Intracellular Stability of Anti-caspase-3 Intrabodies Determines Efficacy in Retargeting the Antigen

Although intracellular antibodies (intrabodies) are being explored as putative therapeutic and research reagents, little is known about the principles that dictate the efficacy of these molecules. In our efforts to address this issue, we generated a panel of five intrabodies, directed against catalytically inactive murine caspase-3, by screening single-chain antibody (Fv) phage display libraries. Here we determined criteria that single-chain Fv fragments must fulfill to act as efficient intrabodies. The affinities of these intrabodies, as measured by surface plasmon resonance, varied ∼5-fold (50–250 nm). Despite their substantial sequence similarity, only two of the five intrabodies were able to significantly accumulate intracellularly. These disparities in intracellular expression levels were reflected by differences in the stability of the purified protein species when analyzed by urea denaturation studies. We observed varied efficiencies in retargeting the antigen murine caspase-3, from the cytosol to the nucleus, mediated by intrabodies tagged with an SV40 nuclear localization signal. Our results demonstrate that the intrinsic stability of the intrabody, rather than its affinity for the antigen, dictates its intracellular efficacy. Although intracellular antibodies (intrabodies) are being explored as putative therapeutic and research reagents, little is known about the principles that dictate the efficacy of these molecules. In our efforts to address this issue, we generated a panel of five intrabodies, directed against catalytically inactive murine caspase-3, by screening single-chain antibody (Fv) phage display libraries. Here we determined criteria that single-chain Fv fragments must fulfill to act as efficient intrabodies. The affinities of these intrabodies, as measured by surface plasmon resonance, varied ∼5-fold (50–250 nm). Despite their substantial sequence similarity, only two of the five intrabodies were able to significantly accumulate intracellularly. These disparities in intracellular expression levels were reflected by differences in the stability of the purified protein species when analyzed by urea denaturation studies. We observed varied efficiencies in retargeting the antigen murine caspase-3, from the cytosol to the nucleus, mediated by intrabodies tagged with an SV40 nuclear localization signal. Our results demonstrate that the intrinsic stability of the intrabody, rather than its affinity for the antigen, dictates its intracellular efficacy. single-chain Fv caspase-3 murine caspase nuclear localization signal dithiothreitol nitrilotriacetic acid 3-[(3-cholamidopropyl)dimethylammonio[-1-propanesulfonic acid phosphate-buffered saline enzyme-linked immunosorbent assay response unit bovine serum albumin fluorescein isothiocyanate The design, cloning, expression, and use of single-chain antibodies (scFv)1 have now become routine in protein engineering. These moieties, although monovalent in comparison with their bivalent precursor IgG, do not suffer from the need to produce and assemble multichain oligomers like other members of the immunoglobulin family. scFv fragments have been shown to display similar (although not identical) kinetic and thermodynamic properties as Fab fragments of whole IgG (1Rajpal A. Kirsch J.F. Proteins Struct. Funct. Genet. 2000; 40: 49-57Crossref PubMed Scopus (7) Google Scholar). In recent years, filamentous phage display libraries of scFv molecules derived from the human IgG repertoire have become increasingly available (2Vaughan T.J. Williams A.J. Pritchard K. Osbourn J.K. Pope A.R. Earnshaw J.C. McCafferty J. Hodits R.A. Wilton J. Johnson K.S. Nat. Biotechnol. 1996; 14: 309-314Crossref PubMed Scopus (861) Google Scholar, 3Sheets M.D. Amersdorfer P. Finnern R. Sargent P. Lindquist E. Schier R. Hemingsen G. Wong C. Gerhart J.C. Marks J.D. Lindqvist E. Proc. Natl. Acad. Sci. U. S. A. 1998; 95: 6157-6162Crossref PubMed Scopus (381) Google Scholar, 4Pini A. Viti F. Santucci A. Carnemolla B. Zardi L. Neri P. Neri D. J. Biol. Chem. 1998; 273: 21769-21776Abstract Full Text Full Text PDF PubMed Scopus (386) Google Scholar). It is now possible to generate a scFv molecule against a target within a few weeks. However, since these libraries are usually generated from non-immunized sources and hence have not undergone affinity maturation, the isolation of high (subnanomolar equilibrium constant for dissociation) affinity scFv fragments from these selections is rare. Among the myriad applications of these molecules, the use of scFv fragments as intrabodies (intracellularly expressed antibodies) has received some attention recently. The aim of such attempts has been to neutralize the function of endogenous target proteins, using intrabodies, by several methods (5Marasco W.A. Gene Ther. 1997; 4: 11-15Crossref PubMed Scopus (86) Google Scholar, 6Lener M. Horn I.R. Cardinale A. Messina S. Nielsen U.B. Rybak S.M. Hoogenboom H.R. Cattaneo A. Biocca S. Eur. J. Biochem. 