British Journal
of
Haematology, 1976,32, 565.
Lysosomal Acid Hydrolases in Human Lymphocyte
Subpopula tions R. D. BARRAND S. PERRY* Ofice ofthe Director, Division of Cancer Treatment, National Cancer Institute, and * 0 8 c e o f the Director, National Institutes of Health, Bethesda, Maryland (Received 5 August 1975 ; acceptedfor publication 9 October 1975)
SUMMARY.Acid phosphatase and beta-glucuronidase activities have been determined cytochemically in T, B and null lymphocytes as part of an effort to characterize the human haemopoietic stem cell. A combination of weak acid phosphatase activity and strong beta glucuronidase staining, in the form of a single large granule, has been shown to be specific for T cells. On the basis of this approach alone, non-T cells could not be further subclassified. Further cytochemical evaluation is being explored. Following our recent identification of the human haemopoietic stem cell in normal peripheral blood (Barr et al, 197~a,b) we have embarked on its further characterization. In this endeavour we have responded to Lajtha’s plea (1973) for a cytochemical assessment of the pluripotent cell. Recent evidence, from work on human and other mammalian material, has suggested that discrete populations of lymphocytes may be distinguished by their differential content of enzymes located in the lysosomes. It appeared appropriate, therefore, to pursue this possibility, by a cytochemical evaluation of acid phosphatase (E.C.3. I .3.2) and beta-glucuronidase (E.C.3.2.1.3I) in purified lymphocyte fractions (T, B and null cells) harvested from normal human peripheral blood. MATERIALS AND METHODS Peripheral venous blood was obtained from 10 normal adult volunteers and heparinized at 40 unitsiml. Mononuclear cell concentrates were prepared by isopycnic sedimentation on Ficoll-Hypaque gradients (Boyum, 1968). Pure thymus-dependent lymphocytes (T cells) were obtained by the formation of rosettes with neuraminidase-treated sheep erythrocytes (Lohrmann et al, 1974);the rosettes were subsequently harvested by density sedimentation and the red cells removed by cold shock lysis (Fallon et al, 1962). Non-E rosetting cells were further separated by velocity sedimentation in a sucrose gradient at unit gravity (Brubaker & Evans, 1969). Concentrates of bursa-equivalent lymphocytes (B cells) were prepared by rosette formation with EAC’ complexes (Shevach et al, 1974). For this purpose sheep erythrocytes (E), labelled with rabbit anti-sheep erythrocytes IgG (A), were obtained from Cordis Laboratories, Miami, Florida 33137, U.S.A., and incubated with fresh serum (C’), from leukaemic Correspondence: Dr Seyniour Perry, Special Assistant to the Director, National Institutes of Health, Bldg. Room 111, Bethesda, Maryland 20014, U.S.A.
565
I,
R. D. Burr and S . Perry
566
AKR mice as a source of complement (Lay & Nussenzweig, 1968). The mice were supplied by Microbiological Associatx, Bethesda, Maryland 20014, U.S.A. Acid phosphatase activity was demonstrated by the method of Rosales et al (1966), and beta glucuronidase by the method of Lorbacher et ul (1967). The combined method of Yam et ul (1971), for the demonstration of chloroacetate and nonspecific esterases, was used initially for confirmatory identification of neutrophils and monocytes respectively. Basophils were specifically recognized by their characteristic metachromasia with toluidene blue (Yam et al, 1971).
RESULTS Composition of mononuclear cell concentrates. Monocytes constituted 25-30% of this material. Since the total mononuclear cell recovery was consistently of the order of 70-80%, the monocyte harvest was always in excess of 100%. E and EAC’ rosetting populations. The proportion of mononuclears forming E-rosettes was highly reproducible, with a range of 65-70%. The mean fraction forming EAC’ rosettes was 31%.
6l-
2
4
6
0 10 12 14 16 18 20 22 24 26 28 30 FRACTION NUMBER
FIG I . Cell size distribution profile, obtained by velocity sedimentation, of mononuclear cells previously depleted of T lymphocytes. Increasing fraction number reflects increasing cell size. Smaller peak consists entirely of debris.
E non-rosetting cells consisted of two distinct populations, when separated, mainly on the basis of differences in size, by velocity sedimentation (Fig I). The small cell fraction (I) was composed of lymphocytes of greater than 99% purity. The second population (11) of higher modal sedimentation velocity had a mean differential cell count of 80.5% monocytes, 15% lymphocytes and 4.5% basophils. EAC’ rosetting cells had a mean differential composition of 55.5% monocytes, 40.5% lymphocytes and 4% basophils, when prepared from whole mononuclear concentrates. When the starting material had first been depleted of T cells, the mean fractional cell countsof the EAC’ rosetting populations were 81.5% monocytes, 15.5% lymphocytes and 3% basophils.
Lysosomal Enzymes in Lymphocytes
2
4
6
8
567
10 12 14 16 18 20 22 24 26 28 30 Fraction Number
FIG 2. Cell size distribution profile of EAC‘ rosetting mononuclear cells.
