796
TH E LANC ET , OCTOBE R
9. Matsanioti s, N ., K arpouz as, J., Economou-Maurou, C. J. Pediat . 1967,70, 810 . 10. H arri s, F., K out soul ieris, E. A rchs Dis. Childh. 1967,42,449. U . Daniels, D . M ., Simon, J. L. J. Pediat, 1968,72, 697. 12. Litm an , N . N. ibid . 1968,73,798. 13. Aarskog, D. Archs D is. Childh . 1968,44,454. 14. Shaheed, W . A., Rosenbloom, L . ibid. 1973,48,917 . 15. M cGirr, E. M ., Murray, 1. P. C.]. clin, Endocr, M etab. 1956, 16 160. 16. D iggle, J . H ., W eetch, R. S. Proc. R . S oc. M ed. 1958,51, 293. 17. N ickey, L . N . Am.]. D is. Child. 1960, 99,680. 18. Johnson, J. E., Cook, A. R.]. ciin, Endocr. Metab. 1962,22,665. 19. Ka y, C. J., Esselb orn , V. M . A m.J. Dis. Child. 1963, 106,411 . 20. Simon, A., Ludwig, c., Gofman, J. W. , Cook, G. H. Am. ]. Psychiat , 1954, 111, 139. ZI. Friedman, A. Pedi atri cs, Spr ingfield, 1955, 16, 55. 22. Bollack, S., Al bau m, J. Am.]. menl oDefi c, 1957,62, 275. 23. Pearse, J . J., Reiss, M ., Suwalski, R. T .]. d in. Endocr, Me /ab. 1963, 23,311 . 24. Saxena, K . M ., Pryles, C. V.]. Pediat, 1965,67,363. 25. D eaton, J. G .J. men t. Defie. R es. 1973, 17, 117. 26. Murphy, B. E . P., Pattee, C. J.J. din. Endocr, Metab. 1964,24, 187. 27 . Roberts, R. C., N ikolai, T. F . Clin. Chem. 1969, 15, 1132. 28. Engel, E., Forbes, A . P. Med icine, 1965,44, 135. 29. Fialkow, P. J. Lancet, 1964, i, 474. 30. Fialkow, P. J. Ann. N . Y . A cad. S ci. 1970, 171,500.
Preliminary Communication
25, 1975
Fig . i -Cryostat sections stai ned with and- act in ant ibodi es.
(a ) Cells from no rm al hum an skin (thin t r a nsplant ) showi ng no immunofl uorescent staining . (b) Squamo us-cell ca rci noma wi th br ight stai ning a t periph er y of each cell. Red uced to ~ fro m x 400 .
ments as seen by electron microscopy. Thi s increase in the amount of contractile proteins distingu ishes cancer cells from the cells of normal tissues. INTROD UCTION
INCREASE OF CONTRACTILE PROTEINS IN HUMAN CANCER CELLS G . G ABBI ANI
Department of Pathology , University of Geneva , Switzerland
P.
TR EN CH EV
E. ] . H O LBORO W
Medical Research Council Rh eumatism Unit , Canadian Red Cross Memorial Hospital, Taploto, Maidenh ead, Berk shire
Human cells from skin and mammarygland carcinomas show strong immunofluorescent staining with specific antibodies against smooth-muscle actin , myosin, heavy meromyosin and light meromyosin , and contain numerous microfila-
Summary
Contractile proteins are present in man y types of nonmuscle cells where they are "generally organised as a filamentous apparatus, and may be involved in cell movement , endoc ytosis, or exocytosis. 1 - ' The amount of filamentous material in a given cell varies considera bly with ch ange in its functional activities or microenvir onment. T hus. microfilament s 4-7 nm in diameter increase in fibroblasts of gran ulation tissue (myofibroblasts)," epidermal cells during the healing of an open wound," and aortic endothel ial cells during the earl y stage of hypertension.8 We found th at human car cinoma cells cont ain more contractile proteins, compared with normal tissues . ME THODS
Tw o techn iques were used: (a) immunofluorescent staini ng
'.j
/ t
-,
(;.,
l'ig. 2- E lect r on mi crograph ot ski n cell s. ' ai l-rom no rma l hu man epithelium wit h cyto pla sm containing tonofilam en ts. (b) From sq ua mo us-cell ca rcino ma wit h m an v mi crofilarncm s a t . ccll pcr iph er v a nd pseudopo d ia-like cell project ion s. Re d uced to f fro m x 25 000 . .
