796 R.C. Montenegro et al. / Toxicology in Vitro 21 (2007) 795–800(Suh et al., 1995). New compounds capable of inducing Ficoll-Hypaque. PBMC were washed and resuspended at adiVerentiation are considered candidate agents for the pre- concentration of 3 £ 105 cells/ml in RPMI 1640 mediumvention and/or treatment of cancer. supplemented with 20% fetal bovine serum, 2 mM gluta- Pisosterol, a triterpene isolated from a Pisolithus tincto- mine, 100 U/ml penicillin, 100 g/ml streptomycin at 37 °Crius, has been reported to have multiple biological proper- with 5% CO2. Phytohemagglutinin (3%) was added at theties. It has been shown to strongly inhibit tumor cell growth, beginning of culture. After 24 h and 72 h of culture, cellsespecially in leukemia and melanoma cell lines, but to lack were treated with the test compound.an eVect on mouse erythrocytes and on the development ofsea urchin eggs (Montenegro et al., 2004). The study of alter- 2.3. Cytotoxicityations in sea urchin egg development is a suitable model fordetecting cytotoxic, teratogenic and antineoplastic activities The cytotoxic activity of pisosterol was evaluatedof new compounds (Jacobs and Wilson, 1986; Costa-Lotufo in PBMC and HL60 using the MTT assay (Mosmann,et al., 2002). This assay can detect such selective agents as 1983). This assay is based on the reduction of the yellow-inhibitors of DNA and RNA synthesis, protein synthesis colored 3-(4Ј,5Ј-dimethylthiazol-2Ј-yl)-2,5-diphenyl-tetra-inhibitors and inhibitors of microtubule assembly (Fusetani, zolium bromide (MTT) by mitochondrial dehydrogenase of1987). These data suggest a selectivity of pisosterol towards metabolically active cells to a purple-blue formazan. Cellsmammalian cells, and also a mechanism of action that were plated in 96-well plates (0.5 £ 105 cells/well for cellswould involve a speciWc structure or pathway, apart from in 100 L of medium). Pisosterol dissolved in 1% DMSOthose typically described for cytotoxic substances, which are was added to each well at a Wnal concentration of 0.39–expected to act on both systems. 25 g/mL. The cells were exposed to the drug for 24 or 72 h. The aim of this study was to evaluate whether the anti- The control group received the same amount of DMSO.proliferative eVects observed for pisosterol are related to Thereafter, the plates were centrifuged and the medium wascell diVerentiation induction using the HL-60 cell line. We replaced by fresh medium (200 L) containing 0.5 mg/mLalso investigated the role of pisosterol in proliferation and/ MTT. After 3 h incubation, the MTT formazan productor apoptosis induction after diVerentiation. Cell diVerentia- was dissolved in 150 L DMSO, and absorbance was mea-tion was determined by nitro blue tetrazolium (NBT) sured using a multiplate reader (Spectra Count, Packard,reducing activity, morphological changes, and -naphthyl Ont., Canada). Drug eVect was quantiWed as the percentageacetate esterase assay. Cell viability was determined by try- of control absorbance of reduced dye at 550 nm.pan blue exclusion, while the proliferation status wasassessed through BrdU incorporation. Cell death was eval- 2.4. Trypan blue exclusionuated by diVerential Xuorescent staining (acridine orange-ethidium bromide). The cytotoxicity of pisosterol in human Cell viability was determined by the trypan blue dyeperipheral blood mononuclear cells (PBMC) was also eval- exclusion test (Renzi et al., 1993) after incubation of HL-60uated. cells (105 cells/mL) with pisosterol at 1, 2 and 5 g/mL. Aliquots were removed from cultures after 24, 48 72 and2. Material and methods 96 h, and cells that excluded trypan blue were counted in a Neubauer chamber. 