Application of the EAC’ rosetting cells (after de-rosetting) to a sucrose gradient produced two readily separable populations of different modal sedimentation velocity (Fig 2). The small cell fraction (111) was composed predominantly of lymphocytes with more than 97% purity and the large cell fraction (IV) was a concentrate of greater than 85% monocytes. EAC’ non-rosetting cells had a mean differential composition of 93.5% lymphocytes, 5.5% basophils and 1% monocytes, when prepared from whole mononuclear concentrates. When the starting material had first been depleted of T cells, the mean fractional cell counts of the EAC’ non-rosetting populations were 73.5% lymphocytes, 25% monocytes and 1.5% basophils. Separation of EAC’ non-rosetting cells by velocity sedimentation (Fig 3) resulted in a lymphocyte concentrate of greater than 95% purity. In primary analyses, fraction 111 was used as pure B cells, and the EAC’ non-rosetting lymphocytes, harvested from T cell depleted mononuclear cell concentrates, were used as null cells (Dickler et a!, 1974). Secondary, less pure, populations of B and T cells were taken
FRACTION NUMBER
FIG 3. Cell size distribution profile of mononuclear cells previously depleted of EAC’ rosetting cells.
R. D. Barr and S . Perry
568
respectively from fraction I and EAC’ non-rosetting lymphocytes harvested from whole mononuclear cell concentrates. W e have previously shown that the haemopoietic stem cells are contained within the null cell population (Barr et a!, I975a, b) and fraction IV, although morphologicallyheterogeneous, does contain stem cells and was used as a secondary null cell concentrate. Acidphosphatase. A simple scoring system was devised for assessment of the degree of positivity of individual cells. Negative: no granules; weak: 1-5 granules; moderate: 6-10 granules; strong : greater than 10granules. In each lymphocyte subpopulation examined, the strength of the staining reaction was remarkably uniform among component cells, in contradistinction to the findings with beta-glucuronidase. With the exception of the findings in one 15 13 12
I
2
4
6
8
10 12 14 16 18 20 22 24 26 28 30
Fraction Number
FIG 4. Cell size distribution profile of pure T lymphocytes.
subject, the results revealed a consistent and distinguishing pattern. Activity in T cells was negative or weak, and in both B and null cells was moderate or strong. In the exceptional subject, for no apparent reason, T and B cells showed a weak reaction, while null cells were moderately positive. Contaminating neutrophils were consistently moderately positive and monocytes strongly positive. No activity was demonstrable in peripheral blood films, and we surmise that this may be due to the buffering action of plasma. Care must be taken to determine that apparent positivity in individual cells, particularly when these are present in clusters, is not due to overlying platelets, which are strongly positive for acid phosphatase. Beta-glucuronidase. The staining pattern in leucocytes was either diffuse or granular. Monocytes and granulocytes revealed diffuse cytoplasmic staining, while the lymphocytes were uniformly granular or negative. The granular pattern was of two types, positive lymphocytes containing either a single large block of cytoplasmic stain (Type a ) , or multiple small granules (Type a). In all subjects more than 90% of the T cells were positive, and the ratio of Type CI
Lysosomal E n z y m e s in Lymphocytes
569
to Type P cells ranged from 0.8 to I.OS:I, and, in each instance, approximately half the positive cells were of each type. The fraction of B cells giving a positive staining reaction was much more variable, ranging from 10to 76%. However, a consistent feature in these cells was the uniform Type /I granular staining. One subject had B cells in which a significant minority had a Type GI pattern. The staining reactions of the null cells closely paralleled those of the B cells in the individual subjects. A possible relationship between lymphocyte beta-glucuronidase activity and cell size was investigated by fractionating pure T cells by velocity sedimentation (Fig 4). Duplicate procedures were performed on material from two subjects. An examination was made of individual fractions constituting the cell-size distribution profile. No difference could be determined in either the total proportion of positive cells or the ratio of Type c1 to Type /I patterns. As with the acid phosphatase stain, demonstrable beta-glucuronidase activity was virtually abolished if the preparatory cold shock lysis time for removal of rosettes was greater than 30 s. Fixed specimens for both stains were successfully stored at 4°C for up to 72 h without apparent loss of activity. However, in the beta-glucuronidase technique, once freezing was performed, staining had to be effected immediately thereafter. Once stained, the acid phosphatase preparations should be examined promptly, for they are photosensitive and the positive staining reaction disappears within a few hours. DISCUSSION Both chemical and cytochemical estimations of the activity of a wide variety of lysosomal enzymes (Weissmann, 1964) have, in the past, been made on various lymphatic tissues in man and other mammals. These studies have, in general, been characterized by their lack of agreement, which in large part may reflect the impurity of the preparations on which the analyses were uniformly made. Thus, studies on human spleen (Li et al, 1972; Muller-Hermelink et a/, 1974) have failed to demonstrate either acid phosphatase or beta-glucuronidase activity in lymphoid cells. An early report cited the absence of acid phosphatase activity in lymphocytes in lymph node sections (Braunstein et al, 1958). Later investigators (Tamaoki & Essner, 1969) claimed to be able to detect acid phosphatase and beta-glucuronidase in thymus-dependent areas. Most recently, it has been reported (Mueller et al, 1975) that there is no differential activity of these enzymes in thymus-dependent and independent areas of the lymph nodes in mice. In peripheral blood lymphocytes, unseparated on the basis of immunological subtypes, lysosomal enzyme activity is the same in males and females (Kidd & Shaw, 197j), but is significantly lower in neonates than in adults (Astaldi et al, 1973). The degree of activity of acid phosphatase and beta-glucuronidase is generally agreed to be similar in such preparations (Flandrin & Daniel, 1974), though it has been suggested (Lorbacher et a!, 1967; Flandrin & Daniel, 1974) that the latter enzyme is present in greater quantity in large lymphocytes. It has wen been proposed (Chichoki et al, 1972) that normal lymphocytes can be categorized into three distinct groups; one negative for both enzymes, a second showing granular positivity for both and a third in which the presence of both enzymes is revealed by a diffuse reaction. Our present findings fail to corroboratc these earlier observations. Evidence from our experiments does not support a correlation between lymphocyte beta-glucuronidase
