Optimization of Biopharmaceuticals

Pharmacists should be knowledgeable about the different approaches used to modify biopharmaceuticals and how they affect stability, dosing, safety, and efficacy.

Biopharmaceuticals are medications made using living systems and biological processes or biotechnology. One in 4 approved drugs is a biopharmaceutical,¹ and they are increasingly prescribed for patients. Biopharmaceutical drug sales had revenues of approximately $46.5 billion in 2009, a 4.5% increase from 2007.² In 2010, 7 of the top 20 drugs sold were biopharmaceuticals, totaling approximately $21.7 billion in total sales.³

Thanks to technological advances, biopharmaceuticals are able to be modified (Figure 1) with the intent of improving stability, dosage regimens, and safety and efficacy compared with their original forms. Many of these medications are complex molecules that have special storage and use requirements in order to maintain their stability. Pharmacists who are knowledgeable about the different approaches used to modify these medications and how those modifications affect stability, dosing, safety, and efficacy are better able to inform patients regarding use, side effects, and storage of these medications.

OPTIMIZED INSULIN PRODUCTS

Insulin is one example of a biologically based medication manipulated using biotechnology. Since its original approval, insulin has been modified in a variety of ways, leading to the availability of numerous optimized insulin products. Table 1 lists several of the insulin products available on the US market. This listing includes the organism the insulin was produced in, how it was modified, and pharmacokinetic parameters of the products.

Prior to the approval of recombinant insulin, patients were treated with bovine- and pork-derived insulin. The initial optimizations to the animal-derived insulins were made to the final formulation using zinc and protamine to decrease insulin solubility, resulting in insulins with longer durations of action.^4-7 More recent optimizations to the biotechnology-derived insulins, in addition to formulation, are modifications to the protein primary structure. Protein modifications can generally be grouped into 4 categories: changes in sequence to a small number of the amino acids making up the protein, deletions of a segment of the amino acids of the protein, synthetic addition of specific chemical moieties such as polyethylene glycol (pegylation) or fatty acids (acylation) to the protein, and modifications by the addition of structures such as glycans to the protein by the host cell. These optimizations improve protein stability, immunogenicity, pharmacokinetics, and ease of use.

Of the available insulin products, there are several that are examples of products with modified amino acid sequences, and some that have also been modified by the addition of chemical structures to the protein. Lispro (Humalog), aspart (Novolog), glulisine (Apidra), glargine (Lantus), and detemir (Levemir) all have 1 to 3 amino acid modifications of the recombinant human insulin primary sequence. Endogenous human insulin consists of 2 chains, with chain A containing 21 amino acids and chain B containing 30. In insulin lispro, the proline at amino acid position 28 and the lysine at position 29 in the B chain have been exchanged—hence the name insulin lispro. Insulin aspart is similarly named because the proline in position B28 has been replaced with aspartic acid. An asparagine in position B3 and a lysine in position B29 of human insulin have been replaced with lysine and glutamic acid, respectively, in insulin glulisine. These modifications reduce the ability of the protein to form dimers and hexamers, leaving the active monomeric form free in the blood.^8-12

In insulin glargine (Lantus), the asparagine in position A21 has been replaced with glycine and 2 additional arginines have been added to the end of the B chain. The product is formulated in acidic pH where it has a net positive charge and is solubilized. Because of the amino acid changes, the solubility of the drug is significantly reduced at physiological pH and upon injection, the protein precipitates. The precipitant slowly dissolves, resulting in near-constant release over 24 hours.⁸ Insulin detemir (Levemir) is an example of an acylated biopharmaceutical. The amino acid at position B30 has been removed and a fatty acid chain chemically attached to B29. This manipulation increases the ability of the insulin to bind to itself as well as to circulating albumin, forming a complex that slowly dissociates over time, resulting in prolonged half-life.¹³

The modifications in lispro, aspart, and glulisine reduce the onset and duration of insulin postinjection, while the modifications in insulin glargine and insulin detemir increase the onset and duration of these insulin products. The fast-acting insulins (lispro, aspart, and glulisine) are used to provide patients with preprandial insulin, while the slow onset insulins (glargine and detemir) are used to provide basal insulin levels.

BIOPHARMACEUTICAL MEDICATIONS

Additional biopharmaceuticals have been modified in ways similar to insulin. The selected examples (Table 2) are biopharmaceutical medications listed in the top 200 drugs by sales and are used to treat diseases characterized by an inability of the patient to produce a vital endogenous protein. This article therefore does not include monoclonal antibodies, vaccines, and diagnostic products. However, 2 life-saving therapies not included in the top 200 drugs are included as examples. Several medications that have been optimized using small amino acid modifications, deletions, chemical modifications, and posttranslational modifications are discussed below.