12-O-Tetradecanoylphorbol-13-ace-2.1. Pisosterol isolation tate (TPA) 0.1 g/mL was used as the positive control. Pisosterol was isolated from the ectomycorrhizal fungus 2.5. Analysis of morphological changesP. tinctorius collected in a Eucalyptus plantation at Brejo,Maranhão in Northeast Brazil as described previously Untreated and pisosterol-treated HL-60 cells were exam-(Montenegro et al., 2004). ined for morphological changes by light microscopy (Olym- pus, Tokyo, Japan). To evaluate morphology, cells were2.2. Cell line and cell culture harvested, transferred to cytospin slides, Wxed with ethanol for 1 min, and stained with hematoxylin–eosin. TPA (0.1 HL-60 human promyelocytic leukemia cells were g/mL) was used as the positive control.obtained from Children’s Mercy Hospital (Kansas City,MO, USA). The cells were maintained in RPMI 1640 2.6. NBT reducing activitymedium supplemented with 10% fetal bovine serum, 2 mMglutamine, 100 U/mL penicillin, 100 g/mL streptomycin at HL-60 cells (105 cells/mL) were cultured with pisosterol37 °C with 5% CO2. The cells were split every 3 day and at 1, 2 and 5 g/mL in RPMI medium containing 10% FBSwere diluted 1 day before each experiment. for 24, 48, 72 and 96 h after which the NBT reducing activ- Heparinized blood (from healthy, non-smoker donors ity was determined by the method described by Kohrokiwho had not taken any drug for at least 15 days prior to et al. (1988). BrieXy, after the incubation, cells were har-sampling) was collected and PBMC were isolated by a vested and mixed with freshly prepared solution containingstandard method of density-gradient centrifugation over 1 g/mL TPA and 2 mg/mL NBT, and incubated for 30 min
R.C. Montenegro et al. / Toxicology in Vitro 21 (2007) 795–800 797at 37 °C. The reaction was stopped by the addition of cold chromatin in the nucleus that distinguish them fromPBS (3 mL). The suspension was centrifuged and resus- necrotic cells, which have a uniform orange color (Cury-pended in PBS and transferred to cytospin slides, and the Boaventura et al., 2004).The cells were then classiWed ascells were then Wxed with ethanol for 1 min, and stained follows: live cells, apoptotic cells and necrotic cells. Thewith eosin. Two hundred cells were counted using light percentage of apoptotic cells and necrotic cells was thenmicroscopy. TPA (0.1 g/mL) was used as the positive calculated.control. 2.10. Statistical analysis2.7. -Naphthyl acetate esterase activity Data obtained from diVerent experiments are presented HL-60 cells (105 cells/mL) were incubated for 72 h with as means § SEM from at least three independent experi-pisosterol at 1, 2 and 5 g/mL in RPMI medium supple- ments and evaluated by analysis of variance (ANOVA) fol-mented with 10% FBS. After 72 h incubation, cells were lowed by Student Newman–Keuls test, using a signiWcanceharvested, transferred to cytospin slides, and the -naph- level of 5%.thyl acetate esterase was carried out using a cytochemicalkit from Sigma (91-A, Saint Louis, MO). DiVerentiated 3. Resultscells were examined by counting two hundred cells usinglight microscopy. TPA (0.1 g/mL) was used as the positive 3.1. EVects of pisosterol on HL-60 diVerentiation: timecontrol. and dose dependence2.8. Antiproliferative eVect (inhibition of DNA synthesis) The eVects of pisosterol on cell viability were deter- mined by trypan blue exclusion after 24, 48, 72 and 96 h of HL-60 cells (105 cells/mL) were added to 24-well tissue exposure (Fig. 1). As shown in Fig. 1, there was a reduc-culture plates (2 mL/well) and treated with pisosterol at 1, 2 tion in cell viability in a time- and dose-dependent manner.and 5 g/mL for 72 h. Doxorubicin (0.02 g/mL) was used With increasing exposure time, there was a signiWcantas the positive control. Ten microliters of 5-bromo-2Ј-deoxy- decrease in cell viability in the presence of pisosterol aturidine (BrdU, 10 mM) were added to each well and incu- 5 g/mL, leading to a reduction of 65%, 80% and 62% afterbated for 3 h at 37 °C before completing the 72 h period of 48, 72 and 96 h, respectively. Moreover, a signiWcantdrug exposure. To assay the amount of BrdU incorporated reduction in cell viability was also observed after 72 hinto DNA (Pera et al., 1977), cells were harvested, trans- exposure for pisosterol at all tested doses. The IC50 valuesferred to cytospin slides, and allowed to dry for 2 h at room obtained from