5 70
R.D. Burr and S. Perry
activity and cell size. The current data indicate that, on the basis of acid phosphatase and betaglucuronidase activity, lymphocytes may be classified as thymus-dependent (T) or thymusindependent (non-T) cells. Cells which give a negative or weak reaction for acid phosphatase and a strong Type a reaction for beta-glucuronidase may, with confidence, be labelled as T cells. Unfortunately, these techniques fail to resolve non-T cell subpopulations and, in particular, no pattern, selective for stem cells, is apparent. A cautionary note should be expressed on the extrapolation of such results obtained in normal T and B cells from peripheral blood, as has been widely practiced with immunological markers. Activated lymphocytes show an increase in lysosomal enzyme activity (Astaldi et al, 1973) as do cells stimulated by phytomitogens (Allison & Mallucci, 1964; Barker & Farnes, 1967; Konig et al, 1973). Furthermore, plasma cells, of B-cell origin, show strong beta-glucuronidase activity (Szniigielski, 1966). Likewise, in pathological conditions, there is no uniform pattern. Hence, in infectious mononucleosis, acid phosphatase activity is strong in the atypical lymphocytes (Li et a!, 1g70), which are derived from T cells (Papamichail et al, 1974; Pattengale et al, 1974). Activities of acid phosphatase and beta-glucuronidase are increased in acute lymphoblastic leukaemia (Sippel1et al, 1975) and reduced in chronic lymphocytic leukaemia (Yam & Mitus, 1968; Li et a/, 1970), which has been characterized as a B-cell neoplasm (Wilson & Nossal, 1971; Perera & Pegrum, 1974). Again, the activity of both enzymes is markedly elevated in Sezary cells (Flandrin & Daniel, 1974), which are considered to be neoplastic T cells (Flandrin & Brouet, 1974).
Recently, additional attempts to provide cytochemical discrimination of T from B cells have been described. The cell membrane associated enzymes, 5-nucleotidase and adenosine triphosphatase have been shown to be selectively present in thymus-independent areas of human lymph nodes (Miiller-Hermelink, 1974). Complementary results have been obtained with the localization of nonspecific acid esterase activity to the thymus-dependent areas in murine lymph nodes. Preliminary evidence, from electron microscopy, suggests that this enzyme is also located on Iysosomes (Mueller et al, 1975). As yet these techniques have not been applied to suspensions of purified cell preparations. Until now the enzyme has formed the basis of what was widely considered to be the most specific cytochemical feature of monocytes (Schmalzl& Braunsteiner, 1970). The results of our own investigations, which are as yet incomplete, demonstrate that virtually all T cells show a granular positivity for the presence of nonspecific esterase, while non-T cells uniformly give a diffuse cytoplasmic stain. The outcome of that study will form the basis of a subsequent report dealing with lymphocytemonocyte interconversion. ACKNOWLEDGMENT
We wish to thank Ms Ada Brooks for technical assistance. REFERENCES ALLISON, A.C. & MALLUCCI, L. (1964) Lysosomes in dividing cells, with special reference to lymphocytes. Lancet, ii, 1371. ASTALDI, G., LISIEWICZ, J., CHICHOKI, T. &BuBLINA,A.
(1973) The lysosomal enzymes in lymphocytes. 11. Activated lymphocytes. Acta Vituminologica et Enzymologicu, 27, 197.
BARKER, B.E. & FARNES,P. (1967)Histochemistry of
Lysosomal Enzymes in Lymphocytes blood cells treated with pokeweed mitogen. Nature, 214, 787. BARR,R.D., WHANC-PENG, J. & PERRY,S. (197~a) Recognition of the human hemopoietic stem cell. Proceedings of the American Association for Cancer Research, 16, 23. BARR,R.D., WHANG-PENG, J. & PERRY,S. (197sb) Hemopoietic stem cells in human peripheral blood. Science, 190, 284. BOYUM, A. (1968) Separation ofleucocytes from blood and bone marrow. Scandinavian Journal of Clinical and Laboratory Investigation, 21,Supplement 97. BRAUNSTEIN, H., FREIMAN, D.G. & GALL,E.A. (1958) A histochemical study of the enzymatic activity of lymph nodes. I. The normal and hyperplastic lymph node. Cancer, 11, 829. W.H. (1969) Separation of BRUBAKER, L.H. 81 EVANS, granulocytes, nionocytes, lymphocytes, erythrocytes, and platelets from human blood and relative tagging with diisopropylfluorophosphate (DFP). Journal ofLaboratory and Clinical Medicine, 73, 1036. J. (1972) The CHICHOKI, T., ASTALDI, G. & LISIEWICZ, lysosomal enzymes in lymphocytes. I. Normal lymphoid tissue. Acta Vitaminologica et Enzymologica, 263. DICKLER, H.B., ADKINSON,N.F. & TERRY,W.D. (1974) Evidence for individual human peripheral blood lymphocytes bearing both B and T cell markers. Nature, 247, 213. FALLON, H.J., FREI,E., 111, DAVIDSON, J.D., TRIER, J.S. & BURK,D. (1962) Leucocyte preparations from human blood: evaluation of their morphologic and metabolic state. Journal of Laboratory and Clinical Medicine, 59, 779. J.-C. (1974) The Sezary cell: FLANDRIN, G. & BROUET, cytologic, cytochemical, and immunologic studies. Mayo Clinic Proceedings, 49, 575. FLANDRIN, G. 81 DANIEL, M.T. (1974) 8-Glucuronidase activity in Sezary cells. Scandinavian Journal of Haematology, 12, 23. KIDD,V.J. & SHAW,M.T. (1975) Cytochemical differences in leucocytes in normal adult males and females. Scandinavian Journal of Haematology, 14, 120.
