A micromethod for determination of terminal deoxynucleotidyl transferase (TdT) in the diagnostic evaluation of acute leukemias

J Cancer Res Clin Oncol 98, 91-104 (1980)

doumat of

~Cancer Research Clinical Oncology @ Springer-Verlag 1980

A Micromethod for Determination of Terminal Deoxynucleotidyl Transferase (TdT) in the Diagnostic Evaluation of Acute Leukemias*

M.J. Modak, R. Mertelsmann, B. Koziner, R. Pahwa, M.A.S. Moore, B.D. Clarkson, and R.A. Good

Memorial Sloan-Kettering Cancer Center, 1275 York Avenue, New York, 10021, USA

Summary. A micromethod for the determination of TdT in peripheral leukocytes and bone marrow cells has been developed that allows unequivocal iden- tification and quantitation of TdT in less than 1 x l06 leukocytes from ALL patients, i.e., in 1 ml of peripheral blood and/or 0.5 ml of bone marrow obtained during routine clinical sampling. The method involves disruption of cell pellet with high salt and detergent followed by centrifugation of extracts at 12,000 x g and partial purification on phosphocellulose matrix by a batch el- ution technique using a standard laboratory microcentrifuge. Using this microassay, TdT activities have been determined in 500 samples of peripheral blood and bone marrow of 240 adult patients with acute leukemias (86 ALL, 108 ANLL, 44 blastic CML, two acute leukemias following P. vera). From an analysis of our data based on TdT activity, cell surface markers and growth patterns in soft agar and observations published in the literature, it can be concluded that the frequencies of TdT + phenotypes in the various clinical- morphological diagnostic groups are approximately 95~ in ALL, 10~ in ANLL, 50~o in AUL, and 35~ in blastic CML. Since the presence of high TdT activity is clearly associated with clinical response to specific forms of chemo- therapy in blastic CML and most probably, also in ANLL, the determination of TdT should be considered in all cases of acute leukemias to objectively define prognostically important subgroups which can not be diagnosed by conventional means.

Key words: TdT micromethod - Acute leukemia

Terminal deoxynucleotidyl transferase (TdT) 2 is a biochemically and biologically unique DNA polymerase which adds deoxyribonucleotides onto an appropriate

* This research was supported in part by the following grants: ACS PDT-95, NCI-CA-17404, NCI-CA-19267, NCI Program Project Grant 3 PO1 CA-20194, NCI Grant CA-08748, the Gar Reich- man Foundation and the Zelda R. Weintraub Cancer Fund, and a research career development award no. 1 KO4-CA-00545 from the National Cancer Institute to MJM Offprint requests to: Dr. M.J. Modak (address see above)

0171-5216/80/0098/0091/$2.80

92 M.J. Modak et al.

primer molecule in the absence of any directing template polynucleotide (Bollum 1974). The distribution of TdT is restricted to thymocytes under physiological conditions in all vertebrate species including man and to a subpopulation of bone marrow lymphocytes, which exhibit characteristics of immature T cells, i.e., prothymocytes (Silverstone et al. 1976). Because of this restricted distribution to early de- velopmental stages of T-lymphocytes (Baltimore 1974; Bollum 1975; Hutton and Bollum 1977; Kung et al. 1975; McCaffrey et al. 1973; Mertelsmann et al. 1978 a, b; 1979a, b; Sarin et al. 1976; Silverstone et al. 1976; Sujimoto and Bollum 1979), which lose TdT activity during final maturation into circulating T cells, it has been speculated that TdT might play a role in the generation of immunological diversity (Baltimore 1974). Although these models have not been experimentally verified, re- cent studies have confirmed the close association of the appearance of TdT and im- munocompetent T cells during ontogeny in chicken embryos (Sujimoto and Bollum 1979), strongly suggesting a physiological role for TdT in the generation of im- munocompetent T cells.

High levels of TdT activity have been found in most cases of ALL, and in approximately 30% of the cases with blastic CML (Hoffbrand et al. 1977; Kung et al. 1978; Mertelsmann et al. 1978a; Sarin et al. 1976). These observation have at- tracted considerable interest in TdT as a potential marker for normal and malignant T-cell-related progenitor cells with both diagnostic and therapeutic implications (Mertelsmann et al. 1978a; Bollum 1978). All reported techniques for biochemical quantiation of TdT activities in human cells require high cell numbers, unobtainable from routine clinical samples. In addition, they require sophisticated equipment, e.g., ultracentrifuge, column chromatography equipment, not readily available in a clinical biochemistry laboratory. We have previously reported an improved assay for TdT which can be performed on a relatively small cell number (~ 107 cells) and does not require ultracentrifugation (Mertelsmann et al. 1978a). However, as a result of the increasing recognition of the clinical significance of TdT determinations in hematopoietic neoplasias (Coleman et al. 1974, 1976; Kung et al. 1978; Marcus et al. 1976; Marks et al. 1978), we have further simplified the procedure and developed a micromethod which allows TdT determinations on less than 104 cells and can be carried out in any clinical laboratory that has access to a liquid scintillatin counter. DEAE-Sephadex column chromatography described earlier (Mertelsmann et al. 1978a) is replaced by batch adsorption of extracts to phosphocellulose matrix and enzyme is eluted in concentrated form. These modi- fications permit handling of as many as 20-30 samples a day. The details of this assay and its application in the diagnostic evaluation and phenotypic characterization of 240 cases of acute leukemias of adults are described in this report.

Methods

Patients

Peripheral blood and/or bone marrow samples of 240 patient with acute leukemias, followed by the adult Hematolgy/Lymphoma Service, Memorial Hospital, were studied at diagnosis and/or at relapse of their disease. Leukemia diagnoses were based on Wright-Giemsa stained smears as well as cytochemical stains of bone marrow aspirates and classified according to the FAB classification (Bennet et al. 1976).

TdT Micromethod in the Diagnosis of Acute Leukemia 93

Normal control bone marrow and peripheral blood samples were obtained from healthy volunteers.