the MTT assay in HL-60 cells after 24 andtemperature. Cells that had incorporated BrdU were 72 h of incubation were 5.3 and 5.9 g/mL, respectively.labeled by direct peroxidase immunocytochemistry utiliz- TPA (0.1 g/mL) also decreased the number of viable cellsing the chromogen diaminobenzidine. Slides were counter- by 78%, 94% and 93% after 48, 72 and 96 h, respectivelystained with hematoxylin, mounted, and coverslipped. (Fig. 1). The microscopic examination of pisosterol-Evaluation of BrdU positivity was performed using light treated cells demonstrated morphological changes, includ-microscopy (Olympus, Tokyo, Japan). Two hundred cells ing the appearance of large cytoplasmic vacuoles (Fig. 2).were counted per sample to determine the percentage of These morphological changes are compatible with cellpositive cells. 1002.9. DiVerential Xuorescent staining with acridine/orange- viable cells (x 104 cells/ml)ethidium bromide 75 * * HL-60 cells (105 cells/mL) were plated onto 24-well tis- 50sue culture plates (2 mL/well) and treated with pisosterol at *1, 2 and 5 g/mL. After 72 h of incubation, cells were pel- 25 * * *leted and resuspended in 25 L phosphate-buVered saline * *(PBS). Afterwards, 1 L of aqueous solution of AO/EB 0 C TPA 1 2 5 (μg/mL)(100 g/mL) was added and cells were observed under a pisosterolXuorescence microscope (Olympus, Tokyo, Japan). Acri- 24h 48h 72h 96hdine orange intercalates into the DNA giving it a greenappearance. Ethidium bromide also intercalates into DNA, Fig. 1. Time-course of HL-60 cells viability in the presence of pisosterol.making it appear orange, but it is only taken up by non- The viability of cells was determined by the trypan blue dye exclusion method after 24, 48, 72 and 96 h of incubation. Data are presented asviable cells. Live cells with intact membranes have a uni- means § SEM from three independent experiments in duplicate forform green color in their nuclei. Early apoptotic cells have human leukemia HL-60. ¤p < 0.05 compared to control by ANOVA fol-chromatin condensation with bright green colored nuclei. lowed by Student–Newman–Keuls test. C, control; TPA, phorbol ester,Late apoptotic cells have bright orange areas of condensed 0.1 g/mL.
798 R.C. Montenegro et al. / Toxicology in Vitro 21 (2007) 795–800Fig. 2. Microscopic analysis of hematoxylin/eosin-stained HL-60 cells. Cells were untreated (A) or treated with pisosterol (1 g/mL, B, and 5 g/mL, C)and examined by light microscopy (200£). Black arrows show large cytoplasmic vacuoles.diVerentiation phenomena (Ramirez et al., 2000). No cyto- 80toxicity was detected in PBMC after treatment with pisos- *terol. * NAE active cells (%) 60 To further conWrm the ability of pisosterol to inducecell diVerentiation, the NBT reduction assay was per- 40formed. After 48 h, cells exposed to pisosterol at all dosesstarted to diVerentiate. Pisosterol-induced cell diVerentia-tion was maximal at 72 h incubation, while TPA achieved 20its maximum eVect at 48 h (Fig. 3). It is clear that pisosterolinduces cell diVerentiation in a time- and dose-dependent 0manner. C TPA 1 2 5 (μg/mL) Pisosterol3.2. Analysis of mechanisms involved in pisosterol-inducedcell diVerentiation Fig. 4. -Naphthyl acetate esterase (NAE) activity in HL-60 cells treated with pisosterol for 72 h. Data are presented as means § SEM from two independent experiments in duplicate. ¤p < 0.05 compared to control by To further evaluate the pathway involved in pisosterol- ANOVA followed by Student–Newman–Keuls test. C, control; TPA,induced HL-60 diVerentiation, -naphthyl acetate esterase phorbol ester, 0.1 g/mL.activity was determined after 72 h of incubation. Pisosterolappeared to induce a monocyte-like diVerentiation, since 70HL-60 cells treated with this compound showed an increas- 60 BrdU positive cells (%)ing esterase activity in a dose-dependent manner, as shown 50in Fig. 4. 