Protein Pharmaceuticals

Cure Arthritis Naturally

Cure Arthritis Naturally

Get Instant Access

Discovery and Preclinical Development

DavindcrS. Gill* Abstract

Proteins are natural molecules that carry out important cellular functions within our bodies. Their precise role is crucial to the maintenance of good health. Malfunctioning proteins or those not produced optimally result in disease. The foundation of biopharmaceutical drug therapy has therefore been to modulate cellular function by targeting specific proteins expressed on or outside the cell. Because most biopharmaceuticals are natural in origin, they are biologically and chemically very different from conventional medicines. In addition to differences in mechanism of action, biopharmaceuticals differ in the process by which they get manufactured and delivered. Because of their large, complex structure, they must often be produced by culturing cells and then purified from a host of cellular components. This can be time-consuming and cosdy. Also, most biopharmaceuticals are given by injection under the skin or by infusion into the veins. This creates significant limitations to their utility. Nonetheless, biopharmaceuticals can be very powerful and selective in disease applications such as in rheumatoid arthritis or cancer. This chapter describes methods by which proteins drugs are discovered, optimized and developed. It also covers novel agents and next generation proteins as well as some of the challenges and opportunities in the area.

Introduction

Most biopharmaceuticals today can be broadly classified into one of three major categories— monoclonal antibodies, fusion proteins or native biologies. Monoclonal antibodies are a growing class of Biopharmaceutical products that includes 21 FDA approved drugs to date.1 The list includes products such as Bevacizumab (Genentech), Panitumumab (Amgen), Rituximab (Biogenldec/Genentech) and Adalimumab (Abbott). In 2006, this class of drugs generated greater than $ 19 billion in annual sales worldwide, a figure that is expected to grow at the rate of 30-40% for the next 5 years.1"2 Fusion proteins are a class of drugs exemplified by Etanercept, the extra cellular domain of the TNFa receptor (p75) fused the Fc domain of immunoglobulin IgG. Since its launch in 1998, Etanercept is used for the treatment of chronic inflammatory diseases such as Rheumatoid arthritis, plaque psoriasis and alkylosing spondylitis. Other examples of this class of drugs include Alefacept (Biogenldec) for treatment of multiple sclerosis and Abatecept (Bristol-Myers Squibb) also for treatment of rheumatoid arthritis. Fusion proteins are an established class of drugs with medicines such as Etanercept projected to reach $5 billion in global sales by 2010.3 Native biologies is a class of drugs that are also referred to as secreted proteins or replacement therapies and it represents the earliest category of biotherapeutic drugs approved by the US FDA. It includes erythropoeitins, interferons, insulins, clotting factors, growth hormones and interleukins. By market size, this class of Biopharmaceutical products dominates the global portfolio of protein-based drugs. Erythropoeitins alone drew close to $10 billion in worldwide sales in 2004 and the number is still growing.4

♦Davinder S. Gill—Wyeth Research, Biological Technologies, 87 CambridgeRark Drive,

Cambridge, Massachusetts 02140, USA. Email: [email protected]

Pharmaceutical Biotechnology, edited by Carlos A. Guzmin and Giora Z. Feuerstein. ©2009 Landes Bioscience and Springer Science+Business Media.

Protein Drug Discovery Screening

Discovery of Biopharmaceuticals, as with any modern medicine, begins with the identification of a molecular "target", one that is closely associated with disease. The target can be either a missing or an overproduced or a malfunctioning protein or its derivative. A therapeutic hypothesis is formulated around the function of the target protein. Using recombinant DNA techniques, a series of candidate therapeutic proteins are then generated to test the hypothesis. The strategy to screen candidates depends on the type of protein being discovered. For fusion proteins and native biologies, screens tend to be relatively small with evaluation of 10-20 different proteins for lead identification. These candidates generally include variants with small (1-2 amino acid) modifications. In the case of receptor fusion proteins changes are made either in the receptor ectodomain or in the hinge region and in the more rare case in Fc portion of the protein. In the case of native biologies, the variants can be single or double amino acid changes introduce to improve expression or stability either as it relates to product formulation or to pharmacokinetics. In some cases, variants can also include engineering of the carbohydrate composition of the protein for improved pharmacokinetics.5 The small panel of proteins is then evaluated in a number of in vitro assays to select the most suitable lead protein.

