitor-I on this receptor-induced GroPIns formation.
Stable transfectants overexpressing either PLA2IV or PLA2IV1–522, a dominant-negative form of PLA2IV, further support this regulatory cascade. Thus, while in the former an increase in receptor-mediated GroPIns formation was seen, in the latter both arachidonic acid release and GroPIns formation were prevented. This appears to be due to efficient translocation to the membrane of the dominant-negative mutant after agonist stimulation, with the consequent competition with the endogenous active enzyme (19)
.
It should be noted that PLA2IV acts as a phospholipase and a lysolipase, as indicated by in vitro studies showing the hydrolysis of PtdIns and LysoPtdIns, and in line with reported observations (31)
. This supports the proposal that one enzyme supports both deacylation steps in PCCl3 cells (Fig. 7)
.
While the above data show that both ATP and NE induce GroPIns formation and arachidonic acid release through PLA2IV activation, the extent of this activation and the effects of a panel of inhibitors acting on this signaling cascade were different. Thus, while ATP and NE produce similar increases in GroPIns levels indicative of a common pathway of activation, the different extent of arachidonic acid release induced by the two agonists suggests that ATP may act on more than one P2 receptor (see supplemental material) and/or be coupled to a different PLA2. Similarly, the [Ca2+]i increase induced by the ionophore A23187 elicited an increase in arachidonic acid that exceeded that of the hormone stimuli (Table 2)
. This could be interpreted as a nonspecific effect of high [Ca2+]i on different PLA2s resulting in a pronounced release of arachidonic acid, as seen in other cell systems (40)
.
Since purinergic and adrenergic regulation of thyroid functions have been well documented (17)
, we analyzed the possibility that this pathway of PIs hydrolysis mediates some of the differentiated functions regulated by these ligands. In this cell system, proliferation and differentiation controls are strictly coupled; TSH-regulated signals promote cell growth and cell differentiation, whereas signals activated by oncogenes of different categories impair differentiation and stimulate TSH-independent proliferation (12
13
14)
. Our data suggest that NE slightly down-regulates mRNA levels of thyroglobulin, PAX8, and TTF1; the combined addition of NE and pyrrophenone amplifies this reduction, suggesting that a metabolite downstream of PLA2IV activity antagonizes this effect. Thus, the recovery from this down-regulation seen for TTF1 mRNA after the addition of GroPIns demonstrates that GroPIns is a good candidate for mediating these effects. The limited extent of these modulations requires further work to better define whether PLA2IV is indeed part of the regulatory pathways involved in controlling differentiation in the thyroid.
We have also shown that ATP and NE clearly stimulate PCCl3 cell growth in a PLA2IV-dependent manner, as their effects were inhibited by pyrrophenone and mimicked by the PLA2 products GroPIns, LysoPtdIns, and arachidonic acid. The micromolar concentrations of GroPIns used agree well with the cytosolic levels of GroPIns in these and other cell lines (144 µM in PCCl3 cells; ref. 41
), and again imply that GroPIns is a good candidate for modulation of adrenergic and purinergic receptor-mediated control of cell growth. On the other hand, LysoPtdIns has been shown to be a mitogen in several cell systems, including thyroid-derived cell lines (42)
, although to exert its function it needs to be released to reach its specific membrane receptor, as seen for other lysolipids (3
, 42)
. While LysoPtdIns release has been reported in ras-transformed cell lines, it was not detected in normal cells (probably due to the lower levels of PIs metabolites present in the intracellular space, which can be metabolically managed within the cell) (3)
. This has also been confirmed in the cell system used in this study, where no detectable release of LysoPtdIns was seen during ATP and NE stimulation (S. Mariggiò et al., unpublished observations). The increased levels of LysoPtdIns in these PCCl3 cells thus support the pathway of formation of GroPIns due to receptor-activated PLA2IV acting on the membrane PIs rather than supporting a role for LysoPtdIns itself in cell growth control.
