Experience with host cell protein impurities in biopharmaceuticals.

In the 40-year history of biopharmaceuticals, there have been a few cases where the final products contained residual host cell protein (HCP) impurities at levels high enough to be of concern. This article summarizes the industry experience in these cases where HCP impurities have been presented in public forums and/or published. Regulatory guidance on HCP impurities is limited to advising that products be as pure as practical, with no specified numerical limit because the risk associated with HCP exposure often depends on the clinical setting (route of administration, dose, indication, patient population) and the particular impurity. While the overall safety and purity track record of the industry is excellent, these examples illustrate several important lessons learned about the kinds of HCPs that co-purify with products (e.g., product homologs, and HCPs that react with product), and the kinds of clinical consequences of HCP impurities (e.g., direct biological activity, immunogenicity, adjuvant). The literature on industry experience with HCP impurities is scattered, and this review draws in to one reference documented examples where the data have been presented in meetings, patents, product inserts, or press releases, in addition to peer-reviewed journal articles. © 2018 American Institute of Chemical Engineers Biotechnol. Prog., 34:828-837, 2018.

Retargeting the Clostridium botulinum C2 toxin to the neuronal cytosol.

Many biological toxins are known to attack specific cell types, delivering their enzymatic payloads to the cytosol. This process can be manipulated by molecular engineering of chimeric toxins. Using toxins with naturally unlinked components as a starting point is advantageous because it allows for the development of payloads separately from the binding/translocation components. Here the Clostridium botulinum C2 binding/translocation domain was retargeted to neural cell populations by deleting its non-specific binding domain and replacing it with a C. botulinum neurotoxin binding domain. This fusion protein was used to deliver fluorescently labeled payloads to Neuro-2a cells. Intracellular delivery was quantified by flow cytometry and found to be dependent on artificial enrichment of cells with the polysialoganglioside receptor GT1b. Visualization by confocal microscopy showed a dissociation of payloads from the early endosome indicating translocation of the chimeric toxin. The natural Clostridium botulinum C2 toxin was then delivered to human glioblastoma A172 and synchronized HeLa cells. In the presence of the fusion protein, native cytosolic enzymatic activity of the enzyme was observed and found to be GT1b-dependent. This retargeted toxin may enable delivery of therapeutics to peripheral neurons and be of use in addressing experimental questions about neural physiology.

Characterization of IXINITY® (Trenonacog Alfa), a Recombinant Factor IX with Primary Sequence Corresponding to the Threonine-148 Polymorph.

The goal of these studies was to extensively characterize the first recombinant FIX therapeutic corresponding to the threonine-148 (Thr-148) polymorph, IXINITY (trenonacog alfa [coagulation factor IX (recombinant)]). Gel electrophoresis, circular dichroism, and gel filtration were used to determine purity and confirm structure. Chromatographic and mass spectrometry techniques were used to identify and quantify posttranslational modifications. Activity was assessed as the ability to activate factor X (FX) both with and without factor VIIIa (FVIIIa) and in a standard clotting assay. All results were consistent across multiple lots. Trenonacog alfa migrated as a single band on Coomassie-stained gels; activity assays were normal and showed <0.002 IU of activated factor IX (FIXa) per IU of FIX. The molecule has >97% γ-carboxylation and underwent the appropriate structural change upon binding calcium ions. Trenonacog alfa was activated normally with factor XIa (FXIa); once activated it bound to FVIIIa and FXa. When activated to FIXa, it was inhibited efficiently by antithrombin. Glycosylation patterns were similar to plasma-derived FIX with sialic acid content consistent with the literature reports of good pharmacokinetic performance. These studies have shown that trenonacog alfa is a highly pure product with a primary sequence and posttranslational modifications consistent with the common Thr-148 polymorphism of plasma-derived FIX.

Recombinant human fibrinogen that produces thick fibrin fibers with increased wound adhesion and clot density.

