J Clin Pathol 1984;37:628-632

Method for the'simultaneous labelling of terminaldeoxynucleotidyl transferase (TdT) and membraneantigensJ TAVARES DE CASTRO, JF SAN MIGUEL, J SOLER, D CATOVSKY

From the MRC Leukaemia Unit, Department ofHaematology, Royal Postgraduate Medical School,Hammersmith Hospital, London W12 OHS

SUMMARY A method for the simultaneous labelling of terminal deoxynucleotidyl transferase(TdT) and membrane antigens is described. TdT is visualised in the cell nucleus with theperoxidase-antiperoxidase (PAP) method, and the immunogold method is used in combinationwith monoclonal antibodies against membrane antigens. The morphology of the labelled cells iswell preserved for analysis by light microscopy and the preparations obtained can be kept perma-nently. This method is useful for the analysis of mixed cell proliferations, particularly inleukaemias.

Microscopical observations that depend on twoimmunological markers and the morphologicalappearances of cells are important for the compari-son of leukaemic cells with corresponding normalcell types.' In acute leukaemia the most useful com-binations involve the staining of the intracellularenzyme terminal deoxynucleotidyl transferase(TdT)2 and one membrane antigen, such as thecommon acute lymphoblastic leukaemia (ALL) orHLA-DR antigens, which are usually performed byimmunofluorescence methods.'We report here a technique that combines the

immunogold method3 for the detection of mem-brane antigens with a peroxidase-antiperoxidasemethod (PAP)4 for the demonstration of TdT andpermits anatysis by light microscopy. The advan-tages of this method are: (a) the observations aremade on conventional microscopes; (b) the cell pre-parations are permanent; and (c) the morphologicaldetail is greater than with immunofluorescence pre-parations, hence facilitating the comparison of mor-phological characteristics with the immunologicphenotype of individual cells.

Material and methods

CELLS AND ANTIBODIESThe antibodies used and their specificities, sources,titres, and volumes applied are detailed in Table 1.

Accepted for publication 1 March 1984

To set up the technique, which is referred to as thecombined technique, the NALM-l cell line'" wasused. This cell line, derived from a lymphoid blastcrisis of chronic granulocytic leukaemia, expressesTdT, the common ALL antigen, and la-like deter-minants. The procedure was then tested in samplesfrom six patients with leukaemia, whose diagnosesare listed in Table 2.

COMBINED TECHNIQUECell separationThe mononuclear cell fraction from leukaemic sam-ples (peripheral blood or bone marrow) was isolatedby centrifugation at 400 g for 30 min on Lympho-prep (Nyegaard, Oslo, Norway). The cells were thenwashed, resuspended in buffer (phosphate buffered-aline (PBS), 1% bovine serum albumin, 0 2%sodium azide, pH 7.4), and counted.

IMMUNOGOLD FOR SURFACE ANTIGENSIn each test a pellet of 106 cells was incubated withthe relevant monoclonal antibody in the dilutionsand volumes detailed in Table 1 for 30 min at roomtemperature. After washing, 40 ,l of neatimmunogold was added to the pellet and incubatedfor 60 min at room temperature. The cells were thenwashed and cytocentrifuge preparations made with2 x 105 cells per slide (Shandon Cytospin, 300 rpm,7 min). It is convenient to mark the cell area with adiamond pen. At this stage the slides were dried andeither stored at room temperature for some days

628

on March 28, 2020 by guest. P

rotected by copyright.http://jcp.bm

j.com/

J Clin P

athol: first published as 10.1136/jcp.37.6.628 on 1 June 1984. Dow

nloaded from

Method for the simultaneous labelling of TdT and membrane antigens

Table 1 Immunological reagents

Antibody Specificity Source Titre Volume Reference

J5 Human common acute lymphoblastic Coulter Electronics Neat 55.d/10' cells 5leukaemia antigen

OKla Human HLA-DR framework Ortho, Raritan, New Jersey, USA Neat 5j&l/10' cells 6RFBI Immature human haemopoietic cells Dr MP Bodger 1/100 of 50p1A101 celis 7

and T lineage cells Department of Immunology, asciticRoyal Free Hospital, London fluid

FMC 8 Differentiation antigen expressed in Dr H Zola Supernatant 50ul/10' cells 8some human blood cells Flinders Medical Center, as supplied

Bedford Park,South Australia

FMC 25 Platelet membrane glycoprotein lb Dr H Zola Supernatant 50;d/10' cells 9as supplied

