Cytochemistry and Ultrastructural Morphometry of Cultured HL60 Myeloid Leukemia Cells1 Richard T

[CANCER RESEARCH 47, 4932-4940, September 15, 1987] Cytochemistry and Ultrastructural Morphometry of Cultured HL60 Myeloid Leukemia Cells1 Richard T. Parmley,2Dianne T. Akin, James C. Barton, Charles S. Gilbert, and Joseph M. Kinkade,Jr.

Departments of Pediatrics and Pathology, University of Texas Health Science Center, San Antonio, Texas 7S284-7810 [R. T. P., C. S. GJ; Department of Biochemistry, Emory University School of Medicine, Atlanta, Georgia 30322 [D. T. A., J. M, K.J; and Department of Medicine, University of Alabama, Veterans Administration Medical Center, and Comprehensive Cancer Center, Birmingham, Alabama 35294 fJ. C. B.J

ABSTRACT oxidase (11, 12), OZI staining of reducing substances (13), HID-reactive sulfated glycoconjugates (14), and periodate-re- Cultured human myeloid leukemia (HL60) cells were characterized active vicinal glycols (15) in normal human promyelocytes. using ultrastructural cytochemical methods and differences identified when cells were compared for low (17 to 47), middle (69 to 100), and Although ultrastructural studies have compared HL60 mor phology with that of normal promyelocytes, and DAB-reactive high (214 to 244) passages or to normal promyelocytes aspirated from bone marrow. Endoplasmic reticulum and transition structures (pre-GoIgi myeloperoxidase and HID-reactive sulfate in granules have compartment) of III,60 cells stained positively for peroxidase using been briefly described in HL60 cells (2, 7, 16,17), these studies diaminobenzidine but stained sparsely for reducing groups with osmium- have not: (a) related these staining characteristics to other zinc iodide. Staining of Golgi elements was relatively indistinct with HL60 cell organdÃes involved in granule genesis; (b) compared diaminobenzidine and strong with osmium-zinc iodide, in comparison to the cytochemical properties of HL60 cells to those of normal freshly harvested promyelocytes which have intense diaminobenzidine promyelocytes; and (c) compared HL60 cells at different cell and osmium-zinc iodide staining of the pre-Golgi and Golgi compart passages. The present study provides these comparisons using ments. Cytoplasmic polyribosomes were more numerous in middle and cytochemical methods previously used to characterize normal high passage cells, whereas dilatation of endoplasmic reticulum was less prominent in these cells. The mean granule size was significantly in promyelocytes. The results demonstrate significant differences creased in low passage cells, and staining of peroxidase was more between HL60 cells and normal promyelocytes, and between prominent by light and electron microscopy when compared to high different passages of HL60 cells. passage cells. Cytoplasmic granules demonstrated strong complex car bohydrate staining, indicating a lack of granule maturation in 111.60cells. MATERIALS AND METHODS Terminally differentiated myeloid cells were more frequent in low passage samples, and some neutrophil granule maturation appeared to occur HL60 cells were grown in continuous suspension culture in RPMI within these cells, whereas all eosinophil granules consistently remained 1640 medium (Sigma Chemical Co., St. Louis, MO) supplemented immature with intense complex carbohydrate staining and lack of crys with 2 mM glutamine (Sigma), 25 HIMsodium bicarbonate (Mallinck- talloid formation. These studies demonstrate significant differences be rodt Chemical Works, St. Louis, MO), penicillin (50 units/ml), strep tween 111611cells and normal promyelocytes, and also passage-dependent tomycin (50 g/ml) (Advanced Biotechnologies Inc., Silver Springs, maturational differences in 1II.60 cells. These differences should be MD), and 10% defined bovine calf serum (HyCIone Laboratories, considered in evaluating parameters of cell growth and maturation and in Logan, UT). Cultures were maintained in an atmosphere of 7.5% CO2 in air at 37Â°C.Cells were divided every 3 to 4 days by adding fresh the biochemical and enzymatic characterization of these cells. medium to dilute the cells to a density of 0.25 x IO6 cells/ml. Cells were counted using a hemocytometer and processed for electron mi INTRODUCTION croscopy while in logarithmic growth phase. Cell size was analyzed using a FACS Analyzer I (Becton-Dickinson, Sunnyvale, CA); single Cultured leukemia cell lines provide a unique