KONIC, E., BRITTINGER, G. & COHNEN,G. (1973) Relation of 1ysosomal fragility in CLL lymphocytes to PHA reactivity. Nature: N e w Biology, 244, 247. LAJTHA, L.G. (1973) In: Hemopoiesis in Culture. Second International Workshop. DHEW Publication No. (NIH) 74-20s. p 401. V. (1968) Receptors for LAY,W.H. 81 NUSSENZWEIG, complement on leukocytes. Journal of Experimental Medicine, 128, 991. LI, C.Y., YAM,L.T. & CROSBY, W.H. (1972) Histochemical characterization of cellular and structural elements of the human spleen. Journal of Histochemistry and Cytochemistry, 20, 1049.
571
LI, C.Y., YAM, L.T. & LAM, K.W. (1970) Acid phosphatase isoenzyme in human leukocytes in normal and pathologic conditions. Journal of Histochemistry and Cytochemistry, 18, 473. LOHRMANN, H.-P., Novmovs, L. & GRAW,R.G., JR (1974) Cellular interactions in the proliferative response of human T and B lymphocytes to phytomitogens and allogeneic lymphocytes. Journal of Experimental Medicine, 139, 1553. LORBACHER, P., YAM, L.T. & MITUS, W.J. (1967) Cytochemical demonstration of &glucuronidase activity in blood and bone marrow cells. Journal of Histochemistry and Cytochemistry, 15,680. H., KELLER, MUELLER, J., BRUNDEL RE, G., BUERKI, H.-U., HESS,M.W. 81 COTTIER,H. (197s) Nonspecific acid esterase activity: a criterion for differentiation of T and B lymphocytes in mouse lymph nodes. European Journal of Immunology, 5,270. MULLER-HERMELINK, H.K. (1974) Characterization of the B-cell and T-cell regions of human lymphatic tissue through enzyme histochemical demonstration of ATPase and 5'-nucleotidase activities. Virchowr Archives, B, Cell Pathology, 16, 371. U. & MULLER-HERMELINK, H.K., HEUSERMANN, STUTTE, H.J. (1974) Enzyme histochemical observations on the localization and structure of the T cell and B cell regions in the human spleen. Cell and Tissue Research, 154, 167. E.J. PAPAMICHAIL, P., SHELDON, P.J. 81 HOLBOROW, (1974) T- and B-cell subpopulations in infectious mononucleosis. Clinical and Experiniental Immunology, 18, I. PATTENGALE, P.K., SMITH,R.W. & PERLIN,E. (1974) Atypical lymphocytes in acute infections mononucleosis. Identification by multiple T and B lymphocyte markers. N e w England Journal uf Medicine, 291, 1045. PERERA, D.J.B. & PEGRUM, G.D. (1974) The lymphocyte in chronic lymphatic Ieukaemia. Lancet, i, 1207. ROSALES, C.L., BENNETT, J.M. 81 RUTENBURG, A.M. (1966) Histochemical demonstration of leucocyte acid phosphatase in health and in disease. British Journal of Haematology, 12, 172. SCHMALZL, F. & BRAUNSTEINER, H. (1970) The cytochemistry of monocytes and macrophages. Series Haematologica, 3, No. 2, 93. SHEVACH, E., EDELSON, R., FRANK, M., LUTZNER, M. & GREEN, I. (1974) A human leukemia cell with both B and T cell surface receptors. Proceedings of the National Academy of Sciences of the United States o j America, 71, 863. SIPPELL,W.G., ANTONOWICZ, I., LAZARUS, H. & SCHWACHMAN, H. (1975) Lysosomal and mitochondrial enzyme activities in human lymphoid cell lines obtained from children with acute lymphoblastic leukemia and controls. Experimental Cell Research, 91,i s z .
572
R. D. Burr and S. Perry
SZMIGIELSKI, S. (1966) An improved method for histochemical demonstration of beta-glucuronidase and beta-galactosidase activity in bone marrow and blood cells. Journal of Laboratory and Clinical Medicine, 67, 709. TAMAOKI, M. & ESSNER, E. (1969) Distribution of acid phosphatase, pglucuronidase and N-acetyl-& glucosaminidase activities in lymphocytes of lymphatic tissue of man and rodents. Journal of Histochemistry and Cytochemisfry, 17, 238. WEISSMA",G. (1964) Lysosomes. Blood, 24, 594.
WILSON, J.D. & NOSSAL, G.J.V. (1971)Identification of human T and B lymphocytes in normal peripheral blood and in chronic lymphocytic leukaemia. Lancet, ii, 788. YAM,L.T., LI, C.Y.& CROSBY, W.H. (1971) Cytochemical identification of monocytes and granulocytes. American Journal ofClinica1 Pathology, 55, 283. YAM,L.T. & MITUS,W.J. (1968) The lymphocyte fi-glucuronidase activity in lymphoproliferative disorders. Blood, 31, 480.
of
Haematology, 1976,32, 565.