2000; 267: 1196-1205Crossref PubMed Scopus (62) Google Scholar, 7Zhu Q. Zeng C. Huhalov A. Yao J. Turi T.G. Danley D. Hynes T. Cong Y. DiMattia D. Kennedy S. Daumy G. Schaeffer E. Marasco W.A. Huston J.S. J. Immunol. Methods. 1999; 231: 207-222Crossref PubMed Scopus (50) Google Scholar). Among these, the misdirected localization of the target to another subcellular region features as the most popular methodology employed. These initial promising results suggest potential applications for intrabodies in functional genomics (8Visintin M. Tse E. Axelson H. Rabbitts T.H. Cattaneo A. Proc. Natl. Acad. Sci. U. S. A. 1999; 96: 11723-11728Crossref PubMed Scopus (169) Google Scholar) and gene therapy (9Marasco W.A. Chen S. Richardson J.H. Ramstedt U. Jones S.D. Hum. Gene Ther. 1998; 9: 1627-1642Crossref PubMed Scopus (37) Google Scholar, 10Deshane J. Siegal G.P. Alvarez R.D. Wang M.H. Feng M. Cabrera G. Liu T. Kay M. Curiel D.T. J. Clin. Invest. 1995; 96: 2980-2989Crossref PubMed Scopus (95) Google Scholar). The ability of intrabodies to generate phenotypic knockouts in vivo is now being explored in several organisms (11Ruberti F. Capsoni S. Comparini A. Di Daniel E. Franzot J. Gonfloni S. Rossi G. Berardi N. Cattaneo A. J. Neurosci. 2000; 20: 2589-2601Crossref PubMed Google Scholar, 12Hassanzadeh Gh G. Devoogdt N. Ghysen A. De Baetselier P. Muyldermans S. Dambly-Chaudière C. FEBS Lett. 1998; 437: 81-86Crossref PubMed Scopus (6) Google Scholar). The aim of this study was to determine the biochemical principles involved in the generation of effective intrabodies. We selected apoptosis (programmed cell death) as our model system, in which specifically caspase-3 was chosen as the target. Caspase-3 (CPP32/CASP3) is a member of a family of thiol proteases involved in programmed cell death (13Stennicke H.R. Salvesen G.S. Biochim. Biophys. Acta. 1998; 1387: 17-31Crossref PubMed Scopus (357) Google Scholar, 14Los M. Wesselborg S. Schulze-Osthoff K. Immunity. 1999; 10: 629-639Abstract Full Text Full Text PDF PubMed Scopus (372) Google Scholar, 15Porter A.G. Jänicke R.U. Cell Death Differ. 1999; 6: 99-104Crossref PubMed Scopus (2858) Google Scholar, 16Xue D. Shaham S. Horvitz H.R. Genes Dev. 1996; 10: 1073-1083Crossref PubMed Scopus (280) Google Scholar, 17Fernandes-Alnemri T. Armstrong R.C. Krebs J. Srinivasula S.M. Wang L. Bullrich F. Fritz L.C. Trapani J.A. Tomaselli K.J. Litwack G. Alnemri E.S. Proc. Natl. Acad. Sci. U. S. A. 1996; 93: 7464-7469Crossref PubMed Scopus (693) Google Scholar). These proteases are initially synthesized as inactive proenzymes that are proteolytically activated by upstream activators upon apoptotic stimulation. Caspase-3 has been implicated as a downstream component of the apoptotic machinery and is involved in the proteolysis of several cellular components. Among the several moieties involved in apoptosis, at the time we initiated the project, it was most facile to generate caspase-3 protein in bulk quantities as antigen for the selection of scFv fragments. Other more attractive targets such as caspase-8 were difficult to overexpress in Escherichia coli due to loss of protein through aggregation. A panel of scFv fragments was isolated from a phage display library directed toward the antigen murine caspase-3 (mCASP3). These scFv fragments were tested for properties such as affinity for the antigen, protein stability, and intracellular accumulation to determine characteristics that define an efficient intrabody. In conjunction with these studies, the scFv fragments fused with an SV40 nuclear localization signal (NLS) were utilized to retarget antigen from the cytoplasm to the nucleus in an effort to correlate intracellular behavior with biochemical parameters. Our data suggest that the stability of an intrabody is the most significant attribute, more so than affinity for the antigen, in dictating its effectiveness intracellularly. The intracellular stability of an intrabody was also found to correlate with the stability of the purified protein, as measured by urea denaturation under standard and reducing conditions. To our knowledge, this is the largest panel of scFv fragments against a common antigen that have been tested for intracellular stability and efficacy as intrabodies thus far. 293T transformed human embryonic kidney cells and BHK-21 baby hamster kidney cells were grown in Dulbecco's modified Eagle's medium (Life Technologies, Inc.) supplemented with 10% fetal bovine serum, 2 mm glutamine, and gentamycin (all from Life Technologies, Inc.). These cells were grown at 37 °C in 5% CO2. For Western blot and co-immunoprecipitation studies, 1 × 106 293T cells were plated on 100-mm dishes and transfected with 10 µg of DNA using 30 µl of Fugene-6 (Roche Molecular Biochemicals). Green fluorescent protein-pcDNA3.1 was used a control, and transfection efficiencies were determined by fluorescence-activated cell sorter analysis of trypsinized 293T cells. 