Screens for monoclonal antibodies can be much larger. While monoclonal antibodies have been around now for more than three decades, their development as therapeutic agents is relatively recent however. Screening for monoclonal antibodies has traditionally relied on the hybridoma process developed by Kohler and Milstein in 1975. The process involves somatically fusing rodent B-cells derived from spleen or lymph node with mouse myeloma cells such as Sp2/0. Successfully fused cells can be screened for antibodies that selectively bind target antigen. Typically, primary screens are based on ELISA or a FACS assay. Increasingly however, the trend has been to carry out functional assays upfront where possible. As an example, a reporter gene assay can be used as a primary screen to identify hybridomas producing ligand-neutralizing antibodies or in the rare case a receptor-activating antibody. However since any initial screening is done using pools of typically 500 hybridoma cells per well, cloning by limited dilution must be carried out to isolate single hybridomas from the pool. This is typically done over one or two rounds of subcloning before the monoclonality of the hybridoma is established. Further, not all hybridomas grow at the same rate. This requires constant monitoring of the pool to ensure all possible hits from the fusion screen are isolated and rescued.

Litde has changed in the hybridoma technique since the days of Kohler and Milstein. While being overall robust, the hybridoma process is not very efficient. Typically only 1-2% of B-cells in the spleen or lymph node of an immunized rodent are actually antigen specific.4 The somatic cell fusion technique used to produce hybridomas is blind to the specificity of the B-cell. Thus, a number of irrelevant B-cells fuse along with relevant clones. Moreover, cell fusions conducted using the classic PEG reagent are inefficient with less than 1 in 20,000 B-cells resulting in a viable hybridoma.7 Techniques such as electric field-induced hybridization techniques have been reported to improve fusion efficiency. Regardless, hybridomas can sometimes be unstable resulting in a loss of antibody producing genes over time.

To overcome these limitations new techniques such as the Selected Lymphocyte Antibody Method (SLAM) have been developed.8 The aim of these approaches is to bypass the somatic cell fusion by direcdy sampling B-cells and to rescue antibody-encoding genes from a cells of interest. The SLAM approach consists of two steps. In the first step, biotinylated target antigen is cova-lently coupled with streptavidin coated sheep red blood cells. Next antibody producing B-cells are mixed with antigen coated sheep blood cells followed by addition of anti-IgG rabbit anti-serum and guinea pig serum as source of complement. Formation of plaques is then visualized under a microscope. In the second step, a desired B-cell, identified from a large pool of lymphocytes by its ability to form hemolytic plaque, is isolated using a micropipette. Antibody producing genes from the single B-cell are then isolated using RT-PCR followed by cloning of the VH and VL

domains. The advantage of the SLAM approach is that it directly samples antibody producing B-cells bypassing the hybridoma fusion step. This allows screening of a much higher number of B-cells (500,000) versus typically 2000-4000 wells in a hybridoma fusion. Rare B-cell clones can therefore be more efficiently isolated from a vast pool of irrelevant clones.

An approach different from the screening of hybridomas is based a completely new technology called phage display that does not require immunization of animals or screening of antibody producing cells. This approach is based on first direcdy isolating antibody encoding genes from B-cells and then cloning them as repertoires of individual antibodies in the form of binding fragments.9 These can be either single-domain antibodies, single-chain Fv of Fab fragments. These repertoires can then be displayed on hosts such as filamentous bacteriophage. The advantage of this system is that it not only obviates the need to carry out lengthy immunizations but it also removes any in vivo biases around immunodominant epitopes on target antigens. Further, since bacteriophage can be cultured more easily than antibody producing mammalian cells, the system lends itself to automation that can be integrated into high-throughput screening methodologies.