tool to examine the proliferation, differentiation induction, and the expression histogram statistics were performed using 1000 gated events. Low passages used were passages 17, 20, 21, 30, 36, and 47; middle passage of a variety of biochemical and immunological markers in cells included passages 69, 75, 83, 98, and 100; and high passage cells hematopoietic cells. Cells of the HL60 line, established from a included passages 214, 218, 234, and 244. Passage was defined as the patient with myeloblastic/promyelocytic leukemia, resemble number of subcultures performed since the line was originally estab normal promyelocytes in their ability to synthesize specialized lished. Cells were subcultured twice weekly since originally obtained at lysosomes containing peroxidase, acid hydrolases, and chon- passage 13 (American Type Culture Collection, Rockville, MD). droitin sulfate (1-4). HL60 cells can be induced to differentiate Light Microscopy. Cells in culture medium were centrifuged at 1500 x g at 4Â°Cfor 15 min, and the supernatant was removed. The cells into segmented neutrophils by a variety of reagents and appear to retain some degree of pluripotency in that their differentia were resuspended in cold phosphate-buffered saline, pH 7.4, and cyto- tion into cells morphologically similar to normal neutrophils, centrifuged smears were prepared on glass slides at 200 rpm for 5 min (Cytospin 2; Shandon Southern Instruments, Inc., Sewickley, PA). The monocytes, macrophages, and eosinophils is possible (5, 6). smears were fixed and stained using the following methods: Wright's HL60 cells can also grow in immunodeficient mice, where they Giemsa; Sudan Black B; peroxidase; PAS; a-naphthyl butyrate esterase; retain some ultrastructural and cytochemical characteristics of ASD-chloroacetate esterase; and acid phosphatase ( 18,19). The smears myeloid cells (7). were examined by routine light microscopy; Cytoplasmic positivity in Numerous studies have examined the ultrastructural mor 100 to 250 cells/smear was evaluated and expressed as a percentage of phology of normal human promyelocytes (8-10). Similarly, total cells examined. The configuration of stain deposits was catego cytochemical studies have examined DAB3-reactive myeloper- rized as diffuse, finely granular, or coarsely granular. Freshly prepared peripheral blood smears from normal subjects were simultaneously Received 12/19/86; revised 6/18/87; accepted 6/22/87. stained as control preparations. The costs of publication of this article were defrayed in part by the payment Electron Microscopy. HL60 cells were centrifuged into a pellet and of page charges. This article must therefore be hereby marked advertisement in fixed with 3% glutaraldehyde-0.1 M cacodylate, pH 7.35, at 4Â°Cfor 60 accordance with 18 U.S.C. Section 1734 solely to indicate this fact. 1This work was supported in part by Veterans Administration medical research min. The cells were then resuspended in 0.1 M cacodylate with 7 g/dl funds and USPHS Grant CA22294 from the National Cancer Institute. sucrose, pH 7.35. Samples for myeloperoxidase staining were incubated 2To whom requests for reprints should be addressed. 3The abbreviations used are: DAB, 3,3'-diaminobenzidine; PAS, periodic acid- 30 min in a DAB solution prepared by adding 5 mg of DAB-4HC1 to 10 ml of 0.05 M Tris-HCI buffer, pH 7.6, containing 0.01% H2O2 (20). Schiff; HID, high iron diamine; OZI. osmium zinc iodide; PA, periodic acid; TCH-SP, thiocarbohydrazide-silver proteinate. Control samples were similarly incubated in medium without H2O2. 4932

Sulfated glycoconjugates were stained en block for 18h with Spicer's low, middle, and high passages had granular positivity which HID solution (21) as described previously for normal promyelocytes was slightly less intense in high passage cells. Peroxidase posi (13). Samples for evaluation of intrinsic density were comparably tivity was present in almost all cells of each passage, but was incubated in a control medium (14), an acid MgCl2 solution at pH 1.4. most intense in low and middle passage cells. With the PAS OZI staining was performed as described previously for marrow cells technique, 95 Â±1% and 96 Â±6% of the cells in low (n = 3) and (13). Specimens for vicinal glycol staining of the Golgi lamellae and middle (n = 3) passages, respectively, had positive staining, granules of HL60 cells were incubated for 3 h with porcine pancreatic

