Immunogenicity is the ability of a foreign substance, such as an antigen or an antibody, to provoke an immune response in the body of a human or other animal.

Wanted immunogenicity is typically related to vaccines, where the injection of an antigen (the vaccine) provokes an immune response against the pathogen (virus, bacteria), protecting the organism from future exposure. Unwanted immunogenicity is an immune response by an organism against a therapeutic antigen (e.g., recombinant protein, or monoclonal antibody). This reaction leads to the production of anti-drug antibodies (ADAs), inactivating the therapeutic effects of the treatment and, in rare cases, inducing adverse effects.

Characterization of the immune response in research and pre-clinical development is necessary to get a valid interpretation of the pre-clinical efficacy and safety data. Immunogenicity should be investigated in the target human population since animal testing and in vitro models cannot precisely predict immune response in humans. EMEA and FDA published relevant guidelines for immunogenicity [1,2].

ADA assay for the assessment of immunogenicity in preclinical and clinical studies are mainly based on direct or bridging Enzyme-Linked Immunosorbent Assay (ELISA). Assay development and validation depend on the availability of a positive control (PC) which is a sample containing ADA against the biological drug of interest. The ADA assay is characterized by a multi-tiered approach, and the crucial assessment during the development and validation of the assay is the evaluation of the cut point (screening, confirmatory, and titration cut point) [3]. Other relevant assessments of the validation are sensitivity, drug tolerance, and specificity.

A case study will be also discussed to show the impact of ADA on the interpretation of PK data in a preclinical study.

References
[1] Developing and Validating Assays for Anti-Drug antibody Detection. U.S. Department of Health and Human Sciences, Food and Drug Administration, Center for Drug Evaluation and Research (CDER), Center for Biologics Evaluation and Research (CBER). January 2019
[2] Guideline on Immunogenicity Assessment of Therapeutic Proteins. European Medicines Agency (EMEA). Committee for Medical Products for Human Use (CHMP). 01 December 2017
[3] V. Devanarayan et al. Recommendations for Systematic Statistical Computation of Immunogenicity Cut Points. The AAPS Journal, 2017, 19 (5), 1487-1498.

From the preclinical discovery point of view, protein molecules are a subject of interest because they can be both new therapeutic targets and new therapeutic modalities.

What happens in the lab when a target molecule of interest is a protein AND the therapeutic modality to match the target is also a protein?

The presentation will start with the fundamentals of protein structure, dynamics, and function and the move to the approaches and the technologies employed at the very early stages of drug discovery to identify and validate protein targets and therapeutic modalities with a specific focus on protein-based modalities, in order to progress from research to development.

This lecture will show the value of an advanced and integrated protein characterization toolbox to gain precious insights on the drug-target complex useful to be forwarded to the next stages of drug discovery. This prior knowledge is essential for setting the basis for understanding protein-based therapeutics structure-function relationship and, accordingly, their efficacy and selectivity, together with the critical quality attributes to be monitored during the development lifecycle.

A selection of the biophysical tools that pave the way for the experimental pathway from the original idea through hit identification, hit-to-lead optimization, lead selection, and passage into the clinical phase will be presented, and a critical assessment based on real case studies will be proposed.

The rise of biopharmaceuticals has revolutionized medicine, and at the forefront lie therapeutic peptides, proteins, and monoclonal antibodies (mAbs). This lecture discusses each modality´s characterization and how this contributes to satisfying product quality, safety, and efficacy.

The presentation introduces each therapeutic modality´s common production processes, exploring techniques like liquid-phase peptide synthesis or solid-phase peptide synthesis for therapeutic peptides or protein expression processes for therapeutic proteins and mAbs.

Ensuring the quality and safety of these complex products is paramount. The lecture discusses the product quality attributes that should be evaluated to ensure proper safety and efficacy for the patients. A broad perspective of quality assessment procedures employed for each modality is provided, revealing their role in analyzing purity and potential product modifications. The lecture also covers proper compliance requirements applicable to these analytical techniques according to their application domain.

The lecture then discusses advanced characterization concepts and how biophysical analysis methods can provide insights into protein higher-order structure, stability, and interaction with target molecules. These advanced techniques offer crucial information to understand molecule´s structure to function relationship.

The final objective of this presentation is to provide the attendees with a deeper appreciation of the characterization processes that are integral parts of therapeutic peptides, proteins, and mAbs drug development lifecycle.

The rapid growth of biotherapeutic industry with more and more complex molecules entering the market forces the need for advanced analytical platforms which can quickly and accurately identify and quantify product quality attributes. Mass spectrometry has the potential to provide more detailed information about the quality attributes of complex products, and MS methods are more sensitive than UV methods for detection of impurities. The multi-attribute method (MAM), a liquid chromatography-mass spectrometry based analytical approach, is an emerging platform which supports biotherapeutics characterization and cGMP testing. The main advantage lies on the ability to monitor multiple quality attributes in a single assay.

Since its introduction, the concept and data handling of MAM has been applied to different workflows, moving to the analysis of protein at intact level, bringing novel advantages and area of application. In this talk, the principles and requirements for correct implementation of MAM will be explored while demonstrating the potential and applicability of the workflow outside QC environment to support bioprocessing pipeline from early stage till batch release and for biosimilarity assessment.

References
[1] R. S. Rogers, N. S. Nightlinger, B. Livingston, P. Campbell, R. Bailey, and A. Balland. MAbs, 2015, 7, 881-90.
[2] S. Millan-Martin, C. Jakes, S. Carillo, J. Bones. J Pharm Biomed Anal. 2023, 234, 115543.
[3] S. Millan-Martin, C. Jakes, S. Carillo, R. Rogers, D. Ren, J. Bones. Nat Protoc. 2023, 18(4), 1056-1089.
[4] S. Millan-Martin, C. Jakes, S. Carillo, L. Gallagher, K. Scheffler, K. Broster, Bones J. Crit Rev Anal Chem. 2023, 1-18.

If disease-modifying therapies for conditions such as Huntington´s disease (HD) are to become a reality, clinicians and researchers need validated biomarkers. Quantitating mutant HTT (mHTT) as an HD biomarker is crucial to enabling the precise measurement of prognosis, progression, and treatment response.

An assay to detect mHTT protein in the cerebrospinal fluid of patients with Huntington´s disease was validated and found to be sensitive enough to detect decreases in mHTT after administering an HTT-lowering new potential drug currently in phase II trials.

The mHTT quantification method was developed on the Erenna^(R) platform and then transferred to a more sensitive SMCxPRO^TM platform (Merck Millipore), keeping the same critical reagents and assay format. The mHTT method was validated on the SMCxPRO^TM platform, following regulatory guidelines, and is a pivotal biomarker to support clinical studies.

Results from validation assessments performed both on artificial CSF and human CSF spiked samples are shown, including the long-term stability of recombinant and endogenous mHTT and for critical reagents (labeled antibodies).

Upon the analysis of clinical study samples, although ISR is not a requirement for biomarker analysis, re-analysis was performed, on a selected set of samples, across different sites and correlation of results was shown to be good.

Additional HTT assays can help to better understand the disease status and treatment effects. For this reason, we validated total HTT assays in CSF and mHTT assays in plasma, PBMC, and Whole Blood.

Based on results from ongoing clinical studies, the relevance of the assays and the sensitivity achieved for analysis of both healthy volunteers´ and HD patients´ samples are explored. The aim is to potentially adapt the assays or select different matrices that are more relevant and less laborious to handle at clinical sites for robust results generation.

In this presentation, we will focus on the development of novel Antibody Drug Conjugates. We will discuss analytical characterization and bioanalytical assays.

First, we will introduce the ADC mode of action. Then, we will present the most important methods for their analytical characterization, taking into consideration the defined critical quality attributes. We will also discuss the main assays needed for the ADC pharmacokinetic profile setup, including both Ligand Binding and LC-MS techniques. Additionally, we will introduce an LC-MS ligand binding hybrid method based on high-resolution mass spectrometry to detect intact protein and determine DAR values.

After an introduction to the ADC mode of action, the most important methods for their analytical characterization will be presented, taking into consideration the defined critical quality attributes. The main assays needed for the ADC pharmacokinetic profile setup will be discussed, including both Ligand Binding and LC-MS techniques. In addition, an LC-MS ligand binding hybrid method, based on high-resolution mass spectrometry, to detect intact protein and determine DAR values will be introduced.

Different free toxin assays will be described, reporting both protein precipitation and solid phase sample preparation techniques.

The Ab-conjugated payload assay will be covered by considering an enzymatic dependent toxin cleavages. This method consists of an immunocapture step that selectively binds IgG antibodies. Subsequently, an enzymatic cleavage step was applied to the immobilized ADC on the plate to generate a free payload. Eluates from the plate were subjected to a protein precipitation technique (PPT) step, and supernatants were finally analyzed by LC-MS/MS to quantify the Ab-conjugated payload.

Data related to qualification of the method will be presented and a comparison of the results obtained using antibody conjugated payload and conjugated antibody methods will be described as well as the application to ADC stability studies.

In addition, an Affinity Capture LC-MS (AFC LC-MS) method that characterizes ADC molecules irrespective of their site of conjugation and linker (cleavable or non-cleavable) type will be presented. This method is particularly useful for ADCs carrying plasma unstable cytotoxic small-molecule drugs, because, in these cases, conjugated payload and free payload LC-MS methods require derivatization of the reactive group on the small-molecule drug [1].

Reference
[1] L.M. Barbero et al., European Journal of Pharmaceutical Sciences, 2023, 188, 106502

Stability is a critical quality attribute of pharmaceutical products and plays a crucial role in the drug development process. Plasma-derived medicinal products (PDMPs) are prepared industrially from human plasma [1] and include products such as albumin, coagulation factors and immunoglobulins, which are life-saving therapeutics for several chronic and acute life-threatening diseases.

Stability studies play a critical role in assuring product quality at all points in the plasma derived product life cycle. Stability studies of these complex biologics present challenges beyond those found for typical small-molecule pharmaceuticals, which may demonstrate non-Arrhenius behavior and degrade during different phases of shelf life. During and after licensure, stability studies on quality attributes provide a critical link between marketed and clinically evaluated plasma derived product, addressing important regulatory concerns by assuring that product quality is maintained throughout the dating period.

Evaluation of shelf life according to ICH guidelines [2], Stability monitoring studies and comparability approach to support process/product changes post-licensure, will be presented. Lean stability [3] case studies used to reduce and optimize the design of standard stability protocols and expedite the generation of stability data without impact to safety, efficacy, or quality of the product, will also be presented.

References
[1] Committee, Social. Directive 2001/82/EC of the European Parliament and of the Council of 6 November 2001 on the Community code relating to veterinary medicinal products. Official Journal L 311.28/11 (2004): 1-66.
[2] Davies, John G., et al. ICH Q5C stability testing of biotechnological/biological products. ICH Quality Guidelines: An Implementation Guide (2017): 345-373.
[3] Adamec, Eric, et al. Lean stability case studies-leveraging science-and risk-based approaches to enable meaningful phase specific pharmaceutical stability strategies. Journal of Pharmaceutical Innovation 16 (2021): 566-574.

Trastuzumab is the first HER2 tracer mAb discovered and represents one of the most significant advances in cancer treatment since HER2 was discovered as a pharmacological target. The use of trastuzumab is well known in the landscape of antibody-drug conjugates (ADCs), which represent the new generation of cancer therapies and have accounted for a growing share of the biologics market over the last decade.[1]

ADCs have the ability to combine the high selectivity of monoclonal antibodies to a given antigen with the potency of small molecules. Hence, it provides the opportunity to deliver drugs to tumor cells while minimizing toxicity to normal tissue, achieving wider therapeutic windows and enhanced pharmacokinetic and pharmacodynamic properties.[2] So far, two ADCs using trastuzumab, trastuzumab emtansine, and trastuzumab deruxtecan, have been approved by the FDA.[3]

Herein, I will present a series of case studies aimed at not compromising the stability and functionality of trastuzumab during the conjugation process in order to facilitate and accelerate ADCs development.

References
[1] D.Y. Oh, Y.J. Bang. HER2-targeted therapies - a role beyond breast cancer. Nat Rev Clin Oncol 2020, 17, 33-48. https://doi.org/10.1038/s41571-019-0268-3
[2] Hafeez U, Parakh S, Gan HK, Scott AM. Antibody-Drug Conjugates for Cancer Therapy. Molecules. 2020 Oct 16;25(20):4764. doi: 10.3390/molecules25204764. PMID: 33081383; PMCID: PMC7587605.
[3] von Arx C, De Placido P, Caltavituro A, Di Rienzo R, Buonaiuto R, De Laurentiis M, Arpino G, Puglisi F, Giuliano M, Del Mastro L. The evolving therapeutic landscape of trastuzumab-drug conjugates: Future perspectives beyond HER2-positive breast cancer. Cancer Treat Rev. 2023 Feb; 113:102500. doi: 10.1016/j.ctrv.2022.102500. Epub 2022 Dec 24. PMID: 36587473.