THE LA NCET, OCTOBER
25,1975
of various contractile proteins bv specific antibodies;4. (b) identification by electron microscopy of filaments (4-7 nm in diameter corresponding to actin, or 10-12 nm in diameter corresponding to myosin10) in the cytoplasm of tumour cells. Surgical specimens of 17 tumours were studied from patients with basal-cell carcinoma (4), squamous-cell carcinoma (3), and infiltrating mammary carcinoma with (5) or without (5) fibrosis. Control specimens came from normal human skin from necropsy material taken three hours alter death (1), from thin skin transplants (5), and from mammoplasty of normal mammary glands (3). The following antibodieswere used: (a) anti-actin autoantibodies obtained from 2 patients with chronic aggressive hepatitis? and (b) antibodies raised in rabbits by singleimmunisation with purified human smooth-muscle actin, myosin, light meromyosin, heavy meromyosin, and tropomyosin, as previously described." S RESULTS
In all control tissues, the immunofluorescent staining of epithelial cells was weak (fig. 1a), and some faint staining of mammary epithelium was visible, especially at the periphery of the cell facing the lumen of alveoli or ducts. In contrast, staining for actin, myosin, and light and heav y meromyosin was strongly positive in cancer cells (fig. 1b). There was no staining for tropomyosin, and cancer cells did not stain with normal serum. Staining was cytoplasmic, but more intense under the cell membrane and around the nucleu s, which itself was not stained. Basal-cell carcinomas were consistently less brightly stained than squamous-cell carcinomas, particularly with anti-actin antibodies. Electron microscopy demonstrated no microfilaments in normal epidermal cells (fig. 2a), except for the usual tonofilarnents (10 nm in diameter), but in mammary epithelium a few microfilaments were present at the peripheral pan of the cells facing the lumen, some extending into microvillous processes. Tumour cells from the skin and mammary gland contained consistently prominent microfilaments, and larger filaments (10-12 nm in diameter) were scattered among the microfilaments. The filamentous structures appeared as a network, and were usually located at the cell periphery and in cellular pseudopodia-like prolongations. These prolongations, as well as the microfilaments, were especially abundant in squamous carcinomas (fig. 2b). Microfilaments could also be seen within the cytoplasm and particularly in perinuclear areas, a distribution similar to that obtained by immunofluorescent staining for actin. DIS C USSION
These findings show that malignant cells contain more contractile proteins than normal cells. In all the specimens we examined, immunofluorescent staining for contractile proteins was brightest in cells at the growing edge of the tumour, and in isolated groups of metastatic cells around blood and lymph vessels. It is tempting to speculate that the increase of contractile proteins in human cancer cells is related to the increased motility of these cells, and hence to their invasiveness. At an y rate, the intensity of staining of tumour cells is a property of potential importance in distinguishing between malignant and benign growths. Requests for reprint s should be addressed 10 E.).H . References at f oot of next colum n
797
Hypothesis POLYAMINES AS PREDICTORS OF SUCCESS AND FAILURE IN CANCER CHEMOTHERAPY DIANE H. RUSSELL BRIAN G. M. DURIE SYDNEY E. SALMON Sect ion of Hematology and Oncology, Department of Internal M edicine, and Department of Pharmacology; Univ ersity of Arizona College of Medicine, Tucson, Arizona 85724, U.S.A.
Polyamines appear to have clinical utility for monitoring cancer therapy. Investigations on the patterns of urinary excretion of polyamines in patients receiving cancer chemotherapy and studies of the tritiated-thymidine labelling index of tumour cells in a su bgroup of these patients form the basis for a hypothesis that spermidine can serve as a marker of tumour-cell kill, whereas putrescine appears to reflect the proliferative behaviour or " growth fraction" of the tumour. Although tests of this hypothesis are necessary to establish the predictive value of polyamines for monitoring cancer therapy, it is clear that polyamines may provide important new information on tumour cytokinetic parameters.
