Chemometrics and integrative LC-MS identify compositional variability in cottonseed hydrolysates and its effects on CHO cell growth and antibody productivity

Chemometrics and integrative LC-MS identify compositional variability in cottonseed hydrolysates and its effects on CHO cell growth and antibody productivity | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article Chemometrics and integrative LC-MS identify compositional variability in cottonseed hydrolysates and its effects on CHO cell growth and antibody productivity Yongjing Xie, Michael Butler This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-9267576/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract Protein hydrolysates have attracted increasing interest as cost-effective media supplements for mammalian cell culture, including Chinese hamster ovary (CHO) cells used widely in biopharmaceutical production. However, the biological basis of their beneficial effects remains poorly understood because of their compositional complexity and batch-to-batch variability. In this study, time-resolved compositional profiles of culture media supplemented with different batches of cottonseed hydrolysates were analysed by liquid chromatography-high resolution mass spectrometry (LC-HRMS) and related to CHO DG44 cell growth, antibody productivity, cellular metabolism, and antibody glycosylation. During 10-day batch cultures, hydrolysate supplementation prolonged high cell viability and significantly enhanced antibody productivity, despite lower peak viable cell densities than the control. Hydrolysate-supplemented cultures also showed reduced lactate and ammonia accumulation, consistent with altered nutrient utilization and metabolic activity. In addition, cottonseed hydrolysates significantly increased antibody galactosylation to varying degrees. Chemometric analysis further linked hydrolysate compositional variability to culture performance and identified 25 signature features associated with cell growth and antibody production. These findings provide complementary insight into how hydrolysate composition relates to functional performance in CHO cell culture and support the development of a method to identify hydrolysates that support high performance biopharmaceutical manufacturing. Chemometrics Chinese hamster ovary cells high resolution tandem mass spectrometry glycan analysis monoclonal antibody protein hydrolysates Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Introduction Chinese hamster ovary (CHO) cells are the predominant mammalian host used for the large-scale commercial production of recombinant therapeutic proteins, vaccines, and monoclonal antibodies. Their widespread use is largely due to their ability to grow to high cell densities in serum-free suspension culture while supporting correct protein folding, assembly, and post-translational modification (Kim et al., 2012 ; Wurm, 2004 ). In the biopharmaceutical industry, improving CHO cell cultivation and increasing productivity while maintaining consistent product quality remain major priorities for reducing manufacturing costs. Considerable advances have been made in bioprocessing through vector design, cell engineering and stable cell line development (Kim et al., 2012 ; Yang et al., 2022 ). However, culture medium development remains one of the most effective strategies for increasing cell density and product titre, and improving product quality attributes (Combe and Sokolenko, 2021 ; Ritacco et al., 2018 ; Yao and Asayama, 2017 ). Protein hydrolysates derived from a range of raw materials have attracted increasing interest as cost-effective supplements for CHO cell culture. They are widely used as serum substitutes and have been reported to enhance cell growth, improve product titre, and influence product quality (André Siemensma, 2010; Franek et al., 2000 ; Rodriguez et al., 2021 ). For example, Farges-Haddani et al. showed that rapeseed-derived hydrolysate fractions stimulated CHO cell growth significantly, enabling the formulation of a protein-free medium with reduced cell death, increased recombinant interferon gamma (IFN-γ) production, and faster adaptation to serum-free conditions (B. Farges-Haddani, 2006). However, the effects of hydrolysates are not uniform across materials or processes. A soy hydrolysate was reported to support high cell growth but not human beta interferon (β-IFN) productivity, whereas a yeast hydrolysate supported lower growth but β-IFN productivity comparable to that obtained with the animal-derived hydrolysate Primatone RL (Spearman et al., 2014 ). Likewise, two cotton-derived hydrolysates were shown to substantially prolong CHO cell growth and enhance monoclonal antibody production (Obaidi et al., 2021 ). Together, these studies demonstrate that protein hydrolysates can provide substantial benefits, but their functional effects are variable. The mechanisms by which hydrolysates influence CHO cell culture performance remain poorly understood. One likely explanation is that they enrich the medium with free amino acids and short peptides that serve as additional nutrient sources (Juliet Lobo-Alfonso, 2008 ; Ng et al., 2020 ). In addition, short peptides may exert more specific bioactivities, including anti-apoptotic, antioxidant, immunomodulatory, and antimicrobial effects, which may support cell survival and growth (Ho et al., 2021 ; Juliet Lobo-Alfonso, 2008 ). Previous studies have also suggested that hydrolysates may alter the expression of metabolic regulators, reduce oxidative stress, influence cell cycle progression, and affect cell morphology and energy metabolism (Du et al., 2022 ; Hu et al., 2018 ). More specifically, in relation to cottonseed hydrolysates, Kumar et al. performed transcriptomic analyses and identified differentially expressed genes, associated with pathways known to prolong cell growth and increase antibody titre in supplemented CHO cultures (Kumar et al., 2021 ). Dhara et al. further combined extracellular metabolomics with tandem mass tag (TMT) proteomics to show that cottonseed hydrolysate supplementation altered cellular functions critical to growth and protein productivity, including metabolism, protein processing, and apoptosis (Dhara et al., 2023 ). A major challenge in the use of protein hydrolysates is their intrinsic compositional variability, which can lead to inconsistent effects on bioprocess performance (Xie and Butler, 2025 ). Hydrolysates are complex mixtures that may contain peptides, amino acids, carbohydrates, minerals, lipids, vitamins, and inorganic salts (Franek et al., 2000 ; Rodriguez et al., 2021 ; Schaafsma, 2009 ). Their composition depends not only on the source material but also on the method and extent of hydrolysis, which may involve acid, alkaline, thermal, or enzymatic processing (Colla et al., 2017 ; Kristinsson and Rasco, 2000 ; Zhang et al., 2019 ). As a result, different hydrolysate batches may elicit substantially different cellular responses, including changes in transcriptomic, metabolomic, and proteomic profiles. This variability presents a significant challenge for industrial bioprocessing, where media formulations are ideally low-cost, animal-component-free, chemically defined, and capable of supporting robust cell growth, high productivity, and consistent product quality. Therefore, reducing or at least understanding batch-to-batch compositional variability in protein hydrolysates is of considerable practical importance for ensuring process consistency and manufacturing reliability. In a previous study, Luo et al. addressed this issue by combining high-resolution nuclear magnetic resonance (NMR) spectroscopy with partial least squares (PLS) analysis to screen intact soy hydrolysate lots for recombinant monoclonal antibody manufacturing, although glycosylation effects were not examined (Luo and Pierce, 2012 ). In the present study, rather than repeating earlier transcriptomic or proteomic approaches, we applied reverse-phase ultra-high-performance liquid chromatography high-resolution electrospray ionization tandem mass spectrometry (RP-UHPLC-HR-ESI-MS/MS) to profile the composition of chemically defined culture media supplemented with different batches of cottonseed hydrolysates produced using different raw materials and/or processing conditions. Dihydrofolate reductase-deficient CHO DG44 cells producing bevacizumab were used as a model system to investigate changes in cell growth, antibody productivity, cellular metabolism, and N-glycan profiles. Chemometric analysis, including partial least squares discriminant analysis (PLS-DA) (Lee et al., 2018 ; Richard G. Brereton, 2014 ) and PatternHunter analysis (Li et al., 2004 ; Ma et al., 2002 ), was used to identify signature components associated with cell growth and bevacizumab production. The main objective of this work was to evaluate the effects of batch-dependent compositional variability in cottonseed hydrolysates on CHO DG44 culture performance and to correlate these effects with LC-MS-derived compositional signatures. This study provides complementary insight into the relationship between hydrolysate composition and functional performance and may support the future identification of key components that enhance CHO cell growth and productivity in biopharmaceutical manufacturing. Materials and Methods 1. Materials and reagents Three batches of each of three cottonseed protein hydrolysates (CPHs) Ultrapep™ cotton (Part NO: S-2803719, labelled as H1A, H1B, H1C), Hypep™ 7504 25LB (Part NO: S-1729241, labelled as H2A, H2B, H2C) and Hypep™ 7504 25LBS (Part NO: S-2556075, labelled as H3A, H3B, H3C) were kindly donated by Kerry lnc. (Beloit, WI 53511, USA). Each type of hydrolysate (H1, H2, H3) were derived from different batches of raw material and/or different processing protocols. Stock solutions for each cottonseed protein hydrolysate (250 mg/mL) were prepared in Gibco™ CD CHO medium (Part NO: Gibco TM 10743029, Fisher Scientific, Dublin, Ireland) supplemented with 4.0 mM Gibco™ L-glutamine (Part NO: Gibco TM 25030024, Fisher Scientific, Dublin, Ireland) and filtered through 0.2 µm sterile syringe filters (Part NO: 15206869, Fisher Scientific, Dublin, Ireland). The stock solutions were stored at 4°C until use. 2. CHO cell line and cell maintenance The stably transfected dihydrofolate reductase (DHFR-) deficient CHO cell line (CHO DG44) expressing a monoclonal antibody was kindly donated by Kerry lnc. (Beloit, WI 53511, USA). The cells were routinely grown in suspension in CD CHO medium supplemented with 4.0 mM L-glutamine (Gibco/ Fisher Scientific, Dublin, Ireland). Cells were cultured in 250 mL Erlenmeyer flasks with a working volume of 50 mL (Part NO: Corning TM 431144, Fisher Scientific, Dublin, Ireland) at an agitation rate of 120 rpm, at 37 ºC and with a 5% CO 2 supply in a shaking incubator (Adolf Kühner AG, CH-4127 Birsfelden (Basel), Switzerland). Cells were passaged at a seeding density at 0.5×10 6 cells/mL twice per week to keep them in exponential growing phase. 3. Experimental design A series of Erlenmeyer flasks (250 mL) each containing 4.0 mM glutamine supplemented CD CHO (49.0 mL) were seeded with cells at an inoculation density of 0.4×10 6 cells/mL. Individual cottonseed hydrolysates were supplemented into the cultures to a final concentration at 5.0 g/L from hydrolysate stock solutions (1.0 mL, 250 mg/mL). Control cultures were supplemented with 1.0 mL of medium without hydrolysate. Anti-clumping agent (Part NO: Gibco TM 0010057DG, Fisher Scientific, Dublin, Ireland) 100 µL was added to each flask. Biological triplicates were prepared for each condition. After incubation for 0.5 h, 0.5 mL cell suspensions were taken to serve as zero time point samples. Subsequently, cell suspension samples (0.5 mL) were taken daily until day 10. Viable cell density and viability were measured by the trypan blue exclusion method (Strober, 2015 ) and cell diameter were performed on a LUNA-II™ automated cell counter (Logos Biosystems, Gyeonggi-do 14055, South Korea). The samples were then centrifuged at 3,000 rpm for 5 min at room temperature with an Eppendorf 5452 Minispin centrifuge (Eppendorf, Hamburg, Germany), and the supernatants were transferred into fresh 1.5 mL tubes and stored at -20 ºC for future analysis. The cultures were grown until day 10 and then centrifuged at 1,500 rpm for 5 min at room temperature with Eppendorf 5804R multipurpose benchtop centrifuge (Eppendorf, Hamburg, Germany). The supernatant which contained bevacizumab antibody was transferred to a 50 mL tube and stored at -20°C until further use. 4. Osmolality measurement The osmolality of the cell culture supernatant (40 µL) harvested at different time points was measured by freezing point depression methods as described previously (Koumantakis and Wyndham, 1989 ) on an OsmoTECH ® PRO multi-sample osmometer (Advanced Instruments Inc) following the manufacturer’s instructions. 5. Media substrates and metabolites profiling To 50 µL of the cell culture supernatant prepared from each time point were added 150 µL of 0.1% formic acid in H 2 O and 200 µL of acetonitrile. After vortexing for 1 min, the mixture was centrifuged at 10,000 rpm for 15 min at room temperature with an Eppendorf 5452 Minispin centrifuge (Eppendorf, Hamburg, Germany). The supernatant (50 µL) was taken without disturbing the pellet and added to 450 µL of 0.1% formic acid in H 2 O and mixed well. The substrates and metabolites in the spent media samples were analysed by reverse phase ultra-high performance liquid chromatography coupled to electrospray ionization tandem mass spectrometry (RP-UHPLC-ESI-MS/MS) using a Vanquish UHPLC coupled to Q Exactive Plus Orbitrap Mass Spectrometry (Thermo Scientific, Waltham, Massachusetts, USA) equipped with InfinityLab Poroshell 120 HPH-C18 column (Pore size: 100 Å, 2.1 mm×150 mm, particle size: 2.7 µm, Part NO: 693775-702(T), Agilent Technologies, Santa Clara, California, USA). Buffer A was 0.1% (v/v) formic acid in water and buffer B was 0.1% (v/v) formic acid in acetonitrile. The samples were maintained at 5.0°C before injection (10.0 µL) into the column held at a temperature of 40°C. After initial system equilibrium for 1.5 min with 100% (v/v) of buffer A, the separation was carried out by a linear gradient of 100 − 90% (v/v) of buffer A within 1.5-6.0 min, followed by a linear gradient of 90 − 65% (v/v) of buffer A for another 4.0 min. From 10.0 to 13.0 min, the gradient of buffer A was reduced further to 5% (v/v) and maintained for another 3.0 min. Subsequently, the gradient of buffer A was increased to 100% (v/v) in 0.1 min. The flow rate was maintained as 0.3 mL/min throughout the chromatographic run. The running conditions for ESI-MS/MS were positive mode, spray voltage of 3.80 kV, capillary temperature of 320°C, aux gas heater temperature of 400°C, sheath and Aux gas flow rate of 40 µL/min and 10 µL/min, and resolution of 70,000. The mass-to-charge ratio (m/z) scan range was 70 − 1,050. All the raw RP-UHPLC-ESI-MS/MS data files were batch processed by open-source software MZmine 3.9.0, and if necessary, the peptides and small molecules were annotated by following the workflow as described (Xie and Butler, 2025 ). The CHO cell line expressed a humanized anti-vascular endothelial growth factor (VEGF) monoclonal antibody (IgG), bevacizumab. Bevacizumab is among the most frequently prescribed therapeutic proteins (Ferrara et al., 2005 ). The antibody (bevacizumab) produced by the cells as well as common media substrates and metabolites including glucose, lactate, glutamine, and ammonia were measured using a Cedex ® Bio Analyzer (Roche Diagnostics Deutschland GmbH, Mannheim, Germany) following the manufacturer’s instructions. 