Final Report Summary - HTP-GLYCOMET (Methods for high-throughput glycoproteomic analysis)
Executive summary
The main aim of the HTP-GlycoMet proposal was to develop the analytical technologies for the high-throughput analysis of glycosylation of individual glycoproteins isolated for body fluids and cell membranes (see also graphical abstract that is submitted separately). In the frame the work package 1 (WP1), monoclonal antibodies against both blood plasma and membrane proteins, and methods for isolation of native forms of these molecules were developed. The optimized method for antibody immobilization on monolithic columns was optimized, and subsequent isolation of corresponding antigens in their native form from plasma (as body fluid) and from cell membranes was performed. For membrane proteins, methods for production of glycoprotein amounts that are sufficient for subsequent glycan analyses were optimized. For this sake, corresponding immobilization techniques and miniaturized monolithic columns and new purification methods for isolation of biologically active proteins from large volume of medium were developed. Monolithic supports with immobilized monoclonal antibodies against above mentioned individual proteins were mounted in a 96-well ELISA plate and were applied for isolation of single antigens. This high-throughput method was validated and also adapted for application of laboratory robotics and for possible use in routine analyses. In the next step, and in the frame of the WP2, oligosaccharides from isolated glycosylated proteins were cleved by corresponding glycosidases. For development of high-throughput analysis, glycosidases were immobilized on monolithic supports, and a method for «in-flow» enzymatic cleavage was developed. In order to understand the role of individual protein glycosylation pattern during biological processes and in health and disease, methods for both glycoprotein analysis and analysis of the structures of glycan parts based on ultra high performace liquid chromatography (UPLC) and mass spectrometry (MS and MS/MS) were applied. By use of these analytical methods, changes of protein glycosylation during pathological processes such as inflamation and/or viral infection were determined. Capillary gel electrophoresis with laser-induced fluorescence (CGE-LIF) was introduced as a complementary high-throughput method for the analysis of glycan structures, and both UPLC and xCGE-LC databases for glycan assigment were developed and collection of necessary data was started. In the next step (WP3) glycoproteins were identified by use of electrosprey ionisation-LC-MS/MS (ESI-LC-MS/MS), and their potential glycosylation sides were assigned. Membrane glycoproteins RAE-1 and MULT-1 were isolated from both citomegalovirus infected and non-infected cells. In order to isolate protein amounts that will be sufficient for analysis of glycan structure, production of these glycoproteins was up-scaled, and processing of large volumes of protein-containing supernatant by use of small monolithic columns was optimized. However, the analysis of MULT1 as well as RAE-1 glycans could still be performed only on an analytical scale due to low amounts of these glycoproteins obtained during isolation. The foreseen high-throughput analysis of these membrane glycoproteins was postponed due to amounts of proteins obtained during isolation that were adequate only for the analytical-experiments, and further scaling-up protocols for production of larger amounts of these proteins are in development. Therefore, optimization of the high-throughput purification protocol by use of UHPLC coupled to MALDI-TOF MS glycan analysis was done by use of the model glycoprotein yeast invertase, which is localized on the cell surface of the yeast Saccharomyces cerevisiae. Large amounts of the yeast can be easily fermented, and invertase can be easily isolated in amounts that are sufficient for large-scale analyses. In the next step, a chromatographic procedure for separation of invertase glycoforms on anion-exchange monolithic supports was developed. The ESI-LC-MS/MS analysis of this model glycoprotein that contains more than 50% oligosaccharides by weight showed thirteen N-glycosylation sites in the amino acid sequence. The chromatographic procedure for separation of invertase glycoforms was that carry different negative charges densities from a natural matrix was developed. These invertase glycoforms were used in the course of the project for development of glycan hydrolyses protocols based on glycosydases immobilized at monolithic supports. This protocol was also applied for high-throughput analysis of glycan structures of other glycoproteins. The aim of the WP4 is to benefit the transfer of the knowledge and technologies between SME and academic partners, and for this sake the exchange of both young and experienced researchers (in form of secondments) took place from the industry to the academia, as well as from the academia to industry (WP5). The foreseen exchange in both directions was completely realized. In the frame of the WP6 four experienced researcher with specific skills were recruited by Genos, UNI-RI, MEDRI and MPI. Their work was successfully performed, and yielded in numerous oral presentations and posters at different scientific meetings, as well in a significant number of publications. Special Issue of the journal Electrophoresis dedicated to monolithic supports was recently edited by the Coordinator of the HTP-GlycoMet project. This issue contains six papers that present our work performed during 4 years of project realization. Different trainings in scientific and technical methods (WP7) applied for isolation of proteins from human and animal plasma, as well as hydrophobic membrane proteins were organized during three scientific meetings (Monolith Summer Schools and Monolith Symposia 2014 and 2015 and Glyco23 International Symposium), and researchers that were recruited by UNIRI and Genos in the frame of this project were participating and also teaching there. Workshops dedicated to the the teaching/training skills, presentation, business and both project writing and project management, as well as career planning (WP8) were organized by MEDRI with participation of a large number of young researchers from both academia and industry. Our collaborative project was presented to the publics by different interview and presentations given by the project coordinator and group leaders (WP9), and project management administrative support and foreseen meetings were organized and reports were submitted on time (WP10). In conclusion, the foreseen work in WP1-WP10 was realized. Due to difficulties in up-scaling of production of membrane glycoproteins from non-infected and virally infected cultured cells, we achieved to assign oligosaccharide structures to the peaks in both UPLC and CGE, but until the end of the project time, the analyses could be performed only on an analytical scale. The method for high-throughput analysis of glycans is developed, and the development of the protocol for further scaling up of membrane proteins is ongoing.
A summary description of project context and objectives
When started, the main aim of the HTP-GlycoMet proposal was to develop technologies for the high-throughput analysis of glycosylation of individual proteins from body fluids and cell membranes as one of most common and most important posttranslation modifications. This knowledge will be applied for the understanding of the machanisms of both immunity and infections. Glycan analysis is extremely demanding from both technological and conceptual aspect and large-scale studies of glycosylation of individual proteins are still a big challenge, and before the start of this common proposal, only one large scale study was performed by the partners of this proposal. Our strategic concept was to combine the complementary expertise in the (i) production of specialised monolithic chromatographic tools for high-throughput fractionation of complex biological fluids (BIASEP), (ii) purification of proteins from both body fluids and membrane proteins by use of high-throughput chromatographic techniques (UNI-RI), (iii) high-throughput glycomic analysis by use of high-performance liquid chromatography (Genos) and multiplexed capillary gel electrophoresis (MPI) with following MS and MS/MS analysis of separated fractions and (iv) expertise in the field of viral immunology and production of specific monoclonal and polyclonal antibodies (MEDRI) in order to achieve significant progress in this direction. All our partners are recognized leaders who already made significant progress beyond the state of the art in their respective fields. Additionally, through the HTP-GlycoMet project secondments on all levels (MER, ER, ESR) were realized in order to capitalize on synergistic effects of this interdisciplinary and transnational collaboration. In addition to the generation of new knowledge and the development of new innovative technologies, and according to one of main goals of this collaboration, significant transfer of know-how between academic and industrial partners was achieved. As expected, our SME partners already started developing new lines of products and services through the HTP-GlycoMet programme, but also through future collaboration with HTP-GlycoMet partners beyond the lifetime of this project. The main objective of the HTP-GlycoMet initiative, the development of robust, reliable high-throughput methods for glycoproteomics during four project years was achieved. When the HTP-GlycoMet initiative started four years ago, the predominating glycomics approach was the analysis of glycans from all glycoproteins as a single pool of structures. We immediatelly started with much more advanced glycoproteomics approach, and our focus was the analysis of the glycosylation of individual glycoproteins that were isolated from large number of samples by use of high-throughput techniques. While glycomics can identify some general glycosylation features like deficient glycan branching, increased fucosylation, as well as changes in sialic acid as terminal sugar, these variations can have very different consequences on individual glycoproteins. For example, increased fucosylation of IgG would decrease antibody-dependent cellular cytotoxicity (ADCC) and consequential tissue damage, while increased fucosylation of other plasma proteins is generally associated with inflammation. Sialylated Fc domains of the immunoglobulins G (IgG) have also antiinflammatory properties, and these changes are essential for the mode of action of these glycoproteins. From the structural point of view, the analysis of individual glycoproteins is less complex than the analysis of the complete tissue glycome, but it requires higher sensitivity of applied analytical systems because the majority of individual glycoproteins are present in very low concentrations. The second major obstacle is the isolation process of individual low abundance glycoproteins from complex mixtures. The combined effects of low amount of available glycoproteins, especially these ones that can be produced only by use of cultured cells (such as membrane proteins) frequently combined with their degradation or inactivaton (or both) during the isolation process, and the high-throughput approach generally hampers the quality control of each sample. Therefore, innovative methods, fast, robust and reliable, are needed to minimise inaccuracies introduced by protein purification during sample preparation. We addressed these two main obstacles and also solved these problems during the project realization by combining complementary expertise of our partners in reaching the project objectives: BIA-SEP (proprietary monolithic technology), MEDRI (production of both membrane proteins by use cultured cells before and after viral infection, and monoclonal antibodies against single proteins), UNI-RI (isolation of individual glycoproteins), Genos (high-throughput glycomic analysis via HPLC/UPLC based method) and MPI (high-throughput glycomic analysis via a system based on multiplexed capillary gel electrophoresis with laser induced fluorescence detection (xCGE-LIF)). These newly developed methods are further implemented to study important aspects of protein glycosylation during viral infection and other pathological processes. The planned experiments were organized in three work packages (WP1-WP3) as described below. To reach these objectives, the partners were in the first step focused on the implementation of existing expertise, already established protocols and available material to reach the proof of principle level, and this work was successfully finalized in the first year of the realization of the project. The next step was the integration of the selected technologies in order to develop a novel approach, an actual prerequisite for the high-throughput implementation. For example, MEDRI has extensive expertise in the analysis and interpretation of molecular mechanisms exploited by cytomegaloviruses in order to subvert the immune system. The current state of the art in the field shows strong evidence that infection leads to change of glycosylation or to misglycosylation of the cellular proteins. However, this is mainly achieved through the site mutations of the amino acids as putative glycosylation sites or by using drugs that prevent different posttranslational modifications. The changes in the total protein glycosylation were monitored, however, due to analyses of complete glycome (and not singular, isolated proteins) little information was available about the exact cellular processes involved. Thus, new protocols had to be developed in order to analyse individual sugar differences on individual isolated glycoproteins, especially their changes under pathological conditions. Likewise, BIA-SEP had a strong base in the development of different high-throughput separation media, and first also experiences in production of method for fast handling of large number of samples. However, the production of affordable customized chromatographic devices based on the antibody of interest and especially in a format allowing a significant number of different samples/conditions was still a challenge. After the proof of principle level in first step, in the next step of project realization, several successful protocols for production of chromatographic media based on specific monoclonal antibodies were devloped. These media were used for glycoprotein analyses starting from crude samples containing sometimes target protein in very low concentrations, and ending in sophisticated glycoanalysis from individual proteins. Comparing glycoproteins from samples of patients with control samples, or samples of non-infected cells with cells after infection with citomegalovirus, changes of protein glycosylation under different pathological pressures were determined. New methodologies that were developed by the partners were presented as a platform for various applications, and fundaments for new products and protocols were established.