Immunogold Goat antibodies to mouse immuno- Janssen Life Sciences Products, Neat 40;LV10' cells 3globulin linked to 40nm colloidal gold Beerse, Belgiumparticles

Anti-TdT Rabbit antiserum to calf TdT Professor FJ Bollum 1/10 10ILI per slide 10Bethesda, Maryland, USA (approx 2 x 105

cells)Qoat antirabbit Antirabbit IgG Miles Laboratories, Slough 1/200 20g1 per slidePAP complex Rabbit antiperoxidase complexed with Bioproducts UCB, 1/300 20y1 per slide

horseradish peroxidase Brussels, Belgium

Table 2 Comparison ofthe combined method with conventional immunofluorescence methods in samples from patientswith leukaemia

Case Diagnosis Percentage ofpositve cells*

TdT J5 OKla RFBI FMC8 FMC251 Mixed BC 10 (10) 65 (70) 42 (50) 25 (33)

Myeloblastic + lymphoid+ megakaryoblastic

2 Lymphoid BC 69 (94) 98 (93) 56 (62) 100 (95)3 Mixed BC 64 (63) 10 (13) 80 (64) 73 (90)

Lymphoid + myeloblastic4 Lymphoid BC 90 (86) 97 (95) 100 (82)5 Mixed acute leukaemia 80 (70 6 (8) 25 (60) 92 (98)

Lymphoid + myeloblastic6 c-ALL 88 (92) 62 (96)

BC = blast crisis of chronic granulocytic leukaemia.c-ALL = common acute lymphoblastic leukaemia.*Percentages for combined method (immunofluorescence results in brackets).

(maximum one week) or processed immediately forTdT.

PAP technique for TdTFixation was done in methanol at 4°C for 45 min.After fixation, slides were washed in PBS for 10 minin a container with buffer and a magnetic stirrer.The cells were incubated successively with anti-TdT,goat antirabbit immunoglobulin, and PAP complex.The 30 min incubations were done at room tempera-ture in a moist chamber and each was followed bywashes in PBS.

After the last wash, slides were briefly immersedin Tris-HCI buffer (0-05 M Tris-HCI 0-9% NaCl,pH 7-4) and peroxidase activity was developed byimmersion for 10 min at room temperature in thedark in diaminobenzidine medium: 0-3 mg/mldiaminobenzidine (Aldrich Chemical Co, Gilling-ham, UK), 0.03% H202 in Tris-HCI buffer.

Counterstaining and mountingAfter washing with Tris-HCI buffer and a briefimmersion in distilled water, the cells were counter-stained with 1% methyl green (5-10 min), washed inwater, dehydrated by immersion in graded ethanols(50, 70, 90, absolute ethanol), cleared with xylene,and mounted permanently with DPX (RaymondLamb, London). The preparations were examinedwith a conventional light microscope.

Endogenous peroxidaseEndogenous peroxidase could be blocked by immer-sion in 0 3% H;O2 in methanol for 30 min immedi-ately after fixation.

ControlsFor negative controls either the monoclonal anti-body in the first step, the anti-TdT in the second, orboth were omitted. We tested the samples by con-

629

on March 28, 2020 by guest. P

rotected by copyright.http://jcp.bm

j.com/

J Clin P

athol: first published as 10.1136/jcp.37.6.628 on 1 June 1984. Dow

nloaded from

630

MM TV- *

Fig. Nalm-1 cells. TdT detection by the

peroxidase-antiperoxidase method. The reaction is present

mainly in the nucleus. A dividing cell shows the reaction

diffusely distributed in the cytoplasm, but absent in the area

containing the chromosomes (arrow). x 000.

Fig. 2 Case 3, combined method: TdT and JS. Largeblasts are TdT negative, JS negative (large arrow); mediumsize blasts TdT positive, predominantly JS negative (smallarrow); and small blasts TdT positive, predominantly JSpositive (arrow head). TdT reaction is stronger in smallblasts. Original magnification x 000.

Tavares de Castro, San Miguel, Soler, Catovsky

ventional immunofluorescence for TdT and therelevant surface antigens. Table 2 shows a compari-son of the results with those obtained by the newmethod.