6. Evaluation of cell growth, antibody productivity, and major substrates exchange rates The integral (area under the curve) viable cell density (IVCD) by trapezoidal integration of the viable cell density over the entire cultivation duration was calculated according to Eq. (1) (Derek Adams, 2007 ; Schellenberg et al., 2022 ) by using GraphPad Software (San Diego, CA, USA.). $$\:{IVCD}_{n}=\underset{t=0}{\overset{t}{\int\:}}VCD\left(t\right)dt\approx\:\sum\:_{i=1}^{n}\left(\frac{{{VCD}_{i-1}+VCD}_{i}}{2}\right)\times\:\left({t}_{i}-{t}_{i-1}\right)\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\left(1\right)$$ Additionally, the antibody titer at the end of the batch cultivation (CMab) was divided by the respective IVCD for comparison of the overall antibody producivity (qMab) according to Eq. (2). \(\:qMab=\frac{CMab}{IVCD}\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\left(2\right)\) Where VCD(t) is the viable cell density (×10 6 cells/mL) at cultured time t (day), t 0 is the initial cultured time (day), IVCD is the integral viable cell density (×10 6 cells·days/mL), CMab is the antibody titer at the end of cultivation (mg/L), qMab is the overall cell-specific antibody productivity (pg/cell/day). 7. Antibody purification and N-glycan analysis Antibody purification was performed by affinity chromatography (Ana Cristina Grodzki, 2010 ; Rodriguez et al., 2020 ) using a Protein A HP Spin Trap antibody purification column (Part NO: 28-9031-32, GE healthcare, Dublin, Ireland) following the manufacturer’s instructions. After centrifugation (Fisher brand mini centrifuge, cat NO: HS120336) for 10 seconds to remove the storage buffer, the column was washed three times with 600 µL of loading buffer (25 mM sodium phosphate buffer, pH 7.0). The cell culture supernatant harvested on day 10 (700 µL) was loaded onto the column and incubated for 5 min with gentle shaking. After centrifugation, another 700 µL of cell culture supernatant was loaded onto the column and incubated for 5 min with gentle shaking. After centrifugation, the column was washed three times with 600 µL loading buffer to remove unbound molecules. Elution was performed by 400 µL of 0.1 M citric acid, pH 3.0. The collected fraction containing the antibody was neutralized by adding 30 µL of 1.0 M Tris-HCl, pH 9.0, followed by buffer exchange against 50 mM 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid (HEPES) buffer, pH 8.0, using an Amicon ® Ultra − 0.5 mL centrifugal filter with molecular weight cut off of 10 kDa (Millipore, Bedford, MA, USA). Antibody was quantified by Cedex ® Bio Analyzer (Roche Diagnostics Deutschland GmbH, Mannheim, Germany) following the manufacturer’s instructions and diluted to a final concentration of 2.0 mg/mL. The InstantPC-labelled N-glycans were prepared by using the AdvanceBio Gly-X N-Glycan Prep with InstantPC kit, 96-ct (Part NO: GX96-IPC, Agilent Technologies, Santa Clara, CA 95051, USA) and profiled by following the protocols described previously without any revision (Xie and Butler, 2022 ; Xie et al., 2021 ). Samples were then stored at -20°C before further use. 8. Chemometric analysis A chemometric method (Biancolillo and Marini, 2018 ) was employed for systematic statistical analysis and differentiation of the chemically defined culture media supplemented with different cottonseed hydrolysates, as well as to identify key components that were associated with cell growth and antibody production under different conditions. Significant differences were determined by Two-way analysis of variance (ANOVA) (Pandis, 2015 , 2016 ) with GraphPad Software (San Diego, CA, USA.). To identify potential correlations between compositional variability of culture media supplemented with different cottonseed hydrolysates and cell growth performance, partial least squares discriminant analysis (PLS-DA) and PatternHunter feature correlation were carried out on a web-based platform MetaboAnalyst ( https://www.metaboanalyst.ca/ ) (Pang et al., 2024 ; Pang et al., 2022 ; Xia et al., 2009 ). Results 1. Compositional variability of culture media supplemented with cottonseed hydrolysates Reverse-phase liquid chromatography–high resolution mass spectrometry (RPLC-HRMS) was used to characterise compositional differences between control culture media and media supplemented with different batches of cottonseed hydrolysates. Characteristic features were detected across an m/z range of 70–1,050 over a 20 min chromatographic run, capturing a broad range of metabolites including amino acids, peptides, carbohydrates, and lipids. Each feature in the analysis was identified by a mass/charge (m/z) ratio at a specific peak retention time and listed as m/z@ retention time (min). The full dataset of detected features is provided in Supplementary Information (ST1). A representative volcano plot comparing hydrolysate-supplemented media (H2B) with control media (Fig. 1 A) showed extensive upregulated features (fold change ≥ 2.0, p ≤ 0.1), corresponding to components introduced by the hydrolysate. A smaller number of downregulated features were also observed, likely reflecting dilution effects during hydrolysate stock preparation. Direct comparison between two hydrolysate batches (H3C vs H1B; Fig. 1 B) revealed substantial compositional differences, with distinct sets of features enriched in each condition. These results confirm pronounced batch-dependent compositional variability in cottonseed hydrolysates. 2. Chemometric differentiation of hydrolysate-supplemented media Sparse partial least squares discriminant analysis (SPLS-DA) was applied to the RPLC-HRMS feature set to assess compositional differences across conditions. At day 0, samples clustered according to hydrolysate type (Fig. 2 A), with clear separation between control media and hydrolysate-supplemented media. The three batches of each hydrolysate type (H1, H2, H3) formed distinct clusters, indicating reproducible within-group similarity and between-group variability. The top 10 signature features that discriminate between the hydrolysate batches and contribute to the first and second components are shown in Figs. 2 B and 2 C, respectively. The features are defined by their mass-to-charge ratio and retention time (m/z@RT). For example, [email protected] and [email protected] were among the most influential variables differentiating hydrolysate groups. Representative extracted ion chromatograms illustrating variation in feature abundance across conditions are shown in Figs. 2 D and 2 E. These results demonstrate that LC-MS-derived feature profiles combined with chemometric analysis can effectively distinguish between hydrolysate batches and define compositional signatures associated with each group. 3. Cottonseed hydrolysates prolonged cell viability and enhanced antibody productivity CHO DG44 cells were cultured in the presence or absence of cottonseed hydrolysates (5.0 g/L) to evaluate their effects on growth and productivity. Control cultures exhibited a typical batch growth profile, with exponential growth (day 0–4), a stationary phase (day 4–7), and a decline phase thereafter (Fig. 3 A). In contrast, hydrolysate-supplemented cultures reached lower peak viable cell densities but maintained high viability (> 90%) throughout the 10-day culture period (Fig. 3 B). Maximum viable cell densities in hydrolysate-supplemented cultures ranged from 2.9 to 4.7 × 10^6 cells/mL, compared to approximately 6 × 10^6 cells/mL in the control. Correspondingly, IVCD values were generally lower in supplemented cultures (19.62–25.63 × 10^6 cells·days/mL) than in the control (29.93 ± 0.98), with the exception of H2A, which showed comparable IVCD (Table 1 ). Despite reduced cell densities, hydrolysate supplementation markedly prolonged culture viability beyond day 7. Table 1 The effect of cottonseed hydrolysate supplements on integral viable cell density, maximum viable cell density, maximum antibody concentration, and specific productivity over the 10-day batch cultivation Hydrolysate supplement IVCD (×10 6 cells·days/mL) VCD max (×10 6 cells/mL) CMab at day 10 (mg/L) qMab (pg/cell/day) Control 29.93 ± 0.98 5.89 ± 0.23 287.00 ± 4.36 9.60 ± 0.44 H1A 20.85 ± 0.75 3.56 ± 0.27 403.00 ± 10.44 19.34 ± 0.41 H1B 20.20 ± 0.83 3.19 ± 0.33 351.67 ± 12.66 17.43 ± 1.17 H1C 20.11 ± 1.08 2.95 ± 0.20 382.00 ± 9.54 19.02 ± 0.78 H2A 28.53 ± 0.80 4.66 ± 0.21 575.67 ± 2.52 20.19 ± 0.54 H2B 25.63 ± 0.73 3.92 ± 0.37 599.00 ± 4.00 23.38 ± 0.51 H2C 23.86 ± 0.12 3.92 ± 0.34 517.67 ± 10.12 21.70 ± 0.33 H3A 25.23 ± 0.97 3.95 ± 0.30 585.00 ± 12.29 23.20 ± 0.45 H3B 19.62 ± 0.54 2.89 ± 0.16 510.67 ± 11.72 26.05 ± 1.29 H3C 21.83 ± 1.13 3.23 ± 0.02 534.67 ± 13.32 24.52 ± 0.99 All the parameters are expressed as mean ± standard deviation of sample (SD). Independent biological triplicate cultures were carried out (n = 3). Cell diameter was consistently higher in hydrolysate-supplemented cultures compared to control (Fig. 3 C). The cell diameter in the control culture decreased notably from day 2 to day 8, to a value of 12.3 µm which was substantially lower than the mean of the hydrolysate-supplemented cultures at 14.5 µm. This difference may be related to the culture osmolality which was substantially higher in the hydrolysate-supplemented cultures (Fig. 3 D), suggesting a link between hydrolysate supplementation, osmotic conditions, and cell size. Antibody production profiles differed substantially between conditions (Fig. 3 E). In control cultures, the antibody titre reached approximately 270 mg/L by day 7 and remained unchanged as viability declined. In contrast, hydrolysate-supplemented cultures continued producing antibody until day 10, resulting in significantly higher final titres. The highest-producing conditions (H2A, H2B, H3A) reached 576–599 mg/L, more than double the control. Cell-specific productivity (qMab) was also significantly increased in supplemented cultures (17.43–26.05 pg/cell/day) compared to control (9.60 ± 0.44 pg/cell/day) (Table 1 ). 4. Impact of hydrolysates on key metabolic substrates and by-products The concentrations of glucose, glutamine, lactate, and ammonia were monitored throughout the culture period (Figs. 3 F–I). Glucose was rapidly consumed in all conditions; however, hydrolysate-supplemented cultures continued to consume glucose beyond day 7, whereas consumption in control cultures plateaued after this point. Lactate accumulation differed markedly between conditions. In control cultures, lactate increased to approximately 28 mM by day 7 and remained elevated. In contrast, hydrolysate-supplemented cultures reached peak lactate concentrations of 13–19 mM by day 4–5, followed by a decline, indicating a shift from lactate production to lactate consumption. By day 10, lactate levels in some supplemented cultures (e.g., H2A) were as low as 6 mM. Glutamine was rapidly consumed during the first 4 days in all cultures, after which residual concentrations (~ 1 mM) remained relatively stable. Ammonia accumulation was consistently higher in control cultures, reaching ~ 9 mM by day 10, compared to 6–7 mM in hydrolysate-supplemented cultures. These results indicate that hydrolysate supplementation is associated with altered metabolic profiles, including reduced accumulation of inhibitory by-products and a shift in lactate metabolism during later culture stages. 5. Effects of hydrolysates on N-glycosylation of bevacizumab N-glycosylation profiles of bevacizumab were analysed by HILIC-FLD. All samples exhibited similar glycan structures, with differences observed primarily in relative abundance (Fig. 4 ). Under control conditions, the dominant glycan species was FA2 (76.2 ± 0.2%), with other glycans (A1, FA1, A2, M5, FA2G1, FA2G2) present at lower levels. No sialylated glycans were detected. Hydrolysate supplementation significantly altered glycan distributions. In particular, increased levels of mono- and di-galactosylated species (FA2G1’, FA2G1, FA2G2) and A2 were observed, accompanied by a decrease in FA1. The abundance of high-mannose glycan M5 varied between hydrolysate groups, with lower levels in H1 and higher levels in H3 relative to control. Statistical analysis confirmed significant differences across conditions (Supplementary Information ST4). These results indicate that cottonseed hydrolysates influence antibody glycosylation, particularly by increasing galactosylation. 6. Correlation of compositional features with cell growth and antibody production To explore relationships between media composition and culture performance, time-resolved LC-MS data from H2B-supplemented cultures were analysed using SPLS-DA and PatternHunter correlation analysis. SPLS-DA revealed clear temporal separation of samples with minimal variation between biological replicates (Fig. 5 A). The top signature features that discriminated between samples included m/z [email protected] , (bevacizumab), and [email protected] (Fig. 5 B). A volcano plot comparing day 10 and day 0 samples (Fig. 5 C) showed extensive changes in feature abundance, reflecting net production and consumption of metabolites during culture. PatternHunter analysis identified 25 features significantly correlated with cell growth and antibody production. Cell growth was positively correlated with features such as [email protected] and [email protected] (annotated as glutathione oxidized and 2'-deoxycytidine, respectively), and negatively correlated with features including [email protected] (annotated as pyridoxine). Antibody production showed positive correlations with features such as [email protected] and [email protected] , and negative correlations with several low-mass metabolites (Fig. 6 ). Distinct correlation patterns were also observed in control cultures, reflecting differences in media composition between supplemented and unsupplemented conditions. These findings demonstrate that LC-MS-derived compositional features can be linked to functional outputs and used to identify candidate markers associated with culture performance. Discussion Protein hydrolysates are widely used as cost-effective supplements in CHO cell culture to enhance productivity. However, their compositional complexity and batch-to-batch variability remain major challenges for understanding and controlling their functional effects (Obaidi et al., 2021 ). In this study, we combined LC-HRMS-based compositional profiling with chemometric analysis to systematically relate the variability of cottonseed hydrolysates to CHO DG44 cell culture performance, including growth characteristics, metabolism, antibody productivity, and glycosylation. We have extended the development of this approach that was originally applied to the compositional profiling analysis of variants of soy hydrolysates (Xie and Butler, 2025 ). The LC-HRMS workflow enabled clear differentiation of culture media supplemented with different hydrolysate batches based on distinct compositional signatures. Unsupervised and supervised chemometric analyses demonstrated that batches clustered according to hydrolysate type, indicating reproducible yet distinct compositional profiles. These findings confirm that even hydrolysates derived from similar sources can exhibit substantial compositional variability, which is likely to contribute to differences in their biological effects. Functionally, cottonseed hydrolysate supplementation did not increase peak viable cell density and, in most cases, resulted in lower maximum cell densities and integral viable cell density (IVCD) compared to the control. However, a key observation was the prolonged maintenance of high cell viability in hydrolysate-supplemented cultures, particularly during the late stages of the batch process. This extended culture longevity, combined with a marked increase in cell-specific productivity (qMab), resulted in substantially higher final antibody titres. These results indicate that the primary benefit of hydrolysate supplementation lies not in promoting rapid cell proliferation, but in enhancing culture longevity and productivity during the stationary and decline phases. Hydrolysate supplementation was also associated with an increase in cell diameter and culture osmolality. This was in good agreement with previous research, in which cell size changed significantly over the course of cell culture process (Seewoster and Lehmann, 