Table l HL60 organelle variation with passage number Ten random consecutive cells were scored for each passage. All organelles were scored for each cell profile. Type I granules contained a lucent rim occupying approximately 25% or more of the granule diameter, whereas type II granules contained uniformly dense material. Type I granules were significantly larger in passage 17 than passages 83 and 214 with P values of <0.01 and 0.05. respectively, when calculated by the Student Â»test.A similar difference in type H granules could not be demonstrated. In general, total granules for passage 17 were larger than passages 83 and 214 with a /'value -0.001 Type I granules were more frequent in passage 17 than passages 83 and 214 with a /"value of <0.01 and <0.005, respectively, whereas a difference could not be as clearly demonstrated for type II and total granules (I' > 0.1). Differences in mitochondria! size with passage number could not be demonstrated, whereas mitochondria! frequency tended to be less in passage 17. (>mr')Passage17 Organelle area cytoplasm)Granulefrequency (per 10 //m;

I0.373 II0.084 II0.246 1 and I1.56Â± II1.44Â± total3.21 Â±0.210" (115)* Â±0.062 (90) (205) Â±0. (106) 1.03 1.63 Â± Â±0.83 83 0.1 16 Â±0.072 (19) 0.026 Â±0.018 (118) 0.038 Â±0.042 (137) 0.202 Â±0. 134 (118) 0.45 Â±0.53 2.75 Â±1.69 3.19 Â±1.52 3.18 Â±1.28 0.069"214Granule0.1 50 Â±0.089 (27)Granule0.053 Â±0.046 (129)Granules Â±0.067 (156)Mitochondria0.259Â±0.9549330.234 Â±0.140179 (123)Organelle0.48 Â±0.45Granule2.25 Â±0.97Granule2.70 Â±1.731.01Mitochondria1.672.15 Mean Â±SD. 6 Numbers in parentheses, number of organelles counted.Â±0.216

^' - A fcÂ£v ".'â€¢ K Ã•3jj ÃŒfe^iÂ£s*

.â€¢ER

Fig. 1. HL60 cells obtained at passage 17 contain numerous cytoplasmic Fig 3. HL60 cells obtained at passage 214 contain cytoplasmic granules which granules. The majority of granules do not appear fully condensed. There are are somewhat fewer, smaller, and more condensed (arrows: Table 1) than in low several segments of dilated rough endoplasma reticulum (ER). Thin section passage cells (cf. Fig. 1). Thin section stained with uranyl acetate and lead citrate, stained with uranyl acetate and lead citrate. Bar, l .im. Â¡Â¡ar] m

M J

Â» / :. . /. G ER

â€¢â€¢

!sA> Fig. 2. An enlarged portion of cytoplasm from a passage 17 III do cell. Fig. 4. Polyribosomes (small arrows) are more numerous, and rough endo- containing predominantly monoribosomes (small arrows), dilated rough endoplasmic reticulum (Â£R)is less dilated and therefore less prominent in high passage plasmic reticulum (ER), Golgi lamellae and vesicles (G), and variably condensed cells (passage 214 shown here) than low passage cells. Golgi apparatus ((7), granules (large arrows). Mitochondrion (A/). Thin section stained with uranyl cytoplasmic granules (large arrows). Thin section stained with uranyl acetate and acetate and lead citrate. Bar, I win. lead citrate. Bar, 1 Â»mi. 4934

X / X

ER S - ER

Ã¨- ER .

Fig. S. This HL60 cell (passage 47) demÃ³nstrales DAB staining in rough endoplasmic reticulum (ER) and granules (arrows) similar to normal promyelocytes ( 10- 12). Thin section not counterstained. Bar, 1 /

^ â€¢>

' .^

Fig. 7. HL60 cells (passage 17) demonstrate intense OZI staining of reducing Fig. 8. A region at higher magnification shows intense OZI staining of the substances in cytoplasmic and Golgi vesicles Â«/]and mitochondria (A/) like abundant vesicles and lamellae in the ru-Golgi area (<â€¢),whereasrra/u-Golgi normal promyelocytes (13) but lack significant nuclear envelope and rough vesicles (/), condensing vacuoles ( V ), and granules (G) lack staining. Mitochon endoplasmi reticulum staining. Thin section not counterstained. Bar, I /mi. dria (A/), nucleus (A/). Thin section not counterstained. Bar, 1 jim.