The recombinant antibody approach begins by mass cloning of antibody genes derived either from an immunized source such as a rodent or even a human exposed to pathogens or infectious agents. With sequencing of the human genome now complete, it is relatively straightforward to design sets of degenerate primers to amplify by PCR genes encoding antibody variable domains. Once the individual gene fragments are cloned the actual format by which they are assembled can vary. The most widely used format is that of the single-chain Fv wherein the heavy and the light chain variable regions of the antibody are linked through a 15-20 amino acid flexible linker. However, several labs have also reported the successful use of the Fab format. The scFv is a less stable molecule and one that is more prone to aggregation than the Fab.10 However, its single gene construct means it is easier to manipulate. Also, scFv s in general express better than Fabs.11 Smaller fragments such as single-domain antibodies have also been successfully cloned as recombinant repertoires however bigger fragments particularly those that contain the Fc portion have proved to be challenging.

Cloned ensembles of antibody genes can then be propagated in bacteria or in yeast.12 By creating a genetic fusion between the antibody binding domain (scFv or Fab) and host surface protein (e.g., phage pill or yeast Aga2) a link between genotype and phenotype is created. This allows mass screening of antibody repertoire through affinity driven selections. Only those phage oryeast that display binders get selected while the rest get washed away. Selected phage or yeast are then rescued and amplified for successive rounds of selection and amplification. At the end of 3-4 such cycles, phage or yeast are then sampled individually for binding activity. Those phage oryeast that display using binding properties are rescued and antibody genes contained within them isolated for further manipulations.

One of the key advantages of such in vitro screening methods is that the antibody selectivity can be driven based on the desired outcome. For example, if antibodies against a particular epitope or domain of the antigen are desired then counter-selections against undesired regions can be carried out. Or if antibodies with slow off-rates are to be selected then phage or yeast can be incubated in buffer at incubation times before elution. In cancer applications, antibodies that bind surface antigens and get internalized may be of value. In vitro approaches allow direct selection of these types of antibodies.13 Perhaps the ultimate screening approach is one where antibody repertoires are direcdy selected in vivo. This has been successfully demonstrated.14 However, its utility to the drug discovery process remains to be proven.

Optimization

Proteins, like small molecule drugs, frequently require optimization before lead candidates can be identified for further development. Often, the optimization is focused towards improvement of binding affinity or selectivity. But it can also include reduction of potential for immunogenicity or improvement of solubility of stability characteristics. For fusion proteins or native biologies since the initial screens are small, sometimes optimization in included upfront as in the case of darbopoetin alfa.5

Over the past decade several different optimization technologies have been developed. It is not possible to review them exhaustively here. However, it is important to highlight a few key technologies that have had the most impact in the protein optimization process. Most protein optimization technologies have essentially relied on two key steps. The first step has been to generate chemical diversity within the protein through a variety of mutagenesis techniques. Depending on the size of the diversity created, the second step is to use any one of a number of "Display" technologies to sort through the libraries and isolate important variants. There are also semi-rational approaches based on protein structural modeling that generate small and focused chemical diversity that can be screened using standard methodologies.

One of the early approaches used to generate diversity was the use of error-prone PCR.15 The approach relied on the low fidelity of the polymerases used in the PCR process to randomly create mutations across a stretch of DNA. When the low fidelity of an enzyme like Taq polymerase is combined with certain PCR conditions such as high Mg2+ concentrations, the error rate can be as high as 0.01 mutation/bp/PCR cycle. The advantage of the error-prone PCR process is that it is simple, it introduces mutations across the gene of interest and can therefore be very usefid for identify structure/function "hotspots". The disadvantage is that biases with certain polymerases have been reported and newer polymerases have been engineered to overcome the problems but still the overall mutational rate achieved is low. A different approach, targeted mutagenesis, works better when there is knowledge of where mutations need to be introduced as in the case antibody CDR regions.16 This approach generates far more chemical diversity than error-prone PCR but in a more restricted portion of the sequence. Targeted mutagenesis can involve any number of oligonucleotide-mediated strategies. This can include spiking wild type nucleotides or codons with mutants or replacing stretches of sequences with nucleotides or codons that are completely randomized such that each nucleotide is substituted by every other possibility. Thus if three nucleotides are substituted with four possibilities at each position (G, A, T or C) then this generates 3+ or 81 possible combinations. For larger sequence stretches the permutations are even larger.