Glycosylation is recognized as a Critical Quality Attribute for protein therapeutics such as monoclonal antibodies, fusion proteins therapeutic proteins, vaccines, due to its extensive impact on product safety and efficacy. Glycan characterization is important in the process of protein drug development, from early stage candidate selection to late stage regulatory submission, and it is also an indispensable part in the evaluation of biosimilarity. Efficient qualitative and quantitative glycan analysis techniques have been increasingly important for discovery, development and quality control of glycoprotein therapeutics.

The multiple levels of glycan heterogeneity poses a daunting analytical challenge. In the development of therapeutic glycoproteins, especially biosimilar products, glycan analysis usually involves the use of complementary methods for assessing specific glycosylation attributes, such as glycosylation site, glycan structure and abundance.

This presentation focuses on the analytical methods used to characterize protein glycosylation, which include hydrophilic interaction liquid chromatography, reversed-phase liquid chromatography, porous graphitic carbon liquid chromatography, lectin-based microarray, capillary electrophoresis, isoelectric focusing, eventually coupled to mass spectrometry.

Advances and novelties in each separation method, as well as associated challenges and limitations, are discussed at the released glycan, glycopeptide, glycoprotein subunit and intact glycoprotein levels.

We are entering a new era where, alongside the now traditional biotechnologically derived molecules such as copies of endogenous proteins or human monoclonal antibodies, new formats of proteins with superior physicochemical and biological properties are increasingly being developed. Antibody-like molecules and fusion proteins are beginning to be used as new therapeutic tools.

In a world where the product is the process, quality control in its broadest sense is crucial for ensuring the consistency, safety, and efficacy of biopharmaceutical products.

The biotech pharmaceutical industry focuses on using living organisms to produce medicines on both small and large scales. This distinguishes the biotech sector from traditional pharmaceuticals in terms of production (raw materials, processes, and end products) and significantly impacts all control activities, including raw materials, producing organisms, process controls, drug characterization, and final release. This distinction generally involves a difference in the molecular weight of the end product (commonly referred to as large vs. small molecules), leading to a substantial disparity in complexity between the two categories of drugs.

In biopharmaceuticals, control systems result from a meticulous and intelligent blend [1] of traditional methods, such as classical raw material and process controls, supplemented by modern, advanced molecular biology and protein analysis techniques. Quality control in the biopharma industry today is very different from that of traditional pharmaceuticals and traditional biologics, which relied solely on biological activity testing via in-vivo methods.

The complexity of quality control systems generally correlates with the size and structural features of the final protein and the difficulty of the production process. It should be noted that, as of now, no technique is refined enough to capture all details of the material being tested. The only alternative is to use multiple techniques simultaneously, examining test samples from different perspectives and with varying granularity, and then integrating all results into a cohesive picture, where all pieces must fit perfectly within the limits of experimental error.

Today, methods range from traditional cell culture virology and electrophoresis to more sophisticated high-resolution mass spectrometry and next-generation sequencing techniques. To consider these activities merely as routine Good Manufacturing Practice is reductive and fails to acknowledge the considerable impact of the know-how, experience, and creativity that a biotech analytical scientist must bring to effectively design an appropriate, up-to-date control system.

In other words, biotech quality control is modern applied science.

Reference
[1] R.L. Garnick, N.J. Solli, and P.A. Papa. The role of quality control in biotechnology: an analytical perspective. Anal Chem, 1988, 60(23), 2546-57. doi: 10.1021/ac00174a004

Collagen is the most abundant protein in the human body. Its unique biosynthetic process and extensive post translational modifications (PTMs) enable the formation of its characteristic triple helix structure, which is further incorporated into fibrils of increasing complexity. These structures provide tensile strength and elasticity to many tissues in the body [1].
Traditionally, collagen has been extracted from animal derivatives. However, this method faces growing constraints due to ethical and religious concerns, as well as the unfavourable environmental impact associated with livestock production. The high water footprint, energy consumption, and the emission of water pollutants involved in this process make developing sustainable sources necessary and urgent. Currently, research is focusing on marine sources of collagen as an alternative to terrestrial ones.
However, the potential of recombinant technology using yeast and bacteria cells as expression systems is more promising [2]. This biotechnological approach may allow for large-scale production while being environmentally friendly, thanks to lower water consumption and the possibility of bioconversion. Moreover, recombinant technology enables the production of animal-component-free collagen, addressing the variability, potential immunogenicity, and infection risks associated with collagen used in medical applications. It also offers traceable processes that ensure products of consistent quality.
In this context, Gnosis by Lesaffre (GbL) aims to obtain recombinant human collagen type II, developing a green-production system by exploiting yeast fermentation. It is important to note that collagen, to be in its native form, requires enzymatic modifications during its biosynthesis, such as the hydroxylation of proline by the enzyme prolyl 4-hydroxylase (P4H), which forms hydroxyproline (Hyp). This modified amino acid is unique to collagen and is fundamental for stabilizing the triple helix [3, 4]. Therefore, it is essential to incorporate not only the human collagen gene into the yeast cell but also the genes for the enzymes responsible for the biological activity of the protein. This involves selecting the expression strain, optimizing the gene transfection method, and selectively isolating the released protein, alongside extensive structural and functional characterization.

This presentation will highlight the milestones achieved over the past three years in the PhD project, conducted in collaboration with GbL and with the participation of the Group of Analytical Pharmaceutical Chemistry at the University of Geneva. The project focused on developing ad hoc analytical methods and exploring cutting-edge technologies to establish a robust analytical platform, essential for supporting the company at each stage of collagen production.

Initially, collagen expression and integrity were verified using SEC-UV for molecular weight determination and RPLC-ESI-MS peptide mapping for amino acid sequence coverage. However, since preliminary analyses of GbL products indicated the presence of collagen only as low molecular weight fragments, the objective was to find a rapid and effective method for monitoring collagen expression efficiency and determining whether degradation occurred during fermentation or purification steps. To address this, an in-solution bead-based immunoaffinity system was developed and validated using protein A-coated magnetic beads for the immobilization of an anti-procollagen monoclonal antibody. The system was designed to selectively isolate human procollagen type II from complex fermentation matrices, facilitating subsequent structural analysis with minimal sample manipulation. Protein A magnetic beads (100 μL) were cross-linked with 20 μg of an anti-human procollagen II IgG, achieving an average immobilization yield of 97.7% and thus creating a stable and reusable support for the specific procollagen capture. Aliquots of 25 μL beads were then exposed to 50 μL of a 0.04 μg/μL recombinant procollagen standard solution (78 kDa). Binding was conducted at 4 �C and physiological pH, while elution used a 0.2 M acidic glycine buffer, pH 2.5. A RPLC-UV method monitored the antigen-antibody interaction, with quantification showing that 0.067 μg of antigen was captured by 25 μL of beads. The method reproducibility was considered satisfactory (RSD 20.89%), and the system maintained 56% binding capacity over 22 days of repeated use. Specificity tests confirmed exclusive binding of procollagen, without non-specific interactions observed. The system was successfully tested on a raw yeast fermentation sample supplemented with synthetic procollagen antigen, demonstrating effective target protein binding even in complex matrices. Further confirmation of method specificity through an RP-LC-HRMS peptide study identified a single peptide from the C-terminus propeptide region as the antibody binding region [5].

Afterward, a RPLC-ESI-MS method was developed for the detection and quantification of all possible hydroxyproline isomers, including both 3-Hyp and 4-Hyp. Although trans-4-L-Hyp (T4L) is the main collagen isomer, other isomers can form due to harsh hydrolysis conditions (such as D-series amino acids) or impaired P4H function. The method involved protein samples hydrolysis with HCl 6 M at 110 �C for 24 hours and the derivatization of the released amino acid with L-FDVA for UV detection. Different chromatographic columns and mobile phase conditions were tested to resolve selected critical pairs of isomers. While methanol-based mobile phases yielded high resolution, they were considered inefficient in terms of analysis time and solvent consumption. By focusing on the use of acetonitrile and formic acid as mobile phase additive, excellent outcomes were achieved with a HALO^(R) ES-C18 2.7 μm 160 Å (150 x 1.5 mm), which was ultimately chosen for method validation. Within the field of recombinant collagen production, which continues to be a challenge, the separation and quantification of all eight possible Hyp isomers may hold great significance. These isomers serve as valuable indicators of quality, efficacy, and specificity of P4H and, therefore, of the overall success and reliability of the recombinant collagen production process. Our RPLC-UV method, coupled to MS, was tested on different collagen samples (native human II, human IV, chicken II and GbL recombinant procollagen II) to gain insight on their Hyp composition. Consistent with existing literature, both the native chicken and human collagen (II and IV) contain over 90% of T4L in addition to less than 5% of its epimer, cis-4-D-Hyp (C4D). Additionally, the analyzed samples exhibit the presence of trans-3-L-Hyp (T3L), the second most notable isomer in collagen. However, in the case of recombinant procollagen, the content of C4D was higher, likely due to distinct folding status compared to fibrillar collagen; this could potentially enhance Hyp accessibility during hydrolysis, leading to C4D higher production. Nonetheless, the substantial T4L content in recombinant procollagen confirmed successful hydroxylation by P4H and highlights its specificity for 4-hydroxyproline, as evidenced by the absence of T3L [6,7].

Finally, the last part of the presentation will be dedicated to the more recent collaboration with the Group of Analytical Pharmaceutical Chemistry at the University of Geneva to obtain, through a comprehensive approach, a robust HILIC method for collagen peptide mapping analysis. HILIC with its unique retention mechanism based on multiple interactions between hydrophilic stationary phases and polar analytes, has emerged in the last decade as a promising alternative for the separation and analysis of biomolecules, both at the intact and at the peptide level. Nevertheless, RPLC is still the most common choice for peptide mapping analysis, with HILIC often considered only as a complementary technique and focusing on glycosylation. [8] Collagen, due to its rich composition in proline residues and their specific hydroxylation to form hydroxyproline as a distinctive post-translational modification, represents an interesting subject for reassessing HILIC potential over traditional RPLC peptide mapping approaches. The presence of hydroxylated proline residues along the collagen sequence, specifically 4-Hyp and the less common 3-Hyp, introduce polar functionalities that can interact favorably with the hydrophilic stationary phases used in HILIC. These hydrophilic interactions may enhance the selectivity of collagen peptides, making HILIC a suitable platform for their analysis. Moreover, collagen peptides often feature multiple potential hydroxylation sites, further complicating their chromatographic behavior. By exploiting the inherent hydrophilicity of collagen peptides and the distinctive retention mechanisms of HILIC, this analytical approach holds substantial promise in revealing the structural characteristics and modifications of collagen. This is crucial considering that even subtle alterations to the hydroxylation pattern can lead to improper protein folding and severe biological consequences, underscoring the importance of accurate analytical characterization.
This study presents a comprehensive and systematic approach aimed at developing a robust HILIC method to specifically address collagen peptide mapping analysis. A set of sixteen model peptides derived from in-silico predicted tryptic peptides (zero missed cleavages), were selected for detailed fundamental studies in HILIC. Their sequences are representative of different physicochemical properties and structural motifs typical of collagen.
The methodology explores both conventional and state-of-the-art HILIC stationary phases with different mobile phase conditions. Additionally, attention is given to the sample diluent and injection mode, as it is known to be particularly critical in HILIC. In this context, classic injection and Performance Optimizing Injection Sequence (POISe) were compared in a systematic way. By elucidating the factors influencing peptide retention and selectivity in HILIC, the aim was to identify the most suitable conditions for the optimal separation of collagen model peptides, while also clarifying the still underexplored chromatographic behavior of peptides in HILIC. This will contribute to expanding the knowledge and applicability of HILIC to peptide analysis.
Finally, following full optimization of the chromatographic separation, hyphenation with mass spectrometry leads to a thorough understanding of the method effectiveness in analyzing collagen digested samples. This involves comprehensive evaluation of peptide sequence coverage and the ability to discern hydroxylation patterns compared to RPLC analysis.

Beyond the ambitious goal of recombinantly producing such a complex and fundamental protein, this project aimed to provide significant support to the company while also advancing academic research in the field of protein analysis, which is currently a highly trending area. The expertise acquired hold potential for broader applications in method development for other biopharmaceuticals as well. This dual focus underscores the project relevance and impact, both in an industrial and academic context.

References
[1] Ricard-Blum S., Ruggiero F., Pathol. Biol. 2005, 53: 430-442.
[2] Kovar K., Looser V., Hyka P., Merseburger T., Meier C. CHIMIA 2010, 64.11: 813-818.
[3] Gorres K.L., Raines R.T. Crit. Rev. Biochem. Mol. Biol. 2010, 45: 106-124.
[4] Taga Y, Tanaka K., Hattori S., Mizuno K., Matrix Biol. Plus. 2021, 10: 100067.
[5] Lioi, M., Tengattini, S., Bagatin, F., Galliani, S., Daly, S., Massolini, G., & Temporini, C. Anal. Bioanal. Chem. 2023, 415: 3155-3166.
[6] Langrock, T., Hoffmann, R. Amino Acid Analysis. 2019, 47-56.
[7] Lioi, M., Tengattini, S., Gotti, R., Bagatin, F., Galliani, S., Massolini, G., Daly S., Temporini, C. J. Chromatogr, A 2024, 464771.
[8] Periat, A., Krull, I. S., Guillarme, D. J. Sep. Sci. 2015, 38: 357-367.

