RSS-Feed abonnieren
DOI: 10.1055/a-2077-6187
A Manufacturing Strategy Utilizing a Continuous-Mode Reactor toward Homogeneous PEGylated Bioconjugate Production[1]
Abstract
Protein PEGylation is a traditional bioconjugation technology that enhances the therapeutic efficacy and in vivo half-life of proteins by the formation of covalent bonds with highly activated ester group linked polyethylene glycol (PEG). However, the high reactivity of these reagents induces a random reaction with lysine residues on the protein surface, resulting in a heterogeneous mixture of PEGylated proteins. Moreover, the traditional batch-mode reaction has risks relating to scalability and aggregation. To overcome these risks of traditional batch-mode PEGylation, a manufacturing strategy utilizing structural analysis and a continuous-flow-mode reaction was examined. A solvent exposure analysis revealed the most reactive lysine of a protein, and the continuous-flow mode modified this lysine to achieve the mono-PEGylation of two different proteins within 2 seconds. This ultrarapid modification reaction can be applied to the gram-scale manufacturing of PEGylated bioconjugates without generating aggregates. A similar trend of the exposure level of protein lysine and mono-selectivity performed by continuous-flow PEGylation was observed, which indicated that this manufacturing strategy has the potential to be applied to the production of a wide variety of bioconjugates.
#
Protein PEGylation is a modification technique that forms a covalent link between polyethylene glycol (PEG) and a protein.[3] This traditional conjugation approach enhances the therapeutic efficacy and safety profile of protein-based biopharmaceutics due to the hydrophilic nature of PEG molecules. Currently, many PEGylated proteins are approved by the US Food and Drug Administration.[4] PEG reagents possessing activated esters, such as N-hydroxysuccinimide (NHS), are widely used in this technique. These reagents modify the lysine residues of proteins to form covalent bonds and create PEGylated conjugates. The conjugation concept is simple; however, most PEGylated proteins currently on the market have wide heterogeneity owing to nonspecific protein modifications, resulting in clinical insufficiencies. Furthermore, this heterogeneous nature of current PEGylated proteins may be problematic in terms of chemistry, manufacturing, and control (CMC) challenges. The wide distribution of PEGylation sites and structural complexity make batch-to-batch consistency analysis challenging. In addition to the heterogeneous nature of PEGylated proteins, the chemical reaction to install PEG poses a risk of aggregation generation.[5] In particular, several sensitive proteins may have limited compatibility with chemical reactions such as PEGylation; therefore, mild reaction conditions with short reaction times should allow for more reliable conjugations. Furthermore, kinetic reactions, such as amidation by activated ester reagents such as NHS, can cause scalability issues.[6] To overcome the aggregation and scale gap issues in protein PEGylation, a variety of chemical reactions are required during the early stages of manufacturing, including screening of functional groups to react with the amino acid residues of the target protein and/or careful process development. A promising option for achieving homogeneous PEGylation is to utilize a continuous-flow-mode reaction. The continuous-flow reaction is a rapidly growing manufacturing process in industry.[7] [8] This process enables chemical reactions in designated systems consisting of tubes, mixers, and pipes. Flow-mode manufacturing equipment can perform sensitive novel chemical reactions that cannot be controlled in traditional stirred-batch reactors.[9] Furthermore, this process is environmentally friendly, as it reduces the risk of accidental exposure to toxic chemicals,[10] and is straightforward to scale-up.[11] Based on these advantages, a flow microreactor (FMR) system can reduce operating expenditures and facilitate automated manufacturing of industrial materials. PEGylation in continuous mode has been attempted by several groups using proprietary systems such as the on-column countercurrent chromatograph,[12] hollow-fiber membrane reactor,[13] [14] and coiled flow inverter reactor.[15] These fundamental studies demonstrated the feasibility of protein PEGylation using an FMR to achieve bioconjugation with native lysine amino acid side chains. However, this type of lysine conjugation produces a heterogeneous mixture of conjugate molecules. For example, typical IgG1 antibody proteins have greater than 80 exposed and reactive lysine residues[16] indicating that careful optimization of the reaction conditions is required despite the use of an FMR system. Amino acid nature differs among protein bases; therefore, careful reaction optimization that is dependent on protein characteristics is required. These results prompted us to investigate a versatile manufacturing strategy that supports continuous-mode PEGylation. To establish a practical strategy for homogeneous PEGylated protein manufacturing, we conducted demonstration studies using traditional lysine-based PEGylation with PEG reagents functionalized by an activated ester group. We hypothesized that the exposed lysine groups would have higher reactivity to PEGylation; therefore, we investigated the relationship between the exposure level of each lysine in the protein and the reaction selectivity. Solvent exposure analysis[17] enabled us to predict the reactivity of each lysine in the protein, and the resultant trends were compared with the site selectivity of PEGylation produced using a continuous-mode reaction. For this demonstration, a V-shaped mixing system that can be easily applied at the manufacturing scale was selected. Feasibility flow system trials enabled the mono-PEGylation of therapeutic proteins such as lysozyme and interleukin-6 (IL-6) within 2 seconds (Figure [1]). Furthermore, this ‘ultrarapid’ and mild reaction condition was successfully applied to a gram-scale synthesis of the PEGylated protein without aggregation generation, whereas batch-mode synthesis revealed scale gap issues.
