Inhibition of phorbol ester-dependent differentiation of human promyelocytic leukemic (HL-60) cells by sphinganine and other long-chain bases.

The effects of long-chain (sphingoid) bases on the phorbol ester-dependent differentiation of HL-60 cells were investigated since these molecules are potent inhibitors of protein kinase C (Hannun, Y. A., Loomis, C. R., Merrill, A. H., Jr., and Bell, R. M. (1986) J. Biol. Chem. 261, 12604-12609). After 24 h, low concentrations of sphinganine (1-5 microM blocked both cell adherence and the inhibition of growth in response to phorbol 12-myristate 13-acetate, as measured by cell number and acid phosphatase activity. Sphinganine and sphingosine decreased adherence by 50% at 1-3 microM; other long-chain bases were effective in parallel to their inhibition of protein kinase C. Sphinganine decreased the binding of [3H]phorbol dibutyrate by the phorbol receptor of HL-60 cells, protein kinase C, and inhibited the response of HL-60 cells to dioctanoylglycerol, a cell permeable activator of this enzyme. Long-chain base uptake by HL-60 cells was demonstrated with [3-3H]sphinganine and within 1-3 days much had been converted to ceramides. By day 3, most of the cells had recovered the ability to adhere and exhibited macrophage characteristics, whereas cells in suspension did not differentiate. The level of free sphinganine in HL-60 cells was determined to be 12.3 +/- 1.2 pmol/10(6) cells. These results establish that sphingoid bases inhibit protein kinase C in HL-60 cells and may function physiologically as negative effectors of this enzyme.

number and acid phosphatase activity. Sphinganine and sphingosine decreased adherence by 60% at 1-3 p~; other long-chain bases were effective in parallel to their inhibition of protein kinase C. Sphinganine decreased the binding of [3HJphorbol dibutyrate by the phorbol receptor of HL-60 cells, protein kinase C, and inhibited the response of HL-60 cells to dioctanoylglycerol, a cell permeable activator of this enzyme. Long-chain base uptake by HL-60 cells was demonstrated with [3-~H]sphing~ine and within 1-3 days much had been converted to ceramides. By day 3, most of the cells had recovered the ability to adhere and exhibited macrophage characteristics, whereas cells in suspension did not differentiate. The level of free sphinganine in HL-60 cells was determined to be 12.3 2 1.2 pmol/106 cells. These results establish that sphingoid bases inhibit protein kinase C in HL-60 cells and may function physiolo~caIly as negative e€fectors of this enzyme.
Diverse biochemical and cellular changes occur in response to extracellular agents that stimulate phosphatidylinositol hydrolysis, mobilization of Ca2+, and activation of a lipiddependent protein kinase termed protein kinase C (1-3).
Diacylglycerols produced from phosphatidylinositols are naturally occurring activators of protein kinase C; but cellular processes have been investigated using the cell permeable activators l-oleoy1-8-acetyIglycero1, l , Z -d i~t~o y l g l y~e r o l , and phorbol diesters, which are structurally related (1-5). Responses to these compounds range from acute metabolic effects, such as activation of the oxidative burst in neutrophils, to much more protracted responses, including stimulation or inhibition of growth and differentiation.
The mechanism of activation of protein kinase C by phos- The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "&vertisernent" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
pholipids, Ca2*, and diacylglycerols (phorbol esters) has been extensively investigated in uitro and in uiuo. In contrast, the ways in which protein kinase C is regulated negatively are less well defined. Sphinganine and other long-chain (sphingoid) bases have been recently discovered to be potent inhibitors of protein kinase C in uitro and in human platelets (6) and to block the oxidative burst in human neutrophils (7).
These molecules have the potential of serving as natural inhibitors in viuo because they, like diacylglycerols, could be mobilized from membrane lipids.