1997 ; Pan et al, 2017 ). Our observed increase in cell size is consistent with previous reports linking hyperosmolar conditions to enlarged CHO cells (Alhuthali et al., 2021 ; Romanova et al., 2021 ). As cell volume increases nonlinearly with diameter, this effect may partially explain the reduced viable cell densities observed in hydrolysate-supplemented cultures, despite comparable overall cellular biomass. These findings highlight a limitation of relying solely on cell number as a measure of culture performance and suggest that cell size should be considered when interpreting growth and productivity metrics. A major metabolic feature of hydrolysate-supplemented cultures was the shift from lactate production to lactate consumption after the exponential growth phase. In contrast, control cultures exhibited continued lactate accumulation, reaching significantly higher concentrations. This metabolic transition has been widely associated with improved culture performance and higher recombinant protein titres. The reduced accumulation of lactate and ammonia in hydrolysate-supplemented cultures likely contributed to the prolonged cell viability observed, as both metabolites are known to exert inhibitory effects on cell growth and productivity (Cruz et al., 2000 ; Kanehisa and Goto, 2000 ; Pereira et al., 2018 ). While the precise molecular mechanisms underlying this metabolic shift remain unclear, it is consistent with previous reports suggesting a transition from high glycolytic flux to more efficient metabolic states associated with improved bioprocess performance (Luo et al., 2012 ). The glycan profile of the antibody produced under control conditions without hydrolysate supplementation was in good agreement with previous studies (Carillo et al., 2020 ; Planinc et al., 2017 ; Seo et al., 2018 ). We observed an increase in galactosylated glycan species following hydrolysate supplementation which was consistent with previous studies reporting enhanced galactosylation in the presence of plant-derived hydrolysates (Obaidi et al., 2021 ). Although the underlying mechanisms were not directly investigated in this study, these changes may reflect altered intracellular metabolism or enzyme activity within the Golgi apparatus (McDonald et al., 2014 ). Given the importance of glycosylation for therapeutic antibody efficacy and stability, these findings highlight an additional dimension through which hydrolysates can influence bioprocess outcomes. By integrating compositional profiling with chemometric analysis, we identified 25 signature features that were positively or negatively associated with cell growth and antibody production. These features, defined by their m/z and retention time, represent potential markers linking hydrolysate composition to functional performance. Tentative annotation of selected features suggested the involvement of metabolites such as glutathione, deoxycytidine, and pyridoxine, although most features remain to be structurally confirmed. Importantly, these results demonstrate the potential of data-driven approaches to identify candidate components within complex hydrolysate mixtures that may contribute to enhanced culture performance. Despite these insights, several limitations should be acknowledged. First, the study was conducted using a single CHO DG44 cell line and a specific chemically defined medium; therefore, the observed effects may not be directly transferable to other cell lines, media formulations, or hydrolysate types. Second, the LC-MS analysis was performed in positive ion mode over a defined m/z range, which may have limited the detection of certain classes of metabolites. Third, the identification of signature features was based primarily on correlation analysis, and causal relationships between specific components and biological effects remain to be established. Future studies integrating targeted metabolomics, functional validation, and multi-omics approaches will be required to confirm the roles of these candidate components. In summary, this study demonstrates that LC-HRMS-based compositional profiling combined with chemometric analysis provides a robust framework for linking hydrolysate variability to CHO cell culture performance. Cottonseed hydrolysates enhanced culture longevity, reduced the accumulation of inhibitory metabolites, increased cell-specific productivity, and altered antibody glycosylation, despite not increasing peak cell density. These findings suggest that hydrolysate supplementation primarily modulates cellular metabolism and productivity rather than simply promoting cell growth. The approach described here offers a pathway toward the systematic identification of functional components in complex media supplements and may support the development of more consistent and effective hydrolysate-based strategies for biopharmaceutical manufacturing. Declarations Author contributions Yongjing Xie: Conceptualization, methodology, investigation, writing-review and editing; Michael Butler: Validation, writing-review and editing. Acknowledgements The protein hydrolysates used in the current study were kindly provided by Kerry lnc. (Beloit, WI 53511, USA). We thank Hans Huttinga, Angel Varelarohena, Brandon Wrage, Michael woods, Derek Carr of Kerry lnc. for their review and helpful comments in the preparation of this manuscript. We thank Caitriona Walsh from NIBRT Core Facility for RP-UHPLC-HR-ESI-MS/MS spectra acquisition. Supplementary Information The data underlying this article are available in the article and in its online Supplementary Information. Funding This work was financially supported by Enterprise Ireland Innovation Partnership Programme (IPP) Award (IP20211007). Conflict of interest statement The authors declare that there is no conflict of interest. Data availability statement The original or derived data underlying this article is available from the corresponding author upon reasonable request. References Alhuthali, S., Kotidis, P., Kontoravdi, C., (2021) Osmolality Effects on CHO Cell Growth, Cell Volume, Antibody Productivity and Glycosylation. Int J Mol Sci 22. Ana Cristina Grodzki, E.B., (2010) Antibody Purification: Affinity Chromatography – Protein A and Protein G Sepharose. In: Constance Oliver, M.C.J. (Ed.), Immunocytochemical Methods and Protocols. Humana Press, New Jersey, USA, pp. 33-41. André Siemensma, J.B., Chris Wilcox, Hans Huttinga, (2010) Towards an Understanding of How Protein Hydrolysates Stimulate More Efficient Biosynthesis in Cultured Cells. In: Vijai K. Pasupuleti, A.L.D. (Ed.), Protein Hydrolysates in Biotechnology. Springer, Dordrecht, pp. 33-54. B. Farges-Haddani, B.T., S. Chenu, I. Chevalot, C. Harscoat, I. Marc, J.L. Goergen, A. Marc, (2006) Peptide fractions of rapeseed hydrolysates as an alternative to animal proteins in CHO cell culture media. Process Biochemistry 41, 2297-2304. Biancolillo, A., Marini, F., (2018) Chemometric Methods for Spectroscopy-Based Pharmaceutical Analysis. Front Chem 6, 576. Carillo, S., Perez-Robles, R., Jakes, C., Ribeiro da Silva, M., Millan Martin, S., Farrell, A., Navas, N., Bones, J., (2020) Comparing different domains of analysis for the characterisation of N-glycans on monoclonal antibodies. J Pharm Anal 10, 23-34. Colla, G., Hoagland, L., Ruzzi, M., Cardarelli, M., Bonini, P., Canaguier, R., Rouphael, Y., (2017) Biostimulant Action of Protein Hydrolysates: Unraveling Their Effects on Plant Physiology and Microbiome. Front Plant Sci 8, 2202. Combe, M., Sokolenko, S., (2021) Quantifying the impact of cell culture media on CHO cell growth and protein production. Biotechnol Adv 50, 107761. Cruz, H.J., Freitas, C.M., Alves, P.M., Moreira, J.L., Carrondo, M.J., (2000) Effects of ammonia and lactate on growth, metabolism, and productivity of BHK cells. Enzyme Microb Technol 27, 43-52. Derek Adams, R.K., and Wei-Shou Hu (2007) Application of Stoichiometric and Kinetic Analyses to Characterize Cell Growth and Product Formation. In: Pörtner, R. (Ed.), Animal Cell Biotechnology Methods and Protocols, Second ed. Humana Press, Totowa, NJ, pp. 269-284. Dhara, V.G., Kumar, S., DeVine, L., Naik, H.M., Cole, R.N., More, A.J., Lau, E.C., Betenbaugh, M.J., (2023) Cottonseed hydrolysate supplementation alters metabolic and proteomics responses in Chinese hamster ovary cell cultures. Biotechnol J 18, e2200243. Du, Q., Zhang, X., Wang, T., Wang, X., (2022) Effects and mechanisms of animal-free hydrolysates on recombination protein yields in CHO cells. Appl Microbiol Biotechnol 106, 7387-7396. Ferrara, N., Hillan, K.J., Novotny, W., (2005) Bevacizumab (Avastin), a humanized anti-VEGF monoclonal antibody for cancer therapy. Biochem Biophys Res Commun 333, 328-335. Franek, F., Hohenwarter, O., Katinger, H., (2000) Plant protein hydrolysates: preparation of defined peptide fractions promoting growth and production in animal cells cultures. Biotechnol Prog 16, 688-692. Ho, Y.Y., Lu, H.K., Lim, Z.F.S., Lim, H.W., Ho, Y.S., Ng, S.K., (2021) Applications and analysis of hydrolysates in animal cell culture. Bioresour Bioprocess 8, 93. Hu, D., Zhao, L., Wang, J., Fan, L., Liu, X., Wang, H., Tan, W.S., (2018) Physiological responses of Chinese hamster ovary cells to a productivity-enhancing yeast extract. J Biosci Bioeng 126, 636-643. Juliet Lobo-Alfonso, P.P., David Jayme, (2008) Benefits and Limitations of Protein Hydrolysates as Components of Serum-Free Media for Animal Cell Culture Applications. In: Vijai K. Pasupuleti, A.L.D. (Ed.), Protein Hydrolysates in Biotechnology. Springer, Dordrecht, pp. 55-78. Kanehisa, M., Goto, S., (2000) KEGG: kyoto encyclopedia of genes and genomes. Nucleic Acids Res 28, 27-30. Kim, J.Y., Kim, Y.G., Lee, G.M., (2012) CHO cells in biotechnology for production of recombinant proteins: current state and further potential. Appl Microbiol Biotechnol 93, 917-930. Koumantakis, G., Wyndham, L.E., (1989) An evaluation of osmolality measurement by freezing point depression using micro-amounts of sample. J Automat Chem 11, 80-83. Kristinsson, H.G., Rasco, B.A., (2000) Fish protein hydrolysates: production, biochemical, and functional properties. Crit Rev Food Sci Nutr 40, 43-81. Kumar, S., Dhara, V.G., Orzolek, L.D., Hao, H., More, A.J., Lau, E.C., Betenbaugh, M.J., (2021) Elucidating the impact of cottonseed hydrolysates on CHO cell culture performance through transcriptomic analysis. Appl Microbiol Biotechnol 105, 271-285. Lee, L.C., Liong, C.Y., Jemain, A.A., (2018) Partial least squares-discriminant analysis (PLS-DA) for classification of high-dimensional (HD) data: a review of contemporary practice strategies and knowledge gaps. Analyst 143, 3526-3539. Li, M., Ma, B., Kisman, D., Tromp, J., (2004) Patternhunter II: highly sensitive and fast homology search. J Bioinform Comput Biol 2, 417-439. Luo, J., Vijayasankaran, N., Autsen, J., Santuray, R., Hudson, T., Amanullah, A., Li, F., (2012) Comparative metabolite analysis to understand lactate metabolism shift in Chinese hamster ovary cell culture process. Biotechnol Bioeng 109, 146-156. Luo, Y., Pierce, K.M., (2012) Development toward rapid and efficient screening for high performance hydrolysate lots in a recombinant monoclonal antibody manufacturing process. Biotechnol Prog 28, 1061-1068. Ma, B., Tromp, J., Li, M., (2002) PatternHunter: faster and more sensitive homology search. Bioinformatics 18, 440-445. McCarthy, D.J., Smyth, G.K., (2009) Testing significance relative to a fold-change threshold is a TREAT. Bioinformatics 25, 765-771. McDonald, A.G., Hayes, J.M., Bezak, T., Gluchowska, S.A., Cosgrave, E.F., Struwe, W.B., Stroop, C.J., Kok, H., van de Laar, T., Rudd, P.M., Tipton, K.F., Davey, G.P., (2014) Galactosyltransferase 4 is a major control point for glycan branching in N-linked glycosylation. J Cell Sci 127, 5014-5026. Meghna Rajvanshi, K.V.V., (2013) Pathway Modeling, Metabolic. In: Werner Dubitzky, O.W., Kwang-Hyun Cho, Hiroki Yokota (Ed.), Encyclopedia of Systems Biology. Springer, New York, USA, pp. 1652–1655. Ng, J.Y., Chua, M.L., Zhang, C., Hong, S., Kumar, Y., Gokhale, R., Ee, P.L.R., (2020) Chlorella vulgaris Extract as a Serum Replacement That Enhances Mammalian Cell Growth and Protein Expression. Front Bioeng Biotechnol 8, 564667. Obaidi, I., Mota, L.M., Quigley, A., Butler, M., (2021) The role of protein hydrolysates in prolonging viability and enhancing antibody production of CHO cells. Appl Microbiol Biotechnol 105, 3115-3129. Ozturk, S.S., Riley, M.R., Palsson, B.O., (1992) Effects of ammonia and lactate on hybridoma growth, metabolism, and antibody production. Biotechnol Bioeng 39, 418-431. Pan, X., Dalm, C., Wijffels, R.H., Martens, D.E., (2017) Metabolic characterization of a CHO cell size increase phase in fed-batch cultures. Appl Microbiol Biotechnol 101, 8101-8113. Pandis, N., (2015) Two-way analysis of variance: Part 1. Am J Orthod Dentofacial Orthop 148, 1078-1079. Pandis, N., (2016) Two-way analysis of variance: Part 2. Am J Orthod Dentofacial Orthop 149, 137-139. Pang, Z., Lu, Y., Zhou, G., Hui, F., Xu, L., Viau, C., Spigelman, A.F., MacDonald, P.E., Wishart, D.S., Li, S., Xia, J., (2024) MetaboAnalyst 6.0: towards a unified platform for metabolomics data processing, analysis and interpretation. Nucleic Acids Res. Pang, Z., Zhou, G., Ewald, J., Chang, L., Hacariz, O., Basu, N., Xia, J., (2022) Using MetaboAnalyst 5.0 for LC-HRMS spectra processing, multi-omics integration and covariate adjustment of global metabolomics data. Nat Protoc 17, 1735-1761. Patterson, T.A., Lobenhofer, E.K., Fulmer-Smentek, S.B., Collins, P.J., Chu, T.M., Bao, W., Fang, H., Kawasaki, E.S., Hager, J., Tikhonova, I.R., Walker, S.J., Zhang, L., Hurban, P., de Longueville, F., Fuscoe, J.C., Tong, W., Shi, L., Wolfinger, R.D., (2006) Performance comparison of one-color and two-color platforms within the MicroArray Quality Control (MAQC) project. Nat Biotechnol 24, 1140-1150. Pereira, S., Kildegaard, H.F., Andersen, M.R., (2018) Impact of CHO Metabolism on Cell Growth and Protein Production: An Overview of Toxic and Inhibiting Metabolites and Nutrients. Biotechnol J 13, e1700499. Planinc, A., Dejaegher, B., Vander Heyden, Y., Viaene, J., Van Praet, S., Rappez, F., Van Antwerpen, P., Delporte, C., (2017) Batch-to-batch N-glycosylation study of infliximab, trastuzumab and bevacizumab, and stability study of bevacizumab. Eur J Hosp Pharm 24, 286-292. Richard G. Brereton, G.R.L., (2014) Partial least squares discriminant analysis: taking the magic away. Journal of Chemometrics 28, 213-225. Richardson, J., Shah, B., Bondarenko, P.V., Bhebe, P., Zhang, Z., Nicklaus, M., Kombe, M.C., (2015) Metabolomics analysis of soy hydrolysates for the identification of productivity markers of mammalian cells for manufacturing therapeutic proteins. Biotechnol Prog 31, 522-531. Ritacco, F.V., Wu, Y., Khetan, A., (2018) Cell culture media for recombinant protein expression in Chinese hamster ovary (CHO) cells: History, key components, and optimization strategies. Biotechnol Prog 34, 1407-1426. Rodriguez, E.L., Poddar, S., Iftekhar, S., Suh, K., Woolfork, A.G., Ovbude, S., Pekarek, A., Walters, M., Lott, S., Hage, D.S., (2020) Affinity chromatography: A review of trends and developments over the past 50 years. J Chromatogr B Analyt Technol Biomed Life Sci 1157, 122332. Rodriguez, R.R., Alhamad, K., Ghosh, S., Bihler, E., Raj, M.S., (2021) Atypical Mycobacterium Abscessus Infection in Stable Chronic Lymphocytic Leukemia: A Case Report and Review of the Literature. Cureus 13, e20574. Romanova, N., Niemann, T., Greiner, J.F.W., Kaltschmidt, B., Kaltschmidt, C., Noll, T., (2021) Hyperosmolality in CHO culture: Effects on cellular behavior and morphology. Biotechnol Bioeng 118, 2348-2359. Schaafsma, G., (2009) Safety of protein hydrolysates, fractions thereof and bioactive peptides in human nutrition. Eur J Clin Nutr 63, 1161-1168. Schellenberg, J., Nagraik, T., Wohlenberg, O.J., Ruhl, S., Bahnemann, J., Scheper, T., Solle, D., (2022) Stress-induced increase of monoclonal antibody production in CHO cells. Eng Life Sci 22, 