Cytotechnology 7: 1-6, 1991. 9 1991 KluwerAcademic Publishers. Printed in the Netherlands.
Increased antigen binding strengths of hybrid antibodies produced by human hybrid hybridomas Hirofumi Tachibana 1, Sanetaka Shirahata 1, Hiroharu Kawahara 2 and Hiroki Murakami 1 1Graduate School of Genetic Resources Technology; 2Department of Food Science and Technology, Faculty of Agriculture, Kyushu University 46-09, 6-10-1 Hakozaki, Fukuoka 812, Japan Received 22 May 1991; accepted in revised form 4 June 1991
Key words: hybrid hybridoma, hybrid antibody, human IgM Abstract
Two cell lines of human hybridomas were fused to generate hybrid antibodies. One human hybridoma cell line was HT2 producing IgM monoclonal antibody (MAb) reactive to carboxy peptidase A (Cpase) and double stranded DNA (ds DNA) and another was SU-l-D2 secreting IgM MAb reactive to ds DNA but not to Cpase. Most hybrid hybridomas obtained by fusion of the two hybridomas secreted hybrid antibodies exhibiting increased antigen binding strengths. All of the hybrid antibodies with increased binding strengths against Cpase and ds DNA contained only the light chains derived from SU-1-D2. These results suggested that increase in the binding strength of the hybrid antibodies resulted from heterogeneous association of H and L chains derived from HT2 and SU-1-D2 cells.
Introduction
Recent developments in animal cell culture technology have enabled the production of large quantities of bioactive proteins by cell culture (Murakami, 1990). Since natural bioactive proteins do not necessarily have ideal properties for utilization as drugs or reagents, it will be advantageous to produce supranatural proteins having superior properties to those of natural proteins. Although natural monoclonal antibody (MAb) recognizes only a single epitope, bifunctional antibodies (BAbs) or multispecific antibodies can bind to two or more different antigens simultaneously and have many advantages for immunotoxin, immunoassay and immunodiagnosis (Nolan et al., 1990; Suresh et al., 1986, Gilliland et al., 1988).
BAbs can be produced by biological or chemical methods with various degrees of success. Biologically produced BAbs have the advantage of being active in their native state and need no chemical alteration. Though biological production of BAbs involves the production of hybrid hybridomas (HyHs), it is difficult to predict and control the heterogeneous or undesirable combination of immunoglobulin (Ig) due to the differential rate of Ig synthesis and differential preference on chain association by HyHs. On IgG class hybrid antibodies, ten molecular species arising from the random association of heavy and light chains can be expected (Suresh et al., 1986). Among them, only one BAb antibody will bind to the antigens with high affinity, and the other 9 molecular species will be inactive or have reduced affinity for the antigen. Since IgM mol-
2 ecules have ten antigen binding sites, bispecific or multispecific IgM hybrid antibodies are expected to be easily produced. However, few papers report production of human IgM hybrid antibodies. Recently, Shinmoto et al. (1991) generated human IgM class hybrid anti-ricin and anti-diphtheria toxin antibodies. The BAb had almost the same reactivity to ricin as parental antibody, whereas the binding strength to diphtheria was about 1/10 to that of parental antibody. In order to produce human IgM hybrid antibodies and clarify their nature, HyHs were produced by fusing the HT2 cells that secrete I g M class antibody reactive to both carboxy peptidase A (Cpase) and double stranded DNA (ds DNA) with SU-1-D2 cells that secrete IgM class anti ds DNA by electroporation. Most of the human IgM class hybrid antibodies produced by the HyHs exhibited increased antigen binding strength. The increase in affinity for antigen was assumed to be caused by heterogeneous association of heavy (H) and light (L) chain derived from parental hybridomas.
Materials and methods
Cells and cell culture Human hybridoma HT2, a G418 resistant clone of HB4C5 (Murakami et al., 1985) secreted antibodies (IgM, ~,) reactive to Cpase and ds DNA. SU-1-D2 producing human MAb (IgM, ~,, •) specific to ds DNA was a mycophenolic acid (MPA) resistant clone of SU-1 (Hashiozume et al., 1987). These cells were maintained in ERDF medium (Kyokuto Seiyaku, Japan) containing 5% fetal bovine serum (FBS) (Hyclone Lab. Co., Utha) in humidified 5% CO2/95% air atmosphere at 37~
Cell fusion HT2 cells were fused with SU-1-D2 cells by electroporation (Denshi Kagaku, ESCF-3007).
Each cell aliquot (1 x 107 cells) was mixed and washed with fusion buffer containing 0.25 M mannitol, 0.1 mM CaC12, 0.1 mM MgCI2 and 0.2 mM Tris-HC1 (pH 7.4), suspended with 0.5 ml of the buffer, plated into a cell chamber and incubated for 10 min at room temperature. The cell mixture was treated twice at the pulse width of 30 ~ts at 3 KV/cm. Cells were washed with ERDF medium and resuspended with ERDF medium containing 15% FBS. The cell suspension was seeded into a 96 well plate at about 104 ceils per well and maintained in a selection medium (15% FBS/ERDF containing 2 mg/ml G418, 2 gg/ml MPA, and 0.25 mg/ml xanthin). Wells with growing colonies were screened for production of antibodies reactive to both Cpase and ds D N A and recloned twice by limiting dilution. HyHs were cultured in ERDF medium containing 5% FBS unless otherwise stated.