Isolation of Cells

One to 10 ml of peripheral blood and/or 0.5-1.0 ml of iliac crest bone marrow were collected in heparin for assays of cell markers and TdT acitivity. For TdT determinations, mononuclear cells were isolated after a 1:1 dilution in normal saline, layered onto Ficoll-Hypaque (Pharmacia Fine Chemicals, Inc., Pis- caraway, N J) gradients and centrifuged at 400 x g for 20 rain according to the method of Boyum (Boyum 1968). Cells were washed once in normal saline, an aliquot counted in a Coulter Counter (Mod- el S) and cells resuspended at a cell concentration of 1-2 x 107/ml in buffer A (50 mM Tris HC1 pH 7.8, 200 mM KC1, 0.5~ triton x-100) (Sigma, Inc., St. Louis, MO, USA), 0.5 mM K-EDTA, 0.1 mg/ml bovine serum albumin (Fraction V, Mils Laboratories, Inc., Elkhart, IN, USA) and 0.1 mM dithiothreitol, added just prior to use). Cell suspensions were sonicated in closed plastic tubes (38 x 12.5 mm, round bottom tubes, Vanguard Inc., Neptune, NJ USA) in the cuphorn of a Heat Systems - Ultrasonics Sonicator (Model W 200R, Plainview, N.Y., potentiometer setting 60~, pulsar setting 50~, 10 bursts) and stored frozen at - 2 0 ~ until used for the TdT assay. Alternatively, cells may be disrupted in high salt (1 M), other components being identical. The usage of high salt obviates the need for sonication although dilution of extracts is required for subsequent phosphocellulose adsorption step (see below).

Isolation of cells was as previously described for cell surface markers (Koziner et al. 1978) and the soft agar culture (CFU-c, Moore 1975a, 1976).

For sedimentation velocity analysis, 5-10 ml of bone marrow from normal volunteers was as- pirated in 1 ml aliquots from multiple sites of the posterior iliac crest. After Ficoll-Hypaque separation, mononuclear cells were further separated by velocity sedimentation at unit gravit using 1,200 ml albumin gradients for 4 h at 4~ a technique which separates cells on the basis of their size (Miller and Phillips 1969; Phillips and Miller 1970). After discarding the bottom 250 ml of the gradient, which did not contain any cells, 35 ml fractions were collected. Cells in each fraction were collected by centrifugation, washed once, counted, and stored frozen until further use. Normal human thymus from chil- dren was obtained during cardiac surgery and cells were separated by teasing and passing through a wire mesh. Thymocytes were washed, counted and indicated cell numbers stored frozen at - 2 0 ~

Frozen calf thymus tissue was suspended at 2 x 107 cells (equivalent to 20 mg tissue) per ml of buffer A and homogenized in a blender (Sorvall Ommi-Mixer, speed control setting 5, 3 x 1 min in ice). The 100,000 g supernatant (1 h, 4 ~ was stored in aliquots at - 2 0 ~ After thawing this supernatant ser- ved as standard and was processed through all further preparative steps of the procedure in parallel to the samples to be tested.

Cell Surface Markers, CFU-e

Analysis of cell surface markers included immunoflurescent staining for surface immunoglobulins (sIg), rosette formation with sheep erythrocytes (E-rosettes) at 4 ~ and 37 ~ and with mouse erythrocytes, human EA rosette technique, determination of phagocytic ability using latex particles as well as adherence procedures and was performed as previously described (Koziner et al. 1978; Mertelsmann et al. 1978b). Analysis of incidence and growth pattern of myeloid committed stem cells in soft agar was performed as described (Moore 1975b). Acute myeloid, myelomonocytic, and monocytic leukemias exhibed a cluster ( < 40 cells per aggregate) growth pattern with no colonies (> 40 cells per aggregate) in the majority of cases while lymphoid and non-lymphoid/non-myeloid leukemias revealed no growth under the culture conditions used or a low incidence of colonies and clusters with a normal growth pattern (Moore 1975a, b; 1976).

Phosphocellulose Adsorption and TdT Assay

Upon thawing, samples were sonicated again, and a clear supernatant was obtained by centrifugation at top speed in 1.5 ml plastic tubes in a microcentrifuge (Beckman or Eppendorf) for 5 min. One hundred ~tl of the clear supernatant (S-10 fraction) was then mixed with 50 ~tl of a 60% (v/v) phosphocellulose slurry prepared from commerical phosphocellulose powder (Whatman P- 11) previously activated by treatment with alkali and acid (Burges 1969) and equilibrated in TGED (50 mM Tris HC1 pH 7.8, 0.5 mM K-EDTA, 10~ glycerol (v/v), 0.1 mg/ml bovine serum albumin, and 0.1 mM dithiothreitol, added just prior to use). After incubation for 15 rain at 4 ~ the phosphocellulose with the TdT bound to it was spun down


in 2 min in the microcentrifnge. The supernatant containing negligible amounts of TdT was discarded, the pelleted phosphoeellulose washed once in 150 g TGED, and TdT was eluted by addition of 50 rtl of 0.6 M KC1 in TGED. For experimental samples, e.g., after further cell separation procedures, cell concentrations as well as washing and ehition buffer volumes may be modified as necessary to allow quantitation of very low amounts of TdT activity. As little as 50 gl of cell suspension could be reliably handled using this technique. After the final spin in the microeentrifuge, 5-10 gl &the supernatant was assayed for TdT activity as previously described, using 20-50 gM (3H) dGTP (sp. act. 750 cpm/pmol) as substrate and 5 ~tg/ml oligo d(pA)a z- 18 as primer, in a total volume of 50 gl (Marcus et al. 1976; Modak et al. 1978). Plaosplaocellulose ehites may be obtained in somewhat larger volumes (~ 200 gl) to avoid losses due to fluid trapping. Lower concentration of non-radioactive dGTP (resulting in the higher specific activity) have been occasionally used when cell number were too low (Table 1). Duplicate assays in the absence and presence of 50 gM ATP were perfornaed. ATP has previously been shown to be a specific inhibitor of TdT aeitivity (Bhalla et al. 1977; Modak 1978). Results were calculated from the dif- ference in incorporation in the absence and presence of ATP and expressed in units/108 cells (1 unit= I nmol (3H)dGMP incorporation/hr at 37 ~ If no TdT activity was detected, results were expressed by giving the lower limit of detectability under the conditions used, e.g., <0.05 units/108 cells.

