EXPERIMENTAL

CELL

RESEARCH

202,

405-411 (19%)

intracellular Localization of Terminal Transferas ROBERTO *Istituto

DI PRIMIO,

*J ORIANA TRUBIANI,~

di Morfoiogia Umana Normale and tlstituto di Citomorfologia Normale e Patologica de1 CNR, FacoltcZ di &i&&a, Unioersitci di Chieti, via dei Vestini 12, I-66100 Chieti, Italy; and SDepartment of Biochemistry, Uniformed Services University of Health Sciences, 4301 Jones Bridge Road, Bethesda, Maryland 20814

Changes in the localization of terminal transferase during the cell cycle in random cultures of human pre-T leukemia line RPMI-8402 were examined by light and electron microscopy on immunoperoxidase-stained preparations. Paraformaldehyde-fixed and saponinpermeabilized human cells were used with a monoclonal anti-human terminal deoxynucleotidyl transferase (TdT) primary reagent to demonstrate changes in enzyme distribution occurring between interphase and mitosis. Nuclear localization is found uniformly during interphase. At metaphase, however, the majority of TdT staining appears randomly distributed in the cytoplasm and traces of TdT staining remain associated with mitotic cbromatin. At later phases, when the daughter cells are forming, the enzyme again appears to be restricted to the new nuclear structure. o 1992 Academic

AND F

Press,

Inc.

INTRODUCTION

Terminal deoxynucleotidyl transferase (TdT) is a DNA polymerase that catalyzes condensation of deoxyribonucleoside triphosphates without using a template [l-4]. Molecular cloning ofcDNA [5], ontogenic studies, investigation of TdT function [6-81, production of polyand monoclonal antibodies ]9, IO], and immunocytochemical study at the light and the electron microscope ievels [ll-131 have provided demonstrations that the presence of the enzyme is restricted to pre-T and pre-B cells during immunopoiesis. This suggests that the role of this particular “creative” DNA polymerase is implicated in the development of the immune function. Different experimental evidence proposes the hypothesis that terminal transferase may act, in early lymphoid cells, as a somatic mutagen to generate diversity in the DNA sequence [?, 8] by the addition of nucleotides to the gene producing N-regions in the immunoglobulin heavy chain or variable regions in the T cell receptor [ 14-17]. The precise reasons for the presence of TdT in primary immunopoietic tissues and in certain acute leu-

I To whom reprimt requests should be addressed.

e have r’ecently demkemias remain to be identified. onstrated by light and electron roscopic immunocytochemistry that TdT resides, at tbe nuclear level during interphase, bound to the ~rotei~~~e~~~s scaffold of the nucleus defined as nuclear matrix [X2], and that, in rat thymocyte subpopulations, the enzyme undergoes intracellular movements related to the uration [13]. The aim of this study is t detailed description of TdT changes t the cell cycle.

Cell lines. The TdT-positive RPMI-8402 [18j and TdT-negative RA-1 (Ramos, ATCC CRL 1596) human ceil lines were used for this study. The cells were grown in RPMI 1640 medium containing 10% fetal bovine serum, 4 m&f L-glutamine, 100 mM Na-pyruvate; and 25 mA4 Hepes. Random cultures were harvested eariy in their growth phase in order to maximize the number of mitotic stages present. Immunochemical and chemical reagents. Two mouse monoclonal anti-human TdT antibodies (Nos. 15 and 71), reacting with different epitopes, were mixed and used as the primary reagent. The characterization of the monoclonal antibody has been described in an earlier publication [lo]. Clone 15 antibody (IgG,,) inhibits enzyme activity, whereas clone 71 (IgG,) does not. Peroxidase-labeied conjugate of goat anti-mouse IgG obtained from Litton Bionetics (Kensington, MD) was used as the secondary reagent. Controi observations were made on TdT-negative cell lines and by substituting PBS for the primary reagent. All other chemical reagents used in this study were reagent grade materials from Sigma (St. Louis, MO), and EM reagents were from Polyscience (Warrington, PA). Light microscopy. Preparations of TdT-positive or -negative cells or mixtures of positive and negative eels were fixed for 45 min at room temperature in a 1% solution of paraforma!dehyde in 0.1 M Na-cacodylate buffer at pH 7.6 immediately after cytocentrifugation. The slides were washed for 2 h with several changes of cacodylate buffer containing 2% sucrose and 0.1 M glycine. They were then blocked for 15 min with 2% normal goat Serum and then with 0.2% H,O, in cacodylate buffer for 5 min. The slides were then allowed to react for 30 min with 0.25 rg/mI of mouse monoclonal anti-human TdT and rinsed well with PBS. Specific reactivity was detected with 0.5 pglml of goat anti-mouse IgG peroxidase conjugate. To visualize the sites of antibody reaction, a substrate solution of 0.05% diaminobenzidine tetrahydrochloride (DAB) containing 0.1)15% H,O, in 0.05 M Tris-HCl bufIer, pH 7.6, was applied for 15 min at room temperature. The slides were counterstained with Mayer’s Xematoxyhn. A Zeiss Photomicroscope III was used for light microscopy andphotomicrography. Electron microscopy. Cells grown as described above were fixed in suspension for 3 h at room temperature with a soluti~on of2% parafor-