Summary
INTRODUCTION
PRECISE markers of tumour behaviour are critical for effective monitoring of cancer therapy. Although the biochemical markers secreted by certain tumours (e.g., myeloma immunoglobulins, human chorionic gonadotropin [H.C.G.], o-fetoprotein, and carcinoembr yonic antigen [C. E.A .) ) have gained clinical utility as indicators of total tumour burden, there is still a great need for more general markers of tumour cytokinetic behaviour and tumour-cell kill in response to treatment. Polyamines may prove clinically useful in this respect since they are intimately involved with cellular growth and death processes in both normal and neoplastic tissues. During cell growth, intracellular polyamine levels are increased early after tissue stimulation. After cell death, extracellular levels are increased. The discovery of increased excretion of the polyamines putrescine, spermi dine, and sper mine in the urine of patients with cancer was reported in 1971.1•2 Since that time other investigators have confirmed that the excretion of at least two of the poly-
DR GABBIANI AND OT H E RS : REFERENC ES
I. Pollard, T . D., Weihung, R. D. c-«. R ev. Biochem . 1974.2, I. 2. Ishikawa, H., Bischoff, R., Hohzer, H.]. Cell Bioi. 1969,43, 3t 2. 3. Gabb ian i, G., Malaisse-Lagae, F., Blondel, B., Orci, L. E ndocrin ology , 1974,95,1 630 . 4. Tren chev, P.• Sneyd, P., Holborow, E. J. Clin . exp o Im mun, 1974,16, 125. 5. Holborow, E. j ., T renchev, P. S.• Dorling, ] ., Webb, ] . A nn. N . Y . A cad. Sc i. 1975,254,489. 6. Gabbi ani, G., Hir schel, B. I .• Ryan, G. B., Statkov, P. R., Maino. G .]. expo M ed. t9 72, 135, 719. 7. Gab bian i, G., Ryan, G. B.]. submicr, Cytot. 1974,6,143. 8. Gabbian i, G., Badonnel M-C., Rona, G_La b. In vest. 1975,32, 227. 9. Gebbian i, G., Ryan, G. 8. , Lamelin J-P., Vassalli, P., Crucha ud. A., Luscher , E. F. A m.]. Path . 1973,72,473. 10. Goldman, R. D.]. cell Bioi. 197t , 51, 763.
TH E LANC ET , OCTOBE R
9. Matsanioti s, N ., K arpouz as, J., Economou-Maurou, C. J. Pediat . 1967,70, 810 . 10. H arri s, F., K out soul ieris, E. A rchs Dis. Childh. 1967,42,449. U . Daniels, D . M ., Simon, J. L. J. Pediat, 1968,72, 697. 12. Litm an , N . N. ibid . 1968,73,798. 13. Aarskog, D. Archs D is. Childh . 1968,44,454. 14. Shaheed, W . A., Rosenbloom, L . ibid. 1973,48,917 . 15. M cGirr, E. M ., Murray, 1. P. C.]. clin, Endocr, M etab. 1956, 16 160. 16. D iggle, J . H ., W eetch, R. S. Proc. R . S oc. M ed. 1958,51, 293. 17. N ickey, L . N . Am.]. D is. Child. 1960, 99,680. 18. Johnson, J. E., Cook, A. R.]. ciin, Endocr. Metab. 1962,22,665. 19. Ka y, C. J., Esselb orn , V. M . A m.J. Dis. Child. 1963, 106,411 . 20. Simon, A., Ludwig, c., Gofman, J. W. , Cook, G. H. Am. ]. Psychiat , 1954, 111, 139. ZI. Friedman, A. Pedi atri cs, Spr ingfield, 1955, 16, 55. 22. Bollack, S., Al bau m, J. Am.]. menl oDefi c, 1957,62, 275. 23. Pearse, J . J., Reiss, M ., Suwalski, R. T .]. d in. Endocr, Me /ab. 1963, 23,311 . 24. Saxena, K . M ., Pryles, C. V.]. Pediat, 1965,67,363. 25. D eaton, J. G .J. men t. Defie. R es. 1973, 17, 117. 26. Murphy, B. E . P., Pattee, C. J.J. din. Endocr, Metab. 1964,24, 187. 27 . Roberts, R. C., N ikolai, T. F . Clin. Chem. 1969, 15, 1132. 28. Engel, E., Forbes, A . P. Med icine, 1965,44, 135. 29. Fialkow, P. J. Lancet, 1964, i, 474. 30. Fialkow, P. J. Ann. N . Y . A cad. S ci. 1970, 171,500.