3 × 104 BHK-21 cells were coated onto coverslips in six-well dishes. 2 µg of DNA was transiently transfected these cells using µl of and µl of (Life Technologies, Inc.). These cells were utilized for The gene was from a library using from The and were utilized to the gene expression The a and a at the of the The were used to generate catalytically inactive by the to and expression was by the which were grown in × 1 at 37 °C These were by 1 for at 30 The was by and by × 1 in of mm 2 mm mm 1 1 and 1 mm was by and the was in to of acid Inc.) was to the and at °C with The was by and with of mm 2 mm mm and The was in a and the protein was with of mm 2 mm mm mm and The was against mm and 2 mm The protein was onto a and with a in mm and 2 The for the scFv fragments were the and of expression a sequence that of the protein the It also two a and a were grown at 37 °C in medium supplemented with and The were isolated by in medium supplemented with 1 mm and and at 30 The cells were by in scFv mm mm 1 1 and 1 mm and purified with in the as that 2 mm was from the against and using the scFv fragments were purified by using a in a protein of the from this were analyzed by to scFv of of the scFv The of the scFv fragments was by of and phage display libraries were to scFv fragments (2Vaughan T.J. Williams A.J. Pritchard K. Osbourn J.K. Pope A.R. Earnshaw J.C. McCafferty J. Hodits R.A. Wilton J. Johnson K.S. Nat. Biotechnol. 1996; 14: 309-314Crossref PubMed Scopus (861) Google Scholar). µg of purified proteolytically inactive protein was coated in and two of selections were to a standard (2Vaughan T.J. Williams A.J. Pritchard K. Osbourn J.K. Pope A.R. Earnshaw J.C. McCafferty J. Hodits R.A. Wilton J. Johnson K.S. Nat. Biotechnol. 1996; 14: 309-314Crossref PubMed Scopus (861) Google Scholar). The from the libraries was grown in and for at 30 scFv were by the in mm and 1 mm These were for affinity for the target by on an were isolated and on their from these were analyzed by DNA by and 1998; PubMed Scopus Google Scholar) were in the The were with 1 and mm response of antibody antibody Inc.) was on the by standard These antibodies were used to of scFv fragments. of 10 to 1 were the at a of 10 and to with the The were to at equilibrium when the of the the by the of the scFv fragments from the data were by the the equilibrium for the medium was from the and the cells were with of were in 1 of mm and mm supplemented with Molecular at °C for The was by and with µg of antibody and µl of protein with constant at The were with being in The were by on and onto by standard Western blot analysis was with a of antibody and a of antibody antibodies at a were utilized in conjunction with the to the A of was with 10 and A of 10 was with the medium and to 293T cells transiently transfected with scFv fragments for The cells were and Western blot analysis with antibody was to determine the intracellular levels of the scFv fragments. were to A Scholar). of urea was in a of of The of the was determined by scFv were in of mm and mm and at denaturation was by the in the intrinsic at on a Life The data were and to denaturation as Biotechnol. Full Text PDF PubMed Scopus Google Scholar). The data were not used to our in the was to a of the scFv fragments isolated from the phage display was to an of intracellular protein stability by the urea denaturation in the of supplemented with mm as A. A. FEBS Lett. 1998; PubMed Scopus Google Scholar). the transiently transfected BHK-21 cells were with and in for at These were with and 10 mm for 10 The cells were with of by with a of in the with to five the cells were with and for 1 The antibodies were antibody S. and antibody in and and with the cells in a at 37 °C for 1 antibodies were by the coverslips with for 10 The cells were with 30 and antibodies Inc.) in at 37 °C for 30 the cells were with × 10 and in and with The cells were on a with a A catalytically inactive of murine caspase-3 was to was to of phage display scFv fragments the screening of expressed was on affinity by a 1 A and the in the and of the purified protein were in the of 3 of have been of protein were from the The of the protein was determined by The observed at is the of protein, as by of the not It is to this in the of a The determined by of the protein was similar to that from the acid sequence the of of the protein as a in as determined by not scFv fragments were by the protein as a target in of screening against phage display scFv libraries. for of the