An Antibody Delivery System for Regulated Expression of Therapeutic Levels of Monoclonal Antibodies In Vivo

Abstract

Monoclonal antibody (mAb) delivery by gene transfer in vivo may be an attractive alternative to current mAb therapies for applications that require long-term therapy. This article describes a transfer system that allows inducible high-level expression of unmodified mAbs in vivo. A recombinant adeno-associated viral (rAAV) vector is used that comprises an expression cassette consisting of a dimerizer-regulated promoter that drives expression of the antibody heavy and light chains linked by a 2A self-processing peptide and a furin cleavage site. Following intravenous injection of the rAAV vector, serum mAb levels >1 mg/ml were attained by administration of the inducer, rapamycin. Antibody expression could be rapidly shut off by discontinuing treatment with rapamycin. By optimizing the furin cleavage sequence, this system generated native antibody in vivo, decreasing the likelihood of a host immune response to foreign sequences. In summary, this optimized mAb expression system allows regulated high-level expression of native full-length mAbs in vivo and may offer a new opportunity for delivery of therapeutic mAbs in the clinic. Monoclonal antibody (mAb) delivery by gene transfer in vivo may be an attractive alternative to current mAb therapies for applications that require long-term therapy. This article describes a transfer system that allows inducible high-level expression of unmodified mAbs in vivo. A recombinant adeno-associated viral (rAAV) vector is used that comprises an expression cassette consisting of a dimerizer-regulated promoter that drives expression of the antibody heavy and light chains linked by a 2A self-processing peptide and a furin cleavage site. Following intravenous injection of the rAAV vector, serum mAb levels >1 mg/ml were attained by administration of the inducer, rapamycin. Antibody expression could be rapidly shut off by discontinuing treatment with rapamycin. By optimizing the furin cleavage sequence, this system generated native antibody in vivo, decreasing the likelihood of a host immune response to foreign sequences. In summary, this optimized mAb expression system allows regulated high-level expression of native full-length mAbs in vivo and may offer a new opportunity for delivery of therapeutic mAbs in the clinic. Monoclonal antibodies (mAbs) are an important class of therapeutic agents used in the treatment of a wide variety of diseases, including cancer, autoimmune disorders, and infectious disease. To date, 20 mAb therapeutics have been approved by the regulatory authorities, and many others are currently in clinical development. Most of these therapies require frequent administration of relatively high doses of the mAb, often via intravenous perfusion, since high and persistent serum levels of antibody are frequently required for optimal clinical efficacy. Antibody delivery using a vector-based gene transfer system in vivo has the potential to address the need for long-term treatment with mAbs. We have previously described a system for the constitutive expression of full-length two-chain antibodies from a single open reading frame (ORF) by linking the heavy and light antibody chains with a 2A self-processing sequence and a furin cleavage site.1Fang J Qian J Yi S Harding TC Tu G VanRoey M et al.Stable antibody expression at therapeutic levels using the 2A peptide.Nat Biotechnol. 2005; 23: 584-590Crossref PubMed Scopus (313) Google Scholar By using an adeno-associated virus 8 (AAV8) vector containing this novel antibody expression cassette, long-term expression of serum mAb levels of >1 mg/ml was achieved in mice after a single vector administration. However, several improvements may be necessary if such an in vivo gene expression system is to have broad clinical utility.2Marasco WA Therapeutic antibody gene transfer.Nat Biotechnol. 2005; 23: 551-552Crossref PubMed Scopus (11) Google Scholar,3Secko D Immunotherapy made more accessible?.CMAJ. 2005; 173: 144Crossref PubMed Scopus (2) Google Scholar First, antibody expression from the original vector was driven by a constitutive cytomegalovirus-based promoter, whereas regulated expression would be highly desirable for the therapeutic use of mAbs; i.e., a means should be provided to switch the expression system off in the event of adverse events and to titrate levels into an optimal range. Second, the antibody heavy chain protein from the previous system contained two additional amino acids at its C-terminus, derived from the furin cleavage site (H chain + Arginine and Alanine),which could potentially alter the activity of the antibody or result in the induction of anti-idiopathic antibodies. Thus, expression of unmodified, native