Preliminary Communication
25, 1975
Fig . i -Cryostat sections stai ned with and- act in ant ibodi es.
(a ) Cells from no rm al hum an skin (thin t r a nsplant ) showi ng no immunofl uorescent staining . (b) Squamo us-cell ca rci noma wi th br ight stai ning a t periph er y of each cell. Red uced to ~ fro m x 400 .
ments as seen by electron microscopy. Thi s increase in the amount of contractile proteins distingu ishes cancer cells from the cells of normal tissues. INTROD UCTION
INCREASE OF CONTRACTILE PROTEINS IN HUMAN CANCER CELLS G . G ABBI ANI
Department of Pathology , University of Geneva , Switzerland
P.
TR EN CH EV
E. ] . H O LBORO W
Medical Research Council Rh eumatism Unit , Canadian Red Cross Memorial Hospital, Taploto, Maidenh ead, Berk shire
Human cells from skin and mammarygland carcinomas show strong immunofluorescent staining with specific antibodies against smooth-muscle actin , myosin, heavy meromyosin and light meromyosin , and contain numerous microfila-
Summary
Contractile proteins are present in man y types of nonmuscle cells where they are "generally organised as a filamentous apparatus, and may be involved in cell movement , endoc ytosis, or exocytosis. 1 - ' The amount of filamentous material in a given cell varies considera bly with ch ange in its functional activities or microenvir onment. T hus. microfilament s 4-7 nm in diameter increase in fibroblasts of gran ulation tissue (myofibroblasts)," epidermal cells during the healing of an open wound," and aortic endothel ial cells during the earl y stage of hypertension.8 We found th at human car cinoma cells cont ain more contractile proteins, compared with normal tissues . ME THODS
Tw o techn iques were used: (a) immunofluorescent staini ng
'.j
/ t
-,
(;.,
l'ig. 2- E lect r on mi crograph ot ski n cell s. ' ai l-rom no rma l hu man epithelium wit h cyto pla sm containing tonofilam en ts. (b) From sq ua mo us-cell ca rcino ma wit h m an v mi crofilarncm s a t . ccll pcr iph er v a nd pseudopo d ia-like cell project ion s. Re d uced to f fro m x 25 000 . .
THE LA NCET, OCTOBER
25,1975
of various contractile proteins bv specific antibodies;4. (b) identification by electron microscopy of filaments (4-7 nm in diameter corresponding to actin, or 10-12 nm in diameter corresponding to myosin10) in the cytoplasm of tumour cells. Surgical specimens of 17 tumours were studied from patients with basal-cell carcinoma (4), squamous-cell carcinoma (3), and infiltrating mammary carcinoma with (5) or without (5) fibrosis. Control specimens came from normal human skin from necropsy material taken three hours alter death (1), from thin skin transplants (5), and from mammoplasty of normal mammary glands (3). The following antibodieswere used: (a) anti-actin autoantibodies obtained from 2 patients with chronic aggressive hepatitis? and (b) antibodies raised in rabbits by singleimmunisation with purified human smooth-muscle actin, myosin, light meromyosin, heavy meromyosin, and tropomyosin, as previously described." S RESULTS
In all control tissues, the immunofluorescent staining of epithelial cells was weak (fig. 1a), and some faint staining of mammary epithelium was visible, especially at the periphery of the cell facing the lumen of alveoli or ducts. In contrast, staining for actin, myosin, and light and heav y meromyosin was strongly positive in cancer cells (fig. 1b). There was no staining for tropomyosin, and cancer cells did not stain with normal serum. Staining was cytoplasmic, but more intense under the cell membrane and around the nucleu s, which itself was not stained. Basal-cell carcinomas were consistently less brightly stained than squamous-cell carcinomas, particularly with anti-actin antibodies. Electron microscopy demonstrated no microfilaments in normal epidermal cells (fig. 2a), except for the usual tonofilarnents (10 nm in diameter), but in mammary epithelium a few microfilaments were present at the peripheral pan of the cells facing the lumen, some extending into microvillous processes. Tumour cells from the skin and mammary gland contained consistently prominent microfilaments, and larger filaments (10-12 nm in diameter) were scattered among the microfilaments. The filamentous structures appeared as a network, and were usually located at the cell periphery and in cellular pseudopodia-like prolongations. These prolongations, as well as the microfilaments, were especially abundant in squamous carcinomas (fig. 2b). Microfilaments could also be seen within the cytoplasm and particularly in perinuclear areas, a distribution similar to that obtained by immunofluorescent staining for actin. DIS C USSION