405 lkpyright 0 1992 .&I rights of reproduction

0014~4327/92 $5.00 by Academic Press, Inc. in any form reserved.

406

DI PRIMIO,

TRUBIANI,

maldehyde in 0.1 M Na-cacodylate buffer, pH 7.6, and then washed several times in the same buffer. The fixed cell suspension was incubated for 30 min in freshly prepared 0.4% saponin and 2% normal goat serum in cacodylate buffer to permeabilize the cells. The fixed and permeabilized cells were washed with 0.1 M Na-cacodylate buffer, pH 7.6, containing 2% sucrose and 0.1 M glycine for 3 h using several changes of the wash buffer. The cells were then incubated with 0.25 pg per ml of mouse monoclonal anti-human TdT in the Na-cacodylate buffer (without sucrose and glycine) for 32 h at 4°C. They were then rinsed well with PBS. The peroxidase conjugate of goat anti-mouse IgG was applied at 0.5 pg per ml for 4 h at room temperature. After the washing with PBS, the cells were incubated for 15 min with the DAB:H,O, substrate solution, rinsed in cacodylate buffer, and postfixed with 1% 0~0, in the same buffer for 1 h at 4’C. The cells were coated with agar and dehydrated with alcohol and toluene before being embedded in Epon 812 for 48 h at 60°C. Thin sections which had been cut using a diamond knife on an LKB microtome were placed on Cu grids and stained briefly with lead citrate. Observations were made in a Zeiss 10A TEM at 60 kV. For standard morphology, fresh cells were fixed in 2% glutaraldehyde in cacodylate buffer for 30 min and postfixed with 1% 0~0, in the same buffer for 1 h at 4°C. The saponin permeabilization procedure and immunostaining were omitted. These preparations were stained overnight with saturated aqueous uranyl acetate. The samples were coated, dehydrated, and embedded as described above. Thin sections were counterstained with lead citrate.

AND

BOLLUM

only in the nucleus as detected by dark brown staining (Figs. la and lb) or dark coloration (Fig. 2). In Figs. 2b, 2c, and 2d, early prophase and prophase are seen. In these cells, the TdT stain appears in both the nucleus and the cytoplasm. The nuclear distribution appears near a condensation of chromatin associated with the nuclear envelope. As the cells go into metaphase (Fig. 2e) and proceed into anaphase and telophase (Figs. 2f and 2g), the intensity of staining of nuclear material appears to decrease as the nuclear envelope disappears and chromatin condensation continues. Some residual stain may be associated with certain areas of the mitotic chromatin. The cytoplasmic staining is rather homogenous and strong. With cytodieresis and eventual reconstruction of the nuclear envelope in the last moments of the division cycle (Figs. 2h-2j), the reaction for the TdT reappears in the nucleus, with only slight positive staining remaining in the cytoplasm. The cytoplasmic staining disappears completely at the end of cytodieresis (Fig. 2j).

Electron Microscopy RESULTS

Light Microscopy Light micrographs showing several fields of mixtures of RPMI-8402 and RA-1 cells stained with the TdT-immunoperoxidase procedure are seen in Figs. la and 2. The immunoenzymatic stain shows that TdT is found only in the nucleus of the majority of the positive cells, with somewhat different localization during recognizable stages of the cell cycle (Figures la, lb, and 2). The negative RA-1 cells, treated both using primary and secondary antibodies, are readily distinguished by the absence of TdT staining (sky-blue in Figs. la and lc or light coloration in Fig. 2) in all stages of the cell cycle. During interphase, the positive cells show the enzyme

FIG. 1.