Assay of antigen binding Antigen binding activities of Ig obtained from HyHs were assayed for antigen binding by ELISA as described elsewhere (Hashizume et al., 1987). Briefly, a 96-well multiplate was coated with ds DNA (Sigma)trapped with protamine sulfate. Wells coated with protamine sulfate only were used as controls, because protamine tended to be bound to IgM. After washing with PBS containing 0.05% Tween 20 (TPBS), wells were blocked with 50% FBS/PBS to prevent nonspecific adsorption of Ig. Igs obtained from HyHs from the cells were added to the wells and incubated for 1 h. After washing the wells with TPBS, peroxidase conjugated anti-human IgM antibody (Tago, CA, USA) was added and incubated for 1 h. Each well was washed and substrate solution (2,2'-azino-di-(3-ethyl benzthiazolin sulfonic acid) for peroxidase was added and absorbance at 405 n m (OD405nm) was measured by a plate reader. The reactivity of antibodies were obtained by subtracting the control OD405nm from OD405nm of DNA coated well. For the reactivity of antibodies to Cpase (Sigma), a plate was coated with Cpase dissolved in 0.05 M sodium bicarbonate.
3 amine and s o d i u m selenite (ITES) ( M u r a k a m i et
Transfer blotting
al., 1982) (Table 1). Antibodies f r o m H y H s were applied to SDSp o l y a c r y l a m i d e gel electrophoresis ( S D S - P A G E ) and transferred to a nitrocellulose filter as described previously (Aihara et al., 1988). T h e blotted filter was b l o c k e d with B l o c k A c e (blockhag reagent, D a i n i p p o n Pharmaceuticals, Japan) for 10 h at r o o m temperature. A f t e r w a s h i n g with T P B S , L chains of blotted Ig w e r e detected with peroxidase conjugated anti h u m a n ~: chain or chain (Tago, CA, USA).
Characterization of antibodies from HyHs
Results
Production of HyHs H T 2 (a G418-resistant H B 4 C 5 clone) was fused with S U - 1 - D 2 (a MPA-resistant SU-1 clone) b y electroporation. H y H s were selected b y the use of drug-resistance to b o t h G418 and M P A . Proliferation o f H y H s was o b s e r v e d in 7 wells of 384 wells seeded, and all o f the H y H s p r o d u c e d Igs. T h e s e H y H s produced Igs in the range f r o m about 50 to 600 ng/ml/105 cells per 3 days for m o r e than 6 months in serum-free E R D F m e d i u m s u p p l e m e n t e d with insulin, transferrin, ethanol-
T h e antibodies o f the all growing wells were e x a m i n e d for the reactivities against C p a s e and ds D N A b y E L I S A . Six o f 7 H y H s clones produced I g M reactive to both Cpase and ds D N A . O f the rest one clone, E10D3 produced I g M reactive to Cpase but not t o ds D N A . I g G produced b y S U - 1 - D 2 and H y H s did not react t o both Cpase and ds D N A (data not shown). The reactivities of these antibodies to C p a s e and ds D N A were c o m p a r e d with those o f H T 2 and S U - 1 - D 2 antibodies. Antibodies produced b y the 7 H y H s reacted to C p a s e stronger t h a n that o f the parental H T 2 (Table 1). For ds D N A , these antibodies except E10D3 reacted stronger than parental H T 2 and SU-1-D2. Figure 1 shows the reactivities o f E 5 H 8 H 1 , E10D3, H T 2 and SU-1D2 antibodies to C p a s e and ds D N A . O f the 6 clones, E 5 H 8 H 1 was e x a m i n e d as representative of the antibodies reactive to b o t h Cpase and ds D N A because o f its high reactivity. E10D3 antib o d y reacted with Cpase stronger than the H T 2 antibody, but reacted with ds D N A m u c h l o w e r
Table t. Characterization and production of antibodies from hybrid hybridomas and parental hybridomas Cells
Anti Cpase (OD405)
Anti ds DNA (OD405)
H chain
L chain
Productivity IgM/105cells (ng/ml)
HT-2 SU-1-D2
0.15 0,005
0.100 0.150
Ix ~t, T
~, k, n
390 109
E5H8G6 E12B8E3 E12B8H6 G8B7 E5H8H9 E10D3 E5H8H1
>0.5 >0.5 >0.5 >0.5 >0.5 0.2 >0.5
>0,3 >0.3 >0,3 >0.3 >0.3 0.005 >0.3
IX,T ~t Ix ].t, T Ix, T Ix, T ~, T
~, ~ ~, ~ ~, ~ ~, ~ ~, ~ ~ ~, K
196 99 363 203 312 59 619
Antibodies produced by HyHs and the parental hybridomas were examined for the reactivity to Cpase and ds DNA, and for the productivity oflgM. Anti Cpase and anti ds DNA assay was done at dilutions ranging from 10 ng/ml to 2 Ixg/mlof antibodies by ELISA. The represented values in the table were at 400 ng/ml. The concentration of total Ig and class of H and L chains were determined by ELISA using anti H or L chain specific antibodies.