Results

The percentage of positive cells by the combinedtechnique correlated well with the results obtainedby conventional immunofluorescence methods withsimilar reagents (Table 2). TdT was detected as abrown reaction present in the nucleus in a fine par-ticular pattern. The intensity of reaction varied fromcell to cell. In some cells a weak reaction was alsopresent in the cytoplasm. In dividing cells the reac-tion was diffusely distributed through the cytoplasm,with the area containing the chromosomes devoid ofreactivity (Fig. 1). In leukaemic samples the TdTreaction tended to be stronger in small blasts with alymphoid appearance (Fig. 2). Inhibition ofendogenous peroxidase with H202 weakened theTdT reaction. When this procedure was not usedrare erythrocytes and myeloid cells were positive;this did not constitute a problem because they couldbe recognised morphologically. Cells reacting withthe monoclonal antibodies had numerous black dotsaround the surface membrane. In the strongly reac-tive cells these dots were confluent and covered var-iable sections of the cell surface as a black lining(Fig. 3). Most of the unreactive cells had no dots.Cells with fewer than three dots were considerednegative. Ultrastructural studies (results not shown)have shown that the individual dots correspond topatches of 40 nm colloidal gold particles.'2The combined method was particularly useful for

analysing the cellular components of leukaemiaswith different types of blasts (cases 1, 3, 5), charac-teristically seen in mixed acute leukaemias andmixed blast crisis of chronic granulocytic leukaemia.In case 1 TdT positive blasts also expressed the anti-gens recognised by OKla and FMC8, but the anti-platelet antibody FMC25 reacted only with TdTnegative blasts. In cases 3 and 5 with mixtures ofmyeloblasts and lymphoblasts large blasts were TdTnegative, J5 negative, and RFB1 positive; mediumsize blasts were TdT positive, predominantly J5negative, and RFB1 positive; and small cells wereTdT positive, predominantly J5 positive, and RFB1positive (Figs. 2 and 3). In the control preparationsall the cells were negative for the omitted antibody.The percentage of positive cells was the same inexperiments where the two steps of the techniquewere performed separately and in combination,which shows that they do not interfere with eachother.

;if: t.

.iiC,l

:t.',.!i

-.111k ".:",..;t:;: -s'

v... Al:

AL

on March 28, 2020 by guest. P

rotected by copyright.http://jcp.bm

j.com/

J Clin P

athol: first published as 10.1136/jcp.37.6.628 on 1 June 1984. Dow

nloaded from

Method for the simultaneous labelling of TdT and membrane antigens

Fig. 3 Case 5, combined method:TdT and RFBI. One large blast isRFBl positive, TdT negative(arrow), the other blasts expressboth markers. Originalmagnification x 1000.

Discussion

Although more time consuming than conventionalimmunofluorescence methods this techniquerequires only the use of a simple light microscopeand provides permanent preparations. We found a

good correlation between the combined techniqueand conventional immunofluorescence methodsboth for TdT and surface antigens, and thespecificity of both components of the method was

also good as a positive reaction in the controls was

never seen. The morphological detail obtained was

good and from each single cell it was possible toobtain three types of information: morphology andexpression of TdT or of a membrane antigen, or

both.This method is particularly useful in the study of

mixed leukaemias and blast crisis of chronicgranulocytic leukaemia, where frequently two or

more different populations of blasts coexist.'3 Inthese cases it is possible: (a) to analyse the differentmorphological characteristics of the blasts that cor-respond to each of the immunological phenotypespresent (cases 3 and 5), and (b) to determine if TdTand a membrane marker can be simultaneously pre-sent in the same cell or are always expressed in dif-ferent blasts (case 1). This technique can be used forthe analysis of single cells in certain tissues-forexample, cerebrospinal fluid and remission bonemarrow-where it may help identify cells with a

leukaemic membrane phenotype and thus distingu-ish them from putative normal precursors.' Anotherpossible area of application could be in the studies ofin vitro differentiation of lymphoid cell lines, wherechanges in the expression of two markers could be

correlated with the morphological evolution.'4In the present work we studied the expression of

five different surface antigens: the same approachcould be extended to any cell membrane antigendetectable by a monoclonal antibody or heteroanti-sera. A possible variant of the technique is thedetection of cytoplasmic immunoglobulin'5 insteadof TdT and its association with the membranephenotype as shown by the immunogold technique.The combination of the immunogold technique withcytochemistry has been reported from this labora-tory both in studies by light microscopy'6 and by elec-tron microscopy.'2

We are grateful to Professor FJ Bollum for the giftof anti-TdT, Dr MF Greaves for the Nalm-l cell line,Dr HP Bodger for the RFBI, and Dr H Zola for theFMC8 and FMC25 monoclonal antibodies. JTC wassupported by the Calouste Gulbenkian Foundation,Portugal, and JFSM by the Science and EducationMinistry of Spain.

