Modifications in Phospholipase D Activity and Isoform Expression

Home , Phospholipase D1, PLD2, Ribonuclease

Leukemia (2000) 14, 2118–2127  2000 Macmillan Publishers Ltd All rights reserved 0887-6924/00 $15.00 www.nature.com/leu Modiﬁcations in phospholipase D activity and isoform expression occur upon maturation and differentiation in vivo and in vitro in human myeloid cells M El Marjou1, V Montalescot1, A Buzyn2 and B Geny1

1INSERM U332, Laboratoire de Signalisation, Inﬂammation et Transformation Cellulaire, ICGM, Paris; and 2INSERM U445, CHU Cochin- Port-Royal, Paris, France

Activation of phospholipase D (PLD) occurs in response to process of maturation and differentiation is important for the various stimuli and results from the activity of two isozymes, development of AML therapies. hPLD1 and hPLD2. PLD activity appears to be involved in several myeloid cell processes during their development and acti- Over the past two decades, phospholipase D (PLD) has vation, including proliferation of myeloblasts in the bone mar- been well documented to play an essential role in signal trans- row and secretion, phagocytosis and NADPH oxidase duction in a large variety of mammalian cells.4 Its activation activation, essential functions of differentiated neutrophils. The leads to hydrolysis of phosphatidylcholine, producing choline present work studies PLD characteristics, activity and both iso- and phosphatidic acid (PA), a phospholipid recognized as a zyme expression during maturation and differentiation of second messenger (see review Ref. 5). Two PLD isoforms have myeloid cells by using three different systems: leukemic mye- been cloned in human cells6,7 and were shown to be differ- loblasts at different stages of maturation, terminally differen- 8,9 tiated neutrophils ex vivo and four human myeloid cell lines, ently expressed in most mammalian cell types and tissues. NB4, HL-60, PLB 985 and U937, induced to differentiate with all- During the past few years, many studies have been focused trans retinoic acid (ATRA), a cyclic adenosine monophosphate on PLD activity which indicate that this enzyme activity might (cAMP) analogue or both agents together. HL-60, a bipotential be involved in numerous physiological processes of myeloid cell line has also been differentiated along the granulocytic cells, such as proliferation, differentiation, phagocytosis, pathway with DMSO and the monocytic pathway with 1,25-dihydroxy vitamin D3. In all these systems, PLD activity increases secretion and NADPH oxidase activation (see review Ref. 4). with maturation and differentiation whatever the inducer used Moreover, in mammals, myeloid cells appear to be amongst and the granulocytic or monocytic pathways. Increase in basal the richest in both PLD isozyme mRNAs.8 Despite major inter- activity which reflects the expression during development of est in the understanding of the relative participation of each both hPLD1 and hPLD2 appears to be mainly related to the for- PLD isoform in physiological processes, their respective roles mer isozyme expression. Association of PLD characteristic in myeloid cell functions are still unknown. changes with maturation and differentiation was also confirmed using two NB4 clones resistant to these processes. To get an insight on PLD changes during myeloid matu- Comparison between PLD characteristics in myeloblasts dur- ration and differentiation, the levels of PLD activity and PLD ing maturation and differentiation ex vivo and in vitro in the isoform expression were measured during these processes. different cell lines demonstrated that NB4 induced to differen- This study was performed in ex vivo circulating human leu- tiate with ATRA represents the best model for further studies kemic myeloblasts at different maturation and differentiation on the specific roles of each PLD isoform in various functions and in four different myeloid cell lines established in vitro of differentiated myeloid cells. Leukemia (2000) 14, 2118–2127. Keywords: myeloid cells; maturation; differentiation; phospho- from patients sufferring from different AML subtypes. HL-60 lipase D; PLD isoform expression cells now recognized to be established from an AML stage M2 (M2)10 and NB4 from a M3 AML carrying a t(15;17) translo- cation11 can be induced to differentiate towards neutrophil- Introduction like cells in the presence of all-trans retinoic acid (ATRA) and/or a cyclic AMP analogue. PLB 985 (M4) and U937 Early myeloid cells are able to proliferate along the pathway (histiocytic), already irreversibly induced toward the mono- of maturation. Upon maturation, they lose this capacity and cytic lineage, differentiate towards monocytes in the presence differentiate mainly into neutrophils, and to a lesser extent into of ATRA. monocytes. Fully differentiated myeloid cells possess several In the present work, PLD activity and PLD isoform properties linked to their final role: prevent infection. Indeed, expression were studied in parallel upon maturation and dif- they are able to ingest and destroy microorganisms.1 ferentiation in ex vivo human leukemic myeloblasts and in Acute myeloblastic leukemia (AML) is due to abnormal pro- several myeloid cell lines induced to mature and differentiate liferation of myeloid cells blocked at a specific stage of matu- in vitro. ration and differentiation classified from M0 to M5 according to the FAB leukemia subtype nomenclature. AML is a frequent disease which still has a serious prognosis. For several years, Materials and methods patients suffering from M3 promyelocytic leukemia have been treated with a high rate of remission with all-trans retinoic RPMI-1640 and medium 199 were purchased from Gibco BRL acid (ATRA), a differentiating agent which forces leukemic (Paisley, UK). Fetal bovine serum was from Dutcher (Vilmorin, cells to differentiate, acquire new functions and stop prolifer- France). Alkyl-sn-glyceryl-3-phosphorylcholine, 1-0-[alkyl- Ј Ј 3 ation.2,3 1 2 - H]-, (30–60 Ci/mmol) (lyso-PAF), phosphatidylcholine Understanding the events taking place during the complex were obtained from NEN (Boston, MA, USA). fMLP, phorbol 12-myristate 13-acetate (PMA), dibutyryladenosine3Ј:5Ј-cyclic monophosphate (dibutyryl-cAMP), dibutyryl-cAM8-(4- Ј Ј Correspondence: B Geny, Unite´ Inserm 332, ICGM, 22 rue Mećhain, chlorophenylthio)adenosine cyclic 3 ,5 -monophosphate (8- 75014 Paris, France; Fax: 33 01 40 51 64 54 CPT-cAMP), dimethyl sulfoxide (DMSO), all-trans retinoic Received 31 March 3000; accepted 16 August 2000 acid (ATRA), 1,25-dihydrocholecalciferol (1␣,25-dihydroxy Modifications in phospholipase D activity and isoform expression M El Marjou et al 2119 vitamin D3), cytochrome C, nitro-blue tetrazolium (NBT), NBT reduction Hank’s and HEPES buffers were purchased from Sigma (St Louis, MO, USA). Ficoll–Hypaque Plus was from Pharmacia Cell maturation was estimated by nitroblue tetrazolium (NBT) Biotech (Upsala, Sweden). RNase protection kit, protease and reduction and by morphological changes after cellular stain- fatty acid-free bovine serum albumine (BSA), protease inhibi- ing with May–Gru¨nwald-Giemsa. Determination of NBT tor cocktail (Complete) was purchased from Boehringer- reduction was performed by incubation of 106 cells in 1 ml Mannheim (Mannheim, Germany). Transfer PVDF membranes fresh culture medium for 20 min at 37°C with 0.1% NBT and were purchased from Amersham (UK). Silica gel 60 plates for 0.3 ␮M PMA. The percentage of cells containing intracellular thin layer chromatography (TLC) analysis were obtained from blue-black formazan deposits was then determined. At least Whatman (Maidstone, UK). Scintillant liquid, Optiphase ‘HiS- 200 cells were examined. afe’ 3, was from EG&G Wallack (Turku, Finland).