List of partners:
1. UNI-RI BIOTECH - http://www.biotech.uniri.hr/en/
2. BIA-SEP - http://www.biaseparations.com/
3. Genos - http://genos.hr/en/
4. MPI - http://www.mpi-magdeburg.mpg.de/2316/en
5. MEDRI - http://capri.com.hr/
Work packages, objectives, deliverables and milestones
WP 1. New tools for high-throughput isolation of individual glycoproteins
O.1.1. To acquire monoclonal antibodies with adequate characteristics for the preparation of new affinity materials
O.1.2. To isolate individual glycoproteins from complex mixtures in a high-throughput setting
D.1.1. New affinity monolithic 96-well plates (M48)
D.1.2. Monoclonal antibodies for affinity purification (M24)
D.1.3. Methods for purification of plasma proteins (M48)
D.1.4. Methods for purification of membrane proteins (M48)
M.1.1. Monoclonal antibodies (M12)
M.1.2. Tools for purification of glycoproteins (M24)
WP 2. Development of high-tFkor
hroughput glycoproteomic analysis
O.2.1. To understand the role of individual protein glycosylation pattern in health and disease
D.2.1. New UPLC methods (M48)
D.2.2. UPLC glycan assignment database (M48)
D.2.3. Structural analysis by MS and MS/MS (M48)
D.2.4. New CGE-LIF methods (M48)
D.2.5. xCGE-LIF glycan assignment database (M48)
M.2.1. Methods for glycoprotein analysis (M24)
WP 3. Analysis of protein glycosylation during viral infection
O.3.1. To explore whether viruses interfere with glycosylation patterns of the immune system receptors and ligands
D.3.1. Lysates of transfectant cell lines non-infected or infected with various recombinant MCMVs (M48)
D.3.2. Production of additional monoclonal antibodies (to viral and cellular targets) for affinity purification (M24)
D.3.3. Immobilization of antibodies and isolation of target glycoproteins (M36)
D.3.4. Comparison of glycosylation profile of immunologically relevant molecules before and after MCMV infection (M48)
D.3.5. Methods for the identification of interaction partners using fusion proteins (M48)
M.3.1. Glycoproteins from virally infected cells (M36)
WP 4. Secondments from industry to academia
O.4.1. To benefit from the transfer of specific, interdisciplinary knowledge and technologies between SME and academic partners (as described in detail in WP1-3), with the particular emphasis on knowledge available at academic institutions that can help SMEs increase their competitiveness on global markets
D.4.1. From BIA-SEP: MER to UNI-RI for 2 months (M24)
D.4.2. From BIA-SEP: two ERs to MEDRI; 4 months each (M24, 48)
D.4.3. From BIA-SEP: ESR to UNI-RI for 5 months (M24, 48)
D.4.4. From BIA-SEP: ESR to MEDRI for 3 months (M36)
D.4.5. From Genos: MER to MEDRI for 2 months (M12, 24, 36, 48)
D.4.6. From Genos: ER to MPI for 3 months (M12)
D.4.7. From Genos: ESR to MEDRI for 2 x 3 months (M36)
D.4.8. From Genos: ER to UNI-RI for 3 months (M36)
D.4.9. From Genos: ESR to MPI for 12 months (M24, 36)
WP 5. Secondments from academia to industry
O.5.1. To benefit from the transfer of specific, interdisciplinary knowledge and technologies between SME and academic partners (as described in detail in WP1-3), with the particular emphasis on knowledge available at SMEs that can help academic researchers increase their competitiveness and improve career prospects
D.5.1. From UNI-RI: MER to BIA-SEP and Genos; 18mts (M12, 24, 36, 48)
D.5.2. From UNI-RI: MER to BIA-SEP for 2 x 3 months and Genos for 3 months (M24, 36, 48)
D.5.3. From MEDRI: MER to BIA-SEP and Genos; 4 mts (M12, 24, 36, 48)
D.5.4. From MEDRI: MER to BIA-SEP and Genos; 2 x 3 months (M12, 24, 36, 48)
D.5.5. From MEDRI: ER to Genos for 6 months (M36)
D.5.6. From MEDRI: ER to BIA-SEP for 3 months (M24)
D.5.7. From MEDRI: ER/Project manager to BIA-SEP and Genos for 4 months (M36, 48)
D.5.8. From MPI: MER to Genos for 2 months (M12, 24, 36, 48)
D.5.9. From MPI: ER to Genos for 4 months (M12, 24, 36, 48)
D.5.10. From MPI: ESR to Genos for 12 months (M24, 36)
WP 6. Recruitment
O.6.1. To enhance the project team through recruitment of experienced researchers with specific skills required to contribute to the accomplishment of the interdisciplinary project objectives
D.6.1. Recruitment of ER for Genos (24 months) (M36)
D.6.2. Recruitment of ER for UNI-RI (24 months) (M36)
D.6.3. Recruitment of ER for MEDRI (24 months) (M36)
D.6.4. Recruitment of ER for MPI (24 months) (M24)
WP 7. Training in scientific and technical methods
O.7.1. To enhance the scientific and technical skills of recruited and seconded fellows through day-today transfer of knowledge in laboratories and specialised theoretical and practical workshops
D.7.1. Completed training in general methods (M48)
D.7.2. Workshop on “CIM technology” (M12)
D.7.3. Workshop on high throughput affinity purification (M48)
WP 8. Training in complementary skills
O.8.1. To enhance the scientific competitiveness and career prospects of recruited and seconded fellows through training in a variety of complementary skills (presentation skills, career planning)
O.8.2. To improve the leadership, managerial and business skills of recruited and seconded fellows through training in a variety of complementary skills (project management, entrepreneurship)
D.8.1. Teaching/ Training skills (M48)
D.8.2. Presentation skills (M48)
D.8.3. Workshop on business, project proposals writing and project management (M36)
D.8.4. Career planning workshop (M48)
WP 9. Dissemination and outreach
O.9.1. To promote the results of the project in the scientific community
O.9.2. To increase the visibility of the project and promote its benefits to policy makers and the general public
D.9.1. Dissemination/outreach activities (M48)
D.9.2. Workshop on “High-throughput glycoproteomics” (M24)
WP 10. Project management
O.10.1. To establish an efficient project managerial system and administrative support
O.10.2. To monitor the overall financial, administrative, legal and ethical aspects and progress of the project, including contracting with REA
D.10.1. Kick-off Meeting and establishment of the project administrative and managerial structure
D.10.2. Consortium meetings (M10, 22, 34, 46)
D.10.3. Progress, midterm, periodic and final reports (M12, 22, 24, 36, 48)
D.10.4. MidTerm Review (M22)
A description of the main science and technology results
The main objective of the HTP-GlycoMet initiative was the development of robust and reliable high-throughput methods for glycoproteomics. In the frame of WP1, BIA-SEP has optimized the immobilization procedure of antibodies onto monolithic chromatographic support (CIM® monoliths). The immobilisation protocol was mostly developed using anti-fibrinogen antibodies (@FIB) - murine IgG monoclonal antibodies, specific for human fibrinogen, a secretory glycoprotein that is synthesized in the liver and has a key role in blood clotting. Glycosylation of fibrinogen is not very well understood and to our knowledge it has been studied only on commercial fibrinogen samples. To meet the demand of high-throughput (HTP) glycomics studies, or glycan analysis of large numbers of samples, the methods used for analysis have to be fast, robust and reproducible. The use of polymethacrylate monolithic columns as a chromatographic support enables flowrate-independent binding capacity and resolution for large biomolecules due to the convective nature of the flow of the mobile phase. Convective Interaction Media (CIM®) monolithic columns have been commercialised by BIA-SEP and have already proven to be an appropriate medium for affinity chromatography, but to our knowledge there was only one previous report by three members of our team (BIA-SEP, Genos and UniRi) combining immunoaffinity sorbents with chromatographic monoliths in a 96-well format before the start of HTP-Glycomet project.
BIA-SEP's first goal was to prepare the immunoaffinity CIM monolith with the highest possible binding capacity for human plasma FIB. Preliminary experiments of mAb immobilisation on monoliths employing non-oriented immobilisation (coupling the protein through its free amino or thiol groups) resulted in binding capacities for fibrinogen below 1 mg/mL (results not shown), which was not enough for the intended application. Therefore, we focused on oriented mAb immobilisation, where the Fab region of the Ab, responsible for the interaction with the antigen, is positioned towards the lumen of the monolithic pore and is available for interaction with its counterpart. Covalent cross linking of the Ab with a protein A chromatographic support or binding of the antibody to a hydrazide support through oxidized sugar moieties are two typical approaches found in literature. We have optimised both coupling protocols for the protein immobilisation onto monolithic analytical column with column volume of 0.1 mL (CIMacTM).
Protein A crosslinking approach requires a protein A-specific mAb. The mAb of our choice was murine IgG2a, which interacts with protein A. High coverage of the monolith surface area with @FIB enabled lower exposure of non-occupied protein A molecules, present on the monolithic surface. No @FIB was detected in the wash and elution fractions after the crosslinking reaction with DMSI and DMPI, proving successful fixation of @FIB to the monolithic support. The dynamic binding capacity (DBC) for pure FIB on the CIMac pA-@FIB column was 4.6 mg/mL of support, that was high enough for isolation of FIB from human plasma. SDS-PAGE analysis of IgG-depleted plasma loaded onto a CIMac pA-@FIB column shows that FIB is efficiently bound to and eluted from the column with high purity. However, some protein A ligands on the monolith surface remained free (not in a complex with @FIB) even after @FIB immobilisation, resulting in potential places, where IgG from applied biological samples could bind, thus decreasing the column specificity. Indeed, when plasma not depleted of IgG was applied on a CIMac pA-@FIB column, a contaminating human IgG was detected in the elution fraction besides FIB. Due to insufficient specificity together with the high price of the protein A-based column, a strategy of binding @FIB to a hydraside-based support was thoroughly studied. Because a typical protocol for immobilisation of antibody onto a hydrazide ligand is time consuming with low coupling yield (buffer exchange after Ab oxidation, concentrating the Ab before binding to the column, using excess Ab in the coupling reaction), the focus of our work was on the optimisation of the coupling step. A unique approach in our new coupling protocol was the introduction of a small amount of cation-exchanging groups on the surface of a hydrazide-based monolith in order to enable preconcentration of positively charged Ab on the chromatographic surface.