Ligand Binding Assays (LBAs) are traditionally the first choice for the quantification of biotherapeutics in biological matrices. LBAs take advantage of the interaction between a ligand, the biotherapeutic, and a binding molecule that could be a target protein or an antibody. When LBAs are used to measure biotherapeutics as part of regulatory decisions regarding their safety and efficacy, methods need to be well characterized and validated in order to ensure reliable data to support regulatory decisions [1]. The market offers different LBA platforms and the choice of the most suitable is driven from one side by costs and laboratory routine and from the other by sensitivity, dynamic range, robustness, and ruggedness. The common thing among all the platforms is that, depending on the analyte and the type of matrix, method development may be challenging: high or low affinity of the reagents used as well as the presence in the matrix of interfering agents such as target, receptors, anti-drug antibodies, and binding proteins can affect the success of the assay.

The characterization of the assay so developed is recommended by the ICH M10 Guideline, and it can include some or all of these parameters: dynamic range, minimum required dilution (MRD), critical reagents, selectivity and specificity, sensitivity, accuracy, precision, recovery, and stability of the analyte.

Once the method has been developed, the subsequent step is the validation. Specificity, selectivity, calibration curve (response function), range (LLOQ to ULOQ), accuracy, precision, carry-over, dilution linearity, and stability prove that the method is suited to the analysis of the study samples.

What about tissues and rare matrices? Should the same process be followed or not? An overview of the current discussions about ´non-standard´ matrices will be presented.

Reference
[1] ICH guideline M10 on bioanalytical method validation and study sample analysis

INTRODUCTION
My Ph.D. project is part of the Health and Quality of Life area of the ″Programma Operativo Nationale″ (PON). The research focused on the circular economy, which aims to recover the value from agri-food waste generated by industrial processing, as opposed to the traditional linear economic model [1]. The purpose is to evaluate food waste′s chemical and functional properties and use them to create new health products. This approach would reduce the impact of waste disposal and incineration on the environment and climate change and lower the transportation costs associated with waste products for manufacturing companies [2,3]. The project has been developed in collaboration with two companies in the Romagna region of Italy. Fruttagel, an agri-food company that provided the agri-food samples, and Valpharma SpA, a pharmaceutical company responsible for designing a new formulation enriched with biomolecules from food waste. According to the scientific literature [4], agri-food waste contains bioactive compounds with wide applications in nutraceutical, cosmetic, and pharmaceutical fields. These bioactive compounds include proteins and polyphenols with many biological activities beneficial for health and well-being [1]. The project aim was to valorize waste from legumes (including peas, beans, green beans, and soy) and fruits (such as peaches, apricots, apples, and tomatoes) from fruit juices, tomato sauce, frozen legumes, and soy milk industrial production at Fruttagel. The agri-food company also provided the relative final product intended for sale for comparison with the relative by-product. Samples analyzed in the project vary based on the cultivation techniques used, including conventional, biological, and ″lotta integrata″ cultivation. Conventional farming involves intensive agriculture using chemicals for plant protection and fertilization, while biological farming follows sustainable agricultural practices based on natural substances and processes [5]. "Lotta integrata" aims to reduce chemical use and environmental impact while finding a balance between economic needs and environmental health [5]. Samples supplied have also been taken from different steps of the production chain (pre- and post-cooking). The valorization of food waste was carried out by developing high-yield extraction methods to isolate bioactive compounds and efficient analytical approaches for their chemical and functional characterization, such as HPLC-DAD and UHPLC-MS chromatographic methods and spectrophotometric assays, such as Total Phenolic Content (TPC) and Total Antioxidant Status (TAS), for polyphenols characterization. To determine the protein content, the Kjeldahl method was performed. Finally, the research also included the determination of pesticide levels to ensure the safety of the vegetal extracts.

Chemical and functional characterization of polyphenols in legumes and fruits
All fruit and legume samples have been ground, followed by cryo-lyophilization (24h, -60°C) to remove all water content. Next, polyphenols were extracted from powdered samples using an Ultrasound Assisted Extraction (UAE) method [6]. The solid/liquid extraction procedure involved six repeated cycles of extraction using different solvents such as methanol, acetone, and acidic water, resulting in a higher extraction yield of polyphenols from vegetal matrices. The obtained fruit and legume extracts were filtered, dried, and weighed for Gravimetric analysis. The results, calculated in terms of extraction yield (%), showed that the amount of fruit extracts obtained from the by-products is significantly lower than the relative final products, except for the biological peach by-product. Concerning legume samples, the differences in extraction yield (%) between the by-product and the final product are not that notable except for soy, where the final product yield is higher than that of the by-product. In any case, it was expected that the final products would have a higher extractable percentage than their relative by-products which is possibly due to different compositions of all by-products.

Characterization of fruit Samples by HPLC-DAD
Then, a chromatographic High-Performance Liquid Chromatography-Diode Array Detection method (HPLC-DAD) was developed to determine the quali-quantitative polyphenolic profile of fruit by-products by comparing them with the relative final products. Our work started by optimizing the separation of 22 polyphenols standard using a Kinetex XB-C18 150 mm x 4,6 mm x 5 μm column kept at 20.0° ± 0.8 and a mobile phase consisting of 100% methanol (solvent A) and 2% acetic acid water solution (solvent B) at the flow rate of 1 mL min-1. The detection was performed at five different wavelengths, representing the maximum absorptions of standards in the mixture. The HPLC-DAD method allowed the separation and identification of all 22 compounds in less than 70 minutes in a gradient mode. The method was then validated in terms of sensitivity, linearity, precision, accuracy, and specificity and it was found to be sensitive, with LoD values between 0.2 and 16.5 μg mL-1 and LoQ values between 0.6 and 49.9 μg mL-1. The accuracy assessed both intra- and inter-day showed an average value of 2.5% and 3.5% respectively. The optimal linearity was confirmed by an R² value of over 0.9994 for all standard compounds, and the recovery showed values between 90.54 and 104.44% [7]. The method was then used for the quali-quantitative profile of fruit samples. All sample peaks were identified by comparing the elution order, retention times, and UV-vis absorption spectra with standards. Peaks with no available reference compounds were identified by comparison with absorption parameters of a phenolic class and quantified using the calibration curves of a phenolic class member with similar UV-vis spectra [8]. For apples, phloridzin was found to be the most concentrated polyphenol in the by-products, comparable to the amount found in the final product [7]. Peach by-products contain significant amounts of chlorogenic acid and its derivatives, especially the biological peach by-products which has the highest concentrations of these hydroxycinnamic acids [7]. Apricot samples are mainly characterized by a high fraction of hydroxycinnamic acids and flavonols, with isoquercitrin being the main compound [7]. Finally, tomato samples show that naringenin is the main polyphenol, with significant quantities found in the tomato by-products [7].

Qualitative analysis of legume Samples by UHPLC-DAD-ESI-MSn
The legume samples′ polyphenol qualitative profiles were analyzed by a UHPLC-DAD-ESI-MSn chromatographic method entirely developed at the Department of Pharmacognosy, at the University of Graz, in Professor Franz Bucar′s research group where I spent my Ph.D. period abroad. This method allowed the determination of the qualitative profile of the legume by-products, which was then compared with the related final products. The separation of the 22 polyphenols standard mixture was achieved using a Kinetex C18 C18, 50x2.1 mm, 1.7 μm column kept at 20.0° ± 0.8. The mobile phase consisted of a 0.1% solution of formic acid in water and 100% methanol, allowing for a total chromatographic run time of 27 minutes in a gradient mode, at the flow rate of 0.25 mL min-1. Samples chromatographic peaks were identified by comparing elution order, retention times, UV-Vis absorption spectra, and mass fragmentation patterns with standards and then with literature values if reference compounds were not available. In the qualitative analysis of peas, a similar profile was found between by-products and the relative final product with the presence of two main compounds with 278 m/z and 308 m/z in negative mode [M-H]-. According to the literature, those two compounds seem to belong to the N-phenylpropenoyl-L-amino acids class [9], but their identification is still in progress. In the case of beans, all samples, both by-products and the final product, show high recurring peaks with m/z 385 in the negative mode [M-H]-, which can be connected to derivatives of feruloyl-glucaric acid or feruloyl-galactaric acid [10]. This is an interesting result because glucaric acid derivatives have been shown to significantly reduce blood cholesterol levels [11]. Green bean by-products exhibit a different profile between cooked and fresh by-products, with the presence of Kaempferol 3-glucuronide (m/z 461 [M-H]-) not detected in cooked by-products but visible in fresh by-products. It seems that the high temperature of the baking process has caused the compound degradation. In general, green beans are rich in flavonoids and one of the most interesting compounds is quercetin-3-O-glucuronide (m/z 477 [M-H]-), which has been shown to have an anti-vascular leakage effect [12]. Finally, soy samples, both the by-product and row materials (seeds), reveal the presence of many isoflavones, analyzed in positive mode [M+H]+, known for their preventive properties on menopausal symptoms [13].

The phenolic content of fruit and legume extracts was measured by the Total Phenolic Content (TPC) spectrophotometric assay using the Folin-Ciocalteu reagent [14]. The total polyphenol content of the fruit by-products containing high pulp, such as apricot and peach, is very high and similar to that of the final products. Specifically, for apricots, the TPC values of all samples are very similar, indicating that the amounts of polyphenols remaining in the by-products are comparable to those in the final products. As expected, the apple by-products show lower but still valuable TPCs compared to the final products. In the case of tomato samples, the TPC values of the by-products are significantly lower than those of the final products. This difference could be explained by the composition of the by-products and the low extraction yield observed in the gravimetric determination [7]. Concerning legumes none of the samples show high differences between the by-product and the final product. This is comparable to the results obtained in the gravimetric determination.
The functional characterization of fruit and legume extracts was performed by applying the Total Antioxidant Status (TAS) colorimetric assay. This spectrophotometric Trolox Equivalent Antioxidant Capacity (TEAC) test measures the inhibition of the radical cation ABTS•+ in the presence of an antioxidant molecule [15]. The fruit samples′ TAS results indicate that, in general, the final products have similar TAS values compared to the by-products, except for the biological peach [7]. The same can be observed in the case of legume samples.
Concerning pesticide analysis, currently, only biological fruit UAE extracts have been tested but the further goal will be to analyze all samples. The method used for pesticide analysis was entirely carried out at the agri-food company Fruttagel where I spent part of my Ph.D in-company period. This method, known as UNI EN 15662 of 2018, comes from UNI (Ente Italiano di Normazione), a private non-profit association that sets technical standards [16]. It involves a multi-method approach using GC-MS and LC-MS for detecting pesticide residues after extraction. The obtained results indicate no pesticides in all biological fruit by-product extracts (apricot, apple, peach), except for the biological tomato peel, where minimal pesticide traces are found within the legal limit of <0.01 ppm [17]. This is an important finding for the future use of extracts as active ingredients in new formulations.

Protein content in Legumes and Fruit by-products
The protein content determination was carried out on all samples, both fruits and legumes. The Kjeldahl method was performed according to the guidelines of the Association of Official Agricultural Chemists International (AOAC) [18]. The freeze-dried powder samples were hydrolyzed with concentrated sulfuric acid at 420 °C for 2 hours, and the proteins were determined by titration. The results indicate that legume by-products are highly rich in protein (10.30 ± 1.90 to 34.50 ± 6.20 grams of protein per 100 grams-1 of DW) than fruit samples (2.20± 4.0 to 15.80 ± 2.80 grams of protein per 100 grams-1 of DW). The obtained results highlight the potential use of legume by-products as a valuable source of protein.

CONCLUSIONS
All analyses in this research project have confirmed that fruit and legume waste materials are rich in bioactive molecules such as polyphenols and proteins. The results from the experiments show that we can obtain ingredients suitable for new nutraceuticals and food supplements from the by-products we studied. The fruit by-products we analyzed turned out to be rich sources of polyphenols known for their bioactive and antioxidant properties, which were found in significant concentrations. The HPLC-DAD method used to analyze the samples for polyphenol content was easy to use, fast, and applicable to a wide range of fruit samples with different chemical complexities. The analysis confirmed that most of the fruit by-products contained valuable polyphenols in fair amounts compared to the industrial end product, and some even had the same compounds after juice processing. Additionally, the UHPLC-DAD-ESI-MSn method used to characterize the qualitative profile of polyphenols in legumes enabled us to determine the presence of numerous polyphenols with interesting biological activities in both the by-product and the final product. In addition, it has been confirmed that the biological fruit extracts from the UAE are pesticide-free which is promising for the rest of the samples. In conclusion, results confirmed that the agri-food by-products can potentially be exploited as a promising source of bioactive ingredients to design new formulations with a wide range of applications in the pharmaceutical, cosmetic, and nutraceutical industries falling within a circular economy perspective.

