» antibodies, and to the detection of activated B-cells.

B-lymphocytes are involved in the humoral, or circulating, immune response to the extracellular phases of bacterial and viral infection. When activated by exposure to specific antigenic or mitogenic stimuli, human B-lymphocytes proliferate and subsequently differentiate into antibody secreting cells. During the stages of activation, B-cells increase in size, synthesize both DNA and RNA, and undergo changes in cell surface structures. Neoplastic changes in B-cells, e.g., leukemias or lymphomas of B-cell origin, also induce changes in the expression of cell surface antigens.

Thorley-Lawson et al., (1982) Cell 30x415-425 describe a B-cell restricted antigen (B-LAST-1) which appears on the surface of B-lymphocytes 2-3 days after stimulation with mitogens or Epstein-Barr Virus (EBV) . B-LAST-1 is a single chain polypeptide with a molecular weight of 45 kd. B-LAST-1 is also found on chronic lymphocytic leukemia (CLL) cells and poorly differentiated lymphoma cells. The isotype of anti-B-LAST-1- monoclonal antibody of murine origin was reported to be IgG2b.

Yokochi et al., (1982) J. Immunol. 128:823-827 report a B-cell antigen, BB-1, which appears on the surface of EBV and mitogen-activated human and some non-human primate B-cells. The molecular weight of the BB-1 antigen is 37,000. Expression was detected.4-5 days post-EBV-activation and peaked at 7 days. BB-1 was also found on 50% of the myelomas tested.

Summary of the Invention The present invention provides a monoclonal antibody recognizing an antigenic determinant on activated human B-cells, the antigen being characterized in that it is a protein distinct from B-LAST-1 (described in Thorley-Lawson et al., (1982) Cell

3 :415-425) and from BB-1 (described in Yokochi et al. , (1982) J.Immunol 128.823-827) . The antibody is substantially unreactive with unstimulated human B-cells. The present invention also provides a substantially pure protein having an antigenic determinant or determinants substantially identical to determinants of a single-chain polypeptide having an apparent molecular weight of approximately 75,000 daltons under reducing conditions and 67,000 daltons under non-reducing conditions, the single-chain polypeptide being a protein on the surface of activated human B-cells.

In another aspect the invention provides a process for preparing the antigenic protein which comprises culturing peripheral or splenic human B-cells activated with Epstein-Barr virus, pokeweed mitogen, protein A, or anti-immunoglobulin, or culturing Burkitt's lymphoma cell line Ramos or the plasma cell leukemia line RPMI 8226, and isolating -the protein from these cells.

The invention also provides kits useful for assaying a biological sample for the presence of cells expressing the antigen of the invention and for assaying a biological sample for the presence of antibody to the cells expressing the antigen of the invention. These kits contain one or more containers, each holding separately detectably labeled or unlabeled antibody or antigen of the invention, and in another compartment, a means for detecting the formation of immunocomplexes.

Description of Preferred Embodiments In the drawings. Fig. 1 is a series of histograms showing the reactivity of anti-Bl and anti-B5 monoclonal antibody to splenic B-cells activated with anti-Ig antibody; Fig. 2 is a pair of histograms showing the reactivity of anti-Bl and anti-B5 monoclonal antibody to splenic B-cells before (A) and after activation with anti-Ig antibody (B) ;

Fig. 3 is a series of histograms showing the reactivity of anti-Bl and anti-B5 monoclonal antibody to splenic B-cells unactivated (A) and activated with anti-Ig antibody (B) , and unactivated monocytes (C) ; and

Fig. 4 is an SDS-PAGE characterization of labeled cell surface proteins immunoprecipitated with anti-B5 monoclonal antibody.

Preparation of Activated B-cells

Normal human splenic B-lymphocytes were cultured at 1.5 X 10 ⁶ cells/ml in RPMI 1640 supplemented with 10% FCS, 2mM glutamine, ImM sodium pyruvate in tissue culture flasks for 1, 3, and 6 days with four different stimuli. 1) Pokeweed mitogen (PWM) : at a final concentration of 1:300. For days, 1, 3, and 6 whole splenic mononuclear cells were used. 2) Anti-Ig: Affinity purified rabbit anti-human Ig was coupled to Affigel 702 beads, specificity was checked by testing Affigel 702 beads which has been conjugated to bovine serum albumin, anti-Bl, or anti-B2 antibody, none of which showed any B cell stimulation. For the day 1 and day 3 stimulations, anti-Ig beads were incubated with highly purified B cells which were obtained by lysing the E rosette negative fraction of splenic mononuclear cells with anti-Mol, anti-Mo2, anti-T4, and anti-T8 followed by complement. The day 6 stimulation utilized unfractionated splenic mononuclear cells. 3)