The possible effects of sphinganine on protein kinase Cdependent differentiation of HL-60 cells into macrophageand monocyte-like cells was investigated. HL-60 cells are a human promyeiocytic cell line that matures into a variety of cell types depending on the stimulant (8)(9)(10). Since the phor-bo1 ester receptor has been found to be protein kinase C (11) and differentiation can be effected by lY2-dioctanoylglycerol (12), it is likely that protein kinase C is involved in differentiation, although other effects of phorbol esters cannot be excluded (13,14).
When HL-60 cells were treated with sphing~ine, there was pronounced inhibition of cell adherence and other markers of diffe~ntiation in response to phorbol 18-myristate 13acetate (PMA'). The characteristics of the inhibition and the structural specificity suggested that these effects were due to inhibition of protein kinase C. Furthermore, sphinganine inhibited phorbol dibutyrate binding by HL-60 cells and differentiation of the cells in response to ~octanoylglycerol. These results suggest that long-chain bases can inhibit protein kinase C physiologically.
CeU Culture-The HL-60 cells (obtained from the American Type Culture Collection, ATCC CCL240) were grown at 37 "C as a suspension culture in 175 cmz Nunc tissue culture flasks (Vangard International, Neptune, NJ). The cells were subcultured at a density of 0.25 X IOe cells/ml and used between passage numbers 30 and 40. Incubation of HL-60 CeUs-Cells were centrifuged at 600 X g for 3 min, and added at 1 X 106cells/well in 12-well culture dishes. Medium, PMA, and long-chain bases (prepared as the 1:l molar complex with fatty acid-free bovine serum albumin) at the indicated concentrations were added for a total volume of 2.0 ml. After the desired times, the cells in suspension were counted and checked for viability (trypan blue exclusion) with a hemacytometer. The attached cells were quantitated by measuring the DNA content (17) and by assaying for acid phosphatase (18).
When 1,2-dioctanoylglycerol was used, the cells were treated essentially as described by Ebeling et al. (12). Sphinganine was added to the cells, then ~~~n o y l g l y c e r o l in 1 pl of ethanol to yield 100 p~, followed by additional maintenance additions of 20 PM diacylglycerol every 2 h for 12 h. ~C p~e~n t of Phorbot by ~p h i~~~~-C o m p e t i t i v e binding assays were conducted as described by Goodwin and Weinberg (19) and Ebeling et al. (12). Approximately 1 x IO6 cells/ml were incubated with 12 nM [*H]phorbol dibutyrate (8.3 Ci/mmol) and varying concentrations of sphinganine (1:l with bovine serum albumin) for 1 h at 37 "C. The cells were recovered on Millipore filters, washed, and counted. The data were corrected for nonspecific binding by subtracting the counts/min obtained in the presence of 300 nM PMA.
Kinetics of [3-3HlSphinganine Uptake by HL-60 Cells-Approximately 1 X lo6 HL-60 cells in 1 ml of medium were mixed with an equal volume of medium containing 2.5 p~ [3-3H]sphinganine (equimolar with bovine serum albumin). After varying time intervals, an aliquot of the cells was removed and counted, and a portion was extracted as described below. The extracts were applied to Silica Gel H plates and developed in CHCIJmethanol/S N NH,OH (40:10:1, v/v/v), air dried, sprayed with Amplify (Amersham Corp.), and subjected to fluorography. Radiolabel was observed in only three regions of the chromatogram, coincident with ceramides near the solvent front, sphinganine with an R p of approximately 0.45, and in a region near the origin that encompassed sphingomyelin and other more polar complex sphingolipids (RF of 0.1-0.2). AnalysC of ~~-c~~~ Bases-From 1-4 X IO7 cells were recovered by centrifugation, washed thrice with phosphate-buffered saline, and extracted immediately by a minor modification of the procedure of Bligh and Dyer (20): 1.5 ml of c~lorofo~/methanol(1:2) were added and mixed thoroughly; 1 ml each of chloroform and water were added and the two phases were separated by centrifugation; the upper phase was discarded, and the chloroform phase was washed twice with water and dried by passage through a small column containing Na,SO,.