References
[1] Campos, D. A.; Gomez-Garcia, R.; Vilas-Boas, A. A.; Madureira, A. R.; Pintado, M. M. Molecules 2020, 25 (2), 320.
[2] Donner, M.; Gohier, R.; de Vries, H. Science of The Total Environment 2020, 716, 137065.
[3] Laufenberg, G.; Kunz, B.; Nystroem, M. Bioresource Technology 2003, 87 (2), 167-198.
[4] Baiano, A. Molecules 2014, 19 (9), 14821-14842.
[5] Boschiero, M.; De Laurentiis, V.; Caldeira, C.; Sala, S. Environmental Impact Assessment Review 2023, 102, 107187.
[6] Kumar, K.; Srivastav, S.; Sharanagat, V. S. Ultrason Sonochem 2021, 70, 105325.
[7] Terenzi, C.; Bermudez, G.; Medri, F.; Davani, L.; Tumiatti, V.; Andrisano, V.; Montanari, S.; De Simone, Antioxidants 2024, 13 (5), 604.
[8] Mesquita, E.; Monteiro, M. Food Research International 2018, 106, 54-63.
[9] Hensel, A.; Deters, A.; Muller, G.; Stark, T.; Wittschier, N.; Hofmann, T. Planta Med 2007, 73 (2), 142-150.
[10] Nguyen, H. T.; Van Der Fels-Klerx, H. J. (Ine).; Van Boekel, M. A. J. S. Food Chemistry 2017, 230, 14-23.
[11] Walaszek, Z.; Szemraj, J.; Hanausek, M.; Adams, A. K.; Sherman, U. Nutrition Research 1996, 16 (4), 673-681.
[12] Cioffi, E.; Comune, L.; Piccolella, S.; Buono, M.; Pacifico, S. Foods 2023, 12 (14), 2646.
[13] Miadokova, E. Interdisciplinary Toxicology 2009, 2 (4), 211-218.
[14] Redmile-Gordon, M. A.; Armenise, E.; White, R. P.; Hirsch, P. R.; Goulding, K. W. T. A Soil Biol Biochem 2013, 67 (100), 166-173.
[15] Davani, L.; Terenzi, C.; Tumiatti, V.; De Simone, A.; Andrisano, V.; Montanari, S. Journal of Pharmaceutical and Biomedical Analysis 2022, 219, 114943.
[16] Chi siamo - UNI - Ente Italiano di Normazione. https://www.uni.com/chi-siamo/ (accessed 2024-06-06).
[17] Regolamento - 396/2005 - IT - EUR-Lex. https://eur-lex.europa.eu/eli/reg/2005/396/oj (accessed 2024-06-06).
[18] Horwitz, W.; Latimer, G. W. Official Methods of Analysis of AOAC International, 18th ed.; AOAC international: Gaithersburg, Md., 2010.

Introduction
Biopharmaceuticals represent a wide class of pharmaceutical products, including recombinant protein therapeutics (monoclonal antibodies, fusion proteins and vaccines) and cellular and gene therapies (C>s), that have significantly grown in the market over the last decade. These products represent a pivotal advancement in modern medicine and biotechnology, encompassing any biological product produced by or derived from living organisms using recombinant DNA (rDNA) technology. This technology involves inserting specific genes into host cells, such as bacteria, plants, or mammalian cells, which then express the desired proteins. The biological origin of these products offers novel mechanisms of action for treating a variety of conditions, such as cancers, autoimmune diseases, genetic disorders, and infectious diseases.
Despite their advantages over conventional small-molecule drugs, biopharmaceuticals encounter several challenges. Their large molecular size and production in living organisms introduce natural variability at the proteome level, inherent in biologics. This complexity is mainly reflected in their structural and conformational dynamics, the presence of numerous process- and product-related impurities during manufacturing, and characteristic micro-heterogeneity, often resulting from sequence variants and post-translational modifications (PTMs). These modifications, known as critical quality attributes (CQAs), affect the physical, chemical, and biological properties of biopharmaceuticals. Major modifications include size, charge, oxidation, and glycosylation. Therefore, the characterization of recombinant proteins and biopharmaceuticals is crucial to ensure quality, safety, and efficacy. Characterization involves comprehensive analysis, even at early stages of drug development, to establish a well-characterized molecule and gain a thorough understanding of the protein′s structure, activity, immunochemical and physicochemical properties, purity/impurities, and the impact of process changes on these attributes, ensuring they meet stringent standards and provide the intended therapeutic benefits.

Analytical Methods for Protein-Based Biotherapeutics Characterization
In the field of protein-based biopharmaceuticals, the nature of the product and the attribute being analyzed determine the appropriate analytical strategy including top-down, bottom-up, or middle-up (specific to mAbs) approaches. Each method offers unique insights at different organizational levels: top-down analysis allows intact protein examination, providing a holistic view of the protein structure and heterogeneity; bottom-up approaches involve enzymatic digestion of proteins into peptides, facilitating detailed peptide-level analysis, identification and localization of post-translational modifications such as glycosylation; middle-up approaches combine elements of both strategies to analyze larger protein fragments balancing detailed characterization with structural context.
Liquid chromatography (LC) and mass spectrometry (MS) generate comprehensive and robust platforms for the detailed characterization of these complex molecules. LC effectively separates individual protein species or peptides based on their physicochemical properties using diverse stationary phases like reversed-phase (RP), ion-exchange (IEX), size-exclusion (SEC), and hydrophilic interaction (HILIC) LC. MS provides accurate mass analysis and structural information. Common ionization sources in MS include electrospray ionization (ESI), ideal for coupling with LC due to its efficiency with large biomolecules in solution, and matrix-assisted laser desorption/ionization (MALDI), suited for analyzing intact proteins and peptides from a solid matrix. In addition, MS can perform fragmentation through methods such as collision-induced dissociation (CID), electron capture dissociation (ECD), higher-energy collisional dissociation (HCD), and in-source decay (ISD) providing information about sequence, PTMs and structural insights. Depending on the product′s purity, MS can be employed either as a standalone technique or coupled with LC for the separation of complex mixtures followed by detailed mass and structural analysis of the separated components, enhancing the overall resolution and sensitivity of the analysis.

The major aim of this project was to develop adequate characterization strategies tailored to the nature and attributes of different protein-based biopharmaceuticals, using LC, MS, or a combination of both techniques and overcome the limitations and challenges arising e.g. by the implementation of complex biotherapeutic proteins formats to obtain their in-depth characterization.

Cell therapy without cells
The first project involved a collaboration with a company, Corion Biothech srl., to characterize their product developed for treating preeclampsia. This condition affects 3-8% of pregnancies and can lead to fetal-maternal mortality and long-term complications like cardiovascular and metabolic diseases. The biopharmaceutical is produced using a novel ″cell therapy-without-cells″ technologyis derived from human placental derived mesenchymal stem cells (hPDMSC) that are purified, expanded, and cultured in vitro. The secretome product of these cells contain bio-active factors (soluble proteins and extracellular vesicles (EVs)) which mediate cell-to-cell signaling able to revert maternal hypertension and proteinuria, improving fetal outcomes, and inhibits the expression of placental biomarkers associated with the syndrome. The goal of the research project was to develop a suitable analytical platform for this specific type of bioproduct, including sample preparation, qualitative and quantitative characterization of protein/EVs components. For secret agreement only results obtained during method optimization on a reference standard sample (RSS), composed of 13 pooled research and development samples, will be reported.
First, a sample preparation method was set-up based on dialysis and concentration steps. Different volumes of RSS (12.5 mL, 25 mL, 50 mL, and 200 mL) were centrifugated with 3kDa membranes, and Bradford assay allowed to determine protein concentration in recovered aliquots. Then, a SEC-UV method was developed to obtain representative profiles for the recovered samples. The SEC-UV chromatograms showed the same characteristic profile, with a region of higher molecular weight components (EVs), one with soluble proteins and the last with lower molecular weight components. A calibration curve was built (670 kDa-14 kDa) to gain information on eluting proteins MWs. The SEC-UV method allowed also to derive quantitative data on protein content, that were cross-validated by Bradford assay. Data showed that an RSS starting volume of 50 mL enables to obtain recovered samples with a total protein concentration around 100 μg/mL, which is appropriate for accurate LC-MS analysis. Finally, three individual production lots were submitted to the optimized sample preparation protocol. The resulting sample solutions, along with one pooled sample, were subjected to peptide mapping by nano-LC-HRMS. The results obtained allowed to identify for the first time the number, the type and the relative abundance of proteins contained in the new free-cell biopharmaceutical, laying the foundation for the structure-activity relationships understanding.

Subsequentially, SEC-UV was used to characterize the vesicular component of the RSS. EVs are lipid bilayer enclosed structures, with a size range from 40 to 1000 nm, containing proteins, lipids, RNAs, metabolites etc. For the purpose, an EVs standard sample (ESS) was analyzed using SEC-UV and the obtained chromatogram compared to that previously obtained on the RSS. In both the high molecular weight regions a peak with the same retention time was observed largely far from the protein-region. To assess identity on these early-eluting species, the peak was isolated from the ESS and prepared for morphology characterization using scansion electron microscopy (SEM). SEM images show different size structures (from 100 to 1000 nm) having spherical round shapes, which seems to indicate the presence of different EVs subpopulations. The data confirm the vesicular nature of species eluting in the range between 5.2 and 7.5 min in SEC-UV trace. As a future prospective, a two-dimensional (2-D) SEC-SEC-UV will be set up to characterize EVs different subpopulations of ESS and RSS by exploiting new SEC stationary phases dedicated to EVs in second dimension [1].

Glycoproteins
Glycosylation plays important roles in modulating the biological functions of glycoproteins, such as protein folding, stability, solubility, and immunogenicity. Natural glycoproteins produced by cultured cells, often carry heterogeneous N-glycans, due to the complexity of N-glycosylation processing in the biosynthesis. For pharmaceutical applications such as biotherapeutics and glycovaccines, preparation of well-defined and ideally homogenous glycoproteins is of paramount importance. Therefore, close attention has been drawn to the development of glycoengineering strategies to better control the glycan structures and the glycosylation profile. Such achievements are realized through methods like the genetic engineering of host cells, but also via chemical or enzymatic remodeling. Alongside these advancements, there is a concurrent need for rapid analytical methods that can efficiently characterize glycoengineered products thereby accelerating the development phase and ensuring quality in the finished product.

Neoglycoprotein advanced monitoring strategy
A MALDI in-source decay (ISD) Fourier-transform ion cyclotron resonance (FT-ICR) MS method was developed for the direct glycosylation analysis of intact glycoproteins without the need for enzymatic release and prior separation. This top-down approach enabled the generation of glycans ISD fragment ions during the ionization process, allowing for the further fragmentation through collision induced dissociation (CID) which allowed the glycans structure corroboration. As previously demonstrated [2, 3] the use of a sodium-doped MALDI matrix was confirmed to be essential for the detection of glycan moieties over protein fragment ions, and not significantly differences were observed in the presence of other cations (K+, Li+). This method was applied on standard mAbs and their glycoengineered variants with the aim to rapidly and easily demonstrate the success of the glycoengineering process. In fact, differences in the glycosylation profile were observed directly from the MS1 spectra for standard and glycoengineered mAbs. This was confirmed by glycoforms quantitation performed considering solely the univocally diagnostic cross-ring ^0,2A-fragment ions among all the glycan fragments. High repeatability and accuracy in glycan profiling was demonstrated, consistent with conventional methods.
The method was also applied to chemically synthesized glycoconjugate vaccines. In detail, a known tuberculosis antigen, from mycobacterium tuberculosis, Ag85B was chemically glycosylated with a 3-aminopropyl Man3 activated via disuccinimidyl glutarate (DSG) to synthesize glycoconjugates [4,5,6]. Glycoconjugates which vary in glycosidic linkage (1,6-1,6 or 1,2-1,6) and carrier protein (wild type or double mutated silencing two glycosylation sites) were considered to study the impact of linker and glycosidic linkage type on fragmentation patterns. All vaccines showed almost identical fragmentation patterns in terms of identity and relative abundance in MS1, likely due to the labile nature of the linker compared to internal linkages and glycosidic bonds. Predominantly, two glycan fragment ions were observed, generated by cleavage at the linker level, differing by the presence of NH2 or OH at the reducing terminal. However, it was not possible to determine the specific positional isomer even post CID. Thus, the MALDI-ISD method was applied on a TIMS-TOF MS, to evaluate the possibility to gain positional information from ion mobility times. In conclusion, the MALDI-ISD MS method is simple, fast, and provides reliable glycosylation profiles of glycoproteins, even though challenges remain in analyzing isomeric glycan structures, with ion mobility showing promise for more detailed structural information.