Protein A: Protein A was used at a final concentration of 10 g/ml. As described above, highly purified B cells were cultured for 1 and 3 days at a concentration of 1.5 X 10 /ml. The day 6 stimulation utilized unfractionated splenic mononuclear cells. 4) EBV: The E rosette negative fraction of splenic mononuclear cells were cultured with EBV (1:4 diluted supernatant from the EBV-producing marmoset cell line B955) for 1, 3, and 6 days. Prior to phenotypic analysis of all activated samples, the cells were harvested and lysed with anti-Mol, anti-Mo2, anti-T4, and anti-T8 followed by complement to clear monocytes and T-cells respectively and enrich the B-cell fraction from the samples. Preparation of Anti-B5 Monoclonal Antibody A 6 week old female BALB/C mouse was immunized i.p. with 5 X 10 cryopreserved B cell diffuse histiocytic lymphoma (DHL) cells in phosphate-buffered saline (PBS) . These tumor cells were of B cell origin in that they expressed monoclonal cell surface IgM, k, _as ell as the B cell associated antigens la, Bl, and

B4. In contrast, these tumor cells were unreactive with monoclonal antibodies directed against the common acute leukemia antigen (CALLA) ; T cell antigens T3, T4, T8, and Til; and the myeloid/monocyte antigens Mol, Mo2, and MY7. Twenty eight days later, the animal was boosted g with 5 X 10 tumor cells i.v. and somatic cell hybridization was carried out 4 days later by the method of Kohler and Milstein (Nature, (1977) 256:495) with modifications as described in Nadler et al. , J.Immunol. (1980) 125:570. Mouse splenocytes (1.5 X 10 ⁸ ) were fused with 30% polyethylene glycol and Dulbecco's MEM with 2 X 10 ⁷ P3/NSl/l-Ag4-l myeloma cells. After fusion, cells were cultured in aminopterin-containing medium at 37° in a 5% CO- humid atmosphere. Ten to 28

days later, approximately 300 macroscopic clones were identified, 125 of which were reactive with the immunizing DHL cells, measured by indirect immunofluorescence. In brief, 0.5 to 1 X 10 viable washed DHL cells were treated with 100 ul of supernatant from hybridoma cultures exhibiting growth, incubated at 4°C for 30 minutes and washed three times. The cells were then treated with 100 ul of 1:50 dilution of goat anti-mouse IgG and goat anti-mouse IgM conjugated with fluorescein isothiocyanate incubated at 4° for 30 minutes, washed three times, analyzed on an EPICS V cell sorter. The percent positive cells were determined using the EASY Immuno-program. Producer clones were then screened on a panel of fractionated peripheral blood and tumor cells. One hybrid clone, designated anti-B5 was found to react with the immunizing DHL cells, several other B cell DHL cell lines, but was unreactive with fractionated peripheral blood mononuclear cells. Hybrid clone anti-B5 was then subcloned three times by limiting dilution and passaged into BALB/C mice to produce a malignant ascites. Supernatant and ascites anti-B5 were shown to have a similar reactivity pattern by indirect immunofluorescence. The B5 ascites demonstrated reactivity with the immunizing DHL cells to a dilution of 1/20,000 that diminished to background at 1/50,000. The B5 antibody was determined to be of the murine IgM isotype. In all subsequent experiments, B5 ascites were used. Deposit

The anti-B5-producing hybridoma cell line, designated Hybridoma B5, has been deposited in the American Type Culture Collection, Rockville, MD, and given ATCC Accession No. HB 8716.

Expression of Activated B-cell Surface Antigen

As shown in Table I, prior to culturing, cells did not express B5. In the presence of media alone, approximately 10-15% of cells expressed B5 after three days of culture, although the viability of these cells was only 10-20%. When stimulated with Protein A, anti-Ig antibody or EBV as shown in Figure 1 (viability approximately 70-80%) about 10-15% of cells weakly expressed B5 after 1 day, while 65% of cells expressed the antigen more intensely by 3 days of culture. The level of expression of B5 by day 6 was similar to day 1, and by day 10 (in the presence of anti-Ig antibody) , B5 antigen expression was again background level. PWM did not appear to induce cells to express B5 as well as anti-Ig antibody, EBV, or protein A, with the number of cells being only two-fold over background.