full-length mAbs is highly desirable. This article describes a regulated antibody delivery system that allows expression of high levels of mAbs in vivo. This system was modified from ARIAD's dimerizer-regulated gene regulation system in which all protein components are of human origin and expression is controlled by rapamycin or analogs with reduced immunosuppressive activity.4Rivera VM Clackson T Natesan S Pollock R Amara JF Keenan T et al.A humanized system for pharmacologic control of gene expression.Nat Med. 1996; 2: 1028-1032Crossref PubMed Scopus (454) Google Scholar,5Rivera VM Gao GP Grant RL Schnell MA Zoltick PW Rozamus LW et al.Long-term pharmacologically regulated expression of erythropoietin in primates following AAV-mediated gene transfer.Blood. 2005; 105: 1424-1430Crossref PubMed Scopus (235) Google Scholar Using a rapamycin-regulated promoter to drive the mAb expression cassette, antibody expression levels of >1 mg/ml were achieved in multiple induction cycles. Importantly, mAb production was efficiently turned off after discontinuation of the inducer drug. Moreover, with the optimization of the mAb expression cassette and furin cleavage sequence, the two additional amino acids at the C-terminus of the heavy chain were eliminated, resulting in the expression of a full-length native mAb product. We were able to demonstrate that no antibodies to the 2A sequence were induced by the use of this system in immune-competent mice. In summary, this optimized mAb expression system allows regulated high-level expression of native full-length mAbs in vivo and may offer a new opportunity for delivery of therapeutic mAbs in the clinic. Previous studies demonstrated that full-length mAbs can be efficiently generated from a single ORF by linking mAb heavy and light chains with a 2A self-processing sequence derived from foot and mouth disease virus and a furin cleavage site. The furin cleavage site sequence used in the original cassette was RAKR.1Fang J Qian J Yi S Harding TC Tu G VanRoey M et al.Stable antibody expression at therapeutic levels using the 2A peptide.Nat Biotechnol. 2005; 23: 584-590Crossref PubMed Scopus (313) Google Scholar Mass spectrometry analysis of the antibody heavy chain generated from such an expression construct showed that it contained two additional amino acid residues, Arginine and Alanine, at its C-terminus. For many secreted proteins, C-terminal sequences are modified by carboxypeptidases (CPs), which remove the basic amino acids, K or R, from the C-termini.6Bowler RP Nicks M Olsen DA Thogersen IB Valnickova Z Hojrup P et al.Furin proteolytically processes the heparin-binding region of extracellular superoxide dismutase.J Biol Chem. 2002; 277: 16505-16511Crossref PubMed Scopus (59) Google Scholar,7Fricker LD Carboxypeptidase E.Annu Rev Physiol. 1988; 50: 309-321Crossref PubMed Scopus (311) Google Scholar The Arginine and Alanine residues of the heavy chain derived from the original expression cassette would not be removed by CP because the second amino acid in the furin cleavage site sequence RAKR is the non-basic amino acid A. However, the furin cleavage site consensus sequence R-X-K/R-R allows flexibility in choosing the amino acids at positions two and three, which permits the evaluation of the efficiency of cleavage of alternative sequences that could result in more homogeneous and unmodified mAbs. Four antibody expression cassettes with the furin cleavage site sequences RRRR, RKRR, RRKR, or RKKR were constructed (Table 1). To avoid heterogeneity at the last amino acid of the heavy chain, which is commonly observed with current mAb production technologies owing to incomplete CP cleavage,8Lewis DA Guzzetta AW Hancock WS Costello M Characterization of humanized anti-TAC, an antibody directed against the interleukin 2 receptor, using electrospray ionization mass spectrometry by direct infusion, LC/MS, and MS/MS.Anal Chem. 1994; 66: 585-595Crossref PubMed Scopus (87) Google Scholar a construct consisting of the heavy chain sequence with the last lysine deleted, followed by a RKRR furin cleavage site, the 2A sequence, and the light chain, was also constructed. This construct was designated (ΔK)RKRR (Table 1). For rapid assessment of the heavy chain C-terminal sequence of antibodies expressed from these constructs, human immunoglobulin IgG4 sequences were used for all cassettes, because they contain a methionine near the C-terminus within the heavy chain sequence that allows cleavage by cyanogen bromide (CNBr) to generate a C-terminal fragment that is suitable in size for mass spectrometry analysis. Plasmids containing the antibody expression cassette driven by a cytomegalovirus-based promoter (CAG promoter) were transfected into Chinese hamster ovary cells. Antibody products from