Lysosomal Acid Hydrolases in Human Lymphocyte
Subpopula tions R. D. BARRAND S. PERRY* Ofice ofthe Director, Division of Cancer Treatment, National Cancer Institute, and * 0 8 c e o f the Director, National Institutes of Health, Bethesda, Maryland (Received 5 August 1975 ; acceptedfor publication 9 October 1975)
SUMMARY.Acid phosphatase and beta-glucuronidase activities have been determined cytochemically in T, B and null lymphocytes as part of an effort to characterize the human haemopoietic stem cell. A combination of weak acid phosphatase activity and strong beta glucuronidase staining, in the form of a single large granule, has been shown to be specific for T cells. On the basis of this approach alone, non-T cells could not be further subclassified. Further cytochemical evaluation is being explored. Following our recent identification of the human haemopoietic stem cell in normal peripheral blood (Barr et al, 197~a,b) we have embarked on its further characterization. In this endeavour we have responded to Lajtha’s plea (1973) for a cytochemical assessment of the pluripotent cell. Recent evidence, from work on human and other mammalian material, has suggested that discrete populations of lymphocytes may be distinguished by their differential content of enzymes located in the lysosomes. It appeared appropriate, therefore, to pursue this possibility, by a cytochemical evaluation of acid phosphatase (E.C.3. I .3.2) and beta-glucuronidase (E.C.3.2.1.3I) in purified lymphocyte fractions (T, B and null cells) harvested from normal human peripheral blood. MATERIALS AND METHODS Peripheral venous blood was obtained from 10 normal adult volunteers and heparinized at 40 unitsiml. Mononuclear cell concentrates were prepared by isopycnic sedimentation on Ficoll-Hypaque gradients (Boyum, 1968). Pure thymus-dependent lymphocytes (T cells) were obtained by the formation of rosettes with neuraminidase-treated sheep erythrocytes (Lohrmann et al, 1974);the rosettes were subsequently harvested by density sedimentation and the red cells removed by cold shock lysis (Fallon et al, 1962). Non-E rosetting cells were further separated by velocity sedimentation in a sucrose gradient at unit gravity (Brubaker & Evans, 1969). Concentrates of bursa-equivalent lymphocytes (B cells) were prepared by rosette formation with EAC’ complexes (Shevach et al, 1974). For this purpose sheep erythrocytes (E), labelled with rabbit anti-sheep erythrocytes IgG (A), were obtained from Cordis Laboratories, Miami, Florida 33137, U.S.A., and incubated with fresh serum (C’), from leukaemic Correspondence: Dr Seyniour Perry, Special Assistant to the Director, National Institutes of Health, Bldg. Room 111, Bethesda, Maryland 20014, U.S.A.
565
I,
R. D. Burr and S . Perry
566
AKR mice as a source of complement (Lay & Nussenzweig, 1968). The mice were supplied by Microbiological Associatx, Bethesda, Maryland 20014, U.S.A. Acid phosphatase activity was demonstrated by the method of Rosales et al (1966), and beta glucuronidase by the method of Lorbacher et ul (1967). The combined method of Yam et ul (1971), for the demonstration of chloroacetate and nonspecific esterases, was used initially for confirmatory identification of neutrophils and monocytes respectively. Basophils were specifically recognized by their characteristic metachromasia with toluidene blue (Yam et al, 1971).
RESULTS Composition of mononuclear cell concentrates. Monocytes constituted 25-30% of this material. Since the total mononuclear cell recovery was consistently of the order of 70-80%, the monocyte harvest was always in excess of 100%. E and EAC’ rosetting populations. The proportion of mononuclears forming E-rosettes was highly reproducible, with a range of 65-70%. The mean fraction forming EAC’ rosettes was 31%.
6l-
2
4
6
0 10 12 14 16 18 20 22 24 26 28 30 FRACTION NUMBER
FIG I . Cell size distribution profile, obtained by velocity sedimentation, of mononuclear cells previously depleted of T lymphocytes. Increasing fraction number reflects increasing cell size. Smaller peak consists entirely of debris.
E non-rosetting cells consisted of two distinct populations, when separated, mainly on the basis of differences in size, by velocity sedimentation (Fig I). The small cell fraction (I) was composed of lymphocytes of greater than 99% purity. The second population (11) of higher modal sedimentation velocity had a mean differential cell count of 80.5% monocytes, 15% lymphocytes and 4.5% basophils. EAC’ rosetting cells had a mean differential composition of 55.5% monocytes, 40.5% lymphocytes and 4% basophils, when prepared from whole mononuclear concentrates. When the starting material had first been depleted of T cells, the mean fractional cell countsof the EAC’ rosetting populations were 81.5% monocytes, 15.5% lymphocytes and 3% basophils.
Lysosomal Enzymes in Lymphocytes
2
4
6
8
567
10 12 14 16 18 20 22 24 26 28 30 Fraction Number
FIG 2. Cell size distribution profile of EAC‘ rosetting mononuclear cells.