Other TdT Assay System

A commercial kit based on the procedure developed by Bollum and collaborators for the determination of TdT in clinical samples is available now (Coleman 1977; Greenwood et al. 1977). Basically, this procedure involves assay of enzyme activity in soluble extracts obtained by centrifugation of disrupted cells at 10,000 x g (S-10) or at 100,000 x g (S-100). Enzyme activity is determined by using high substrate concentration (1 mM) requiring use of large quantities of radioisotope, oligo (dA)~o as a primer and Mg z+ as an effective divalent cation (Coleman 1977; Gordon et al. 1978). To compare the results obtained by our procedure with those obtained by conventional procedure, we have used commercial kits obtained from Bethesda Research Laboratory, Bethesda, MD. All the reagents were supplied in this kit and S-10 extract obtained from various sources was used as source of enzyme.

DNA Polymerase

To determine the presence of DNA polymerases, in the cell extracts the assay system consisting of activated DNA, four deoxynucleoside triphosphate, dithiothreitol and Mg a§ was used (Modak 1979). These assays also included 50 ~tM ATP to suppress TdT catalysis if any.

Deoxyribonuclease Assays

The presence and quantitation of DNase in the cell extracts before and after phophocelhilose adsorption of enzyme was determined by monitoring the degradation of 3H-labeled Poly dT. For this purpose, both Mg 2§ and Mn z+ were used as an effective divalent cation (Modak and Marcus 1977).

Results

Methodology

E n z y m e recover i e s o b t a i n e d f r o m h u m a n b o n e m a r r o w m o n o n u c l e a r cells a f t e r ve loc i ty s e d i m e n t a t i o n (cf. Fig . 1), f r o m h u m a n t h y m u s a n d f r o m a p a t i e n t w i t h a c u t e l y m p h o b l a s t i c l e u k e m i a (ALL) were f o u n d p r o p o r t i o n a l to t he n u m b e r o f cells

a s s a y e d o v e r a r a n g e o f 1 z l 0 s to 107 cells pe r assay (Tab le 1). T h e t i m e c o u r s e o f

(a l l ) d G M P i n c o r p o r a t i o n was l inea r for u p to 90 min . T h e K m v a l u e for T d T

u n d e r the c o n d i t i o n s used was 22.5 ~tM. B a c k g r o u n d was a p p r o x i m a t e l y 100 cpm.

Fu l l va lues r a n g e d f r o m 300 to 200,000 cpm. T d T ac t iv i ty f r o m all t issues s tud ied was i nh ib i t ed by m o r e t h a n 8 0 ~ o f t he c o n t r o l v a l u e in the p re sence o f 50 ~tM A T P (da t a n o t shown) , c o n f i r m i n g t h a t the e n z y m e u n d e r assay was T D T (Bha l l a et al. 1977; M e r t e l s m a n n et al. 1978a). R e c o v e r y e x p e r i m e n t s us ing pur i f i ed T d T o r ex-

TdT Micromethod in the Diagnosis of Acute Leukemia

Fig. 1. Distribution of TdT in normal bone marrow cells separated according to sedimentation velocity. Mononuclear bone marrow cells from a normal donor were further separated by sedimentation velocity in an albumin gradient at unit gravity (see Methods). Fastest sedimenting, i.e., largest cells at the left, slowest sedimenting, i.e., smallest cells at the right

500

400

,o

300 o

3 = 200

o_

E 100

/ ,

2 i0 15 20 25 Fraction number

95

5

S x

3 ~ -6

E

Table 1. Effect of cell number of TdT activity in human thymic cells

Source of cells Number of cells Incorporation/assay Specific activity ( x 10 s) (pmol/h) a (units/108 cells)

Human thymus 1 1 1.00 2 2.3 1.15 5 5.5 1.10

10 10.5 1.05 20 20 1.00 50 48.0 0.96

100 86.0 0.86 Human bone marrow, 1 0.1 - - Velocity sed. no. 17 5 1.8 0.36

10 3.2 0.32 25 8.2 0.33 50 15.8 0.32

Acute lymphoblastic leukemia 2 0.1 - - 5 1 0.20

10 2.1 0.21 50 9.8 0.19

Assays were performed as described under Methods except that cells were stored - 2 0 ~ and assayed in the presence of dGTP (specific activity 3500 cpm/pmol) " Actual assay used only 20 pl extract out of a total of 200 g]

as cell pellets at

t rac ts f r o m T d T - p o s i t i v e cells m i x e d wi th T d T - n e g a t i v e cells gave p r o p o r t i o n a t e yields. W i t h i n the l imi ts o f cell n u m b e r s used ( m a x i m u m 5 x 106 pe r 50 gl o f phos -

phoce l l u lo se s lurry) a c o m p l e t e r e c o v e r y o f e n z y m e f r o m p h o s p h o c e l l u l o s e was ach ieved . R e s i d u a l T d T ac t iv i ty o b t a i n e d d u r i n g r e p e a t ex t r ac t ions o f the p h o s -

p h o c e l l u l o s e pel le t a m o u n t e d to 1 0 ~ - 3 0 ~ o f the ini t ia l ex t rac t , m o s t p r o b a b l y the

resul t o f f lu id t r a p p e d b e t w e e n p h o s p h o c e l l u l o s e par t ic les a n d r ema ins in the test tubes. S l ight ly l ower specif ic act ivi t ies t h o u g h qua l i t a t i ve ly iden t ica l resul ts were

96

Table 2. Comparison of two assay systems for TdT in various cell extracts

M.J. Modak et al.