TABLE "C EXPRESSION OF B5 AND Bl ANTICENS OH SPLENIC B CELLS STIMULATED IN VITSO*

Dav Anti ^g en - tested Z positive cells when stinula _t ed tn vitro

Media Protein A Anti-IgG E3V PWM '

Bl 80 80 80 B5 80 80 <5 <5 < 5 <5 <5

B • _B lC 74 _. 65 8 -0 8 «4-» 5 J6O

^B5 6 38 14 12 • is Bl ^* - - 4 8 go 80 69

^B5 10 43 68 64 25

⁶ Bl 46 48 57 71 ^• 31

^B5 . 2 5 16 24 4

*one of three experiments showing identical patterns of induction of B5 expression.

In order to further demonstrate that B5 was expressed on activated B cells, splenic mononuclear cells enriched for B cells were stimulated for 3 days with anti-Ig conjugated to beads. Viable cells were harvested and labelled with directly fluoresceinated anti-Bl and directly biotinylated anti-B5 developed with Texas-Red-Avidin, and then evaluated by dual laser flow cytometric analysis. As shown in Figure 2, panel A unstimulated cells only expressed the Bl antigen and failed to express B5. After three days in culture with anti-Ig, all cells expressing B5 co-expressed the Bl antigen.

B5 is expressed on cell lines and tumor cells of B-cell origin. As shown in Table 2, anti-B5 was reactive with cell lines of B lineage including all EBV transformed ly phoblastoid B cell lines, Burkitt's lymphoma lines, four of five DHL lines and the plasma cell leukemia cell line RPMI 8226. The non-T-cell ALL line Laz 221, and the CML blast crisis line Nalm 1, both known to be of early neoplastic B cell origin, were unreactive with anti-B5. No reactivity was found on T-cell lines or an Ia+ T-cell clone. These results as well as the lack of expression of B5 π^eloid cell lines HL-60, KG-1, U937, and Sϋ-DHL 1 and the erythroid cell line K562 indicate that anti-B5 has restricted reactivity to cells of B cell derivation.

TABLE 2.

The reactivity of anti-B5 with a variety of B-cell malignancies was next investigated (Table 3). This series of neoplasms represent stages of normal B cell differentiation. All of the non-T-cell acute lymphoblastic leukemias (ALL) tested were of ^" B-cell origin by expression of la and B4. None of these early neoplastic B-cells expressed B5. About half of the B cell chronic lymphocytic leukemias (CLL) and DHLs examined, expressed B5, with a smaller percentage of poorly differentiated lymphocytic lymphomas (PDL) expressing the antigen. With the lack of expression of B5 on the Waldenstrom's acroglobulinemia cells and myelomas examined, the expression of B5 antigen was limited to cells which correspond to the mid stages of normal B cell differentiation. Cells from patients with T cell derived ALL, CLL, and T cell non-Hodgkin's lymphoma, including lymphoblastic lymphoma and DHL were unreactive with anti-B5. The B5 antigen was also not expressed on cells from patients with acute myeloblastic ^" leukemia (AML) . These observations confirm the B cell specificity of B5 and indicate that B5 is expressed on populations of B lymphocytes in the mid stages of normal B cell differentiation.

TABLE 3

Disease iff of patient samples # of patients reactive with monoclonal antibody

B Cell B5 B4 Bl la T3

Non-T ALL 21 0

B-CLL 21

24 8

DHL 12 21 24 0

18 24

8 24

PDL-N 18 0

7 18

5 18

PDL-D 7 0

8 7

0 7 0

Haldenstrom'β 8

2 8

0 8

Myeloma 2 0

2 2

0 2 0 0

0 0 0

T Cell

ALL 8

, CLL 0 0 3 0 0 ND 8

NHL* 0 5 0 0 0 3 0 0 2 5

Myeloid

AML/AMMOL 12

12

* Waldenstrom ^' 6 macroglobulinemia

* Includes lymphoblastic lymphoma and T cell DHL

- 10 - Reactivity of Anti-B5 Monoclonal Antibody

The reactivity of anti-B5 to unactivated fractionated peripheral blood cells, and normal lymphoid and myeloid tissues was examined. Less than 1% of peripheral blood mononuclear cells (PBMC) isolated by Ficoll-Hypaque density sedimentation expressed the antigen, (Table 4) whereas they demonstrated significant reactivity with monoclonal antibodies directed against B-cell, T-cell, and monocyte antigens.