supernatants were purified and separated by sodium dodecyl sulfate polyacrylamide gel electrophoresis. The heavy chain bands were excised from the gel and digested with CNBr for mass spectrometry analysis. The mass spectrometry analysis showed that native antibodies, without additional amino acid residues remaining at the C-terminus of the heavy chain, were generated in >99% of mAb protein from expression cassettes that contained three different furin cleavage sites, RRRR, RKRR, and RRKR (Table 1). In contrast, the antibody product from the RKKR furin cleavage site expression cassette exhibited only partial cleavage, with various C-terminal residues of RKK in approximately 18.3% of antibody protein product (Table 1) as estimated by mass spectrum peak areas. The construct containing (ΔK)RKRR also resulted in the complete removal of the extra amino acid residues at the heavy chain C-terminus. Thus, furin cleavage sites consisting of all basic amino acids facilitated expression of homogeneous native full-length antibody product in vitro from HF2AL expression cassettes using a single ORF.Table 1Mass spectrometry analyses of C-terminal amino acid residues in antibody heavy chains derived from the HF2AL constructs containing various furin cleavage sitesConstructCleavage efficiencyH chain + RRRRComplete cleavageH chain + RKRRComplete cleavageH chain + RRKRComplete cleavageH chain + RKKRComplete: 81.7% +R:6.4% +RK:6.8% +RKK:5.1%H chain (ΔK) + RKRRComplete cleavagePercentages of cleavage were estimated by peptide peak areas in mass spectra. Complete cleavage: C-terminal fragment without additional amino acid residues in >99% of antibody product. Open table in a new tab Percentages of cleavage were estimated by peptide peak areas in mass spectra. Complete cleavage: C-terminal fragment without additional amino acid residues in >99% of antibody product. The dimerizer-regulated gene regulation system was used to regulate mAb expression from recombinant adeno-associated viral (rAAV) vectors. This system has three individual components: the activation domain fusion protein, the DNA-binding domain fusion protein, and the inducible promoter followed by the transgene sequence.4Rivera VM Clackson T Natesan S Pollock R Amara JF Keenan T et al.A humanized system for pharmacologic control of gene expression.Nat Med. 1996; 2: 1028-1032Crossref PubMed Scopus (454) Google Scholar,9Ye X Rivera VM Zoltick P Cerasoli Jr., F Schnell MA Gao G et al.Regulated delivery of therapeutic proteins after in vivo somatic cell gene transfer.Science. 1999; 283: 88-91Crossref PubMed Scopus (274) Google Scholar,10Rivera VM Ye X Courage NL Sachar J Cerasoli Jr, F Wilson JM et al.Long-term regulated expression of growth hormone in mice after intramuscular gene transfer.Proc Natl Acad Sci USA. 1999; 96: 8657-8662Crossref PubMed Scopus (166) Google Scholar The DNA-binding domain and activation domain fusion proteins dimerize in the presence of the inducer molecule, resulting in the activation of the promoter. Owing to the limited packaging capacity of rAAV vectors (<5,000 base pairs), all three regulatory elements plus the antibody expression cassette could not be fit into a single rAAV vector construct. A dual vector system that included the regulatory elements and antibody HF2AL ORF in two separate vectors was therefore developed (Figure 1). The first vector construct, rAAV BiDi AD-DBD, encodes the activation domain and the DNA-binding domain, which are driven by a bi-directional promoter as described by Nguyen et al.11Nguyen M Tu G Gonzalez-Edick M Rivera VM Jooss KU et al.Rapamycin-regulated control of anti-angiogenic tumor therapy following rAAV-mediated gene transfer.Mol Ther. 2007; (published online 23 January 2007 (doi:10.1038/sj.mt.6300079))Google Scholar The second construct, rAAV8 Z8IL2p DC101 HF2AL, consists of a regulated promoter followed by the HF2AL ORF for the DC101 mAb, a rat mAb against mouse vascular endothelial growth factor receptor 2 developed by Imclone Systems (New York, NY).12Prewett M Huber J Li Y Santiago A O'Connor W King K et al.Antivascular endothelial growth factor receptor (fetal liver kinase 1) monoclonal antibody inhibits tumor angiogenesis and growth of several mouse and human tumors.Cancer Res. 1999; 59: 5209-5218PubMed Google Scholar This mAb was used as a demonstration antibody since its activity is readily evaluated in mice. Based on the results obtained from the optimization study on furin cleavage sites, the (ΔK)RKRR sequence was used in the AAV8 HF2AL construct for regulated antibody expression. Antibody expression occurs when an individual cell is simultaneously transduced with the two vectors in the presence of the inducer rapamycin. The regulated expression system was first evaluated in vitro by transient transfection of the rAAV plasmids into Huh7 cells. In the presence of rapamycin, this system resulted in a more than 100-fold induction of antibody expression (data not shown). AAV8 was chosen as the gene transfer vector on the basis of its significant liver- and muscle-mediated gene transfer efficiency.13Gao GP Alvira MR Wang L Calcedo R Johnston J Wilson JM Novel adeno-associated viruses from rhesus monkeys as vectors for human gene therapy.Proc Natl Acad Sci USA. 2002; 99: 11854-11859Crossref PubMed Scopus (1232) Google Scholar,14Wang Z Zhu T Qiao C Zhou L Wang B Zhang J et al.Adeno-associated virus serotype 8 efficiently delivers genes to muscle and heart.Nat Biotechnol. 