Source of Cell con- Assay A" Assay B b TdT centration

dGTP Incorporation cpm pmol cpm pmol

Percent relative efficiency

B/A Calf thymus 5 • 106 2,650 10.6 6,600 8.8 83 Calf thymus 2 • 107 8,400 33.6 9,600 12.8 38 Calf thymus 1 x 108 59,800 239.0 32,800 47 20 CML-LB1 2 x 107 6,728 26.9 37,480 26.8 100 CML-LB2 2 x 10 v 5,947 23.8 46,965 33.6 141 CML-LB3 1 x 106 < 50 < 0.2 851 1.1 >500 ALL-1 5 X 106 1,205 4.8 8,050 5.7 118 ALL-2 1 • 106 < 50 < 0.2 1,205 1.6 >800 AML-1 2 x 10 v 213 0.8 < 50 - - ~ 4 AML-2 1.5x 107 388 1.5 < 50 ~ 2

Assays were carried out as described in Materials and Methods. Incubation time was 30 min at 37 ~ Five microliters of homogenate (S-10) or phosphocellulose eluate was used as an enzyme source for assay system A and B, respectively a Assay A represents conventional assay which used 1 mM 3H-dGTP and Mg 2 + as divalent cation. CPM values are connected for background obtained in the absence of oligo dA. Background varied from 300 to 1,800 cpm for various cell extracts b Assay B represents values obtained with our assay system Using 50 gM 3H-dGTP and Mn z+ as divalent cation. CPM values are corrected for background (The values obtained in the presence of 50 IxM ATP are considered as background since inclusion of ATP reduces the TdT activity by > 90%)

obta ined when samples were frozen and thawed in the presence of high salt instead of sonication.

TdT Assay Using Conventional Procedure (Geenwood et al. 1977)

To compare the efficiency of our procedure with the one that does no t utilize phosphocellulose adsorpt ion step, we determined the TdT activities using pre-and post- phosphocellulose extracts as an enzyme source. In addit ion, the assay system used to assay individual extracts also differed significantly. In our assay only 50 ~tM substrate is used, while the convent iona l assay uses 1 m M d G T P in cacodylate based system. Table 2 lists the values of TdT obta ined f rom various leukemic cell extracts under the two assay condit ions. Wi th calf thymus as an enzyme source, S-10 fract ion apparent ly does no t need adsorpt ion to phosphocellulose step to ob- serve T d T catalysis. In fact, with the limited a m o u n t of phosphocellulose used, a higher concent ra t ion of cells, enzyme recovery is significantly reduced. However, with mos t of leukemic cell extracts, phosphocellulose adsorpt ion step appears to improve the TdT recoveries. The increase in the TdT activity of extracts obta ined by phosphocellulose batch adsorpt ion procedure has been part ial ly due to removal of nucleases, other con tamina t ing proteins and D N A polymerases (data no t shown). Fur thermore , our procedure has consistenly showed extremely low backgrounds for the cell extracts which were TdT negative (e.g., AML). Conven t iona l procedures have yielded false positive values due p robab ly to high backgrounds .

TdT Micromethod in the Diagnosis of Acute Leukemia

Table 3. TdT activity in normal human tissues and calf thymus

97

Tissue Samples studied Specific activity (Units/108 cells) SD n

Mean Range

Human tissues Peripheral blood 10 < 0.018 < 0.010-< 0.030 0.009 Bone marrow 12 0.048 < 0.010- 0.109 0.034 Lymph nodes 2 < 0.030 < 0.020-< 0.040 - - Tonsils 2 0.020 < 0.020- 0.032 - - Spleen 4 < 0.015 < 0.010- 0.020 - - Thymus 2 1.08 0.78 - 1.39 - -

Calf thymus Extract I 13 11.3 9.2 - 13.3 1.42 Extract II 14 14.5 11.4 - 18.2 2.33

All separations and assays performed under standard conditions as decribed under Methods

These results demonstrate superiority of phosphocellulose adsorption procedure for unambiguious determination of TdT.

Normal Tissues

No TdT activity was detected in any normal peripheral blood samples, while TdT activities in normal bone marrow specimen ranged from < 0.01 to 0.109 (Table 3). Similarly, no TdT activity was detected in non-malignant nodes, uninvolved spleens obtained during staging laparotomies, while occasionally low TdT activities were detected in some inflamed tonsils (Table 3) as previously reported (Modak et al. 1978). The TdT activity exhibiting subpopulation of human bone marrow cells could be further enriched using cell separation according to sedimentation velocity (Miller and Phillips,1969 and Fig. 1), as previously reported by Barr et al. (1976) except that every single cell fraction has been assayed by us. Although only small numbers of cells were obtained in each fraction, these samples were found adequate for reproducible quantitation of TdT activities using the methodology described. We have been able to demonstrate the increased expression of the early T-cell marker HTLA in this cell population in respose to thymopoietin (Mertelsmann et al. 1979a), strongly suggesting the pre-T-cell character of this cell population.

TdT Acitivites in Acute Leukemias

Cell marker phenotypes and clinical diagnoses of 240 adult patients with acute leukemias are shown in Table 4. The cases were classified by clinical, morphological, and cytochemical criteria into (1) acute lymphoblastic leukemia (ALL), (2) acute non-lymphoblastic leukemia (ANLL), which included all categories (M1- M6) of the FAB classification (Bennett et al. 1976) as well as the acute undifferentiated leukemias (AUL), and (3) into myeloproliferative syndromes in blastic transformation or acute phase (MPS-AP), which included 44 cases of blast phase CML,


and two cases of polycythemia vera terminating in an acute leukemia. In addition, all leukemias were classifield according to phenotypic features of the predominating cell type as determined by TdT activity, CFU-c incidence and growth pattern and surface markers (sIg, E-rosettes). Th~ lymphoid "non-b, non-T" phenotype was defined as exhibiting high levels of TdT activity (0.5 to 34 units/108 cells), no E- rosetting, negative sIg and no or low normal CFU-c incidence'and growth pattern. The thymic T-cell phenotype ("thy-T") exhibited E-rosetting in addition to high levels of TdT acitivty while the peripheral blood T-cell type (PB-T) revealed E- rosette formation only. The B-cell phenotype was TdT negative and exhibited surface immunoglobulins, no E-rosetting and no or low normal CFU-c growth.

