J. Mol.
Riol.
(1992) 223. 381-382
Crystallization of the Fab Fragments of Anti-peptide Monoclonal Antibodies and a Complex with Peptide Robert E. Griest-j-, Philip D. Jeffreyt, Garry L. Taylor?,5 and Anthony R. Rees-fLaboratory of Molecular Biophysics The Rex Richards Building, South Parlcs Road &ford OX1 ,?Qli. Grea,t Britnin (Received 14 August 1991: accepted X October 1991) The antigen-binding fragments of four monoclonal antibodies that cross-react with both the “loop” peptide of hen egg-white lysozyme (residues 57 to 84) against which they were raised, and with the native protein (HEL) have been crystallized. One of these fragments also crystallizes as a complex with the peptide antigen. Krywords:
crystallization:
lysozyme;
antibodies;
The preparation of a panel of monoclonal antibodies against the “loop” peptide. comprising residues 57 to 84 of HEL, and the characterization of their epitopes has been reported (Darsley & Rees? 1985n,b). These monoclonal antibodies were obtained from mice immunized with a synthetic conjugate of a proteolytic fragment of HEL coupled to bovine serum albumin, and bind both whole native HEL and the loop peptide fragment. We have been att,empting to crystallize the uncomplexed antigen binding fragment (Fab), Fabpeptide and Fab-HEL complexes. Such a series of three-dimensional structures would elucidate the mechanisms and differences in binding intact protein and the more flexible peptide antigen. Additionally, the Fab structures would provide a test of the “maximum overlap” method used to model the Fv regions of the Gloop antibodies (de la Paz et al., 1984). Protein for crystallization was prepared with hpbridomas grown as ascites tumours in Fl hybrid SWRxRalb/c mice primed with pristane. The antibodies were purified from the ascites fluid by SO?/; saturated ammonium sulphate precipitation followed by ion-exchange chromatography on diethylaminoethyl-Sephadex (Harrison & Mage, 1967). Antibodies were digested with papain
peptide
antigen;
Fab fragments
(Porter. 1959) using an IgG to enzyrne ratio of 500 : 1 (w/w) at pH 7.0 for t’he readily cleaved antibody Gloop4. 100: 1 at pH 7.0 for Gloopl and GloopP. and 20: 1 at pH 5.5 for the more resistant Gloop5, all for four hours at 37°C. The cleavage reaction was st,opped with 3 mg of iodoacetamide per millilitre of reaction mix. Fab fragments were separated from partially digested IgG and Fc fragments by passing the concentrated digest through a 75 cm x 2.6 cm S-200 gel filtration column (20 mMsodium phosphate buffer, pH 7.5). This purificat.ion scheme gives complete separation of Fab from Fc fragments and partially digested IgG. Analysis by high-resolut,ion ion-exchange chromatography (Mono-Q column, Pharmacia) revealed t’hat some digests contained more than one ionic species of Fat). In particular, Gloop4 only crystallized after further purification on a Mono-Q column. The Fab fragments were subjected to crystallization trials in hanging drops at room temperature using the method of vapour diffusion (Wlodawer & Hodgson, 1975). All buffers and solutions used were filtered through 0.45 pm Millio filters. Protein concentrations from 1.5 to 20 mg/ml in 0.01 to 004 MTris. HCl or sodium phosphate buffers were used. Polyethylene glycol 6000 (Sigma Co.) was used as the precipitant. Initial concentrations in the drop of 5 to 1004 (w/v) polyethylene glycol were equilibrated over wells of 15, 20 and 25% (w/v) polvet,hylene glycol. The effects of both pH (5.0 to 9.5. in steps of 0.5) and NaCl concentration (60 to 0.4 M in 0.1 M steps) were tested. Gloopl (IgG2a,K), Gloop2 and Gloop4 (both IgC2b,rc) all crystallized within two weeks at pH 6.0 to pH 7.5 and after longer periods Gloopl also crystallized a,t as low as pH 5.0. There was a trend of
t Present address: Department of Biochemistry. University of Bat,h, Claverton Down, Bath BA2 TAT, Great Britain. 1 Present address: Bristol-Myers Squibb Pharmaceutical Research Institute. P.0. Box 4000, Princeton, NJ 08543-4000, U.S.A. Q Author to whom all correspondence should be addressed.