2005; 23: 321-328Crossref PubMed Scopus (502) Google Scholar rAAV8 BiDi AD-DBD and rAAV8 Z8IL2p DC101 HF2AL vectors were generated as described in Materials and Methods. The vectors were mixed at a 1:1 ratio and injected intravenously (2.5 × 1011 vector genomes (vg)/mouse per vector) into NCR nu/nu mice on day 0. From day 18, mice were injected intraperitoneally twice weekly for 3 weeks with 3 mg/kg of rapamycin to induce expression of the mAb (Figure 2a). Serum levels of mAb were measured on day 17, before treatment with rapamycin, to obtain baseline antibody serum levels. Low levels (<0.2 μg/ml) of DC101 mAb were detected, suggesting that the promoter was not significantly leaky in the absence of the inducer. In contrast, rapamycin treatment resulted in a rapid increase of serum mAb levels, which reached 1,000 μg/ml 2 weeks after initiation of rapamycin. Once serum mAb levels reached a plateau 3 weeks after initiating induction, rapamycin treatment was discontinued to evaluate how quickly mAb levels returned to baseline levels. mAb levels returned to <1 μg/ml by 4 weeks after discontinuation of the inducer drug. Similar results were seen in two additional induction cycles, demonstrating the consistency of the inducer effect (Figure 2a). After the last induction cycle, baseline levels of serum mAb levels (<0.2 μg/ml) were reached 6 weeks after discontinuation of rapamycin treatment (Figure 2a). Thus, the regulated mAb expression system described allows rapamycin-induced mAb expression to increase to levels approximately 4 logs over baseline in vivo as well as the complete shut-off of expression after discontinuation of rapamycin administrations. Next it was determined how serum mAb levels were affected by vector or rapamycin dose. rAAV8 BiDi AD-DBD and rAAV8 Z8IL2p DC101 HF2AL vectors were administered intravenously at either 2.5 × 1011 vg/mouse per vector (high dose) or 5 × 1010 vg/mouse per vector (low dose). Each group received either a high (3 mg/kg) or a low (0.3 mg/kg) dose of the inducer rapamycin, and serum mAb levels were followed over time. As shown in Figure 2b, serum mAb levels were dependent on both the rAAV vector and rapamycin doses. In animals that received the high rAAV vector and rapamycin dose, maximum serum mAb levels of >2,400 μg/ml were achieved. In the group that was injected with the high rAAV vector but low rapamycin dose, mAb levels peaked at 1,300 μg/ml. Serum mAb levels in mice that received the low rAAV vector dose and either the high or low rapamycin dose were significantly lower, reaching peak levels of only 20 μg/ml, suggesting that serum mAb levels are more affected by rAAV vector dose than rapamycin dose. Since the (ΔK)RKRR construct resulted in complete removal of the additional amino acids from the C-terminus of the heavy chain in vitro, this expression cassette was used to evaluate the homogeneity of the DC101 mAb heavy chain after in vivo gene transfer. To allow purification of DC101 mAb from mouse serum for mass spectrometry analysis, a construct including a 6xHistidine tag (His-tag) attached to the C-terminus of the antibody light chain was generated (Figure 1c). The mAb with the His-tag can be purified from mouse serum by nickel or cobalt metal affinity column purification after gene transfer. Unlike human IgG, the DC101 antibody, which is a rat IgG1 antibody, does not contain a methionine near the C-terminus of the crystallizable fragment portion of the antibody. To facilitate C-terminal sequence analysis, the construct was mutated by changing a leucine (L) to a methionine at position 19 from the C-terminus of the heavy chain (Figure 3a). This mutated methionine was used for CNBr digestion to generate a C-terminal peptide that is conveniently sized for mass spectrometry analysis (Figure 3a). An rAAV8 vector encoding this modified DC101 antibody ORF, including the furin cleavage site (ΔK)RKRR, under the control of a CAG promoter was generated and injected at 2.5 × 1011 vg/mouse into NCR nu/nu mice intravenously or intramuscularly. Three weeks after AAV injection, serum DC101 mAb levels reached 600 μg/ml as determined by enzyme-linked