427-436. Seewoster, T., Lehmann, J., (1997) Cell size distribution as a parameter for the predetermination of exponential growth during repeated batch cultivation of CHO cells. Biotechnol Bioeng 55, 793-797. Seo, N., Polozova, A., Zhang, M., Yates, Z., Cao, S., Li, H., Kuhns, S., Maher, G., McBride, H.J., Liu, J., (2018) Analytical and functional similarity of Amgen biosimilar ABP 215 to bevacizumab. MAbs 10, 678-691. Spearman, M., Lodewyks, C., Richmond, M., Butler, M., (2014) The bioactivity and fractionation of peptide hydrolysates in cultures of CHO cells. Biotechnol Prog 30, 584-593. Strober, W., (2015) Trypan Blue Exclusion Test of Cell Viability. Curr Protoc Immunol 111, A3 B 1-A3 B 3. Torres, M., Hawke, E., Hoare, R., Scholey, R., Pybus, L.P., Young, A., Hayes, A., Dickson, A.J., (2025) Deciphering molecular drivers of lactate metabolic shift in mammalian cell cultures. Metab Eng 88, 25-39. Wurm, F.M., (2004) Production of recombinant protein therapeutics in cultivated mammalian cells. Nat Biotechnol 22, 1393-1398. Xia, J., Psychogios, N., Young, N., Wishart, D.S., (2009) MetaboAnalyst: a web server for metabolomic data analysis and interpretation. Nucleic Acids Res 37, W652-660. Xie, Y., Butler, M., (2022) Construction of an InstantPC-derivatized glycan glucose unit database: A foundation work for high-throughput and high-sensitivity glycomic analysis. Glycobiology 32, 289-303. Xie, Y., Butler, M., (2025) Compositional profiling of protein hydrolysates by high resolution liquid chromatography-mass spectrometry and chemometric analysis. Food Chem 487, 144756. Xie, Y., Mota, L.M., Bergin, A., O'Flaherty, R., Jones, A., Morgan, B., Butler, M., (2021) High-throughput and high-sensitivity N-Glycan profiling: A platform for biopharmaceutical development and disease biomarker discovery. Anal Biochem 623, 114205. Yang, Y., Li, Z., Li, Q., Ma, K., Lin, Y., Feng, H., Wang, T., (2022) Increase recombinant antibody yields through optimizing vector design and production process in CHO cells. Appl Microbiol Biotechnol 106, 4963-4975. Yao, T., Asayama, Y., (2017) Animal-cell culture media: History, characteristics, and current issues. Reprod Med Biol 16, 99-117. Zhang, Y., Tu, D., Shen, Q., Dai, Z., (2019) Fish Scale Valorization by Hydrothermal Pretreatment Followed by Enzymatic Hydrolysis for Gelatin Hydrolysate Production. Molecules 24. Additional Declarations No competing interests reported. Supplementary Files 20260320Supplementryinformation6thdraft.xlsx Cite Share Download PDF Status: Posted Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-9267576","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":622488022,"identity":"2006f231-bb49-4a86-9150-232c3cfc548d","order_by":0,"name":"Yongjing Xie","email":"","orcid":"","institution":"National Institute for Bioprocessing Research and Training","correspondingAuthor":false,"prefix":"","firstName":"Yongjing","middleName":"","lastName":"Xie","suffix":""},{"id":622488023,"identity":"4e625e1f-7bc7-492b-823a-ca3539162840","order_by":1,"name":"Michael Butler","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA9UlEQVRIiWNgGAWjYDACZgY2IGnBYMDAwHiAgcGGgU0CJGxAUIsEUA0zA1BLGkgLYwNeLQyoWg6D2EAteIBuO/OzBz8qJBjMJfIPHPi453wen3Tv8wcMBTY4tZgdZjM37DkjwWA5I5nh4Ixnt4vZZI4bAh2WhkcLD5sEbxvQYTeSGQ7zHLid2CaRBvLLYbxaJP/+g2r5c+AcTMt/vFqkeRugWhgOHIBpOYDPL2bSMsckeCx7Hhsc7DmQnNgmc4xxRoJBMm4t5w8/k3xTYyNnzp748MGPA3aJ82e3MXz48McOpxYY4EHlJhDUMApGwSgYBaMAHwAAenVQ2dD1sqoAAAAASUVORK5CYII=","orcid":"","institution":"National Institute for Bioprocessing Research and Training","correspondingAuthor":true,"prefix":"","firstName":"Michael","middleName":"","lastName":"Butler","suffix":""}],"badges":[],"createdAt":"2026-03-30 13:25:58","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-9267576/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-9267576/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":107088497,"identity":"3ff4d7a9-716a-45e2-926f-1db00addd69e","added_by":"auto","created_at":"2026-04-16 15:26:08","extension":"jpeg","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":622050,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eVolcano plots showing compositional differences in chemically defined culture media supplemented with cottonseed hydrolysates.\u003c/strong\u003e Volcano plots were generated to identify significant changes in LC-MS feature abundance (area under the curve, AUC) between conditions. (A) Comparison of culture media supplemented with hydrolysate H2B and control media without hydrolysate. (B) Comparison between hydrolysate-supplemented media H3C and H1B. The x-axis represents log2 fold change (log2(FC)), and the y-axis represents −log10(p-value). Thresholds are indicated at p = 0.1 and log2(FC) = ±2.0. Features exceeding both thresholds are considered significantly changed. Upregulated features are shown in red, downregulated features in purple, and non-significant features (p \u0026gt; 0.1 and −2.0 \u0026lt; log2(FC) \u0026lt; 2.0) in grey. Only the top 5 feature labels are shown. Features are labelled as m/z@retention time (min).\u003c/p\u003e","description":"","filename":"floatimage1.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-9267576/v1/6bc13aafaba3dd0a93f766cc.jpeg"},{"id":107088498,"identity":"cef03f6d-fc55-4ddb-b79e-1db344d790f5","added_by":"auto","created_at":"2026-04-16 15:26:08","extension":"jpeg","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":315549,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eChemometric differentiation of culture media supplemented with different batches of cottonseed hydrolysates by unsupervised cluster analysis. \u003c/strong\u003e\u0026nbsp;(A) Sparse partial least squares discriminant analysis (SPLS-DA) score plot showing clustering of cell-free culture media at day 0 with and without hydrolysate supplementation (5.0 g/L). (B) Loading plot for the first component showing the top 10 discriminant features. (C) Loading plot for the second component showing the top 10 discriminant features. (D–E) Representative plots showing LC-MS AUC values for selected features ([email protected] and [email protected]) contributing to the first and second components, respectively. Features are labelled as m/z@retention time (min).\u003c/p\u003e","description":"","filename":"floatimage2.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-9267576/v1/fd9e40ef820a1a679a625ec4.jpeg"},{"id":107088520,"identity":"00e99831-67fb-4b09-8966-52a70751b112","added_by":"auto","created_at":"2026-04-16 15:26:13","extension":"jpeg","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":583676,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eEffects of cottonseed hydrolysates on CHO DG44 cell growth, antibody production, and metabolism.\u003c/strong\u003e Time-course profiles of CHO DG44 cultures grown in the absence (control) and presence of different cottonseed hydrolysate batches (5.0 g/L) over 10 days. (A) Viable cell density (×10⁶ cells/mL), (B) cell viability (%), (C) cell diameter (µm), (D) osmolality (mOsm/kg), (E) bevacizumab concentration (mg/L), (F) glucose (mmol/L), (G) lactate (mmol/L), (H) glutamine (mmol/L), and (I) ammonia (mmol/L). Samples were collected daily. Symbols: control (●), H1A (■), H1B (▲), H1C (▼), H2A (◆), H2B (○), H2C (□), H3A (△), H3B (▽), H3C (◇). Data are presented as mean ± SD (n = 3). Complete datasets are provided in Supplementary Information ST3.\u003c/p\u003e","description":"","filename":"floatimage3.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-9267576/v1/4b1ae00cc5a99ac66fcadc0f.jpeg"},{"id":107088512,"identity":"9e721726-9766-4026-bff1-2d856912b908","added_by":"auto","created_at":"2026-04-16 15:26:11","extension":"jpeg","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":640701,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eEffects of cottonseed hydrolysates on bevacizumab N-glycosylation.\u003c/strong\u003e Relative abundance (AUC%) of major N-glycan species measured by HILIC-FLD. (A) Representative glycan distribution for bevacizumab produced under control conditions (mean of biological triplicates). (B–I) Comparison of individual glycan species: (B) A1, (C) FA1, (D) A2, (E) FA2, (F) M5, (G) FA2G1’, (H) FA2G1, and (I) FA2G2 across control and hydrolysate-supplemented cultures (5.0 g/L). Data are presented as mean ± SEM (n = 3). Statistical significance (two-way ANOVA) is indicated as * p \u0026lt; 0.05, ** p \u0026lt; 0.01, *** p \u0026lt; 0.001; non-significant differences (p \u0026gt; 0.05) are not shown. Glycan symbols: square, N-acetylglucosamine; green circle, mannose; yellow circle, galactose; triangle, fucose. Full statistical analysis is provided in Supplementary Information ST4.\u003c/p\u003e","description":"","filename":"floatimage4.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-9267576/v1/b3fe8c2fb9118c2642e9042e.jpeg"},{"id":107088521,"identity":"5beb230b-2570-4e1f-b8ee-789fd31755bb","added_by":"auto","created_at":"2026-04-16 15:26:14","extension":"jpeg","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":576054,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eTime-dependent compositional changes in culture media supplemented with cottonseed hydrolysate H2B.\u003c/strong\u003e (A) SPLS-DA score plot showing temporal separation of spent media samples over the 10-day culture. (B) Loading plot of the first component showing the top 10 discriminant features contributing to temporal differentiation. (C) Volcano plot comparing media composition at day 10 and day 0. The x-axis represents log2 fold change (log2(FC)), and the y-axis represents −log10(p-value). Thresholds are set at p = 0.1 and log2(FC) = ±2.0. Significantly increased features (net production) are shown in red, decreased features (net consumption) in purple, and non-significant features in grey. Only the top 10 feature labels are shown. Features are labelled as m/z@retention time (min).\u003c/p\u003e","description":"","filename":"floatimage5.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-9267576/v1/ad1fc5a3045ffbb1a8350b47.jpeg"},{"id":107088499,"identity":"549b6788-618d-4513-9711-6a928b38d5ee","added_by":"auto","created_at":"2026-04-16 15:26:08","extension":"jpeg","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":501727,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eCorrelation of compositional features with cell growth and antibody production.\u003c/strong\u003e PatternHunter analysis identifying LC-MS features correlated with CHO DG44 cell growth and bevacizumab production during 10-day batch culture. (A) Features correlated with cell growth in hydrolysate (H2B)-supplemented cultures. (B) Features correlated with bevacizumab production in H2B-supplemented cultures. (C) Features correlated with cell growth in control cultures. (D) Features correlated with bevacizumab production in control cultures.\u003c/p\u003e","description":"","filename":"floatimage6.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-9267576/v1/fd0f0b9022c65da0884efc16.jpeg"},{"id":108005910,"identity":"43b05a93-6672-4f67-94ec-c80b8035c44e","added_by":"auto","created_at":"2026-04-28 12:50:44","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":2989617,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-9267576/v1/97f6bf16-58eb-453a-b6a0-c415e7e7782c.pdf"},{"id":107088513,"identity":"c45f973b-ae81-4041-92a9-5be08d2a741c","added_by":"auto","created_at":"2026-04-16 15:26:11","extension":"xlsx","order_by":0,"title":"","display":"","copyAsset":false,"role":"supplement","size":16765403,"visible":true,"origin":"","legend":"","description":"","filename":"20260320Supplementryinformation6thdraft.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-9267576/v1/af619f8f77d5706454fdf1d6.xlsx"}],"financialInterests":"No competing interests reported.","formattedTitle":"Chemometrics and integrative LC-MS identify compositional variability in cottonseed hydrolysates and its effects on CHO cell growth and antibody productivity","fulltext":[{"header":"Introduction","content":"\u003cp\u003eChinese hamster ovary (CHO) cells are the predominant mammalian host used for the large-scale commercial production of recombinant therapeutic proteins, vaccines, and monoclonal antibodies. Their widespread use is largely due to their ability to grow to high cell densities in serum-free suspension culture while supporting correct protein folding, assembly, and post-translational modification (Kim et al., \u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e2012\u003c/span\u003e; Wurm, \u003cspan citationid=\"CR55\" class=\"CitationRef\"\u003e2004\u003c/span\u003e). In the biopharmaceutical industry, improving CHO cell cultivation and increasing productivity while maintaining consistent product quality remain major priorities for reducing manufacturing costs. Considerable advances have been made in bioprocessing through vector design, cell engineering and stable cell line development (Kim et al., \u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e2012\u003c/span\u003e; Yang et al., \u003cspan citationid=\"CR60\" class=\"CitationRef\"\u003e2022\u003c/span\u003e). However, culture medium development remains one of the most effective strategies for increasing cell density and product titre, and improving product quality attributes (Combe and Sokolenko, \u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e2021\u003c/span\u003e; Ritacco et al., \u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e2018\u003c/span\u003e; Yao and Asayama, \u003cspan citationid=\"CR61\" class=\"CitationRef\"\u003e2017\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eProtein hydrolysates derived from a range of raw materials have attracted increasing interest as cost-effective supplements for CHO cell culture. They are widely used as serum substitutes and have been reported to enhance cell growth, improve product titre, and influence product quality (Andr\u0026eacute; Siemensma, 2010; Franek et al., \u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e2000\u003c/span\u003e; Rodriguez et al., \u003cspan citationid=\"CR46\" class=\"CitationRef\"\u003e2021\u003c/span\u003e). For example, Farges-Haddani et al. showed that rapeseed-derived hydrolysate fractions stimulated CHO cell growth significantly, enabling the formulation of a protein-free medium with reduced cell death, increased recombinant interferon gamma (IFN-γ) production, and faster adaptation to serum-free conditions (B. Farges-Haddani, 2006). However, the effects of hydrolysates are not uniform across materials or processes. A soy hydrolysate was reported to support high cell growth but not human beta interferon (β-IFN) productivity, whereas a yeast hydrolysate supported lower growth but β-IFN productivity comparable to that obtained with the animal-derived hydrolysate Primatone RL (Spearman et al., \u003cspan citationid=\"CR52\" class=\"CitationRef\"\u003e2014\u003c/span\u003e). Likewise, two cotton-derived hydrolysates were shown to substantially prolong CHO cell growth and enhance monoclonal antibody production (Obaidi et al., \u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e2021\u003c/span\u003e). Together, these studies demonstrate that protein hydrolysates can provide substantial benefits, but their functional effects are variable.