381 OOZZ--2836/92/010381-02
$03.00/O
0 1992 Academic
Press Limited
mow nucleation sites and smaller crystals at both lower pH and higher salt concentration. forming clusters of very thin plates or thin branched needles. There was a requirement for some salt. for in the absence of Na(‘l, crystallization seldom occurred and then only after several months. producing cryst,als with irregular morphologies. The best conditions cent,red around pH 7.0 to pH 7.5 with 0.1 wXaC1 present: Qloopl gavel prisms (up to 14 mm x 0.4 mm x 0.2 mm): C:loopP produced long regular needles (up t’o 1.5 mm x 0~25 mm x 0.15 mm); and Cloop4 gave similar needles (up to 0.2 mm X 0.1 mm X 0.1 mm). (:loops (IgGl,K) behaved rather differently and usually formed an oily phase during crystallization trials and has onI> twice ever yielded crystals, one of which (I.4 mm x (h3 mm x 0.3 mm) was suitable for data collection. (:loop2 presented a different problem, in that two of the cryst,al forms are morphologically indistinguisb able until after data collection and processing. ljatasets for each crystal form were collecated on a Nicolet/Xentronics area detector and processed using the XENGEN suite of programs (A. Howard. 19X8, Genex (‘orporation. Gaithersburg. MD. IT.S.A,). Table 1 lists the space groups and diffrav Con limits of the crystal forms characterized. Synthetic loop peptide has been used in cocrystallization attempts with Fah fragments: both (:loop2 and Gloop5 have yielded crystals morphologically distinct from those grown under similar conditions in t’he absence of peptide. Growth ot GloopSpeptide crystals in t,he presence of *4(ilabelled peptide has confirmed the presence of peptide in these crystals. The Gloop2 complex diffracts to at least 4.0 !I (1 Lh = 0.1 rim). but the space group has not been characterized. Att,empts to soak peptide into uncomplexed Fab cryst)als resulted in the disordering of these crystals. Molecular replacement results for the three
crystal forms of nncomplexed (sloop:! and for Gloopl indicate that there are t,wo Fa,b molcc~ules per cell for ea,ch of t’hese crystal forms. I’,,, values (Matthews. 1968) listed in Table 1 suggest four and 12 molecules per cell for Gloop4 and Gloop5, respt+ timely. The structures of Gloopl and thv three crystal forms of (:loop:! have been determined and will be reported elsewhere. The authors thank Professor Sir David L’hilliJjs for his support mtl rrr~ouragement. L$:r arr grateful to I)rs Elspeth (barman. Yvonne Jones and David Stuart, for their assistance in data collection and processing. and to Dr Andrew Howard for pointing out the init,ial incorrrc~t assignment of the (iloop spare’ group.
References Daralry. 31. J. & Rees. A. It. (198.5a). Sucleotidr seyurnf&es of‘ tive anti-lysozyme monof~lonal antibodies. EXBO -1. 4. 33!%398. IIarsley. M. .I. & Rees. A. R. (19856). Three distinct epltopes within the lofq) region of hen egg Ivsozymr defined with rnonorlonal antibodies. EM/%7 -1. 4. X-392. dr la Paz, P.. Sut’ton. 13. .J.. Parsley, hl. .I. & Rers. A. R. (1984). ModrIling of’thr fvmbining sites of’thrrr antilysozyme monoclonal antibodies and of the com$es between onv of the antibodies and its el)itoF)tl. E’NHO J. 5, 4ldbU:i. Harrison. E. T. 8: Magr. .\I. (:. (1965). Isolat~ion and characterisation of’shrrp yl- anfl y,-immuno~lol~ulins and their yot&ytic f’ragments. /Gochin/. IC)@ys.
itctrr. 147. 52-59. Mat’thews. R. W. (1968). Solvent f*ontrnt of r~rotrin crystals. .J. A2101.Hid. 33. 491-497. Porter. R. R. (1959). The hydrolysis of’ rabbit, y-globulin and antibodies wit,h crystalline Jjapain. Hioc!~m. J. 73. 119~127. L~Yodawrr. A. & Hodgson. K. 0. (1975). Crystallisation and farystal data of Monellin. f’roc. Snt. drrrrl. ,Yci.. I-.S.d. 72, 398-399.