Lysozyme has only six lysine groups in its sequence; therefore, the PEGylated product has a relatively low distribution (from mono- to hexa-conjugates). This inexpensive (USD 23 per gram; Millipore Sigma, accessed Jan 29, 2023) protein has been subjected to continuous PEGylation in several previous studies.[14] To predict the reactivity difference of each lysine, we conducted solvent exposure analysis of the lysozyme structure reported in the Protein Data Bank (PDB) (Figure [2]).[17] [18] The results indicated that although all six lysozyme lysines were well exposed to the solvent, the solvent-accessible surface areas (SASAs) differed slightly. The most exposed lysine (Lys116) showed a 10% higher SASA level than that of the second most exposed lysine (Lys97).
In addition to lysozyme, IL-6 levels were analyzed. IL-6 has a higher molecular weight and a greater number of lysine residues (14) than lysozyme. PEGylation of IL-6 enhanced the therapeutic efficacy of native IL-6;[19] however, only limited pharmaceutical applications were identified due to the heterogeneous nature of the conjugate.[20] First, we attempted to apply the same calculation procedure as that used for lysozyme; however, 52–60 amino acid disorders were observed in the IL-6 structure reported in the PDB (PBD: 1ALU) [Supporting Information (SI), Figure S1].[21] To understand the actual structure, we utilized the AlphaFold-2 database,[22] [23] which contains a modified IL-6 monomer structure (AlphaFold Protein Structure Database: AF-P05231-F1-model_v4) that can be used for structural and SASA analyses. These analyses showed that all 14 lysines of IL-6 were well exposed to the solvent; however, the SASA levels of each lysine differed slightly. The most exposed lysine (Lys159) showed an SASA that was only 4% higher than that of the second most exposed lysine (Lys98). These analyses revealed that IL-6 is a more challenging target than lysozyme for demonstrating this strategy.
Next, we attempted to determine whether a continuous-flow system could be used to identify small reactivity differences and perform selective lysine modifications. Several studies[11] [24] have reported that the FMR reaction is suitable for kinetically controlled reactions; therefore, our first attempt used PEG reagent screening. NOF Corporation, a main PEG reagent company in the bioconjugation field, provides reagents in different PEG units.[25] Batch-mode reaction screening revealed that methoxy-PEG-CH2-COO-NHS (5 kDa, catalog number SUNBRIGHT ME-050AS) showed the highest reactivity. The PEGylation reaction was completed in less than 1 minute using SUNBRIGHT ME-050AS, whereas the other PEG reagents continued to react after 3 minutes (SI, Figure S2). These reactivity trends were similar to that of the half-life of these PEG reagents[25] and supported the hetero-functional group placing in neighboring positions, which enhanced the reactivity of the NHS group and reduced hydrolysis resistance.