To accelerate the immobilization protocol, a sample with oxidized @FIB was diluted 15 times with 50 mmol/L MES, pH 5.2 instead of using time consuming buffer exchange and antibody concentration. Decreasing the pH to below the pI of most Abs enabled ionic interactions with the negatively-charged monolithic support. Due to known very sharp adsorption isotherms of monolithic supports in ion exchange mode, a low concentration of @FIB in the loading solution (0.067 mg/mL) does not prevent efficient adsorption of the mAb when pumping the solution through the column. Consequently, the protein was efficiently concentrated on the monolith surface (no @FIB in the flow through) without any need for recirculation. Covalent binding of @FIB was completed after the column was incubated at room temperature for at least 10 hours. The amount of immobilised @FIB using the proposed protocol was controlled by adjusting the pH of the loading buffer. Increasing the pH from 5.2 to 5.8 and then to 6.4 decreased the ionic interaction, influencing the amount of adsorbed mAb (see Table 1, attached separatelly). Consequently, the binding capacity for FIB followed the binding capacity for @FIB, corroborating the need for the highest @FIB coupling possible in order to produce the most efficient immunoaffinity monolith.
Coupling efficiencies and mAb utilisation using CIMac-@FIB columns were in the range of those obtained with the CIMac pA-@FIB, but the new coupling protocol on a hydrazide-based monolithic support enabled considerably simpler and less cost intensive immobilisation as well as higher specificity towards FIB.
WP1 was devoted to high-throughput method development for fast isolation of glycoproteins from a large number of samples. These tasks were acomplished through the development of a „proof-of-principle“ monolith-based FIB immunoaffinity 8-well plate. Large-pore monoliths are more amenable to high-throughput applications due to their higher permeability and decreased possibility of clogging. Therefore, we used 2.1 μm average pore size diameter monoliths for preparation of CIMac-@FIB 8-well plates, as compared to 1.3 μm average pore size diameter monoliths which were used for CIMac-@FIB columns. On the other hand, large pores decrease the surface area of the monolith resulting in approximately 1/3 decreased binding capacity for the antigen. For this reason we used monoliths of larger volume (200 μL) for the CIMac-@FIB 8-well plate preparation, compared to the monoliths used for the column format (100 μL). @FIB was successfully bound in each well and the CIMac-@FIB 8-well plate was characterized for pure FIB binding capacity with the same reagents as with the CIMac-@FIB columns (see Figure 1, attached separately). The average binding capacity was between 3.5 and 4.5 mg/mL, which is in accordance with the 1/3 lower binding capacity compared to the 1.3 µm pore size monoliths. The RSD across the 8 wells was 2.2%.
Within WP2, the end goal in terms of CIM-@FIB 8-well plate application was to explore the potential biological differences in FIB N-glycosylation between individuals. The developed isolation procedure (GENOS) was applied to purify FIB from plasma samples of ten individuals (five men and five women of approximately same age) taken on the same day. Variability of the method was tested by using 18 aliquots of a standard plasma sample and isolations were performed in a period of three days. Figure 2 shows the average %Area and standard deviation of each FIB N-glycan peak in individuals (biological variability) and standards (three-day method variability). It is evident that the biological variability of FIB N-glycan peaks in ten healthy human individuals used in this study is larger than the variability of the same N-glycan peaks in the standard plasma sample, that is, of the method itself. Therefore, the developed semi-HTP procedure for immunoaffinity purification and N-glycan analysis of FIB from human plasma described in this study can be used to detect variations in FIB N-glycosylation in healthy individuals and, thus, most probably also in disease.
After a successful development of CIM-@FIB 8-well plate we continued with the development of HTP purification of human transferrin, another member of glycoprotein family, that was found in human plasma in medium abundance. To this aim, mice were immunized with human transferrin. Spleen cells of immunized mice were collected, fused with SP2/0 myeloma cells at ratio 1:1 and subsequently seeded onto 96-well tissue culture plates (this process is typically referred to as ‘fusion’). In the first test, cultures were screened for antibodies reactive against human transferrin by using an ELISA test. A number of positive hybridoma cell lines was further screened for their antibody production yield and capacity to adapt to serum free medium as well as for the ability of secreted antibodies to immunoprecipitate human transferrin from human plasma. Large scale Transferin.09 clone production was performed and monoclonal antibodies were purified. Transferin.09 monoclonal antibody was purified from serum free medium using GE ÄKTA Prime Plus Liquid Chromatography System and Protein G columns. Protein quality was confirmed by Coomassie blue staining (see Figures 3 and 4). The most efficient coupling reaction of the Transferin.09 clone was confirmed to be based on hydrazide coupling chemistry, as was already the case for fibrinogen. The first immunoaffinity chromatographic support was prepared in the format of analytical column with column volume of 0.1 mL (CIMac-@Tf column).
Since the final application for the prepared columns is isolation of the protein from plasma sample, we evaluated the column's specificity using a HPLC system and 1 mL of 10 times diluted plasma sample in bind/elute mode. In order to test the column efficiency at high-throughput conditions, we used a high flow rate of 1.0 mL min-1. The bound proteins were eluted using 0.1 mol L-1 formic acid, pH 2.4. The analysis of collected fractions by SDS-PAGE showed that the purity of the elution fraction was above 95 %. Bound Tf was eluted from the column in 500 µL of elution buffer, corresponding to 5 column volumes. Within WP2, glycoproteomic analysis of isolated glycoproteins, a HILIC-UPLC glycan analysis was performed for all fractions since it can detect antibody leakage in the wash after column storage or during Tf isolation procedure. There was no detectable antibody leakage during the washing step before sample application and the majority of the bound Tf eluted from the column in first 0.5 mL of 0.1 mol L-1 formic acid, pH 3.0 (see Figure 5).
Initial amount of Tf is important for reproducible glycosylation analysis by HILIC-UPLC and around 300 μg of purified Tf is required from a single capture step from blood plasma. Therefore, binding capacity for pure Tf dissolved in PBS buffer was determined for prototype CIMac-@Tf columns. The DBC50 values were between 3.5 and 3.9 mg of Tf per mL of support, enabling isolation of more than 300 μg of Tf in a single run using 0.1 mL CIMac-@Tf column. After successful characterisation of prototype Tf column, a complete monolithic 96-well plate with immobilised @Tf was developed. After @Tf immobilization the CIMac-@Tf 96-well plate was characterized for pure human Tf binding capacity with the same reagents as with the CIMac-@Tf columns. The average amount of eluted Tf calculated from absorbance reading at 280 nm was 300 µg per well with relative standard deviation (RSD) of 9.1 % for the whole plate, what is comparable with commercially available CIM® protein G or protein A 96-well plates. The Tf elution capacity was within previously determined requirements for amount of purified Tf and newly prepared CIMac-@Tf 96-well plate could be used for larger population studies.
In the frame of WP2, Transferrin was successfully isolated from around 1000 human plasma samples in the population study of Korčula island (Croatia). Comparative transferrin and immunoglobulin G N-glycosylation in this population study have been analyzed by ultra-performance liquid chromatography (UPLC, see Figure 6). Preliminary results indicate that N-glycosylation of transferrin varies depending on age and sex of individual. However, while immunoglobulin G N-glycosylation is more influenced by age, transferrin N-glycosylation is more dependent on sex of an individual. Additionally, immunoglobulin G and transferrin N-glycosylation associate with different biochemical parameters. Anti-transferrin plate that has been developed during this project has therefore been shown to be robust and re-usable for up to 20 times without a loss of capacity. In order to determine N-glycan composition in each chromatographic peak, UPLC peaks have been fractionated, ethyl-esterified and analyzed with matrix-assisted laser desorption/ionization (MALDI) mass spectrometry technique.
In the second period, and within WP1, an upgrade from purification of plasma proteins towards cell proteins was done (transmembrane protein MULT-1 and PPI-anchored protein RAE-1). We started development of a monoclonal antibody to the RAE-1 protein using the methodology described for the development of the anti-human transferrin antibody. First, we had to produce the immunogen, RAE-1-Fc fusion protein (RAE-1 protein fused to human IgG1 Fc domain, see Figure 7). The protein was successfully cloned, produced and purified. Next, we showed that the developed RAE-1γ.01 clone is able to precipitate native RAE-1 protein from cell lysates. The same clone is suitable for immunoblotting of RAE-1. In addition, we were able to produce a large quantity of purified RAE-1γ.01 clone needed for immobilization to affinity monolithic columns.
By use of affinity monolithic columns with immobilized Rae1γ-Fc, we could purify a significant amount of this low-abundant single membrane protein from a native source (see Figure 7), that is suitable for downstream analyses, such as mass spectrometry analysis of their glycans. Immunoaffinity chromatographic monoliths against RAE-1 GPI anchored glycoprotein were developed (CIMmic @RAE-1 column) as a part of Deliverable D.1.4 and D.3.3. with the main goal to compare glycosylation profiles of immunologically relevant molecules before and after MCMV infection (WP3).
As in the case of glycoprotein isolation from plasma, a hydrazide-coupling chemistry was proven to be the most ideal for the immobilisation of @RAE-1 antibody onto a CIM chromatographic support. Instead of using CIMac base support for preparation of a prototype immunoaffinity column, a novel column format was developed by BIA-SEP. It is a small, single monolithic column, made exclusively from plastic, which was designed according to the needs of a typical biotechnological/biochemical laboratory. The characteristics of novel column format were defined together with partner MEDRI for applications, where 96-well format is not applicable. The new member of CIM family was named CIMmic and is composed of a plastic housing, which enables easy exchange of monolithic disks. The volume of disk is only 50 uL, because as such it could be used for immobilisation of low abundacy proteins and expensive ligands. Additionally, a small volume design results in very concentrated eluted fractions, what simplifies the subsequent analysis. The column could be operated on HPLC instrument or manually by pumping the liquid by syringe, therefore such column is applicable even outside typical analytical laboratory.
The CIMmic @RAE-1 column functionality was confirmed with bind-elute tests with soluble Fc fusion protein (Fc-RAE-1), which was prepared from the supernatant of transfected HEK 293T cells using standard purification procedures. Additionally, binding capacity for Fc-RAE-1 was determined on anti-RAE-1 column to be approximately 5 mg of fusion protein per mL of chromatographic support. The result confirmed the possibility of purifying at least 200 µg of antigen using 50 µL column format in a single chromatographic step, what should be sufficient amount of protein needed for N-glycosylation studies.
BIA-SEP, together with MEDRI, isolated a small quantity of antigen from lysates of RAE-1 transfected cells, as confirmed by western blot analysis and HPLC-MS/MS analysis. Thus, the method was successful for analysis of mammalian membrane proteins. Next, our efforts were focused to the optimisation of sample preparation as well as chromatographic protocol for immunoaffinity isolation of larger (micro to milligram) quantities of membrane proteins from eukaryotic cells. One of the main issues was prevention of column clogging during sample application. Due to the high complexity of a lysate sample, an aggregation of biological molecules occurred within the pores of monoliths, negatively influencing the monolith permeability. A working solution was the addition of an acidification step inducing precipitation of impurities following by filtration through a 0.22 um filter, what drastically improved column permeability during the sample loading as a first chromatographic step. (D.1.4.).