The extracts were saponified in methanolic KOH (0.1 M, and incubated at 37 "C for 1 h) to remove es~r-con~ining g~ycerolipi~. The 2,4-dinitrophenyl derivatives were prepared according to Braun and Snell (21) as follows. The lipids were dissolved in 50 rl of methanol/ ether (l:l), then 0.5 ml of 0.2% f l u o r~i n i t r o~n z e n e (Sigma) in methanol/ether (1:l) and 0.5 ml of 2 M &BO3 (pH 9.6) were added. After incubating for 1 h at 37 "C, 2 ml of ether and 2 ml of water were added. The ether was collected and the aqueous layer was reextracted with an additional 1 ml of ether. The combined ether extracts were washed with 2 ml of water and dried through NaZSO,, and the solvent was removed under a stream of NO. The recovery of 13H]sphinganine (60%) was used to correct for losses during extraction. For each experiment, the 2,4-dinitrophenyl derivatives of standard sphingosine, sphinganine (dihy~osphingosiRe), and ph~osphingosine (Sigma) were also prepared. as means zk S.D. and the significance of differences between groups was evaluated with the Student's t test for unpaired data.

RESULTS
Effect of S~~~~u n~n e on Celt Growth-The effect of sphinganine on HL-60 cell viability and growth was investigated because sphinganine has been reported to alter growth and to be cytotoxic for Chinese hamster ovary cells (23). Sphinganine was chosen over sphingosine because the former is available commercially as a homogeneous compound whereas the latter is a mixture of various homologs.
Untreated cells doubled during the first 24 h (Fig. I), and this was not changed by 1 pM sphinganine, but both 2.5 and 5 HM limited growth. None of the cells exhibited a loss of cell viability for the first 24 h. By the second day, all concentrations of sphinganine were still somewhat inhibitory; 1 and 2.5 pht inhibited growth without cytotoxicity whereas 5 JLM resulted in significant cell death. The change in total cell numbers between days 2 and 3 indicated that growth inhibition had ceased for the cells in 1 and 2.5 p~ sphinganine. This may have been due to removal of sphinganine by metabolism (see below).
These effects d e~n d e d on both the cell number, which probably reflected surface dilution (6), and the sphinganine to albumin ratio (data not shown); therefore, these parameters were kept constant except where noted.
Effects on PMA-induced Adherence and Growth Inhibition-Upon adding 9 nM PMA, growth of the HL-60 cells was inhibited by 70% within 24 h and the majority of the cells (61%) attached to the Petri dish (Table I), which is typical for this cell line (10). When 1 p~ sphinganine was also added, the cells continued to grow and only 26% of the total adhered. These data establish that sphinganine prevented PMA-induced growth inhibition at a concentration where sphinganine itself did not affect growth (cf. Table I and Fig. 1).
Because the celis in sp~inganine cont~nued to grow, the number available for adherence in response to PMA was higher. This results in similar numbers of adherent cells (ie. 0.51 X IO6 for PMA plus sphinganine versus 0.76 X lo5 for PMA alone, or a 33% difference) while the adherent cells as a percent of the total was much lower (ie. a 74% difference). Adherence has been expressed as the percent of the total viable cells to normalize for differences in growth.
Acid Phosphatase Activities of Treated Cells-For a more quantitative index of differentiation, the acid phosphatase activities of suspended and attached cells were compared (Table 11). Essentially all of the acid phosphatase activity was associated with the cells in suspension until PMA treatment, when varying percentages were transferred from the media to the dish. Sphinganine caused a concentration-dependent increase in the activity remaining in suspension and a decrease in the adherent activities, which reflect inhibition of cell adherence.
Inhibition of Phrbol Dibutyrate Binding by Sphiqanine-The effect of sphinganine on phorbol dibutyrate binding was investigated because the blockage of attachment may be due to inhibition of protein kinase C, which is thought to be the phorbol ester receptor of HL-60 cells (11). Spinganine blocked [3Hlphorbol dibutyrate binding (Fig. 2), with 50% inhibition at approximately 15 phi. Addition of 20 pM bovine serum albumin alone did not alter binding significantly (8%). This concentration of sphinganine was higher than that resulting in 50% inhibition of attachment, but a higher cell number and shorter incubation time was used for binding. Therefore, less sphinganine would have been taken up by the cells in the binding experiment, and the effective concentration in the membrane would probably also be lower.