TABLE 4

REACTIVITY OF ANTI-B5 WITH RESTING LYMPHOID AND MYELOID CELLS

Cell ϋ of testβ of cells expressing antigen

•

B5 Bl ^' T3 Mol l

Peripheral blood

PBMC 4 1+ 1 5 + 2. 50 5 31 + 8 18 + 4

E+ (T) 6 1 + 1 1 + 1 90 5 8 + 3 2 + 1

E- nonadherant (B) 3 1 + 1 22 + 4 2 + 1 24 + 3 42 +3

Monocyte 3 1 + 1 2 + 1 9 + 3 54 + 6 71 + 5

Granulocyte 5 0+ 1 1 + 1 0+ 1 88 + 10 1 + 1

RBC 3 0 + 1 0 + 1 0 + 1 0 + 0 . 0 + 0

Platelet 3 0+ 1 0 + 1 0 + 1 0 + 1 0+ 1

- 11 -

T-cells isolated by E rosetting similarly lacked detectable B5 expression. The E rosette negative fraction containing B-cells, monocytes, and null cells was further enriched for B-cells by adherence. The B-cell-enriched PBMC were stained with directly fluoresceinated anti-Bl and directly biotin-conjugated anti-B5 developed with Avidin Texas Red. Utilizing dual laser flow cytometic analysis, cells were examined as shown in Figure 3, before (Panel A) and after 3 days of stimulation with anti-Ig antibody (Panel B) . As seen in Panel A very few dual labelled cells were observed, whereas in Panel B clearly demonstrated B1+B5+ cells could be detected. In addition, monocytes were stained with directly fluoresceinated anti-Mol and directly biotin-conjugated anti-B5 developed with Avidin Texas Red (Panel C) and similarly analyzed. Adherent monocytes were noted to weakly express B5 on 10-20% of cells analyzed. However, when monocytes were incubated in 10% human serum for 1 hour prior to phenotyping, no cells appeared to co-express B5 and Mol. Similarly granulocytes, RBC and platelet preparations lacked reactivity with anti-B5. The percentage of B5 bearing Ficoll-Hypaque mononuclear cells within lymphoid tissues is enumerated in Table 5.

- 12 -

TABLE 5 REACTIVITY OF ANTI-B5.WITH LYMPHOID TISSUES * ^of tests ^. Z of cells reactive with

L ^y mph node 4

Spleen (whole) 7

Spleen (E-) . 3

Tonsil 3 .

Thy us 3

Bone marrow 3 .

- - Mononuclear cells isolated from normal lymph node, tonsil, and spleen were weakly reactive with anti-B5, with less than 6% of cells analyzed being positive. The E-population of normal spleen of which 70-80% of cells

> express the Bl antigen, similarly weakly expressed B5. Mononuclear cells from thymus and bone marrow were unreactive with anti-B5.

Reactivity of anti-B5 monoclonal antibody to activated fractionated peripheral blood cells was also

10 determined. B-cells were prepared by E rosetting, adherence, and then lysis of the remaining cells with anti-T-cell (T4 and T8) and anti-monocyte (Mol and Mo2) antibodies and complement. This B cell enriched fraction (50% B1+) was cultured in the presence of

-- ^■ -- anti-Ig antibody conjugated to beads. At 3 days, these cells (60% B1+) were harvested and the viability ranged between 60 and 80%. These cells were considered activated since they were proliferating as measured by

3 uptake of H TdR (stimulation index = 5 - 10) and

²⁰ morphologically approximately 2/3 of cells now were enlarged with a lymphoblastoid appearance. The cells were then examined by indirect immunofluorescence and flow cytometric analysis for B5 expression. In contrast to resting peripheral blood B-cells, approximately 25%

²⁵ of cells now expressed B5. In order to demonstrate that B5 expression was limited to activated B-cells, cells were labelled with anti-Bl directly conjugated to fluorescein and anti-B5 conjugated to biotin then developed with Texas-Red-Avidin. Utilizing dual laser

30 flow cytometric analysis, it was clearly shown that the majority of unsti ulated cells only expressed Bl with rare cells expressing Bl and B5 (Figure 3) . However, after 3 days of culture with anti-Ig antibody, 25% of cells expressed B5 and all cells expressing B5