40 * The antiproliferative eVect of pisosterol was determinedby inhibition of DNA synthesis based on BrdU incorpora- 30tion into HL-60 proliferating cells. It appears that pisos- 20terol-treated cells continued to proliferate at all doses tested 10after 72 h, with a slight decrease at 5 g/mL (Fig. 5). Cell 0death was then evaluated by AO/EB Xuorescent staining. C D 1 2 5 (μg/mL) Pisosterol 75 Fig. 5. EVect of pisosterol on the 5-bromo-2Ј-deoxyuridine (BrdU) incor- * poration by HL-60 cells after 72 h of incubation. Negative control (C) was * * NBT positive cells (%) * * treated with the vehicle used for diluting the test substance. Doxorubicin 50 * (0.02 g/mL) was used as positive control (D). Data are presented as * mean £ values § SEM from three independent experiments. ¤p < 0.05 compared to control by ANOVA followed by Student–Newman–Keuls 25 * test. * * 0 (μg/mL) Apoptosis was observed in pisosterol-treated cells in a C TPA 1 2 5 dose-dependent manner, with a signiWcant increase at the Pisosterol highest dose tested (5 g/mL) (Fig. 6). 24h 48h 72h 96hFig. 3. Time-course of NBT reduction in HL-60 cells treated with pisos- 4. Discussionterol. DiVerentiation induction was determined by the NBT reductionafter 24, 48, 72 and 96 h of incubation. Data are presented as Deregulation between cell proliferation, apoptosis andmeans § SEM from three independent experiments in duplicate forhuman leukemia HL-60. ¤p < 0.05 compared to control by ANOVA fol- diVerentiation leads to the development of malignant celllowed by Student–Newman–Keuls test. C, control; TPA, phorbol ester, clones. Based on the understanding of tumor kinetics, new0.1 g/mL. strategies, such as induction of diVerentiation and apopto-
R.C. Montenegro et al. / Toxicology in Vitro 21 (2007) 795–800 799 100 simultaneously incorporate bromodeoxiudridine (BrdU) into DNA and exhibit a proliferating cell nucleus (PCNA) * 75 * (Liu et al., 1999). Some sublines continue to proliferate and show induced expression of CD14 and other monocytic cells (%) 50 markers when treated with 1 ,25-dihydroxyvitamin D3 (Studzinski et al., 1997). Our results with pisosterol are in * * 25 agreement with all the above observations, where treated cells continued to incorporate BrdU while undergoing 0 C D 1 2 5 (μg/mL) diVerentiation. Hematopoietic cell maturation and subse- Pisosterol quent induction of apoptosis seem to be related. However, induction of diVerentiation in hematopoietic cells has been Viable cells Apoptotic Cells Necrotic Cells associated with increased expression of certain anti-apopto-Fig. 6. EVects of pisosterol on HL-60 cell viability determined by acridine tic proteins, including Bcl-2, A1, and Mcl-1 (Leszczynieckaorange and ethidium bromide-staining (AO/EB) after 72 h incubation. et al., 2001). This appears to occur with pisosterol-inducedC, negative control; D, doxorubicin (0.02 g/mL). Data are presented as diVerentiation, since the cells still proliferate. Induction ofmeans § SEM from three experiments. ¤p < 0.05 compared to control byANOVA followed by Student–Newman–Keuls test. such anti-apoptotic proteins may serve to delay cell death in terminally diVerentiated cells. Nevertheless, leukemia cells exposed to diVerentiation inducers ultimately undergosis, have recently emerged in the Welds of cancer chemopre- an apoptotic process, generally as a rather late event. Dur-vention and chemotherapy. In this context, leukemia cells ing the early stages of diVerentiation, leukemia cells may becould be considered a very useful model in the development more, rather than less, resistant to apoptosis, suggestingof novel cell diVerentiation inducers. that a reciprocal relationship exists between these processes The HL-60 cell line was established from an acute mye- (Leszczyniecka et al., 2001). Our results indicate that HL60loid leukemia patient and provides a useful model system cells treated with pisosterol triggers apoptosis after 72 hfor studying leukemia cell diVerentiation and can also be treatment.monitored by changes in morphological, biochemical, and Considering that cancer treatment has severe side eVects,immunological properties (Yam et al., 1971). Early studies it is very