A Monoclonal Antibody, H2, Defines a New Surface Antigen Expressed on Human Lymphocytes

Scand. J. Immunol. 30, 573-581. 1989

A Monoclonal Antibody, H2, Defines a New Surface Antigen Expressed on Human Lymphocytes J. G R U N E W A L D , C. H. J A N S O N . M. J. T E H R A N I . A. P O R W I T . Y. MATSUO, H. M E L L S T E D T & H. W I G Z E L L Department of Immunology. Karolinska Institute; Departments of Pathology and Oncology, Karolinska Hospital. Stockholm, Sweden and Fujisaki Cell Center. Hayashibara Biochemical Laboratories, Inc.. Okayama. Japan

Grunewald. J., Janson, C.H.. Tehrani. M J . . Porwil. A.. Malsuo. Y.. Mellstedt. H. & Wigzell. H. A. Monoclonal Aniihudy, H2. Defmcs a New Surface Antigen E.sprcsscd on Human Lymphocytes. Scand. J. Immunol. 30. 573-581. 1989 We have previously described a monoclonal antibody (MoAb). H2, which recognized a tumourunique antigen on a human T-ccll chronic lymphaiic leukaemia (T-CLL. CD3.4'). However, further characterization of H2 has revealed a reactivity with the majority of T lymphocytes and a minority of B lymphocytes, some malignant T cells and a few cell lines of leukaemia or of hematopoietic tumour origin. The molecular weight of the antigen (80,000) precipitated by the MoAb H2 from Ihe cell lines NALM-6 and Reh corresponded lo ihat previously found, When PBL were slimulaled with PHA, lL-2, or Con A a reduced reactivity of H2 could be seen. The MoAb H2 was submitted to Ihc Fourth International Conference on Human Leucocyte Differentiation Antigens. Vienna, 1989. H2 did not cluster in any of the 7H clusters of differentiation (CD I 78) discussed at the conference, indicating its unique reactivity. This suggests that we have defined a new antigen on lymphocytes with a possible role along the restingprotiferating axis. Carl Harald Janson, MD. Department of fmmunology. Karolinska Institute. Box 60400. S-104 01 Stockholm. Sweden

In the process of producing anti-idiotypic monoclonal antibodies (MoAb) for therapy of T-cell malignancies, we identified not only MoAb reactive with the variable part of the T-cell receptor for antigen, but also MoAb H2 and 1VF7 reactive with two other cell membrane structures [5]. This reactivity was initially considered tumourunique, because H2 and IVF7 only sremed to react with the tumour cells against which they were produced. However, further characterization of MoAb H2 and 1VF7 by flow cylometry revealed that they also stained subpopulations of lymph node and peripheral blood lymphocytes and a few cell lines of leukaemia or hematopoietic tumour origin. In this paper we describe the reactivity pattern of H2, and argue that it recognizes a new antigen on the cell membrane of human lymphocytes with a possible role along the activation pathway.

MATERIALS AND

METHODS

Monoclonal antibody H2. The production and selection of the MoAb H2 has previously been described [5]. Briefly, tumour cells from a patient with a T-ccll chronic lymphatic leukaemia were separated and used for immunization of BALB/c mice. Spleen cells were fused with Sp 2/0 myeloma cells, and hybridomas producing antibodies with a tumour-unique reactivity were selected using a micropanning technique [9] and indirect immunofluorcscence. MoAh H2 was of IgG t isotype. An IgG2a spontaneous switch variant was selected by repeated hmiting dilution and an lgG2a-subclass-specific ELISA. The MoAb H2 was produced as hybridoma culture supernatant or a.scites and puriHed over a protein A-Sepharose column (Pharmacia. Uppsala. Sweden). F(ab'); fragments of H2 were produced by pepsin digestion [14]. Biotinylated MoAb H2 was produced according to standard procedures (14]. Preparation of celh. Peripheral blood mononuclear cells were obtained from healthy blood donors by a Ficoll-Hypaque (Pharmacia) gradient ccntrifugation [ I ]. The cells were washed and resuspended in complete

573

574

J, Grunewald et al.