To test this possibility, the effects of sphinganine on growth, viability, and adherence were evaluated with the higher cell numbers used in the binding experiments. Beginning with lo6 cells/ml, 25 phi sphinganine had no effect on cell growth (after 24 h, cell numbers in the absence and presence of sphinganine were 1.71 f 0. 18

Sphinganine-Dioctanoylglycerol, a cell-permeant activator
of protein kinase C, also induces HL-60 cell differentiation (12). Dioctanoylglycerol was added at 100 p~ with or without sphinganine in an initial loading dose, and additional dioctanoylglycerol was given to the cells in maintenance doses (20 PM) every 2 h for 16 h total. Sphinganine reduced adherence by 50% at approximately 5 p~ (Table 111). This trend was observed in three separate experiments; however, more precise comparison were thwarted by variability in the response of the cells to dioctanoylglycerol, which must be added as described above to elicit differentiation (12).
Structural Specificity of the Inhibition-The concentration dependence of sphinganine inhibition of PMA-induced attachment is shown in Fig. 3; results of similar experiments using other long-chain bases are also shown. Sphinganine at 3 p~ caused 50% inhibition and sphingosine, the predominant long-chain base found in mammalian sphingolipids (24), caused 50% inhibition at 1 pM. Stearylamine, which is S t N Cturally related but lacks the 1,3-dihydroxy groups, effected similar inhibition at 10 p~. Other evidence for the minimal involvement of the 3-hydroxyl was similar inhibition by 3ketosphinganine (not shown).
Both the free amino and the long alkyl chain were important. Ceramides from bovine brain and N-palmitoyldihydrosphingosine were not inhibitory, nor was N-acetylsphinganine at up to 500 p~. Octylamine did not inhibit, nor did another short-chain analog of sphinganine, 1,3-dihydroxy-2-amino-3phenylpropane.
Effects of Sphinganine on Cell Morphology and Histochemical Parameters-Expression of most other signs of HL-60 cell differentiation requires longer time periods after treatment with PMA. For these experiments, the cells were examined on day 3 for adherence and acid phosphatase activity (Table IV) and morphology and the marker enzymes a-naptholacetate esterase and acid phosphatase (Table V).  (Table IV). The majority of the cells had lost promyelocyte morphology, and resembled macrophages with visibly higher a-naphtholacetate esterase activity (Table V). Acid phosphatase activities were higher for adherent cells when expressed as activity per l@ cells, increasing 3-fold upon addition of PMA with and without sphinganine. This is a typical response of HL-60 cells to PMA (25).
None of the few viable cells in suspension had clear signs of differentiation.
Kinetics of Sphinganine Uptake and Metabolism-The rate of disappearance of [3H]sphinganine from the culture medium and its appearance in the cells is shown in Fig. 4. Approximately half of the sphinganine was taken up from the medium by the cells within 6 h. Free sphinganine accounted for most of the cellular radiolabel (Fig. 5). A large fraction was rapidly incorporated into ceramides but little was found in more polar sphingolipids (sphingomyelin and glycolipids). By days 2 and 3, only 0.07 and 0.06 nmol of free [3H]sphinganine was associated with the cells. The radioactivity that could not be accounted for in lipids was in the aqueous phase and may reflect degradation, which produces 3H20. enous level of long-chain bases in HL-60 celIs was quantitated by high performance liquid chromatography. Sphingosine was the major free long-chain base detected (Le. >80% of the total) and was present at 12.3 f 1.2 pmol/106 cells. Since leukocytes contain approximately 5 nmol of sphingolipids/1O6 cells, this corresponds to about 0.2% of the total long-chain bases present in the cells (26). Previous studies have found that free long-chain bases are not artifacts of the isolation or derivatization procedures (22).