Human fibrinogen is a biomaterial used in surgical tissue sealants, scaffolding for tissue engineering, and wound healing. Here we report on the post-translational structure and functionality of recombinant human FI (rFI) made at commodity levels in the milk of transgenic dairy cows. Relative to plasma-derived fibrinogen (pdFI), rFI predominantly contained a simplified, neutral carbohydrate structure and >4-fold higher levels of the γ'-chain transcriptional variant that has been reported to bind thrombin and Factor XIII. In spite of these differences, rFI and pdFI were kinetically similar with respect to the thrombin-catalyzed formation of protofibrils and Factor XIIIa-mediated formation of cross-linked fibrin polymer. However, electron microscopy showed rFI produced fibrin with much thicker fibers with less branching than pdFI. In vivo studies in a swine liver transection model showed that, relative to pdFI, rFI made a denser, more strongly wound-adherent fibrin clot that more rapidly established hemostasis.

High throughput quantification of N-glycans using one-pot sialic acid modification and matrix assisted laser desorption ionization time-of-flight mass spectrometry.

Appropriate glycosylation of recombinant therapeutic glycoproteins has been emphasized in biopharmaceutical industries because the carbohydrate component can affect safety, efficacy, and consistency of the glycoproteins. Reliable quantification methods are essential to ensure consistency of their products with respect to glycosylation, particularly sialylation. Mass spectrometry (MS) has become a popular tool to analyze glycan profiles and structures, showing high resolution and sensitivity with structure identification ability. However, quantification of sialylated glycans using MS is not as reliable because of the different ionization efficiency between neutral and acidic glycans. We report here that amidation in mild acidic conditions can be used to neutralize acidic N-glycans still attached to the protein. The resulting amidated N-glycans can then be released from the protein using PNGase F, and labeled with permanent charges on the reducing end to avoid any modification and the formation of metal adducts during MS analysis. The N-glycan modification, digestion, and desalting steps were performed using a single-pot method that can be done in microcentrifuge tubes or 96-well microfilter plates, enabling high throughput glycan analysis. Using this method we were able to perform quantitative MALDI-TOF MS of a recombinant human glycoprotein to determine changes in fucosylation and changes in sialylation that were in very good agreement with a normal phase HPLC oligosaccharide mapping method.

N-glycosylation microheterogeneity and site occupancy of an Asn-X-Cys sequon in plasma-derived and recombinant protein C.

Human protein C (hPC) is glycosylated at three Asn-X-Ser/Thr and one atypical Asn-X-Cys sequons. We have characterized the micro- and macro-heterogeneity of plasma-derived hPC and compared the glycosylation features with recombinant protein C (tg-PC) produced in a transgenic pig bioreactor from two animals having approximately tenfold different expression levels. The N-glycans of hPC are complex di- and tri-sialylated structures, and we measured 78% site occupancy at Asn-329 (the Asn-X-Cys sequon). The N-glycans of tg-PC are complex sialylated structures, but less branched and partially sialylated. The porcine mammary epithelial cells glycosylate the Asn-X-Cys sequon with a similar efficiency as human hepatocytes even at these high expression levels, and site occupancy at this sequon was not affected by expression level. A distinct bias for particular structures was present at each of the four glycosylation sites for both hPC and tg-PC. Interestingly, glycans with GalNAc in the antennae were predominant at the Asn-329 site. The N-glycan structures found for tg-PC are very similar to those reported for a recombinant Factor IX produced in transgenic pig milk, and similar to the endogenous milk protein lactoferrin, which may indicate that N-glycan processing in the porcine mammary epithelial cells is more uniform than in other tissues.

Analysis of the N-glycans of recombinant human Factor IX purified from transgenic pig milk.