Superoxide ion production assay Purification of human neutrophils, monocytes and NADPH oxidase activity was determined by using the leukemic cells − reduction of cytochrome C by O2 as indicator. Reduction of cytochrome C was measured as the change in absorption at Protocol on human cells were approved by the CCPPRB 550 nm over 20 min at 20 s intervals in a spectrophotometer (Comite´ Consultatif pour la Promotion des Recherches en (Perkin Elmer Lambda 4 apparatus; Perkin Elmer, Norwalk, Biologie). Informed consent was obtained from all blood CT, USA). Cells (2 × 106) were resuspended in PBS containing donors and patients. 30 mM glucose, 1 mM MgCl2 and 0.5 mM CaCl2 and then pre- Neutrophils were isolated from venous blood of five healthy incubated with 50 ␮M of cytochrome C for 5 min at 37°C. blood donors. The platelet-rich plasma was first removed. Superoxide ion formation was initiated by the addition of N- Neutrophils and monocytes were isolated by dextran sedi- formyl Met-Leu-Phe (fMLP) (1 ␮M) or PMA (1 ␮M). mentation and Ficoll–Hypaque centrifugation. Neutrophils present in the pellet were further purified by elimination of contaminating erythrocytes by hypotonic shock. Purified neutrophils were resuspended in HBSS containing 20 mM HEPES, Determination of phospholipase D activity in intact pH 7.4, 0.1% fatty acid-free BSA, and 5.6 mM glucose. Cell cells viability was measured by trypan blue exclusion and was greater than 95%. Peripheral blood monocytes present with PEt measurement: PLD activity was measured in intact other monocytic cells at the interface of the gradient were sep- cells as previously described.14 Briefly, after washing, cells arated from lymphocytes by adhesion to plastic for 2–3 h and were labelled for 30 min at 37°C with 3H-lyso-PAF (2 ␮Ci/ml) recovered for further processes by a 10-min incubation with in an isotonic medium buffered with Hepes, pH 7.4. After 10 EDTA. min preincubation in the same medium containing 1.5% etha- Peripheral blood mononuclear cells (PBMC) of 12 leukemic nol, PLD activity measurement was performed using 106 cells patients were collected between November 1995 and March (100 ␮l aliquots). The enzymatic reaction was stopped after 1998 before the initiation of chemotherapy. The PBMC were 10 min incubation at 37°C by addition of 500 ␮lof separated from total blood samples of patients with hyperleu- chloroform/methanol (1:1, v/v) and samples were processed kocytosis (WBC Ͼ10 × 109/l) by Ficoll–Hypaque gradient as previously reported.14 (Pharmacia Biotech) and frozen in 10% DMSO, 10% human AB serum. All the PBMC samples contained a majority of tumor cells with less than 10% residual lymphocytes or nor- RNase protection assay mal hematopoietic cells. Total RNA was isolated from cell lines, leukemia cells and neutrophils using Qiagen’s RNeasy kit according to manufac- Cell culture and induction of differentiation turer’s instruction (Qiagen, Hilden, Germany). Probes were made of a 398 nt hPLD1 cDNA fragment (nt 382–780) and a 360 nt hPLD2 cDNA fragment (nt 2094– NB4 cell line and two RA-resistant clones, NB4-R1and NB4- 2454). Ribonuclease protection was performed as described R2, were obtained from Dr Michel Lanotte (Inserm-CNRS, in the manufacturer’s kit. An anti-sense labelled probe RNA Paris, France). All other myeloid cell lines, HL60, PLB-985 was obtained from the in vitro transcription of each linearized and U-937 cells purchased from ATCC (USA) were cultured plasmid using T7 RNA polymerase. Total RNAs (5 ␮g for PLD1 at 37°C in RPMI 1640 supplemented with 10% fetal bovine and 20 ␮g for PLD2) were hybridized with approximatively serum, penicillin (100 IU/ml) streptomycin (100 ␮g/ml) and L- 30000 c.p.m. of radiolabelled probe overnight at 45°C. Diges- glutamine (2 mM). tion conditions were 4 ␮g/ml RNase A and 20 U/ml RNase Differentiation of NB4 cells and both clones, was initiated T1 for 30 min at 30°C. Protected fragments were purified by by the addition of 1 ␮M ATRA or the combination of 1 ␮M proteinase K digestion, phenol-chloroform extraction, and ATRA and 200 ␮M 8-CPT-cAMP. HL-60 cells were induced ethanol precipitation. Purified samples were then separated to differentiate towards neutrophils by addition of 1 ␮M ATRA on 7 M urea sequencing gels. Glyceraldehyde-3-phosphate or 500 ␮M dibutyryl-cAMP as described by Breitmann and dehydrogenase (GADPH) was observed not to vary upon Chaplinski, respectively12,13 and towards monocytes in the maturation and differentiation (unpublished results), and was presence of 100 ␮M 1,25-dihydroxy vitamin D3. PLB-985 and used for normalization. Results were quantified using a phos- U937 cells were differentiated with 1 ␮M ATRA. phorimager (Moldyn, Cambridge, MA, USA).