Riol.
(1992) 223. 381-382
Crystallization of the Fab Fragments of Anti-peptide Monoclonal Antibodies and a Complex with Peptide Robert E. Griest-j-, Philip D. Jeffreyt, Garry L. Taylor?,5 and Anthony R. Rees-fLaboratory of Molecular Biophysics The Rex Richards Building, South Parlcs Road &ford OX1 ,?Qli. Grea,t Britnin (Received 14 August 1991: accepted X October 1991) The antigen-binding fragments of four monoclonal antibodies that cross-react with both the “loop” peptide of hen egg-white lysozyme (residues 57 to 84) against which they were raised, and with the native protein (HEL) have been crystallized. One of these fragments also crystallizes as a complex with the peptide antigen. Krywords:
crystallization:
lysozyme;
antibodies;
The preparation of a panel of monoclonal antibodies against the “loop” peptide. comprising residues 57 to 84 of HEL, and the characterization of their epitopes has been reported (Darsley & Rees? 1985n,b). These monoclonal antibodies were obtained from mice immunized with a synthetic conjugate of a proteolytic fragment of HEL coupled to bovine serum albumin, and bind both whole native HEL and the loop peptide fragment. We have been att,empting to crystallize the uncomplexed antigen binding fragment (Fab), Fabpeptide and Fab-HEL complexes. Such a series of three-dimensional structures would elucidate the mechanisms and differences in binding intact protein and the more flexible peptide antigen. Additionally, the Fab structures would provide a test of the “maximum overlap” method used to model the Fv regions of the Gloop antibodies (de la Paz et al., 1984). Protein for crystallization was prepared with hpbridomas grown as ascites tumours in Fl hybrid SWRxRalb/c mice primed with pristane. The antibodies were purified from the ascites fluid by SO?/; saturated ammonium sulphate precipitation followed by ion-exchange chromatography on diethylaminoethyl-Sephadex (Harrison & Mage, 1967). Antibodies were digested with papain
peptide
antigen;
Fab fragments
(Porter. 1959) using an IgG to enzyrne ratio of 500 : 1 (w/w) at pH 7.0 for t’he readily cleaved antibody Gloop4. 100: 1 at pH 7.0 for Gloopl and GloopP. and 20: 1 at pH 5.5 for the more resistant Gloop5, all for four hours at 37°C. The cleavage reaction was st,opped with 3 mg of iodoacetamide per millilitre of reaction mix. Fab fragments were separated from partially digested IgG and Fc fragments by passing the concentrated digest through a 75 cm x 2.6 cm S-200 gel filtration column (20 mMsodium phosphate buffer, pH 7.5). This purificat.ion scheme gives complete separation of Fab from Fc fragments and partially digested IgG. Analysis by high-resolut,ion ion-exchange chromatography (Mono-Q column, Pharmacia) revealed t’hat some digests contained more than one ionic species of Fat). In particular, Gloop4 only crystallized after further purification on a Mono-Q column. The Fab fragments were subjected to crystallization trials in hanging drops at room temperature using the method of vapour diffusion (Wlodawer & Hodgson, 1975). All buffers and solutions used were filtered through 0.45 pm Millio filters. Protein concentrations from 1.5 to 20 mg/ml in 0.01 to 004 MTris. HCl or sodium phosphate buffers were used. Polyethylene glycol 6000 (Sigma Co.) was used as the precipitant. Initial concentrations in the drop of 5 to 1004 (w/v) polyethylene glycol were equilibrated over wells of 15, 20 and 25% (w/v) polvet,hylene glycol. The effects of both pH (5.0 to 9.5. in steps of 0.5) and NaCl concentration (60 to 0.4 M in 0.1 M steps) were tested. Gloopl (IgG2a,K), Gloop2 and Gloop4 (both IgC2b,rc) all crystallized within two weeks at pH 6.0 to pH 7.5 and after longer periods Gloopl also crystallized a,t as low as pH 5.0. There was a trend of
t Present address: Department of Biochemistry. University of Bat,h, Claverton Down, Bath BA2 TAT, Great Britain. 1 Present address: Bristol-Myers Squibb Pharmaceutical Research Institute. P.0. Box 4000, Princeton, NJ 08543-4000, U.S.A. Q Author to whom all correspondence should be addressed.