Application of the EAC’ rosetting cells (after de-rosetting) to a sucrose gradient produced two readily separable populations of different modal sedimentation velocity (Fig 2). The small cell fraction (111) was composed predominantly of lymphocytes with more than 97% purity and the large cell fraction (IV) was a concentrate of greater than 85% monocytes. EAC’ non-rosetting cells had a mean differential composition of 93.5% lymphocytes, 5.5% basophils and 1% monocytes, when prepared from whole mononuclear concentrates. When the starting material had first been depleted of T cells, the mean fractional cell counts of the EAC’ non-rosetting populations were 73.5% lymphocytes, 25% monocytes and 1.5% basophils. Separation of EAC’ non-rosetting cells by velocity sedimentation (Fig 3) resulted in a lymphocyte concentrate of greater than 95% purity. In primary analyses, fraction 111 was used as pure B cells, and the EAC’ non-rosetting lymphocytes, harvested from T cell depleted mononuclear cell concentrates, were used as null cells (Dickler et a!, 1974). Secondary, less pure, populations of B and T cells were taken
FRACTION NUMBER
FIG 3. Cell size distribution profile of mononuclear cells previously depleted of EAC’ rosetting cells.
R. D. Barr and S . Perry
568
respectively from fraction I and EAC’ non-rosetting lymphocytes harvested from whole mononuclear cell concentrates. W e have previously shown that the haemopoietic stem cells are contained within the null cell population (Barr et a!, I975a, b) and fraction IV, although morphologicallyheterogeneous, does contain stem cells and was used as a secondary null cell concentrate. Acidphosphatase. A simple scoring system was devised for assessment of the degree of positivity of individual cells. Negative: no granules; weak: 1-5 granules; moderate: 6-10 granules; strong : greater than 10granules. In each lymphocyte subpopulation examined, the strength of the staining reaction was remarkably uniform among component cells, in contradistinction to the findings with beta-glucuronidase. With the exception of the findings in one 15 13 12
I
2
4
6
8
10 12 14 16 18 20 22 24 26 28 30
Fraction Number
FIG 4. Cell size distribution profile of pure T lymphocytes.
subject, the results revealed a consistent and distinguishing pattern. Activity in T cells was negative or weak, and in both B and null cells was moderate or strong. In the exceptional subject, for no apparent reason, T and B cells showed a weak reaction, while null cells were moderately positive. Contaminating neutrophils were consistently moderately positive and monocytes strongly positive. No activity was demonstrable in peripheral blood films, and we surmise that this may be due to the buffering action of plasma. Care must be taken to determine that apparent positivity in individual cells, particularly when these are present in clusters, is not due to overlying platelets, which are strongly positive for acid phosphatase. Beta-glucuronidase. The staining pattern in leucocytes was either diffuse or granular. Monocytes and granulocytes revealed diffuse cytoplasmic staining, while the lymphocytes were uniformly granular or negative. The granular pattern was of two types, positive lymphocytes containing either a single large block of cytoplasmic stain (Type a ) , or multiple small granules (Type a). In all subjects more than 90% of the T cells were positive, and the ratio of Type CI
Lysosomal E n z y m e s in Lymphocytes
569
to Type P cells ranged from 0.8 to I.OS:I, and, in each instance, approximately half the positive cells were of each type. The fraction of B cells giving a positive staining reaction was much more variable, ranging from 10to 76%. However, a consistent feature in these cells was the uniform Type /I granular staining. One subject had B cells in which a significant minority had a Type GI pattern. The staining reactions of the null cells closely paralleled those of the B cells in the individual subjects. A possible relationship between lymphocyte beta-glucuronidase activity and cell size was investigated by fractionating pure T cells by velocity sedimentation (Fig 4). Duplicate procedures were performed on material from two subjects. An examination was made of individual fractions constituting the cell-size distribution profile. No difference could be determined in either the total proportion of positive cells or the ratio of Type c1 to Type /I patterns. As with the acid phosphatase stain, demonstrable beta-glucuronidase activity was virtually abolished if the preparatory cold shock lysis time for removal of rosettes was greater than 30 s. Fixed specimens for both stains were successfully stored at 4°C for up to 72 h without apparent loss of activity. However, in the beta-glucuronidase technique, once freezing was performed, staining had to be effected immediately thereafter. Once stained, the acid phosphatase preparations should be examined promptly, for they are photosensitive and the positive staining reaction disappears within a few hours. DISCUSSION Both chemical and cytochemical estimations of the activity of a wide variety of lysosomal enzymes (Weissmann, 1964) have, in the past, been made on various lymphatic tissues in man and other mammals. These studies have, in general, been characterized by their lack of agreement, which in large part may reflect the impurity of the preparations on which the analyses were uniformly made. Thus, studies on human spleen (Li et al, 1972; Muller-Hermelink et a/, 1974) have failed to demonstrate either acid phosphatase or beta-glucuronidase activity in lymphoid cells. An early report cited the absence of acid phosphatase activity in lymphocytes in lymph node sections (Braunstein et al, 1958). Later investigators (Tamaoki & Essner, 1969) claimed to be able to detect acid phosphatase and beta-glucuronidase in thymus-dependent areas. Most recently, it has been reported (Mueller et al, 1975) that there is no differential activity of these enzymes in thymus-dependent and independent areas of the lymph nodes in mice. In peripheral blood lymphocytes, unseparated on the basis of immunological subtypes, lysosomal enzyme activity is the same in males and females (Kidd & Shaw, 197j), but is significantly lower in neonates than in adults (Astaldi et al, 1973). The degree of activity of acid phosphatase and beta-glucuronidase is generally agreed to be similar in such preparations (Flandrin & Daniel, 1974), though it has been suggested (Lorbacher et a!, 1967; Flandrin & Daniel, 1974) that the latter enzyme is present in greater quantity in large lymphocytes. It has wen been proposed (Chichoki et al, 1972) that normal lymphocytes can be categorized into three distinct groups; one negative for both enzymes, a second showing granular positivity for both and a third in which the presence of both enzymes is revealed by a diffuse reaction. Our present findings fail to corroboratc these earlier observations. Evidence from our experiments does not support a correlation between lymphocyte beta-glucuronidase