Glycosylation of recombinant proteins is of particular importance because it can play significant roles in the clinical properties of the glycoprotein. In this work, the N-glycan structures of recombinant human Factor IX (tg-FIX) produced in the transgenic pig mammary gland were determined. The majority of the N-glycans of transgenic pig-derived Factor IX (tg-FIX) are complex, bi-antennary with one or two terminal N-acetylneuraminic acid (Neu5Ac) moieties. We also found that the N-glycan structures of tg-FIX produced in the porcine mammary epithelial cells differed with respect to N-glycans from glycoproteins produced in other porcine tissues. tg-FIX contains no detectable Neu5Gc, the sialic acid commonly found in porcine glycoproteins produced in other tissues. Additionally, we were unable to detect glycans in tg-FIX that have a terminal Galalpha(1,3)Gal disaccharide sequence, which is strongly antigenic in humans. The N-glycan structures of tg-FIX are also compared to the published N-glycan structures of recombinant human glycoproteins produced in other transgenic animal species. While tg-FIX contains only complex structures, antithrombin III (goat), C1 inhibitor (rabbit), and lactoferrin (cow) have both high mannose and complex structures. Collectively, these data represent a beginning point for the future investigation of species-specific and tissue/cell-specific differences in N-glycan structures among animals used for transgenic animal bioreactors.

Efficient, thermally stable, second order nonlinear optical response in organic hybrid covalent/ionic self-assembled films.

A covalent/electrostatic layer-by-layer self-assembly method was used to achieve polar ordering of a water soluble, reactive dye in the fabrication of nonlinear optical (NLO) films. We observed a quadratic relationship between the second harmonic intensity I2(omega) and bilayer number for all films made with Procion Brown MX-GRN, demonstrating that the polar ordering of the chromophores is consistent in each successive bilayer. As the ionic strength of the dye deposition solution was increased to 0.5 M NaCl, the of the films increased by approximately 250% to 50 x 10(-9) esu, with a corresponding average chromophore tilt angle of 38 degrees . This was attributed to increased shielding of the dye charges which led to higher chromophore density in the bilayers. The electrooptic coefficient for films of 50 bilayers fabricated at 0.5 M NaCl was 14 +/- 2 pm/V. Importantly, these films exhibited excellent thermal stability, with only a 10% decrease in (I2(omega))(1/2) after 36 h at 85 degrees C and then 24 h at 150 degrees C. Furthermore, the (I2(omega))(1/2) recovered completely upon cooling to room temperature. These results with a commodity textile dye point to the potential value of this class of reactive chromophores and this self-assembly method for fabrication of electrooptic materials at ambient conditions from aqueous solutions.

Cloning strategy for producing brush-forming protein-based polymers.

Brush-forming polymers are being used in a variety of applications, and by using recombinant DNA technology, there exists the potential to produce protein-based polymers that incorporate unique structures and functions in these brush layers. Despite this potential, production of protein-based brush-forming polymers is not routinely performed. For the design and production of new protein-based polymers with optimal brush-forming properties, it would be desirable to have a cloning strategy that allows an iterative approach wherein the protein based-polymer product can be produced and evaluated, and then if necessary, it can be sequentially modified in a controlled manner to obtain optimal surface density and brush extension. In this work, we report on the development of a cloning strategy intended for the production of protein-based brush-forming polymers. This strategy is based on the assembly of modules of DNA that encode for blocks of protein-based polymers into a commercially available expression vector; there is no need for custom-modified vectors and no need for intermediate cloning vectors. Additionally, because the design of new protein-based biopolymers can be an iterative process, our method enables sequential modification of a protein-based polymer product. With at least 21 bacterial expression vectors and 11 yeast expression vectors compatible with this strategy, there are a number of options available for production of protein-based polymers. It is our intent that this strategy will aid in advancing the production of protein-based brush-forming polymers.

Unnatural proteins for the control of surface forces.