In the development of continuous reactions, several parameters must be considered[26] [27] such as the tube length, time, pH, and temperature of the reaction. However, the appropriate parameter depends on the target protein behavior and this experimental design approach is beyond the scope of the current feasibility study. The purpose of this study was to demonstrate the developability of this modification strategy (SASA analysis scouting the target lysine followed by a continuous-flow reaction). For this purpose, we selected a tentative condition to apply to the PEGylation. The most important factors affecting mixability in flow systems are the geometry and diameter of the mixer unit, and we selected conditions previously reported as effective for early selection.[11,26] The iodide–iodate reaction, termed the Dushman reaction, is commonly used to evaluate mixing efficiencies[28,29] and we previously confirmed the high reproducibility of this method,[26] whereby V-shaped mixers produced the most efficient mixability for several geometric types (Vortex-shaped, T-shaped, and V-shaped; SI, Figure S3). Therefore, a V-shaped mixer was selected for PEGylation. Diameter is also a critical factor that affects the mixability of the flow system. Based on our previous study,[11] a diameter of 0.25 mm was selected for use in the present investigation. Previously, our group succeeded in performing a tandem reaction (reduction of disulfide bonds of an antibody followed by conjugation with a cytotoxic drug) in continuous mode to produce antibody–drug conjugates (ADCs). A 0.25 mm diameter V-shaped mixer sufficiently converted naked antibodies into ADCs, while a mixer with a diameter greater than 0.5 mm did not reach the target drug-to-antibody ratio. We expected that the 0.25 mm diameter V-shaped mixer that was applied to a complicated tandem bioconjugation to produce ADC could also be applied to a single reaction PEGylation. The flow system consisted of two V-shaped mixers (Mixer-1 and Mixer-2 in Figure [1], see also Figure S3 in the SI) and two 0.25 mm reactors (Reactor-1 and Reactor-2 in Figure [1], see also Figure S3 in the SI). The diameter of the reactors was 1.0 mm, and using a high flow rate (8 mL/min for lysozyme and 2 mL/min for PEG reagent), PEGylation was completed in 1.17 s (residence time in Reactor-1). The detailed calculations are below:
• reactor length; 250 mm = 0.25 m
• reactor diameter; 1 mm = 1 × 10–3 m
• reactor volume; 0.196 mL
• flow rate; 8 mL/min (lysozyme) and 2 mL/min (PEG reagent)
• reaction time = 1.17 s = 0.196 (mL)/10 (mL/min)
This rapid reaction mode enabled the production of gram-scale PEGylated lysozyme within 15 minutes. The conversion yield and mono-PEGylation selectivity were analyzed using RP-HPLC (Figure [3] and SI, Figure S4). Additionally, a direct comparison was conducted between the PEGylated lysozyme produced using the batch-mode approach and that synthesized through the continuous-flow system.
Continuous-flow mode converted 54% of the lysozyme into mono-PEGylated conjugates, showing a mono-selectivity of 78%. In contrast, batch mode provided less than 40% mono-conversion. In addition, the reproducibility and scalability of the batch mode were clearly problematic. In larger-scale syntheses, the mono-conversion rate decreased, and an overreaction was observed. In addition to the conversion yield, the ineffective mixability of the batch mode triggered aggregate generation [greater than 44% by size exclusion chromatography (SEC) analysis;[30] SI, Figure S5], whereas no aggregates were observed in the gram-scale conjugates produced in continuous mode. We also performed a comparative SDS-PAGE analysis of the two PEGylated proteins to confirm their mono-selectivity. Due to the complex structure of bioconjugates, it is recommended to use several analytical methods to obtain more accurate results.[31] A comparison of the SDS-PAGE results supported the mono-selectivity of the PEGylated conjugates produced by an FMR (SI, Figure S7).
Next, continuous mono-PEGylation was performed to modify IL-6 (Figure [4]). Similar comparisons were obtained for the batch and continuous-flow modes as those observed with lysozyme. Continuous-flow mode converted 30% of IL-6 into mono-PEGylated conjugates with a mono-selectivity of 46%, while batch mode presented several issues (low mono-selectivity, reproducibility, and scalability, and high aggregation generation (greater than 54% by SEC analysis; SI, Figure S5). IL-6 has more lysine residues, most of which are more exposed to solvent than those of lysozyme. Based on previously reported results,[32] [33] lysine residues with more than 20–25 SASAs can be defined as solvent exposed. This means that all lysine residues of both lysozyme and IL-6 used in this study are well exposed. However, the highest SASA value for lysozyme lysine was 118.9 for Lys116, whereas IL-6 has eight lysine residues with higher SASA values. In addition, scatter plots and box plots were created for clear comparison of the variability of SASA values for each lysine residue in the respective proteins (SI, Figure S6). It can be hypothesized that if the variability of SASA values for each lysine residue is low, it may be difficult to achieve selective PEGylation. However, the results showed that lysozyme has a lower variability in SASA values compared to IL-6, indicating that lysozyme has a higher prevalence of lysine residues with similar SASA values. On the other hand, IL-6 shows a higher variability in SASA values, but has a higher prevalence of highly exposed lysine residues compared to lysozyme, as mentioned earlier. Therefore, it can be speculated that the decrease in selectivity may be attributed to the higher prevalence of highly exposed lysine residues. Therefore, this exposure level difference caused a relatively lower mono-conversion rate in the IL-6 modification. Although this speculation requires further examples for conclusive assertions, the present results provide valuable guidance for evaluating the relationship between the kinetics-dominant reaction mediated by active groups such as NHS ester and SASA. We also performed an SDS-PAGE analysis comparison of these two PEGylated proteins, showing mono-selectivity of the PEGylated conjugates (SI, Figure S7).