SMALL AMINO ACID MODIFICATIONS

Amino acid modifications of protein drugs can produce safer, more effective, and stable medications. These modifications may be 1 or 2 amino acid changes that either modify the physicochemical properties of the final product or are used as precursors onto which chemical moieties are added. Examples of simple amino acid modifications include filgrastim (Neupogen) and interferon beta-1b (Betaseron).

Filgrastim was modified compared with the endogenous human granulocyte colony-stimulating factor (hG-CSF) by adding a methionine at the beginning of the amino acid sequence. The methionine is required for the expression of the hG-CSF gene in Escherichia coli and has no effect on the biological activity of the protein.¹⁴ Filgrastim is nonglycosylated because it is produced in E coli, in contrast to the naturally glycosylated human protein. Filgrastim was further modified by chemical addition of polyethylene glycol (PEG) to make pegfilgrastim, resulting in increased half-life. In interferon beta-1b, the amino acid sequence has been altered from that of human interferon beta by substituting serine for cysteine in position 17. This substitution prevents disulfide linkages at position 17, decreasing aggregation of the protein and enhancing the stability and effectiveness of the product.¹⁵ Betaseron also differs from interferon beta 1a (Avonex and Rebif) produced in mammalian cells in that it is non-glycosylated, as it is produced in E coli. The biologic and antitumor activities of natural interferon beta (interferon beta-1a) and interferon beta-1b are similar.^16-18

OPTIMIZATION THROUGH DELETION

Deletions of nonessential domains of a protein not required for biological activity may improve the ease of manufacturing, stability, and pharmacokinetic properties of the product. Recombinant human factor VIII is a protein involved in blood clotting used to treat hemophilia. It has several amino acid domains, one of which is not required for its procoagulant activity.¹⁹ Removal of this domain results in increased expression of the protein, allowing for higher yield in manufacturing.^19,20

Another example of optimization through deletion is tissue plasminogen activator (tPA), which breaks down blood clots by converting plasminogen to plasmin, which in turn converts insoluble fibrin in blood clots into soluble products. Reteplase (Retavase) is an optimized version of tPA lacking several domains. The deleted domains reduce the affinity of the modified protein for receptors involved in its clearance, increasing the drug’s half-life. One of the deleted domains also binds to fibrin, affecting its potency. Although reteplase has lower in vitro potency compared with alteplase (unmodified tPA) in dissolving blood clots, its in vivo potency is higher due to its increased half-life._^21,22 The increased half-life of reteplase also means that the product can be administered using bolus injections, in contrast to alteplase (Activase) which has a dosing regimen of an initial bolus injection followed by 2 IV infusions over 30 and 90 minutes.

CHEMICAL ADDITION

Drug delivery design has always been an important part of the development process for biopharmaceuticals. Once administered, therapeutic proteins are susceptible to enzymatic degradation, reducing their bioavailability and halflife as well as having rapid renal clearance due to their size.²³ Therapeutic proteins can also be antigenic. Furthermore, they have limited solubility with the potential for precipitation in solution, limiting formulation options and routes of administration.²⁴

Pegylation, the covalent attachment of polyethylene glycol (PEG) to macromolecules, was developed in the 1970s based on early efforts of Abuchowski and Davis.²⁵ Schering-Plough was the first pharmaceutical company to utilize this process, with the production of peginterferon alfa-2b (PegIntron) for the treatment of chronic hepatitis C in 2001.²⁴ Today, pegylation is a standard drug delivery technology. PEG is a common ingredient in familiar products such as cosmetics, ointments, creams, suppositories, and even some food products.²⁶ Data available due to its widespread use indicate PEG is safe due to the absence of immunogenicity and antigenicity²³ and is relatively non-toxic, biocompatible, and highly soluble in water and other organic solvents.²⁵ These characteristics make PEG valuable in the formulation of PEG-protein conjugates.