Innovative neoglycoprotein synthetic strategy
The goal of the last part of the research project is to produce a glycoengineered vaccine starting from recombinant non-glycosylated protein antigens like Ag85B. This innovative approach combines chemoselective glycosylation of a protein with minimal synthetic glycans (N-acetyl glucosamine) used as acceptors in a transglycosylation reaction catalyzed by endoglycosidases in the presence of a donor. These enzymes can catalyze both the hydrolysis of the chitobiose core of N-glycans between two N-acetyl glucosamine (GlcNAc) residues and transglycosylation reactions that transfer an activated N-glycan (donor) to a protein bearing a single GlcNAc residue (acceptor).
A protein model, RNase A, was glycosylated with GlcNAc chemically activated with different functional groups (amino, aldehyde, and 2-iminomethoxyethyl). Additionally, RNase B was enzymatically deglycosylated using wild-type endoglycosidase M to obtain the natural substrate. Both natural and synthetic acceptors then underwent a transglycosylation reaction catalyzed by endoglycosidases in the presence of a synthetic activated glycan. Two enzyme variants were selected for this process: the wild-type endo M, which is more suited for hydrolysis while it can perform transglycosylation only with small substrates, and the mutated variant N175Q, optimized for transglycosylation.
This study aims to evaluate the suitability of the enzymes in transglycosylation reaction when a linker is introduced in the acceptor, and to investigate the linker′s impact on enzyme recognition of synthetic acceptors compared to natural ones. In order to monitor the glycosylation yields an HILIC-UV-MS method was set up to analyze, at the intact glycoprotein level, the reaction mixture and to define the glycan pattern. The most effective GlcNAc functionalization strategy, identified based on its ability to provide an optimal combination of chemical and chemoenzymatic glycosylation, will be applied to glycosylate Ag85B, to develop a novel way to obtain a neo-glycoconjugate vaccine.

References
[1] Zheng H. et al. Anal Chem. 2020; 92(13), 9239-9246
[2] Urakami et al. ACS Omega 2022, 7, 43, 39280-39286
[3] Nicolardi S. et al. Anal. Chem. 2022, 94, 12, 4979-4987
[4] Tengattini S., et al. Pharmaceutics. 2023,15(5):1321
[5] Rinaldi F., et al. RSC Adv. 2018, 8(41), 23171-23180
[6] Bernardini R., et al. Biol Direct. 2024;19(1):11

Introduction
The advent of ′omics sciences has represented a revolution in many fields of biological research [1].
Among these, lipidomics and metabolomics are gaining increasing interest due to their close connection with the organism′s phenotype. As the main analytical goal of omics approaches is to identify and/or quantify as many compounds as possible within a studied system, powerful analytical techniques and bioinformatics tools are required. Mass spectrometry (MS) is one of the most powerful analytical platforms for the analysis of the lipidome in complex biological matrices and offers promising new insights in this field [2]. Lipids are a large class of biomolecules involved in many biological processes with signaling, biophysical, and metabolic functions. The lipidome is complex, consisting of many species that share the same elemental composition but have different structural and physicochemical properties [3]. Its comprehensive analysis is, therefore, technically challenging, and advanced LC-MS lipidomics workflows aim to better address this complexity. Lipidomics approaches can potentially be applied in all therapeutic areas, including cardiovascular diseases, metabolic syndrome, and inflammatory diseases. This is because an increasing number of human diseases are associated with significant lipidome alterations, and changes in lipid profile have been identified as a major risk factor for many of them [4].

This PhD project aims to unravel the complexity of the lipidome and assemble this knowledge into the description of biological systems to highlight lipid changes in pathology or response to drug treatment. To this end, we have optimized MS-based untargeted analytical workflows for the comprehensive profiling of lipids in different biological samples, starting from sample preparation to data mining and data integration. As LC-MS generates high-quality lipidomic data, excellent performance of both parts of the analytical system is essential to achieve these benefits. However, this must always be supported by proper study design, sample and QC preparation, and data analysis. Throughout the PhD project, efforts have been made to critically evaluate all these steps and to identify the workflow that best suits the biological question to be answered.

During my PhD I applied the optimized untargeted LC-MS lipidomic pipelines to: i) phenotypic drug discovery to test novel bioactive compounds in in vitro cell models ii) the study of lipidomic signatures in in vivo models, and iii) the development of a workflow for high-throughput annotation and structural characterization of oxidized lipid molecular species in complex biological matrices.

i) Phenotypic screening drug discovery to test novel bioactive compounds in in vitro cell models

This study aimed to apply untargeted lipidomics to test the lipid-lowering activity of bioactive compounds and the underlying mechanism. Once better understood, lipid metabolism can be targeted pharmacologically.

Evaluation of the efficacy of Scutellaria baicalensis in the prevention of non-alcoholic fatty liver disease by lipidomics studies
Phytochemicals from natural extracts represent an important source of bioactive compounds potentially useful in the treatment of metabolic diseases characterized by major lipidome reprogramming. These include non-alcoholic fatty liver disease (NAFLD), a metabolic disorder caused by the accumulation of lipids in the liver as a result of an imbalance between lipid synthesis, uptake, and export [5]. There are currently no approved pharmacological treatments for NAFLD. Scutellaira baicalensis, a plant used in Chinese officinal medicine for its hypolipidemic properties, has shown several biological activities associated with lipid and cholesterol-lowering and anti-inflammatory effects, mainly due to its major constituent, baicalin [6].However, the mechanism of action responsible for these effects is not fully understood. In this study, the in vitro lipid-lowering effect of baicalin and the whole extract was evaluated in HepG2 cells by HRMS untargeted lipidomics (positive and negative ESI mode), and the results were examined by multivariate statistical analysis using fenofibrate, a PPAR α agonist, as a reference.
Oleic acid (OA), a monounsaturated fatty acid, was used to induce lipid accumulation in HepG2 cells to create an in vitro phenotypic model of NAFLD. OA treatment resulted in significant lipid droplet formation and dysregulated lipid profiles, including increased levels of triacylglycerols (TGs) and ceramides (Cers), which are associated with the metabolic disorder. Untargeted lipidomic analysis showed that both Scutellaria baicalensis and baicalin could reverse these changes by reducing TG levels and increasing diacylglycerols (DGs). The extract also improved the levels of phosphatidylcholines (PCs) and phosphatidylethanolamines (PEs), suggesting their role in maintaining cellular lipid homeostasis. PCs and PEs are essential for hepatocyte health and their dysregulation is associated with NAFLD. In addition, levels of lysophospholipids, which are critical for inflammation, were increased by treatment, supporting their role in related metabolic disorders.
Overall, this study highlights the importance of integrating lipidomics into phenotypic screening to elucidate the molecular effects of lipid-lowering compounds. Finally, our results were consistent with proteomic changes (parallel project), showing that Scutellaria baicalensis extract is effective in modulating cellular pathways and biological processes related to lipid metabolism, transport, and accumulation in hepatocytes.

ii) Application of HPLC-MS and UHPLC-MS to the study of lipidome signatures in in vivo models

Lipidome investigation of carnosine′s effect on nude mice skin to prevent UV-A damage
The skin is composed of multiple layers with varying lipid content, each with unique and essential biological functions for maintaining skin properties and homeostasis [7]. The barrier function of the skin is largely due to the stratum corneum, the outer layer of the epidermis, which provides an effective barrier to the environment and protects the body from external factors such as bacteria, radicals, or UV radiation. Alterations in the skin′s lipid profile caused by these agents can have serious consequences for skin health and have been implicated in numerous skin diseases [8]. UV-A radiation induces chronic inflammation and the production of reactive oxygen species (ROS), thereby promoting skin photoaging. Several compounds have been used to protect the skin from UV radiation, among which carnosine, an endogenous β-alanyl-L-histidine dipeptide, has shown promising antioxidant and carbonyl scavenger properties that significantly prevent skin aging, making it a compelling ingredient to consider for dermatological use [9]. Based on proteomic data showing that carnosine treatment was able to ameliorate calcium signaling and fibrogenic processes altered by UV-A exposure [10], the aim of this work was to describe by HRMS the changes in the skin lipidome of nude mice exposed to UV-A w/wo treatment with carnosine. This project was carried out during the second year of the PhD in collaboration with Prof. Salvayre of the University of Toulouse (France), who performed the treatment of the animals and collected skin samples.
Lipids from skin samples were extracted, analyzed by HPLC-MS/MS, and processed by MS-DIAL [11]. Metaboanalyst and Ingenuity Pathway Analysis (Qiagen) software were then used for statistical and network analysis. Semi-quantitative analysis of lipids revealed several changes in barrier composition after UV-A irradiation, with or without carnosine treatment, as evidenced by the marked alteration of TGs, free fatty acids (FA), Cers, and phospholipid profile [12]. According to the proteomics results, lipidomic network analyses confirmed a significant modulation of calcium signaling together with an increase in ROS production and TNF signaling caused by UV-A and completely reversed by carnosine treatment. Collectively, these findings support the activity of carnosine to prevent UV-A damage by reducing inflammation and dysregulation of the lipid skin barrier. Finally, we integrated proteomics and lipidomics data to further investigate UV-A skin damage and carnosine effects through multi-scale networks involving protein-lipid interactions. The results supported previous findings from protein and lipid analysis and allowed us to gain a deeper insight into skin biological processes.

Analysis of plasma and liver lipidome profile changes induced by carnosine treatment in obese rats
Obesity is a heterogeneous condition often associated with other diseases, including cardiovascular and metabolic disorders. The liver is often implicated in the pathology of obesity, as prolonged caloric excess in this organ manifests as lipid deposition, which can lead to the development of NAFLD. Hepatic lipid excess and dyslipidemia are highly correlated with the production of lipotoxic intermediates resulting from the presence of excess TGs. This is thought to be involved in the induction of mitochondrial dysfunction, inflammation, and oxidative stress in hepatocytes. Carnosine, due to its antioxidant and carbonyl scavenging properties[9], could help to ameliorate this pathological phenotype. Based on the observed proteomic changes in liver samples (parallel project), we applied 2 min UHPLC-MS/MS lipidomics to obtain untargeted profiles of lipid extracts from plasma and liver tissue of lean and obese rats treated w/wo carnosine. This project was carried out during my visiting period at the ETH (Switzerland) under the supervision of Prof. Nicola Zamboni.
Lipid extracts from plasma and liver were obtained by single-phase liquid extraction (MeOH/MTBE/Chloroform) and separated on RP C18. Lipids were analyzed in both positive and negative ionization modes using ZenoTOF 7600 with CID and operating in DDA mode (top 8). Lipid annotation was based on SLAW [13] output, while multivariate statistical analysis was performed using R.
The untargeted LC-MS profiling of lipids captured a broad and detailed profile of the lipidomic changes observed between the experimental groups. Although the main separation of lipid profiles was by phenotype, supervised multivariate models allowed a clear differentiation between the untargeted LC-MS profiles of obese rats treated with carnosine, both in plasma and liver matrices. Accumulation of TGs was observed in both liver and plasma samples from obese rats. Significant differences among DG and Cer species were also observed, confirming their role in the progression of liver steatosis and fibrotic processes. Interestingly, the pattern of these lipids was partially restored after treatment with carnosine. Particularly interesting findings from these data were also observed in the PC classes, with some lipid species selectively increased after carnosine treatment.

iii) High-throughput annotation and structural characterization of oxidized lipid molecular species in complex biological matrices.

High-throughput detection of oxidized lipids in plasma and liver samples
As a next step of the previous work, the analyses of oxidized lipids could provide a more complete perspective on the efficacy of carnosine in ameliorating the obese phenotype. To this end, we looked at the functional modification of lipids to gain more information on their regulatory capacity. Lipids can undergo numerous modifications through the introduction of small chemical groups via enzymatic and non-enzymatic reactions, which ultimately affect their structure, function, and reactivity[14]. Among these, lipid oxidation has been extensively documented in the context of oxidative stress-related disorders. Technically, the discovery and structural elucidation of oxidized lipids are still lacking, as their identification in biological samples is hampered by their low natural abundance and structural diversity. To deal with this complexity, advanced analytical and computational tools are required.
During my visiting period at the ETH, we optimized a workflow combining a rapid 2-min untargeted lipidomic analysis with our newly assembled pipeline for lipid and epilipid annotation to support high-throughput detection of oxidized lipid molecular species in liver and plasma samples. Lipid extracts were analyzed by UHPLC-MS/MS following the method previously described. Specifically, for oxidized lipid annotation, the MS1 match was extended up to 10 oxygens, while acyl chain MS2 fragments were generated by fatty acid residue-specific fragments and neutral losses to create an MS2 library with all possible combinations of product ions. From the lipid species detected in the untargeted analysis, we were able to annotate 414 (liver samples) and 275 (plasma samples) lipids with putative oxidation. Interestingly, obese rats have distinct epilipidomic signatures enriched in oxygenated 18:1, 18:2, and 18:3 acyl chains of TGs. Unique truncated FA were identified as oxygen cleavage products and the pattern of these lipids was partially restored after treatment with carnosine.