Yet another approach that combines the breadth of error-prone PCR and the depth of targeted mutagenesis is Directed Evolution.17 Here, codons are designed to introduce all possible mutations at a given position in linear sequence. However, rather than carry this through in a combinatorial fashion all along the sequence, parallel synthesis is conducted to construct small pools of mutagenic sequence. These pools are then assayed using traditional screens to isolate functionally important variants.

Two more mutagenesis approaches are worth mentioning. One involves randomly shuffling DNA to create chemical diversity.18 The process begins by first digesting DNA into small fragments using a restriction endonuclease such as DNase I. Digested fragments are then randomly combined and amplified using PCR to recreate the full length gene. The diversity of sequences created by this process is sorted for functionally important clones. The process is then repeated until the protein is optimized, for example, until a certain improvement in binding affinity is achieved.

For some of the mutagenesis approaches discussed above there is the limitation that the chemical diversity generated can only be partially sampled. That is because the number ofpossibie permutations rise exponentially with linear sequence sampled whereas the maximum number of variants that can tested is around 10'° due to limitations in bacterial transformations. Some of these limitations are alleviated by the use of cell-free systems as described below. But another approach has been to use semi-rational or computational methods to sample protein sequence in silico before any experimental works begins.19 The process begins with construction of a protein structural model followed by analysis of all possible sequences that permit the fold predicted by the model. Using computational resources a large number of possible sequences are screened and those predicted to retain the protein functional fold given the highest score. Only the most optimal sequences predicted by the algorithm are synthesized and experimentally tested. Thus, sequence space is reduced to fewer functionally relevant sequences.

As discussed above in those cases where very large chemical diversity is generated one needs a selection tool to sort through the enormous collection of variants. Three types of "Display" technologies have been developed and used successfully in the drug discovery process. All three technologies are based on a simple but a powerftd principle: a robust link between sequence and function. In the case of phage display, fusing genes of interest with anyone of the genes encoding a phage coat protein creates this link. Several proteins have been used successfully including pill, pVIII, pVI and pIX although pill remains the most widely used.20 The number of different classes of proteins displayed on phage is probably the widest. This includes peptides, single-domain antibodies, scFy's, Fabs, growth factors, cytokines and receptor ectodomains. Thus, as a library selection tool, phage display is probably the most versatile.

Ribosome display is a cell free system in which a gene ofinterest is transcribed and translated in vitro in a cell free environment using£. colt, rabbit reticulocyte or wheat germ extracts.21 Translated polypeptides are folded in vitro and them made to remain anchored to the ribosome using long polypeptide tethers that are part of the translated sequence. By doing so, the folded protein is linked to the mRNA sequence encoding it. These protein-ribosome-mRNA complexes are then sorted through steps of affinity selection followed by rescue of functionally important complexes. Rescued mRNA is then reverse-transcribed to produce cDNA which can then be amplified by PCR followed by iterative rounds of in vitro translation and selection. The main advantage of ribosome display is that it is cell-free which means sequence diversity larger than phage libraries can be sampled (~1014). Further, steps to carry out bacterial transformations and preparation of viral stocks are obviated. Therefore, multiple parallel proteins can be optimized The disadvantage is that the system is not as robust due to presence of ribosomes and mRNA. Therefore, selection conditions have to be extensively optimized.