Another putative mediator of this PLA2IV effect could be arachidonic acid or one of its metabolites. Indeed, in FRTL5 thyroid cells, NE-dependent regulation of cell proliferation can be mediated in part by PGE2 but not by the tromboxanes or HETEs (33)
. Similarly, activation of the TSH receptor by immunoglobulins of Graves patients results in activation of PLA2, followed by arachidonic acid release and metabolism via the cyclooxygenase pathway, also leading to increased thymidine incorporation (reviewed in ref. 43
). In the present study the prostaglandins are not involved in the control of cell proliferation, since inhibition of the cyclooxygenase pathway did not prevent adrenergic and purinergic stimulation of cell growth; direct addition of PGE1, PGE2, and TXB2 was also ineffective. However, the direct addition of arachidonic acid, which leads to the production of other metabolites through lipooxygenase and cytochrome P450 activities (44)
, also stimulated cell proliferation. Although this is in line with studies reported above on the role of the arachidonic acid metabolism in the control of thyroid cell growth, it does not appear to be correlated to the role of the adrenergic and purinergic receptors in this cell system. In fact, while these receptors induce the same extent of cell growth, which correlates well with the similar extents of GroPIns and LysoPtdIns production, the increase in arachidonic acid is dissociated from this regulation, as it is more pronounced after purinergic stimulation (ATP 100 µM, 18-fold increase) than after adrenergic stimulation (2-fold increase). This, together with the observation that the 30-fold increase in arachidonic acid release produced by the calcium ionophore A23187 had relatively little effect on cell proliferation (2-fold increase; S. Mariggiò et al., unpublished observations), led to the conclusion that specific GroPIns production, rather than a more general arachidonic acid release, mediates the receptor-activated PLA2IV involvement in the control of cell proliferation. The molecular mechanism involved is presently under investigation. It has been excluded that GroPIns may regulate the classical cAMP-dependent pathway, since it does not affect the levels of this second messenger (ref. 8
and S. Mariggiò et al., unpublished observations). Some proteins that may represent the GroPIns targets are under study, including cytosolic factors involved in transcription regulation of gene expression. Additional studies are needed to define this aspect.
The data presented are therefore consistent with the proposal that GroPIns mediates the effects of ATP and NE on cell growth (Fig. 7)
. GroPIns is hydrophilic and freely diffusible within cells; it can also permeate the cell membrane through a transporter that has been characterized in yeast (45)
, explaining its effects when administered exogenously to cells (3
, 6
, 8
9
10
11)
. An orthologue of this transporter has recently been identified and characterized in mammals (S. Mariggiò et al., unpublished observations). GroPIns can also interact with a series of cytosolic proteins but not with a specific membrane receptor (45
and S. Mariggiò et al., unpublished observations). Furthermore, GroPIns is relatively stable both within the cell and when added to the extracellular medium, in line with the proposal that it is active per se and not after phosphorylation to GroPIns4P or reacylation to LysoPtdIns and PtdIns (3
, 29)
. This conclusion is further supported by the observation that GroPIns4P does not affect proliferation either in these PCCl3 cells or in fibroblasts, where it can only potentiate growth factor effects (9)
; instead, GroPIns4P is very active in regulating the organization of the actin cytoskeleton (10)
.
In summary, we have identified a novel pathway of cell growth regulation that involves receptor-induced activation of PLA2IV and specific hydrolysis of the membrane PIs to form GroPIns. This suggests that PLA2IV specificity in living cells is also exerted at the level of substrate identification and the concomitant release of specific metabolites that can control cell proliferation. These findings extend the regulatory pathways controlled by the PIs and their various metabolites in cells, widening the role of these inositol-containing compounds in cell regulation.
ACKNOWLEDGMENTS
We thank I. Kudo for providing the PLA2IVwt and PLA2IV1–522 plasmids, C. Leslie for the pAcHLT-PLA2IV vector, K. Seno for the pyrrophenone inhibitor, M. Di Girolamo, C. P. Berrie, and R. M. Melillo for useful discussions, G. Di Tullio for preparing the recombinant PLA2IV, E. Fontana for preparation of the figures, and the Italian Association for Cancer Research (AIRC, Milan, Italy), Telethon Italia (Italy), the MIUR (Italy), and the Marie Curie Industry Host Fellowship of the European Community "Quality of life and Management of Living Resources" Programme, Contract Number QLK5-CT-2000-60088 for financial support. B.M.F. is a fellow of the Italian Foundation for Cancer Research (FIRC, Milan Italy).
Received for publication February 2, 2006. Accepted for publication July 6, 2006.
REFERENCES
, pyrrophenone. Biochem. J. 363,727-735[CrossRef][Medline] inhibits stimulus-induced arachidonate release, whereas that of type VI Ca2+-independent phospholipase A2 augments spontaneous fatty acid release. J. Biol. Chem. 275,18248-18258
subunits in a rat osteosarcoma cell line. Mol. Pharmacol. 48,532-539[Abstract]
0.01).