\u003c/p\u003e \u003cp\u003eThe mechanisms by which hydrolysates influence CHO cell culture performance remain poorly understood. One likely explanation is that they enrich the medium with free amino acids and short peptides that serve as additional nutrient sources (Juliet Lobo-Alfonso, \u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e2008\u003c/span\u003e; Ng et al., \u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e2020\u003c/span\u003e). In addition, short peptides may exert more specific bioactivities, including anti-apoptotic, antioxidant, immunomodulatory, and antimicrobial effects, which may support cell survival and growth (Ho et al., \u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e2021\u003c/span\u003e; Juliet Lobo-Alfonso, \u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e2008\u003c/span\u003e). Previous studies have also suggested that hydrolysates may alter the expression of metabolic regulators, reduce oxidative stress, influence cell cycle progression, and affect cell morphology and energy metabolism (Du et al., \u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e2022\u003c/span\u003e; Hu et al., \u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e2018\u003c/span\u003e). More specifically, in relation to cottonseed hydrolysates, Kumar et al. performed transcriptomic analyses and identified differentially expressed genes, associated with pathways known to prolong cell growth and increase antibody titre in supplemented CHO cultures (Kumar et al., \u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e2021\u003c/span\u003e). Dhara et al. further combined extracellular metabolomics with tandem mass tag (TMT) proteomics to show that cottonseed hydrolysate supplementation altered cellular functions critical to growth and protein productivity, including metabolism, protein processing, and apoptosis (Dhara et al., \u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e2023\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eA major challenge in the use of protein hydrolysates is their intrinsic compositional variability, which can lead to inconsistent effects on bioprocess performance (Xie and Butler, \u003cspan citationid=\"CR58\" class=\"CitationRef\"\u003e2025\u003c/span\u003e). Hydrolysates are complex mixtures that may contain peptides, amino acids, carbohydrates, minerals, lipids, vitamins, and inorganic salts (Franek et al., \u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e2000\u003c/span\u003e; Rodriguez et al., \u003cspan citationid=\"CR46\" class=\"CitationRef\"\u003e2021\u003c/span\u003e; Schaafsma, \u003cspan citationid=\"CR48\" class=\"CitationRef\"\u003e2009\u003c/span\u003e). Their composition depends not only on the source material but also on the method and extent of hydrolysis, which may involve acid, alkaline, thermal, or enzymatic processing (Colla et al., \u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e2017\u003c/span\u003e; Kristinsson and Rasco, \u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e2000\u003c/span\u003e; Zhang et al., \u003cspan citationid=\"CR62\" class=\"CitationRef\"\u003e2019\u003c/span\u003e). As a result, different hydrolysate batches may elicit substantially different cellular responses, including changes in transcriptomic, metabolomic, and proteomic profiles. This variability presents a significant challenge for industrial bioprocessing, where media formulations are ideally low-cost, animal-component-free, chemically defined, and capable of supporting robust cell growth, high productivity, and consistent product quality. Therefore, reducing or at least understanding batch-to-batch compositional variability in protein hydrolysates is of considerable practical importance for ensuring process consistency and manufacturing reliability. In a previous study, Luo et al. addressed this issue by combining high-resolution nuclear magnetic resonance (NMR) spectroscopy with partial least squares (PLS) analysis to screen intact soy hydrolysate lots for recombinant monoclonal antibody manufacturing, although glycosylation effects were not examined (Luo and Pierce, \u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e2012\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eIn the present study, rather than repeating earlier transcriptomic or proteomic approaches, we applied reverse-phase ultra-high-performance liquid chromatography high-resolution electrospray ionization tandem mass spectrometry (RP-UHPLC-HR-ESI-MS/MS) to profile the composition of chemically defined culture media supplemented with different batches of cottonseed hydrolysates produced using different raw materials and/or processing conditions. Dihydrofolate reductase-deficient CHO DG44 cells producing bevacizumab were used as a model system to investigate changes in cell growth, antibody productivity, cellular metabolism, and N-glycan profiles. Chemometric analysis, including partial least squares discriminant analysis (PLS-DA) (Lee et al., \u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e2018\u003c/span\u003e; Richard G. Brereton, \u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e2014\u003c/span\u003e) and PatternHunter analysis (Li et al., \u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e2004\u003c/span\u003e; Ma et al., \u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e2002\u003c/span\u003e), was used to identify signature components associated with cell growth and bevacizumab production. The main objective of this work was to evaluate the effects of batch-dependent compositional variability in cottonseed hydrolysates on CHO DG44 culture performance and to correlate these effects with LC-MS-derived compositional signatures. This study provides complementary insight into the relationship between hydrolysate composition and functional performance and may support the future identification of key components that enhance CHO cell growth and productivity in biopharmaceutical manufacturing.\u003c/p\u003e"},{"header":"Materials and Methods","content":"\n\u003ch3\u003e1. Materials and reagents\u003c/h3\u003e\n\u003cp\u003eThree batches of each of three cottonseed protein hydrolysates (CPHs) Ultrapep\u0026trade; cotton (Part NO: S-2803719, labelled as H1A, H1B, H1C), Hypep\u0026trade; 7504 25LB (Part NO: S-1729241, labelled as H2A, H2B, H2C) and Hypep\u0026trade; 7504 25LBS (Part NO: S-2556075, labelled as H3A, H3B, H3C) were kindly donated by Kerry lnc. (Beloit, WI 53511, USA). Each type of hydrolysate (H1, H2, H3) were derived from different batches of raw material and/or different processing protocols. Stock solutions for each cottonseed protein hydrolysate (250 mg/mL) were prepared in Gibco\u0026trade; CD CHO medium (Part NO: Gibco \u003csup\u003eTM\u003c/sup\u003e 10743029, Fisher Scientific, Dublin, Ireland) supplemented with 4.0 mM Gibco\u0026trade; L-glutamine (Part NO: Gibco \u003csup\u003eTM\u003c/sup\u003e 25030024, Fisher Scientific, Dublin, Ireland) and filtered through 0.2 \u0026micro;m sterile syringe filters (Part NO: 15206869, Fisher Scientific, Dublin, Ireland). The stock solutions were stored at 4\u0026deg;C until use.\u003c/p\u003e\n\u003ch3\u003e2. CHO cell line and cell maintenance\u003c/h3\u003e\n\u003cp\u003eThe stably transfected dihydrofolate reductase (DHFR-) deficient CHO cell line (CHO DG44) expressing a monoclonal antibody was kindly donated by Kerry lnc. (Beloit, WI 53511, USA). The cells were routinely grown in suspension in CD CHO medium supplemented with 4.0 mM L-glutamine (Gibco/ Fisher Scientific, Dublin, Ireland). Cells were cultured in 250 mL Erlenmeyer flasks with a working volume of 50 mL (Part NO: Corning \u003csup\u003eTM\u003c/sup\u003e 431144, Fisher Scientific, Dublin, Ireland) at an agitation rate of 120 rpm, at 37 \u0026ordm;C and with a 5% CO\u003csub\u003e2\u003c/sub\u003e supply in a shaking incubator (Adolf K\u0026uuml;hner AG, CH-4127 Birsfelden (Basel), Switzerland). Cells were passaged at a seeding density at 0.5\u0026times;10\u003csup\u003e6\u003c/sup\u003e cells/mL twice per week to keep them in exponential growing phase.\u003c/p\u003e\n\u003ch3\u003e3. Experimental design\u003c/h3\u003e\n\u003cp\u003eA series of Erlenmeyer flasks (250 mL) each containing 4.0 mM glutamine supplemented CD CHO (49.0 mL) were seeded with cells at an inoculation density of 0.4\u0026times;10\u003csup\u003e6\u003c/sup\u003e cells/mL. Individual cottonseed hydrolysates were supplemented into the cultures to a final concentration at 5.0 g/L from hydrolysate stock solutions (1.0 mL, 250 mg/mL). Control cultures were supplemented with 1.0 mL of medium without hydrolysate. Anti-clumping agent (Part NO: Gibco \u003csup\u003eTM\u003c/sup\u003e 0010057DG, Fisher Scientific, Dublin, Ireland) 100 \u0026micro;L was added to each flask. Biological triplicates were prepared for each condition. After incubation for 0.5 h, 0.5 mL cell suspensions were taken to serve as zero time point samples. Subsequently, cell suspension samples (0.5 mL) were taken daily until day 10. Viable cell density and viability were measured by the trypan blue exclusion method (Strober, \u003cspan citationid=\"CR53\" class=\"CitationRef\"\u003e2015\u003c/span\u003e) and cell diameter were performed on a LUNA-II\u0026trade; automated cell counter (Logos Biosystems, Gyeonggi-do 14055, South Korea). The samples were then centrifuged at 3,000 rpm for 5 min at room temperature with an Eppendorf 5452 Minispin centrifuge (Eppendorf, Hamburg, Germany), and the supernatants were transferred into fresh 1.5 mL tubes and stored at -20 \u0026ordm;C for future analysis. The cultures were grown until day 10 and then centrifuged at 1,500 rpm for 5 min at room temperature with Eppendorf 5804R multipurpose benchtop centrifuge (Eppendorf, Hamburg, Germany). The supernatant which contained bevacizumab antibody was transferred to a 50 mL tube and stored at -20\u0026deg;C until further use.\u003c/p\u003e\n\u003ch3\u003e4. Osmolality measurement\u003c/h3\u003e\n\u003cp\u003eThe osmolality of the cell culture supernatant (40 \u0026micro;L) harvested at different time points was measured by freezing point depression methods as described previously (Koumantakis and Wyndham, \u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e1989\u003c/span\u003e) on an OsmoTECH \u0026reg; PRO multi-sample osmometer (Advanced Instruments Inc) following the manufacturer\u0026rsquo;s instructions.\u003c/p\u003e\n\u003ch3\u003e5. Media substrates and metabolites profiling\u003c/h3\u003e\n\u003cp\u003eTo 50 \u0026micro;L of the cell culture supernatant prepared from each time point were added 150 \u0026micro;L of 0.1% formic acid in H\u003csub\u003e2\u003c/sub\u003eO and 200 \u0026micro;L of acetonitrile. After vortexing for 1 min, the mixture was centrifuged at 10,000 rpm for 15 min at room temperature with an Eppendorf 5452 Minispin centrifuge (Eppendorf, Hamburg, Germany). The supernatant (50 \u0026micro;L) was taken without disturbing the pellet and added to 450 \u0026micro;L of 0.1% formic acid in H\u003csub\u003e2\u003c/sub\u003eO and mixed well.\u003c/p\u003e \u003cp\u003eThe substrates and metabolites in the spent media samples were analysed by reverse phase ultra-high performance liquid chromatography coupled to electrospray ionization tandem mass spectrometry (RP-UHPLC-ESI-MS/MS) using a Vanquish UHPLC coupled to Q Exactive Plus Orbitrap Mass Spectrometry (Thermo Scientific, Waltham, Massachusetts, USA) equipped with InfinityLab Poroshell 120 HPH-C18 column (Pore size: 100 \u0026Aring;, 2.1 mm\u0026times;150 mm, particle size: 2.7 \u0026micro;m, Part NO: 693775-702(T), Agilent Technologies, Santa Clara, California, USA). Buffer A was 0.1% (v/v) formic acid in water and buffer B was 0.1% (v/v) formic acid in acetonitrile. The samples were maintained at 5.0\u0026deg;C before injection (10.0 \u0026micro;L) into the column held at a temperature of 40\u0026deg;C. After initial system equilibrium for 1.5 min with 100% (v/v) of buffer A, the separation was carried out by a linear gradient of 100\u0026thinsp;\u0026minus;\u0026thinsp;90% (v/v) of buffer A within 1.5-6.0 min, followed by a linear gradient of 90\u0026thinsp;\u0026minus;\u0026thinsp;65% (v/v) of buffer A for another 4.0 min. From 10.0 to 13.0 min, the gradient of buffer A was reduced further to 5% (v/v) and maintained for another 3.0 min. Subsequently, the gradient of buffer A was increased to 100% (v/v) in 0.1 min. The flow rate was maintained as 0.3 mL/min throughout the chromatographic run. The running conditions for ESI-MS/MS were positive mode, spray voltage of 3.80 kV, capillary temperature of 320\u0026deg;C, aux gas heater temperature of 400\u0026deg;C, sheath and Aux gas flow rate of 40 \u0026micro;L/min and 10 \u0026micro;L/min, and resolution of 70,000. The mass-to-charge ratio (m/z) scan range was 70\u0026thinsp;\u0026minus;\u0026thinsp;1,050. All the raw RP-UHPLC-ESI-MS/MS data files were batch processed by open-source software MZmine 3.9.0, and if necessary, the peptides and small molecules were annotated by following the workflow as described (Xie and Butler, \u003cspan citationid=\"CR58\" class=\"CitationRef\"\u003e2025\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eThe CHO cell line expressed a humanized anti-vascular endothelial growth factor (VEGF) monoclonal antibody (IgG), bevacizumab. Bevacizumab is among the most frequently prescribed therapeutic proteins (Ferrara et al., \u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e2005\u003c/span\u003e). The antibody (bevacizumab) produced by the cells as well as common media substrates and metabolites including glucose, lactate, glutamine, and ammonia were measured using a Cedex \u0026reg; Bio Analyzer (Roche Diagnostics Deutschland GmbH, Mannheim, Germany) following the manufacturer\u0026rsquo;s instructions.\u003c/p\u003e\n\u003ch3\u003e6. Evaluation of cell growth, antibody productivity, and major substrates exchange rates\u003c/h3\u003e\n\u003cp\u003eThe integral (area under the curve) viable cell density (IVCD) by trapezoidal integration of the viable cell density over the entire cultivation duration was calculated according to Eq.\u0026nbsp;(1) (Derek Adams, \u003cspan citationid=\"CR10\"\u003e2007\u003c/span\u003e; Schellenberg et al., \u003cspan citationid=\"CR49\"\u003e2022\u003c/span\u003e) by using GraphPad Software (San Diego, CA, USA.).\u003c/p\u003e\n\u003cdiv id=\"Equa\"\u003e\n \u003cdiv format=\"TEX\" id=\"FileID_Equa\" name=\"EquationSource\"\u003e$$\\:{IVCD}_{n}=\\underset{t=0}{\\overset{t}{\\int\\:}}VCD\\left(t\\right)dt\\approx\\:\\sum\\:_{i=1}^{n}\\left(\\frac{{{VCD}_{i-1}+VCD}_{i}}{2}\\right)\\times\\:\\left({t}_{i}-{t}_{i-1}\\right)\\:\\:\\:\\:\\:\\:\\:\\:\\:\\:\\:\\:\\:\\:\\:\\:\\:\\:\\:\\:\\left(1\\right)$$\u003c/div\u003e\n\u003c/div\u003e\n\u003cp\u003eAdditionally, the antibody titer at the end of the batch cultivation (CMab) was divided by the respective IVCD for comparison of the overall antibody producivity (qMab) according to Eq.\u0026nbsp;(2).\u003c/p\u003e\n\u003cp\u003e\u003cspan\u003e\u0026nbsp;\u003cspan\u003e\\(\\:qMab=\\frac{CMab}{IVCD}\\:\\:\\:\\:\\:\\:\\:\\:\\:\\:\\:\\:\\:\\:\\:\\:\\:\\:\\:\\:\\:\\:\\:\\:\\:\\:\\:\\:\\:\\:\\:\\:\\:\\:\\:\\:\\:\\:\\:\\:\\:\\:\\:\\:\\:\\:\\:\\:\\:\\:\\:\\:\\:\\:\\:\\:\\:\\:\\left(2\\right)\\)\u003c/span\u003e\u0026nbsp;\u003c/span\u003e\u003c/p\u003e\n\u003cp\u003eWhere \u003cem\u003eVCD(t)\u003c/em\u003e is the viable cell density (\u0026times;10\u003csup\u003e6\u003c/sup\u003e cells/mL) at cultured time \u003cem\u003et\u003c/em\u003e (day), t\u003csub\u003e0\u003c/sub\u003e is the initial cultured time (day), IVCD is the integral viable cell density (\u0026times;10\u003csup\u003e6\u003c/sup\u003e cells\u0026middot;days/mL), CMab is the antibody titer at the end of cultivation (mg/L), qMab is the overall cell-specific antibody productivity (pg/cell/day).