Leukemia Modiﬁcations in phospholipase D activity and isoform expression M El Marjou et al 2120 Immunoprecipitation and Western blot analysis

Cells (6 × 107 cells/ml) were washed with PBS and solubilized on ice for 15 min in lysis buffer: Tris-HCl (50 mM), pH 7.4,

NaCl (150 mM), MgCl2 (1.5 mM), EGTA (5 mM), 10% glycerol, 1% triton X-100, NaVO4 (2 mM), phenylmethylsulfonyl ﬂuor- ide (1 mM), aprotinin (5 ␮g/ml), leupeptin (5 ␮g/ml) and centri- fuged at 10000 g for 10 min at 4°C. The resulting supernatants were directly immunoanalyzed by Western blotting for their content in PLD2 and after immunoprecipitation with anti- PLD1 antiserum as described.15 Labelling was performed as described in the ECL protocol. Immunoreactive bands were scanned and intensity was quantiﬁed using a computerized program (ImageQuant).

Results

In vivo, maturation and differentiation of myeloid leukemic cells are associated with a marked increase in PLD activity and a differential increase in each PLD messenger expression

Myeloblasts were obtained from the peripheral blood of patients sufferring from AML at different stages of maturation and differentiation, using the FAB leukemia subtype nomenclature. Changes concerning PLD characteristics during these processes were analyzed. As reported in Figure 1A, intrinsic PLD activity was found to increase with cell matu- Figure 1 PLD activity and PLD isoform mRNA expression in ration and differentiation from stage M0 to M5. M0 to M3 relation to maturation stage in leukemic myeloid cells ex vivo. Cells leukemic cells are arrested at stages of maturation which pre- isolated as described in the Methods section were rapidly thawed, cede commitment towards PMN or monocyte, whereas M4 washed in PBS at 37°C and preincubated in complete RPMI 1640 and M5 myeloblats are already committed to monocytic medium for at least 4 h before use. More than 90% of cells were alive in such condition. Cells were labelled for 30 min with 3H-lyso-PAF differentiation. (2 ␮Ci/ml), washed and preincubated at 37°C for 10 min in an isotonic To study if such variations could be related to an increase medium buffered with HEPES, pH 7.4 containing 0.1% BSA. Intrinsic in PLD expression, mRNA expression of both isoforms was PLD activity was then measured in intact cells over a 20 min at 37°C estimated using RNAse protection, a sensitive measurement in the presence of 1.5% ethanol. PLD activity was estimated as technique. As reported in Figure 1b and c, mRNA expression amount of radioactivity incorporated into phosphatidylethanol (PEt), of hPLD1 and hPLD2 was increased according to the stage of expressed as the percentage of phosphatidylcholine (PC) hydrolysed. All determinations were carried out in duplicate. Results are the mean maturity and differentiation of the leukemic myeloid cells ± s.e. obtained from triplicate aliquots of circulating myeloblasts from from M0 to M5, hPLD1 being more increased than hPLD2 three different patients with AML at the same maturation stage. UTP- (Figure 1b and c). Differences observed in the expression of labelled human PLD1, PLD2 and GADPH probes were hybridized to both PLD isoform messengers might explain the difference in RNA extracted from myeloblasts isolated from patients with AML at basal PLD activity. different stages of maturation. Five ␮g of RNA were used for PLD1 ␮ In vivo differentiation of all myeloid cells, whatever their (a) and 20 g of RNA for PLD2 (b) RNAse protection assays, respectively. All determinations were carried out on cells from two or three final commitment appears to involve modifications concern- different patients. Results were estimated as the ratio of GADPH ing PLDs which might be important for specific functions in mRNA level and the amount of each isoform in immature M0 myelob- differentiated myeloid cells. lasts was considered as equivalent to 1. Studies on maturation and differentiation have been restric- ted for years by the availability of ex vivo myeloid cells. At present, several human myeloid cell lines have been estab- leukemia, induces differentiation of NB4 cells towards neutro- lished in continuous culture from patients suffering from AML phil-like cells in 4–5 days. Association of ATRA and 8-CPT- and represent interesting tools for such studies. Thus, we stud- cAMP also allows the NB4 cell line to mature and differentiate; ied the development of PLD activity and changes in both inducers together having a more rapid effect than ATRA expression of PLD1 and PLD2 RNA messengers and proteins alone as full differentiation is obtained within 2–3 days.16 in in vitro established myeloid cell lines. In the present work, we checked that this was the case in our hands as measured by the percentage of cells reducing NBT, a good index for cell maturation. A maximal percentage The NB4 cell line acquires neutrophil-like functions of NBT-positive cells was obtained after 4 days in the presence upon maturation and differentiation of ATRA and after 2 days in the presence of ATRA plus 8- CPT-cAMP (Figure 2a and b). The NB4 cell line, established from malignant M3 promyelo- A second index of maturation and differentiation towards cytes, is a model to study maturation and differentiation neutrophil-like cells is the acquisition of a functional NADPH towards cells presenting neutrophil-like characteristics. oxidase that can be activated either nonspecifically with PMA Indeed, ATRA, used for the treatment of patients with acute M3 or specifically via fMLP receptors. As shown in Figure 2c and