DISCUSSION
Naturally occurring long-chain (sphingoid) bases inhibit the phorbol ester-and diacylglycerol-induced adherence of HL-60 cells. Since sphinganine has been found to inhibit protein kinase C in uitro, and the structural specificity (6) was similar to the in~bition of differentiation, it appears that the effects of long-chain bases on HL-60 cells are due to inhibition of this enzyme? Further evidence for this was the displacement by sphinganine of phorbol dibutyrate from its receptor, protein kinase C (11).
Inhibition of protein kinase C by other "lipoidal amines," which include palmitoylcarnitine, polyamines, and CP-46,665-1, an antineoplastic compound, have been reported (27, 28). Treatment of intact HL-60 cells with palmitoylcarnitine blocked PMA-induced cell adhesion but not acid phosphatase activity (291, which suggests that only some of the effects of PMA on HL-60 cells are mediated via protein kinase C. A similar conclusion has been drawn from the different effects of PMA and 1-oleoyl-2-acetylglycerol on HL-60 cells, since both activate the C kinase but only the former induced differentiation (13,14).
In contrast to these findings, di~tanoylglycerol acts as a good analog of naturally occurring diacylglycerol activators of protein kinase C (12) and caused differentiation of HL-60 cells much like PMA (8). This suggests that PMA is inducing differentiation via protein kinase C activation. A possible ~x p~a n a t~o n for the differences is that PMA and dioctanoylglycerol may be accessible to protein kinase C in all cellular compartments (phosphopro~ins are found in cytoplasm and in or around the nucleus) (30) whereas 1-oleoyl-2-acetylgly-cero1 is not. Precedents for such intracellular sorting of lipids 3 Inhibition of protein kinase C by sphinganine and steawiamine " " may also explain the toxicity of these compounds (23,35).
have been seen in the work of Pagano and Sleight (31). Ebeling et al. (12) found that dioctanoylglycerol displaced all of the phorbol dibutyrate from its receptor in HL-60 cells, whereas 1-oleoyl-2-acetylglycerol only displaced half.
Our results do not establish whether or not all of the effects of PMA involved protein kinase C. Since adherent cells differentiated in dishes containing sphinganine, protein kinase C may be responsible for adherence and growth inhibition and the other phenotypic markers arise from other effects of PMA. Alternatively, the lack of differentiation of cells in suspension may indicate that growth inhibition and adherence are prerequisites for these changes. If so, once cells have adhered, sphinganine would have no further effect on differentiation.
Since several amine inhibitors are now known, the relevance of these molecules to the normal in vivo regulation of protein kinase C is worth considering. The levels of free iongchain bases in HL-60 cells suggest that these molecules are endogenous inhibitors of protein kinase C in uiuo.4 Unfortunately, we do not know the subcellular localization of the sphinganine which makes further comparison of intracellular and extracellular concentrations ambiguous.
We have recently demonstrated that free long-chain bases are not detected early in the biosynthesis of sphingolipids from [14C]serine,2 but appear at later times. This indicates that they arise from sphingolipid breakdown, rather than as intermediates of long-chain base formation. Furthermore, the majority of the newly synthesized long-chain bases were degraded with a half-life of approximately 8 h, which is much more rapid than the presumed rate of sphingolipid turnover (22).
The source of the free long-chain bases of HL-60 cells is unknown, but might be GM3 ganglioside, which increases during ~fferentiation of HL-60 cells and affects differentiation when added exogenously (32,331. Since the level of free long-chain bases for normal, immature leukocytes cells is not known, it i s possible that the amount observed in HL-60 cells is high and accounts for its arrested d~ferentiat~on. Sphingolipids provide a logical counterbalance to the activation of protein kinase C by diacylglycerols because they are primarily found in the plasma membrane, often interact with receptors, and are well known to undergo changes with differentiation and transformation (34). Long-chain bases may additionally provide an endogenous inhibitor to prevent the "accidental" activation of the C kinase by diacylglycerols that arise from biosynthetic or degradatory pathways and thereby influence the level of diacylglycerol necessary to overcome inhibition.