4
0.8~
-
Cpase
F
E5H8HI
--
0.6 0.4 0.2 0
C~ 0
I su-l-D
0
E5HSHI
0.8 -
d
0.6-
m
0.4S U - I -D2
0.2-
H'I'2
0
10
100
1000
IgM C o n c e n t r a t i o n
(ng/ml)
Fig. 1. ELISAanalysiscomparing the reactivityof antibodies from HyHs with those from the parental hybridomas. The hybrid (E10D3 and E5H8H1) and parental (HT2 and SU-1-D2) antibodies were assayedfor antibodybinding to Cpase and ds DNA. Antibodysamples were diluted with 5% FBS/PBS and aliquots were assayedfor antibodybinding.
than those of HT2 and SU-l-D2. Reactivities of E5H8H1 antibody to Cpase and ds DNA were much higher than those of the parental cells (HT2 and SU-1-D2) and E10D3.
Classification of light chains of hybrid antibodies To examine the difference in reactivities of these antibodies, the L chain classes of the antibodies were determined. Six clones, except for E10D3, produced antibodies having both tc and )~ L chains (Table 1). E10D3 produced only IgM-~: antibody, and )~ chain was not detected in either the cytoplasm of the cell or in culture medium. SU-1-D2 produced both ~ and ~ chains as L chain, and the molecular weights of these chains were identical with that of human serum L
chains. On the other hand, HT2 cells produced IgM-)~ antibody, and the ~ chain was glycosylated (data not shown). Since the molecular weight of the )~ chain of HT2 antibody was higher than that of the )~ chain of SU-1-D2, both L chains could be electrophoretically distinguished from each other. Therefore, the L chains of E5H8H1 and E10D3 antibodies and their parental HT2 and SU-1-D2 antibodies were assayed by Western blotting. As shown in Fig. 2, the molecular weight of ~ chain of E5H8H1 antibody was the same size to that of SU-1-D2 antibody, and the ~, chain having the same molecular size as that of HT2 was not detected. The ~: chains of E5H8H1 and E10D3 antibodies had the same molecular size as that of SU-1-D2. These results indicate that the L chains of E5H8H1 antibody consisted of ~: and )~ chain derived from only SU-1-D2, not with the ~ chain derived from HT2, and E10D3 produced the only ~: chain derived from SU-1-D2 as L chain. These results are summarized in Table 2. Since all IX chains of these antibodies had the same molecular size as that of human serum IgM, it was difficult to distinguish the origin of the tx chains of antibodies produced by E5H8H1 and E10D3 by Western blotting (data not shown).
Discussion
HyHs produce hybrid type Ig molecules arising from the random association of H and L chains (Shinmoto et al., 1988). Only homologous pairs of H and L chain will produce the correct three dimensional configuration of the high affinity antigen binding site. Heterogeneous combinations will be inactive or have reduced affinity for the antigen. The experiments by Klinman indicated that the loss of antibody binding activity after recombination of serum anti-DNP antibody was due to the heterogeneous combination of H and L chains (Klinman et al., 1971). A slight increase in antigen binding resulted from recombining antigen-specific H chains and non-specific L chains (Hong et al., 1966). We have obtained hybrid antibodies with i n -
Fig. 2. SDS-PAGE analysis of antibodies obtained from parental hybridomas and HyHs. Antibody samples were electrophoresed and
transfer blotted. The blotted filter was reacted with peroxidase conjugated anti human ~. chain (A) or • chain (B) antibodies. Lane 1, human serum IgM; lane 2, HT2; lane 3, SU-1-D2; lane 4, EI0D3; lane 5, E5H8H1. Table 2. Light chain class of hybrid and two parental antibodies
Cells
(HT2 derived)
(SU-1-D2 derived)
(SU-1-D2 derived)
HT2 SU-1-D2
+ -
+
+
E10D3 E5H8H1
-
+
+ +
creased antigen binding strength for antigens b y fusing the parental h y b r i d o m a s , H T 2 and SU-1D2. Figure 3 shows the possible c o m b i n a t i o n o f H and L chains in H y H s . L chains o f E 5 H 8 H 1 , one o f the hybrid antibodies consisted o f only ~: and ~, chains derived f r o m S U - 1 - D 2 , and h o u n d to C p a s e and ds D N A with higher affinity than parental H T 2 and S U - 1 - D 2 antibodies. S U - 1 - D 2 antibody was scarcely reactive to Cpase. T h e s e results suggested that high antigen binding o f E 5 H 8 H 1 was due to the c o m b i n a t i o n o f H chain derived f r o m H T 2 and L chain derived f r o m SU-1-D2. H y b r i d a n t i b o d y E 1 0 D 3 has only K chain derived f r o m S U - 1 - D 2 , and reacted slightly higher to C p a s e than H T 2 but not to ds D N A .