phage libraries were high the of the significantly These phage were and phage were the for the were than for the We to the selections to only two in to in the scFv fragments by the libraries. a of the from the two it was determined that only the library from this library were and to These scFv fragments were tested for antigen affinity by the that to affinity for the antigen signal The of these was tested using other and as antigen in similar The results that the isolated scFv fragments were for their antigen, with against the other caspase family members not The in these was determined by using 2 that the were of members from these were members within a by DNA However, upon DNA and acid with other the and were members from the five of scFv fragments were isolated and and expression The scFv fragments were in the and of and A protein sequence for the five scFv fragments selected is shown in The of the five scFv fragments are with the constant and the the the the in the and five scFv fragments were and in E. coli with to protein and were and analyzed by to the and of the purified protein not The of these scFv fragments was determined by to not The and of the signal sequence of five scFv fragments were determined by The affinity of these scFv fragments for the antigen was determined by surface plasmon scFv fragments were tested for using a antigen, as The scFv fragments affinity for the antigen tagged with an of to the on scFv and were utilized to on the to the of of the the equilibrium for were determined by the analysis kinetic was by kinetic behavior observed for some of the The affinities for the antigen varied from to for only and only an for this scFv The of is an since of for of and were with at to data for the in antigen with antibody and using scFv fragments as were We that the for these scFv fragments are in this system, the of the antigen the of the for of the scFv fragments and the antigen as determined by for an for an in a Western of transiently transfected cell was to initially determine the intracellular expression levels of the scFv fragments. that was the most the five scFv fragments. The for the of intracellular accumulation of the other scFv fragments was not of the of M. Biochem. Sci. 1996; Full Text PDF PubMed Scopus Google Scholar) not and the substantial sequence the scFv were utilized to the of of these intracellular scFv fragments. 293T cells with a of these we were able to the intracellular of the scFv fragments A and through The in is the denaturation were to determine the stability of scFv These were possible only with of the five scFv fragments due to the of for and In the of denaturation for and were similar and than that of These data with by Western and were the only scFv fragments However, the the expression levels of the two scFv fragments is not from the two similar under conditions. It is more with the data in reducing the of In the of reducing for was significantly than for and of the in stability of scFv fragments by urea denaturation in the of mm denaturation in the of mm in a To the of the scFv and a co-immunoprecipitation was 293T cells transiently with the intrabodies and a were and with antibody The of the were by and with antibody by Western The data demonstrate the and affinity of for its target as shown by co-immunoprecipitation and The was in and levels of in the from cells with the other scFv fragments The are the We to on the Western blot and that its intracellular levels by the methodology were to the intracellular The caspase-3 precursor has been shown to have a and subcellular in cells M. S. A. J. Cell Biol. 1998; PubMed Scopus (372) Google Scholar). To the and affinity of the scFv and an SV40 D. B. Richardson A. Full Text PDF PubMed Scopus Google Scholar) was fused to the of is a protein in and scFv in the nucleus, as in transiently and we observed the subcellular retargeting of the antigen that only to data are shown in in the we observed the retargeting of antigen the The in the of the in which the and nuclear are The data also demonstrate that an intrabody with an SV40 is to a protein The of this intracellular was by with antigen and another intrabody not with and The scFv fragments tagged with SV40 were with the antigen in BHK-21 cells to their intracellular efficacy in retargeting antigen from the cytoplasm to the data from for and scFv fragments are shown in A of and that the of retargeting efficiencies is of in the in the as by in the in efficiencies is by the of intracellular accumulation of the scFv fragments as by in the with scFv were due to expression levels of the intrabody. phage display libraries has become a standard for the isolation of affinity scFv fragments that molecule and (2Vaughan T.J. Williams A.J. Pritchard K. Osbourn J.K. Pope A.R. Earnshaw J.C. McCafferty J. Hodits R.A. Wilton J. Johnson K.S. Nat. Biotechnol. 