Leukemia Modifications in phospholipase D activity and isoform expression M El Marjou et al 2121 d, undifferentiated cells were unable to generate superoxide ions. After induction of differentiation both stimuli provoked superoxide ions generation. Indeed, within 20 s, fMLP led to the production of superoxide ions, maximal after 5 min and reaching a level similar to that produced by normal neutrophils (data not shown). By contrast, PMA gave rise to superoxide ions generation after a 90 s delay which was prolonged for 30 to 40 min (data not shown). fMLP- and PMA-induced production of superoxide ions increased with differentiation as reflected by the percentage of NBT-positive cells. In cells differentiated for 5 days with ATRA (Figure 2c), stimulation of NADPH oxidase with PMA was greater than in cells differentiated for 3 days with ATRA and 8-CPT-cAMP (Figure 2d). Thus, NB4 cells were found to acquire in vitro essential functions of differentiated neutrophils. Therefore, we studied whether we would find in this cell line similar changes in PLD activity and isozyme expression observed in leukemic cells.

Figure 2 Measurements of NBT reduction and superoxide ion production in NB4 cells upon differentiation with ATRA or ATRA plus 8- CPT-cAMP. NB4 cell differentiation was induced either with ATRA alone (a and c) or ATRA plus 8-CPT-cAMP (b and d). Maturation and differentiation were estimated as the percentage of cells able to reduce NBT (a and b) and by cell acquisition of a functional NADPH oxidase as measured by superoxide ion production (c and d) when stimulated in the presence of 1 ␮M fMLP (ﬁlled squares) or 1 ␮M PMA (empty squares). Superoxide ion production was expressed in nmoles of ions formed per min per 106 cells. All determinations were carried out in duplicate or triplicate. Results are the mean ± s.e. from at least three different experiments.

Figure 4 Expression of PLD1 and PLD2 RNA messengers in NB4 cells upon differentiation. Human PLD1 and PLD2 and GADPH Figure 3 Time-course of intrinsic PLD activity in NB4 cells upon probes were hybridized to RNA extracted from NB4 cells, at different differentiation. Basal PLD activity in intact NB4 cells was measured days after induction of differentiation with ATRA (a and c) or ATRA upon differentiation with ATRA (a) or with ATRA plus 8-CPT-cAMP plus 8-CPT-cAMP (b and d). PLD1 mRNA (a and b) and PLD2 mRNA (b) as detailed in Figure 1. Results are the mean ± s.e. of three different (c and d) were quantiﬁed using RNAse protection assays as detailed experiments in which all determinations were made in duplicate and in Figure 1. All determinations were carried out on three different are expressed as in Figure 1. occasions and results are the mean ± s.e.

Leukemia Modiﬁcations in phospholipase D activity and isoform expression M El Marjou et al 2122

Figure 5 Expression of PLD1 and PLD2 proteins in NB4 resistant clones upon differentiation. NB4 cells were either undifferentiated or differentiated into neutrophil-like cells for different periods in the presence of ATRA (a and c) or ATRA plus 8-CPT-cAMP (b and d). PLD1 (a and b) and PLD2 (c and d) proteins were detected in cellular extracts from NB4 and quantiﬁed by densitometry. All determinations were carried out on three different cell lysates and results are the mean ± s.e.

In vitro maturation and differentiation of NB4 cells cAMP, an increase in PLD1 messenger expression was pro- with ATRA or ATRA plus 8-CPT-cAMP is gressively obtained and reached a similar level within 3 days accompanied by an increase in PLD activity to that obtained with ATRA alone (Figure 4b). PLD2 mRNA expression also increased with differentiation whichever the Intrinsic PLD activity was studied in the NB4 cell line induced inducer(s) used but to a lesser extent than PLD1 (Figure 4c to differentiate with ATRA alone or in combination with 8- and d), reproducing results obtained with patients myelo- CPT-cAMP. When ATRA was the sole inducer, PLD activity blasts. increased after a 2 day delay and reached a maximal level Since PLD activity is related to protein expression, we within 5 days (Figure 3a). When both inducers were used quantified the cell lysate content of each isoform along the together, no delay was observed in PLD activity increase differentiation process. As suspected from mRNA levels, which was progressive and maximal after 3 days (Figure 3b). expression of the PLD1 isoform increased with maturation and differentiation (Figure 5a and b). In similar conditions, PLD2 protein expression increased but to a lesser extent than PLD1 Expression of PLD1 and PLD2 messengers and (Figure 5c and d). proteins increases upon in vitro maturation and Increased expression of both proteins during this process differentiation of NB4 cells with ATRA or ATRA plus could reasonably be considered as responsible for the 8-CPT-cAMP observed increase in intrinsic PLD activity. Although changes in PLD characteristics observed during in vitro-induced differ- PLD1 mRNA expression was not modified for the first 24 h entiation in the NB4 cell line appear to be similar to those following ATRA addition (not shown) and increased after 2 observed upon in vivo maturation and differentiation of neu- days to give a dramatic change when full maturation was trophils, it cannot be excluded that the inducers used in vitro reached (Figure 4a). When ATRA was added with 8-CPT- might be responsible for PLD modifications without any

Leukemia Modiﬁcations in phospholipase D activity and isoform expression M El Marjou et al 2123

Figure 6 PLD activity in R1-NB4 and R2-NB4 clones resistant to maturation and differentiation and messenger RNA and protein expression of PLD isozymes in R1 NB4 clone. PLD activity (a) was measured in R1 (empty bars) and R2 (hatched bars) clones derived from the NB4 cell line which are resistant to ATRA- and ATRA plus 8-CPT-cAMP-induced differentiation, respectively. Cells were treated with ATRA (1 ␮M) for 4 days, or by ATRA plus 8-CPT-cAMP (0.2 mM) for 3 days and intrinsic PLD activity was estimated as in Figure 1. mRNA expression of PLD1 (b) and PLD2 (c) and protein expression of PLD1 and PLD2 (d and e, respectively) were estimated in ATRA- and ATRA plus 8-CPT-cAMP- differentiated cells as indicated. mRNA and protein isozyme measurements were carried out as in Figures 2 and 5.