This suggested that the hybrid Ig molecule consisted of ~t chain derived f r o m H T 2 and ~: chain derived f r o m S U - 1 - D 2 which could bind to Cpase but loses affinity for ds D N A (as indicated in Fig. 3-A). F o r c o m p a r i s o n o f the two hybrid antibodies, we attempted to clarify the Ig m o l e c u l e s with higher affinity. E 5 H 8 H 1 was reactive to Cpase with higher affinity than E10D3, and bound to ds D N A stronger than S U - l - D 2 . On classes o f the L chains, E 5 H 8 H 1 p r o d u c e d Ig molecules consisting o f ~. and ~: chains. Therefore, it is thought that the higher affinity o f E 5 H 8 H 1 for both antigens was due to the Ig m o l e c u l e H chain derived f r o m H T 2 and ~ chain derived f r o m S U - 1 - D 2 (as indicated in Fig. 3,D). These find-
6 112"2
Acknowledgement
SU- I -D2
A n U Cpz~se § A n t i d s D N A +~-
/~tJ Cpase AnU ds DNA
W e a r e g r a t e f u l to Dr. S. H a s h i z u m e generous gift of the SU-1-D2 cells.
++
f o r his
References
I
I
EIOD3
A~ti Cpase .ann ds DNA
E5HSH1 ++ -
AnU Cpase Anti ds DNA
+++ 4-++
Fig. 3. The expected Ig molecules (monomelic) with increased antigen binding by heterogeneous combination of H and L chains. The binding affinity was symbolized by +++, ++, + and - . The pairing patterns of H and L chains were designed A,B,C and D.
ings suggest that hybrid antibodies with increased affinity and changed specificity can be generated by heterogeneous combination of H and L chain in H y H s . T h e h y b r i d a n t i b o d i e s c a n b e p r o d u c e d b y c h e m i c a l m e t h o d s ( B r e n n a n et al., 1985). H o w e v e r , b i o l o g i c a l l y p r o d u c e d h y b r i d antibodies have the advantage of maintenance of l o n g - t e r m p r o d u c t i o n ( N o l a n et al., 1990). H y H s o b t a i n e d h e r e c a n g r o w a n d p r o d u c e h y b r i d antib o d i e s in s e r u m - f r e e c u l t u r e f o r m o r e t h a n 6 m o n t h as o f this t i m e . G e n e r a t i o n o f h y b r i d antibodies by HyHs may be used for generation of not only bifunctional antibodies but also antibodies with high affinity.
Aihara K, Yamada K, Murakarni H, Nomura Y and Omura H (1988) Production of human-human hybfidomas secreting monoclonal antibodies reactive to breast cancer cell lines. In Vitro Cell. Develop. Biol. 24" 959-962. Brennan M, Davison PF and Paulus H (1985) Preparation of bispecific antibodies by chemical recombination of monoclonal immunoglobulin G1 fragments. Science 229: 81-83. Gilliland LK, Clark MR and Waldmann H (1988) Universal bispecific antibody for targeting tumor cells for destruction by cytotoxic T cells. Proc. Natl. Acad. Sci. USA 85:7719-7723. Hashizume S, Murakami H, Kamei M, Hirose S, Shirai T, Yamada K and Omura H (1987) Specificity of anti-polynucleotide monoclonal antibodies from human-human hybridomas. In Vitro Cell. Develop. Biol. 23: 53-56. Hong R and Nissonoff A (1966) Heterogeneity in the complementation of polypeptide subunits of a purified antibody isolated from an individual rabbit. J. Immunol. 96: 622-628. Klinman NR (1971) Regain of homogeneous binding activity after recombination of chains of 'monofocal' antibody. L Immunol. 106: 1330--1337. Murakami H (1990) What should be focused in the study of cell culture technology for production for bioactive proteins. Cytotechnology 3: 3-7. Murakami H, Hashizume S, Ohashi H, Shinohara K, Yasumoto K, Nomoto K and Omura H (1985) Human-human hybridomas secreting antibodies specific to human lung carcinoma. In Vitro Cell. Develop. Biol. 21: 593-596. Murakami H, Masui H, Sam GH, Sueoka N, Chow TP and Kano-Sueoka T (1982) Growth of hybridoma cells in serumfree medium: Ethanolamine is an essential component. Proc. Natl. Acad. Sci. USA 79: 1158-1162. Nolan O and O'Kennedy R (1990) Bifunctional antibodies: Concept, production and applications. Biochim. Biophys. Acta 1040: 1-11~ Shinmoto H, Dosako S, Tachibana H, Yamada K, Shirahata S and Murakami H (1991) Generation of hybrid hybfidomas secreting human IgM class hybrid antificin and antidiphtefia toxin antibodies. Hum. Antibod. Hybridomas 2: 39-41. Shinmoto H, Murakami H, Dosako S, Yamada K and Omura H (1988) Human hybrid immunoglobulin M containing immunoglobulin A. In Vitro Cell. Develop. Biol. 24" 505-510. Suresh MR, Cuello AC and Milstein C (1986) Bispecific monoclonal antibodies from hybrid hybridomas. Methods in Enzymol. 121: 210--228.