Two of 21 cases of "T cell" ALL did not exhibit detectable TdT activity and probably represent malignant proliferations of a more mature T cell type ("IPB- T"). Fourteen of 108 cases of ANLL exhibited high levels of TdT activity within the range seen in ALL (0.2-12.4 units/108 cells). In one half of these, the growth pattern in soft agar suggested the presence of leukemic myeloid progenitor cells, while in the other seven cases no evidence for a myeloid component could be detected by these techniques. Cell marker data in the latter category with a "non-B, non-T, TdT + " phenotype were suggestive of their "T-cell" lineage in spite of a conventional morphological diagnosis of ANLL. All but one patient of 10 cases of TdT + ANLL achieved a clinical response to vincristine and prednisone with significant reduction or disappearance of bone marrow and peripheral blood blast cells. Sixteen of 44 caes of blast phase CML exhibited high levels of TdT acitivity (0.2-64.0 units/108 cells). Two of these also exhibited a myeloid leukemic growth pattern in agar, suggesting the coexistence of more than one leukemic phenotype as we have observed in some cases of ANLL (Mertelsmann et al. 1978b). Response to therapy was significantly (p<0.01) associated with the TdT+ phenotypes with a complete response in six of eight cases who received a regimen containing vincris- fine and prednisone. These cases reverted temporarily to the chronic phase of their disease, in several cases extending beyond two years. Two cases of polycythemia vera terminating in an acute leukemia also exhibited high levels of TdT activity. One of these patients received vincristine and prednisone and achieved a remision of several months duration.

Discussion

Although antibodies of reproducible specifity and sensitivity for TdT have been described for demonstration of TdT in clinical specimes by immunofluorescence (Bol- lure 1978; Marks et al. 1978), detection and quantitation of this enzyme by biochemical techniques still has the advantages of being independent of limited supplies of antibody and of objective quantiation of TdT activity. In addition, the assay described here can be reproducibly carried out with the same low numbers of cells as required for the immunofluorescent assay. Since there is disagreement between biochemical determination and the demonstration of TdT by immunofluorescence in approximately 10% of the cases (unpubl. data), we would favor the ATP-controlled biochemical assay until cross-reactivities and non-specific staining reactions can be excluded by using appropriately characterized monoclonal antibodies. An addition- al advantage of the biochemical assay described here is that it is performed with stan-


dard laboratory equipment and does not require sophisticated techniques, such as fluorescence microscopy or fluorescence activated cell sorting. The comparatively large number of cells needed for the biochemical assays most widely used (Cole- mann et al. 1974, 1976, 1977, 1978; Gordon et al. 1978; Hutton and Coleman 1976; Kung et al. 1978; Penit et al. 1977; Satin et al. 1976; Shaw et al. 1978; Srivastava et al. 1978; Vogler et al. 1978) had made TdT estimation previously difficult in patients with low peripheral counts.

In addition to its clinical importance in diagnosis and perhaps prognosis (Don- lonet al. 1977; Filippa et al. 1978; Meyskens and Jones 1978), TdT determinations have important applications in the study of normal and malignant T-cell differentiation. The most widely used procedures for TdT extraction involve treatment of separated cells with buffer containing high concentrations of salt and nonionic de- tergents, followed by ultracentrifugation to obtain a 100,000 x g supernatant (S- 100). The S-100 is either directly assayed for TdT content (Greenwood et al. 1977) or is further fractionated by phosphocellulose chromatography (Marcus et al. 1976; Sarin et al. 1976; Silverstone et al. 1976). Assay of the S-100 fraction requires the use of very high concentrations of both radioactive and non-radioactive substrate to overcome the effect of impurities in the enzyme fraction (Greenwood et al. 1977). A kinetic study for each enzyme determination is also required to achieve a reliable estimate of specific activity (Bentler and Kuhl 1978; Greenwood et al. 1977). The advantage of phosphocellulose chromatography is that a partially puri- fied enzyme fraction is obtained which is largely free of nucleotides, DNA polymerases, phosphatases, and nucleases. With our procedure, several estimates of TdT can be carried out simultaneously (20 or more samples/day) and results are available within 24h. Furthermore, any laboratory with an access to a liquid scintilla- tion counter can apply this protocol since most, if not all, equipment required is available in a standard clinical biochemistry laboratory and all reagents are available commercially.

Although this assay allows TdT determination over a broad range of cell numbers, it is recommended that analysis of clinical samples be carried out at cell concentrations of 1-2 x 107/ml to keep the protein concentration constant. Since at this concentration only a minimum of 1.5 x 106 cells are required for the duplicate assays and duplicate controls in the presence of ATP, most routinely provided 0.5- 1.5 ml bone marow and 5-10 ml blood, samples can be tested even in leukopenic situations. A comparative study of TdT activities obtained by our procedure with the one that utilizes high substrate concentration and crude enzyme (S-10 or S-100 actions) clearly demonstrates the advantage of the partial purification step, partic- ularly with small number of cells (Table 2). It may also be pointed out here that the amount of radioactive dGTP used in our assay is at least one order of magnitude lower than that used for the assay of crude enzyme which requires 1 mM dGTP concentration of sufficiently high specific activity. Since the Km for dGTP under our assay condition (with Mn 2 + as an effective divalent cation) is 20-25 gM, use of 50 gM dGTP allows optimal incorporation rates. Furthermore, absolute radioactivity numbers (cpm incorporation) obtained in our assay system when compared to the conventional assay are much higher (Table 2) due to the higher specific activity (3-to 4-fold) of dGTP used, although total dGTP concentration is 20-fold lower. Levels of TdT activity observed ranged from 0.010 to 86.2 unit/108


cells, demonstrating the sensitivty of this assay over four orders of magnitude. The lower limit of detection was approximately 0.05 units/108 cells when 1.5 • 106 cells were assayed.