immunosorbent assay (ELISA). The mAb was purified from the serum using a cobalt affinity column, and western blot analysis showed heavy and light chain bands with expected molecular weights at 55 and 25 kd without additional bands, suggesting that 2A and furin cleavage are complete (Figure 3b). For mass spectrometry analysis, the heavy chain band was excised from the sodium dodecyl sulfate polyacrylamide gel electrophoresis gel and subjected to CNBr digestion. A single peptide peak with a molecular weight at the expected size of 2,004 d was detected following either intravenous or intramuscular administration of the rAAV vectors (Figure 3d and e). No evidence was seen in mass spectrometry analysis of heavy chain C-terminal peptides containing additional amino acid residues derived from the furin cleavage site (i.e., -R, -RK, -RKR, -RKRR; see Figure 3c). In summary, in vivo gene transfer to muscle or liver of an rAAV8 vector expressing DC101 from the (ΔK)RKRR expression cassette results in high-level expression of unmodified/native full-length mAb. To evaluate potential host immune responses to the 2A peptide derived from foot and mouth disease virus, an rAAV8 CAG HF2AL vector consisting of a constitutive CAG promoter, the furin cleavage sequence RAKR, and the 2A self-processing sequence was injected intravenously into immune-competent C57BL/6 mice at doses of 2 × 1011 or 0.5 × 1011 vg/mouse. Sera were evaluated at different time points for up to 196 days for DC101 expression levels as well as for the presence of anti-2A-specific antibodies. Serum DC101 levels up to 2,900 and 350 μg/ml were reached in the high- and low-dose-treated mice, respectively (Figure 4a), but no anti-2A peptide–specific antibodies were detected on evaluation by a 2A-specific ELISA assay (Figure 4b). In addition, stable mAb expression in immunocompetent mice for at least 6 months suggests that 2A peptide–specific T-cell responses were also not elicited in animals after gene transfer. T-cell responses to any of the expression cassette sequences would be expected to eliminate vector-transduced cells, leading to transient gene transfer and consequent declines in the level of mAb expression. In this study a regulated expression system was developed that allows inducible high-level expression of a full-length native mAb following a single intravenous injection of a pair of rAAV8 vectors. Serum mAb levels of >1,000 μg/ml were reproducibly achieved in multiple induction cycles using rapamycin as the inducer. Furthermore, serum mAb levels were stable for at least 8 weeks (the duration of the experiment) when rapamycin was continuously administered. A similar regulated promoter has been used in the context of AAV-mediated gene transfer in non-human primates to induce expression of erythropoietin for more than 6 years without apparent host immune responses to the regulatory proteins or the therapeutic protein.5Rivera VM Gao GP Grant RL Schnell MA Zoltick PW Rozamus LW et al.Long-term pharmacologically regulated expression of erythropoietin in primates following AAV-mediated gene transfer.Blood. 2005; 105: 1424-1430Crossref PubMed Scopus (235) Google Scholar These observations suggest that the regulated mAb expression system described here potentially offers patients a lifelong mAb therapy that requires only a single administration of an rAAV vector. After intravenous administration, high liver transduction is observed using AAV8 vectors,13Gao GP Alvira MR Wang L Calcedo R Johnston J Wilson JM Novel adeno-associated viruses from rhesus monkeys as vectors for human gene therapy.Proc Natl Acad Sci USA. 2002; 99: 11854-11859Crossref PubMed Scopus (1232) Google Scholar,15Nakai H Fuess S Storm TA Muramatsu S Nara Y Kay MA Unrestricted hepatocyte transduction with adeno-associated virus serotype 8 vectors in mice.J Virol. 2005; 79: 214-224Crossref PubMed Scopus (253) Google Scholar which is important because only those hepatocytes that are transduced by both vectors simultaneously can express antibody upon rapamycin induction. The high DC101 expression levels achieved in this study indeed suggest that a high number of hepatocytes were transduced by both vectors. Notably, antibody expression from this system can be shut off rapidly, returning to baseline levels within 4–6 weeks after discontinuation of the inducer rapamycin. The time to return to baseline is, of course, likely to depend on both the half-life of the mAb generated and the peak expression levels achieved. Prolonged and uncontrolled expression of therapeutic proteins by gene transfer in patients is of concern,2Marasco WA Therapeutic antibody gene transfer.Nat Biotechnol. 