Structural characterization of oxidized lipids using advanced MS approach
Currently, no single analytical method can fully characterize a lipid, and even in the field of MS-based lipidomics, more than one MS experiment is usually required to obtain complete structural information. To date, conventional tandem MS methods such as CID can often leave gaps in the structural elucidation of lipids and epilipids. To identify the exact localization of unsaturation and oxidation positions within the acyl chains, we can adopt a different type of fragmentation mechanism, such as electron-activated dissociation (EAD) [15]. EAD can fragment virtually any C-C bond in the acyl chain, allowing saturation and oxidation to be identified by mass shifts of H (1.00) and O (16.00) respectively. The integration of EAD fragmentation into the commercially available mass spectrometer has enabled the acquisition of EAD-based spectra on a fast LC-MS timescale. During my stay at ETH, I had the opportunity to apply such an advanced MS approach for comprehensive structural characterization of oxidized lipids using ZenoTOF 7600, applying multiple reaction monitoring MRM method (reaction time 30ms, accumulation time 500ms). With EAD we were able to generate accurate annotations of all selected oxidized lipids, providing information on chain length, location of oxygen and double bonds, regioisomers, and cis/trans configuration.

Taken together, the results presented demonstrate the potential of MS-untargeted lipidomic approaches to address lipidome and epilipidome changes and to guide future research on the role of lipids in biological systems, both in physiological and pathological conditions or upon drug treatment.

References [1] Joyce AR, Palsson B. Nat Rev Mol Cell Biol 2006;7:198-210
[2] Cajka T, Fiehn O. Toward Merging Anal Chem 2016;88:524-45.
[3] Wenk MR. Cell 2010;143:888-95.
[4] Wishart DS. Nat Rev Drug Discov 2016;15:473-84.
[5] Bessone F, Razori MV, Roma MG. Cellular and Molecular Life Sciences 2019;76:99-128.
[6] Zhao T, Tang H, Xie L, Zheng Y, Ma Z, Sun Q, et al. Journal of Pharmacy and Pharmacology 2019;71:1353-69.
[7] de Szalay S, Wertz PW. Int J Mol Sci 2023;24:3145.
[8] Bouwstra JA, Ponec M. Biomembranes 2006;1758:2080-95.
[9] Aldini G, Facino RM, Beretta G, Carini M. BioFactors 2005;24:77-87.
[10] Radrezza S, Carini M, Baron G, Aldini G, Negre-Salvayre A, D′Amato A Free Radic Biol Med 2021;173:97-103.
[11] Tsugawa H, Cajka T, Kind T, Ma Y, Higgins B, Ikeda K, et al. Nat Methods 2015;12:523-6.
[12] Zoanni B, Aiello G, Negre-Salvayre A, Aldini G, Carini M, D′Amato A. Int J Mol Sci 2023;24:10009.
[13] Delabriere A, Warmer P, Brennsteiner V, Zamboni N. Anal Chem 2021;93:15024-32.
[14] Criscuolo A, Nepachalovich P, Garcia-del Rio DF, Lange M, Ni Z, Baroni M, et al. Nat Commun 2022;13:6547.
[15] Campbell JL, Baba T. Anal Chem 2015;87:5837-45.

Host Cell Proteins (HCP) are process-related impurities considered a Critical Quality Attribute for biopharmaceuticals [1]. Their monitoring during process development and batch release is fundamental since they may copurify with the biopharmaceutical drug product. HCP can impact the drug in different ways. Residual proteases can potentially degrade the product, reducing the shelf life; biologically active molecules can give secondary and not wanted effects, and some proteins may induce immunogenicity. In the current GMP setting, ELISA methodology is normally implemented for HCP absolute quantification. However, Mass Spectrometry is recently becoming more and more popular for HCP analysis especially in process development phases [2,3].

References
[1] M. Jones, N. Palackal, Biotechnol Bioeng., 2021, 118 (8), 2870-2885.
[2] H. Falkenberg, D. M. Wakdera-Lupa, Biotechnol Prog., 2019, 35 (3), e2788
[3] J. Guo, R. Kufer, MAbs, 2023, 15 (1), 2213365

Introduction

Cannabis sativa L. is a plant with a very complex chemical composition and with a high potential from a pharmaceutical point of view. It includes both psychoactive and non-psychoactive (hemp) varieties, based on the cannabinoid profile. Particularly, non-psychotropic C. sativa is characterized by low levels of Δ⁹-tetrahydrocannabinol (Δ⁹-THC), which is usually below 0.2-0.3%. In the fresh plant material the typical components are represented by cannabinoic acids, mainly cannabidiolic acid (CBDA) and cannabigerolic acid (CBGA) in hemp, followed by their decarboxylated or neutral counterparts, such as cannabidiol (CBD) and cannabigerol (CBG) [1,2]. Besides these cannabinoids, other classes are present in the plant, even if the full cannabinoid composition of the plant is still under investigation. CBD is a bioactive compound, which has demonstrated to possess several biological activities, including the antioxidant, anti-inflammatory, neuroprotective and antiepileptic ones [1,3]. Moreover, several studies have highlighted its antiproliferative activity on different cancer cell lines, even if its mechanism/s of action is still under investigation [4].

In addition to cannabinoids, hemp contains other chemical classes of bioactive compounds, such as polyphenols and policosanols (PCs).

Regarding polyphenols, different classes of phenolics have been identified in the plant [5]. Cannflavins are the typical isoprenoid flavones of C. sativa, with cannflavin A and B (CFL-A and CFL-B) as the most representative ones [5]. These compounds have been demonstrated to possess antioxidant, anti-inflammatory, antiparasitic, and antiviral activities [5,6]. However, there are still few studies of their antiproliferative activity, making the investigation of their bioactivity an important topic.

Finally, PCs are long-chain aliphatic alcohols extracted from different natural waxes. They are characterized by a carbon chain length ranging from 22 to 36 carbon atoms. These compounds are present in different natural matrixes, including C. sativa [7,8]. PCs have shown diverse biological properties, even if the available data regarding their antioxidant and anti-inflammatory activity are limited.

In the light of all the above, my PhD project is a multi-disciplinary study, based on the extraction and analysis of non-psychotropic C. sativa, for the recovery of bioactive compounds belonging to the chemical classes of cannabinoids, polyphenols and PCs. The activity of the enriched extracts obtained from the above-mentioned material, together with pure compounds, was assessed for an array of biological activities, with the focus on antioxidant, anti-inflammatory and antiproliferative ones, using In vitro assays. The results will be described starting from the more polar to the less polar compounds investigated.

Identification of phenolic compounds from non-psychoactive Cannabis sativa L. by UHPLC-HRMS and In vitro assessment of the antiproliferative activity against colorectal cancer cell lines

In the first part of my project, the attention was aimed at obtaining a polyphenol-enriched fraction (PEF) from decarboxylated non-psychotropic C. sativa inflorescences to evaluate its antiproliferative activity against human colon adenocarcinoma cell lines, in comparison with a conventional anticancer drug currently used in chemotherapy. Colorectal cancer (CRC) is indeed one of the most diagnosed cancers in high-income countries. One of the main concerns in CRC is that it can easily develop multidrug resistance, with consequent reduction or inefficacy of current anticancer drugs. Some polyphenol, such as quercetin, has already demonstrated to possess antiproliferative activity against colon cancer cell lines by modulating the expression of cannabinoid receptor 1 (CB1) [9]. Cannflavins, which are chemically related to quercetin, represent new possible compounds to be investigated for the treatment of CRC.

In the light of this, a new extraction and purification method was developed in this work for C. sativa polyphenols [10]. Decarboxylated inflorescences were submitted to a dynamic pre-maceration with n-hexane, to remove lipophilic compounds, followed by a dynamic maceration with a MeOH/acetone (90:10 v/v) with 0.1% HCOOH solution. The extract was then purified by preparative flash column chromatography under normal phase conditions to remove cannabinoids and other coeluting compounds [10]. Cannabinoids free fractions were then combined and brought to dryness, obtaining PEF.

Then, an UHPLC-HRMS method was developed and optimized to fully characterize the extract [10]. After the untargeted qualitative analysis, 32 different phenolic compounds were identified, by comparing their fragmentation pattern with the ones already described in literature. The same chromatographic conditions were applied to quantify the main components of PEF, using HPLC-UV. The main components resulted to be CFL-A, CFL-B and N-trans-feruloyltyramine, with a concentration of 10.7 ± 0.8, 8.1 ± 0.3 and 17.7 ± 2.2 mg/g, respectively.

Secondly, the biological activity of PEF and of its main components was assessed on human CRC cell lines Caco-2 and SW480 [10]. Cells were cultured on a 96-well plate and treated with PEF and pure compounds for 24 and 48 h. The best results were achieved after 48 h of treatment in SW480 cell line with PEF and CFL-A, which resulted to have IC₅₀ values of 3.6 µg gallic acid equivalent (GAE)/mL and 34.7 µM, respectively. This value is lower than the one obtained with the chemotherapy drug cisplatin, making both PEF and CFL-A possible new therapeutic products to be further investigated for their bioactivity against CRC.

Extraction, analysis and In vitro evaluation of the antiproliferative activity of cannabinoids from non-psychoactive Cannabis sativa L. against glioblastoma multiforme cancer cell lines

In the second part of the project, the research was addressed at the extraction and full characterization of a fraction enriched in cannabinoids (CEF) from non-psychotropic C. sativa, to evaluate its antiproliferative activity against human glioblastoma cancer cell lines, in comparison with conventional anticancer drugs currently used in chemotherapy. Glioblastoma multiforme (GBM) is one of the most frequent malignant primary tumours, characterized by high proliferation, invasion, migration, angiogenesis and resistance to conventional anticancer drugs. Since CBD is able to cross the blood brain barrier and to decrease cancer cell proliferation, this compound deserves to be further investigated for the treatment of this pathology [11].

To this aim, CEF was obtained by dynamic maceration with ethanol of the plant material (inflorescences). The extract was then qualitatively characterized with a dedicated UHPLC-HRMS method, while and HPLC-UV was used to quantify the main compounds. After the untargeted qualitative analysis, 28 cannabinoids, including the minor ones, were identified, by comparing the obtained fragmentation pattern with the ones already described in literature. However, its main component resulted to be CBD, with a concentration of 403.1 ± 9.9 mg/g.

CEF was then tested for its activity against GBM cancer cell lines, being U87MG and T98G. After 48 h of treatment CEF showed an IC₅₀ value around 20 and 25 µg/mL in U87MG and T98G cells, respectively. CBD (being the main component of CEF) was also tested, providing IC₅₀ values around 20 and 25 µM in U87MG and T98G cells, respectively, after 48 h of treatment. Temozolomide, being the current anticancer drug used in chemotherapy, gave IC₅₀ values higher than 100 µM in both cell lines at 24 and 48 h of treatment.

Then, a cell migration assay was performed to check whether the cannabinoids were able to decrease cell migration. To achieve this, cells were cultured in a 6-well plate and, after 24 h, treated with CEF at the concentration of 20 µg/mL. The migration assay showed a decrease of cell mobility, when cells are exposed to cannabinoids. Taking in consideration the Random Mobility Coefficients (RMCs), representing the distance covered by a single cell over a given period of time, the means of the obtained RMCs of U87MG cells treated with CEF was significantly lower than the one of control (p < 0.05). CEF was in fact able to drop the RMC value by 83%, in comparison to the RMC mean value of control.

The experimental data indicate cannabinoids as promising candidates for the treatment of GBM. The existence of a possible new target for CBD, involved in cell mobility and migration is plausible. Now the research is focused to a deeper investigation of the mechanism/s of action of cannabinoids, with a focus on CBD.

Extraction, purification and In vitro assessment of the antioxidant and anti-inflammatory activity of policosanols from non-psychoactive Cannabis sativa L.

This third part of the project was aimed at the extraction and purification of PCs from an innovative source, represented by a wax obtained from supercritical fluid extraction (SFE) from the inflorescences of non-psychotropic C. sativa. Then, the obtained purified PCs were tested for their In vitro antioxidant and anti-inflammatory activity.

Initially, PCs were obtained by microwave-assisted trans-esterification and hydrolysis combined in a single step, followed by preparative liquid chromatography under normal phase conditions [12]. The purified product was characterized using a new method based on high-performance liquid chromatography (HPLC) with an evaporative light scattering detector (ELSD) [13]. The quantitative analysis of the purified mixture from hemp wax revealed hexacosanol (C26OH) and octacosanol (C28OH) as the main compounds, with a concentration of 181 ± 0.2 and 130.5 ± 2 mg/g, respectively.

The antiproliferative activity of PCs was initially assessed against a panel of human cancer cell lines, but the results indicated no cytotoxic activity. In vitro cell-free and cell-based antioxidant and anti-inflammatory assays were then performed on the purified mixture [12]. The results indicated an inhibition of intracellular reactive oxygen species (ROS) production, a reduction of nuclear factor kappa B (NF-?B) activation and of the activity of the neutrophil elastase. Immunoblotting assays allowed us to hypothesize the mechanism of action of the compounds of interest, given the higher levels of MAPK-activated protein kinase 2 (MK2) and heme oxygenase-1 (HO-1) protein expression in the PC pretreated HaCaT cells. Even if more research is needed to unveil other molecular mechanisms involved in PC activity from non-psychotropic C. sativa, the results of this work suggest that these compounds may have potential for use in oxinflammation processes.

Conclusions

As an overall conclusion, during this PhD project it was possible to prepare and fully characterize three different extracts from non-psychotropic C. sativa, enriched in different chemical classes of compounds, that were tested for their bioactivity together with pure molecules.