\ M /

N /

Fig. 9. The majority of cytoplasmic granules (arrows) in HL60 cells (passage 47) demonstrate prominent HID-TCH-SP staining of sulfated glycoconjugates seen in immature granules of normal promyelocytes (14), whereas few, if any, granules lack staining. This finding suggests a lack of granule maturation in these cells. Mitochondria (A/). Specimen not osmicated: thin section not counterstained. Bar, l Â¿im. Fig. 10. This differentiated HL60 cell from passage 17 contains a multilobed or folded nucleus (Â¿V)withcondensed chromatin. HID-TCH-SP staining of sulfated glycoconjugates is masked in presumed mature primary granules (large arrows). Smaller tertiary granules are (small arrows) stained, however. Specimen not osmicated; thin section not counterstained. Bar, 1 mu 4936

Downloaded from cancerres.aacrjournals.org on September 30, 2021. © 1987 American Association for Cancer Research. HL60 ULTRASTRUCTURE staining in these sites in the present and previous studies (13). Eosinophil Differentiation. Less than 5% of cells contained The majority of mitochondria demonstrated intense matrix large (0.4 to 0.7 pm) granules with a central globular area of staining with OZI. lesser density consistent with immature eosinophil granules. HID-TCH-SP stained sulfated glycoconjugates prominently None of these granules contained a distinct crystalloid. Several in almost all cytoplasmic granules in HL60 cells from early, cells with eosinophil-like granules also contained granules that mid, and late passages (Fig. 9). The lack of granule maturation resembled condensing neutrophil primary granules (Fig. 13). as evidenced by the loss of HID-TCH-SP staining in HL60 Both granule types were DAB positive (Fig. 14). Occasional cells contrasted to that observed in normal marrow promyelo- cells corresponding to those with eosinophil-like differentiation cytes similarly studied and reported previously (14). Either a in morphological preparations had large granules whose sul homogeneous or rim staining pattern was observed in these fated glycoconjugates always stained intensely with HID-TCH- granules; this could depend on the plane of sectioning. Endo- SP in a rim or homogeneous distribution (Fig. 15). plasmic reticulum and distinct Golgi staining could not be demonstrated. The plasmalemma and mitochondria uniformly DISCUSSION lacked staining. A few differentiated cells with several nuclear lobes contained small HID reactive granules (Fig. 10) similar The present studies demonstrate significant morphological to tertiary granules described in normal myeloid cells (14). differences in HL60 cells related to passage number. High These granules generally corresponded in size and shape to the passage HL60 cells in this study closely resemble previously scattered DAB-negative granules seen in similar cells. Staining described very early promyelocytes from directly sampled mar appeared masked in some larger primary granules of these row cells (9) in their content of abundant polyribosomes and differentiated cells (Fig. 10). relatively small granules. Low passage cells contain fewer PA-TCH-SP stained vicinal glycols with increasing intensity polyribosomes, more segments of dilated rough endoplasmic from the cis- to trans-face of the Golgi lamellae (Fig. 11). reticulum, and larger cytoplasmic granules as described previ Endoplasmic reticulum lacked distinct staining. The majority ously for promyelocytes in a later stage of development (8-10). of cytoplasmic granules and numerous cytoplasmic vesicles Although both high and low passage cells resembled promye demonstrated moderate staining. Mitochondria lacked staining. locytes at various stages of development, no HL60 cells ob The plasmalemma stained moderately. Differences in staining served in this study had the abundant granules observed in with passage number could not be demonstrated. However, as either normal late promyelocytes (9, 10) or cells from patients noted above, the reactive granules appeared considerably with acute promyelocytic leukemia (25, 26). Also, HL60 cells smaller in high passage cells. In spontaneously differentiated in this study lack the morphologically distinct, defensin-rich cells PA-TCH-SP stained numerous cytoplasmic granules of dense granules seen in normal promyelocytes (24). These find variable size, possibly representing secondary and/or tertiary ings are consistent with a recent reappraisal of the original bone granules or immature primary granules. No TCH-SP staining marrow from which the HL60 line was derived suggesting, in was observed in control specimens incubated in acid MgCl2 fact, that the patient had acute myeloblastic leukemia rather followed by TCH-SP (Fig. 12) instead of HID-TCH-SP (Fig. than promyelocytic leukemia (4). The passage-dependent dif 12). Control specimens for PA-TCH-SP in which periodate ferences observed in this study indicate culture conditions, and oxidation was omitted lacked staining. repetitive passages may create a selective pressure for growth