2005; 23: 551-552Crossref PubMed Scopus (11) Google Scholar,3Secko D Immunotherapy made more accessible?.CMAJ. 2005; 173: 144Crossref PubMed Scopus (2) Google Scholar,16Bakker JM Bleeker WK Parren PW Therapeutic antibody gene transfer: an active approach to passive immunity.Mol Ther. 2004; 10: 411-416Abstract Full Text Full Text PDF PubMed Scopus (28) Google Scholar especially after stable gene transfer of novel therapeutic mAbs that have the potential to induce adverse events that would require discontinuation of therapy.17Adams GP Weiner LM Monoclonal antibody therapy of cancer.Nat Biotechnol. 2005; 23: 1147-1157Crossref PubMed Scopus (975) Google Scholar The mAb expression system described here can be readily switched on and off, offering a flexible means of generating high levels of therapeutic antibodies in vivo. Moreover, serum mAb levels can be adjusted to some extent by altering the rapamycin dose, offering an opportunity to fine-tune the system after gene transfer has occurred to achieve targeted serum mAb levels for therapeutic intervention. Use of the 2A self-processing sequence allowed the construction of a single rAAV vector that contains the entire mAb expression cassette capable of efficient mAb expression at high levels with a balanced heavy and light chain ratio.1Fang J Qian J Yi S Harding TC Tu G VanRoey M et al.Stable antibody expression at therapeutic levels using the 2A peptide.Nat Biotechnol. 2005; 23: 584-590Crossref PubMed Scopus (313) Google Scholar Importantly, the 2A sequence, which is derived from foot and mouth disease virus, did not elicit 2A-specific immune responses in immunocompetent animals following gene transfer of rAAV in vivo. In addition, no peptide derived from the 2A sequence was identified that strongly and stably binds to the HLA-A2 pocket when the 2A sequence was evaluated using an HLA-A2-specific epitope prediction program (data not shown), suggesting that this sequence most likely does not induce immune responses in HLA-A2-positive patients. In previous studies, the furin cleavage site RAKR was used to facilitate the removal of 2A peptide–derived amino acids from the antibody heavy chain during protein secretion.1Fang J Qian J Yi S Harding TC Tu G VanRoey M et al.Stable antibody expression at therapeutic levels using the 2A peptide.Nat Biotechnol. 2005; 23: 584-590Crossref PubMed Scopus (313) Google Scholar However, the final antibody product contained two additional amino acid residues derived from the furin cleavage sequence at the C-terminus of the antibody heavy chain that could affect the activity of the antibody. Accordingly, the use of alternative furin cleavage sequences consisting of only basic amino acids, which can be efficiently cut by CPs, was evaluated in the context of various mAb expression cassettes. It was demonstrated that the furin sequences RKRR, RRRR, and RRKR all resulted in homogeneous, unmodified antibody product in vitro. Use of the furin cleavage sequence RKRR resulted in unmodified antibody product even in animals in which high antibody serum levels of >600 μg/ml were achieved, suggesting that both furin- and CP-mediated cleavage are highly efficient in liver and muscle cells, which are preferred target organs for rAAV-mediated gene transfer. In fact, furin was originally added to the 2A-containing mAb expression cassette because furin is known for its efficient processing of pre-prohormones or other pre-proteins. After furin cleavage, the proteins generally contain 2–4 C-terminal basic amino acids (R or K) that are removed by CPs to generate the mature proteins. Thus, the 2A/furin-containing antibody expression cassettes evaluated in this study exploited a naturally occurring protein maturation machinery to produce high serum levels of native mAb from the HF2AL single ORF after rAAV-mediated gene transfer in vivo. Finally, the dimerizer-regulated gene regulation system has a number of advantages over other technologies that allow pharmacological regulation of protein expression in vivo. All the functional components of this system are derived from human proteins and therefore are less likely to be immunogenic in a clinical setting.5Rivera VM Gao GP Grant RL Schnell MA Zoltick PW Rozamus LW et al.Long-term pharmacologically regulated expression of erythropoietin in primates following AAV-mediated gene transfer.Blood. 2005; 105: 1424-1430Crossref PubMed Scopus (235) Google Scholar,10Rivera VM Ye X Courage NL Sachar J Cerasoli Jr, F Wilson JM et al.Long-term regulated expression of growth hormone in mice after intramuscular gene transfer.Proc Natl Acad Sci USA. 1999; 96: 8657-8662Crossref PubMed Scopus (166) Google Scholar Gene expression from this system is activated by the orally bioavailable small molecule rapamycin, which is approved as an immunosuppressive agent for organ transplantation. Furthermore, the oral route is convenient for long-term administration in patients. Although intermittent dosing of rapamycin is expected to be minimally immunosuppressive, the dimerizer system can also be controlled using rapamycin analogs with substantially reduced immunosuppressive activity.5Rivera VM Gao GP Grant RL Schnell MA Zoltick PW Rozamus LW et al.Long-term pharmacologically regulated expression of erythropoietin in primates following AAV-mediated gene transfer.Blood. 