A relatively new entrant into the display technology field is yeast display.22 In this approach a the genetic linkage between structure and function is created by fusing a gene of interest to the Aga2p gene which encodes the adhesion subunit of the yeast agglutinin protein. Unlike phage or ribosomes, yeast is a eukaryote which means that folding and posttranslational modification is different and closer to that in mammalian cells. Further, through the fusion process described above 10,000-100,000 copies of the protein ofinterest are displayed per cell. This means quantitative screening of variants can be carried out by fluorescence activated cell sorting permitting both equilibrium and kinetic selections. A limitation of yeast display is smaller levels of transformations than permissible in phage or ribosome display thereby limiting the amount of sequence space that can be sampled.

Novel Agents

As mentioned in the introduction, most biopharmaceutical products can be classified as monoclonal antibodies, receptor fusion proteins or native biologies. However, there are some products that do not fall in any of these categories. Two of them are worth mentioning under the category of novel agents.

The first of this novel set of proteins are antibody drug conjugates. These novel agents are part protein and part small molecule drugs. The novelty comes from the role played by each component in producing a safe, pharmacological effect. It is well recognized in fields such as Oncology that if chemotherapeutic drugs could be made safer such that they more selectively kill cancer cells than normal cells, it would gready improve their safety profile. Antibody drug conjugates do just that. An excellent example of this mechanism is Gemtuzumab Ozogamicin which is approved as monotherapy for treatment of relapsed acute myeloid leukemia.23 This drug consists of three components: a humanized monoclonal antibody directed against the CD33 antigen expressed on myeloid cells, a hydrolysable bifunctional linker and calicheamicin, a potent cytotoxic drug. By itself, calicheamicin is too toxic and cannot be used without serious side effect. Without the drug the antibody by itself has no effect. However, coupling the two allows the drug to be selectively delivered to target positive cells minimizing nonhematologic toxicity. Upon binding to CD33, the conjugate gets internalized into endosomes whereupon the linker holding the calicheamicin gets hydrolysed. Calichaemicin gets released resulting in cell death. Since the approval of Gemtuzumab, this type of approach has been used for the development of several conjugates for both hematologic malignancies as well as for solid tumors. In such cases a variety of cell surface antigens such as ErbB2 or CD30 have been targeted using chimeric or humanized monoclonal antibodies conjugated to other cytotoxic drugs such as maytansinoid or aurastatin.24"25

Another class of biopharmaceutical drugs that are novel agents is one composed of proteins that exert their pharmacologic effect through the implantation of a device. An example ofthis class of products is recombinant human bone morphogenetic protein -2 (rhBMP2).2<; rhBMP2 is a member of the TGF{3 superfamily that has strong osteogenic properties. This protein is too potent to be delivered systemically. But when administered along with absorbable collagen sponge (ACS) as amatrix, rhBMP2/ACS is effective at inducing de novo bone formation. The drug has been approved by the FDA for three distinct indications in the orthopedic area—interbody spinal fusions, open tibial fractures and for autogenous bone grafts. The choice of the delivery device for rhBMP2 has a strong effect on its clinical activity. It was the development of a suitable carrier that took up a significant time in the clinical development of rhBMP2.26 Given that several other members of the TGFp superfamily are candidates as drugs in the muscle, tendon and bone repair areas, lessons learnt from the development of rhBMP2 would help these programs move forward.

Challenges and Opportunities

Perhaps, where protein drug discovery most differs from discovery of conventional drugs is the attention that is paid to maintenance of human composition of the protein drug, to efficient protein synthesis in cultured cells and to the preparation of the drug for delivery by injection. Because conventional drugs are synthetic organic molecules, these considerations normally do not apply. Another difference is that certain tests carried out to verify potential side effects of conventional medicines are not necessary for Biopharmaceuticals. This is because Biopharmaceuticals do not inhibit the function of normal human proteins found in the liver (e.g., cytochyeme p450) or in other critical organs such as the heart (e.g., hERG). Outside of this, criteria to establish safety and efficacy are quite similar between protein and conventional drugs.