H N

Fig. 11. PA-TCH-SP stains vicinal glycols in the Golgi apparatus (G) and granules (arrows) of HL60 cells (passage 47). The staining in the majority of granules further suggests a lack of granule maturation seen in normal promyelocytes (15). Specimen not osmicated; this section not counterstained. Nucleus (N). Bar, 1 /jtn. Fig. 12. Control cells, not treated with PA or HID, but incubated in TCH-SP lack the silver deposits seen in the PA-TCH-SP- and HID-TCH-SP-stained cells (cf. Figs. 9 and 11). Cytoplasmic granules (arrows), nucleus (jV). Specimen not osmicated; thin section not counterstained. Bar, I pm. 4937

Fig. 13. This differentiated HL60 cell from passage 17 contains abundant large granules with increased density consistent with forming eosinophil granules (Â£). Smaller granules with dense matrix material consistent with mature neutrophil primary granules (small arrow) are present (enlarged in inset). Less condensed larger granules consistent with immature neutrophil primary granules are also present (large arrows). Nuclear lobes ( N). glycogen (Gl). Thin section stained with uranyl acetate and lead citrate. Bar, 1 Â¡an.

Ã„. Fig. 14. DAB intensely stains both eosino- 1 phil-like granules (/:') and neutrophil-likegran ules (arrows) in this differentiated cell from passage 17. Thin section not counterstained. Bar, 1 jim.

Fig. 15. HID-TCH-SP intensely stains sulfated glycoconjugates in eosinophil-like gran ules (arrows) of HL60 cells. Granule matura â€¢. tion defined as the loss of HID-TCH-SP stain ing and the development of crystalloids as seen . in normal eosinophilic myelocytes (27) was not observed. Nucleus (A^. Specimen not osmicated; thin section not counterstained. Bar, 1

of more immature cells already present in the culture or result talloid (9-11) and loss of carbohydrate staining (14, 15), proc in dedifferentiation of myeloid cells. Nevertheless, regardless esses not observed in HL60 cells from either low or high of passage, HL60 cells demonstrate unique differences from passages. Development of crystalloids and loss of carbohydrate normal promyelocytes in granule number and type, lack of staining have also been observed in normal human eosinophil significant Golgi staining with DAB, and decreased endo- granule maturation (27). Like neutrophils, eosinophils derived plasmic reticulum staining with OZI. from HL60 cells lack granule maturation as evidenced by Maturation of cytoplasmic granules has been definitively persistent HID-TCH-SP staining of sulfated glycoconjugates described in normal promyelocytes and involves a sequence of or glycosaminoglycans and the lack of crystalloid formation. A morphological changes including development of a central crys chromosome 16 anomaly has been associated with a lack of 4938