\u003c/p\u003e\n\u003ch3\u003e7. Antibody purification and N-glycan analysis\u003c/h3\u003e\n\u003cp\u003eAntibody purification was performed by affinity chromatography (Ana Cristina Grodzki, \u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2010\u003c/span\u003e; Rodriguez et al., \u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e2020\u003c/span\u003e) using a Protein A HP Spin Trap antibody purification column (Part NO: 28-9031-32, GE healthcare, Dublin, Ireland) following the manufacturer\u0026rsquo;s instructions. After centrifugation (Fisher brand mini centrifuge, cat NO: HS120336) for 10 seconds to remove the storage buffer, the column was washed three times with 600 \u0026micro;L of loading buffer (25 mM sodium phosphate buffer, pH 7.0). The cell culture supernatant harvested on day 10 (700 \u0026micro;L) was loaded onto the column and incubated for 5 min with gentle shaking. After centrifugation, another 700 \u0026micro;L of cell culture supernatant was loaded onto the column and incubated for 5 min with gentle shaking. After centrifugation, the column was washed three times with 600 \u0026micro;L loading buffer to remove unbound molecules. Elution was performed by 400 \u0026micro;L of 0.1 M citric acid, pH 3.0. The collected fraction containing the antibody was neutralized by adding 30 \u0026micro;L of 1.0 M Tris-HCl, pH 9.0, followed by buffer exchange against 50 mM 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid (HEPES) buffer, pH 8.0, using an Amicon \u0026reg; Ultra\u0026thinsp;\u0026minus;\u0026thinsp;0.5 mL centrifugal filter with molecular weight cut off of 10 kDa (Millipore, Bedford, MA, USA). Antibody was quantified by Cedex \u003csup\u003e\u0026reg;\u003c/sup\u003e Bio Analyzer (Roche Diagnostics Deutschland GmbH, Mannheim, Germany) following the manufacturer\u0026rsquo;s instructions and diluted to a final concentration of 2.0 mg/mL. The InstantPC-labelled N-glycans were prepared by using the AdvanceBio Gly-X N-Glycan Prep with InstantPC kit, 96-ct (Part NO: GX96-IPC, Agilent Technologies, Santa Clara, CA 95051, USA) and profiled by following the protocols described previously without any revision (Xie and Butler, \u003cspan citationid=\"CR57\" class=\"CitationRef\"\u003e2022\u003c/span\u003e; Xie et al., \u003cspan citationid=\"CR59\" class=\"CitationRef\"\u003e2021\u003c/span\u003e). Samples were then stored at -20\u0026deg;C before further use.\u003c/p\u003e\n\u003ch3\u003e8. Chemometric analysis\u003c/h3\u003e\n\u003cp\u003eA chemometric method (Biancolillo and Marini, \u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e2018\u003c/span\u003e) was employed for systematic statistical analysis and differentiation of the chemically defined culture media supplemented with different cottonseed hydrolysates, as well as to identify key components that were associated with cell growth and antibody production under different conditions. Significant differences were determined by Two-way analysis of variance (ANOVA) (Pandis, \u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e2015\u003c/span\u003e, \u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e2016\u003c/span\u003e) with GraphPad Software (San Diego, CA, USA.). To identify potential correlations between compositional variability of culture media supplemented with different cottonseed hydrolysates and cell growth performance, partial least squares discriminant analysis (PLS-DA) and PatternHunter feature correlation were carried out on a web-based platform MetaboAnalyst (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.metaboanalyst.ca/\u003c/span\u003e\u003cspan address=\"https://www.metaboanalyst.ca/\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e) (Pang et al., \u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e2024\u003c/span\u003e; Pang et al., \u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e2022\u003c/span\u003e; Xia et al., \u003cspan citationid=\"CR56\" class=\"CitationRef\"\u003e2009\u003c/span\u003e).\u003c/p\u003e"},{"header":"Results","content":"\n\u003ch3\u003e1. Compositional variability of culture media supplemented with cottonseed hydrolysates\u003c/h3\u003e\n\u003cp\u003eReverse-phase liquid chromatography\u0026ndash;high resolution mass spectrometry (RPLC-HRMS) was used to characterise compositional differences between control culture media and media supplemented with different batches of cottonseed hydrolysates. Characteristic features were detected across an m/z range of 70\u0026ndash;1,050 over a 20 min chromatographic run, capturing a broad range of metabolites including amino acids, peptides, carbohydrates, and lipids. Each feature in the analysis was identified by a mass/charge (m/z) ratio at a specific peak retention time and listed as m/z@ retention time (min). The full dataset of detected features is provided in Supplementary Information (ST1).\u003c/p\u003e \u003cp\u003eA representative volcano plot comparing hydrolysate-supplemented media (H2B) with control media (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eA) showed extensive upregulated features (fold change\u0026thinsp;\u0026ge;\u0026thinsp;2.0, p\u0026thinsp;\u0026le;\u0026thinsp;0.1), corresponding to components introduced by the hydrolysate. A smaller number of downregulated features were also observed, likely reflecting dilution effects during hydrolysate stock preparation. Direct comparison between two hydrolysate batches (H3C vs H1B; Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eB) revealed substantial compositional differences, with distinct sets of features enriched in each condition. These results confirm pronounced batch-dependent compositional variability in cottonseed hydrolysates.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e\n\u003ch3\u003e2. Chemometric differentiation of hydrolysate-supplemented media\u003c/h3\u003e\n\u003cp\u003eSparse partial least squares discriminant analysis (SPLS-DA) was applied to the RPLC-HRMS feature set to assess compositional differences across conditions. At day 0, samples clustered according to hydrolysate type (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eA), with clear separation between control media and hydrolysate-supplemented media. The three batches of each hydrolysate type (H1, H2, H3) formed distinct clusters, indicating reproducible within-group similarity and between-group variability.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eThe top 10 signature features that discriminate between the hydrolysate batches and contribute to the first and second components are shown in Figs.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eB and \u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eC, respectively. The features are defined by their mass-to-charge ratio and retention time (m/z@RT). For example, [email protected] and [email protected] were among the most influential variables differentiating hydrolysate groups. Representative extracted ion chromatograms illustrating variation in feature abundance across conditions are shown in Figs.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eD and \u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eE.\u003c/p\u003e \u003cp\u003eThese results demonstrate that LC-MS-derived feature profiles combined with chemometric analysis can effectively distinguish between hydrolysate batches and define compositional signatures associated with each group.\u003c/p\u003e\n\u003ch3\u003e3. Cottonseed hydrolysates prolonged cell viability and enhanced antibody productivity\u003c/h3\u003e\n\u003cp\u003eCHO DG44 cells were cultured in the presence or absence of cottonseed hydrolysates (5.0 g/L) to evaluate their effects on growth and productivity. Control cultures exhibited a typical batch growth profile, with exponential growth (day 0\u0026ndash;4), a stationary phase (day 4\u0026ndash;7), and a decline phase thereafter (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eA). In contrast, hydrolysate-supplemented cultures reached lower peak viable cell densities but maintained high viability (\u0026gt;\u0026thinsp;90%) throughout the 10-day culture period (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eB).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eMaximum viable cell densities in hydrolysate-supplemented cultures ranged from 2.9 to 4.7 \u0026times; 10^6 cells/mL, compared to approximately 6 \u0026times; 10^6 cells/mL in the control. Correspondingly, IVCD values were generally lower in supplemented cultures (19.62\u0026ndash;25.63 \u0026times; 10^6 cells\u0026middot;days/mL) than in the control (29.93\u0026thinsp;\u0026plusmn;\u0026thinsp;0.98), with the exception of H2A, which showed comparable IVCD (Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e). Despite reduced cell densities, hydrolysate supplementation markedly prolonged culture viability beyond day 7.\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab1\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 1\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eThe effect of cottonseed hydrolysate supplements on integral viable cell density, maximum viable cell density, maximum antibody concentration, and specific productivity over the 10-day batch cultivation\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"5\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eHydrolysate supplement\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eIVCD\u003c/p\u003e \u003cp\u003e(\u0026times;10\u003csup\u003e6\u003c/sup\u003e cells\u0026middot;days/mL)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eVCD\u003csub\u003emax\u003c/sub\u003e\u003c/p\u003e \u003cp\u003e(\u0026times;10\u003csup\u003e6\u003c/sup\u003e cells/mL)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003eCMab at day 10\u003c/p\u003e \u003cp\u003e(mg/L)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c5\"\u003e \u003cp\u003eqMab\u003c/p\u003e \u003cp\u003e(pg/cell/day)\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eControl\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e29.93\u0026thinsp;\u0026plusmn;\u0026thinsp;0.98\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e5.89\u0026thinsp;\u0026plusmn;\u0026thinsp;0.23\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e287.00\u0026thinsp;\u0026plusmn;\u0026thinsp;4.36\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e9.60\u0026thinsp;\u0026plusmn;\u0026thinsp;0.44\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eH1A\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e20.85\u0026thinsp;\u0026plusmn;\u0026thinsp;0.75\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e3.56\u0026thinsp;\u0026plusmn;\u0026thinsp;0.27\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e403.00\u0026thinsp;\u0026plusmn;\u0026thinsp;10.44\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e19.34\u0026thinsp;\u0026plusmn;\u0026thinsp;0.41\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eH1B\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e20.20\u0026thinsp;\u0026plusmn;\u0026thinsp;0.83\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e3.19\u0026thinsp;\u0026plusmn;\u0026thinsp;0.33\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e351.67\u0026thinsp;\u0026plusmn;\u0026thinsp;12.66\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e17.43\u0026thinsp;\u0026plusmn;\u0026thinsp;1.17\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eH1C\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e20.11\u0026thinsp;\u0026plusmn;\u0026thinsp;1.08\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e2.95\u0026thinsp;\u0026plusmn;\u0026thinsp;0.20\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e382.00\u0026thinsp;\u0026plusmn;\u0026thinsp;9.54\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e19.02\u0026thinsp;\u0026plusmn;\u0026thinsp;0.78\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eH2A\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e28.53\u0026thinsp;\u0026plusmn;\u0026thinsp;0.80\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e4.66\u0026thinsp;\u0026plusmn;\u0026thinsp;0.21\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e575.67\u0026thinsp;\u0026plusmn;\u0026thinsp;2.52\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e20.19\u0026thinsp;\u0026plusmn;\u0026thinsp;0.54\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eH2B\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e25.63\u0026thinsp;\u0026plusmn;\u0026thinsp;0.73\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e3.92\u0026thinsp;\u0026plusmn;\u0026thinsp;0.37\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e599.00\u0026thinsp;\u0026plusmn;\u0026thinsp;4.00\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e23.38\u0026thinsp;\u0026plusmn;\u0026thinsp;0.51\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eH2C\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e23.86\u0026thinsp;\u0026plusmn;\u0026thinsp;0.12\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e3.92\u0026thinsp;\u0026plusmn;\u0026thinsp;0.34\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e517.67\u0026thinsp;\u0026plusmn;\u0026thinsp;10.12\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e21.70\u0026thinsp;\u0026plusmn;\u0026thinsp;0.33\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eH3A\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e25.23\u0026thinsp;\u0026plusmn;\u0026thinsp;0.97\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e3.95\u0026thinsp;\u0026plusmn;\u0026thinsp;0.30\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e585.00\u0026thinsp;\u0026plusmn;\u0026thinsp;12.29\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e23.20\u0026thinsp;\u0026plusmn;\u0026thinsp;0.45\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eH3B\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e19.62\u0026thinsp;\u0026plusmn;\u0026thinsp;0.54\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e2.89\u0026thinsp;\u0026plusmn;\u0026thinsp;0.16\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e510.67\u0026thinsp;\u0026plusmn;\u0026thinsp;11.72\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e26.05\u0026thinsp;\u0026plusmn;\u0026thinsp;1.29\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eH3C\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e21.83\u0026thinsp;\u0026plusmn;\u0026thinsp;1.13\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e3.23\u0026thinsp;\u0026plusmn;\u0026thinsp;0.02\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e534.67\u0026thinsp;\u0026plusmn;\u0026thinsp;13.32\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e24.52\u0026thinsp;\u0026plusmn;\u0026thinsp;0.99\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colspan=\"5\" nameend=\"c5\" namest=\"c1\"\u003e \u003cp\u003eAll the parameters are expressed as mean\u0026thinsp;\u0026plusmn;\u0026thinsp;standard deviation of sample (SD).\u003c/p\u003e \u003cp\u003eIndependent biological triplicate cultures were carried out (n\u0026thinsp;=\u0026thinsp;3).\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003eCell diameter was consistently higher in hydrolysate-supplemented cultures compared to control (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eC). The cell diameter in the control culture decreased notably from day 2 to day 8, to a value of 12.3 \u0026micro;m which was substantially lower than the mean of the hydrolysate-supplemented cultures at 14.5 \u0026micro;m. This difference may be related to the culture osmolality which was substantially higher in the hydrolysate-supplemented cultures (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eD), suggesting a link between hydrolysate supplementation, osmotic conditions, and cell size.\u003c/p\u003e \u003cp\u003eAntibody production profiles differed substantially between conditions (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eE). In control cultures, the antibody titre reached approximately 270 mg/L by day 7 and remained unchanged as viability declined. In contrast, hydrolysate-supplemented cultures continued producing antibody until day 10, resulting in significantly higher final titres. The highest-producing conditions (H2A, H2B, H3A) reached 576\u0026ndash;599 mg/L, more than double the control. Cell-specific productivity (qMab) was also significantly increased in supplemented cultures (17.43\u0026ndash;26.05 pg/cell/day) compared to control (9.60\u0026thinsp;\u0026plusmn;\u0026thinsp;0.44 pg/cell/day) (Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e).\u003c/p\u003e\n\u003ch3\u003e4. Impact of hydrolysates on key metabolic substrates and by-products\u003c/h3\u003e\n\u003cp\u003eThe concentrations of glucose, glutamine, lactate, and ammonia were monitored throughout the culture period (Figs.