We then performed 24 successful purification rounds of RAE-1 protein from infected and uninfected cells, both primary MEF mouse embryonic fibroblasts and NIH 3T3 cell line. There were no detectable unspecific bands in elution from immunoaffinity column when we used a lysate of cells that do not express RAE-1 (negative control). Finally, after protocol optimisation, we succeeded in isolation of approximately 1 mg of transduced RAE-1 and approximately 200 ug of transfected RAE-1 (Deliverable D.3.4.). The RAE-1 protein generated by transduction was isolated from cells infected with a previously constructed recombinant mouse cytomegalovirus in which the m152 viral gene, otherwise responsible for RAE-1 downregulation, was exchanged for the RAE-1 gene (the virus was constructed by replacing the m152 ORF, in the BAC-cloned MCMV genome, with a cassette comprising RAE-1 ORF under the control of HCMV immediate-early (IE) promoter). Therefore, the isolated RAE-1 represents the RAE-1 form present in infected cells that can be compared to the RAE-1 protein from infected cells, where exactly the same RAE-1 ORF (GenBank accession no. AAI32033) was cloned into the pB45-Neo plasmid and used to transfect NIH 3T3 (ATCC CRL-1658) cells. Flow rate of 40 column volumes per minute was used to pump more than 100 mL of lysate through 50 µL column (see Figure 8), what is impressive result, enabling high-throughput purification of a low-abundant protein in micro to milligram scale in every biotechnological laboratory.
N-glycosylation analysis was performed for transduced RAE-1 (MEF cells) for the first time (D.1.4.). Coomassie blue-stained gel (the elution fraction from one of chromatographic purifications) was sent to partner MPI. After confirming the identity of the SDS-PAGE band to be RAE-1 by HPLC-MS/MS analysis, an in-gel deglycosilation was performed using PNGase F, followed by extraction of N-glycans from gel. The analysis of MULT1 as well as RAE-1 glycans could be performed only on an analytical scale due to low amounts of this glycoprotein obtained during isolation. The foreseen high-throughput analysis of these membrane glycoproteins was postponed due to low amounts of the protein obtained during isolation that were adequate only for the analytical-experiments. Therefore, optimization of the high-throughput purification protocol (UHPLC) coupled to MALDI-TOF MS glycan analysis was done by use of the model protein, namely yeast invertase, which is a yeast cell-surface protein that may be easily isolated in large amounts. In particular, we developed a chromatographic procedure that combines step elution with sample displacement chromatography on anion exchange CIM monolith for separation of glycoforms. The analysis of the model-enzyme invertase that contains more than 50% oligosaccharides by weight showed 13 N-glycosylation sites in the invertase amino acid sequence. It is known that high mannose N-glycans may locally contain an outer shell composed of oligo-mannose composed of up to 60 mannose units and glycoforms of enzyme bearing different amounts of negative charges attached to glyco-components, thus, exhibit different stability. For industrial application, only glycoforms with highest stability should be selected to increase the overall efficiency of the catalytic industrial process. The chromatographic procedure developed was therefore used for separation of invertase glycoforms that carry different negative charges densities from a natural matrix, namely S. cerevisiae. These invertase glycoforms were used in the course of the project for development of glycan hydrolyses protocols based on glycosydases immobilized at monolithic supports.
We optimized miniaturized supports with immobilized glycosidases for high-throughput glycosylation analysis by use of mass spectrometry. The immobilized glycosidases activity was tested by characterization of separated glycoforms and confirmation of the protein amino acid sequence by mass spectrometry. In addition, optimized protocols for mass spectrometry analysis were tested and proved to be efficient to yield evidence on differences between separated forms of invertase that are due to the variety of identified glycan structures. Analysis of N-glycan invertase structure was done in a high-throughput fashion. Results showed that outer shell oligo-mannose units are highly complex and that ionization efficiency obtained so far was expectedly rather low. Tested protocols proved efficient to yield evidence on differences between separated forms of invertase that are due to the variety of identified glycan structures. This means that different glycoforms can be successfully separated from mixtures (including those from natural sources) by developed UHPLC protocol and mapped by use of MALDI-TOF MS. Moreover, we delivered protocols and procedures for purification and separation of plasma IgG and IgM and their subsequent high-throughput glycomic analysis by use of mass spectrometry-based methods. Developed chromatographic procedures rely on combined step elution and displacement chromatography on monolithic matrices for separation of protein glycoforms (highly glycosylated enzyme invertase was utilized as a model protein). Major portion of results in this field provided evidence on N-glycoforms that carry different charge due to presence of charged glycans and that impact stability of glycoproteins.
Furthermore, in the frame of WP3, we performed N-glycan analysis using multiplexed capillary gel electrophoresis with laser-induced fluorescence (xCGE-LIF). The main goal was to annotate N-glycan structures released from mouse IgG. This was performed using exoglycosidase digestions and xCGE-LIF analysis. Additionally, N-glycans released from mouse plasma were analyzed and the major structures were annotated using the same method. Structural assignment of glycans has been reached through fractionation of glycan peaks and exoglycosidase digestions approach. About 48% of all N-glycans are of the type high-mannose, while really high percentage of the complex N-glycans contain no terminal galactose. Obtained results will help to expand knowledge about mouse IgG glycome which is still lagging behind human IgG glycome. Developed methods and gained know-how will allow us to understand the role of individual protein glycosylation pattern in health and disease.
Furthermore, we have prepared new samples of purified RAE-1, namely RAE-1 from transduced NIH cells, as well as RAE-1 from transfected NIH cells (for sample preparation scheme, cf. Figure 9 that is submitted separately). Unfortunately, due to the low amount of sample obtained form transfected cells not all exoglycosidase digests needed to completely characterize the differences between uninfected and infected samples have been performed (see Figures 10 and 11).
Optimisation of protein G, protein A and protein L immobilisation on CIM monoliths has been performed and was used for the construction of optimised protein G 96-well plates. In addition, 96-well plates for isolation of immunoglobulin G (IgG) and immunoglobulin M (IgM) with mounted anion-exchange (DEAE) mini-disk (100µL) followed by mini-disk with immobilized amino acid tryptophan were developed. Here, the scheme that was earlier developed in the project leader’s laboratory for both preparative and high-throughput analytical work was adopted and optimized for high-throughput and semi-preparative isolation of low-abundance proteins clotting factor VIII (FVIII) and von Willebrand (vWF) factor from human plasma by use of ion-exchange and affinity chromatography on monoliths. Also, protein L immobilized on monolithic discs was tested as an alternative capture agent to proteins A and G in protocols for isolation of IgG and IgM from serum. High-throughput laboratory analysis (TECAN robot), and mass spectrometers MALDI-ToF/ToF and LC-ESI-MS/MS were used for high-throughput isolation and analysis of glycoproteins. We have succeeded to prepare CIM monoliths with high utilisation of protein L or G and with dynamic binding capacities for IgG around 10 mg/mL. We have successfully developed a scheme for fast and simultaneous affinity purification of α (A), γ (G) and µ (M) immunoglobulins from human serum through protein A, G and L affinity monolith chromatography. A novel 96-well plate with immobilised protein G or protein L enables more than 50 purification cycles of plasma sample without losing its activity.
Because protein G has different selectivity towards different groups of antibodies, we developed and optimised an immobilisation of complementary affinity ligand – protein L. So far we have succeeded to prepare CIM monoliths with high utilisation of protein L and with binding capacities for IgG around 10 mg/mL, what is comparable with CIM protein G monoliths. Using immobilised protein L on monolithic support in combination with protein A or protein G monoliths enables simple fractionation of antibody pool. Additionally, protein L enables isolation of rat antibodies, which do not interact with protein A or G. The protein L-based monolithic columns and 96-well plates were tested and successfully utilised for high-throughput isolation of these antibodies within WP3. Additionally, this method was adapted for the application of laboratory robotics.
Within WP3, we determined N-glycan structure of rat IgG isolated using protein L by different analytical approaches (HILIC-UPLC and mass spectrometry). N-glycosylation characterization of rat IgG will enable further research on the impact of different factors (e.g. different drugs and diseases) on IgG glycosylation in animal models. Next to that, we characterized N-glycans in brain tissue samples from different brain regions in four different species: human, chimpanzee, macaque and rat. Obtained results will enable analysis of brain tissue glycosylation patterns specific to brain regions, species and stages of development.
In the frame of WP2, development of high-throughput glycoproteomic analysis, analytical approaches for glycosylation analysis of the purified proteins by UPLC were developed. Optimal separation conditions were established, and the exact composition of glycan structures within each UPLC peak was determined for IgG and IgM. We optimized protocols for purification of three individual glycoproteins from human plasma (fibrinogen, transferrin and haptoglobin) for which new affinity monolithic columns were developed within WP1. Different column batches in combination with purification conditions (buffer, volume of buffer, flow rate, quantity and quality of starting material, etc.) were tested in order to achieve successful and reproducible purification of each glycoprotein. 8-well (out of 96-well) plate for fibrinogen isolation was used for successful isolation and glycan analysis of fibrinogen from approximately 100 plasma samples. Variation of glycan composition of fibrinogen standards was shown to be smaller than variation of glycan composition of fibrinogen isolated from ten individuals, which implies notable natural variability of fibrinogen glycome. Order of glycoprotein isolation (IgG, fibrinogen, transferrin and haptoglobin) from the same plasma sample was optimised, and composition of glycan structures in each UPLC peak was determined. By development and optimization of purification and hydrophilic interaction liquid chromatography (HILIC) separation of N-glycans released from human fibrinogen, transferrin and haptoglobin, methods for glycoprotein analysis have been realized in full.
N- and O-glycans can be analysed via „Porous Graphitized Carbon“[PGC]-Liquid Chromatography [LC]. In order to establish this technique a sample preparation method prior mass spectrometry analysis for N-and O-glycans was introduced using standards. A Quadrupole-Time-Of-Flight (Q-TOF) hybrid mass spectrometer was used for the analysis of the self-produced standards, first, an intensive parameter optimization was necessary. The instrument settings had then to be further optimized for the nano-LC setup of the PGC-LC separation that was coupled online to the mass spectrometer. This included the determination and optimization of the complete set of MS and MS/MS parameters for N-and O-glycan analysis for PGC nano-LC tandem MS runs. After establishment, initial analyses of samples were successfully run. However, due to problems with the spray stability of this Q-TOF instrument (low acetonitrile concentrations in the separation gradient; negative-ion mode) the PGC nano-LC set up was also introduced to a second Q-TOF instrument from another vendor.