relationship to these processes. Thus, we took advantage of in PLD isoform messenger and protein expression occurred two NB4 clones resistant to induction of differentiation with only in cells susceptible to differentiation, ie in clone R1 ATRA or ATRA plus 8-CPT-cAMP16,17 to determine whether induced to mature and differentiate with ATRA plus 8-CPT- PLD changes were linked to these physiological processes. cAMP. Within 3 days, a major increase in hPLD1 mRNA expression occurred (Figure 6b) and a lesser increase in hPLD2 mRNA (Figure 6c). Both protein isozymes were also Changes in PLD activity and PLD expression are increased (Figure 6d and e). In contrast, no modiﬁcation was associated with the process of differentiation observed in the levels of either isoform mRNA in clone R1 cultured with RA alone, nor in clone R2 under all conditions As reported in Figure 6a, intrinsic PLD activity was not differ- studied (data not shown). ent in ATRA-resistant clones, R1 or R2, treated only with this From the present study, the NB4 cell line appears to be a inducer for 5 days. When each clone was cultured with ATRA good model for PLD studies during maturation of myeloid plus 8-CPT-cAMP for 3 days, a dramatic increase in PLD cells. However, other myeloid cell lines have been established activity was observed in R1 cells whereas no modiﬁcation and also differentiate in vitro. Thus, we compared such lines occurred in the non-inducible R2 clone. In parallel, changes with NB4 cells and circulating human neutrophils ex vivo.

Leukemia Modiﬁcations in phospholipase D activity and isoform expression M El Marjou et al 2124 lines, PLD activity and PLD isoform mRNAs expression were measured before and after differentiation with ATRA and/or cAMP analogues. In undifferentiated cells, intrinsic PLD activity was found to be higher in PLB 985 (M4) and U937 (promonocytic) cell lines than in HL-60 (M2) and NB4 (M3) cells (Table 1). Upon differentiation with ATRA, intrinsic PLD activity was only slightly increased in the two mature cell lines (PLB 985 and U937) to reach the level of PLD activity present in mature neutrophils after differentiation. In contrast, intrinsic PLD activity was strongly increased with maturation and differentiation in HL-60 and NB4 cells (Table 1). After differentiation, PLD isoform mRNAs were increased in all cell lines studied and the increase in PLD1 mRNA was more marked than that of PLD2 mRNA. Indeed, PLD1 messengers increased 3.5 to 5.5 times in these three cell lines, whereas the PLD2 messengers only increased 1.5 to 2.8 times. Moreover, the increase in each messenger isoform was somewhat less marked upon differentiation in PLB 985 and U937 (around three- to four-fold increase for PLD1 and 1.5–1.8-fold increase for PLD2 messengers) than in HL-60 (5.5- and 2.8-fold increase, respectively) or NB4. HL-60 cell line can be differentiated towards granulocyte- or monocyte-like cells depending on the inducer used. In this cell line, intrinsic PLD activity and PLD isoform expression were estimated after differentiation towards neutrophils in the presence of ATRA, dibutyryl-cAMP or DMSO or towards monocytes with 1,25-dihydroxy vitamin D3. Measurements were performed after 4 days in the presence of each inducer except for dibutyryl-cAMP which induced full differentiation in 3 days. Maturation and differentiation generated by all inducers led to an increase in intrinsic PLD activity (Figure 7a). PLD1 level was increased upon differentiation whichever the differentiating agent used. In contrast, in this cell line, PLD2 expression was differently expressed depending on the inducer used for differentiation. ATRA did not modify PLD2 protein expression, while DMSO increased it and cAMP provoked a diminution. Differentiation towards monocyte-like cells generated with 1,25-dihydroxy vitamin D3 also decreased this isozyme expression. Thus, in myeloid cells, maturation and differentiation, whichever the granulocytic or monocytic pathway appear to be accompanied by an increase in intrinsic PLD activity, which is likely to be essentially related to PLD1 isoform expression.

PLD activity in fully differentiated human myeloid cells: circulating neutrophils and monocytes Figure 7 PLD activity and protein expression of PLD isozymes in HL-60 cells after differentiation into granulocyte- and monocyte-like In vivo, myeloid cells differentiate in the bone marrow. In cells with several inducers. PLD activity was measured as PEt formed blood, neutrophils and monocytes represent fully differen- in HL-60 cells (a, empty bars). Differentiation towards granulocyte- tiated cells with specific functions. Thus, PLD activity was also like cells was induced with ATRA (1␮M), DMSO (1.2%) for 4 days or studied ex vivo in neutrophils and monocytes isolated from dibutyryl-cAMP (0.5 mM) for 3 days and towards monocyte-like cells healthy donors. A high basal PLD activity was observed in with 1,25-dihydroxy vitamin D3 (0.1 ␮M) for 4 days. Intrinsic PLD activity was estimated as in Figure 1; determinations were carried out these cells representing about four times that in M2 or M5 on three different occasions and results are the mean ± s.e. PLD1 (b) circulating myeloblasts. Moreover, in polymorphonuclear and PLD2 (c) proteins were detected in cellular extracts from undiffer- neutrophils which are abundant, PLD1 and PLD2 mRNA con- entiated and differentiated HL-6O cells and quantified by densitome- tent could be measured and were found to represent 2.5 and try as in Figure 5. 1.4 times, respectively, the amount measured in M2 cells. Increases in intrinsic PLD activity and in both isoform messenger expressions are common features upon Discussion maturation and differentiation in myeloid cell lines and in ex vivo human myeloid cells A role for PLD activity in cell maturation and differentiation Intrinsic PLD activity has been studied in three other cell lines has been suggested but, so far, is poorly documented. The established from AML at various maturation states. In all cell present work is the first demonstration that maturation and