Differences in activities between diagnostic categories always exceeded one order of magnitude in bone marrow or blood samples. This makes false positive or false negative results highly unlikely. The absolute specific activities obtained with this assay are slightly lower than with the previously used ion filtration chromatography technique (Mertelsmann et al. 1978), but there were no qualitative differences between the two techniques. The ease, speed and reproducibility of the phosphocellulose micromethod makes this assay superior for most applications. Differ- ences between specific activities reported by other investigators (Coleman et al. 1976, 1978; Kung et al. 1978; Penit et al. 1977) and those determined in this laboratory (Konziner et al. 1977, 1978; Mertelsmann et al. 1978a, b; 1979a, b) have been discussed previously (Mertelsmann et al. 1978a). Although it has been claimed that certain extraction and test conditions are significantly superior to others (Cole- man 1977), this claim was in part based on the combination of incompatible extraction and assay conditions. Nevertheless, final conclusions drawn from individual laboratories with individual sets of standards appear to yield qualita- tively very similar results (Table 4).

Analyses of normal human tissues confirmed that TdT activity is restricted to thymocytes and to a T-cell related subpopualtion of bone marrow mononuclear cells. Enriched myeloid, erythroid, megakaryocytic, and putative pluripotent stem cell fractions from human, primate, or rodent sources (unpublished) as well as mouse pre-B cells (Paige et al. 1978) did not exhibit detectable leels of TdT activity.

While all myeloproliferative syndromes (MPS: polycythemia vera, chronic myeloid leukemia, CML) and myelodysplastic syndromes (MDS: refractory anemia with or without an excess of blasts, CMMOL) were found to be TdT negative during the chronic phase of their disease, approximately 30~ of blast phase CML and 15~ of other MDS and MPS terminating in an acute leukemia were found to exhibit high levels of TdT activity (Table 4). It is of clinical significance that the great majority of TdT + acute leukemias preceded by MPS or MDS have responded by going into complete remission upon treatment with vincristine and prednisone (data not shown, Mertelsmann et al. 1979b; Srivastava et al. 1978; Vogler et al. 1978).

In acute leukemias, highest levels of TdT activity were present in more than 90~o of "non-T, non B" and T cell ALL, while only 2~ of ALL cases were of TdT negative, T-cell type. The later cases, probably represent leukemic of lymphoma- tous proliferation of a more mature T cell. Recently a previously unrecognized ALL phenotype exhibiting intracytoplasmic IgM ("pre B" phenotype) has been described, which in some cases was found also to be associated with high levels of TdT activity (Shaw et al. 1978; Vogler et al. 1978). The majority of cases with AML were found to be TdT negative while approximately 10~ of cases with a morphological diagnosis of AML, also exhibited high levels of TdT activity. More exten- sive phenotypic analysis revealed evidence of an involvement of more than one cell lineage in these leukemias (Mertelsmann et al. 1978b).

Based on our own observations and those described in the literature (Table 4) a consistent pattern for TdT distribution in acute leukemias has emerged. Because

O

o m"

O

e~

Tab

le 4

. Sta

tus

of T

dT i

n ac

ute

leuk

emia

, re

view

of

the

lite

ratu

re a

nd o

ur d

ata

Cli

nica

l dia

gnos

is

Sou

rce

of in

form

atio

n (N

umbe

r of

TdT

+ c

ases

+ T

otal

num

ber

stud

ied)

a b

c d

e f

g h

Tot

al

(% T

dT

+)

AL

L,

all

phen

otyp

es

73/7

7 62

/65

25/2

8 37

/40

21/2

3 8/

8

4/6

81/

86

311/

333

(94%

) A

LL

, non

-B, n

on-T

30

/33

12/1

4 29

/31

14/1

6 6/

6

62/

62

153/

162

(94%

) A

LL

, T

14/1

4 11

/12

8/ 9

7/

7

2/ 2

2/

2 19

/ 21

63

/ 67

(9

4%)

AN

LL

2/

55

0/22

0/

17

2/ 3

3/

27

0/3

14/1

08

21/2

33

(9%

) ~"

A

UL

6/

10

8/17

10

/17

1/I

25/

45

(55%

) C

ML

, bl

asti

c 24

/72

2/ 5

11

/23

4/5

16/

44

57/1

49

(38%

)

The

fol

low

ing

refe

renc

es h

ave

serv

ed a

s a

sour

ce o

f inf

orm

atio

n su

mm

ariz

ed i

n th

is t

able

, a:

Kun

g et

al.

1978

; b:

Col

eman

et

al.

1976

, 197

8; c

: P

enit

et

al.

1977

; d:

Hof

fbra

nd e

t al

. 19

77; e

: S

riva

stav

a et

al.

1978

; f:

Gor

don

et a

l. 19

78;

g: S

arin

et

al.

1976

, and

h:

our

data


of the diagnostic and therapeutic implications, we recommend that TdT activities be determined in all cases of acute leukemia at diagnosis and at relapse. Further studies are required to determine the clinical value of following the remission status in ALL, LBL or other TdT positive neoplasias (Mertelsmann et al. 1978a). With the methodology described here the TdT assay should become more widely available and thus contribute to improved patient management and to new insights into the pathophysiology of human leukemia.

Acknowledgement. We appreciate the cooperation of the physicians of Memorial Hospital and the skillful technical assistance of Ms. L. Barnett, Ms. Bhatt and Ms. I. Mertelsmann.