Downloaded from cancerres.aacrjournals.org on September 30, 2021. © 1987 American Association for Cancer Research. HL60 ULTRASTRUCTURE crystalloid formation in human leukemic eosinophils (28), and between HL60 cells should be considered in studies that use a similar anomaly has been reported to occur in HL60 cells these cells to evaluate mechanisms for the regulation of cell despite the production of eosinophil major basic protein by proliferation and differentiation. these latter cells (2, 6). A variety of basic proteins are major components of neutrophil (29, 30) and eosinophil granules and ACKNOWLEDGMENTS form crystalloids in the latter cell type (31). Glycosaminogly- cans, stained here with the HID-TCH-SP method, may interact The authors wish to acknowledge the technical assistance of RenÃ©e and/or complex with basic proteins (32), and the persistent Koenig, Gloria Goode, and Anne Sereni. sulfate staining and lack of crystalloid formation in HL60 cells may reflect an abnormality in this complex formation during REFERENCES granule maturation. 1. Collins, S. .1.. Gallo, R. C., and Gallagher, R. E. Continuous growth and HL60 cells contain a prominent Golgi complex which can be differentiation of human myeloid leukemic cells in suspension culture. Nature demonstrated both with the methanolic uranyl acetate and lead (Lond.). 270: 347-349, 1977. 2. Gallagher. R., Collins, S., Trujillo, J., McCredie, K., Ahearn, M., Tsai, S., citrate counterstains used for ultrastructural morphology and Metzgar, R., Aulakh, G., Ting, R., Ruscetti, F.. and Gallo, R. Characteriza with the OZI method. However, unlike normal promyelocytes tion of the continuous, differentiating myeloid cell line (HL-60) from a patient with acute promyelocytic leukemia. Blood, 54: 713-733, 1979. (11, 12) HL60 cells from low and high passages in the present 3. Reiss, M., Maniglia, C. A., and Sartorelli, A. C. Reversible effects of retinoic study lacked prominent Golgi staining for peroxidase. A simi acid on glycosaminoglycan synthesis during differentiation of HL-60 leuke lar, but often less striking, decrease in Golgi staining is also mia cells. Cancer Res., 45: 2092-2097, 1985. 4. Dalton. W. T., Ahearn, M. J., McCredie, K. B., Freireich, E. J.. Trujillo, J. observed in directly sampled cells from patients with acute M., and Stass, S. A. The HL 60 cell line was derived from a patient with FAB-M2 and not FAB-M3. Blood. 6Â«:245a, 1986. myeloid leukemia (25, 33). The decreased staining of Golgi, 5. Breitman, T. R., Selonick, S. !â€¢'..andCollins, S. J. Induction of differentiation despite moderately strong DAB staining of endoplasmic retic- of the human promyelocytic leukemia cell line (HL-60) by retinoic acid. ulum, suggests that either rapid transit of peroxidase through Proc. Nati. Acad. Sci. USA, 77: 2936-2940. 1980. the Golgi apparatus occurs and is not detected with the present 6. Fischkoff, S. A.. Pollak, A., Gleich. G. J., Testa, J. R., Misawa, S., and Reber, T. J. Eosinophilic differentiation of the human promyelocytic leuke methodology, or that peroxidase fails to traverse the entire mia cell line HL-60. J. Exp. Med., 160: 179-196, 1984. Golgi apparatus in HL60 cells. The latter interpretation is 7. Machado, E. A., Gerard, D. A., Lozzio, C. B., Mitchell, J. R., and Olde, D. consistent with our recent observation in HL60 cells (34, 35) W. Proliferation and differentiation of human myeloid leukemic cells in immunodeficient mice: electron microscopy and cytochemistry. Blood. 63: that monensin causes the accumulation of peroxidase in a low- 1015-1022,1984. density compartment distinct from DAB-negative, monensin- 8. Scott, R. !â€¢'...andHorn, R. G. Ultrastructural aspects of neutrophil granulo- induced Golgi vacuoles. cyte development in humans. Lab. Invest., 23: 202-215, 1970. 9. Ackerman, G. A. The human neutrophilic promyelocyte: a correlated phase HL60 cells also differ from normal cells (14) in their marked and electron microscopic study. Z. Zellforsch., 118:467-481, 1971. decrease in OZI-reactive reducing substances in the nuclear 10. Bainton, D. F., Ullyot, J. L., and Farquhar, M. G. The development of neutrophilic polymorphonuclear leukocytes in human bone marrow. J. Exp. envelope and endoplasmic reticulum, in contrast to their intense Med., 134:907-934, 1971. Golgi staining. The discrepancy in Golgi and endoplasmic 11. Dunn. W. B., Hardin, J. H., and Spicer, S. S. Ultrastructural localization of myeloperoxidase in human neutrophil and rabbit heterophil