381 OOZZ--2836/92/010381-02
$03.00/O
0 1992 Academic
Press Limited
mow nucleation sites and smaller crystals at both lower pH and higher salt concentration. forming clusters of very thin plates or thin branched needles. There was a requirement for some salt. for in the absence of Na(‘l, crystallization seldom occurred and then only after several months. producing cryst,als with irregular morphologies. The best conditions cent,red around pH 7.0 to pH 7.5 with 0.1 wXaC1 present: Qloopl gavel prisms (up to 14 mm x 0.4 mm x 0.2 mm): C:loopP produced long regular needles (up t’o 1.5 mm x 0~25 mm x 0.15 mm); and Cloop4 gave similar needles (up to 0.2 mm X 0.1 mm X 0.1 mm). (:loops (IgGl,K) behaved rather differently and usually formed an oily phase during crystallization trials and has onI> twice ever yielded crystals, one of which (I.4 mm x (h3 mm x 0.3 mm) was suitable for data collection. (:loop2 presented a different problem, in that two of the cryst,al forms are morphologically indistinguisb able until after data collection and processing. ljatasets for each crystal form were collecated on a Nicolet/Xentronics area detector and processed using the XENGEN suite of programs (A. Howard. 19X8, Genex (‘orporation. Gaithersburg. MD. IT.S.A,). Table 1 lists the space groups and diffrav Con limits of the crystal forms characterized. Synthetic loop peptide has been used in cocrystallization attempts with Fah fragments: both (:loop2 and Gloop5 have yielded crystals morphologically distinct from those grown under similar conditions in t’he absence of peptide. Growth ot GloopSpeptide crystals in t,he presence of *4(ilabelled peptide has confirmed the presence of peptide in these crystals. The Gloop2 complex diffracts to at least 4.0 !I (1 Lh = 0.1 rim). but the space group has not been characterized. Att,empts to soak peptide into uncomplexed Fab cryst)als resulted in the disordering of these crystals. Molecular replacement results for the three
crystal forms of nncomplexed (sloop:! and for Gloopl indicate that there are t,wo Fa,b molcc~ules per cell for ea,ch of t’hese crystal forms. I’,,, values (Matthews. 1968) listed in Table 1 suggest four and 12 molecules per cell for Gloop4 and Gloop5, respt+ timely. The structures of Gloopl and thv three crystal forms of (:loop:! have been determined and will be reported elsewhere. The authors thank Professor Sir David L’hilliJjs for his support mtl rrr~ouragement. L$:r arr grateful to I)rs Elspeth (barman. Yvonne Jones and David Stuart, for their assistance in data collection and processing. and to Dr Andrew Howard for pointing out the init,ial incorrrc~t assignment of the (iloop spare’ group.
References Daralry. 31. J. & Rees. A. It. (198.5a). Sucleotidr seyurnf&es of‘ tive anti-lysozyme monof~lonal antibodies. EXBO -1. 4. 33!%398. IIarsley. M. .I. & Rees. A. R. (19856). Three distinct epltopes within the lofq) region of hen egg Ivsozymr defined with rnonorlonal antibodies. EM/%7 -1. 4. X-392. dr la Paz, P.. Sut’ton. 13. .J.. Parsley, hl. .I. & Rers. A. R. (1984). ModrIling of’thr fvmbining sites of’thrrr antilysozyme monoclonal antibodies and of the com$es between onv of the antibodies and its el)itoF)tl. E’NHO J. 5, 4ldbU:i. Harrison. E. T. 8: Magr. .\I. (:. (1965). Isolat~ion and characterisation of’shrrp yl- anfl y,-immuno~lol~ulins and their yot&ytic f’ragments. /Gochin/. IC)@ys.
itctrr. 147. 52-59. Mat’thews. R. W. (1968). Solvent f*ontrnt of r~rotrin crystals. .J. A2101.Hid. 33. 491-497. Porter. R. R. (1959). The hydrolysis of’ rabbit, y-globulin and antibodies wit,h crystalline Jjapain. Hioc!~m. J. 73. 119~127. L~Yodawrr. A. & Hodgson. K. 0. (1975). Crystallisation and farystal data of Monellin. f’roc. Snt. drrrrl. ,Yci.. I-.S.d. 72, 398-399.