2005; 105: 1424-1430Crossref PubMed Scopus (235) Google Scholar In summary, the regulated antibody delivery system described in this study is able to generate native mAbs at therapeutic levels in vivo, offering a novel delivery system for long-term therapy of full-length mAbs. Plasmid construction. Plasmid rAAV BiDi AD-DBD was generated from pAATBi-Ci-sVEGFR1/R2 (ref. 1Fang J Qian J Yi S Harding TC Tu G VanRoey M et al.Stable antibody expression at therapeutic levels using the 2A peptide.Nat Biotechnol. 2005; 23: 584-590Crossref PubMed Scopus (313) Google Scholar) by removing the regulated promoter and transgene using standard cloning methods. To generate the rAAV plasmid encoding a regulated promoter and transgene, the rAAV backbone including the Z8IL-2p promoter was isolated from the pAATBi-Ci-sVEGFR1/R2 plasmid. The heavy and light chain sequences of the rat anti-mouse VEGFR2 antibody (DC101) were cloned from DC101 hybridoma cells (American Type Culture Collection, Manassas, VA).1Fang J Qian J Yi S Harding TC Tu G VanRoey M et al.Stable antibody expression at therapeutic levels using the 2A peptide.Nat Biotechnol. 2005; 23: 584-590Crossref PubMed Scopus (313) Google Scholar The single ORF of the antibody heavy chain, furine cleavage site, 2A, and antibody light chain was assembled by sewing polymerase chain reaction, using a 24–amino acid 2A sequence of APVKQTLNFDLLKLAGDVESNPGP.1Fang J Qian J Yi S Harding TC Tu G VanRoey M et al.Stable antibody expression at therapeutic levels using the 2A peptide.Nat Biotechnol. 2005; 23: 584-590Crossref PubMed Scopus (313) Google Scholar,18Ryan MD King AM Thomas GP Cleavage of foot-and-mouth disease virus polyprotein is mediated by residues located within a 19 amino acid sequence.J Gen Virol. 1991; 72: 2727-2732Crossref PubMed Scopus (347) Google Scholar The antibody HF2AL single ORF was flanked with EcoRI sites and cloned into the rAAV backbone downstream of the Z8IL-2p-regulated promoter. For optimization studies of furin cleavage sites, antibody HF2AL ORFs with various furin cleavage site sequences (RRRR, RKRR, RRKR, RKKR, and (ΔK)RKRR) and mutated heavy chain sequences were generated using the QuikChange Site-Directed Mutagenesis kit (Stratagene, La Jolla, CA) following the manufacturer's instructions. Cell culture and transfection. To evaluate mAb expression, plasmid DNA was purified from Escherichia coli using the Plasmid Mega kit (Qiagen, Valencia, CA) and transfected into HEK293 or Chinese hamster ovary cells (American Type Culture Collection, Manassas, VA) in 6-well tissue culture plates or 10-cm culture dishes with FuGene 6 (Roche, Indianapolis, IN), following the manufacturer's instructions. Twenty-four hours after transfection, cell culture medium was removed and the cells were re-fed. After 48 hours, the supernatants were collected and cell lysates were prepared using the Ripa buffer (1 mM ethylene glycol tetraacetic acid, 1% NP-40, 0.25% Na deoxycholate, and 2 mM vanadate) supplied with protease inhibitors (Roche, Indianapolis, IN). Antibody product in cell culture supernatants was purified by protein A or G column (GE Healthcare, Piscataway, NJ). Recombinant AAV vector preparation. HEK293 cells (American Type Culture Collection, Manassas, VA) were transfected using the calcium phosphate method with the pAAV vector plasmid in combination with the AAV8 serotype helper plasmid p5e18-VD2/8 (ref. 13Gao GP Alvira MR Wang L Calcedo R Johnston J Wilson JM Novel adeno-associated viruses from rhesus monkeys as vectors for human gene therapy.Proc Natl Acad Sci USA. 2002; 99: 11854-11859Crossref PubMed Scopus (1232) Google Scholar) and pXX-6 (ref. 19Xiao X Li J Samulski RJ Production of high-titer recombinant adeno-associated virus vectors in the absence of helper adenovirus.J Virol. 