medium, consisting of RPM I l640(Gibco. Paisley. UK) v:ontaimng5%fela]cairserum(FCS)(Gibco). 198 U/ml penicillin (Sigma. Si Louis, Mo.. USA), 69 U/ml streplomycin (Sigma). 2 mM i.-glulamale (Sigma). Macrophagcs and monocyles were enriched by plasticadherence by incubuting lO" periphcrnl blood lymphocylcs (PBL) in 10 ml medium on plaslic Petri dishes ((^ = 90 mm) lor 60 min al .17 C. Non-adherenl cells were enriched for T cells by the E rosetling technique [6]. Natural killer (NK) cells were purified by fractionation ofFicolt Hypaque-scparaled lymphocytes on a discontinuous Percolf gradient (Pharmacia) essentially as described (13]- Granulocytes were obtained from the erythrocyle fraction after Kicoll-Uypaque gradient centrifugation by lysing ihe erythrocyies wiih hypotonic shock [14]. PBL were incubated at a concentration of 0.5 X 10^/ml in medium containing 5".i. hcat-inaciivated (56 C for 30 min) autoiognus scrum with 2 /ig/ml phytohacmagglulinin (PHA) (Sigma). 10 U/ml recombinant inlerleukin 2 (rlL-2) (Du Pont. Glenolden. Pa.. USA). 5/ig/mlconcanavalin A(Con AKSigma). or 100 U/ml gamma inlerferon (IFN-)')(Gcnentech.San Francisco. Calif. USA) at 37 C in 6% CO;. The cells were analysed by indirect immunofluorescence at time 0 and after 8. 24 and 72 h. The analysis was carried oui with F(ab'): fragments of H2 in order not lo interfere wiih ihe background which usually is noted on activated cells due ID I'c-rcceplor expression, Mononuclear ceils were separated from tonsils by Ficoil-Hypaque gradient centrifugation Macrophages and T cells were removed by plastic adherence and sheep red blood cell (SRBC)-rosetting. respectively. Remaining cells were cenlrifuged on a discontinuous Percoll gradient and two B-cell populations were collected, a high density population (small B cells) and a low density population (large B cells). B lymphocytes were also prepared from PBL and enriched by the same method as used for B-cell preparation from tonsil. PBL from healthy blood donors with antibodies to cyiomcgalovirus (CMV) were separated. T cells were removed by SRBC-rosetting. Supernatant from mycoplasma-free B 95-8 cells | l l ] was added to the nonrosetting cells and incubated overnight at 37 C in 6";. CO>. Epstein Barr virus (EBV) transformed cells were cloned and clones producing antibodies against CMV were selected by ELISA technique. Tbymocyles were obtained by resuspending thymus biopsies from children undergomg cardiac surgery. Bone marrow cells were obtained from healthy donors. Umbilical cord blood was taken at partus. The respective mononuclear cell fractions were separated by Ficoil-Hypaque gradient cenlrifugation. Cell line.-i. Leukaemia and haematopoietic tumourderived cell lines (Table I) were analysed for ihe reactivity with H2 by indirect immunofluorescence. Immunofluorescente siaining and flow cytomelric analysis. Staining of single-ceil suspensions was carried oul by indirect immunofluorcsccnce (IFL), using fluorescein isothiocyanate (FITC)-conjugated F(ab'): fragments of rabbit anti-mouse Ig (Dakopatts. Glostrup. Denmark) in the second step. The following MoAb were used in the lirststep: Leu-M.l. Leu-2(CD8). Leu-3 (CD4J, Uu-4(CD3), Lcu-ll tCD16j, Leu-I6tCD20J