Leukemia Modiﬁcations in phospholipase D activity and isoform expression M El Marjou et al 2125 Table 1 Intrinsic PLD activity in human myeloid cell lines and in peripheral blood polymorphonuclear neutrophils and monocytes

Treatment of cells Cell line (drug and time) HL-60 NB4 PLB 985 U 937 Polymorphnuclear Peripheral (M2) (M3) (M4) (Histiocytic) neutrophils (PMN) monocytes (ex vivo) (ex vivo)

PEt formation (% of total PC) No treatment — 0.05 ± 0.02 0.08 ± 0.03 0.44 ± 0.04 0.36 ± 0.01 0.6 ± 0.18 0.69 ± 0.28 ATRA 5 days 0.23 ± 0.01 0.40 ± 0.01 0.65 ± 0.04 0.66 ± 0.04 ATRA + 8-CPT-cAMP 3 days 0.29 ± 0.04 — — dibutyryl-cAMP 3 days 0.22 ± 0.03

Intrinsic PLD activity was measured in four cell lines established from human leukemic cells and in circulating polymorphonuclear neutrophils and monocytes from healthy blood donors. Enzyme activity was estimated before and after maturation and differentiation in the presence of ATRA for 5 days, dibutyryl-cAMP (HL-60) and ATRA plus 8-CPT-cAMP (NB4) for 3 days. All determinations were carried out on three different occasions (five for PMN) and results are the mean ± s.e. differentiation in ex vivo human myeloid cells are brain, PLD1 mRNA expression, but not that of PLD2, was accompanied by an increase in PLD activity related to an shown to be markedly increased during embryonic develop- increase in the expression of isozyme messengers, greater for ment.21 Similarly, a major increase in mRNA expression of PLD1 than for PLD2. Patients with acute promyelocytic leuke- PLD1 was only observed during 1,25-dihydroxyvitamin D3- mias (M3) can be initially treated with ATRA which causes induced differentiation in epidermal keratinocytes.22 Differen- the leukemic cells to acquire a mature phenotype2,3,18,19 and tiation of PC12 cell was reported to lead to an increase in improves the outcome.3,19 Therefore, it could be of interest in PLD mRNA which concerns both isoforms or only PLD1 these patients to examine PLD activity and isozyme according to the inducer used, cAMP and nerve-growth factor, expression in neutrophils and monocytes. Moreover, changes respectively.23 Yoshimura et al24 observed that differentiation in PLD activity and/or in PLD isoform expression in neutro- of C6 glioma cells provoked a transient increase in PLD1 RNA phils could be an important feature for understanding neutro- messenger expression and a major decrease in PLD2 mRNA phil functions during severe infections or major inflammatory in response to dibutyryl cAMP and theophilline. In F9 terato- reaction and might suggest a new drug target for affecting carcinoma cells, PLD1 protein expression was measured and neutrophil function. found to be either not changed when cells were differentiated Using four different myeloid cell lines established from into visceral endoderm in response to ATRA or majorly human leukemic cells representing different stages of matu- decreased when differentiation was induced by ATRA plus ration and differentiation, we confirmed that PLD activity and dibutyryl-cAMP into parietal endoderm.25 Although the expression of both isoforms also increased with maturation amount in mRNA of one or both isoform was modified in all and differentiation. Indeed, HL-60 and NB4 cell lines, estab- studies, increases and/or decreases were observed during dif- lished from M2 and M3 AML respectively, contained less ferentiation suggesting different roles for PLD isoforms intrinsic PLD activity than those derived from M5 and histio- expression/activity in the different cell types. cytic leukemias, PLB 985 and U935, respectively. Moreover, The major effect for PLD activity seems to be production of in vitro differentiation also promotes an increase in PLD phosphatidic acid (PA) which acts as a second messenger.5 activity and in mRNA expression of both isoforms in the four This minor and transiently formed phospholipid, shown to be 26,27 cell lines. In NB4 cells, we show for the first time that the involved in PI4,5P2 synthesis, appears to be essential for increase in mRNA expression of each PLD isozyme is trans- PLD activity which requires this polyphosphoinositide for its lated into increased amounts of PLD1 and PLD2 proteins. In activity.28 PA production has also been demonstrated to be the present work, resistant clones, R1 and R2, nonresponsive essential for adhesion,29 granule secretion and NADPH oxi- to particular inducers were essential tools to ascertain the dase activation,30 functions essential for neutrophil migration relationship between PLD development and myeloid cell to the site of infection and for bacteriostatic and bacterio- maturation and differentiation and to discard a role for the lytic activities. agent used to induce differentiation in PLD changes. So far, upon differentiation of myeloid cells, it has been A previous study by Nakashima et al20 showed that upon reported that increased PLD activity was reponsible for the granulocytic differentiation of HL-60 with ATRA or dibutyryl- activation of cPLA2 in U937.31 In contrast, HL-60 cell differ- cAMP, PLD2 mRNA was increased 20-fold and PLD1 only 4– entiation induced with two different agents, ATRA and PMA 7-fold as measured by RT-PCR. Differences between the was previously shown to decrease the activity of another 31–33 results of that study and our own might be due to diffferent phospholipase, PIP2-PLC. sublines used in each laboratory or to different techniques The specific roles of each PLD isoform in the functions of used for mRNA measurements. The latter hypothesis is more differentiated myeloid cells are not yet documented. How- likely as we observed that in the four cell lines studied and ever, levels of PLD1 after HL-60 maturation and differentiation in circulating leukemic myeloblasts studied ex vivo, the with several inducers suggest that this isoform might be that increase of PLD1 mRNA was more marked than the increase required for neutrophil and monocyte functions. Neutrophils of PLD2 mRNA during maturation and differentiation. mRNA are abundant in the peripheral blood however their half-life expression of the two PLD isozymes has been investigated is short (24 h), moreover they contain high levels of proteases, during differentiation in several tissues and cell types. In rat two properties which render work with this human material