1996; 14: 309-314Crossref PubMed Scopus (861) Google Scholar, 3Sheets M.D. Amersdorfer P. Finnern R. Sargent P. Lindquist E. Schier R. Hemingsen G. Wong C. Gerhart J.C. Marks J.D. Lindqvist E. Proc. Natl. Acad. Sci. U. S. A. 1998; 95: 6157-6162Crossref PubMed Scopus (381) Google Scholar, 4Pini A. Viti F. Santucci A. Carnemolla B. Zardi L. Neri P. Neri D. J. Biol. Chem. 1998; 273: 21769-21776Abstract Full Text Full Text PDF PubMed Scopus (386) Google Scholar). the of the library dictates the ability to for In our we found that the library scFv fragments substantial screening of and to the of in the of scFv fragments in this Technologies, The affinity of the scFv fragments isolated the from to and to the of selection this with the It is that a scFv with a affinity and in the library our of the five scFv fragments isolated from the phage display library significant intracellular accumulation of was than the other several for the intracellular expression of the scFv fragments. the scFv were in the expression expression was by the and the significant DNA sequence the that stability was to the of the levels of expression Our data that protein stability is the that the intracellular expression levels of these scFv fragments. The reducing of the cytoplasm Biol. Google Scholar) the of in scFv fragments S. F. M. P. Cattaneo A. 1995; PubMed Scopus Google Scholar). A of were to was to determine the of in the intracellular stability of Western of isolated from transiently transfected 293T cells the accumulation of to levels than the not the stability of scFv which expression levels the of It is possible that for the on the of intracellular protein as by A. A. FEBS Lett. 1998; PubMed Scopus Google Scholar). of scFv was by and determined to not The urea denaturation studies, in the of reducing the differences in intracellular stability the scFv fragments. for scFv fragments with to high stability in such reducing and a facile for effective intrabodies. We found that the intracellular stability of an intrabody dictates its efficacy more than its affinity for the was when we observed the in the antigen by scFv the stability of the nm). on the other is able to from the cytoplasm to the nucleus due to its substantial accumulation in the reducing of the results have been with intrabodies expressed in Q. Zeng C. Huhalov A. Yao J. Turi T.G. Danley D. Hynes T. Cong Y. DiMattia D. Kennedy S. Daumy G. Schaeffer E. Marasco W.A. Huston J.S. J. Immunol. Methods. 1999; 231: 207-222Crossref PubMed Scopus (50) Google Scholar) and A. A. D. A. A. A. J. Biol. Chem. 2000; Full Text Full Text PDF PubMed Scopus Google Scholar) cells. The in which and high affinity intrabodies become the of the stability and affinity parameters. to do this is by a phage display library on a more such as Western of subcellular from transiently transfected cells and on whole cell to determine the intracellular accumulation of isolated a of the to for functional intrabodies that their antigen in vivo (8Visintin M. Tse E. Axelson H. Rabbitts T.H. Cattaneo A. Proc. Natl. Acad. Sci. U. S. A. 1999; 96: 11723-11728Crossref PubMed Scopus (169) Google Scholar). Our ability to a and affinity scFv directed against caspase-3 with substantial intracellular that this methodology to generate intrabodies against other intracellular from our with Inc.) the isolation and efficacy of an intrabody in the antigen several subcellular Q. Zeng C. Huhalov A. Yao J. Turi T.G. Danley D. Hynes T. Cong Y. DiMattia D. Kennedy S. Daumy G. Schaeffer E. Marasco W.A. Huston J.S. J. Immunol. Methods. 1999; 231: 207-222Crossref PubMed Scopus (50) Google Scholar). In the we to the that the effectiveness of intrabodies by the stability, and intracellular accumulation of the largest panel of scFv fragments generated against an antigen thus in We found a the stability of the proteins, as measured by urea denaturation studies, and their intracellular our to the ability of our intrabody in endogenous caspase-3 the of the endogenous in have results not is not due to the of other in apoptosis other downstream such as and Our data other that isolated from display similar of apoptosis as their T. R. R.A. Proc. Natl. Acad. Sci. U. S. A. 1998; 95: PubMed Scopus Google Scholar). It is also possible that endogenous caspase-3 to the nucleus not to apoptosis and of its The need for functional of putative by has the of that fulfill this are several being explored to in this Our results of suggest that the of intracellular stability of the scFv fragments in the reducing of the cytoplasm are synthesized and is to their efficacy. directed at such the of intrabodies as an effective for functional We and for in the Hynes and for with the phage display system, A. J. for with the studies, and Cong for with the and the at We are also to of for on the