important to determine how harmful new drugsdemonstrated that certain compounds, including interferon may be to normal dividing cells such as proliferating lym-(Degos, 1990; Harris et al., 1985), retinoids (Drach et al., phocytes (Zuco et al., 2002; Anazetti et al., 2003). Interest-1994) and 1 ,25-dihydroxyvitamin D3 (KoeZer et al., ingly, despite the fact that HL60 leukemia cells and1984), are potent inducers of HL-60 cell diVerentiation, and lymphocytes are derived from white blood cell lineages,appear to be clinically eVective against diVerent types of pisosterol seems to be selective for tumor cells, since nocancer (Leszczyniecka et al., 2001). DiVerent compounds toxicity against normal cells was observed in our assay.can induce changes in cell surface antigen expression and The mechanisms underlying the eVect of pisosterol inacquire diVerent functions indicative of cell diVerentiation leukemia cells indicates the induction of a monocytic cell-into granulocyte-like or monocyte/macrophage-like cells like diVerentiation. The further characterization of the(Ujihara et al., 1998). The changes in cell viability and mor- mechanisms involved in pisosterol-induced diVerentiationphology were the Wrst indications showing that pisosterol will allow the identiWcation of the key molecular target forinduces cell diVerentiation. The ability of treated cells to this compound to facilitate the development of new phar-reduce NBT implies functional diVerentiation, and the macological tools with potential therapeutic value in theresults demonstrated a time- and dose-dependent eVect on management of leukemia with fewer side eVects.NBT reduction, which was maximal at 72 h with 5 g/mLpisosterol. The NBT assay is rather non-speciWc, since both Acknowledgmentsmonocytes and granulocytes can stain positive. Therefore,the treated cells were assayed for -naphthyl acetate We thank CNPq, CAPES, FUNCAP, FINEP, BNB/esterase activity. The enzyme is found in monocytes, macro- FUNDECI and PRONEX for the Wnancial support in thephages and hystiocytes, and is virtually absent in granulo- form of grants and fellowship awards. We are grateful tocytes (Li et al., 1973). Our results indicate that HL-60 Silvana França dos Santos for technical assistance. We alsotreated with pisosterol triggers diVerentiation towards a thank Dr. A. Leyva for English language editing of themonocytic-like cell. manuscript. In some cells, a period of mitotic expansion, known asmaturation division, follows the addition of diVerentiationinducers. HL-60 cells typically undergo several maturation Referencesdivisions, and each starting cell generates about a dozen Anazetti, M.C., Melo, P.S., Duran, N., Haun, M., 2003. Comparative cyto-mature and non-dividing progeny. DiVerentiating myeloid toxicity of dimethylamide-crotonin in the promyelocytic leukemia cellcells start to acquire mature characteristics while they are line (HL60) and human peripheral blood mononuclear cells. Toxicol-still proliferating (Drayson et al., 2001). HL-60 cells can ogy 188, 261–274.
800 R.C. Montenegro et al. / Toxicology in Vitro 21 (2007) 795–800Costa-Lotufo, L.V., Cunha, G.M.A., Farias, P.A.M., Viana, G.S.B., Cunha, Liu, Q., VanHoy, R.W., Zhou, J.H., Dantzer, R., Freund, G.G., Kelley, K.A.M., Pessoa, C., Moraes, M.O., Silveira, E.R., Gramosa, N.V., Rao, K.W., 1999. Elevated cyclin E levels, inactive retinoblastoma protein, V.S.N., 2002. The cytotoxic and embryotoxic eVects of kaurenoic acid, and suppression of the p27(KIP1) inhibitor characterize early develop- a diterpene isolated from Copaifera langsdorYi oleo-resin. Toxicon 40, ment of promyeloid cells into macrophages. Molecular Cell Biology 19, 1231–1234. 