Address for offprints: Hirofumi Tachibana, Graduate School of Genetic Resources Technology, Faculty of Agriculture, Kyushu University 46-09, Hakozaki 6-10-1, Fukuoka 812, Japan
Increased antigen binding strengths of hybrid antibodies produced by human hybrid hybridomas Hirofumi Tachibana 1, Sanetaka Shirahata 1, Hiroharu Kawahara 2 and Hiroki Murakami 1 1Graduate School of Genetic Resources Technology; 2Department of Food Science and Technology, Faculty of Agriculture, Kyushu University 46-09, 6-10-1 Hakozaki, Fukuoka 812, Japan Received 22 May 1991; accepted in revised form 4 June 1991
Key words: hybrid hybridoma, hybrid antibody, human IgM Abstract
Two cell lines of human hybridomas were fused to generate hybrid antibodies. One human hybridoma cell line was HT2 producing IgM monoclonal antibody (MAb) reactive to carboxy peptidase A (Cpase) and double stranded DNA (ds DNA) and another was SU-l-D2 secreting IgM MAb reactive to ds DNA but not to Cpase. Most hybrid hybridomas obtained by fusion of the two hybridomas secreted hybrid antibodies exhibiting increased antigen binding strengths. All of the hybrid antibodies with increased binding strengths against Cpase and ds DNA contained only the light chains derived from SU-1-D2. These results suggested that increase in the binding strength of the hybrid antibodies resulted from heterogeneous association of H and L chains derived from HT2 and SU-1-D2 cells.
Introduction
Recent developments in animal cell culture technology have enabled the production of large quantities of bioactive proteins by cell culture (Murakami, 1990). Since natural bioactive proteins do not necessarily have ideal properties for utilization as drugs or reagents, it will be advantageous to produce supranatural proteins having superior properties to those of natural proteins. Although natural monoclonal antibody (MAb) recognizes only a single epitope, bifunctional antibodies (BAbs) or multispecific antibodies can bind to two or more different antigens simultaneously and have many advantages for immunotoxin, immunoassay and immunodiagnosis (Nolan et al., 1990; Suresh et al., 1986, Gilliland et al., 1988).
BAbs can be produced by biological or chemical methods with various degrees of success. Biologically produced BAbs have the advantage of being active in their native state and need no chemical alteration. Though biological production of BAbs involves the production of hybrid hybridomas (HyHs), it is difficult to predict and control the heterogeneous or undesirable combination of immunoglobulin (Ig) due to the differential rate of Ig synthesis and differential preference on chain association by HyHs. On IgG class hybrid antibodies, ten molecular species arising from the random association of heavy and light chains can be expected (Suresh et al., 1986). Among them, only one BAb antibody will bind to the antigens with high affinity, and the other 9 molecular species will be inactive or have reduced affinity for the antigen. Since IgM mol-
2 ecules have ten antigen binding sites, bispecific or multispecific IgM hybrid antibodies are expected to be easily produced. However, few papers report production of human IgM hybrid antibodies. Recently, Shinmoto et al. (1991) generated human IgM class hybrid anti-ricin and anti-diphtheria toxin antibodies. The BAb had almost the same reactivity to ricin as parental antibody, whereas the binding strength to diphtheria was about 1/10 to that of parental antibody. In order to produce human IgM hybrid antibodies and clarify their nature, HyHs were produced by fusing the HT2 cells that secrete I g M class antibody reactive to both carboxy peptidase A (Cpase) and double stranded DNA (ds DNA) with SU-1-D2 cells that secrete IgM class anti ds DNA by electroporation. Most of the human IgM class hybrid antibodies produced by the HyHs exhibited increased antigen binding strength. The increase in affinity for antigen was assumed to be caused by heterogeneous association of heavy (H) and light (L) chain derived from parental hybridomas.
Materials and methods
Cells and cell culture Human hybridoma HT2, a G418 resistant clone of HB4C5 (Murakami et al., 1985) secreted antibodies (IgM, ~,) reactive to Cpase and ds DNA. SU-1-D2 producing human MAb (IgM, ~,, •) specific to ds DNA was a mycophenolic acid (MPA) resistant clone of SU-1 (Hashiozume et al., 1987). These cells were maintained in ERDF medium (Kyokuto Seiyaku, Japan) containing 5% fetal bovine serum (FBS) (Hyclone Lab. Co., Utha) in humidified 5% CO2/95% air atmosphere at 37~
Cell fusion HT2 cells were fused with SU-1-D2 cells by electroporation (Denshi Kagaku, ESCF-3007).
Each cell aliquot (1 x 107 cells) was mixed and washed with fusion buffer containing 0.25 M mannitol, 0.1 mM CaC12, 0.1 mM MgCI2 and 0.2 mM Tris-HC1 (pH 7.4), suspended with 0.5 ml of the buffer, plated into a cell chamber and incubated for 10 min at room temperature. The cell mixture was treated twice at the pulse width of 30 ~ts at 3 KV/cm. Cells were washed with ERDF medium and resuspended with ERDF medium containing 15% FBS. The cell suspension was seeded into a 96 well plate at about 104 ceils per well and maintained in a selection medium (15% FBS/ERDF containing 2 mg/ml G418, 2 gg/ml MPA, and 0.25 mg/ml xanthin). Wells with growing colonies were screened for production of antibodies reactive to both Cpase and ds D N A and recloned twice by limiting dilution. HyHs were cultured in ERDF medium containing 5% FBS unless otherwise stated.