Color plate of cytocentrifuge

preparation

The standard EM morphology for RPMI-8402 cells is shown in Fig. 3a. These cells show a large nucleus containing more euchromatin than heterochromatin and a well-defined nucleolar structure. The cell surface is rather smooth with occasional membrane projections. The cytoplasm contains the usual particulate structures like mitochondria, ribosomes, and polyribosomes. No clearly defined, rough, or smooth endoplasmic reticulum can be discerned in the preparations at the magnifications shown. The immunoenzymatic procedure does not provide as high a degree of cytoplasmic fine structure as that of the glutaraldehyde-fixed cells; however, certain features of the TdT staining in the nuclear material can be described. The EM in Fig. 3b confirms the data obtained by light microscopy. Most of the TdT positive cells in interphase have a localization that appears

of positive RPMI-8402

(b), negative RA-1 (c), and mixture of positive and negative cells

(a). All preparations were stained with mouse monoclonal anti-human TdT and goat anti-mouse IgG peroxidase conjugated as the secondary antibody, reacted with DAB, and counterstained with Mayer’s hematoxylin. All positive cells showed a brown specific immunoprecipitate localized to the nucleus. Nuclei of TdT-negative cells (b and arrows in a) are only stained by Mayer’s hematoxylin. Arrowheads indicate positive cells during different phases of mitosis; the immunoprecipitate appears scattered to the cytoplasm. Bar, 10 pm. FIG. 2. Cytospin analysis of same preparation showed in Fig. la. In A the dark nucleus of positive cells and the light nucleus of negative cells can be distinguished, as well as the mitosis of positive cells (arrow). Bar, 7 pm. B-J reflect the distribution of enzyme during different stages of the cell cycle in RPMI-8402 cells. Bar, 5 pm. FIG. 3. (A) Normal ultrastructure of glutaraldehyde-fixed and uranyl acetate and lead citrate-stained RPM18402 cells. (B) EM photograph of RPM18402 positive cell during the interphase, stained by the immunoperoxidase technique described under Materials and Methods. The immunoprecipitate appears localized to interchromatinic region (arrow). (C) Negative RA-1 cell appears unstained. The nucleolus (N) and some blocks of etherochromatine (thin arrow) are stained by DAB-osmium and lead citrate reaction. (D) EM photograph of late prophase with initial disappearance of nuclear envelope (thin arrow) and condensation of chromatin. The TdT appears confined to cytoplasmic level only. (E and F) RPMI-8402 cells at methaphase and cytodieresis show positive cytoplasm and patches of peroxidase reaction product in several areas within the chromosome structure (arrow). Mitotic spindles, arrowhead. Bar, 3 pm.

IMMUNOLOCALIZATION

OF TERMINAL

TRANSFERASE

407

408

DI PRIMIO,

TRUBIANI,

AND

BOLLUM

IMMUNOLOCALIZATION

OF TERMINAL

TRANSFERASE

410

DI PRIMIO,

TRUBIANI,

to be associated with interchromatinic regions. The nucleolus and the nucleolar organizer were negative and no staining of the cytoplasm was noted in interphase nuclei. RA-1 cells are completely negative (Fig. 3~). Only the DAB-osmium staining gives rise to the positive staining at nucleolar level as we have previously described [12-131. In the TdT positive mitotic cells in Figs. 3d, 3e, and 3f, the enzyme is randomly distributed in the cytoplasm, and patches of stain apparently related to some structure in the chromosome are seen in the nuclear region. It is possible to observe a thin, granular line of reaction product delineating a chromosomal arrangement, possibly at late metaphase or anaphase (Fig. 3e). At this stage, the cells contain parts of the mitotic apparatus, including condensed chromosomes and mitotic spindles. During the early phase of cytodieresis (Fig. 3f), the forming daughter cells showed an amount of TdT randomly distributed at the cytoplasmic level. The nuclear structure that is reconstituting appears to show TdT positivity around it as observed at the first stage of mitosis (Fig. 3f). These EM observations provide confirmation of our interpretations of the light observations.

DISCUSSION

The data presented, at the light and the electron microscope levels, provide a more detailed description of the cell cycle-dependent localization of TdT that changes from essentially nuclear during interphase to mostly cytoplasmic during mitosis. Even if the analysis has been carried out on unsynchronized cells because of the impossibility at the present of obtaining a wellsynchronized RPMI-8402 cell line, the immunocytochemical features depict clear differences between interphasic and mitotic distribution of TdT. These results are seen in all TdT-positive cells that we have examined. The nuclear envelope thus appears to be the structure that facilitates the nuclear accumulation of thispartitular enzyme, which seems to be retained in the nucleus by affinity for a specific structure localized in interchromatinic regions reported to be nuclear matrix [12]. During mitosis, cells undergo considerable modification of cellular compartmentation and the TdT distribution seems to reflect this modification. Thus it is tempting to speculate that the appearance of TdT in the cytoplasm in early prophase and its persistence there throughout most of the mitotic cycle is simply the result of the dissolution of the nuclear envelope and the release of a preexisting enzyme into the cytoplasm. The results of this study do not prove that, however, and the final interpretation may be considerably more complex. Some of the requirements for nuclear localization of proteins have been defined. The accumulation and re-