and eosinophil reticulum staining is consistent with the previous observation leukocytes. Blood, 32: 935-944, 1968. (36) that Golgi staining occurs independently of endoplasmic 12. Ackerman, G. A., and Clark, M. A. Ultrastructural localization of peroxidase reticulum staining. This is accentuated in other cells in which activity in normal human bone marrow cells. Z. Zellforsch., 117: 463-475, 1971. inhibition of protein synthesis eliminates endoplasmic reticu 13. Clark, M. A., and Ackerman, G. A. Osmium-zinc iodide reactivity in human lum staining, but has little effect on Golgi staining (36). Thus, blood and bone marrow cells. Anat. Ree., 770: 81-96, 1971. the relative lack of OZI staining in endoplasmic reticulum of 14. Parmley, R. T., Hurst, R. E., Takagi. J., Spicer, S. S., and Austin, R. L. Glycosaminoglycans in human neutrophils and leukemic myeloblasts: ultra- HL60 cells suggests decreased protein synthesis in comparison structural, cytochemical, immunologie, and biochemical characterization. with normal promyelocytes in which endoplasmic reticulum Blood, 61: 257-266, 1983. staining is intense (13). 15. Fittschen, C., Parmley, R. T.. Austin, R. L., and Crist, W. M. Vicinal glycol- staining identifies secondary granules in human normal and ChÃ©diak-Higashi The appearance of some differentiated cells with granules neutrophils. Anat. Ree., 205: 301-311, 1983. resembling both immature eosinophil granules and maturing 16. Luikart, S. D., Sackrison, J. L., and Thomas, V. Altered glycosaminoglycan production by HL-60 cells treated with 4-methylumbelliferyl-/3-D-xyloside. neutrophil primary granules suggests the HL60 cells can si Blood, 66: 866-872, 1985. multaneously differentiate along eosinophil and neutrophil 17. Ilimori. T., Tanaka, T., and Ohnuma, T. Ultrastructural peroxidase cyto pathways. A similar observation has been made previously in chemistry of three established human myelogenous leukemia cell lines, HL- 60, KG-1, and ML-2. Leuk. Res., 9: 913-919, 1985. HL60 cells (6), and intramyeloid infidelity has been described 18. Nelson, D. A., and Davey, F. R. Leukocyte peroxidase: Sudan Black B in patients with myeloid leukemia (37). Nevertheless, additional staining; Leukocyte esterases. In: W. J. Williams, E. Beutler, A. J. Erslev, and M. A. Lichtman (eds.), Hematology, Ed. 3. pp. 1646, 1648-1649. 1651- studies are needed to confirm this hypothesis. 1654. New York: McGraw-Hill, 1983. The differences in the morphology and cytochemistry among 19. Davey, F. R., and Nelson, D. A. Leukocytic acid phosphatase. In: W. J. HL60 cells and those of normal promyelocytes also appear to Williams. E. Beutler, A. J. Erslev, and M. A. Lichtman (eds.). Hematology, be applicable to directly sampled leukemic cells which are highly Ed. 3, pp. 1649-1651. New York: McGraw-Hill, 1983. 20. Graham, R. C., and Karnovsky, M. J. The early stages of absorption of heterogeneous. Although HID-TCH-SP staining of sulfated injected horseradish peroxidase in the proximal tubules of mouse kidney: glycoconjugates may be intense in directly sampled leukemic ultrastructural cytochemistry by a new technique. J. Histochem. Cytochem., 14: 291-302, 1966. cells, granule maturation appears to occur even in Auer rods 21. Spicer, S.S. Diamine methods for differentiating mucosubstances histochem- (14). Similarly, the lack of DAB Golgi staining and OZI endo ically. J. Histochem. Cytochem., 13: 211-234, 1965. plasmic reticulum staining observed in HL60 cells in this study 22. Sannes, P. L., Spicer, S. S., and Katsuyama, T. Ultrastructural localization of sulfated complex carbohydrates with a modified ion diamine procedure. is not observed in directly sampled promyelocytes (10, 13). J. Histochem. Cytochem., 27: 1108-1 111. 1979. Also, we have observed that leukemic myeloid cells directly 23. ThiÃ©ry,J. P. Mise en Ã©videncedes polysaccharides sur coupes fines en sampled from patients frequently demonstrate intense OZI microscopie Ã©lectronique.J. Microsc., 6:987-1018, 1967. 24. Rice, W. G., Ganz, T.. Kinkade, J. M., Jr., Selsted, M. E., Lehrer, R. I., and staining of endoplasmic reticulum4 in contrast to HL60 cells. Parmley, R. T. Defensin-rich dense granules of human neutrophils. Blood, In summary, our results indicate that age-related differences in press. 1987. 25. Breton-Gorius, J. B., and Houssay, D. Auer bodies in acute promyelocytic ' R. T. Parmley, unpublished observation. leukemia: demonstration of their Pine structure and peroxidase localization. 4939