Leukemia Modifications in phospholipase D activity and isoform expression M El Marjou et al 2126 more difficult. In contrast, monocytes have a longer half-life, Morel P, Lamy T, Tilly H, Guerci A, Maloisel F, Bordessoule D, but their number in peripheral blood is low. Therefore, for Sadoun A, Tiberghein P, Fegueux N, Daniel M-T, Chomienne C, functional studies, it is interesting to utilize a myeloid cell line Degos L. All-trans retinoic acid followed by intensive chemotherapy gives a high complete remission rate and may prolong able to acquire most properties of mature differentiated cells remission in newly diagnosed acute promyelocytic leukemia: a in vitro. Several myeloid cell lines established from AML at pilot study on 26 cases. Blood 1992; 80: 2176–2181. various stages are known to differentiate in vitro. They pro- 4 Liscovitch M, Czarny M, Fiucci G, Tang X. Phospholipase D: mol- duce neutrophil-like or monocyte-like cells and thus, rep- ecular and cell biology of a novel gene family. Biochem J 2000; resent useful tools for futher studies on PLD providing that 345: 401–415. their isozyme characteristics are close to those in ex vivo cells. 5 English D, Cui Y, Siddiqui RA. Messenger functions of phosphatidic acid. Chem Phys Lipids 1996; 80: 117–132. Our results concerning PLD in four myeloid cell lines and in 6 Hammond SM, Altshuller YM, Sung T-C, Rudge SA, Rose K, Enge- ex vivo myeloid cells indicate that NB4 would be a good brecht J, Morris AJ, Frohman MA. Human ADP-ribosylation acti- model to study further the roles of PLD isoforms during matu- vated phosphatidylcholine-specific phospholipase D defines a ration, differentiation and in the functions of neutrophil-like new and highly conserved gene family. J Biol Chem 1995; 270: cells. In this cell line, ATRA appears to be the best granulo- 26640–26643. cytic inducer for development of PLD characteristics close to 7 Lopez I, Arnold RS, Lambeth JD. Cloning and initial characteriz- those of neutrophils. Moreover, its maturation and differen- ation of a human phospholipase D2 (hPLD2). J Biol Chem 1998; 273: 12846–12852. tiation with this agent also lead to expression of fMLP recep- 8 Colley WC, Altshuller YM, Sue-Ling CK, Copeland NG, Gilbert 34 tors and lactoferrin expression located in specific granules. DJ, Jenkins NA, Branch KD, Tsirka SE, Bollag RJ, Bollag WB, Froh- Major differences were observed between NB4 and HL-60 in man MA. Cloning and expression analysis of murine phospho- the level of PLD2 protein isoform but not in mRNA. This dis- lipase D1. Biochem J 1997; 326: 745–753. crepancy might be due to differences in RNA protease devel- 9 Park S-K, Provost JJ, Bae CD, Ho W-T, Exton JH. Cloning and opment between these two cell lines along the pathway of characterization of phospholipase D from rat brain. J Biol Chem 1997; 272: 29263–29271. maturation and differention. 10 Dalton WT, Aheearn M, McCredie K, Freidreich EJ, Stass SA, Tru- In conclusion, we have demonstrated that leukemic jillo JM. HL60 cell line was derived from a patient with FAB-M2 myeloid cells possess two different PLD isozymes whose and not FAB-M3. Blood 1988; 71: 242–247. expression increases with maturation and differentiation. 11 Lanotte M, Martin-Thouvenin V, Najman S, Ballerini P, Valensi F, Although, PLD activity is known to be involved in specific Berger R. NB4, a maturation inducible cell line with t(15/17) functions of differentiated myeloid cells, little is known, so far, marker isolated from a human acute promyelocytic leukemia about the respective participation of the two PLD isozymes in (M3). Blood 1990; 77: 1080–1086. 12 Breitman TR, Selonick SE, Collins SJ. Induction of differentiation these functions. It seems likely that each of them is important of the human promelocytic leukemia cell line (HL-60) by retinoic for different physiological processes in relation to their spe- acid. Proc Natl Acad Sci USA 1980; 77: 2936–2940. cific regulation and/or subcellular localization. However, is 13 Chaplinski TJ, Niedel JE. Cyclic nucleotide-induced maturation of each specific PLD isoform responsible for certain functions? human promyelocytic leukemia cells. J Clin Invest 1982; 70: Do they both participate in the same functions but in different 953–964. subcellular locations? More work is required to answer these 14 Geny B, Cockcroft S. Synergistic activation of phospholipase D by protein kinase C- and G-protein-mediated pathways in streptolysin questions. The NB4 cell line appears to be a good model to O-permeabilized HL-60 cells. Biochem J 1992; 284: 531–538. fully understand the relative participation of PLD1 and PLD2 15 Marcil J, Harbour D, Naccache PH, Bourgoin S. Human phospho- isoforms in myeloid cell functions that require PLD activity. lipase D1 can be tyrosine phosphorylated in HL-60 granulocytes. J Biol Chem 1997; 272: 20660–20664. 16 Ruchaud S, Duprez E, Gendron MC, Houge G, Genieser HG, Jas- Acknowledgements torff B, Doskeland SO, Lanotte M. Two distinctly regulated events, priming and triggering, during retinoic-induced maturation and resistance of NB4 promyelocytic leukemia cell line. Proc Natl This work was supported by Association de Recherche sur le Acad Sci USA 1994; 91: 8428–8432. Cancer (ARC) Grant No. 9452. MEM was funded by a fellow- 17 Duprez E, Benoit G, Flexor M, Lillehaug JR, Lanotte M. A mutated ship la Fondation pour la Recherche Me´dicale (FRM) and VM PML/RARA found in the retinoid maturation resistant NB4 sub- is a fellow from l’Association de Recherche sur la Polyarthrite clone NB4-R2, blocks RARA and wild-type PML/RARA transcrip- (ARP). We thank Michel Lanotte for providing NB4 cells and tional activities. Leukemia 2000; 14: 255–261. 18 Warrell RPJ, Frankel SR, Miller WH Jr, Scheinberg DA, Itri LM, resistant clones, Michael Frohman and David Lambeth for Hittelman WN, Vyas R, Andreeff M, Tafuri A, Jakubowski A, Gab- providing hPLD1 and hPLD2 cDNA, respectively, Sylvain rilove J, Gordon M, Dmitrovsky, E. Differentiation therapy of acute Bourgoin for the gift of specific antibodies to PLD1 and PLD2, promyelocytic leukemia with tretinoin (all-trans retinoic acid). Muriel Gaudry and Axel Perianin for their help in setting the New Engl J Med 1991; 324: 1385–1393. technique for superoxide ion measurement, Sebastien Bergeot 19 Huang ME, Ye YC, Chen SR, Chai JR, Lu JX, Zhoa L, Gu LJ, Wang for some technical assistance, Babette Weksler and Jean-Paul ZY. Use of all-trans retinoic acid in the treatment of acute promye- Mira for reading the manuscript and stimulating discussions. locytic leukemia. Blood 1988; 72: 567–572. 20 Nakashima S, Ohguchi K, Frohman MA, Nozawa Y. Increased mRNA expression of phospholipase D (PLD) isozymes during granulocytic differentiation of HL60 cells. Biochim Biophys Acta References 1998; 1389: 173–177. 21 Zhao D, Berse B, Holler T, Cermak JM, Blusztajn JK. Developmen- 1 Ali H, Haribabu B, Richardson RM, Snyderman R. Mechanisms of tal changes in phospholipase D activity and mRNA levels in rat inflammation and leukocyte activation. Adv Rheumatol 1997; 81: brain. Dev Brain Res 1998; 109: 121–127. 1–28. 22 Griner RD, Quin F, Jung E, Sue-Ling CK, Crawford KB, Mann- 2 Castaigne S, Chomienne C, Daniel MT, Ballerini P, Berger R Blakeney R, Bollag RJ, Bollinger Bollag W. 1,25-dihydroxyvitamin Fenaux P, Degos L. All-trans retinoic acid as a differentiation ther- D3 induces phospholipase D-1 expression in primary mouse epi- apy for acute promyelocytic leukemia: I. clinical results. Blood dermal keratinocytes. J Biol Chem 1999; 274: 4663–4670. 1990; 76: 1704–1709. 23 Hayakawa K, Nakashima S, Ito Y, Mizuta K, Miyata H, Nozawa 3 Fenaux P, Castaigne S, Dombret H, Archimbaud, E, Duarte M, Y. Increased expression of phospholipase D1 mRNA during