PEF was obtained after the development and optimization of a dedicated extraction procedure and a reliable analytical method based on UHPLC-HRMS. This extract and pure CFL-A were tested against CRC, providing promising results, even if further research is needed to elucidate their mechanism of action.

CEF was fully characterized using UHPLC-HRMS and tested against GBM cell lines, being able to reduce cell viability and their migration after the treatment. Taking this as a starting point for future bioassays, further studies are ongoing to identify the mechanism of action of CBD, the main compound of the extract.

Finally, PCs from hemp wax were purified with an innovative extraction method and then submitted to chemical analysis using HPLC-ELSD. The results of the biological assays clearly indicated their ability to decrease oxinflammation process.

Two paths of potential new drug development against cancer cell proliferation and oxinflammation have emerged from this work. One approach is focused on developing standardized products that are directly derived from non-psychoactive C. sativa. The other is based on a single molecule approach, whereby individual compounds with therapeutic potential are identified and tested for pharmaceutical development.

References
[1] Pellati F., Borgonetti V., Brighenti V., Biagi M., Benvenuti S., Corsi L. BioMed Res. Int. 2018, 2018, 1-15.
[2] Pellati F., Brighenti V., Sperlea J., Marchetti L., Bertelli D., Benvenuti S. Molecules. 2018, 23, 2639.
[3] Costa A.-M., Senn L., Anceschi L., Brighenti V., Pellati F., Biagini G. Pharmaceuticals. 2021, 14, 1259.
[4] Anceschi L., Codeluppi A., Brighenti V., Tassinari R., Taglioli V., Marchetti L., Roncati L. Alessandrini A., Corsi L., Pellati F. Phytother. Res. 2022, 36, 914-927.
[5] Pollastro F., Minassi A., Fresu L.G. Curr. Med. Chem. 2018, 25, 1160-1185
[6] Bautista J.L., Yu S., Tian L. ACS Omega. 2021, 6, 5119-5123.
[7] Venturelli A., Brighenti V., Mascolo D., Pellati F. J. Pharm. Biomed. Anal. 2019, 172, 200-205.
[8] Montone C.M., Aita S.E., Cannazza G., Cavaliere C., Cerrato A., Citti C., Mondello L., Piovesana S., Lagana A., Capriotti A.L. Talanta. 2021, 235, 122778.
[9] Mirazimi S.M.A., Dashti F., Tobeiha M., Shahini A., Jafari R., Khoddami M., Sheida A.H., EsnaAshari P., Aflatoonian A.H., Elikaii F., Zakeri M.S., Hamblin M.R., Aghajani M., Bavarsadkarimi M., Mirzaei H. Front. Pharmacol. 2022, 13, 860209.
[10] Caroli C., Brighenti V., Cattivelli A., Salamone S., Pollastro F., Tagliazucchi D., Pellati F. J. Pharm. Biomed. Anal. 2023, 236, 115723.
[11] Buchalska B., Kaminska K., Owe-Larsson M., Cudnoch-Jedrzejewska A. Pharmacol. Rep. 2024, 76, 223-234.
[12] Caroli C., Baron G., Cappellucci G., Brighenti V., Della Vedova L., Fraulini F., Oliaro-Bosso S., Alessandrini A., Zambon A., Lusvardi G., Aldini G., Biagi M., Corsi L., Pellati F. Heliyon. 2024, 10, e30291. [13] Brighenti V., Venturelli A., Caroli C., Anceschi L., Gjikolaj M., Durante C., Pellati F. J. Pharm. Biomed. Anal. 2023, 234, 115547.

Natural products deriving from plants have always been recognised as an important source of novel therapeutic leads, both directly or as models for the design and semisynthesis of new drugs.
Due to the increased interest in this field, the development of different innovative techniques for the identification, isolation, characterization, and biological screening is fundamental. For this reason, it is important to discover and design new methodologies to acquire plant materials, keeping threatened and vulnerable species in mind, to scale up the supply of bioactive molecules and to identify and execute suitable high-throughput screening bioassays [1].

In light of the above, the aim of my PhD project was to firstly develop innovative extraction techniques for the pre-concentration and isolation of active principles from complex plant matrices using solid phase adsorption (SPA), with pharmaceutical, nutraceutical and cosmeceutical approaches. Layered double hydroxides of variable composition are biocompatible lamellar clays with great potential for the selective extraction of active compounds, also representing a powerful tool to enhance their chemical stability, bioactivity and bioavailability, thus indeed facilitating their use in multiple formulations.
Analytical techniques such HPLC-DAD-FLD, GC-MS, LC-MS and mono- and bi-dimensional NMR were used for the quantification, characterization and structural elucidation of the recovered and isolated secondary metabolites.
In case of low concentration of a product in the plant matrix, the chemical synthesis of the natural compounds and of their semisynthetic derivatives was performed, all by adhering to the green chemistry principles, by using non-toxic, non-polluting and affordable solvents and reagents.
The final goal consisted in the set-up of novel and alternative chemical assays by using hyphenated techniques for the preliminary biological investigation and the monitoring of the effect on enzymatic activity of the recovered natural and semisynthetic products.

Herein, we will focus on the last part of my PhD project, which involved the development and the validation of a preliminary screening of potential anti-inflammatory, neuroprotective, and anti-cancer agents by assessing the effects of glutathione peroxidase (GPx)-like catalytic activities of natural and semisynthetic compounds.

Glutathione peroxides (GPxs) are a family of oxidoreductases able to catalyse the reductions of various hydroperoxides (e.g., H₂O₂) to H₂O and/or the respective alcohols using glutathione (GSH) as the reducing agent, ensuring protection against oxidative stress [2]. Eight isoforms are known: isoforms 1-4 containing selenocysteine (Sec) in the active site, while isoforms 5-8 are featured by the presence of cysteine (Cys) [3]. Numerous studies have highlighted how a decrease in their activity can be linked with the pathogenesis of numerous acute and chronic inflammation-induced diseases, making GPx mimicking agents a possible therapeutic option [4,5,6,7].
In this context Iwaoka′s assay has been extensively exploited to record the mimicking GPxs effects of natural, semisynthetic and synthetic chemicals, especially for organoselenium and organotellerium compounds. The assay consists in the monitoring of the oxidation of 1,4-dithiothreitol to 4,5-dihydroxy-1,2-dithiane in the presence of H2O2 and a catalyst, using deuterated methanol as a solvent to successfully record, at different time, the 1H NMR spectra of the solution [8].
Although this methodology has undoubtedly great potential for the evaluation of the GPxs-like activity of numerous compounds, two main drawbacks must be pointed out. Firstly, the need of a versatile NMR instrumentation having reproducible settings, and then the necessity to use deuterated MeOH, and not cheaper solvents like deuterated chloroform, to obtain significant 1H NMR spectra. Thus, both factors could limit the usefulness of this test.
With the aim of trying to extend the applicability of the Iwaoka′s assay, two methodologies were developed and validated, using cheaper and widespread hyphenated techniques such as high-performance liquid chromatography with diode array detection (HPLC-DAD) and gas-chromatography coupled to mass spectrometry (GC-MS), granting easier tuning and calibration than the NMR spectrometer. At the same time, this allowed to widen the category of GPx-mimicking agents to be assessed by performing the test in low polarity and not deuterated solvents.

The first step consisted in the evaluation of the effectiveness of the alternative GC-based and HPLC-based methodologies to monitor the oxidation of 1,4-dithiothreitol to 4,5-dihydroxy-1,2-dithiane with H₂O₂ and Ebselen, a well-known GPx-mimicking agent, as a catalyst [9].
The assay′s reaction consisted in a slight modification of the original protocol for both the hyphenated analysis, and the progress of it was monitored by the quantification of the formation of 4,5-dihydroxy-1,2-dithiane coupled to the determination of the residual content of 1,4-dithiothreitol, recorded by sampling the reaction mixtures at fixed times ( 0 min - 30 min - 1 h - 2 h - 3 h - 4 h - 24 h) . The original protocol with 1H-NMR spectroscopy was also performed as reference [8].
No significant difference between the average concentration values of the three methodologies was recorded, thus confirming the concrete possibility of employing this revised version of the Iwaoka′s assay for the characterization of GPx-like effects of known and novel natural, semisynthetic and purely synthetic compounds, which consisted in the second part of this project [10].

A panel of natural and semisynthetic molecules comprising coumarins, flavonoids, ferulic acid, cannabinoids and their derivatives, were tested for the new revised version of the assay, providing an overall look on their activity as GPx mimicking agents.

In conclusion, a novel, alternative and rapid chemical assays to preliminary evaluate putative pharmacological activities was set-up and developed, and then used for the testing of a series of natural compounds and their derivatives.

References
[1] Chaachouay N., Zidane L. Drugs Drug Candidates. 2024, 3: 184-207.
[2] Brigelius-Flohe R., Maiorino M. Biochim. Biophys. Acta. 2013, 1830: 3289-3303.
[3] Ursini F., Maiorino M., Brigelius-Flohe R., Aumann K.D., Roveri A., Schomburg D., Flohe L. Methods Enzymol. 1995, 252: 38-53.
[4] De Haan J.B., Cooper M.E. Front. Bio. 2011, S3: 709-729.
[5] Zhang M.L., Wu H.T., Chen W.J., Xu Y., Ye Q.Q., Shen J.X., Liu J. J. Transl. Med. 2020, 18: 247.
[6] Sharma G., Shin E.J., Sharma N., Nah S.Y., Mai H.N., Nguyen B.T., Jeong J.H., Lei X.G., Kim H.C. Food Chem. Toxicol. 2021, 148: 111945.
[7] Hariharan S., Dharmaraj S., Inflammopharmacol. 2020, 28: 667-695.
[8] Kumakura F., Mishra B., Priyadarsini K.I., Iwaoka M. Eur. J. Org. Chem. 2010, 440-445.
[9] Santi C., Scimmi C., Sancineto, L. Molecules. 2021, 26: 4230.
[10] Fiorito S., Epifano F., Palumbo L., Collevecchio, C., Genovese, S. J. Pharm. Biomed. Anal. 2022, 212: 114652.

In a panorama focused on eco-sustainability and minimizing the impact of agri-food waste on a national and global scale, this doctoral project is aimed at the conversion of agricultural and food chain waste biomasses into functional products with relevant nutraceutical properties and, therefore, into functional products for personal care [1]. This purpose will be achieved through an analytical, functional, and high-tech approach, which fits within the mobilization of industry for a clean and circular economy in the European green deal, that covers different research areas of the NRP 2021-2027 and is in accordance with the Green Chemistry paths defined by the National Strategy for Intelligent Specialization (SNSI) 2014-2020, referring to the transformation of biomasses deriving from agriculture, food waste, organic waste, algae and micro-organisms into chemical substances and products [2-4].

Specifically, the research program is integrated into a scientific collaboration between DISFARM research laboratories and the company Distillerie Bonollo - nutraceutical division. The company, with its Bonollo Unique System, which is the result of a combination of agricultural experience and the most innovative distillation systems, aims at enhancing the pomace at every stage of processing, from harvesting to controlled storage of the raw material. After distillation, the fermented pomace represents the principal waste biomass of the company and Bonollo daily strives to valorize it after obtaining the grappa, separating the pomace with an industrial facility already in use into grape seeds, which represent 10% of the total mass and are reused for the oil industry, and pomace skins, which represent 90% of the total mass, as a sustainable fuel to produce steam for distillation.

The research design consists into the following phases:
1. Biomass identification: comprehensive assessment of discarded products′ qualitative and quantitative composition, guided by technical, scientific, and economic considerations.
2. Development of green technologies: collaborative efforts to develop eco-friendly methods for isolating active fractions, integrating academia′s expertise with industry′s scale-up capabilities.
3. Metabolomic studies: based on the use of highly advanced analytical instrumentation are aimed at the definition of a precise qualitative and quantitative composition of the biomasses highlighting the content of chemical compounds of nutritional/nutraceutical interest.
4. Functional studies: evaluation of biomasses′ activity through the application of different in vitro models to assess their potential in application in various health concerns.
5. Definition of the activity and the mechanism of action: based on the evidence obtained in the previous steps, the ultimate objective is to identify within the complex mixture the molecule or class of compounds responsible for the in vivo activity (hit compound).