References

'Janossy G, Bollum FJ, Bradstock KF, Ashley J. Cellularphenotypes of normal and leukemic hemopoietic cells deter-mined by analysis with selected antibody combinations. Blood1980;56:430-41.

2 Bollum FJ. Terminal deoxynucleotidyl transferase as ahematopoietic cell marker. Blood 1979;54: 1203-15.

De Waele M, De Mey J, Moeremans M, De Brabander M, VanCamp B. Immunogold staining method for the light micro-scopic detection of leukocyte cell surface antigens withmonoclonal antibodies: its application to the enumeration oflymphocyte subpopulations. J Histochem Cytochem1982;31:376-81.

631

on March 28, 2020 by guest. P

rotected by copyright.http://jcp.bm

j.com/

J Clin P

athol: first published as 10.1136/jcp.37.6.628 on 1 June 1984. Dow

nloaded from

632

Hecht T, Forman SJ, Winkler US, et al. Histochemical demon-stration of terminal deoxynucleotidyl transferase in leukemia.Blood 1981;58:856-8.

Ritz J, Pesando JM, Notis-McConartyJ', Lazarus H, SchlossmanSF. A monoclonal antibody to human acute lymphoblasticleukaemia antigen. Nature 1980;283:583-5.

6 Reinherz EL, Kung PC, Pesando JM, Ritz J, 'Goldstein G,Schlossman SF. Ia determinants on human T-ceil subsetsdefined by monoclonal antibody: activation stimuli requiredfor expression. J Exp Med 1979;150: 1472-82.

7Bodger MP, Francis GE, Delia D, Granger SM, Janossy G. Amonoclonal antibody specific for immature human hemopoie-tic cells and T-lineage cells. J Immunol 1981;127:2269-74.

Brooks DA, Bradley J, Zola H. A differentiation antigenexpressed selectively by a proportion of human blood celis:detection with a monoclonal antibody. Pathology 1983;14:5-11.

Zola H, McNamara PJ, Beckman IGR, et al. Monoclonal anti-bodies against antigens of the human platelet surface: prepara-tion and properties. Pathology 1984;16:73-8.

Bollum FJ. Antibody to terminal deoxynucleotidyl transferase.Proc Natl Acad Sci USA 1975;72:4119-22.

Minowada J, Tsubota T, Greaves MF, Walters TR. A non-T,non-B human leukemia cell line (NALM-1): establishment ofthe cell line and presence of leukemia-associated antigens. JNad Cancer Inst 1977;59:83-7.

Tavares de Castro, San Miguel, Soler, Catovsky12 Robinson D, Tavares de Castro J, Polli N, CYBrien M, Catovsky

D. Simultaneous demonstration of membrane antigens andcytochemistry at ultrastructural level: a study with theimmunogold method, acid phosphatase and myeloperoxidase.Br J Haematol 1984 (in press).

3San Miguel JF, Tavares de Castro J, Matutes E, et al. Character-isation of blast cells in chronic granulocytic leukaemia in trans-formation, acute myelofibrosis and undifferentiatedleukaemias. II. Studies with monoclonal antibodies and termi-nal transferase. Br J Haematol (submitted).

Le Bien TW, Boilum FJ, Yasmineh WG, Kersey JH. Phorbolester-induced differentiation of a non-T, non-B leukemic cellline: model for human lymphoid progenitor cell development.J Immunol 1982;128:1316-20.

'5 Mason DY, Farrel C, Taylor CR. The detection of intracellularantigens in human leucocytes by immunoperoxidase staining.Br J Haematol 1975;31:361-70.

16 Crockard A, Catovsky D. Cytochemistry of normal lymphocytesubsets defined by monoclonal antibodies and immunocolloi-dal gold. Scand J Haematol 1983;30:433-43.

Requests for reprints to: Dr D Catovsky, MRC LeukaemiaUnit, Royal Postgraduate Medical School, HammersmithHospital, Ducane Road, London W12 OHS, England.

on March 28, 2020 by guest. P

rotected by copyright.http://jcp.bm

j.com/

J Clin P

athol: first published as 10.1136/jcp.37.6.628 on 1 June 1984. Dow

nloaded from