In conclusion, a manufacturing strategy utilizing SASA analysis and continuous-flow-process-mediated PEGylation was achieved using two proteins that have potential for clinical use. The selected 0.25 mm diameter V-shaped mixer performed rapid (1.17 s) protein modification to achieve mono-selective PEGylation without inducing appreciable aggregation. All flow processes were conducted using a scaled-down manufacturing approach with a sequential mixing system. Furthermore, these early-stage (not thoroughly optimized) reaction conditions were able to generate gram-scale PEGylated lysozyme within 15 minutes. The exposed lysine trend calculated by SASA analysis was similar to that of mono-selective production in the continuous mode. The results described herein indicate that the strategy of using an SASA with continuous-flow chemistry has the potential for application in a wide variety of protein modifications.[34]
To establish a robust and reliable PEGylation using the continuous mode, some challenges still remain. For example, process development including understanding critical process parameters, PEG reagent screening to replace NHS ester having risk of overreactions,[35] detailed conjugation site analysis by peptide mapping, and normal operating range identification was not complete in the present study. Further investigations to establish the ideal continuous-mode-mediated PEGylation are currently underway.
Lysozyme (chicken egg white) was purchased from Sigma-Aldrich (USA). IL-6 protein was expressed and purified as previously reported.[36] The PEG reagent (methoxy-PEG-CH2-COO-NHS, 5 kDa, catalog number SUNBRIGHT ME-050AS) was purchased from NOF Corporation (Japan). All other chemical reagents were purchased from Sigma-Aldrich (USA).
#
SASA Calculations
The SASA calculations were performed using the Bioluminate software suite (Bioluminate, version 2022-2, Schrödinger, Inc.). The initial structures of lysozyme (PDB: 1DPX)[37] and IL-6 (AlphaFold Protein Structure Database: AF-P05231-F1-model_v4)[22] [23] were protonated and minimized, and the SASA score was calculated using the Residue Analysis module.
#
Molecular Modeling
The model structure of the proteins was generated as described previously.[38]
#
Experimental Procedure for PEGylation Using a Batch Reactor
PEG reagent in DMSO (10.5 mg/mL, 1.66 mg or 16.6 mg) was added to a solution of protein (0.5 mg or 5 mg) in 20 mM borate buffer (pH 9.0). The mixture was then incubated for 5 min at 20 °C. An excess of 50 mM glycine and 1 M acetate buffer (pH 4.7) was added to adjust the pH of the reaction mixture (to approximately pH 7.0) and the mixture was stirred for an additional 15 min. Subsequently, the reaction mixture was purified using a PD-10 desalting column and elution with 50 mM acetate buffer (pH 5.5).
#
Experimental Procedure for PEGylation Using a Flow Reactor
V- and T-shaped stainless-steel mixers with inner diameters of 0.25 mm (Sankoh-seiki, Tokyo, Japan) were used as Mixer-1 and Mixer-2, respectively, as shown in Figure [1] (see also Figure S3 in the SI). Reactor-1 (1.0 mm i.d., 0.25 m length) and Reactor-2 (1.0 mm i.d., 0.5 m length) were also made of stainless steel. Borate buffer (50 mM) containing the protein (1.05 mg/mL for lysozyme, 1.47 mg/mL for IL-6, pH 9.0) was added to Mixer-1 through Flow-1 (flow rate: 8 mL/min). PEG reagent in DMSO (3.5 mg/mL, 4.2 g for lysozyme, 690 mg for IL-6) was added to Mixer-1 through Flow-2 (flow rate: 2 mL/min). The output mixture from Reactor-1 and that delivered from Flow-3 were mixed in Mixer-2. Glycine in phosphate buffer (50 mM, excess, pH 7.4) was added to Mixer-2 through Flow-3. The output mixture of Reactor-2 was eluted into a fraction collector, to which an excess of 1 M acetate buffer (pH 4.7) was added for neutralization. This elution was combined and purified using large desalting columns as previously reported.[39]
#
RP-HPLC Methods
RP-HPLC analysis was performed on a PLRP-S 300A column, 3 μm, 150 mm × 4.6 mm (Agilent Technologies, USA), connected to an Agilent 1260 HPLC system containing a binary gradient pump, temperature-controlled column compartment, autosampler, and diode array detector. The system was run at 1.0 mL/min at 70 °C using 0.1% TFA in water (mobile phase A, MPA) and 0.1% TFA in acetonitrile (mobile phase B, MPB), and absorbance was monitored at 280 nm (reference wavelength at 450 nm). All protein samples (2.0 mg/mL, 20 μL) were injected into the system sequentially and eluted with a 25 min method consisting of a 2 min isocratic hold at 24% MPB, an 18 min linear gradient from 24% to 56% MPB, a 3 min wash using 95% MPB, and a 2 min reequilibration at 24% MPB.