(Becton-Dickinson. Mountain View. Calif.. USA) and HB95 (American Type Culture Collection. Rockvillc. Md.. USA). In double fluoresccna' siaining. FITC or phycoerythrin (PE)-conjugated Leu-M.3. Lcu-2. Leu-3, Leu-4, Leu-11, Leu-16 were used. Biolinylated MoAb H2 was used together with a strcptavidin FITC or streptavidin PE conjugate. Fluorescent staining was analysed with a FACS-Star (low cytometer (BectonDickinson. Sunnyvale. Calif.. USA) and Hewlett Packard 300 computer (Palo Alto. Calif.. USA). 10* cells were analysed, and side and forward scatter was used lo gate different cell populations. Dead ceii populations stained with propidium-iodide were gated out. Normal mou.sc serum (NMS) or biotinylaled NMS diluted i :50() in PBS were used as negative controls. F(ab')2 fragments of an irrelevant antibody was used as a negative control together with the H2 F(ab ); fragments. In double staining, due to weak reactivity with MoAb H2. optimal electronic compensation for fluorescence one (FITC) and two (PB) was set. Biop.ue.s. Lymph node or skin biopsies, taken for diagnostic purposes, were in parl fixed in B5 fixative for routine histopathology. in part frozen in liquid nitrogen and then stored at - 7 0 C. Biopsies of normal heart, brain, kidney and stomach tissue were taken at autopsy, performed not later than 8 h after death, and frozen in liquid nitrogen. Mononuclear white blood cells from patients with leukaemia were obtained by FicoilHypaque gradient centrifugation and stored in liquid nitrogen until used. The Kiel classification was used

l!0|. Intnmmistainings. The immunophenotype of lymphomas was determined by immunofiuorescence staining of cell suspensions as described previously [ 12]. or immunohistochemical staining of fro/en sections. For the latter, 6-mm cryosccUons were cui, fixed in cold acetone and stored at - 2 0 C. The alkahne-phosphatase antialkali ne-pho.sphatase method [3| was used. In brief, the sections were incubated with appropriately diluted MoAb (30 min. 20 C) and then with rabbit anti-mouse serum (Dakopalts) (1:40, 15 min, 20 C), followed by the APAAP complex (Dakopatts) (1:20, 15 min. 20 C). The last two steps were repeated to enhance the reaction. The colour enzymatic product was developed wiih Vector Alkaline Phosphatase Kit I (red) or 111 (blue) (Vector Lab.. Burlingame, Calif., USA). For all MoAb the dilution series were performed to determine the best staining conditions. Either the IgGl or IgG2a isotype (spontaneous switch variant) of MoAb H2 was used for the slainings. In double stainings the procedure was repeated with another MoAb (anti-Leu3 or «nti-Leu2 (Becton-Dickinson)) and an alternative kit to obtain a different enzymatic reaction colour product. The control procedures included slainings with omission of the first and second moAb. Measuring NK cell acliviiy. Cytoloxic assay was performed according to standard procedures [14]. Briefly. PBL were passed over nylon wool [14]. washed twice, and used as effector cells. The cell lines K562 or U937 were used as target cells, labelled with 100 ;iCi of Na:^'CrOj for 1.5 b at 37 C and 6",. CO;, washed twice, stained with trypan blue, and adjusted to the cell concentration giving the desired effector to target cell

MoAb H2 Defines New Surface Antigen ratios. All assays were done in triplicate. Cells were incubated in V-bottomcd microtiire plates (Costar. Cambridge, Ma.ss.. USA) at 37 C and 6% COj for 4 h, plates spun at 190 g and 100 /il of the supernatant counted on a -/-counter. Target cells incubated in 0.05% NP40 or complete medium served as controls for maximum and spontaneous release, respectively, To analyse the effecl of MoAb H2 on NK killing, ihe effector cells were cither pre-incubaled with ihe antibody at a concentration of 10 /jg/ml. washed twice, or the antibody was added at the start of the assay. As negative control NMS or an irrelevant antibody of the same isotype was used. Immunoprecipitation of cell-surface antigens. The relative molecular weights of the cell-surface molecules were determined by immunoprecipitation of lysates from '-''l-labelled cell lines of haematopoietic origin. Reh (non-T non-B-ALL) or NALM-6 (pre-B-ALL), using 7 ]5"-M gradient sodium dodecyi sulphate polyacrylamide gel electrophoresis (SDS PAGE) and autoradiography |5, 8]. NP-40 (0.5%) (Fiuka. Buciis. Switzerland) was used as detergent in the lysate bufler under reducing conditions.