Assay of antigen binding Antigen binding activities of Ig obtained from HyHs were assayed for antigen binding by ELISA as described elsewhere (Hashizume et al., 1987). Briefly, a 96-well multiplate was coated with ds DNA (Sigma)trapped with protamine sulfate. Wells coated with protamine sulfate only were used as controls, because protamine tended to be bound to IgM. After washing with PBS containing 0.05% Tween 20 (TPBS), wells were blocked with 50% FBS/PBS to prevent nonspecific adsorption of Ig. Igs obtained from HyHs from the cells were added to the wells and incubated for 1 h. After washing the wells with TPBS, peroxidase conjugated anti-human IgM antibody (Tago, CA, USA) was added and incubated for 1 h. Each well was washed and substrate solution (2,2'-azino-di-(3-ethyl benzthiazolin sulfonic acid) for peroxidase was added and absorbance at 405 n m (OD405nm) was measured by a plate reader. The reactivity of antibodies were obtained by subtracting the control OD405nm from OD405nm of DNA coated well. For the reactivity of antibodies to Cpase (Sigma), a plate was coated with Cpase dissolved in 0.05 M sodium bicarbonate.
3 amine and s o d i u m selenite (ITES) ( M u r a k a m i et
Transfer blotting
al., 1982) (Table 1). Antibodies f r o m H y H s were applied to SDSp o l y a c r y l a m i d e gel electrophoresis ( S D S - P A G E ) and transferred to a nitrocellulose filter as described previously (Aihara et al., 1988). T h e blotted filter was b l o c k e d with B l o c k A c e (blockhag reagent, D a i n i p p o n Pharmaceuticals, Japan) for 10 h at r o o m temperature. A f t e r w a s h i n g with T P B S , L chains of blotted Ig w e r e detected with peroxidase conjugated anti h u m a n ~: chain or chain (Tago, CA, USA).
Characterization of antibodies from HyHs
Results
Production of HyHs H T 2 (a G418-resistant H B 4 C 5 clone) was fused with S U - 1 - D 2 (a MPA-resistant SU-1 clone) b y electroporation. H y H s were selected b y the use of drug-resistance to b o t h G418 and M P A . Proliferation o f H y H s was o b s e r v e d in 7 wells of 384 wells seeded, and all o f the H y H s p r o d u c e d Igs. T h e s e H y H s produced Igs in the range f r o m about 50 to 600 ng/ml/105 cells per 3 days for m o r e than 6 months in serum-free E R D F m e d i u m s u p p l e m e n t e d with insulin, transferrin, ethanol-
T h e antibodies o f the all growing wells were e x a m i n e d for the reactivities against C p a s e and ds D N A b y E L I S A . Six o f 7 H y H s clones produced I g M reactive to both Cpase and ds D N A . O f the rest one clone, E10D3 produced I g M reactive to Cpase but not t o ds D N A . I g G produced b y S U - 1 - D 2 and H y H s did not react t o both Cpase and ds D N A (data not shown). The reactivities of these antibodies to C p a s e and ds D N A were c o m p a r e d with those o f H T 2 and S U - 1 - D 2 antibodies. Antibodies produced b y the 7 H y H s reacted to C p a s e stronger t h a n that o f the parental H T 2 (Table 1). For ds D N A , these antibodies except E10D3 reacted stronger than parental H T 2 and SU-1-D2. Figure 1 shows the reactivities o f E 5 H 8 H 1 , E10D3, H T 2 and SU-1D2 antibodies to C p a s e and ds D N A . O f the 6 clones, E 5 H 8 H 1 was e x a m i n e d as representative of the antibodies reactive to b o t h Cpase and ds D N A because o f its high reactivity. E10D3 antib o d y reacted with Cpase stronger than the H T 2 antibody, but reacted with ds D N A m u c h l o w e r
Table t. Characterization and production of antibodies from hybrid hybridomas and parental hybridomas Cells
Anti Cpase (OD405)
Anti ds DNA (OD405)
H chain
L chain
Productivity IgM/105cells (ng/ml)
HT-2 SU-1-D2
0.15 0,005
0.100 0.150
Ix ~t, T
~, k, n
390 109
E5H8G6 E12B8E3 E12B8H6 G8B7 E5H8H9 E10D3 E5H8H1
>0.5 >0.5 >0.5 >0.5 >0.5 0.2 >0.5
>0,3 >0.3 >0,3 >0.3 >0.3 0.005 >0.3
IX,T ~t Ix ].t, T Ix, T Ix, T ~, T
~, ~ ~, ~ ~, ~ ~, ~ ~, ~ ~ ~, K
196 99 363 203 312 59 619
Antibodies produced by HyHs and the parental hybridomas were examined for the reactivity to Cpase and ds DNA, and for the productivity oflgM. Anti Cpase and anti ds DNA assay was done at dilutions ranging from 10 ng/ml to 2 Ixg/mlof antibodies by ELISA. The represented values in the table were at 400 ng/ml. The concentration of total Ig and class of H and L chains were determined by ELISA using anti H or L chain specific antibodies.