CGE protocol for analysis of APTS-labled glycans was optimized for a high-throughput format (96-well plates). Optimization of sample preparation included adjustment of amount of released glycans and ratio of dye to glycans. For enrichment and clean-up of the labelled glycans, HILIC and size-exclusion protocols were tested. The procedure of CGE run was optimized as well with regard to the total time of the run, length of capillaries used and injection conditions. Analysis of glycosylation of antigen-specific antibodies using CGE and LC-MS methods was performed. Capillary flow high performance liquid chromatography (with porous graphitized carbon as stationary phase) coupled via online electrospray ionization to tandem iontrap mass spectrometry with collision induced dissociation (capPGC-HPLC-ESI-IT-CID-MS(/MS)) based separation and subsequent fragmentation of released glycans was implemented. On close inspection with standards such as fetuin, few problems were noticed with the workflow. This resulted in a low resolution of the chromatogram of the glycan peaks that were analyzed which, in turn resulted in poor recovery of a population of highly sialylated fetuin glycans. Furthermore, analysis of the fetuin glycome revealed an underrepresentation of other sialylated glycans, which pointed to an over-acidification of glycans. After troubleshooting and introduction of a nano-flow PGC-HPLC-column with the appropriate modifications to the chromatography setup including installation of a new to the UPLC system, the installed system was able to detect with high sensitivity the representative glycans of various standards. The data was confirmed with xCGE-LIF.
Furthermore, a new MS-based fragmentation mode was introduced. It has been observed that using high-energy regimes (high normalized collision energy HCD) was sufficient not only to show sugar fragmentation, and thus confirm the presence of glycopeptides, but also the fragment data can be also be used to determine the sequence on which the glycan resides. The intention of this work was to apply the optimized glycomic and glycoproteomic platform to the analysis of influenza virus A/PR/8/34 that is derived from MDCK adhesion and suspension cell lines. Qualitative glycomic analysis of influenza A/PR/8/34.MDCKADH and influenza A/PR/8/34.MDCKSUS were performed. The intention was to create a glycan library from which to make the assignment of their residence on the influenza protein easier to assign.
All planned secondments, both from industry to academia and vice versa, were realized within WP4 and WP5. Following researches have been recruited within WP 6: Dr. Uros Andjelkovic (UNIRI), Dr. Blaž Nemec (BIA-SEP), Dr. Terry Nguyen-Khuong (MPI), Dr. Olga Schvechuk (MEDRI), Dr. Olga Zaitseva (GENOS). In the frame of WP 7 and in accordance with the project plan, training in various research methods occurred in every project partner institution as a part of the research process, both by home staff training visiting researchers, and by visiting MERs and ERs transferring knowledge to the receiving institution. Within WP 8, all researchers from the HTP-GlycoMet network were involved in various teaching and training activities, such as laboratory demonstrating, tutorials/seminars; undergraduate project supervision and PhD project supervision; supervision of technical staff, in accordance with the project plan. Career planning workshop was not organized as a separate event, but as a part of the partner meetings.
Different trainings in scientific and technical methods applied for isolation of proteins from human and animal plasma, as well as hydrophobic membrane proteins were organized during three scientific meetings (Monolith Summer Schools and Monolith Symposia 2014 and 2015 and Glyco23 International Symposium), and researchers that were recruited by UNIRI and Genos in the frame of this project were participating and also teaching there. Our collaborative project was presented to the publics by different interview and presentations given by the project coordinator and group leaders (WP9), and project management administrative support and foreseen meetings were organized and reports were submitted on time (WP10).
The potential impact and the main dissemination activities and exploitation of results
The project deals with a very a very specific research field confined to glycoproteins and glycan structures in general. This research field is in close relationship with advanced in analitical methods and development of dedicated databases for glycan structures assignements. With inovative tools, the field would immediately boost the biomedical research area. Indeed, it is widely recognized that disease, including viral infection, cancer and chronic disease of aging population have been witnessing a steady stagnation in the field of new therapeutic approaches since years. Glycoproteins are major biomarkers of these groups of disease and advancements in their mapping in general might provide major breakthroughs in the field of improved diagnostics, prognostics and development of current and new therapies.
The HTP-GlycoMet project is a collaborative effort of three academic and two SME partners with expertise in complementary fields of science and technology. Individual partners have in the recent past been involved in several bilateral collaboration projects, but these have been of limited scope. The project implemented a high-level research plan aimed at bringing together the scientific expertise and infrastructural capacities of the partners and thus result in three immediate sets of outcomes. The first benefit of the project is an increased knowledge of complementary scientific fields: for example, Genos will obtain from UNI-RI and MEDRI further insights into the role of protein glycosylation in health research, with an emphasis on viral infections. On the other hand, from SMEs involved in the project UNI-RI and MEDRI will acquire reliable and tailor-made methods for specialised chromatographic tools (BIA-SEP) and high-throughput glycomic analysis (Genos). In addition, new ultra-sensitive xCGE-LIF methods for glycoproteomic analysis were developed by joint efforts of BIA-SEP, Genos and MPI. Thus, an important benefit is the increased research and business performance of the partners involved. Second, the realization of the HTP-GlycoMet project resulted in an extensive exchange of personnel between the commercial and academic partners as well as the recruitment of excellent experienced researchers and consequently, significantly strengthen the human resource capacity of all partners. Third, the transfer of knowledge between academic and industrial partners yielded in further development of innovative products and services. This, together with the actions like joint applications to additional EU and other calls for proposals, will ensure the continuation of collaboration beyond the lifetime of the project, and the establishment of new inter-sectoral collaboration projects. This consortium intends to gradually expand to other academic and industrial organisations involved in the research of various pathophysiological processes with the aim of providing glyco-analytical tools to the biomedical community. Resources which are generated through this project will provide proof of principle for the application of the developed glyco-analytical technologies for the study of immunity and infection. This will be continued through joint applications of HTP-GlycoMet partners to new calls expected in Horizon 2020 and other funding schemes. In the wider sense, collaboration with industrial partners from the Zagreb region (Croatia) and Vipava Valley region (Slovenia) will considerably improve the utilisation of technologies available at UNI-RI and MEDRI (Rijeka). This is relevant taking into account the lack of biotech SMEs into which these technologies could be successfully transferred in the immediate surroundings of the city of Rijeka. The implementation of this project could also boost knowledge transfer activities from UNI-RI and MEDRI to the business sector in the future. Owing to the extensive engagement of two research-intensive SMEs in this project, a new line of products and services is now in development. By a large part these is a collaborative effort which will require continuous future collaboration. The relevance of the project for the enhancement of innovation potential of the European Research Area is closely tied to the fact that after completion of the human genome project, and the glycoproteomics has become one of the next large challenges for biomedical science. It is now evident that genome is only a basic database of life and that several layers of complex omics integrate to form life in its full complexity. The same is true for complex human diseases. Glycoproteomic analysis is still globally deficient and is more and more required for research, but also in quality control and development of new biological drugs. Over 200,000 individuals were included in large genomic studies in the past 5 years, and many of these cohorts also have plasma samples available. In addition, many new cohorts are currently being collected and they are all highly interested in further phenotypisation of their patients, what is a great research and market potential. Moreover, newly developed technologies will be applied to study function of several proteins which are important in the process of viral infection (WP3). The expertise in antibody and fusion protein immobilisation was successfully deployed to extract the immunologically relevant membrane glycoproteins. This will greatly improve our current understanding of the interaction of the viral proteins and their ligands as well as the importance of their glycosylation profile. The impact on the innovation potential of the European Research Area is significant because of the identification of possible targets for diagnostics and treatment of viral infections and immunological processes in general. Second, UNI-RI and MEDRI will benefit from secondments to BIA-SEP, since these will be aimed at acquiring specific know-how on potential avenues for commercial exploitation of the existing unique collection of antibodies generated at MEDRI. These activities are expected to result in improved capabilities for commercialization of university-generated technologies and industry-oriented R&D at UNIRI and MEDRI, which currently have no spin-off companies in the field of biotechnology. A better exploitation of the strong biomedical science base available at these institutions will contribute to strengthening of the innovation potential of the European Research Area. Apart from the training in state-of-the-art laboratory techniques, training in complementary skills, such as presentation, proposals writing, entrepreneurship, business planning and leadership skills will also be organised. This will enhance the competences and career prospects of the recruited and seconded researchers. This has become increasingly important in the present science system, which demands from researchers the continuous transfer of basic research results into the commercial sector, or understanding how they can transform ideas into innovative products and services that address European and global health challenges. In addition to secondments, workshops on basic CIM technology and high-throughput affinity purification that will be organised in frame of the project will significantly facilitate the sharing of knowledge between the participants and external lecturers and participating researchers. B.5.2. Plans for exploitation of results and Dissemination strategy Exploitation of results Participation of two companies which are already global leaders in their respective fields warrants the successful exploitation of the research results. New technologies developed in the HTPGlycoMet network will enable Genos and BIA-SEP (in collaboration with UNI-RI, MEDRI and MPI) to preserve their current positions of market leaders in this rapidly advancing field by securing enabling IPR for the whole series of products and services. BIA-SEP is successfully selling a wide range of monolithic chromatographic support tools in EU, US and Asia. The development of new affinity 96- well products for rapid high-throughput purification of glycoproteins will enable them to enter the fields of high-throughput research and monitoring of production of recombinant proteins (therapeutic antibodies, other biological drugs, enzymes for cosmetic industry, etc.). Both of these fields desperately require methods for rapid high-throughput analysis of protein glycosylation and are promising markets, which will only grow in the future. Genos Ltd is currently the largest commercial provider of highthroughput glycomic analysis with sales in Europe, US, Australia and China. The majority of analyses are being done for researchers conducting a large population or patient stratification studies that need better characterisation of their patients. The demand for glycomic analysis is increasing and to preserve its leading position, Genos needs to develop new innovative products, which cannot be easily copied by low-cost companies in Asia. The third potential line of exploitation of research results stems from the experiments planned in WP3. In this work package important physiological processes during viral infection will be studied from a new angle enabled by technological development in WP1 and WP2. Contrary to WPs 1 and 2, where planned deliverables have clear potential for commercial exploitation, in WP3 it is not possible to predict whether the outcome will have some commercial potential or not. However, the results of this work package will be closely monitored and all potentially exploitable IP will be properly protected. The development of potential new therapeutic targets for combating viral infection exceeds the exploitation potential of our SME partners. Dissemination Among network participants. Scientific and training progress will be presented at the annual network meetings. In addition, the project web site will have a confidential section, which will be restricted to project participants and will contain confidential information relevant for the realisation of the project. To the scientific community / biotechnology industry. The dissemination of the results during and after the project will be through publications in high impact peer-reviewed journals and in contributions to scientific conferences (including three workshops on high-throughput technology that will be organized jointly with The Monoliths conferences in 2014 and 2016 and IGO congress in 2015 as detailed in WP7 and WP9). Partnerships with industry and European networks will be fostered, in particular at meetings and other events within the partnership’s own projects. To the general public. A dedicated web site will be set up for the project by IT services of UNI-RI that will contain all relevant and up-to-date information about the project progress, public research reports and relevant publications. Relevant media will be informed about the important project achievements and will be invited to interview project team members. Preparation of articles for the popularisation of science and participation of recruited and seconded researchers and other team members in different events, such are summer schools, festivals of science, and open door events, will help promote and better understand science and technology, in particular in the student population (from elementary schools, high schools and universities). Each recruited fellow will contribute to at least one outreach activity per year. As required by Annex II of the grant agreement, we will ensure that all publications and presentations by members of the project consortium - including all funded fellows - acknowledge the EU financial support received. This acknowledgement
will specifically refer to the Marie Curie Industry-Academia Pathways and Partnerships action, as well as the project number and acronym.