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eF\u0026ndash;I). Glucose was rapidly consumed in all conditions; however, hydrolysate-supplemented cultures continued to consume glucose beyond day 7, whereas consumption in control cultures plateaued after this point. Lactate accumulation differed markedly between conditions. In control cultures, lactate increased to approximately 28 mM by day 7 and remained elevated. In contrast, hydrolysate-supplemented cultures reached peak lactate concentrations of 13\u0026ndash;19 mM by day 4\u0026ndash;5, followed by a decline, indicating a shift from lactate production to lactate consumption. By day 10, lactate levels in some supplemented cultures (e.g., H2A) were as low as 6 mM. Glutamine was rapidly consumed during the first 4 days in all cultures, after which residual concentrations (~\u0026thinsp;1 mM) remained relatively stable. Ammonia accumulation was consistently higher in control cultures, reaching\u0026thinsp;~\u0026thinsp;9 mM by day 10, compared to 6\u0026ndash;7 mM in hydrolysate-supplemented cultures. These results indicate that hydrolysate supplementation is associated with altered metabolic profiles, including reduced accumulation of inhibitory by-products and a shift in lactate metabolism during later culture stages.\u003c/p\u003e\n\u003ch3\u003e5. Effects of hydrolysates on N-glycosylation of bevacizumab\u003c/h3\u003e\n\u003cp\u003eN-glycosylation profiles of bevacizumab were analysed by HILIC-FLD. All samples exhibited similar glycan structures, with differences observed primarily in relative abundance (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e). Under control conditions, the dominant glycan species was FA2 (76.2\u0026thinsp;\u0026plusmn;\u0026thinsp;0.2%), with other glycans (A1, FA1, A2, M5, FA2G1, FA2G2) present at lower levels. No sialylated glycans were detected. Hydrolysate supplementation significantly altered glycan distributions. In particular, increased levels of mono- and di-galactosylated species (FA2G1\u0026rsquo;, FA2G1, FA2G2) and A2 were observed, accompanied by a decrease in FA1. The abundance of high-mannose glycan M5 varied between hydrolysate groups, with lower levels in H1 and higher levels in H3 relative to control. Statistical analysis confirmed significant differences across conditions (Supplementary Information ST4). These results indicate that cottonseed hydrolysates influence antibody glycosylation, particularly by increasing galactosylation.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e\n\u003ch3\u003e6. Correlation of compositional features with cell growth and antibody production\u003c/h3\u003e\n\u003cp\u003eTo explore relationships between media composition and culture performance, time-resolved LC-MS data from H2B-supplemented cultures were analysed using SPLS-DA and PatternHunter correlation analysis. SPLS-DA revealed clear temporal separation of samples with minimal variation between biological replicates (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eA). The top signature features that discriminated between samples included m/z [email protected], (bevacizumab), and [email protected] (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eB).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eA volcano plot comparing day 10 and day 0 samples (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eC) showed extensive changes in feature abundance, reflecting net production and consumption of metabolites during culture. PatternHunter analysis identified 25 features significantly correlated with cell growth and antibody production.\u003c/p\u003e \u003cp\u003eCell growth was positively correlated with features such as [email protected] and [email protected] (annotated as glutathione oxidized and 2'-deoxycytidine, respectively), and negatively correlated with features including [email protected] (annotated as pyridoxine). Antibody production showed positive correlations with features such as [email protected] and [email protected], and negative correlations with several low-mass metabolites (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003e).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eDistinct correlation patterns were also observed in control cultures, reflecting differences in media composition between supplemented and unsupplemented conditions. These findings demonstrate that LC-MS-derived compositional features can be linked to functional outputs and used to identify candidate markers associated with culture performance.\u003c/p\u003e"},{"header":"Discussion","content":"\u003cp\u003eProtein hydrolysates are widely used as cost-effective supplements in CHO cell culture to enhance productivity. However, their compositional complexity and batch-to-batch variability remain major challenges for understanding and controlling their functional effects (Obaidi et al., \u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e2021\u003c/span\u003e). In this study, we combined LC-HRMS-based compositional profiling with chemometric analysis to systematically relate the variability of cottonseed hydrolysates to CHO DG44 cell culture performance, including growth characteristics, metabolism, antibody productivity, and glycosylation. We have extended the development of this approach that was originally applied to the compositional profiling analysis of variants of soy hydrolysates (Xie and Butler, \u003cspan citationid=\"CR58\" class=\"CitationRef\"\u003e2025\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eThe LC-HRMS workflow enabled clear differentiation of culture media supplemented with different hydrolysate batches based on distinct compositional signatures. Unsupervised and supervised chemometric analyses demonstrated that batches clustered according to hydrolysate type, indicating reproducible yet distinct compositional profiles. These findings confirm that even hydrolysates derived from similar sources can exhibit substantial compositional variability, which is likely to contribute to differences in their biological effects.\u003c/p\u003e \u003cp\u003eFunctionally, cottonseed hydrolysate supplementation did not increase peak viable cell density and, in most cases, resulted in lower maximum cell densities and integral viable cell density (IVCD) compared to the control. However, a key observation was the prolonged maintenance of high cell viability in hydrolysate-supplemented cultures, particularly during the late stages of the batch process. This extended culture longevity, combined with a marked increase in cell-specific productivity (qMab), resulted in substantially higher final antibody titres. These results indicate that the primary benefit of hydrolysate supplementation lies not in promoting rapid cell proliferation, but in enhancing culture longevity and productivity during the stationary and decline phases.\u003c/p\u003e \u003cp\u003eHydrolysate supplementation was also associated with an increase in cell diameter and culture osmolality. This was in good agreement with previous research, in which cell size changed significantly over the course of cell culture process (Seewoster and Lehmann, \u003cspan citationid=\"CR50\" class=\"CitationRef\"\u003e1997\u003c/span\u003e; Pan et al, \u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e2017\u003c/span\u003e). Our observed increase in cell size is consistent with previous reports linking hyperosmolar conditions to enlarged CHO cells (Alhuthali et al., \u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e2021\u003c/span\u003e; Romanova et al., \u003cspan citationid=\"CR47\" class=\"CitationRef\"\u003e2021\u003c/span\u003e). As cell volume increases nonlinearly with diameter, this effect may partially explain the reduced viable cell densities observed in hydrolysate-supplemented cultures, despite comparable overall cellular biomass. These findings highlight a limitation of relying solely on cell number as a measure of culture performance and suggest that cell size should be considered when interpreting growth and productivity metrics.\u003c/p\u003e \u003cp\u003eA major metabolic feature of hydrolysate-supplemented cultures was the shift from lactate production to lactate consumption after the exponential growth phase. In contrast, control cultures exhibited continued lactate accumulation, reaching significantly higher concentrations. This metabolic transition has been widely associated with improved culture performance and higher recombinant protein titres. The reduced accumulation of lactate and ammonia in hydrolysate-supplemented cultures likely contributed to the prolonged cell viability observed, as both metabolites are known to exert inhibitory effects on cell growth and productivity (Cruz et al., \u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e2000\u003c/span\u003e; Kanehisa and Goto, \u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e2000\u003c/span\u003e; Pereira et al., \u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e2018\u003c/span\u003e). While the precise molecular mechanisms underlying this metabolic shift remain unclear, it is consistent with previous reports suggesting a transition from high glycolytic flux to more efficient metabolic states associated with improved bioprocess performance (Luo et al., \u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e2012\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eThe glycan profile of the antibody produced under control conditions without hydrolysate supplementation was in good agreement with previous studies (Carillo et al., \u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e2020\u003c/span\u003e; Planinc et al., \u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e2017\u003c/span\u003e; Seo et al., \u003cspan citationid=\"CR51\" class=\"CitationRef\"\u003e2018\u003c/span\u003e). We observed an increase in galactosylated glycan species following hydrolysate supplementation which was consistent with previous studies reporting enhanced galactosylation in the presence of plant-derived hydrolysates (Obaidi et al., \u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e2021\u003c/span\u003e). Although the underlying mechanisms were not directly investigated in this study, these changes may reflect altered intracellular metabolism or enzyme activity within the Golgi apparatus (McDonald et al., \u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e2014\u003c/span\u003e). Given the importance of glycosylation for therapeutic antibody efficacy and stability, these findings highlight an additional dimension through which hydrolysates can influence bioprocess outcomes.\u003c/p\u003e \u003cp\u003eBy integrating compositional profiling with chemometric analysis, we identified 25 signature features that were positively or negatively associated with cell growth and antibody production. These features, defined by their m/z and retention time, represent potential markers linking hydrolysate composition to functional performance. Tentative annotation of selected features suggested the involvement of metabolites such as glutathione, deoxycytidine, and pyridoxine, although most features remain to be structurally confirmed. Importantly, these results demonstrate the potential of data-driven approaches to identify candidate components within complex hydrolysate mixtures that may contribute to enhanced culture performance.\u003c/p\u003e \u003cp\u003eDespite these insights, several limitations should be acknowledged. First, the study was conducted using a single CHO DG44 cell line and a specific chemically defined medium; therefore, the observed effects may not be directly transferable to other cell lines, media formulations, or hydrolysate types. Second, the LC-MS analysis was performed in positive ion mode over a defined m/z range, which may have limited the detection of certain classes of metabolites. Third, the identification of signature features was based primarily on correlation analysis, and causal relationships between specific components and biological effects remain to be established. Future studies integrating targeted metabolomics, functional validation, and multi-omics approaches will be required to confirm the roles of these candidate components.\u003c/p\u003e \u003cp\u003eIn summary, this study demonstrates that LC-HRMS-based compositional profiling combined with chemometric analysis provides a robust framework for linking hydrolysate variability to CHO cell culture performance. Cottonseed hydrolysates enhanced culture longevity, reduced the accumulation of inhibitory metabolites, increased cell-specific productivity, and altered antibody glycosylation, despite not increasing peak cell density. These findings suggest that hydrolysate supplementation primarily modulates cellular metabolism and productivity rather than simply promoting cell growth. The approach described here offers a pathway toward the systematic identification of functional components in complex media supplements and may support the development of more consistent and effective hydrolysate-based strategies for biopharmaceutical manufacturing.\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eAuthor contributions\u003c/strong\u003e\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eYongjing Xie: Conceptualization, methodology, investigation, writing-review and editing; Michael Butler: Validation, writing-review and editing.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAcknowledgements\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe protein hydrolysates used in the current study were kindly provided by Kerry lnc. (Beloit, WI 53511, USA). We thank Hans Huttinga, Angel Varelarohena, Brandon Wrage, Michael woods, Derek Carr of Kerry lnc. for their review and helpful comments in the preparation of this manuscript. We thank Caitriona Walsh from NIBRT Core Facility for RP-UHPLC-HR-ESI-MS/MS spectra acquisition.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eSupplementary Information\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe data underlying this article are available in the article and in its online Supplementary Information.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eFunding\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis work was financially supported by Enterprise Ireland Innovation Partnership Programme (IPP) Award (IP20211007).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConflict of interest statement\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe authors declare that there is no conflict of interest.\u003cstrong\u003e\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eData availability statement\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe original or derived data underlying this article is available from the corresponding author upon reasonable request.\u0026nbsp;\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n\u003cli\u003eAlhuthali, S., Kotidis, P., Kontoravdi, C., (2021) Osmolality Effects on CHO Cell Growth, Cell Volume, Antibody Productivity and Glycosylation. Int J Mol Sci 22.\u003c/li\u003e\n\u003cli\u003eAna Cristina Grodzki, E.B., (2010) Antibody Purification: Affinity Chromatography \u0026ndash; Protein A and Protein G Sepharose. In: Constance Oliver, M.C.J. (Ed.), Immunocytochemical Methods and Protocols. Humana Press, New Jersey, USA, pp. 33-41.\u003c/li\u003e\n\u003cli\u003eAndr\u0026eacute; Siemensma, J.B., Chris Wilcox, Hans Huttinga, (2010) Towards an Understanding of How Protein Hydrolysates Stimulate More Efficient Biosynthesis in Cultured Cells. In: Vijai K. Pasupuleti, A.L.D. (Ed.), Protein Hydrolysates in Biotechnology. Springer, Dordrecht, pp. 33-54.\u003c/li\u003e\n\u003cli\u003eB. Farges-Haddani, B.T., S. Chenu, I. Chevalot, C. Harscoat, I. Marc, J.L. Goergen, A. Marc, (2006) Peptide fractions of rapeseed hydrolysates as an alternative to animal proteins in CHO cell culture media. Process Biochemistry 41, 2297-2304.\u003c/li\u003e\n\u003cli\u003eBiancolillo, A., Marini, F., (2018) Chemometric Methods for Spectroscopy-Based Pharmaceutical Analysis. Front Chem 6, 576.\u003c/li\u003e\n\u003cli\u003eCarillo, S., Perez-Robles, R., Jakes, C., Ribeiro da Silva, M., Millan Martin, S., Farrell, A., Navas, N., Bones, J., (2020) Comparing different domains of analysis for the characterisation of N-glycans on monoclonal antibodies. J Pharm Anal 10, 23-34.