The early monoclonals made using the hybridoma process were murine in origin. When tested in clinical trials these proteins were quickly recognized as 'foreign" by the human immune system and eliminated. There was thus the need for approaches to reduce the potential immunogenic-ity of proteins by reducing their nonhuman content. Early attempts in this regard were oriented towards development of chimeric antibodies. These molecules possessed the minimal binding domains derived from mouse origin with the remainder of the constant domain derived from human IgG. Chimerization reduced the risk associated with immunogenicity and successful product launches were made possible—Rituximab for treatment of non-Hodgkins lymphoma and Cetuximab for head and neck cancer. However, chimeric antibodies did not completely rule out the risk for immunogenicity as there still remained significant amounts of mouse protein in the molecule. Three new technologies arose to fulfill this need and all three have now resulted in successful products.

In a series of advances, the murine content of antibodies was reduced using protein engineering-techniques. The first development was to "humanize" mouse antibodies. In an approach pioneered by Winter and colleagues complementarity-determining regions (CDRs) of murine antibodies were grafted on human framework regions such that the CDRs were the only mouse protein in the antibody.27 However, a problem with this method was the observation that simply grafting CDRs of mouse antibodies on human antibody framework regions resulted in a loss of binding affinity. This is bccause the CDRs fold in the form of loops which must be correcdy positioned by the frameworks for optimal binding. In a second approach pioneered by Queen, problems associated with CDR grafting were successfully solved.28 The Queen approach required changes at key framework positions using sequence alignment between the mouse donor sequence and the human acceptor sequence and also by the use of computer models. This approach was vasdy successful in overcoming the problems associated with antibody humanization. Indeed, the first generation of antibodies approved by the FDA was humanized antibodies such as Trastuzumab, Omaluzumab and Natalizumab.

However, the early nineties saw the emergence of two novel technologies that shaped the discovery of the current generation of monoclonal antibodies. The first was the development ofphage display technology discussed above. While the early proof-of-concept experiments conducted to validate this technology were based on construction of peptide libraries of single-domain antibodies, the field rapidly expanded to included phage display libraries of scFv and Fab fragments. Construction of large diverse phage repertoires allowed scientists to bypass the hybridoma process and obtain antibodies through completely in vitro approaches. However, antibodies isolated using this approach were not always of the highest affinity as they had not gone through the somatic hypermutation process that antibodies from immunized sources had. Larger repertoires or those derived from autoimmune sources were made to mitigate the problem but the affinity problem was not completely solved.

A second technology that arose was the development of transgenic mouse strains harboring loci ofhuman Ig genes. In this technology, genes encoding endogenous mouse Igwere first inactivated using gene-targeting techniques. This was followed by the systematic introduction of large chunks of the human Ig loci (H, k and X) using yeast artificial chromosomes into the mouse germline.29 Reconstituted mice stably expressed human immunoglobulins at normal levels and had normal B-cell development. Most importandy, when challenged with a human protein as an immunogen, these mice mounted a strong immune response of fully human antibodies which included class switch recombination as well as somatic hypermutation. Thus, creation of such transgenic animals obviated the need to first create a mouse monoclonal antibody and then to carry through human-ization. The fact that transgenic mice mounted both a primary and a secondary response meant that high affinity, fully human antibodies of the IgG class could be readily obtained. However, since the process of making such antibodies relied on immunization and screening of hybridomas limitations to the traditional hybridoma approach still applied.

Next Generation Proteins

The nature of biopharmaceutical drug discovery has undergone a fundamental change in the past 5-7 years. Drawing upon the success of first and second-generation protein pharmaceuticals, namely secreted factors such as erythropoetin and receptor fragments such as etarnecept, the industry has increasingly shifted towards the development of monoclonal antibodies as therapeutic agents. Recent product launches by major biotech companies as well as late stage biopharmaceutical pipeline candidates are now mosdy monoclonal antibodies. Genentechs Bevacizumab and Wyeths Bapineuzumab are excellent examples of this trend. However, the industry has also faced several challenges.