5 70
R.D. Burr and S. Perry
activity and cell size. The current data indicate that, on the basis of acid phosphatase and betaglucuronidase activity, lymphocytes may be classified as thymus-dependent (T) or thymusindependent (non-T) cells. Cells which give a negative or weak reaction for acid phosphatase and a strong Type a reaction for beta-glucuronidase may, with confidence, be labelled as T cells. Unfortunately, these techniques fail to resolve non-T cell subpopulations and, in particular, no pattern, selective for stem cells, is apparent. A cautionary note should be expressed on the extrapolation of such results obtained in normal T and B cells from peripheral blood, as has been widely practiced with immunological markers. Activated lymphocytes show an increase in lysosomal enzyme activity (Astaldi et al, 1973) as do cells stimulated by phytomitogens (Allison & Mallucci, 1964; Barker & Farnes, 1967; Konig et al, 1973). Furthermore, plasma cells, of B-cell origin, show strong beta-glucuronidase activity (Szniigielski, 1966). Likewise, in pathological conditions, there is no uniform pattern. Hence, in infectious mononucleosis, acid phosphatase activity is strong in the atypical lymphocytes (Li et a!, 1g70), which are derived from T cells (Papamichail et al, 1974; Pattengale et al, 1974). Activities of acid phosphatase and beta-glucuronidase are increased in acute lymphoblastic leukaemia (Sippel1et al, 1975) and reduced in chronic lymphocytic leukaemia (Yam & Mitus, 1968; Li et a/, 1970), which has been characterized as a B-cell neoplasm (Wilson & Nossal, 1971; Perera & Pegrum, 1974). Again, the activity of both enzymes is markedly elevated in Sezary cells (Flandrin & Daniel, 1974), which are considered to be neoplastic T cells (Flandrin & Brouet, 1974).
Recently, additional attempts to provide cytochemical discrimination of T from B cells have been described. The cell membrane associated enzymes, 5-nucleotidase and adenosine triphosphatase have been shown to be selectively present in thymus-independent areas of human lymph nodes (Miiller-Hermelink, 1974). Complementary results have been obtained with the localization of nonspecific acid esterase activity to the thymus-dependent areas in murine lymph nodes. Preliminary evidence, from electron microscopy, suggests that this enzyme is also located on Iysosomes (Mueller et al, 1975). As yet these techniques have not been applied to suspensions of purified cell preparations. Until now the enzyme has formed the basis of what was widely considered to be the most specific cytochemical feature of monocytes (Schmalzl& Braunsteiner, 1970). The results of our own investigations, which are as yet incomplete, demonstrate that virtually all T cells show a granular positivity for the presence of nonspecific esterase, while non-T cells uniformly give a diffuse cytoplasmic stain. The outcome of that study will form the basis of a subsequent report dealing with lymphocytemonocyte interconversion. ACKNOWLEDGMENT
We wish to thank Ms Ada Brooks for technical assistance. REFERENCES ALLISON, A.C. & MALLUCCI, L. (1964) Lysosomes in dividing cells, with special reference to lymphocytes. Lancet, ii, 1371. ASTALDI, G., LISIEWICZ, J., CHICHOKI, T. &BuBLINA,A.
(1973) The lysosomal enzymes in lymphocytes. 11. Activated lymphocytes. Acta Vituminologica et Enzymologicu, 27, 197.
BARKER, B.E. & FARNES,P. (1967)Histochemistry of
Lysosomal Enzymes in Lymphocytes blood cells treated with pokeweed mitogen. Nature, 214, 787. BARR,R.D., WHANC-PENG, J. & PERRY,S. (197~a) Recognition of the human hemopoietic stem cell. Proceedings of the American Association for Cancer Research, 16, 23. BARR,R.D., WHANG-PENG, J. & PERRY,S. (197sb) Hemopoietic stem cells in human peripheral blood. Science, 190, 284. BOYUM, A. (1968) Separation ofleucocytes from blood and bone marrow. Scandinavian Journal of Clinical and Laboratory Investigation, 21,Supplement 97. BRAUNSTEIN, H., FREIMAN, D.G. & GALL,E.A. (1958) A histochemical study of the enzymatic activity of lymph nodes. I. The normal and hyperplastic lymph node. Cancer, 11, 829. W.H. (1969) Separation of BRUBAKER, L.H. 81 EVANS, granulocytes, nionocytes, lymphocytes, erythrocytes, and platelets from human blood and relative tagging with diisopropylfluorophosphate (DFP). Journal ofLaboratory and Clinical Medicine, 73, 1036. J. (1972) The CHICHOKI, T., ASTALDI, G. & LISIEWICZ, lysosomal enzymes in lymphocytes. I. Normal lymphoid tissue. Acta Vitaminologica et Enzymologica, 263. DICKLER, H.B., ADKINSON,N.F. & TERRY,W.D. (1974) Evidence for individual human peripheral blood lymphocytes bearing both B and T cell markers. Nature, 247, 213. FALLON, H.J., FREI,E., 111, DAVIDSON, J.D., TRIER, J.S. & BURK,D. (1962) Leucocyte preparations from human blood: evaluation of their morphologic and metabolic state. Journal of Laboratory and Clinical Medicine, 59, 779. J.-C. (1974) The Sezary cell: FLANDRIN, G. & BROUET, cytologic, cytochemical, and immunologic studies. Mayo Clinic Proceedings, 49, 575. FLANDRIN, G. 81 DANIEL, M.T. (1974) 8-Glucuronidase activity in Sezary cells. Scandinavian Journal of Haematology, 12, 23. KIDD,V.J. & SHAW,M.T. (1975) Cytochemical differences in leucocytes in normal adult males and females. Scandinavian Journal of Haematology, 14, 120.