These findings show that malignant cells contain more contractile proteins than normal cells. In all the specimens we examined, immunofluorescent staining for contractile proteins was brightest in cells at the growing edge of the tumour, and in isolated groups of metastatic cells around blood and lymph vessels. It is tempting to speculate that the increase of contractile proteins in human cancer cells is related to the increased motility of these cells, and hence to their invasiveness. At an y rate, the intensity of staining of tumour cells is a property of potential importance in distinguishing between malignant and benign growths. Requests for reprint s should be addressed 10 E.).H . References at f oot of next colum n
797
Hypothesis POLYAMINES AS PREDICTORS OF SUCCESS AND FAILURE IN CANCER CHEMOTHERAPY DIANE H. RUSSELL BRIAN G. M. DURIE SYDNEY E. SALMON Sect ion of Hematology and Oncology, Department of Internal M edicine, and Department of Pharmacology; Univ ersity of Arizona College of Medicine, Tucson, Arizona 85724, U.S.A.
Polyamines appear to have clinical utility for monitoring cancer therapy. Investigations on the patterns of urinary excretion of polyamines in patients receiving cancer chemotherapy and studies of the tritiated-thymidine labelling index of tumour cells in a su bgroup of these patients form the basis for a hypothesis that spermidine can serve as a marker of tumour-cell kill, whereas putrescine appears to reflect the proliferative behaviour or " growth fraction" of the tumour. Although tests of this hypothesis are necessary to establish the predictive value of polyamines for monitoring cancer therapy, it is clear that polyamines may provide important new information on tumour cytokinetic parameters.
Summary
INTRODUCTION
PRECISE markers of tumour behaviour are critical for effective monitoring of cancer therapy. Although the biochemical markers secreted by certain tumours (e.g., myeloma immunoglobulins, human chorionic gonadotropin [H.C.G.], o-fetoprotein, and carcinoembr yonic antigen [C. E.A .) ) have gained clinical utility as indicators of total tumour burden, there is still a great need for more general markers of tumour cytokinetic behaviour and tumour-cell kill in response to treatment. Polyamines may prove clinically useful in this respect since they are intimately involved with cellular growth and death processes in both normal and neoplastic tissues. During cell growth, intracellular polyamine levels are increased early after tissue stimulation. After cell death, extracellular levels are increased. The discovery of increased excretion of the polyamines putrescine, spermi dine, and sper mine in the urine of patients with cancer was reported in 1971.1•2 Since that time other investigators have confirmed that the excretion of at least two of the poly-
DR GABBIANI AND OT H E RS : REFERENC ES
I. Pollard, T . D., Weihung, R. D. c-«. R ev. Biochem . 1974.2, I. 2. Ishikawa, H., Bischoff, R., Hohzer, H.]. Cell Bioi. 1969,43, 3t 2. 3. Gabb ian i, G., Malaisse-Lagae, F., Blondel, B., Orci, L. E ndocrin ology , 1974,95,1 630 . 4. Tren chev, P.• Sneyd, P., Holborow, E. J. Clin . exp o Im mun, 1974,16, 125. 5. Holborow, E. j ., T renchev, P. S.• Dorling, ] ., Webb, ] . A nn. N . Y . A cad. Sc i. 1975,254,489. 6. Gabbi ani, G., Hir schel, B. I .• Ryan, G. B., Statkov, P. R., Maino. G .]. expo M ed. t9 72, 135, 719. 7. Gab bian i, G., Ryan, G. B.]. submicr, Cytot. 1974,6,143. 8. Gabbian i, G., Badonnel M-C., Rona, G_La b. In vest. 1975,32, 227. 9. Gebbian i, G., Ryan, G. 8. , Lamelin J-P., Vassalli, P., Crucha ud. A., Luscher , E. F. A m.]. Path . 1973,72,473. 10. Goldman, R. D.]. cell Bioi. 197t , 51, 763.