#
Other Instruments/Analytical Methods
The concentration of proteins was determined using the Slope Spectroscopy® method with a SoloVPE system.[40] Size exclusion chromatography[30] was performed as previously reported.
#
#
Conflict of Interest
The authors declare no conflict of interest.
Acknowledgment
The authors wish to thank their colleagues from Ajinomoto Co., Inc., as follows: Mr.
Michiya Kanzaki, Mr. Yasuhiro Takenaka, Dr. Seiichi Sato, and Dr. Tatsuya Okuzumi
for their helpful comments and suggestions regarding this study.
The manuscript was written with contributions from all authors. All authors approved
the final version of the manuscript. Y.N. and Y.E. contributed equally.
Supporting Information
- Supporting information for this article is available online at https://doi-org.accesdistant.sorbonne-universite.fr/10.1055/a-2077-6187.
Included are figures showing schematic diagrams of the mixers, and SEC and RP-HPLC
chromatograms, as described in the text.
- Supporting Information
-
References
- 1 The submitted (pre-acceptance) version of this manuscript was published as a preprint in ChemRxiv: Nakahara Y, Endo Y, Takahashi K, Kawaguchi T, Kato K, Matsuda Y, Nagaki A. ChemRxiv 2023; preprint
- 2 Current address: Ajinomoto Bio-Pharma Services, 11040 Roselle Street, San Diego, CA 92121, USA.
- 3a Turecek PL, Bossard MJ, Schoetens F, Ivens IA. J. Pharm. Sci. 2016; 105: 460
- 3b Mishra P, Nayak B, Dey RK. Asian J. Pharm. Sci. 2016; 11: 337
- 3c Mao L, Russell AJ, Carmali S. Bioconjugate Chem. 2022; 33: 1643
- 4 Armengol ES, Unterweger A, Laffleur F. Drug Dev. Ind. Pharm. 2022; 48: 129
- 5 Rajan R, Ahmed S, Sharma N, Kumar N, Debas A, Matsumura K. Mater. Adv. 2021; 2: 1139
- 6 Rossetti I, Compagnoni M. Chem. Eng. J. 2016; 296: 56
- 7 Baumann M, Moody TS, Smyth M, Wharry S. Org. Process Res. Dev. 2020; 24: 1802
- 8 Hughes DL. Org. Process Res. Dev. 2020; 24: 1850
- 9 Takahashi Y, Nagaki A. Molecules 2019; 24: 1532
- 10 Movsisyan M, Delbeke EI, Berton JK, Battilocchio C, Ley SV, Stevens CV. Chem. Soc. Rev. 2016; 45: 4892
- 11 Nakahara Y, Mendelsohn BA, Matsuda Y. Org. Process Res. Dev. 2022; 26: 2766
- 12 Ingold O, Pfister D, Morbidelli M. React. Chem. Eng. 2016; 1: 218
- 13 Shang X, Ghosh R. J. Membr. Sci. 2014; 451: 177
- 14 Madadkar P, Selvaganapathy PR, Ghosh R. Biomicrofluidics 2018; 12: 044114
- 15 Kateja N, Nitika, Dureja S, Rathore AS. J. Biotechnol. 2020; 322: 79
- 16 Matsuda Y, Leung M, Tawfiq Z, Fujii T, Mendelsohn BA. Anal. Sci. 2021; 37: 1171
- 17 Yamazaki S, Shikida N, Takahashi K, Matsuda Y, Inoue K, Shimbo K, Mihara Y. Bioorg. Med. Chem. Lett. 2021; 51: 128360
- 18 Matsuda Y, Chakrabarti A, Takahashi K, Yamada K, Nakata K, Okuzumi T, Mendelsohn BA. J. Chromatogr. B: Anal. Technol. Biomed. Life Sci. 2021; 1177: 122753