Lab. Invest., 2Â«:135-141, 1973. of the guinea pig eosinophil major basic protein to the core of the granule. 26. Parmley, R. T., Dahl, G. V., Austin, R. L., Gauthier, P. A., and Denys, F. J. Cell Biol., 77: 702-713, 1978. R. Ultrastructure and cytokinetics of leukemic myeloblasts containing giant 32. Lindahl, I .. and HÃ¶Ã¶k,M.Glycosaminoglycans and their binding to biolog granules. Cancer Res., 39: 3834-3844. 1979. ical macromolecules. Annu. Rev. Biochem., 47: 385-411, 1978. 27. Parmley, R. T., Takagi, M., Spicer, S. S., Thrasher, A., and Denys, F. R. 33. Bainton, D. F., Friedlander, L. M., and Shohet, S. B. Abnormalities in Ultrastructural visualization of complex carbohydrates in eosinophilic leu granule formation in acute myelogenous leukemia. Blood, 49:693-703,1977. kocytes. Am. J. Anat., 165: 53-67, 1982. 34. Aiken, D. T., and Kinkade. J. M., Jr. Evidence for the involvement of an 28. I i HiMU.M. M., Larson, R. A., Bitter, M. A., Vardiman, J. W., Golomb, H. acidic compartment in the processing of myeloperoxidase in human promye- M . and Rowley. J. D. Association of inversion of chromosome 16 with locytic leukemia HL60 cells. Arch. Biochem. Biophys.. 255:428-436, 1987. abnormal marrow eosinophils in acute myelomonocytic leukemia. N. Engl. 35. Akin, D. T., Kinkade, J. M., Jr., and Parmley, R. T. Biochemical and J. Med.. 309: 630-636, 1983. ultrastructural effects of monensin on the processing, intracellular transpon 29. Zeya, H. 1., and Spitznagel, J. K. Cationic proteins of polymorphonuclear and packaging of myeloperoxidase into low and high density compartments leukocyte lysosomes. II. Composition, properties, and mechanism of anti of human leukemia (HL60) cells. Arch. Biochem. Biophys., in press, 1987. bacterial action. J. Bacterio!., 91: 755-762, 1966. 36. Locke, M.. and Huie. P. The mystery of the unstained Golgi complex 30. Ganz, T.. Selsted. M. E., Szklarek, D., Harwig, S. S. L.. Daher, K., Bainton, cisternae. J. Histochem. Cytochem.. 31: 1019-1032. 1983. D. F., and Lehrer, R. I. Defensins: natural peptide antibiotics of human 37. Smith, L. J., Curtis. J. E.. Messner, H. A.. Senn. J. S.. Furthmayer, H., and neutrophils. J. Clin. Invest., 76: 1427-1435, 1985. McCulloch, E. A. Lineage infidelity in acute leukemia Blood, 61: 1138- 31. Lewis, D. M., Lewis, J. C., Loegering, D. A., and Gleich, G. J. Localization 1145. 1983.

4940

Richard T. Parmley, Dianne T. Akin, James C. Barton, et al.

Cancer Res 1987;47:4932-4940.

Updated version Access the most recent version of this article at: http://cancerres.aacrjournals.org/content/47/18/4932

E-mail alerts Sign up to receive free email-alerts related to this article or journal.

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Subscriptions Department at [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://cancerres.aacrjournals.org/content/47/18/4932. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.