1998; 72: 2224-2232Crossref PubMed Google Scholar). Forty-eight hours after transfection, cells were harvested using phosphate-buffered saline/ethylenediaminetetraacetic acid (10 mM) and lysed by three freeze–thaw cycles in cell lysis buffer (150 mM NaCl, 50 mM HEPES, pH 7.6). Lysates were treated with 250 U/ml benzonase for 15 minutes at 37 °C and cellular debris was removed by centrifugation. The cleared cell lysate was fractionated by ammonium sulfate precipitation and the rAAV virions were isolated on two sequential CsCl gradients. The gradient fractions containing rAAV were dialyzed against sterile phosphate-buffered saline containing CaCl2 and MgCl2 and stored at −80 °C. Viral titers were determined by dot-blot analysis. Antibody expression in vivo by AAV vector–mediated gene transfer. Female NCR nu/nu and C57BL/6 mice (6–8 weeks old) were obtained from Taconic (Oxnard, CA). All mice were housed under specific pathogen-free conditions and treated according to the Institute for Laboratory Animal Research Guide for the Care and Use of Laboratory Animals. rAAV vectors at a specified vg per mouse were injected into mice via caudal veins (intravenous, 200 μl/injection) or quadriceps (intramuscular, 50 μl/injection). Mice were bled by alternate retro-orbital puncture at each scheduled time point. Rapamycin (LC Laboratories, Woburn, MA) was injected intraperitoneally at 3 or 0.3 mg/kg at the indicated time points. Rapamycin stock was prepared as a 25 mg/ml solution in N,N-dimethylacetamide (Sigma, St. Louis, MO). For formulation of the 3 mg/kg dose, 6 volumes of rapamycin stock were mixed with 10 volumes of PEG-400 (Sigma) and 84 volumes of 85% Tween-80 dissolved in distilled water. For formulation of the 0.3 mg/kg dose, the 3 mg/kg dose solution was diluted tenfold with a mixture of PEG-400 and 85% Tween-80 (1:9 vol/vol). For evaluation of the presence of an antibody response to the 2A peptide, rAAV8 CAG DC101 HF2AL vector, which contains the 2A sequence and a furin cleavage site RAKR, was injected into C57BL/6 immune-competent mice intravenously at 2 × 1011 vg/mouse. Mice were bled at various time points for evaluation of DC101 antibody serum levels and detection of 2A-specific antibodies by ELISA-based assays. ELISA and western blots. The DC101 antibody concentrations were determined using a commercial ELISA assay kit for rat IgG1 (Bethyl Laboratories, Montgomery, TX). For western blot analysis, proteins were fractionated by polyacrylamide gel electrophoresis using pre-cast Tris–glycine gels (Invitrogen, Carlsbad, CA) under reducing conditions and transferred to nitrocellulose membranes. The membranes were blocked with 5% non-fat dry milk and incubated with a goat anti-rat IgG antibody conjugated to horseradish peroxidase (EMD, San Diego, CA). Protein bands were visualized by exposure on X-ray films (Kodak, Rochester, NY)after the membranes were treated with ECL enhanced chemiluminescence solution (Pierce, Rockford, IL). For evaluation of anti-2A antibodies in sera, 100 μl of cell lysates (1 μg/ml total protein) from Chinese hamster ovary cells transfected with a human antibody HF2AL expression plasmid were coated on ELISA plates. Mouse sera were diluted to 1:50, added to the wells, and detected using a horseradish peroxidase–conjugated goat anti-mouse IgG (GE Healthcare, Piscataway, NJ). To serve as an anti-2A antibody positive control, rabbit anti-2A serum was generated by immunizing rabbits with synthesized 2A peptide (Genemed, South San Francisco, CA). Expression and purification of His-tagged DC101 antibody. To purify DC101 antibody (rat IgG1) from mouse serum, an AAV CAG DC101 HF2AL construct with 6xHis at the C-terminus of the light chain was constructed. This construct contained the DC101 heavy chain with a deleted terminal K, the furin cleavage site RKRR, 2A sequence, and DC101 light chain, driven by the constitutive CAG promoter.20Niwa H Yamamura K Miyazaki J Efficient selection for high-expression transfectants with a novel eukaryotic vector.Gene. 1991; 108: 193-199Crossref PubMed Scopus (4584) Google Scholar The rAAV8 vector for the mAb construct was prepared as described above and injected into NCR nu/nu mice at 2 × 1011 vg/mouse via caudal veins. Sera were harvested at day 21 and mAb concentration was determined by ELISA. His-tagged mAb was purified from mouse sera using Talon cobalt affinity resin (BD Clontech, Palo Alto, CA). C-terminal sequence analysis by mass spectrometry. Purified antibodies were fractionated by sodium dodecyl sulfate polyacrylamide gel electrophoresis under reducing conditions. The gel samples containing the antibody heavy chains were excised, sliced into 1 × 1 mm2 pieces, and processed for CNBr digestion. Analyses were performed on a Kratos Axima QIT (Shimadzu Instruments, Columbia, MD) matrix-assisted laser desorption/ionization mass spectrometer. Peptides were analyzed in positive ion mode in mid-mass range (700–3,000 d). Database searches were performed using Mascot (Matrix Sciences, Boston, MA). The authors would like to thank Peter Working for critical reading of the manuscript and Jeff Waugh, Debbie Farson, and the Vivarium staff at Cell Genesys for technical assistance. The authors also thank John Leszyk at the University of Massachusetts Medical School for mass spectrometry analysis and Tim Clackson and Victor Rivera at ARIAD Pharmaceuticals, Inc., for discussion and assistance.