KONIC, E., BRITTINGER, G. & COHNEN,G. (1973) Relation of 1ysosomal fragility in CLL lymphocytes to PHA reactivity. Nature: N e w Biology, 244, 247. LAJTHA, L.G. (1973) In: Hemopoiesis in Culture. Second International Workshop. DHEW Publication No. (NIH) 74-20s. p 401. V. (1968) Receptors for LAY,W.H. 81 NUSSENZWEIG, complement on leukocytes. Journal of Experimental Medicine, 128, 991. LI, C.Y., YAM,L.T. & CROSBY, W.H. (1972) Histochemical characterization of cellular and structural elements of the human spleen. Journal of Histochemistry and Cytochemistry, 20, 1049.
571
LI, C.Y., YAM, L.T. & LAM, K.W. (1970) Acid phosphatase isoenzyme in human leukocytes in normal and pathologic conditions. Journal of Histochemistry and Cytochemistry, 18, 473. LOHRMANN, H.-P., Novmovs, L. & GRAW,R.G., JR (1974) Cellular interactions in the proliferative response of human T and B lymphocytes to phytomitogens and allogeneic lymphocytes. Journal of Experimental Medicine, 139, 1553. LORBACHER, P., YAM, L.T. & MITUS, W.J. (1967) Cytochemical demonstration of &glucuronidase activity in blood and bone marrow cells. Journal of Histochemistry and Cytochemistry, 15,680. H., KELLER, MUELLER, J., BRUNDEL RE, G., BUERKI, H.-U., HESS,M.W. 81 COTTIER,H. (197s) Nonspecific acid esterase activity: a criterion for differentiation of T and B lymphocytes in mouse lymph nodes. European Journal of Immunology, 5,270. MULLER-HERMELINK, H.K. (1974) Characterization of the B-cell and T-cell regions of human lymphatic tissue through enzyme histochemical demonstration of ATPase and 5'-nucleotidase activities. Virchowr Archives, B, Cell Pathology, 16, 371. U. & MULLER-HERMELINK, H.K., HEUSERMANN, STUTTE, H.J. (1974) Enzyme histochemical observations on the localization and structure of the T cell and B cell regions in the human spleen. Cell and Tissue Research, 154, 167. E.J. PAPAMICHAIL, P., SHELDON, P.J. 81 HOLBOROW, (1974) T- and B-cell subpopulations in infectious mononucleosis. Clinical and Experiniental Immunology, 18, I. PATTENGALE, P.K., SMITH,R.W. & PERLIN,E. (1974) Atypical lymphocytes in acute infections mononucleosis. Identification by multiple T and B lymphocyte markers. N e w England Journal uf Medicine, 291, 1045. PERERA, D.J.B. & PEGRUM, G.D. (1974) The lymphocyte in chronic lymphatic Ieukaemia. Lancet, i, 1207. ROSALES, C.L., BENNETT, J.M. 81 RUTENBURG, A.M. (1966) Histochemical demonstration of leucocyte acid phosphatase in health and in disease. British Journal of Haematology, 12, 172. SCHMALZL, F. & BRAUNSTEINER, H. (1970) The cytochemistry of monocytes and macrophages. Series Haematologica, 3, No. 2, 93. SHEVACH, E., EDELSON, R., FRANK, M., LUTZNER, M. & GREEN, I. (1974) A human leukemia cell with both B and T cell surface receptors. Proceedings of the National Academy of Sciences of the United States o j America, 71, 863. SIPPELL,W.G., ANTONOWICZ, I., LAZARUS, H. & SCHWACHMAN, H. (1975) Lysosomal and mitochondrial enzyme activities in human lymphoid cell lines obtained from children with acute lymphoblastic leukemia and controls. Experimental Cell Research, 91,i s z .
572
R. D. Burr and S. Perry
SZMIGIELSKI, S. (1966) An improved method for histochemical demonstration of beta-glucuronidase and beta-galactosidase activity in bone marrow and blood cells. Journal of Laboratory and Clinical Medicine, 67, 709. TAMAOKI, M. & ESSNER, E. (1969) Distribution of acid phosphatase, pglucuronidase and N-acetyl-& glucosaminidase activities in lymphocytes of lymphatic tissue of man and rodents. Journal of Histochemistry and Cytochemisfry, 17, 238. WEISSMA",G. (1964) Lysosomes. Blood, 24, 594.
WILSON, J.D. & NOSSAL, G.J.V. (1971)Identification of human T and B lymphocytes in normal peripheral blood and in chronic lymphocytic leukaemia. Lancet, ii, 788. YAM,L.T., LI, C.Y.& CROSBY, W.H. (1971) Cytochemical identification of monocytes and granulocytes. American Journal ofClinica1 Pathology, 55, 283. YAM,L.T. & MITUS,W.J. (1968) The lymphocyte fi-glucuronidase activity in lymphoproliferative disorders. Blood, 31, 480.