International conferences/events open to external researchers:
Monolith Summer School (2014), Portorož, Slovenia - BIA-SEP and UniRI
International Symposium on Glycoconjugates, (2015), Split, Croatia - GENOS
Monolith Summer School (2016), Portorož, Slovenia - BIA-SEP and UniRi
GlycoBiotech (2017), Berlin, Germany – MPI
Address of the project public website: www.glycomet.uniri.hr/en
.
The main aim of the HTP-GlycoMet proposal was to develop the analytical technologies for the high-throughput analysis of glycosylation of individual glycoproteins isolated for body fluids and cell membranes (see also graphical abstract that is submitted separately). In the frame the work package 1 (WP1), monoclonal antibodies against both blood plasma and membrane proteins, and methods for isolation of native forms of these molecules were developed. The optimized method for antibody immobilization on monolithic columns was optimized, and subsequent isolation of corresponding antigens in their native form from plasma (as body fluid) and from cell membranes was performed. For membrane proteins, methods for production of glycoprotein amounts that are sufficient for subsequent glycan analyses were optimized. For this sake, corresponding immobilization techniques and miniaturized monolithic columns and new purification methods for isolation of biologically active proteins from large volume of medium were developed. Monolithic supports with immobilized monoclonal antibodies against above mentioned individual proteins were mounted in a 96-well ELISA plate and were applied for isolation of single antigens. This high-throughput method was validated and also adapted for application of laboratory robotics and for possible use in routine analyses. In the next step, and in the frame of the WP2, oligosaccharides from isolated glycosylated proteins were cleved by corresponding glycosidases. For development of high-throughput analysis, glycosidases were immobilized on monolithic supports, and a method for «in-flow» enzymatic cleavage was developed. In order to understand the role of individual protein glycosylation pattern during biological processes and in health and disease, methods for both glycoprotein analysis and analysis of the structures of glycan parts based on ultra high performace liquid chromatography (UPLC) and mass spectrometry (MS and MS/MS) were applied. By use of these analytical methods, changes of protein glycosylation during pathological processes such as inflamation and/or viral infection were determined. Capillary gel electrophoresis with laser-induced fluorescence (CGE-LIF) was introduced as a complementary high-throughput method for the analysis of glycan structures, and both UPLC and xCGE-LC databases for glycan assigment were developed and collection of necessary data was started. In the next step (WP3) glycoproteins were identified by use of electrosprey ionisation-LC-MS/MS (ESI-LC-MS/MS), and their potential glycosylation sides were assigned. Membrane glycoproteins RAE-1 and MULT-1 were isolated from both citomegalovirus infected and non-infected cells. In order to isolate protein amounts that will be sufficient for analysis of glycan structure, production of these glycoproteins was up-scaled, and processing of large volumes of protein-containing supernatant by use of small monolithic columns was optimized. However, the analysis of MULT1 as well as RAE-1 glycans could still be performed only on an analytical scale due to low amounts of these glycoproteins obtained during isolation. The foreseen high-throughput analysis of these membrane glycoproteins was postponed due to amounts of proteins obtained during isolation that were adequate only for the analytical-experiments, and further scaling-up protocols for production of larger amounts of these proteins are in development. Therefore, optimization of the high-throughput purification protocol by use of UHPLC coupled to MALDI-TOF MS glycan analysis was done by use of the model glycoprotein yeast invertase, which is localized on the cell surface of the yeast Saccharomyces cerevisiae. Large amounts of the yeast can be easily fermented, and invertase can be easily isolated in amounts that are sufficient for large-scale analyses. In the next step, a chromatographic procedure for separation of invertase glycoforms on anion-exchange monolithic supports was developed. The ESI-LC-MS/MS analysis of this model glycoprotein that contains more than 50% oligosaccharides by weight showed thirteen N-glycosylation sites in the amino acid sequence. The chromatographic procedure for separation of invertase glycoforms was that carry different negative charges densities from a natural matrix was developed. These invertase glycoforms were used in the course of the project for development of glycan hydrolyses protocols based on glycosydases immobilized at monolithic supports. This protocol was also applied for high-throughput analysis of glycan structures of other glycoproteins. The aim of the WP4 is to benefit the transfer of the knowledge and technologies between SME and academic partners, and for this sake the exchange of both young and experienced researchers (in form of secondments) took place from the industry to the academia, as well as from the academia to industry (WP5). The foreseen exchange in both directions was completely realized. In the frame of the WP6 four experienced researcher with specific skills were recruited by Genos, UNI-RI, MEDRI and MPI. Their work was successfully performed, and yielded in numerous oral presentations and posters at different scientific meetings, as well in a significant number of publications. Special Issue of the journal Electrophoresis dedicated to monolithic supports was recently edited by the Coordinator of the HTP-GlycoMet project. This issue contains six papers that present our work performed during 4 years of project realization. Different trainings in scientific and technical methods (WP7) applied for isolation of proteins from human and animal plasma, as well as hydrophobic membrane proteins were organized during three scientific meetings (Monolith Summer Schools and Monolith Symposia 2014 and 2015 and Glyco23 International Symposium), and researchers that were recruited by UNIRI and Genos in the frame of this project were participating and also teaching there. Workshops dedicated to the the teaching/training skills, presentation, business and both project writing and project management, as well as career planning (WP8) were organized by MEDRI with participation of a large number of young researchers from both academia and industry. Our collaborative project was presented to the publics by different interview and presentations given by the project coordinator and group leaders (WP9), and project management administrative support and foreseen meetings were organized and reports were submitted on time (WP10). In conclusion, the foreseen work in WP1-WP10 was realized. Due to difficulties in up-scaling of production of membrane glycoproteins from non-infected and virally infected cultured cells, we achieved to assign oligosaccharide structures to the peaks in both UPLC and CGE, but until the end of the project time, the analyses could be performed only on an analytical scale. The method for high-throughput analysis of glycans is developed, and the development of the protocol for further scaling up of membrane proteins is ongoing.
A summary description of project context and objectives
When started, the main aim of the HTP-GlycoMet proposal was to develop technologies for the high-throughput analysis of glycosylation of individual proteins from body fluids and cell membranes as one of most common and most important posttranslation modifications. This knowledge will be applied for the understanding of the machanisms of both immunity and infections. Glycan analysis is extremely demanding from both technological and conceptual aspect and large-scale studies of glycosylation of individual proteins are still a big challenge, and before the start of this common proposal, only one large scale study was performed by the partners of this proposal. Our strategic concept was to combine the complementary expertise in the (i) production of specialised monolithic chromatographic tools for high-throughput fractionation of complex biological fluids (BIASEP), (ii) purification of proteins from both body fluids and membrane proteins by use of high-throughput chromatographic techniques (UNI-RI), (iii) high-throughput glycomic analysis by use of high-performance liquid chromatography (Genos) and multiplexed capillary gel electrophoresis (MPI) with following MS and MS/MS analysis of separated fractions and (iv) expertise in the field of viral immunology and production of specific monoclonal and polyclonal antibodies (MEDRI) in order to achieve significant progress in this direction. All our partners are recognized leaders who already made significant progress beyond the state of the art in their respective fields. Additionally, through the HTP-GlycoMet project secondments on all levels (MER, ER, ESR) were realized in order to capitalize on synergistic effects of this interdisciplinary and transnational collaboration. In addition to the generation of new knowledge and the development of new innovative technologies, and according to one of main goals of this collaboration, significant transfer of know-how between academic and industrial partners was achieved. As expected, our SME partners already started developing new lines of products and services through the HTP-GlycoMet programme, but also through future collaboration with HTP-GlycoMet partners beyond the lifetime of this project. The main objective of the HTP-GlycoMet initiative, the development of robust, reliable high-throughput methods for glycoproteomics during four project years was achieved. When the HTP-GlycoMet initiative started four years ago, the predominating glycomics approach was the analysis of glycans from all glycoproteins as a single pool of structures. We immediatelly started with much more advanced glycoproteomics approach, and our focus was the analysis of the glycosylation of individual glycoproteins that were isolated from large number of samples by use of high-throughput techniques. While glycomics can identify some general glycosylation features like deficient glycan branching, increased fucosylation, as well as changes in sialic acid as terminal sugar, these variations can have very different consequences on individual glycoproteins. For example, increased fucosylation of IgG would decrease antibody-dependent cellular cytotoxicity (ADCC) and consequential tissue damage, while increased fucosylation of other plasma proteins is generally associated with inflammation. Sialylated Fc domains of the immunoglobulins G (IgG) have also antiinflammatory properties, and these changes are essential for the mode of action of these glycoproteins. From the structural point of view, the analysis of individual glycoproteins is less complex than the analysis of the complete tissue glycome, but it requires higher sensitivity of applied analytical systems because the majority of individual glycoproteins are present in very low concentrations. The second major obstacle is the isolation process of individual low abundance glycoproteins from complex mixtures. The combined effects of low amount of available glycoproteins, especially these ones that can be produced only by use of cultured cells (such as membrane proteins) frequently combined with their degradation or inactivaton (or both) during the isolation process, and the high-throughput approach generally hampers the quality control of each sample. Therefore, innovative methods, fast, robust and reliable, are needed to minimise inaccuracies introduced by protein purification during sample preparation. We addressed these two main obstacles and also solved these problems during the project realization by combining complementary expertise of our partners in reaching the project objectives: BIA-SEP (proprietary monolithic technology), MEDRI (production of both membrane proteins by use cultured cells before and after viral infection, and monoclonal antibodies against single proteins), UNI-RI (isolation of individual glycoproteins), Genos (high-throughput glycomic analysis via HPLC/UPLC based method) and MPI (high-throughput glycomic analysis via a system based on multiplexed capillary gel electrophoresis with laser induced fluorescence detection (xCGE-LIF)). These newly developed methods are further implemented to study important aspects of protein glycosylation during viral infection and other pathological processes. The planned experiments were organized in three work packages (WP1-WP3) as described below. To reach these objectives, the partners were in the first step focused on the implementation of existing expertise, already established protocols and available material to reach the proof of principle level, and this work was successfully finalized in the first year of the realization of the project. The next step was the integration of the selected technologies in order to develop a novel approach, an actual prerequisite for the high-throughput implementation. For example, MEDRI has extensive expertise in the analysis and interpretation of molecular mechanisms exploited by cytomegaloviruses in order to subvert the immune system. The current state of the art in the field shows strong evidence that infection leads to change of glycosylation or to misglycosylation of the cellular proteins. However, this is mainly achieved through the site mutations of the amino acids as putative glycosylation sites or by using drugs that prevent different posttranslational modifications. The changes in the total protein glycosylation were monitored, however, due to analyses of complete glycome (and not singular, isolated proteins) little information was available about the exact cellular processes involved. Thus, new protocols had to be developed in order to analyse individual sugar differences on individual isolated glycoproteins, especially their changes under pathological conditions. Likewise, BIA-SEP had a strong base in the development of different high-throughput separation media, and first also experiences in production of method for fast handling of large number of samples. However, the production of affordable customized chromatographic devices based on the antibody of interest and especially in a format allowing a significant number of different samples/conditions was still a challenge. After the proof of principle level in first step, in the next step of project realization, several successful protocols for production of chromatographic media based on specific monoclonal antibodies were devloped. These media were used for glycoprotein analyses starting from crude samples containing sometimes target protein in very low concentrations, and ending in sophisticated glycoanalysis from individual proteins. Comparing glycoproteins from samples of patients with control samples, or samples of non-infected cells with cells after infection with citomegalovirus, changes of protein glycosylation under different pathological pressures were determined. New methodologies that were developed by the partners were presented as a platform for various applications, and fundaments for new products and protocols were established.