- 19 Tsutsumi YE, Tsunoda SS, Kamada H, Kihira T, Kaneda Y, Ohsugi Y, Mayumi T. Thromb. Haemost. 1997; 77: 168
- 20 Tsunoda S, Ishikawa T, Watanabe M, Kamada H, Yamamoto Y, Tsutsumi Y, Hirano T, Mayumi T. Br. J. Haematol. 2001; 112: 181
- 21 Somers W, Stahl M, Seehra JS. EMBO J. 1997; 16: 989
- 22 Jumper J, Evans R, Pritzel A, Green T, Figurnov M, Ronneberger O, Tunyasuvunakool K, Bates R, Žídek A, Potapenko A, Bridgland A, Meyer C, Kohl SA. A, Ballard AJ, Cowie A, Romera-Paredes B, Nikolov S, Jain R, Adler J, Back T, Petersen S, Reiman D, Clancy E, Zielinski M, Steinegger M, Pacholska M, Berghammer T, Bodenstein S, Silver D, Vinyals O, Senior AW, Kavukcuoglu K, Kohli P, Hassabis D. Nature 2021; 596: 583
- 23 Varadi M, Anyango S, Deshpande M, Nair S, Natassia C, Yordanova G, Yuan D, Stroe O, Wood G, Laydon A, Žídek A, Green T, Tunyasuvunakool K, Petersen S, Jumper J, Clancy E, Green R, Vora A, Lutfi M, Figurnov M, Cowie A, Hobbs N, Kohli P, Kleywegt G, Birney E, Hassabis D, Velankar S. Nucleic Acids Res. 2022; 50: D439
- 24 Konermann L. J. Phys. Chem. A 1999; 103: 7210
- 25 NOF Corporation (accessed May 8, 2023). https://www.nofamerica.com/store/index.php?dispatch=categories.view&category_id=7
- 26 Endo Y, Furusawa M, Shimazaki T, Takahashi Y, Nakahara Y, Nagaki A. Org. Process Res. Dev. 2019; 23: 635
- 27 Taylor CJ, Baker A, Chapman MR, Reynolds WR, Jolley KE, Clemens G, Smith GE, Blacker AJ, Chamberlain TW, Christie SD. R, Taylor BA, Bourne RA. J. Flow Chem. 2021; 11: 75
- 28 Reckamp JM, Bindels A, Duffield S, Liu YC, Bradford E, Ricci E, Susanne F, Rutter A. Org. Process Res. Dev. 2017; 21: 816
- 29 Ehrfeld W, Golbig K, Hessel V, Löwe H, Richter T. Ind. Eng. Chem. Res. 1999; 38: 1075
- 30 Fujii T, Reiling C, Quinn C, Kliman M, Mendelsohn BA, Matsuda Y. Explor. Targeted Anti-Tumor Ther. 2021; 2: 576
- 31 Tawfiq Z, Matsuda Y, Alfonso MJ, Clancy C, Robles V, Leung M, Mendelsohn BA. Anal. Sci. 2020; 36: 871
- 32 Savojardo C, Manfredi M, Martelli PL, Casadio R. Front. Mol. Biosci. 2021; 7: 626363
- 33 Kajander T, Kahn PC, Passila SH, Cohen DC, Lehtiö L, Adolfsen W, Warwicker J, Schell U, Goldman A. Structure 2000; 8: 1203
- 34 An FMR has the advantage of high throughput to find appropriate reaction conditions. As a hypothetical case study, we have attempted to compare the estimated development time for both batch and FMR cases. In the case of batch development, it is assumed that each parameter study takes 10 minutes to find ideal scale-up conditions, so a total of 300 minutes is required if 30 parameter studies are performed. On the other hand, in the case of FMR development, the run time for each parameter study is less than 1 minute. Therefore, a total of only 30 minutes is required to complete 30 parameter studies.