RESULTS Reactivity of MoAb H2 with haematopoietic cells The MoAb H2 reacted with approximately 6O'^i of Ficoll-Hypaquc-separated lytnphocytegated PBL (Fig. la). 75% of CD3*, 90% of CD4*, atid 50% of CD8* lytnphocytes (Fig. 2). In one representative double staining experiment, carried out with lymphocyte-gated PBL, 93% of H2* cells were CD3^ (Fig. 2a), 68% were CD4* (Fig. 2b). and 20% were CD8* (Fig. 2c). Of the H2* cells 3% were CD20* (data not shown). Analysing B-cell-enriched populations from tonsils (Fig. 3a. b) or PBL (Fig. 3c), a low reactivity of 5-! 0% with H2 was found. Small B cells from tonsils showed a reactivity of 10-15'i^.. large B cells 5% and among B-cell-enriched PBL a 510% reactivity was seen (Fig. 3). EBV-transformed B cells from a normal donor (Fig. I h) and bone marrow cells (Fig. ig) as well as cell populations enriched for granulocytcs (Fig. Ib). NK cells (Fig. lc), or monocytes (Fig. Id), were all negative in their reactivity with H2. Of lymphocyte-gated thymocytes 85''/ii were CD3* and 75% were H2* (Fig. ic). Of umbilical cord lymphocytes. 80% reacted with H2(Fig. If). Figs 1-3 are taken from representative experiments. Different cell lines of haematopoietic origin were screened for MoAb H2 reactivity (Table I). The T-cell lines HPB-MLT (TcR a/^+), DND-41

575

TABLE I. Reactivity of MoAb H2 with cell lines of leukaemia or haematopoietic tumour origin Cell type T-cell. CD3*.x/J ' y5HPB-ALL C D 4 * . C D 8 \ T-ALL CD4 '.CD8*.T-LY HPB-MLT CD4 ' . C D S ' . T-ALL JM TALL-I C D 4 VCD8'.T-ALL H9 CD4 ' . T A L L MOLT 16 C D 4 ' (weak). T-ALL C D 4 * (weak). T-ALL JURKAT CCRF-CEM C D 4 ' . T - A L L HUT 78 C D 4 '-SS T45 1301 TALL T-ccll. CD3*.a^ - , y 5 •< C D 4 *. T-ALL DND4I MOLT 13 T-ALL C D 4 •. T-ALL PEER T-cell. CD3.ct/J ~.y6 RPMI 8402 C D 4 * (weak). T-ALL MOLT 3 C D 4 '.CDS'.T-ALL C D 4 \ C D 8 + .T-ALL MOLT 4 CCRF-HSB2 KE37 C D 4 ' . T-ALL C D 4 '.CD8+.T-CLL SKW3 MOLT 15 A-MoL MT-1 ATL CD4 • . M F HUT iO2 CD4 C5/MJ T-cell malignancies C D 4 ' . T-CLL L.S. CD4 M-CLL E.L. CD8 \ T-CLL l.U. W. CD8 * (weak). T-ALL CD8 \ T-CLL N. B-cell EBV '.BL RAJI EBV *. BL DAUDI EBV . BL RAEL NAMALWA EBV ' . BL EBV '.BL P3H3 EBV ' , BL BL72 EBV . BL RAMOS EBV . BL BL28 EBV , BL BL41 EBV transf. B cells LSBI EBV transf. B cells ESBI NAD 20 B-LCL NALM 6 B-ALL KOPN-K B-ALL BAL-KH B-ALL 123 EBV transf. B cells 223 EBV transf. B cells Non-T- non-B-ccU

H2 reactivity

— _ — _ _ _ -

—

— — _ — _

+ — -f (weak) -

_

+ (weak) -t- (weak) _ -f (weak) — — _ _ _

-^ -1— -

KM3

REH NALM 16 Erythroblastoid K 562 Myelomonocytic U937P U937cl I.cl2 .cI4. ci \b HL60

— — _ -

576

J, Grunewald et ai (b)

•-. ( a )

( c)

(d)

Ml

M tail

, flllUll

111 m i l

i iilmil—L .1 J.ilUU

i l l iiii«

(h)

I I IIIHll

I I I tlHH

Fluorocanc* Inlaniity (log)

Fici. I. FACS profiles of PBL (a), granulocytes (b), NK cells (c). monocylcs (d), thymocytes (e). umbilical cord mononuclear cells (f), bone marrow mononuclear ceils (g). and EBV-transformcd B cells (h) slained wiih MoAb H2 i •), positive control (• • • J, and normal mouse scrum ( ...).

(TcR 7/