4
0.8~
-
Cpase
F
E5H8HI
--
0.6 0.4 0.2 0
C~ 0
I su-l-D
0
E5HSHI
0.8 -
d
0.6-
m
0.4S U - I -D2
0.2-
H'I'2
0
10
100
1000
IgM C o n c e n t r a t i o n
(ng/ml)
Fig. 1. ELISAanalysiscomparing the reactivityof antibodies from HyHs with those from the parental hybridomas. The hybrid (E10D3 and E5H8H1) and parental (HT2 and SU-1-D2) antibodies were assayedfor antibodybinding to Cpase and ds DNA. Antibodysamples were diluted with 5% FBS/PBS and aliquots were assayedfor antibodybinding.
than those of HT2 and SU-l-D2. Reactivities of E5H8H1 antibody to Cpase and ds DNA were much higher than those of the parental cells (HT2 and SU-1-D2) and E10D3.
Classification of light chains of hybrid antibodies To examine the difference in reactivities of these antibodies, the L chain classes of the antibodies were determined. Six clones, except for E10D3, produced antibodies having both tc and )~ L chains (Table 1). E10D3 produced only IgM-~: antibody, and )~ chain was not detected in either the cytoplasm of the cell or in culture medium. SU-1-D2 produced both ~ and ~ chains as L chain, and the molecular weights of these chains were identical with that of human serum L
chains. On the other hand, HT2 cells produced IgM-)~ antibody, and the ~ chain was glycosylated (data not shown). Since the molecular weight of the )~ chain of HT2 antibody was higher than that of the )~ chain of SU-1-D2, both L chains could be electrophoretically distinguished from each other. Therefore, the L chains of E5H8H1 and E10D3 antibodies and their parental HT2 and SU-1-D2 antibodies were assayed by Western blotting. As shown in Fig. 2, the molecular weight of ~ chain of E5H8H1 antibody was the same size to that of SU-1-D2 antibody, and the ~, chain having the same molecular size as that of HT2 was not detected. The ~: chains of E5H8H1 and E10D3 antibodies had the same molecular size as that of SU-1-D2. These results indicate that the L chains of E5H8H1 antibody consisted of ~: and )~ chain derived from only SU-1-D2, not with the ~ chain derived from HT2, and E10D3 produced the only ~: chain derived from SU-1-D2 as L chain. These results are summarized in Table 2. Since all IX chains of these antibodies had the same molecular size as that of human serum IgM, it was difficult to distinguish the origin of the tx chains of antibodies produced by E5H8H1 and E10D3 by Western blotting (data not shown).
Discussion
HyHs produce hybrid type Ig molecules arising from the random association of H and L chains (Shinmoto et al., 1988). Only homologous pairs of H and L chain will produce the correct three dimensional configuration of the high affinity antigen binding site. Heterogeneous combinations will be inactive or have reduced affinity for the antigen. The experiments by Klinman indicated that the loss of antibody binding activity after recombination of serum anti-DNP antibody was due to the heterogeneous combination of H and L chains (Klinman et al., 1971). A slight increase in antigen binding resulted from recombining antigen-specific H chains and non-specific L chains (Hong et al., 1966). We have obtained hybrid antibodies with i n -
Fig. 2. SDS-PAGE analysis of antibodies obtained from parental hybridomas and HyHs. Antibody samples were electrophoresed and
transfer blotted. The blotted filter was reacted with peroxidase conjugated anti human ~. chain (A) or • chain (B) antibodies. Lane 1, human serum IgM; lane 2, HT2; lane 3, SU-1-D2; lane 4, EI0D3; lane 5, E5H8H1. Table 2. Light chain class of hybrid and two parental antibodies
Cells
(HT2 derived)
(SU-1-D2 derived)
(SU-1-D2 derived)
HT2 SU-1-D2
+ -
+
+
E10D3 E5H8H1
-
+
+ +
creased antigen binding strength for antigens b y fusing the parental h y b r i d o m a s , H T 2 and SU-1D2. Figure 3 shows the possible c o m b i n a t i o n o f H and L chains in H y H s . L chains o f E 5 H 8 H 1 , one o f the hybrid antibodies consisted o f only ~: and ~, chains derived f r o m S U - 1 - D 2 , and h o u n d to C p a s e and ds D N A with higher affinity than parental H T 2 and S U - 1 - D 2 antibodies. S U - 1 - D 2 antibody was scarcely reactive to Cpase. T h e s e results suggested that high antigen binding o f E 5 H 8 H 1 was due to the c o m b i n a t i o n o f H chain derived f r o m H T 2 and L chain derived f r o m SU-1-D2. H y b r i d a n t i b o d y E 1 0 D 3 has only K chain derived f r o m S U - 1 - D 2 , and reacted slightly higher to C p a s e than H T 2 but not to ds D N A .