- 35 Fujii T, Matsuda Y, Seki T, Shikida N, Iwai Y, Ooba Y, Takahashi K, Isokawa M, Kawaguchi S, Hatada N, Watanabe T, Takasugi R, Nakayama A, Shimbo K, Mendelsohn BA, Okuzumi T, Yamada K. Bioconjugate Chem. 2023; 34: 728
- 36 Ejima D, Watanabe M, Sato Y, Date M, Yamada N, Takahara Y. Biotechnol. Bioeng. 1999; 62: 301
- 37 Weiss MS, Palm GJ, Hilgenfeld R. Acta Crystallogr., Sect. D 2000; 56: 952
- 38 Matsuda Y, Seki T, Yamada K, Ooba Y, Takahashi K, Fujii T, Kawaguchi S, Narita T, Nakayama A, Kitahara Y, Mendelsohn BA, Okuzumi T. Mol. Pharmaceutics 2021; 18: 4058
- 39 Matsuda Y, Tawfiq Z, Leung M, Mendelsohn BA. ChemistrySelect 2020; 5: 8435
- 40 Seki T, Yamada K, Ooba Y, Fujii T, Narita T, Nakayama A, Kitahara Y, Mendelsohn BA, Matsuda Y, Okuzumi T. Front. Biosci.-Landmark 2022; 27: 234
Corresponding Authors
Publikationsverlauf
Eingereicht: 22. Februar 2023
Angenommen nach Revision: 19. April 2023
Accepted Manuscript online:
19. April 2023
Artikel online veröffentlicht:
30. Mai 2023
© 2024. Thieme. All rights reserved
Georg Thieme Verlag KG
Rüdigerstraße 14, 70469 Stuttgart, Germany
-
References
- 1 The submitted (pre-acceptance) version of this manuscript was published as a preprint in ChemRxiv: Nakahara Y, Endo Y, Takahashi K, Kawaguchi T, Kato K, Matsuda Y, Nagaki A. ChemRxiv 2023; preprint
- 2 Current address: Ajinomoto Bio-Pharma Services, 11040 Roselle Street, San Diego, CA 92121, USA.
- 3a Turecek PL, Bossard MJ, Schoetens F, Ivens IA. J. Pharm. Sci. 2016; 105: 460
- 3b Mishra P, Nayak B, Dey RK. Asian J. Pharm. Sci. 2016; 11: 337
- 3c Mao L, Russell AJ, Carmali S. Bioconjugate Chem. 2022; 33: 1643
- 4 Armengol ES, Unterweger A, Laffleur F. Drug Dev. Ind. Pharm. 2022; 48: 129
- 5 Rajan R, Ahmed S, Sharma N, Kumar N, Debas A, Matsumura K. Mater. Adv. 2021; 2: 1139
- 6 Rossetti I, Compagnoni M. Chem. Eng. J. 2016; 296: 56
- 7 Baumann M, Moody TS, Smyth M, Wharry S. Org. Process Res. Dev. 2020; 24: 1802
- 8 Hughes DL. Org. Process Res. Dev. 2020; 24: 1850
- 9 Takahashi Y, Nagaki A. Molecules 2019; 24: 1532
- 10 Movsisyan M, Delbeke EI, Berton JK, Battilocchio C, Ley SV, Stevens CV. Chem. Soc. Rev. 2016; 45: 4892
- 11 Nakahara Y, Mendelsohn BA, Matsuda Y. Org. Process Res. Dev. 2022; 26: 2766
- 12 Ingold O, Pfister D, Morbidelli M. React. Chem. Eng. 2016; 1: 218
- 13 Shang X, Ghosh R. J. Membr. Sci. 2014; 451: 177
- 14 Madadkar P, Selvaganapathy PR, Ghosh R. Biomicrofluidics 2018; 12: 044114
- 15 Kateja N, Nitika, Dureja S, Rathore AS. J. Biotechnol. 2020; 322: 79
- 16 Matsuda Y, Leung M, Tawfiq Z, Fujii T, Mendelsohn BA. Anal. Sci. 2021; 37: 1171
- 17 Yamazaki S, Shikida N, Takahashi K, Matsuda Y, Inoue K, Shimbo K, Mihara Y. Bioorg. Med. Chem. Lett. 2021; 51: 128360
- 18 Matsuda Y, Chakrabarti A, Takahashi K, Yamada K, Nakata K, Okuzumi T, Mendelsohn BA. J. Chromatogr. B: Anal. Technol. Biomed. Life Sci. 2021; 1177: 122753
- 19 Tsutsumi YE, Tsunoda SS, Kamada H, Kihira T, Kaneda Y, Ohsugi Y, Mayumi T. Thromb. Haemost. 1997; 77: 168
- 20 Tsunoda S, Ishikawa T, Watanabe M, Kamada H, Yamamoto Y, Tsutsumi Y, Hirano T, Mayumi T. Br. J. Haematol. 2001; 112: 181
- 21 Somers W, Stahl M, Seehra JS. EMBO J. 1997; 16: 989