According to this research plan and with the collaboration of Bonollo, white grape skins (WSG), that represent the 90% of the total discarded biomass at Bonollo, are investigated to evaluate the possibility of the development of a new sustainable product with potential health benefits.
White grape skins (WSG) were firstly pulverized in liquid nitrogen, and the resulting powder was extracted using different techniques, a hydro-alcoholic solution, and a method with water. The hydro-alcoholic extract (HAE) underwent further treatment with Sepabeads SP207 resin to enhance its polyphenol content. The polyphenolic profiles of both extracts were characterized using hyphenated techniques. Qualitative analysis with LC-HRMS putatively identified 51 compounds in HAE and 39 compounds in the aqueous extract (AE). Subsequent quantitative analysis and determination of total polyphenol content by HPLC-PDA confirmed a richer profile in HAE compared to AE, while spectroscopic analyses were also conducted to further characterize in terms of tannins, anthocyanins, and radical scavenging capacity the two extracts [5]. Furthermore, HAE and AE were evaluated on two different cell lines with gene reporters for the transcription factors Nrf2 and NfkB to assess their antioxidant and anti-inflammatory activity, respectively. Interestingly, results showed that AE, despite its less abundant polyphenolic composition, exhibited a stronger anti-inflammatory effect (from 1 μg/mL) compared to HAE, entirely independent of the NRF2 pathway, as AE was unable to modulate NRF2 activity. The industrial scaling up for the production of a commercial product was performed from AE product due to its simpler and more environmentally friendly extraction method and its potent anti-inflammatory activity. This industrial extract, named VITUVA^(R), underwent all analyses and tests conducted on the parent AE extract to assess if AE properties were maintained after the scaling-up process. Further studies will focus on the polyphenolic composition of AE and VITUVA^(R) to clarify the mechanism of action. Outcomes obtained demonstrate that WGS can effectively be considered a valuable source of polyphenols for use as nutraceuticals to prevent/treat various inflammatory conditions, emphasizing the importance of circular economy principles and the valorization of natural by-products.
According to this scientific approach, grape seeds have already been identified as a biomass rich in bioactive compounds. Specifically, Bonollo has developed a green production technology based on water extraction and tangential filtration currently applied to grape seeds for the sustainable extraction of bioactive components. This technique was applied for the extraction of proanthocyanidins (PACs), resulting in a registered and commercially available powder (ECOVITIS^(R)) [6].
After the characterization of chemical composition and biological activity in in vitro models [7], ECOVITIS^(R) was evaluated in an in vivo metabolomics study and in a qualitative proteomics platform to understand the bioavailability of ECOVITIS phytochemicals and to elucidate the mechanism of action related to the protective effects of grape seed procyanidins on health.
Proanthocyanidins (PACs) are plant-derived secondary metabolites consisting of flavan-3-ol oligomers (n = 2-4) and polymers (n ≥ 5), composed of catechin/epicatechin and gallocatechin/epigallocatechin units. They exhibit various health benefits, including gastrointestinal defense against pathogens, reduction of inflammation and oxidative stress in the gastric and colonic mucosa, and potential prevention of colorectal cancer. Systemically, PACs act as neuroprotectants and have reported anti-obesity, anti-carcinogenic, anti-diabetic, and anti-inflammatory effects. While monomeric flavan-3-ols and dimeric PACs are partially absorbed in the GI tract, oligomeric and polymeric PACs are not bioavailable orally but interact with colonic microbiota, yielding metabolites such as phenyl-γ-valerolactones and hydroxyphenyl-valeric acids. These metabolites may undergo phase II metabolism in hepatocytes, becoming conjugated derivatives detectable in urine, potentially influencing physiological scenarios. The aim of this first part of the project was to elucidate the comprehensive pathway underlying the health-protecting effects of proanthocyanidins (PACs). Using ECOVITIS^(R) as a standardized grape seed extract (GSE) rich in oligo-polymeric procyanidins, the metabolic conversion of PACs has confirmed in both the gastrointestinal tract and systemically through in vitro and in vivo approaches, and the main PACs catabolite identified was 5-(3′,4′-dihydroxyphenyl)-γ-valerolactone (VL), produced by colonic microbiota. Also, the bioavailability of VL was demonstrated by detecting corresponding phase II metabolites (sulfate and glucuronide) in urine samples from GSE-treated volunteers. Cell biology evidence, supported by proteomic analyses, suggested that VL, after a redox transition to quinone, acts as an electrophilic activator of the nuclear factor E2-related factor 2 (Nrf2). PACs thus exhibit a health-protective effect by activating cell-protective, anti-stress mechanisms, primarily involving the Nrf2 pathway, fitting into the broader framework of para-hormetic effects, ultimately contributing to the preservation of redox steady-state and cellular homeostasis [8].
Despite recent progress in understanding the mechanisms of action and bioavailability of dimeric proanthocyanidins (PACs) and their intestinal metabolism resulting mainly in 5-(3′,4′-dihydroxyphenyl)-γ-valerolactone (VL), further research is needed to fully elucidate these processes, as the pre-ADMET properties of the catabolite which have not been previously evaluated due to the absence of a commercially available standard [9].
Additionally, as PACs and flavonoids are increasingly recognized as potent nutraceuticals, hence investigating potential herb-drug interactions (HDI) when taken with prescription medications becomes crucial. This study integrated in the doctoral project aims to address this gap by exploring the absorption and efflux mechanisms, metabolic stability, cytotoxicity, and drug-like properties of VL.
Caco2 and Wt-MDCK in vitro models for the evaluation of the absorption have shown that VL undergoes a rapid metabolism and absorption, mainly as its sulfate phase II conjugate (VLS), which enters the systemic circulation and minimally activates the BCRP efflux transporter. In human S9 fraction, VL is subjected to metabolization into its glucuronic phase II conjugates (VLG1 and VLG2) with a total conversion rate of 80% and a half-life of 8.72 minutes. In human microsome fraction, VL is metabolized at a slower rate (half-life of 23.08 minutes), indicating that oxidative metabolism of VL is of secondary importance. Investigation into PXR activation and CYP3A4/CYP1A2 inhibition did not yield significant results, suggesting that VL has no risk of herb-drug interaction (HDI) when PACs and flavonoids-rich supplements are co-administered with prescription drugs. In conclusion, the analytical, functional, and high-tech approach applied in this doctoral project proved to be effective in highlighting the potential value of agri-food by-products, in developing functional ingredients and nutraceuticals, and promoting circular economy principles and sustainable utilization of resources in an academic and industrial environment.

References
[1] Uso efficiente delle risorse nelle imprese vitivinicole. Carla Creo, Giuliana Ansanelli, Patrizia Buttol, Cristian Chiavetta, Sara Cortesi, Laura Cutaia, Paola Nobili, Paola Sposato. 2018 ENEA [2] https://ec.europa.eu/info/strategy/priorities-2019-2024/european-green-deal_it
[3] A Multiannual Financial Framework (MFF) for the Climate (Project no.: 81230639 / 17.9045.0-002.36)
[4] https://www.agenziacoesione.gov.it/s3-smart-specialisation-strategy/strategia-nazionale-di-specializzazione-intelligente/
[5] Della Vedova L, Ferrario G, Gado F, Altomare A, Carini M, Morazzoni P, Aldini G, Baron G. Liquid Chromatography-High-Resolution Mass Spectrometry (LC-HRMS) Profiling of Commercial Enocianina and Evaluation of Their Antioxidant and Anti-Inflammatory Activity. Antioxidants (Basel). 2022 Jun 16;11(6):1187. doi: 10.3390/antiox11061187. PMID: 35740083; PMCID: PMC9231191.
[6] Morazzoni P, Vanzani P, Santinello S, Gucciardi A, Zennaro L, Miotto G, Ursini F. Grape Seeds Proanthocyanidins: Advanced Technological Preparation and Analytical Characterization. Antioxidants (Basel). 2021 Mar 9;10(3):418. doi: 10.3390/antiox10030418. PMID: 33803398; PMCID: PMC8001487.
[7] A New Grape Seed Extract Pharma Standard Supplement (ECOVITISTM) Prevents and Controls Borderline Hypertension and Endothelial Dysfunction MR Cesarone et al. Medical & Clinical Research, ISSN: 2577 - 8005
[8] G. Baron, A. Altomare, L. Della Vedova, F. Gado, O. Quagliano, S. Casati, N. Tosi, L. Bresciani, D. Del Rio, G. Roda, A. D′Amato, C. Lammi, A. Macorano, S. Vittorio, G. Vistoli, L. Fumagalli, M. Carini, A. Leone, M. Marino, C. Del Bo′, G. Miotto, F. Ursini, P. Morazzoni, G. Aldini, Unraveling the parahormetic mechanism underlying the health-protecting effects of grapeseed procyanidins, Redox Biology, Volume 69, 2024, 102981, ISSN 2213-2317.
[9] Phenyl-γ-valerolactones and phenylvaleric acids, the main colonic metabolites of flavan-3-ols: synthesis, analysis, bioavailability, and bioactivity; Pedro Mena, Letizia Bresciani, Nicoletta Brindani, Iziar A. Ludwig, Gema Pereira-Caro, Donato Angelino, Rafael Llorach, Luca Calani, Furio Brighenti, Michael N. Clifford, Chris I. R. Gill, Alan Crozier, Claudio Curti and Daniele Del Rio.

KEYWORDS. cross-linking mass spectrometry; epitope/paratope mapping; data-driven integrative modeling; drug discovery & development, biotherapeutics

ABSTRACT. Elucidating antibody-antigen complexes at the atomic level is of utmost interest for understanding immune responses and designing better therapies [1]. Cross-linking mass spectrometry (XL-MS) has emerged as a powerful tool for mapping protein-protein interactions, suggesting valuable structural insights [2]. However, the use of XL-MS studies to enable epitope/paratope mapping of antibody-antigen complexes is still limited up to now. XL-MS data can be used to drive integrative modeling of antibody-antigen complexes, where cross-links information serves as distance restraints for the precise determination of binding interfaces. In this approach, XL-MS data are employed to identify connections between binding interfaces of the antibody and the antigen, thus informing molecular modeling. Current literature provides minimal input about the impact of XL-MS data on the integrative modeling of antibody-antigen complexes. Here we applied XL-MS to retrieve information about binding interfaces of three antibody-antigen complexes. We leveraged XL-MS data to perform integrative modeling using HADDOCK [3] (active-passive residues and distance restraints strategies) and AlphaLink2 [4, 5]. We then compared these three approaches with initial predictions of investigated antibody-antigen complexes by AlphaFold Multimer [6, 7]. This work emphasizes the importance of cross-linking data in resolving conformational dynamics of antibody-antigen complexes, ultimately enhancing the design of better protein therapeutics and vaccines.

References
[1] Sela-Culang, I.; Kunik, V.; Ofran, Y. The Structural Basis of Antibody-Antigen Recognition. Front. Immunol. 2013, 4 (OCT), 1-13. https://doi.org/10.3389/fimmu.2013.00302.
[2] Piersimoni, L.; Kastritis, P. L.; Arlt, C.; Sinz, A. Cross-Linking Mass Spectrometry for Investigating Protein Conformations and Protein-Protein Interactions-A Method for All Seasons. Chem. Rev. 2022, 122 (8), 7500-7531. https://doi.org/10.1021/acs.chemrev.1c00786.
[3] Van Zundert, G. C. P.; Rodrigues, J. P. G. L. M.; Trellet, M.; Schmitz, C.; Kastritis, P. L.; Karaca, E.; Melquiond, A. S. J.; Van Dijk, M.; De Vries, S. J.; Bonvin, A. M. J. J. The HADDOCK2.2 Web Server: User-Friendly Integrative Modeling of Biomolecular Complexes. J. Mol. Biol. 2016, 428 (4), 720-725.
[4] Stahl, K.; Graziadei, A.; Dau, T.; Brock, O.; Rappsilber, J. Protein Structure Prediction with In-Cell Photo-Crosslinking Mass Spectrometry and Deep Learning. Nat. Biotechnol. 2023, 1-24. https://doi.org/10.1038/s41587-023-01704-z.
[5] Stahl, K.; Brock, O.; Rappsilber, J.; Der, S.-M.; Mensch, S. Modelling Protein Complexes with Crosslinking Mass Spectrometry and Deep Learning. bioRxiv 2023, 1-13. https://doi.org/https://doi.org/10.1101/2023.06.07.544059.
[6] Jumper, J.; Evans, R.; Pritzel, A.; Green, T.; Figurnov, M.; Ronneberger, O.; Tunyasuvunakool, K.; Bates, R.; ��dek, A.; Potapenko, A.; Bridgland, A.; Meyer, C.; Kohl, S. A. A.; Ballard, A. J.; Cowie, A.; Romera-Paredes, B.; Nikolov, S.; Jain, R.; Adler, J.; Back, T.; Petersen, S.; Reiman, D.; Clancy, E.; Zielinski, M.; Steinegger, M.; Pacholska, M.; Berghammer, T.; Bodenstein, S.; Silver, D.; Vinyals, O.; Senior, A. W.; Kavukcuoglu, K.; Kohli, P.; Hassabis, D. Highly Accurate Protein Structure Prediction with AlphaFold. Nature 2021, 596 (7873), 583-589. https://doi.org/10.1038/s41586-021-03819-2.
[7] Evans, R.; O�Neill, M.; Antropova, N.; Senior, A.; Green, T.; Zidek, A.; Bates, R.; Blackwell, S.; Yim, J.; Ronneberger, O.; Bodenstein, S.; Zielinkski, M.; Brigland, A.; Potapenko, A.; Cowie, A.; Tunyasuvunakool, K.; Jain, R.; Clancy, E.; Kohli, P.; Jumper, J. and; Hassabis, D. Protein Complex Prediction with AlphaFold-Multimer. bioRxiv 2021, 1-25. https://doi.org/https://doi.org/10.1101/2021.10.04.463034.https://doi.org/10.1016/j.jmb.2015.09.014.