Abbreviations

ALL = acute lymphoblastic leukemia; ANLL = acute non-lymphoblastic leukemia; ATP = adenosine triphosphate; AUL = acute undifferentiated leukemia; B cell = bone marrow related lymphocyte; CFU-c = Colony-forming unit in soft agar culture (myeloid committed stem cell); CM: = chronic myelogenous leukeima; E-rosette = sheep erythrocyte rosette; (3H)dGTP = tritium-labeled deoxyguanosine triphosphate, incorporated into nucleic acid as monophosphate: (3H) dGMP; HTLA = human T-lymphocyte antigen; MDS = myelodysplastic syndrome; MPS = myeloproliferative syndrome; MPS-AP, in acute or blastic phase; Oligo d(pA)~2-~8 = polymer of deoxyadenylic acid, chain length I2-18 residues; PB-T = peripheral blood T cell; sIg = surface immunoglobulins; T cell = thymus-related lymphocyte; TdT = terminal deoxynucleotidyl transferase; TdT = specific TdT activity in bone marrow mononuclear cells <0.10 units/108 cells; TdT+, TdT activity>0.10 units/108 cells; thy-T = thymocyte

References Baltimore D (1974) Is terminal deoxynucleotidyl transferase a somatic mutagen in lymphocytes? Nature

248:409-411 Barr RD, Satin PS, Perry SM (1976) Terminal transferase in human bone-marrow lymphocytes. Lancet

1:508-509 Bennett JM, Catovsky D, Daniel M-T,,Glandrin G, Galton DAG, Gralniek HR, Sultan C (1976) Pro-

posals for the classification of the acute leukemias. Br J Haematol 33:451:458 Beutler E, Kuhl W (1978) An assay for terminal deoxynucleotidyl transferase in leukocytes and bone

marrow. Am J Clin Pathal 70:733-737 Bhalla RB, Schwartz, MK, Modak MJ (1977) Selective inhibition of terminal deoxynucleotidyl trans-

ferase (TdT) by adenosine ribonucleotide triphosphate (ATP) and its application in the detection of TdT in human leukemia. Biochem Biophys Res Commun 76:1056 1061

Bollum FJ (1974) Terminal deoxynucleotidyl transferase. In Boyer PD (ed) The Enzymes, Vol 10. Aca- demic Press, New York, pp 145-171

Bollum FJ (1975) Terminal deoxynucleotidyl transferase: Source of immunological diversity? In: Zahu RK (ed) Karl-August-Forster Lectures, Vol 14. Franz Steiner Verlag, Wiesbaden, S 1-47

Bollum FJ (1978) Terminal deoxynucleotidyl transferase: Biological studies. Adv. Enzymol. 37:347-374 Boyum A (1968) Isolation of mononuclear cells and granulocytes from human blood. Scand J Clin Lab

Invest 21 [-Suppl 97]:77-89 Burgess RR (1969) Purification and properties of DNA-dependent RNA polymerase from Excherichia

coli. J Biol Chem 244:6160-6168 Coleman MS, Greenwood MF, Hutto J J, Bollum FJ, Lampkin B, Holland P (1976) Serial observation

on terminal deoxynucleotidyl transferase activity and lymphoblast surface markers in acute Iym- phoblastic leukemia. Cancer Res 36: 36:120

Coleman MS, Greenwood MF, Hutton JJ, Holland P, Lampkin B, Krill C, Kastelic JE (1978) Adenosine deaminase, terminal deoxynucleotidyl transferase (TdT), and cell surface markers in childhood acute leukemia. Blood 52:1125-1131

Coleman MS, Hutton JJ, DeSimone P, Bollum FJ (1974) Terminal deoxyribonucleotidyl transferase in human leukemia. Proc Natl Sci USA 71:4404-4408


Coleman MS (1977) Terminal deoxynucleotidyl transferase: Characterization of extraction and assay conditions from human and calf tissue. Arch Biochem Biophys 182:525-532

Donlon JA, Jaffe ES, Braylan RC (1977) Terminal deoxynucleotidyl transferase activity in malignant lymphomas. N Engl J Med 297:461-464

Filippa CA, Lieberman PH, Eflandson RA, Koziner B, Siegal FP, Turnbull A, Zimring A, Good RA (1978) A study of malignant lymphomas using light- and ultramicroscopic, cytochemical, and im- munogical techniques. Correlation with clinical features. Am J Med 64:259-268

Gordon DS, Hutton JJ, Smalley RV, Meyer LM, Vogler WR (1978) Terminal deoxynucleotidyl transferase (TdT), cytochemistry, and membrane receptors in adult acute leukemia. Blood 52:107%1088

Greenwood MF, Coleman MS, Hutton J J, Lampkin B, Krill C, Bollum F J, Holland P (1977) Terminal deoxynucleotidyl transferase distribution in neoplastic and hematopoietic cells. J Clin Invest 59:889-899

Hoffbrand AV, Ganeshaguru K, Janossy G, Greaves MG, Catovsky D, WoodruffRD (1977) Terminal deoxynucleotidyl transferase levels and membrane phenotypes in diagnosis of acute leukemia. Lancet 2:520-523

Hutton J J, Coleman MS (1976) Terminal deoxynucleotidyl transferase measurements in the differential diagnosis of adult leukemias. Br J Haematol 34:447-456

Hutton J J, Bollum FJ (1977) Terminal deoxynueleotidyl transferase is present in athymic nude mice. Nucleic Acids Res 4:457-460

Hutton JJ, Bollum FJ (1978) Terminal transferase and DNA polymerase in leukemia. In: Rudon RW (ed) Biological markers of Neoplasia. Elsevier/North-Holland, New York

Koziner B, Filippa DA, Mertelsmann R, Gupta S, Clarkson B, Good RA, Siegal FP (1977) Charac- terization of Malignant Lymphomas in Leukemic Phase by Multiple Differentiation Markers of Mononuclear Cells. Am J Med 63:556--567

Koziner B, Mertelsmann R, Filippa DA (1978) Adult lymphoid neoplasias ofT and null cell types. In: Clarkson B, Marks PA, Till J (eds) Differentation of Normal and Neoplastic Hematopoietic Cells. Conferences on Cell Proliferation Vol. 5, pp 843-857