While hugely successful, monoclonal antibodies are still very large, complex molecules that require significant engineering at the molecular level to be effective. Enabling technologies to carry out this type of engineering are often held by small biotech companies where access can be restructured. More importandy however, in the recent months, Big Pharma has serially acquired biotech companies that pioneered the development of these technologies.

Another trend that has affected the biopharmaceutical industry has been the rapid advances in protein sequence, structure and function coupled with the commercial need for newer, cost-effective protein therapies. This trend has led to the development of technologies to exploit novel protein scaffolds as therapeutic precursors. Strategies include engineered fragments of monoclonal antibodies as well as nonantibody scaffolds. These strategies have recendy been reviewed elsewhere and are referenced here.

Conclusion

The US Food and Drug Administration approved the first protein drug developed using recombinant DNA technology {human insulin) in 1982. Thus, the protein drug industry is relatively young when compared with the traditional drug industry. As discussed in this chapter, significant new technologies developed in the recent years promise to allow rapid advancement of protien drugs. Biopharmaceuticals have enjoyed tremendous success in the recent years judged by the surge of approvals by the FDA. This success draws upon technological advances made in the field of protein therapeutics but also on a greater realization that there are significant opportunities within pharmaceutical drug development to exploit the power of biopharmaceuticals. Indeed, we are now witnessing the emergence of protein drug development opportunities in areas such as metabolic disorders, Alzheimer's disease and osteoporosis. Traditionally, these areas were reserved for conventional drug discovery. As we move forward into the new millennium, it is hoped that the synergy between biopharmaceuticals and conventional medicines can be further leveraged for safer and more cost-effective treatments. If successful, these therapies will address significant unmet medical needs whose aggregate cost to the healthcare systems worldwide runs into tens of billions of dollars each year.

References

1. Das RC, Morrow JK, Antibody Technologies Rise to New Challenges. Am Biotech Lab 2007; 25(8).

2. Maggon K. Monoclonal antibody "gold rush". Curr Med Chem 2007; 14(18); 1978-87.

3. Lawrence S. Billion dollar babies—biotech drugs as blockbusters. Nat Biotechnol 2007; 25(4):380-2.

4. Pavlou AK, Reichert JM. Recombinant protein therapeutics-success rates, market trends and values to 2010. Nat Biotechnol 2004; 22(12):1513-9.

5. Elliott S, Lorenzini T, Asher S et al. Enhancement of therapeutic protein in vivo activities through glycoengineering. Nat Biotechnol 2003; 21(4):414-21.

6. Kodituwakku AP, Jess a p C, Zola H et al. Isolation of antigen-specific B-cells. Immunol Cell Biol 2003; 81(3):163-70.

7. Pirofski L, Casadevall A, Rodriguez L et al. Current state of the hybridoma technology. J Clin Immunol 1990; 10(6 Suppl):5S-12S.

8. Babcook JS, Leslie KB, Olsen OA et al. A novel strategy for generating monoclonal antibodies from single, isolated lymphocytes producing antibodies of defined specificities. Proc Natl Acad Sci USA 1996; 93(15):7843-8.

9. Marks JD, Hoogenboom HR, Bonnert TP ct al. By-passing immunization. Human antibodies from V-gene libraries displayed on phage. J Mol Biol 1991; 222(3):581-97.

10. Rothlisberger D, Honegger A, Pluckthun A. Domain interactions in the Fab fragment: a comparative evaluation of the single-chain Fv and Fab format engineered with variable domains of different stability. J Mol Biol 2005; 347(4):773-89.