A functional group is usually added to PEG before it is conjugated to a protein. The addition of the functional group allows PEG to be attached to the protein through a reactive amino acid such as N-terminal amino groups or C-terminal carboxylic acid. Most often PEG is conjugated with lysines and N-terminal amino acid groups. Lysine is favored due to its abundance in the amino acid sequence of protein molecules. Multiple PEG molecules can be conjugated to a single protein to take greater advantage of the benefits of this conjugation.²⁷

Data indicate pegylated protein biopharmaceuticals have improved patient safety profiles, as they have low toxicity, immunogenicity, and antigenicity.²⁸ PEG conjugates mask immunogenic structures of proteins and decrease proteolytic attack by enzymes.²⁴ Often the half-lives of pegylated proteins are increased compared with the proteins from which they are derived due to their increased size, reducing elimination by glomerular filtration.²⁶ PEG-protein conjugations have greater stability and solubility due to their physicochemical properties compared with native proteins. ²⁴ Overall, pegylated protein drugs display better pharmacokinetic and pharmacodynamic profiles compared with nonpegylated proteins. Currently, pegylation is being used to modify peptides, oligonucleotides, antibody fragments, and other small organic molecules as well.²⁶ Efforts are being made to further understand the structure of pegylated proteins in order to develop second-generation pegylated biopharmaceuticals with increased bioavailability and efficacy.²³

POST-TRANSLATIONAL MODIFICATIONS

Glycosylation is the enzymatic addition of glycans to proteins by the cells in which the protein is expressed. The glycans can be mono-, oligo-, or polysaccharides that can vary greatly in size and are bound to the protein via N or O links. Glycoproteins are not encoded in the genes of an organism; rather, there are genes in organisms that code for enzymes and transporters that synthesize and build the glycans to be bound to proteins. Glycosylation is a form of protein posttranslational modification.²⁹

Various cellular systems are used to produce glycosylated proteins. A gene coding for the protein of interest is transfected into a cell, where it is expressed and then glycosylated by the cell’s glycosylation mechanisms. Mammalian cells, particularly Chinese hamster ovary (CHO) cells, have been used for the past 20 years to produce glycosylated proteins. Mammalian cells are preferred over nonmammalian cell systems because they glycosylate proteins much as they are in humans. Unfortunately, mammalian cell systems are not able to produce proteins in large amounts and are not stable expression systems.³⁰

Other cell systems are used to overcome these problems, including yeast and plant cells. Yeast cells produce large amounts of proteins but do not exactly replicate the glycosylation of mammalian cells.³⁰ Transgenic plants are able to produce large quantities of proteins at relatively low cost. This system is limited, however, as the plant proteins are difficult to isolate and the isolates may contain unwanted plant compounds.³⁰ As in yeast, plant cells are unable to exactly mimic mammalian cell glycosylation due to differences in the glycosylation mechanisms, limiting the use of these plant cell products. Also of concern when using transgenic plants for producing biopharmaceutical proteins is the potential for crosspollination of transgenic plants with native plants, causing loss of control of the production and potential exposure to the recombinant protein.^30,31

Erythropoietin is produced in the human kidney and stimulates erythrocyte production and maturation. It is naturally glycosylated in 3 locations along the amino acid sequence. Several recombinant human erythropoietin products are available. Epoetin alfa (Procrit and Epogen) is a recombinant protein produced in CHO cells with the identical amino acid sequence as endogenous human erythropoietin. Because it is expressed by CHO cells, it has slightly different glycans at the glycosylation sites compared with human erythropoietin. Darbepoetin (Aranesp) is a recombinant erythropoietin also expressed in CHO cells, but with an altered amino acid sequence compared with native human erythropoietin. The altered amino acid sequence results in an increase in the number of glycosylations from the native 3 to 5. This increase in the number of glycans decreases its receptor-mediated uptake and breakdown by target cells, effectively increasing its half-life compared with epoetin alfa.^32,33

CONCLUSIONS

Since the approval of the first recombinant protein in 1982, the number of available biopharmaceuticals has rapidly increased. The amino acid modifications, deletions, and chemical changes that can be made to recombinant proteins have improved recombinant proteins, giving them enhanced safety, efficacy, bioavailability, and stability compared with the unaltered protein products.

As the biotechnology industry progresses and produces even more unique biopharmaceutical products, it becomes increasingly important that health care providers keep abreast of these new products in order to provide the best therapeutic options to patients. Protein drugs are less stable compared with small molecule drugs, and as such, the labeling on many of these products includes warnings such as “do not shake,” “protect from light,” “do not freeze,” etc. The ability of pharmacists to integrate their understanding and knowledge about the therapeutic applications and the stability of these drugs is what will drive the health care team forward in helping to reach better therapeutic outcomes.