Kung PC, Silverstone AE, McCaffrey RP, Baltimore D (1975) Murine terminal deoxynucleotidyl transferase: Cellular distribution and response to cortisone. J Exp Med 141:855-865

Kung PC, Long JC, McCaffrey RP, Ratlif RL, Harrison TA, Baltimore D (1978) Terminal deoxynucleotidyl transferase in the diagnosis of leukemia and malignant lymphoma. Am J Med 64:788-794

Marcus SL, Smith SW, Jarowski CE, Modak MJ (1976) Terminal deoxyribonucleotidyl transferase activity in acute und undifferentiated leukemia. Biochem Biophys Res Commun 70:37-44

Marks SM, Baltimore D, McCaffrey RP (1978) Terminal transferase as a predictor of initial respon- siveness to vincristine and prednisone in blastic chronic myeolgenous leukemia. N Engl J Med 298:81~814

McCaffrey R, Smoler DF, Baltimore D (1973) Terminal deoxynucleotidyl transferase in a case of childhood acute 13nnphoblastic leukemia, Proc Natl Acad Sci USA 70:521-525

Mertelsmann R, Koziner B, Filippa DA, Grossbard E, Incefy G, Moore MAS, Clarkson BD (1979a) Clinical significance of TdT, cell surface markers and CFU-c in 297 patients with hematopoietic neoplasias. In: Neith R, Gallo RC, Hofschneider PH, Manweiler K (eds) Modern Trends in Hu- man Leukemia III, Springer, Berlin Heidelberg New York, pp 131-138

Mertelsmann R, Mertelsmann I, Koziner B, Moore MAS, Clarkson BD (1978a) Improved biochemical assay for terminal deoxynucleotidyl transferase in human blood cells: Results in 89 adult patients with lymphoid neoplasias and malignant lymphomas in leukemic phase. Leuk Res 2:57-69

Mertelsmann R, Koziner B, Ralph P, Filippa DA, McKenzie S, Arlin ZA, Gee TS, Moore MAS, Clark- son BD (1978b) Evidence for distinct lymphocytic and monocytic populations in a patient with terminal transferase-positive acute leukemia. Blood 51:1051-1056

Mertelsmann R, Moore MAS, Clarkson BD (1979b) Sequential marrow culture studies and terminal deoxynucleotidyl transferase activities in myelodysplastic syndromes. In: Schmalzl F, HeUriegel KP (eds), Preleukemia. Springer, Berlin Heidelberg New York, pp 106-117

Meyskens FL Jr, Jones SE (1978) Terminal transferase: Its evolving role (Editorial) N Engl J Med 298:945-946

Miller RG, Phillips RA (1969) Separation of cells by velocity sedementation. J Cell Physiol 73:191-202 Modak M J, Gupta S, Good RA (1978) Terminal deoxynucleotidyl transferase in human tonsils. Lancet

2:481~482

104 M,J. Modak et al.

Modak MJ (1978) Biochemistry of terminal deoxynucleotidyl transferase: Mechanism of inhibition by adenosine 5'-triphosphate. Biochemistry 17:3116-3120

Modak MJ (1979) Biochemistry of terminal deoxynucleotidyl transferase: Mechanism of Managenese- dependent inhibition of TdT by deoxyadenosine Y-triphosphate and biological implications. Bio- chemistry 18:2679-2685

Modak MJ, Marcus SL (1977) Purification and properties of Rauscher leukenia virus DNA polymerase. J Biol Chem 252:1 I- t9

Moore MAS (1975a) Marrow Culture - a new approach to classification of leukemias. Blood Cells 1:149

Moore MAS (1976) Prediction of relapse and remission in AML by marrow culture criteria. Blood Cells 2:10%124

Moore MAS (1975b) Agar culture studies in CML and blastic transformation. Sem Haemat 8:11-27 Paige C J, Kincade RW, Ralph P (1978) Mutine B cell leukemia line with inducible surface immunoglob-

ulin expresion. J Immunol 121:641-647 Penit C, Brouet JC, Rouget P (1977) Terminal Deoxynucleotidyl Transferase in Acute LymPhoblastic

Leukemias and Chronic T Cell Proliferations. Leukemia Res 1:345~50 Phillips RA, Miller RG (1970) Seperation of hematopoietic stem ceils from cells causing graft versus

host disease. J Immunol 105:1168-1176 Satin PS, Anderson PN, Gallo RC (1976) Terminal deoxynucleotidyl transferase activities in human

blood leukocytes and lymphoblast cell lines: High levels in lymphoblast cell lines and in blast cells of some patients with chronic myelogenous leukemia in acute phase. Blood 47:11-20

Shaw MT, Dwyer JM, Allaudeen HS, Weitzmann HA (1978) Terminal deoxyribonucleotidyl transferase activity in B cell acute lymphocytic leukemia~ Blood 51:181-187

Silverstone AE, Cantor H, Goldstein G, Baltimore D (1976) Terminal deoxynucleotidyl transferase is found in prothymocytes. J Exp Med 144:543-548

Srivastava BIS, Khan SA, Monowada J, Freeman A (1978) High terminal deoxy-nucleotidyl transferase activity in pediatric patients with acute lymphocytic and acute myelocytic leukemias. Int J Can- cer 22:4-9

Sujimoto M, Bollum FJ (1979) Terminal deoxynucleotidyl transferase (TdT) in chick embryo lymphoid tissues. J Immunol 122:392-397

Vogler LB, Christ, WM, Bockman DE, Pearl ER, Lawton AR, Cooper MD (1978) Pre-B-cell leukemia. A new phenotype of childhood lyphoblastic leukemia. N Engl J Med 298:872-878

Received April 15, 1980/Accepted July 14, 1980

Documents

A micromethod for determination of terminal deoxynucleotidyl transferase (TdT) in the diagnostic evaluation of acute leukemias