11. Hust M, Jostock T, Menzel C et al. Single chain Fab (scFab) fragment.BMC Biotechnol 2007; 7:14.

12. Feldhaus MJ, Siegel RW, Opresko LK et al. Flow-cytometric isolation of human antibodies from a nonimmune Saccharomyccs ccrevisiae surface display library. Nat Biotechnol 2003; 21(2): 163-70.

13. Poul MA, Becerril B, Nielsen UB et al. Selection of tumor-specific internalizing human antibodies from phage libraries. J Mol Biol 2000; 301(5):1149-61.

14. Trepel M, Arap W, Pasqualini R. In vivo phage display and vascular heterogeneity: implications for targeted medicine. Curr Opin Chem Biol 2002; 6(3):399-404.

15. Hawkins RE, Russell SJ, Winter G. Selection of phage antibodies by binding affinity. Mimicking affinity maturation. J Mol Biol 1992; 226(3) :889-96.

16. Rajpal A, Beyaz N, Haber L et al. A general method for greatly improving the affinity of antibodies by using combinatorial libraries. Proc Natl Acad Sci USA 2005; 102(24):8466-71.

17. Vasserot AP, Dickinson CD, Tang Y et al. Optimization of protein therapeutics by directed evolution. Drug Discov Today 2003; 8(3): 118-26.

18. Stemmer WP. Rapid evolution of a protein in vitro by DNA shuffling. Nature 1994; 370(6488):389-91.

19. Hayes RJ, Bentzien J, Ary ML et al. Combining computational and experimental screening for rapid optimization of protein properties. Proc Natl Acad Sci USA 2002; 99(25):15926-31.

20. Sidhu SS, Feld BK, Weiss GA. M13 bacteriophage coat proteins engineered for improved phage display. Methods Mol Biol 2007; 352:205-19.

21. Lipovsek D, Pluckthun A. In vitro protein evolution by ribosome display and mRNA display. J Immunol Methods 2004; 290(l-2):51-67-

22. Boder ET, Wittrup KD. Yeast surface display for screening combinatorial polypeptide libraries. Nat Biotechnol 1997; 15(6):553-7.

23. Hamann PR, Hinman LM, Hollander I et al. Gemtuzumab ozogamicin, a potent and selective anti-CD33 antibody-calicheamicin conjugate for treatment of acute myeloid leukemia. Bioconjug Chem 2002; 13(l):47-58.

24. Liu C, Tadayoni BM, Bourret LA et al. Eradication of large colon tumor xenografts by targeted delivery of maytansinoids. Proc Nad Acad Sci USA 1996; 93(16):8618-23.

25. Doronina SO, Toki BE, Torgov MY et al. Development of potent monoclonal antibody auristatin conjugates for cancer therapy. Nat Biotechnol 2003; 21(7):778-84.

26. Seeherman H, Wozney JM. Delivery of bone morphogenetic proteins for orthopedic tissue regeneration. Cytokine Growth Factor Rev 2005; 16(3):329-45.

27. Jones PT, Dear PH, Foote J et ai. Replacing the complementarity-determining regions in a human antibody with those from a mouse. Nature 1986; 321(6069):522-5.

28. Queen C, Schneider WP, Selick HE et al. A humanized antibody that binds to the interleukin 2 receptor. Proc Natl Acad Sci USA 1989; 86(24): 10029-33.

29. Mendez MJ, Green LL, Corvalan JR et al. Functional transplant of megabase human immunoglobulin loci recapitulates human antibody response in micc. Nat Genet 1997; 15(2): 146-56.

30. Gill DS and Damlc NK. Biopharmaccutical drug discovery using novel protein scaffolds. Curr Opin Biotechnol. 2006 17(6):653-8.

Was this article helpful?

0 0
Natural Arthritis Relief

Natural Arthritis Relief

Natural Arthritis Relief details a unique method of reversing Rheumatoid Arthritis Symptoms by removing numerous arthritis triggers as well as toxins using a simple 5 step natural process.

Get My Free Ebook


Post a comment