We introduce a new method for the stabilization of colloidal particles via the synthesis and adsorption of unnatural proteins. Biosynthesis of protein-based polymers offers the advantages of preparation of complex sequences through control of the primary sequence, monodisperse polymers, ease of combinatorial search for anchor blocks, environmentally friendly synthesis, use of water as the solvent, and incorporation of a palette of known natural proteins. We have synthesized an unnatural protein with the sequence thioredoxin-Pro(39)Glu(10) for modification of the forces between alumina particles. The polyglutamate sequence, Glu(10), is anionic (pH > 3) and is designed to anchor the protein to positively charged solids, e.g. alumina in water (pH < 9). The polyproline sequence, Pro(39), is neutral. The thioredoxin is a recombinant form of the natural globular protein with a histidine patch (His-patch-thioredoxin) and is zwitterionic. The combined thioredoxin-Pro(39) sequence is hydrophilic with pI approximately 6.3. This block is designed to remain in solution, thereby providing a steric barrier to the approach of two particles in a range of salt and pH conditions. Ellipsometry experiments show that thioredoxin-Pro(39)Glu(10) does adsorb to alumina. Force measurements with the atomic force microscopy (AFM) colloid probe technique show that adsorption of thioredoxin-Pro(39)Glu(10) leads to repulsive forces that decay exponentially with the separation between the surfaces and are independent of salt concentration in the range 0.001-0.1 M KNO(3). This demonstrates that the repulsive forces are not electrostatic. We hypothesize that the repulsion is due to confinement and loss of solvent for the adsorbed polymer; the forces are similar to those expected for a polymer brush. Force measurements between thioredoxin-coated alumina surfaces also show a repulsive force, but the force has a decay length that is consistent with electrostatic double-layer forces: the thioredoxin has not neutralized the surface charge of the underlying alumina. Our results point to interesting future experiments where recombinant DNA technology could be used to synthesize fusion proteins containing useful natural proteins and an anchor. This may allow preparation, via single-step aqueous self-assembly, of anchored proteins that maintain their natural structure. Our technique is not limited to homopolymer blocks; more complex primary sequences can be used.

Peak compression and resolution for electrophoretic separations in diverging microchannels.

We report the results of experiments and simulations on electrokinetic flow in diverging microchannels (with cross-sectional area that increases with distance along the channel). Because of conservation of mass and charge, the velocity of an analyte in the channel decreases as the channel cross-section increases. Consequently, the leading edge of a band of sample moves more slowly than the trailing edge and the sample band is compressed. Sample peak widths, rather than increasing diffusively with time, can then be controlled by the geometry of the channel and can even be made to decrease with time. We consider the possibility of using this peak compression effect to improve the resolution of electrophoretic separations. Our results indicate that for typical separations that are dispersion limited, this peak compression effect is more than offset by the decreased distance between peaks, and the separation resolution in diverging channels is worse than that found for straight channels at the same applied voltage. For separations in very short channels or at very high field strengths, however, when the separation efficiency is injection limited, the peak compression effect is dominant and diverging channels can then be used to achieve improved separation resolution.

Purification of recombinant DNA-derived factor IX produced in transgenic pig milk and fractionation of active and inactive subpopulations.

Transgenic animal bioreactors can be engineered to make gram per liter quantities of complex recombinant glycoproteins in milk. However, little is known about the limitations in post-translational processing that occurs for very complex proteins and how this impacts the task of purification. We report on the purification of recombinant factor IX (rFIX) from the milk of transgenic pigs having an expression level of 2-3 g rFIX/(l(-1) h(-1)), an expression level that is about 20-fold higher than previously reported. This purification process efficiently recovers highly active rFIX and shows that even complex mixtures like pig milk, which contains 60 g/l total endogenous milk protein and multiple subpopulations of rFIX, can be processed using conventional, non-immunoaffinity chromatographic methods. Without prior removal of caseins, heparin-affinity chromatography was used to first purify the total population of rFIX at greater than 90% yield. After the total population was isolated, the biologically active and inactive subpopulations were fractionated by high-resolution anion exchange chromatography using an ammonium acetate elution. Capillary isoelectric focusing of the active and inactive rFIX fractions demonstrated that the active subpopulations are the most acidic.