6229–6239.Cury-Boaventura, M.F., Pompeia, C., Curi, R., 2004. Comparative toxicity Montenegro, R.C., Jimenez, P.C., Feio Farias, R.A., Andrade-Neto, M., of oleic acid and linoleic acid on Jurkat cells. Clinical Nutrition 23, Silva Bezerra, F., Moraes, M.E., de Moraes, M.O., Pessoa, C., Costa- 721–732. Lotufo, L.V., 2004. Cytotoxic activity of pisosterol, a triterpene isolatedDegos, L., 1990. DiVerentiating agents in the treatment of leukemia. Leu- from Pisolithus tinctorius (Mich.: Pers.) Coker & Couch, 1928. kemia Research 14, 717–719. Zeitschrift für Naturforschung C 59, 519–522.Drach, J., McQueen, T., Engel, H., AndreeV, M., Robertson, K.A., Collins, Mosmann, T., 1983. Rapid colorimetric assay for cellular growth and sur- S.J., Malavasi, F., Mehta, K., 1994. Retinoic acid-induced expression of vival: application to proliferation and cytotoxicity assays. Journal of CD38 antigen in myeloid cells is mediated through retinoic acid recep- Immunological Methods 16, 55–63. tor-alpha. Cancer Research 54, 1746–1752. Pera, F., Mattias, P., Detzer, K., 1977. Methods for determining the prolif-Drayson, M.T., Michell, R.H., Durham, J., Brown, G., 2001. Cell prolifera- eration kinetics of cells by means of 5-bromodeoxyuridine. Cell Tissue tion and CD11b expression are controlled independently during HL60 Kinetics 10, 255–264. cell diVerentiation initiated by 1,25alpha-dihydroxyvitamin D(3) or all- Ramirez, C.D., Sleiman, R.J., Catchpoole, D.R., Stewart, B.W., 2000. Mor- trans-retinoic acid. Experimental Cell Research 266, 126–134. phological and molecular evidence of diVerentiation during etoposide-Fusetani, N., 1987. Marine metabolites which inhibit development of echi- induced apoptosis in human lymphoblastoid cells. Cell Death and noderm embryos. In: Scheur, P.J. (Ed.), Biorganic Marine Chemistry. DiVerentiation 7, 548–555. Springer-Verlag, Berlin, Heidelberg, pp. 61–92. Renzi, D., Valtolina, M., Foster, R., 1993. The evaluation of a multi-end-Harris, P.E., Ralph, P., Gabrilove, J., Welte, K., Karmali, R., Moore, M.A., point cytotoxicity assay system. ATLA 21, 89–96. 1985. Distinct diVerentiation-inducing activities of gamma-interferon Studzinski, G.P., Rathod, B., Wang, Q.M., Rao, J., Zhang, F., 1997. Uncou- and cytokine factors acting on the human promyelocytic leukemia cell pling of cell cycle arrest from the expression of monocytic diVerentia- line HL-60. Cancer Research 45, 3090–3095. tion markers in HL60 cell variants. Experimental Cell Research 232,Jacobs, R.S., Wilson, L., 1986. Fertilized sea urchin egg as a model for 376–387. detecting cell division inhibitors. In: Aszalor, A. (Ed.), Modern Analy- Suh, N., Luyengi, L., Fong, H.H., Kinghorn, A.D., Pezzuto, J.M., 1995. sis of Antibiotics. Marcel Dekker, Inc., New York, pp. 481–493. Discovery of natural product chemopreventive agents utilizing HL-60KoeZer, H.P., Amatruda, T., Ikekawa, N., Kobayashi, Y., DeLuca, H.F., cell diVerentiation as a model. Anticancer Research 15, 233–239. 1984. Induction of macrophage diVerentiation of human normal and Ujihara, M., Nomura, K., Yamada, O., Demura, H., 1998. Establishment leukemic myeloid stem cells by 1,25-dihydroxyvitamin D3 and its Xuo- of a stable HL60 subline having the potential for monocytic diVerentia- rinated analogues. Cancer Research 44, 5624–5628. tion using granulocyte-macrophage colony-stimulating factor: possibleKohroki, J., Muto, N., Tanaka, T., Itoh, N., Inada, A., Tanaka, K., Dur- use for the study of monocytic diVerentiation and oxidative stress. Ath- ham, J., Brown, G., 1988. Induction of diVerentiation and apoptosis by erosclerosis 139, 301–306. dithizone in human myeloid leukemia cell lines. Leukemia Research 22, Yam, L.T., Li, C.Y., Crosby, W.H., 1971. Cytochemical identiWcation of 405–412. monocytes and granulocytes. American Journal of Clinical PathologyLeszczyniecka, M., Roberts, T., Dent, P., Grant, S., Fisher, P.B., 2001. 55, 283–290. DiVerentiation therapy of human cancer: basic science and clinical Zuco, V., Supino, R., Righetti, S.C., Cleris, L., Marchesi, E., Passer- applications. Pharmacology and Therapeutics 90, 105–156. ineGambacorti, C., Formelli, F., 2002. Selective cytotoxicity of betuli-Li, C.Y., Lam, K.W., Yam, L.T., 1973. Esterase in human leukocytes. Jour- nic acid on tumor cell lines, but not on normal cells. Cancer Letters nal of Histochemistry and Cytochemistry 21, 1–12. 175, 17–25.