- 14 - co-expressed Bl. In contrast, T cells isolated by E rosetting, were cultured with PHA for six days. Viable cells were isolated and the cell surface phenotype examined after 2 and 6 days of stimulation. These cells were uniformly T cells by their strong-expression of Til and were activated as determined by their expression of la (30% of cells expressed la at day 6) and IL-2 receptor (90% of cells expressed IL-2R at day 2, 70% at day 6) . These activated T-cells demonstrated no

10 detectable B5 antigen. Similarly monocytes were activated overnight with PHA-leukocyte conditioned medium (PHA-LCM) , and although these cells strongly expressed Mol, Mo2, and la, they did not express B5. Biochemical Characterization of B5

15 Burkitt's lymphoma cell line Ramos, and the plasma cell leukemia cell line RPMI 8226, were used for the isolation of the B5 cell surface antigen. A modification of the lactoperoxidase conjugation iodination technique as described by Boyd et al., J.

²⁰ Immunol (1981) 126:2461 was used to label cell surface proteins with I 125. The iodinated cells obtained from this procedure were washed twice with cold PBS and lysed on ice with cell lysis buffer (50mM Tris HCl, 0.4M NaCl,

1% Triton X-100, 2mM DMSF, 5mM EDTA, 50mM iodoacetamide,

²⁵ pH 8). ^' After 30 minutes the lysate was centrifuged at

800g for 10 minutes to remove unlysed cells, nuclei, and other insoluble material. The supernatant was frozen at -80°C until analyzed by immunoprecipitation.

Cell supernatants and cell lysates were

^* -*-- centrifuged at 10,000g for 30 minutes and transferred to fresh test tubes. Prior to immunoprecipitation the cell lysates were mixed with 20mg of rabbit anti-human Ig antibody and pre-cleared four times; twice, for 1 hour at 4°C, with Pansorbin once with Sepharose 4B beads

^*■** -' and finally with preformed complex of rabbit anti-mouse

- 15 - antibody and an irrelevant mouse immunoglobulin. The pre-cleared samples were mixed with either: 1) Anti-B5 rabbit anti-mouse Ig complexes. 2) An irrelevant monoclonal IgM complex with rabbit anti-mouse Ig. 3) B5-conjugated Sepharose 4B or 4) Irrelevant

IgM-conjugated Sepharose 4B. The mixtures were held on ice for 2 hours at 0°C after which the samples were centrifuged at 10,000g for 5 minutes and the supernatants discarded. The pellets were washed four times in 1% Triton X-100/0.2% sodium deoxycholate in RIPA buffer (0.2 sodium phosphate, 5mM EDTA, 5mM EGTA, ImM NaF, pH 7.4). As shown in Figure 4, the precipitates were analyzed by 10% SDS-polyacrylamide gel electrophoresis under non-reducing (lanes 1-5) and reducing (50mM dithiothreitol) (lanes 6-10) conditions. Ramos cell lines: Anti-B5 conjugated to Sepharose 4B (lanes 1 and 5) was compared with an irrelevant antibody conjugated to Sepharose4B (lanes 2 and 6) . RPMI 8226 cell line: Anti-B5 rabbit anti-mouse Ig preformed complexes (lanes 4 and 8) were compared to irrelevant antibody rabbit anti-mouse Ig preformed complexes (lanes 3 and 7). The apparent molecular weights (m.w.) of B5 at 75 kilodaltons (Kd) under reducing and 67Kd under non-reducing conditions, reflect the presence of interchain disulfide bonds. (This biochemical characterization of the B5 antigen shows that it is a single chain cell surface protein.)

Use

The monoclonal antibody of the invention can be labeled with a detectable label, e.g., a radiolabel by conventional procedures, and provide a quantitative measurement of activated B-cells in biological samples or in vivo.

- 16 -

Because of its specificity for neoplasms of B-cell origin corresponding to the mid-stage of B-cell differentiation, the monoclonal antibody of the invention can be used to detect the presence of these cell types in biological samples. The monoclonal antibody of the invention can be used as a diagnostic aid in characterizing the cell type of various lymphomas and leukemias arising from B-cells. In addition, in vivo imaging using rodiolabeled monoclonal antibody of 0 the invention can provide a noninvasive means for detecting and localizing these cell types, e.g., lymphoid tumors.

The monoclonal antibody of the invention will also be useful in defining the role of activated B-cells ^* - ^* in autoimmune diseases, infections and other diseases which are characterized by activated B-cells, e.g., organ rejection.

Other embodiments are within the following claims. 0