AND

BOLLUM

lease of nuclear proteins appears to be controlled by specific “nuclear localization” amino acid sequences [19], and TdT contains a sequence in residues 11-17 [5] that is closely related to SV40 and oncogene nuclear localization sites [20]. It is not known whether the amino acid sequence within the transported protein is sufficient, or whether other protein and nucleic acid interactions are necessary for the transport and retention of nuclear proteins. Since it is not known at what stage in the cell cycle TdT is made, this study does not rule out the possibility that new TdT synthesis is proceeding in the cytoplasm in late S and early prophase, producing the appearance of movement out of the nucleus. The “paling” of the TdT stain in the mitotic chromatin leads to the hypothesis that the enzyme is being excluded from the condensing chromatin in agreement with previous data showing that TdT is localized in interchromatin regions [12]. Some stain does remain within the condensed chromosome structure. Whether this TdT is simply trapped in interstices or specifically bound can not be determined at present. Further study will be required to understand the intracellular redistributions of terminal transferase that occur in mitotic cells and differentiating thymocytes. It is likely that a transient linkage to nuclear matrix proteins plays a pivotal role in the cell cycle dependent TdT translocation. A combination of structural and biochemical approaches will probably be necessary to resolve this phenomenon. We are indebted to Mrs. Christel Augl for expert assistance with cell culture. This research was supported by NIH Grant CA-23262 and Italian CNR grant.

REFERENCES 1.

Bollum, F. J. (1963) Progress in Nucleic pp l-66, Academic Press, New York.

2. Kato, K., Goncalves,

J. M., Houts, (1967) J. Biol. Chem. 242,2780-2789.

Acid Research,

G. E., and Bollum,

Vol. 1, F. J.

3. Basu, M., Hedge, M. V., and Modak, M. J. (1983) Biochem. Biophys. Res. Conmun. 111,1105-1112. 4. Chang, L. M. S., Bollum, F. J. (1990) J. Biol. Chem. 265,1743617440.

5. Peterson,

R. C., Cheung, L. C., Mattaliano, R. J., Chang, L. M. S., and Bollum, F. J. (1984) Proc. Natl. Acad. Sci. USA 81,

4363-4367. 6. Bollum, F. J., Chang, L. M. S. (1986) Advances in Cancer Research, Vol. 47, pp 37-71, Academic

Press, New York.

7. Baltimore, D. (1974) Nature 248,409-411. 8. Chang, L. M. S., Rusquet-Valerius, R., Roy, N. K., Cheung, N. C., Bollum,

F. J. (1989) Biochem.

Biophys.

Res. Commun.

165,271-277. 9. Bollum, F. J. (1975) Proc. Natl. Acad. Sci. USA 72,4119-4125. 10.

Bollum,

F. J., Augl, C., and Chang, L. M. S. (1984) J. Biol. Chem.

259,5848-5855.

IMMUNOLOCALIZATION 11. 12. 13. 14. 15. 16.

OF TERMINAL

Di Primio, R., and Bollum, F. J. (1987) Hemat. Pathol. 3, 173181. Di Primio, R., Trubiani, O., and Bollum, F. J. (1991) Histochemistry 96, 59-64. Di Primio, R., Trubiani, O., and Bollum, F. J. (1992) Thymus 19,183-190. Kunkel, T. A., Copinathan, K. P., Dube, D. K., Snow, E. T., and Loeb, L. A. (1986) Proc. Natl. Acad. Sci. USA 83, 1867-1871. Alt, F. W., and Baltimore, D. (1982) Proc. Natl. Acad. Sci. USA 79,4118-4122. Desiderio, S. V., Yancopoulos, G. D., Paskind, M., Thomas, E.,

Received January 24, 1992 Revised version received May 12, 1992

TRANSFERASE

$11

Boss, M. A., Landau, N., Ah, F. W., and Baltimore, Nature 3 11, 752-755.

D. (1984)

17.

Yancopoulos, G. D., Biackwell, F. W. (1986) Cell 44, 251-259.

T. K., Suh, H., Hood, L., Ah,

18.

Muang, C. C., Hou, Y., Woods, L. K., Moore, G. E., and Minowada, J. (1974) J. Natl. Cancer inst. 53, 655-661.

19.

Kalderon, D., Richardson, W. D., Markham, A. E. (1984) Nature 3 11, 33-38.

26.

Dang, C. V., andLee, 18023.

A. F., and Smith,

W. M. F. (1989) J. Biol. Chem. 264,18019-