SPT

Table 1. Examples of Insulin Products

Category

Name

Brand

Onset (hr)*

Time to Peak (hr)*

Duration (hr)*

Frequency*

Optimization

Effect

Host cell

Short-Acting

Insulin Injection Regular

Humulin R

0.5

0.8-2

4-12 (U-100); Up to 24 (U-500)

30-60 min BAM

Zinc formulation

Increases insulin dimer/hexamer formation and duration. Stabilizes insulin from degradation

E coli

Novolin R

S cerevisiae

Intermediate-Acting

Insulin Isophane Suspension (NPH)

Humulin N

1-2

6-10

14-24

1-2 times daily**

Zinc and protamine formulation

Slower onset of action and longer duration of activity compared to regular insulin

E coli

Novolin N

S cerevisiae

Rapid-Acting

Insulin Lispro

Humalog

0.25-0.5

0.5-1.5

≤5

15 min BAM

Pro(B28)Lys; Lys(B29)Pro

More rapid onset and shorter duration of action than regular insulin

E coli

Insulin Glulisine

Apidra

0.2-0.5

0.6-2

3-4

15 min BAM or 20 min ASM

Asn(B3)Lys; Lys(B29)Glu

E coli

Insulin Aspart

NovoLog

0.2-0.3

0.7-0.8

3-5

5-10 min BAM

Pro(B28)Asp

S cerevisiae

Intermediate —Long-Acting

Insulin Detemir

Levemir

3-4

6-8

6-23

1-2 times daily

Thr(B30) deleted; C14 fatty acid addition at Lys(B29) (acylation)

Prolonged action

S cerevisiae

Long-Acting

Insulin Glargine

Lantus

3-4

none

24

Once daily

Asn(A21)Gly; Arg-Arg added to C-terminus of B chain.

E coli

*Pharmacokinetic information for subcutaneous administration

**Commercial combinations of solution and NPH formulations are also available to modulate onset, peak, and duration of action

References: Dailymed (dailymed.nlm.nih.gov) and Lexicomp (online.lexi.com)

Figure 1. Protein Optimization Approaches

References:

1. Redwan EM, Matar SM, El-Aziz G, Serour EA. Synthesis of the human insulin gene: protein expression, scaling up and bioactivity. Prep Biochem Biotechnol. 2008;38(1):24-39.

2. Aggarwal S. What’s fueling the biotech engine—2009-2010. Nat Biotechnol. 2010;28:1165-1171.

3. Bartholow M. Top 200 prescription drugs of 2010. Pharmacy Times. Published online May 16, 2011.

4. Hagedorn H, Jensen B, Krarup N, Wodstrup I. Protamine insulate. JAMA. 1936;106:177-180.

5. Hallas-Møller K, Petersen K, Schlichtkrull J. Crystalline and amorphous insulin-zinc compounds with prolonged action. Science. 1952;116(3015):394.

6. Owens DR. Insulin preparations with prolonged effect. Diabetes Technology & Therapeutics. 2011;13(S1):5-14.

7. Scott D, Fisher A. Studies on insulin with protamine. J Pharmacol Exp Ther. 1936;58(1):78.

8. Bolli GB, Owens DR. Insulin glargine. Lancet. 2000;356(9228):443.

9. Freeman JS. Insulin analog therapy: improving the match with physiologic insulin secretion. JAOA. 2009;109(1):26.

10. Holleman F, Hoekstra JB. Insulin lispro. N Engl J Med. 1997;337(3):176-183.

11. Apidra [package insert]. Bridgewater, NJ; sanofi-aventis US LLC; 2009.

12. Setter SM, Corbett CF, Campbell RK. Insulin aspart: a new rapid-acting insulin analog. Ann Pharmacother. 2000;34(12):1423.

13. Havelund S, Plum A, Ribel U, et al. The mechanism of protraction of insulin detemir, a long-acting, acylated analog of human insulin. Pharm Res. 2004;21(8):1498-1504.

14. Frampton JE, Lee CR, Faulds D. Filgrastim: a review of its pharmacological properties and therapeutic efficacy in neutropenia. Drugs. 1994;48(5):731-760.

15. Hawkins M, Horning S, Konrad M, et al. Phase I evaluation of a synthetic mutant of β-interferon. Cancer Res. 1985;45(11, part 2):5914.

16. Von Hoff DD, Huong AM. Effect of recombinant interferon-beta ser on primary human tumor colony-forming units. 1988;8:813-820.

17. Mahaley MS Jr, Dropcho EJ, Bertsch L, Tirey T, Gillespie GY. Systemic beta-interferon therapy for recurrent gliomas: a brief report. 1989;71:639-641.

18. Miles SA, Wang HJ, Cortes E, Carden J, Marcus S, Mitsuyasu RT. Beta-interferon therapy in patients with poor-prognosis kaposi sarcoma related to the acquired immunodeficiency syndrome (AIDS): a phase II trial with preliminary evidence of antiviral activity and low incidence of opportunistic infections. 1990;112:582-589.