\u003c/li\u003e\n\u003cli\u003eColla, G., Hoagland, L., Ruzzi, M., Cardarelli, M., Bonini, P., Canaguier, R., Rouphael, Y., (2017) Biostimulant Action of Protein Hydrolysates: Unraveling Their Effects on Plant Physiology and Microbiome. Front Plant Sci 8, 2202.\u003c/li\u003e\n\u003cli\u003eCombe, M., Sokolenko, S., (2021) Quantifying the impact of cell culture media on CHO cell growth and protein production. Biotechnol Adv 50, 107761.\u003c/li\u003e\n\u003cli\u003eCruz, H.J., Freitas, C.M., Alves, P.M., Moreira, J.L., Carrondo, M.J., (2000) Effects of ammonia and lactate on growth, metabolism, and productivity of BHK cells. Enzyme Microb Technol 27, 43-52.\u003c/li\u003e\n\u003cli\u003eDerek Adams, R.K., and Wei-Shou Hu (2007) Application of Stoichiometric and Kinetic Analyses to Characterize Cell Growth and Product Formation. In: P\u0026ouml;rtner, R. (Ed.), Animal Cell Biotechnology Methods and Protocols, Second ed. Humana Press, Totowa, NJ, pp. 269-284.\u003c/li\u003e\n\u003cli\u003eDhara, V.G., Kumar, S., DeVine, L., Naik, H.M., Cole, R.N., More, A.J., Lau, E.C., Betenbaugh, M.J., (2023) Cottonseed hydrolysate supplementation alters metabolic and proteomics responses in Chinese hamster ovary cell cultures. Biotechnol J 18, e2200243.\u003c/li\u003e\n\u003cli\u003eDu, Q., Zhang, X., Wang, T., Wang, X., (2022) Effects and mechanisms of animal-free hydrolysates on recombination protein yields in CHO cells. Appl Microbiol Biotechnol 106, 7387-7396.\u003c/li\u003e\n\u003cli\u003eFerrara, N., Hillan, K.J., Novotny, W., (2005) Bevacizumab (Avastin), a humanized anti-VEGF monoclonal antibody for cancer therapy. Biochem Biophys Res Commun 333, 328-335.\u003c/li\u003e\n\u003cli\u003eFranek, F., Hohenwarter, O., Katinger, H., (2000) Plant protein hydrolysates: preparation of defined peptide fractions promoting growth and production in animal cells cultures. Biotechnol Prog 16, 688-692.\u003c/li\u003e\n\u003cli\u003eHo, Y.Y., Lu, H.K., Lim, Z.F.S., Lim, H.W., Ho, Y.S., Ng, S.K., (2021) Applications and analysis of hydrolysates in animal cell culture. Bioresour Bioprocess 8, 93.\u003c/li\u003e\n\u003cli\u003eHu, D., Zhao, L., Wang, J., Fan, L., Liu, X., Wang, H., Tan, W.S., (2018) Physiological responses of Chinese hamster ovary cells to a productivity-enhancing yeast extract. J Biosci Bioeng 126, 636-643.\u003c/li\u003e\n\u003cli\u003eJuliet Lobo-Alfonso, P.P., David Jayme, (2008) Benefits and Limitations of Protein Hydrolysates as Components of Serum-Free Media for Animal Cell Culture Applications. In: Vijai K. Pasupuleti, A.L.D. (Ed.), Protein Hydrolysates in Biotechnology. Springer, Dordrecht, pp. 55-78.\u003c/li\u003e\n\u003cli\u003eKanehisa, M., Goto, S., (2000) KEGG: kyoto encyclopedia of genes and genomes. Nucleic Acids Res 28, 27-30.\u003c/li\u003e\n\u003cli\u003eKim, J.Y., Kim, Y.G., Lee, G.M., (2012) CHO cells in biotechnology for production of recombinant proteins: current state and further potential. Appl Microbiol Biotechnol 93, 917-930.\u003c/li\u003e\n\u003cli\u003eKoumantakis, G., Wyndham, L.E., (1989) An evaluation of osmolality measurement by freezing point depression using micro-amounts of sample. J Automat Chem 11, 80-83.\u003c/li\u003e\n\u003cli\u003eKristinsson, H.G., Rasco, B.A., (2000) Fish protein hydrolysates: production, biochemical, and functional properties. Crit Rev Food Sci Nutr 40, 43-81.\u003c/li\u003e\n\u003cli\u003eKumar, S., Dhara, V.G., Orzolek, L.D., Hao, H., More, A.J., Lau, E.C., Betenbaugh, M.J., (2021) Elucidating the impact of cottonseed hydrolysates on CHO cell culture performance through transcriptomic analysis. Appl Microbiol Biotechnol 105, 271-285.\u003c/li\u003e\n\u003cli\u003eLee, L.C., Liong, C.Y., Jemain, A.A., (2018) Partial least squares-discriminant analysis (PLS-DA) for classification of high-dimensional (HD) data: a review of contemporary practice strategies and knowledge gaps. Analyst 143, 3526-3539.\u003c/li\u003e\n\u003cli\u003eLi, M., Ma, B., Kisman, D., Tromp, J., (2004) Patternhunter II: highly sensitive and fast homology search. J Bioinform Comput Biol 2, 417-439.\u003c/li\u003e\n\u003cli\u003eLuo, J., Vijayasankaran, N., Autsen, J., Santuray, R., Hudson, T., Amanullah, A., Li, F., (2012) Comparative metabolite analysis to understand lactate metabolism shift in Chinese hamster ovary cell culture process. Biotechnol Bioeng 109, 146-156.\u003c/li\u003e\n\u003cli\u003eLuo, Y., Pierce, K.M., (2012) Development toward rapid and efficient screening for high performance hydrolysate lots in a recombinant monoclonal antibody manufacturing process. Biotechnol Prog 28, 1061-1068.\u003c/li\u003e\n\u003cli\u003eMa, B., Tromp, J., Li, M., (2002) PatternHunter: faster and more sensitive homology search. Bioinformatics 18, 440-445.\u003c/li\u003e\n\u003cli\u003eMcCarthy, D.J., Smyth, G.K., (2009) Testing significance relative to a fold-change threshold is a TREAT. Bioinformatics 25, 765-771.\u003c/li\u003e\n\u003cli\u003eMcDonald, A.G., Hayes, J.M., Bezak, T., Gluchowska, S.A., Cosgrave, E.F., Struwe, W.B., Stroop, C.J., Kok, H., van de Laar, T., Rudd, P.M., Tipton, K.F., Davey, G.P., (2014) Galactosyltransferase 4 is a major control point for glycan branching in N-linked glycosylation. J Cell Sci 127, 5014-5026.\u003c/li\u003e\n\u003cli\u003eMeghna Rajvanshi, K.V.V., (2013) Pathway Modeling, Metabolic. In: Werner Dubitzky, O.W., Kwang-Hyun Cho, Hiroki Yokota (Ed.), Encyclopedia of Systems Biology. Springer, New York, USA, pp. 1652\u0026ndash;1655.\u003c/li\u003e\n\u003cli\u003eNg, J.Y., Chua, M.L., Zhang, C., Hong, S., Kumar, Y., Gokhale, R., Ee, P.L.R., (2020) Chlorella vulgaris Extract as a Serum Replacement That Enhances Mammalian Cell Growth and Protein Expression. Front Bioeng Biotechnol 8, 564667.\u003c/li\u003e\n\u003cli\u003eObaidi, I., Mota, L.M., Quigley, A., Butler, M., (2021) The role of protein hydrolysates in prolonging viability and enhancing antibody production of CHO cells. Appl Microbiol Biotechnol 105, 3115-3129.\u003c/li\u003e\n\u003cli\u003eOzturk, S.S., Riley, M.R., Palsson, B.O., (1992) Effects of ammonia and lactate on hybridoma growth, metabolism, and antibody production. Biotechnol Bioeng 39, 418-431.\u003c/li\u003e\n\u003cli\u003ePan, X., Dalm, C., Wijffels, R.H., Martens, D.E., (2017) Metabolic characterization of a CHO cell size increase phase in fed-batch cultures. Appl Microbiol Biotechnol 101, 8101-8113.\u003c/li\u003e\n\u003cli\u003ePandis, N., (2015) Two-way analysis of variance: Part 1. Am J Orthod Dentofacial Orthop 148, 1078-1079.\u003c/li\u003e\n\u003cli\u003ePandis, N., (2016) Two-way analysis of variance: Part 2. Am J Orthod Dentofacial Orthop 149, 137-139.\u003c/li\u003e\n\u003cli\u003ePang, Z., Lu, Y., Zhou, G., Hui, F., Xu, L., Viau, C., Spigelman, A.F., MacDonald, P.E., Wishart, D.S., Li, S., Xia, J., (2024) MetaboAnalyst 6.0: towards a unified platform for metabolomics data processing, analysis and interpretation. Nucleic Acids Res.\u003c/li\u003e\n\u003cli\u003ePang, Z., Zhou, G., Ewald, J., Chang, L., Hacariz, O., Basu, N., Xia, J., (2022) Using MetaboAnalyst 5.0 for LC-HRMS spectra processing, multi-omics integration and covariate adjustment of global metabolomics data. Nat Protoc 17, 1735-1761.\u003c/li\u003e\n\u003cli\u003ePatterson, T.A., Lobenhofer, E.K., Fulmer-Smentek, S.B., Collins, P.J., Chu, T.M., Bao, W., Fang, H., Kawasaki, E.S., Hager, J., Tikhonova, I.R., Walker, S.J., Zhang, L., Hurban, P., de Longueville, F., Fuscoe, J.C., Tong, W., Shi, L., Wolfinger, R.D., (2006) Performance comparison of one-color and two-color platforms within the MicroArray Quality Control (MAQC) project. Nat Biotechnol 24, 1140-1150.\u003c/li\u003e\n\u003cli\u003ePereira, S., Kildegaard, H.F., Andersen, M.R., (2018) Impact of CHO Metabolism on Cell Growth and Protein Production: An Overview of Toxic and Inhibiting Metabolites and Nutrients. Biotechnol J 13, e1700499.\u003c/li\u003e\n\u003cli\u003ePlaninc, A., Dejaegher, B., Vander Heyden, Y., Viaene, J., Van Praet, S., Rappez, F., Van Antwerpen, P., Delporte, C., (2017) Batch-to-batch N-glycosylation study of infliximab, trastuzumab and bevacizumab, and stability study of bevacizumab. Eur J Hosp Pharm 24, 286-292.\u003c/li\u003e\n\u003cli\u003eRichard G. Brereton, G.R.L., (2014) Partial least squares discriminant analysis: taking the magic away. Journal of Chemometrics 28, 213-225.\u003c/li\u003e\n\u003cli\u003eRichardson, J., Shah, B., Bondarenko, P.V., Bhebe, P., Zhang, Z., Nicklaus, M., Kombe, M.C., (2015) Metabolomics analysis of soy hydrolysates for the identification of productivity markers of mammalian cells for manufacturing therapeutic proteins. Biotechnol Prog 31, 522-531.\u003c/li\u003e\n\u003cli\u003eRitacco, F.V., Wu, Y., Khetan, A., (2018) Cell culture media for recombinant protein expression in Chinese hamster ovary (CHO) cells: History, key components, and optimization strategies. Biotechnol Prog 34, 1407-1426.\u003c/li\u003e\n\u003cli\u003eRodriguez, E.L., Poddar, S., Iftekhar, S., Suh, K., Woolfork, A.G., Ovbude, S., Pekarek, A., Walters, M., Lott, S., Hage, D.S., (2020) Affinity chromatography: A review of trends and developments over the past 50 years. J Chromatogr B Analyt Technol Biomed Life Sci 1157, 122332.\u003c/li\u003e\n\u003cli\u003eRodriguez, R.R., Alhamad, K., Ghosh, S., Bihler, E., Raj, M.S., (2021) Atypical Mycobacterium Abscessus Infection in Stable Chronic Lymphocytic Leukemia: A Case Report and Review of the Literature. Cureus 13, e20574.\u003c/li\u003e\n\u003cli\u003eRomanova, N., Niemann, T., Greiner, J.F.W., Kaltschmidt, B., Kaltschmidt, C., Noll, T., (2021) Hyperosmolality in CHO culture: Effects on cellular behavior and morphology. Biotechnol Bioeng 118, 2348-2359.\u003c/li\u003e\n\u003cli\u003eSchaafsma, G., (2009) Safety of protein hydrolysates, fractions thereof and bioactive peptides in human nutrition. Eur J Clin Nutr 63, 1161-1168.\u003c/li\u003e\n\u003cli\u003eSchellenberg, J., Nagraik, T., Wohlenberg, O.J., Ruhl, S., Bahnemann, J., Scheper, T., Solle, D., (2022) Stress-induced increase of monoclonal antibody production in CHO cells. Eng Life Sci 22, 427-436.\u003c/li\u003e\n\u003cli\u003eSeewoster, T., Lehmann, J., (1997) Cell size distribution as a parameter for the predetermination of exponential growth during repeated batch cultivation of CHO cells. Biotechnol Bioeng 55, 793-797.\u003c/li\u003e\n\u003cli\u003eSeo, N., Polozova, A., Zhang, M., Yates, Z., Cao, S., Li, H., Kuhns, S., Maher, G., McBride, H.J., Liu, J., (2018) Analytical and functional similarity of Amgen biosimilar ABP 215 to bevacizumab. MAbs 10, 678-691.\u003c/li\u003e\n\u003cli\u003eSpearman, M., Lodewyks, C., Richmond, M., Butler, M., (2014) The bioactivity and fractionation of peptide hydrolysates in cultures of CHO cells. Biotechnol Prog 30, 584-593.\u003c/li\u003e\n\u003cli\u003eStrober, W., (2015) Trypan Blue Exclusion Test of Cell Viability. Curr Protoc Immunol 111, A3 B 1-A3 B 3.\u003c/li\u003e\n\u003cli\u003eTorres, M., Hawke, E., Hoare, R., Scholey, R., Pybus, L.P., Young, A., Hayes, A., Dickson, A.J., (2025) Deciphering molecular drivers of lactate metabolic shift in mammalian cell cultures. Metab Eng 88, 25-39.\u003c/li\u003e\n\u003cli\u003eWurm, F.M., (2004) Production of recombinant protein therapeutics in cultivated mammalian cells. Nat Biotechnol 22, 1393-1398.\u003c/li\u003e\n\u003cli\u003eXia, J., Psychogios, N., Young, N., Wishart, D.S., (2009) MetaboAnalyst: a web server for metabolomic data analysis and interpretation. Nucleic Acids Res 37, W652-660.\u003c/li\u003e\n\u003cli\u003eXie, Y., Butler, M., (2022) Construction of an InstantPC-derivatized glycan glucose unit database: A foundation work for high-throughput and high-sensitivity glycomic analysis. Glycobiology 32, 289-303.\u003c/li\u003e\n\u003cli\u003eXie, Y., Butler, M., (2025) Compositional profiling of protein hydrolysates by high resolution liquid chromatography-mass spectrometry and chemometric analysis. Food Chem 487, 144756.\u003c/li\u003e\n\u003cli\u003eXie, Y., Mota, L.M., Bergin, A., O\u0026apos;Flaherty, R., Jones, A., Morgan, B., Butler, M., (2021) High-throughput and high-sensitivity N-Glycan profiling: A platform for biopharmaceutical development and disease biomarker discovery. Anal Biochem 623, 114205.\u003c/li\u003e\n\u003cli\u003eYang, Y., Li, Z., Li, Q., Ma, K., Lin, Y., Feng, H., Wang, T., (2022) Increase recombinant antibody yields through optimizing vector design and production process in CHO cells. Appl Microbiol Biotechnol 106, 4963-4975.\u003c/li\u003e\n\u003cli\u003eYao, T., Asayama, Y., (2017) Animal-cell culture media: History, characteristics, and current issues. Reprod Med Biol 16, 99-117.\u003c/li\u003e\n\u003cli\u003eZhang, Y., Tu, D., Shen, Q., Dai, Z., (2019) Fish Scale Valorization by Hydrothermal Pretreatment Followed by Enzymatic Hydrolysis for Gelatin Hydrolysate Production. Molecules 24.\u003c/li\u003e\n\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":true,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"Chemometrics, Chinese hamster ovary cells, high resolution tandem mass spectrometry, glycan analysis, monoclonal antibody, protein hydrolysates","lastPublishedDoi":"10.21203/rs.3.rs-9267576/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-9267576/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eProtein hydrolysates have attracted increasing interest as cost-effective media supplements for mammalian cell culture, including Chinese hamster ovary (CHO) cells used widely in biopharmaceutical production. However, the biological basis of their beneficial effects remains poorly understood because of their compositional complexity and batch-to-batch variability. In this study, time-resolved compositional profiles of culture media supplemented with different batches of cottonseed hydrolysates were analysed by liquid chromatography-high resolution mass spectrometry (LC-HRMS) and related to CHO DG44 cell growth, antibody productivity, cellular metabolism, and antibody glycosylation. During 10-day batch cultures, hydrolysate supplementation prolonged high cell viability and significantly enhanced antibody productivity, despite lower peak viable cell densities than the control. Hydrolysate-supplemented cultures also showed reduced lactate and ammonia accumulation, consistent with altered nutrient utilization and metabolic activity. In addition, cottonseed hydrolysates significantly increased antibody galactosylation to varying degrees. Chemometric analysis further linked hydrolysate compositional variability to culture performance and identified 25 signature features associated with cell growth and antibody production. These findings provide complementary insight into how hydrolysate composition relates to functional performance in CHO cell culture and support the development of a method to identify hydrolysates that support high performance biopharmaceutical manufacturing.\u003c/p\u003e","manuscriptTitle":"Chemometrics and integrative LC-MS identify compositional variability in cottonseed hydrolysates and its effects on CHO cell growth and antibody productivity","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2026-04-16 15:25:45","doi":"10.21203/rs.3.rs-9267576/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"c0ab1638-f7a1-43af-94b1-6d46f843630a","owner":[],"postedDate":"April 16th, 2026","published":true,"recentEditorialEvents":[{"type":"decision","content":"Revision requested","date":"2026-05-08T08:38:54+00:00","index":"","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2026-05-04T22:54:31+00:00","index":18,"fulltext":""},{"type":"editorInvitedReview","content":"","date":"2026-05-03T16:46:02+00:00","index":17,"fulltext":""}],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[],"tags":[],"updatedAt":"2026-05-08T08:53:27+00:00","versionOfRecord":[],"versionCreatedAt":"2026-04-16 15:25:45","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-9267576","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-9267576","identity":"rs-9267576","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}