Leukemia Modiﬁcations in phospholipase D activity and isoform expression M El Marjou et al 2127 cAMP- or NGF-induced differentition in PC12 cells. Neurosci Lett 30 Erickson RW, Langel-Peveri P, Traynor-Kaplan AE, Heyworth PG, 1999; 265: 127–130. Curnutte JT. Activation of human neutrophil NADPH oxidase by 24 Yoshimura S-I, Nakashima S, Ohguchi K, Sakai H, Shinoda J, Sakai phosphatidic acid or diacylglycerol in a cell-free system. J Biol N, Nozawa Y. Differential mRNA expression of phospholipase D Chem 1999; 274: 22243–22250. (PLD) isozymes during cAMP-induced differentiation in C6 glioma 31 Burke JR, Davern LB, Gregor KR, Owczarczak LM. Differentiation cells. Biochem Biophys Res Commun 1996; 225: 494–499. of U937 cells enables a phospholipase D-dependent pathway of

25 Min DS, Shin KS, Kim E-G, Kim SR, Yoon SH, Kim M-S, Jo Y- cytosolic phospholipase A2 activation. Biochem Biophys Res H. Down-regulation of phospholipase D during differentiation of Commun 1999; 260: 232–239. mouse F9 teratocarcinoma cells. FEBS Lett 1999; 454: 197–200. 32 Geny B, Stutchfield J, Cockcroft. S. Phorbol ester inhibits polypho- 26 Moritz A, De Graan PNE, Gispen WH, Wirtz KWA. Phosphatidic sphoinositide phosphodiesterase activity stimulated by either Ca2+, acid is a specific activator of phosphatidylinositol-4-phosphate fluoride or GTP analogue in HL-60 membranes and in permea- kinase. J Biol Chem 1992; 267: 7207–7210. bilised HL-60 cells. Cell Signal 1989; 1: 165–172. 27 Jenkins GH, Fisette PL, Anderson RA. Type I phosphatidylinositol 33 Geny B, Cost H, Barreau P, Basset M, Le Peuch C, Abita J-P, 4-phosphate 5 kinase isoforms are specifically stimulated by phos- Cockcroft S. The differentiating agent, retinoic acid, causes an phatidic acid. J Biol Chem 1994; 269: 11547–11554. early inhibition of inositol lipid-specific phospholipase C activity 28 Liscovitch M, Chalifa V, Pertile P, Chen C-S, Cantley LC. Novel in HL-60 cells. Cell Signal 1991; 3: 11–23. function of phosphatidylinositol 4,5-bisphosphate as a cofactor for 34 Gre´goire C, Welch H, Astarie-Dequeker C, Maridonneau-Parini I. brain membrane phospholipase D. J Biol Chem 1994; 269: Expression of azurophil and specific granule proteins during differ- 21403–21406. enciation of NBA cells in neutrophils. J Cell Physiol 1998; 175: 29 Tool ATJ, Blom M, Roos D, Verhoeven AJ. Phospholipase D- 203–210. derived phosphatidic acid is involved in the activation of the CD11b/CD18 integrin in human eosinophils. Biochem J 1999; 340: 95–101.

Leukemia