19. Toole JJ, Pittman DD, Orr EC, Murtha P, Wasley LC, Kaufman RJ. A large region (approximately equal to 95 kDa) of human factor VIII is dispensable for in vitro procoagulant activity. Proceedings of the National Academy of Sciences. 1986;83(16):5939.

20. Pittman D, Alderman E, Tomkinson K, Wang J, Giles A, Kaufman R. Biochemical, immunological, and in vivo functional characterization of B-domain-deleted factor VIII. Blood. 1993;81(11):2925.

21. Martin U, Sponer G, Strein K. Differential fibrinolytic properties of the recombinant plasminogen activator BM 06.022 in human plasma and blood clot systems in vitro. Blood Coagul Fibrinolysis. 1993;4:235-242.

22. Martin U, Fischer S, Kohnert U, et al. Pharmacokinetic properties of an escherichia-coli-produced recombinant plasminogen activator (BM 06.022) in rabbits. Thromb Res. 1991;62:137-146.

23. Zacchigna M, Cateni F, Drioli S, Bonora GM. Multimeric, multifunctional derivatives of poly (ethylene glycol). Polymers. 2011;3(3):1076-1090.

24. Kang JS, DeLuca PP, Lee KC. Emerging PEGylated drugs. Expert Opin Emerg Drugs. 2009;14(2):363-380.

25. Davis F. Commentary: the origin of pegnology. Adv.Drug Del.Rev. 2002;54(4):457-458.

26. Harris JM, Chess RB. Effect of pegylation on pharmaceuticals. Nature Reviews Drug Discovery. 2003;2(3):214-221.

27. Roberts MJ, Bentley MD, Harris JM. Chemistry for peptide and protein PEGylation. 2002;54:459-476. Adv Drug Deliv Rev. 2002;54(4):459-476.

28. Veronese FM, Pasut G. PEGylation, successful approach to drug delivery. Drug Discov Today. 2005;10(21):1451-1458.

29. Varki A, Esko JD, Colley KJ. Cellular organization of glycosylation. Essentials of Glycobiology. 2nd ed. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory Press; 2009:chapter 3.

30. Houdebine LM. Genetically modified animals. introduction. Comp Immunol Microbiol Infect Dis. 2009;3-45-46.

31. Karg SR, Kallio PT. The production of biopharmaceuticals in plant systems. Biotechnol Adv. 2009;27(6):879-894.

32. Jelkmann W. Recombinant EPO production--points the nephrologist should know. 2007;22:2749-2753.

Nephrol. Dial. Transplant.

2009;24(5):1366-1368.

33. Goodman LS, Gilman A, Hardman JG, Gilman AG, Limbird LE. Goodman & Gilman’s The Pharmacological Basis of Therapeutics. 11th ed. New York: McGraw-Hill, Health Professions Division; 2006.

About the Authors

Marcela A. Saenz is currently a second-year student at the University of the Incarnate Word (UIW) Feik School of Pharmacy (FSOP) in San Antonio, Texas. She holds a bachelor of science from St. Mary’s University in San Antonio. Previously, she was employed by the otolaryngology department at the University of Texas Health Science Center where she worked as a research assistant investigating lipid drug targeting studies for various types of cancer. She holds positions in the FSOP student chapter of the APhA and SSHP-ASHP and is a member of SNPhA and Phi Delta Chi.Anjelica Seifert is currently a second-year pharmacy student at the UIW FSOP. She worked for several years at DPT Laboratories Inc as a quality control scientist and with the research and development team on finding new drug formulations. She is currently a member of Phi Delta Chi, SNPhA, and the American Chemical Society.Helen E. Smith, RPh, MS, PhD, is an assistant professor in the department of pharmaceutical sciences in the UIW FSOP. She has developed toxicology and pharmacogenomics curricula appropriate for pharmacy students. Dr. Smith has published articles and contributed book chapters on pharmacogenomics and toxicogenomics. She serves on several FSOP and UIW committees, and is a member of the National Society of Toxicology and the American Association of Colleges of Pharmacy.Adeola O. Grillo, PhD, is an assistant professor in the department of pharmaceutical sciences at the FSOP. Previously, she worked for several years in the pharmaceutical industry as a formulation scientist. Her research interests include utilizing design of experiments in the conformational analysis, formulation development, and characterization of protein therapeutics. She is a member of the American Chemical Society, American Association of Pharmaceutical Scientists, and American Association of Colleges of Pharmacy.This article originated from a class project at the UIW FSOP. The authors would like to acknowledge the class of 2014 for its involvement.