List of partners:
1. UNI-RI BIOTECH - http://www.biotech.uniri.hr/en/
2. BIA-SEP - http://www.biaseparations.com/
3. Genos - http://genos.hr/en/
4. MPI - http://www.mpi-magdeburg.mpg.de/2316/en
5. MEDRI - http://capri.com.hr/
Work packages, objectives, deliverables and milestones
WP 1. New tools for high-throughput isolation of individual glycoproteins
O.1.1. To acquire monoclonal antibodies with adequate characteristics for the preparation of new affinity materials
O.1.2. To isolate individual glycoproteins from complex mixtures in a high-throughput setting
D.1.1. New affinity monolithic 96-well plates (M48)
D.1.2. Monoclonal antibodies for affinity purification (M24)
D.1.3. Methods for purification of plasma proteins (M48)
D.1.4. Methods for purification of membrane proteins (M48)
M.1.1. Monoclonal antibodies (M12)
M.1.2. Tools for purification of glycoproteins (M24)
WP 2. Development of high-tFkor
hroughput glycoproteomic analysis
O.2.1. To understand the role of individual protein glycosylation pattern in health and disease
D.2.1. New UPLC methods (M48)
D.2.2. UPLC glycan assignment database (M48)
D.2.3. Structural analysis by MS and MS/MS (M48)
D.2.4. New CGE-LIF methods (M48)
D.2.5. xCGE-LIF glycan assignment database (M48)
M.2.1. Methods for glycoprotein analysis (M24)
WP 3. Analysis of protein glycosylation during viral infection
O.3.1. To explore whether viruses interfere with glycosylation patterns of the immune system receptors and ligands
D.3.1. Lysates of transfectant cell lines non-infected or infected with various recombinant MCMVs (M48)
D.3.2. Production of additional monoclonal antibodies (to viral and cellular targets) for affinity purification (M24)
D.3.3. Immobilization of antibodies and isolation of target glycoproteins (M36)
D.3.4. Comparison of glycosylation profile of immunologically relevant molecules before and after MCMV infection (M48)
D.3.5. Methods for the identification of interaction partners using fusion proteins (M48)
M.3.1. Glycoproteins from virally infected cells (M36)
WP 4. Secondments from industry to academia
O.4.1. To benefit from the transfer of specific, interdisciplinary knowledge and technologies between SME and academic partners (as described in detail in WP1-3), with the particular emphasis on knowledge available at academic institutions that can help SMEs increase their competitiveness on global markets
D.4.1. From BIA-SEP: MER to UNI-RI for 2 months (M24)
D.4.2. From BIA-SEP: two ERs to MEDRI; 4 months each (M24, 48)
D.4.3. From BIA-SEP: ESR to UNI-RI for 5 months (M24, 48)
D.4.4. From BIA-SEP: ESR to MEDRI for 3 months (M36)
D.4.5. From Genos: MER to MEDRI for 2 months (M12, 24, 36, 48)
D.4.6. From Genos: ER to MPI for 3 months (M12)
D.4.7. From Genos: ESR to MEDRI for 2 x 3 months (M36)
D.4.8. From Genos: ER to UNI-RI for 3 months (M36)
D.4.9. From Genos: ESR to MPI for 12 months (M24, 36)
WP 5. Secondments from academia to industry
O.5.1. To benefit from the transfer of specific, interdisciplinary knowledge and technologies between SME and academic partners (as described in detail in WP1-3), with the particular emphasis on knowledge available at SMEs that can help academic researchers increase their competitiveness and improve career prospects
D.5.1. From UNI-RI: MER to BIA-SEP and Genos; 18mts (M12, 24, 36, 48)
D.5.2. From UNI-RI: MER to BIA-SEP for 2 x 3 months and Genos for 3 months (M24, 36, 48)
D.5.3. From MEDRI: MER to BIA-SEP and Genos; 4 mts (M12, 24, 36, 48)
D.5.4. From MEDRI: MER to BIA-SEP and Genos; 2 x 3 months (M12, 24, 36, 48)
D.5.5. From MEDRI: ER to Genos for 6 months (M36)
D.5.6. From MEDRI: ER to BIA-SEP for 3 months (M24)
D.5.7. From MEDRI: ER/Project manager to BIA-SEP and Genos for 4 months (M36, 48)
D.5.8. From MPI: MER to Genos for 2 months (M12, 24, 36, 48)
D.5.9. From MPI: ER to Genos for 4 months (M12, 24, 36, 48)
D.5.10. From MPI: ESR to Genos for 12 months (M24, 36)
WP 6. Recruitment
O.6.1. To enhance the project team through recruitment of experienced researchers with specific skills required to contribute to the accomplishment of the interdisciplinary project objectives
D.6.1. Recruitment of ER for Genos (24 months) (M36)
D.6.2. Recruitment of ER for UNI-RI (24 months) (M36)
D.6.3. Recruitment of ER for MEDRI (24 months) (M36)
D.6.4. Recruitment of ER for MPI (24 months) (M24)
WP 7. Training in scientific and technical methods
O.7.1. To enhance the scientific and technical skills of recruited and seconded fellows through day-today transfer of knowledge in laboratories and specialised theoretical and practical workshops
D.7.1. Completed training in general methods (M48)
D.7.2. Workshop on “CIM technology” (M12)
D.7.3. Workshop on high throughput affinity purification (M48)
WP 8. Training in complementary skills
O.8.1. To enhance the scientific competitiveness and career prospects of recruited and seconded fellows through training in a variety of complementary skills (presentation skills, career planning)
O.8.2. To improve the leadership, managerial and business skills of recruited and seconded fellows through training in a variety of complementary skills (project management, entrepreneurship)
D.8.1. Teaching/ Training skills (M48)
D.8.2. Presentation skills (M48)
D.8.3. Workshop on business, project proposals writing and project management (M36)
D.8.4. Career planning workshop (M48)
WP 9. Dissemination and outreach
O.9.1. To promote the results of the project in the scientific community
O.9.2. To increase the visibility of the project and promote its benefits to policy makers and the general public
D.9.1. Dissemination/outreach activities (M48)
D.9.2. Workshop on “High-throughput glycoproteomics” (M24)
WP 10. Project management
O.10.1. To establish an efficient project managerial system and administrative support
O.10.2. To monitor the overall financial, administrative, legal and ethical aspects and progress of the project, including contracting with REA
D.10.1. Kick-off Meeting and establishment of the project administrative and managerial structure
D.10.2. Consortium meetings (M10, 22, 34, 46)
D.10.3. Progress, midterm, periodic and final reports (M12, 22, 24, 36, 48)
D.10.4. MidTerm Review (M22)
A description of the main science and technology results
The main objective of the HTP-GlycoMet initiative was the development of robust and reliable high-throughput methods for glycoproteomics. In the frame of WP1, BIA-SEP has optimized the immobilization procedure of antibodies onto monolithic chromatographic support (CIM® monoliths). The immobilisation protocol was mostly developed using anti-fibrinogen antibodies (@FIB) - murine IgG monoclonal antibodies, specific for human fibrinogen, a secretory glycoprotein that is synthesized in the liver and has a key role in blood clotting. Glycosylation of fibrinogen is not very well understood and to our knowledge it has been studied only on commercial fibrinogen samples. To meet the demand of high-throughput (HTP) glycomics studies, or glycan analysis of large numbers of samples, the methods used for analysis have to be fast, robust and reproducible. The use of polymethacrylate monolithic columns as a chromatographic support enables flowrate-independent binding capacity and resolution for large biomolecules due to the convective nature of the flow of the mobile phase. Convective Interaction Media (CIM®) monolithic columns have been commercialised by BIA-SEP and have already proven to be an appropriate medium for affinity chromatography, but to our knowledge there was only one previous report by three members of our team (BIA-SEP, Genos and UniRi) combining immunoaffinity sorbents with chromatographic monoliths in a 96-well format before the start of HTP-Glycomet project.
BIA-SEP's first goal was to prepare the immunoaffinity CIM monolith with the highest possible binding capacity for human plasma FIB. Preliminary experiments of mAb immobilisation on monoliths employing non-oriented immobilisation (coupling the protein through its free amino or thiol groups) resulted in binding capacities for fibrinogen below 1 mg/mL (results not shown), which was not enough for the intended application. Therefore, we focused on oriented mAb immobilisation, where the Fab region of the Ab, responsible for the interaction with the antigen, is positioned towards the lumen of the monolithic pore and is available for interaction with its counterpart. Covalent cross linking of the Ab with a protein A chromatographic support or binding of the antibody to a hydrazide support through oxidized sugar moieties are two typical approaches found in literature. We have optimised both coupling protocols for the protein immobilisation onto monolithic analytical column with column volume of 0.1 mL (CIMacTM).