- 22 Jumper J, Evans R, Pritzel A, Green T, Figurnov M, Ronneberger O, Tunyasuvunakool K, Bates R, Žídek A, Potapenko A, Bridgland A, Meyer C, Kohl SA. A, Ballard AJ, Cowie A, Romera-Paredes B, Nikolov S, Jain R, Adler J, Back T, Petersen S, Reiman D, Clancy E, Zielinski M, Steinegger M, Pacholska M, Berghammer T, Bodenstein S, Silver D, Vinyals O, Senior AW, Kavukcuoglu K, Kohli P, Hassabis D. Nature 2021; 596: 583
- 23 Varadi M, Anyango S, Deshpande M, Nair S, Natassia C, Yordanova G, Yuan D, Stroe O, Wood G, Laydon A, Žídek A, Green T, Tunyasuvunakool K, Petersen S, Jumper J, Clancy E, Green R, Vora A, Lutfi M, Figurnov M, Cowie A, Hobbs N, Kohli P, Kleywegt G, Birney E, Hassabis D, Velankar S. Nucleic Acids Res. 2022; 50: D439
- 24 Konermann L. J. Phys. Chem. A 1999; 103: 7210
- 25 NOF Corporation (accessed May 8, 2023). https://www.nofamerica.com/store/index.php?dispatch=categories.view&category_id=7
- 26 Endo Y, Furusawa M, Shimazaki T, Takahashi Y, Nakahara Y, Nagaki A. Org. Process Res. Dev. 2019; 23: 635
- 27 Taylor CJ, Baker A, Chapman MR, Reynolds WR, Jolley KE, Clemens G, Smith GE, Blacker AJ, Chamberlain TW, Christie SD. R, Taylor BA, Bourne RA. J. Flow Chem. 2021; 11: 75
- 28 Reckamp JM, Bindels A, Duffield S, Liu YC, Bradford E, Ricci E, Susanne F, Rutter A. Org. Process Res. Dev. 2017; 21: 816
- 29 Ehrfeld W, Golbig K, Hessel V, Löwe H, Richter T. Ind. Eng. Chem. Res. 1999; 38: 1075
- 30 Fujii T, Reiling C, Quinn C, Kliman M, Mendelsohn BA, Matsuda Y. Explor. Targeted Anti-Tumor Ther. 2021; 2: 576
- 31 Tawfiq Z, Matsuda Y, Alfonso MJ, Clancy C, Robles V, Leung M, Mendelsohn BA. Anal. Sci. 2020; 36: 871
- 32 Savojardo C, Manfredi M, Martelli PL, Casadio R. Front. Mol. Biosci. 2021; 7: 626363
- 33 Kajander T, Kahn PC, Passila SH, Cohen DC, Lehtiö L, Adolfsen W, Warwicker J, Schell U, Goldman A. Structure 2000; 8: 1203
- 34 An FMR has the advantage of high throughput to find appropriate reaction conditions. As a hypothetical case study, we have attempted to compare the estimated development time for both batch and FMR cases. In the case of batch development, it is assumed that each parameter study takes 10 minutes to find ideal scale-up conditions, so a total of 300 minutes is required if 30 parameter studies are performed. On the other hand, in the case of FMR development, the run time for each parameter study is less than 1 minute. Therefore, a total of only 30 minutes is required to complete 30 parameter studies.
- 35 Fujii T, Matsuda Y, Seki T, Shikida N, Iwai Y, Ooba Y, Takahashi K, Isokawa M, Kawaguchi S, Hatada N, Watanabe T, Takasugi R, Nakayama A, Shimbo K, Mendelsohn BA, Okuzumi T, Yamada K. Bioconjugate Chem. 2023; 34: 728
- 36 Ejima D, Watanabe M, Sato Y, Date M, Yamada N, Takahara Y. Biotechnol. Bioeng. 1999; 62: 301
- 37 Weiss MS, Palm GJ, Hilgenfeld R. Acta Crystallogr., Sect. D 2000; 56: 952
- 38 Matsuda Y, Seki T, Yamada K, Ooba Y, Takahashi K, Fujii T, Kawaguchi S, Narita T, Nakayama A, Kitahara Y, Mendelsohn BA, Okuzumi T. Mol. Pharmaceutics 2021; 18: 4058
- 39 Matsuda Y, Tawfiq Z, Leung M, Mendelsohn BA. ChemistrySelect 2020; 5: 8435
- 40 Seki T, Yamada K, Ooba Y, Fujii T, Narita T, Nakayama A, Kitahara Y, Mendelsohn BA, Matsuda Y, Okuzumi T. Front. Biosci.-Landmark 2022; 27: 234