This suggested that the hybrid Ig molecule consisted of ~t chain derived f r o m H T 2 and ~: chain derived f r o m S U - 1 - D 2 which could bind to Cpase but loses affinity for ds D N A (as indicated in Fig. 3-A). F o r c o m p a r i s o n o f the two hybrid antibodies, we attempted to clarify the Ig m o l e c u l e s with higher affinity. E 5 H 8 H 1 was reactive to Cpase with higher affinity than E10D3, and bound to ds D N A stronger than S U - l - D 2 . On classes o f the L chains, E 5 H 8 H 1 p r o d u c e d Ig molecules consisting o f ~. and ~: chains. Therefore, it is thought that the higher affinity o f E 5 H 8 H 1 for both antigens was due to the Ig m o l e c u l e H chain derived f r o m H T 2 and ~ chain derived f r o m S U - 1 - D 2 (as indicated in Fig. 3,D). These find-
6 112"2
Acknowledgement
SU- I -D2
A n U Cpz~se § A n t i d s D N A +~-
/~tJ Cpase AnU ds DNA
W e a r e g r a t e f u l to Dr. S. H a s h i z u m e generous gift of the SU-1-D2 cells.
++
f o r his
References
I
I
EIOD3
A~ti Cpase .ann ds DNA
E5HSH1 ++ -
AnU Cpase Anti ds DNA
+++ 4-++
Fig. 3. The expected Ig molecules (monomelic) with increased antigen binding by heterogeneous combination of H and L chains. The binding affinity was symbolized by +++, ++, + and - . The pairing patterns of H and L chains were designed A,B,C and D.
ings suggest that hybrid antibodies with increased affinity and changed specificity can be generated by heterogeneous combination of H and L chain in H y H s . T h e h y b r i d a n t i b o d i e s c a n b e p r o d u c e d b y c h e m i c a l m e t h o d s ( B r e n n a n et al., 1985). H o w e v e r , b i o l o g i c a l l y p r o d u c e d h y b r i d antibodies have the advantage of maintenance of l o n g - t e r m p r o d u c t i o n ( N o l a n et al., 1990). H y H s o b t a i n e d h e r e c a n g r o w a n d p r o d u c e h y b r i d antib o d i e s in s e r u m - f r e e c u l t u r e f o r m o r e t h a n 6 m o n t h as o f this t i m e . G e n e r a t i o n o f h y b r i d antibodies by HyHs may be used for generation of not only bifunctional antibodies but also antibodies with high affinity.
Aihara K, Yamada K, Murakarni H, Nomura Y and Omura H (1988) Production of human-human hybfidomas secreting monoclonal antibodies reactive to breast cancer cell lines. In Vitro Cell. Develop. Biol. 24" 959-962. Brennan M, Davison PF and Paulus H (1985) Preparation of bispecific antibodies by chemical recombination of monoclonal immunoglobulin G1 fragments. Science 229: 81-83. Gilliland LK, Clark MR and Waldmann H (1988) Universal bispecific antibody for targeting tumor cells for destruction by cytotoxic T cells. Proc. Natl. Acad. Sci. USA 85:7719-7723. Hashizume S, Murakami H, Kamei M, Hirose S, Shirai T, Yamada K and Omura H (1987) Specificity of anti-polynucleotide monoclonal antibodies from human-human hybridomas. In Vitro Cell. Develop. Biol. 23: 53-56. Hong R and Nissonoff A (1966) Heterogeneity in the complementation of polypeptide subunits of a purified antibody isolated from an individual rabbit. J. Immunol. 96: 622-628. Klinman NR (1971) Regain of homogeneous binding activity after recombination of chains of 'monofocal' antibody. L Immunol. 106: 1330--1337. Murakami H (1990) What should be focused in the study of cell culture technology for production for bioactive proteins. Cytotechnology 3: 3-7. Murakami H, Hashizume S, Ohashi H, Shinohara K, Yasumoto K, Nomoto K and Omura H (1985) Human-human hybridomas secreting antibodies specific to human lung carcinoma. In Vitro Cell. Develop. Biol. 21: 593-596. Murakami H, Masui H, Sam GH, Sueoka N, Chow TP and Kano-Sueoka T (1982) Growth of hybridoma cells in serumfree medium: Ethanolamine is an essential component. Proc. Natl. Acad. Sci. USA 79: 1158-1162. Nolan O and O'Kennedy R (1990) Bifunctional antibodies: Concept, production and applications. Biochim. Biophys. Acta 1040: 1-11~ Shinmoto H, Dosako S, Tachibana H, Yamada K, Shirahata S and Murakami H (1991) Generation of hybrid hybfidomas secreting human IgM class hybrid antificin and antidiphtefia toxin antibodies. Hum. Antibod. Hybridomas 2: 39-41. Shinmoto H, Murakami H, Dosako S, Yamada K and Omura H (1988) Human hybrid immunoglobulin M containing immunoglobulin A. In Vitro Cell. Develop. Biol. 24" 505-510. Suresh MR, Cuello AC and Milstein C (1986) Bispecific monoclonal antibodies from hybrid hybridomas. Methods in Enzymol. 121: 210--228.
Address for offprints: Hirofumi Tachibana, Graduate School of Genetic Resources Technology, Faculty of Agriculture, Kyushu University 46-09, Hakozaki 6-10-1, Fukuoka 812, Japan