Protein A crosslinking approach requires a protein A-specific mAb. The mAb of our choice was murine IgG2a, which interacts with protein A. High coverage of the monolith surface area with @FIB enabled lower exposure of non-occupied protein A molecules, present on the monolithic surface. No @FIB was detected in the wash and elution fractions after the crosslinking reaction with DMSI and DMPI, proving successful fixation of @FIB to the monolithic support. The dynamic binding capacity (DBC) for pure FIB on the CIMac pA-@FIB column was 4.6 mg/mL of support, that was high enough for isolation of FIB from human plasma. SDS-PAGE analysis of IgG-depleted plasma loaded onto a CIMac pA-@FIB column shows that FIB is efficiently bound to and eluted from the column with high purity. However, some protein A ligands on the monolith surface remained free (not in a complex with @FIB) even after @FIB immobilisation, resulting in potential places, where IgG from applied biological samples could bind, thus decreasing the column specificity. Indeed, when plasma not depleted of IgG was applied on a CIMac pA-@FIB column, a contaminating human IgG was detected in the elution fraction besides FIB. Due to insufficient specificity together with the high price of the protein A-based column, a strategy of binding @FIB to a hydraside-based support was thoroughly studied. Because a typical protocol for immobilisation of antibody onto a hydrazide ligand is time consuming with low coupling yield (buffer exchange after Ab oxidation, concentrating the Ab before binding to the column, using excess Ab in the coupling reaction), the focus of our work was on the optimisation of the coupling step. A unique approach in our new coupling protocol was the introduction of a small amount of cation-exchanging groups on the surface of a hydrazide-based monolith in order to enable preconcentration of positively charged Ab on the chromatographic surface.
To accelerate the immobilization protocol, a sample with oxidized @FIB was diluted 15 times with 50 mmol/L MES, pH 5.2 instead of using time consuming buffer exchange and antibody concentration. Decreasing the pH to below the pI of most Abs enabled ionic interactions with the negatively-charged monolithic support. Due to known very sharp adsorption isotherms of monolithic supports in ion exchange mode, a low concentration of @FIB in the loading solution (0.067 mg/mL) does not prevent efficient adsorption of the mAb when pumping the solution through the column. Consequently, the protein was efficiently concentrated on the monolith surface (no @FIB in the flow through) without any need for recirculation. Covalent binding of @FIB was completed after the column was incubated at room temperature for at least 10 hours. The amount of immobilised @FIB using the proposed protocol was controlled by adjusting the pH of the loading buffer. Increasing the pH from 5.2 to 5.8 and then to 6.4 decreased the ionic interaction, influencing the amount of adsorbed mAb (see Table 1, attached separatelly). Consequently, the binding capacity for FIB followed the binding capacity for @FIB, corroborating the need for the highest @FIB coupling possible in order to produce the most efficient immunoaffinity monolith.
Coupling efficiencies and mAb utilisation using CIMac-@FIB columns were in the range of those obtained with the CIMac pA-@FIB, but the new coupling protocol on a hydrazide-based monolithic support enabled considerably simpler and less cost intensive immobilisation as well as higher specificity towards FIB.
WP1 was devoted to high-throughput method development for fast isolation of glycoproteins from a large number of samples. These tasks were acomplished through the development of a „proof-of-principle“ monolith-based FIB immunoaffinity 8-well plate. Large-pore monoliths are more amenable to high-throughput applications due to their higher permeability and decreased possibility of clogging. Therefore, we used 2.1 μm average pore size diameter monoliths for preparation of CIMac-@FIB 8-well plates, as compared to 1.3 μm average pore size diameter monoliths which were used for CIMac-@FIB columns. On the other hand, large pores decrease the surface area of the monolith resulting in approximately 1/3 decreased binding capacity for the antigen. For this reason we used monoliths of larger volume (200 μL) for the CIMac-@FIB 8-well plate preparation, compared to the monoliths used for the column format (100 μL). @FIB was successfully bound in each well and the CIMac-@FIB 8-well plate was characterized for pure FIB binding capacity with the same reagents as with the CIMac-@FIB columns (see Figure 1, attached separately). The average binding capacity was between 3.5 and 4.5 mg/mL, which is in accordance with the 1/3 lower binding capacity compared to the 1.3 µm pore size monoliths. The RSD across the 8 wells was 2.2%.
Within WP2, the end goal in terms of CIM-@FIB 8-well plate application was to explore the potential biological differences in FIB N-glycosylation between individuals. The developed isolation procedure (GENOS) was applied to purify FIB from plasma samples of ten individuals (five men and five women of approximately same age) taken on the same day. Variability of the method was tested by using 18 aliquots of a standard plasma sample and isolations were performed in a period of three days. Figure 2 shows the average %Area and standard deviation of each FIB N-glycan peak in individuals (biological variability) and standards (three-day method variability). It is evident that the biological variability of FIB N-glycan peaks in ten healthy human individuals used in this study is larger than the variability of the same N-glycan peaks in the standard plasma sample, that is, of the method itself. Therefore, the developed semi-HTP procedure for immunoaffinity purification and N-glycan analysis of FIB from human plasma described in this study can be used to detect variations in FIB N-glycosylation in healthy individuals and, thus, most probably also in disease.
After a successful development of CIM-@FIB 8-well plate we continued with the development of HTP purification of human transferrin, another member of glycoprotein family, that was found in human plasma in medium abundance. To this aim, mice were immunized with human transferrin. Spleen cells of immunized mice were collected, fused with SP2/0 myeloma cells at ratio 1:1 and subsequently seeded onto 96-well tissue culture plates (this process is typically referred to as ‘fusion’). In the first test, cultures were screened for antibodies reactive against human transferrin by using an ELISA test. A number of positive hybridoma cell lines was further screened for their antibody production yield and capacity to adapt to serum free medium as well as for the ability of secreted antibodies to immunoprecipitate human transferrin from human plasma. Large scale Transferin.09 clone production was performed and monoclonal antibodies were purified. Transferin.09 monoclonal antibody was purified from serum free medium using GE ÄKTA Prime Plus Liquid Chromatography System and Protein G columns. Protein quality was confirmed by Coomassie blue staining (see Figures 3 and 4). The most efficient coupling reaction of the Transferin.09 clone was confirmed to be based on hydrazide coupling chemistry, as was already the case for fibrinogen. The first immunoaffinity chromatographic support was prepared in the format of analytical column with column volume of 0.1 mL (CIMac-@Tf column).
Since the final application for the prepared columns is isolation of the protein from plasma sample, we evaluated the column's specificity using a HPLC system and 1 mL of 10 times diluted plasma sample in bind/elute mode. In order to test the column efficiency at high-throughput conditions, we used a high flow rate of 1.0 mL min-1. The bound proteins were eluted using 0.1 mol L-1 formic acid, pH 2.4. The analysis of collected fractions by SDS-PAGE showed that the purity of the elution fraction was above 95 %. Bound Tf was eluted from the column in 500 µL of elution buffer, corresponding to 5 column volumes. Within WP2, glycoproteomic analysis of isolated glycoproteins, a HILIC-UPLC glycan analysis was performed for all fractions since it can detect antibody leakage in the wash after column storage or during Tf isolation procedure. There was no detectable antibody leakage during the washing step before sample application and the majority of the bound Tf eluted from the column in first 0.5 mL of 0.1 mol L-1 formic acid, pH 3.0 (see Figure 5).
Initial amount of Tf is important for reproducible glycosylation analysis by HILIC-UPLC and around 300 μg of purified Tf is required from a single capture step from blood plasma. Therefore, binding capacity for pure Tf dissolved in PBS buffer was determined for prototype CIMac-@Tf columns. The DBC50 values were between 3.5 and 3.9 mg of Tf per mL of support, enabling isolation of more than 300 μg of Tf in a single run using 0.1 mL CIMac-@Tf column. After successful characterisation of prototype Tf column, a complete monolithic 96-well plate with immobilised @Tf was developed. After @Tf immobilization the CIMac-@Tf 96-well plate was characterized for pure human Tf binding capacity with the same reagents as with the CIMac-@Tf columns. The average amount of eluted Tf calculated from absorbance reading at 280 nm was 300 µg per well with relative standard deviation (RSD) of 9.1 % for the whole plate, what is comparable with commercially available CIM® protein G or protein A 96-well plates. The Tf elution capacity was within previously determined requirements for amount of purified Tf and newly prepared CIMac-@Tf 96-well plate could be used for larger population studies.
In the frame of WP2, Transferrin was successfully isolated from around 1000 human plasma samples in the population study of Korčula island (Croatia). Comparative transferrin and immunoglobulin G N-glycosylation in this population study have been analyzed by ultra-performance liquid chromatography (UPLC, see Figure 6). Preliminary results indicate that N-glycosylation of transferrin varies depending on age and sex of individual. However, while immunoglobulin G N-glycosylation is more influenced by age, transferrin N-glycosylation is more dependent on sex of an individual. Additionally, immunoglobulin G and transferrin N-glycosylation associate with different biochemical parameters. Anti-transferrin plate that has been developed during this project has therefore been shown to be robust and re-usable for up to 20 times without a loss of capacity. In order to determine N-glycan composition in each chromatographic peak, UPLC peaks have been fractionated, ethyl-esterified and analyzed with matrix-assisted laser desorption/ionization (MALDI) mass spectrometry technique.
In the second period, and within WP1, an upgrade from purification of plasma proteins towards cell proteins was done (transmembrane protein MULT-1 and PPI-anchored protein RAE-1). We started development of a monoclonal antibody to the RAE-1 protein using the methodology described for the development of the anti-human transferrin antibody. First, we had to produce the immunogen, RAE-1-Fc fusion protein (RAE-1 protein fused to human IgG1 Fc domain, see Figure 7). The protein was successfully cloned, produced and purified. Next, we showed that the developed RAE-1γ.01 clone is able to precipitate native RAE-1 protein from cell lysates. The same clone is suitable for immunoblotting of RAE-1. In addition, we were able to produce a large quantity of purified RAE-1γ.01 clone needed for immobilization to affinity monolithic columns.
By use of affinity monolithic columns with immobilized Rae1γ-Fc, we could purify a significant amount of this low-abundant single membrane protein from a native source (see Figure 7), that is suitable for downstream analyses, such as mass spectrometry analysis of their glycans. Immunoaffinity chromatographic monoliths against RAE-1 GPI anchored glycoprotein were developed (CIMmic @RAE-1 column) as a part of Deliverable D.1.4 and D.3.3. with the main goal to compare glycosylation profiles of immunologically relevant molecules before and after MCMV infection (WP3).
As in the case of glycoprotein isolation from plasma, a hydrazide-coupling chemistry was proven to be the most ideal for the immobilisation of @RAE-1 antibody onto a CIM chromatographic support. Instead of using CIMac base support for preparation of a prototype immunoaffinity column, a novel column format was developed by BIA-SEP. It is a small, single monolithic column, made exclusively from plastic, which was designed according to the needs of a typical biotechnological/biochemical laboratory. The characteristics of novel column format were defined together with partner MEDRI for applications, where 96-well format is not applicable. The new member of CIM family was named CIMmic and is composed of a plastic housing, which enables easy exchange of monolithic disks. The volume of disk is only 50 uL, because as such it could be used for immobilisation of low abundacy proteins and expensive ligands. Additionally, a small volume design results in very concentrated eluted fractions, what simplifies the subsequent analysis. The column could be operated on HPLC instrument or manually by pumping the liquid by syringe, therefore such column is applicable even outside typical analytical laboratory.
The CIMmic @RAE-1 column functionality was confirmed with bind-elute tests with soluble Fc fusion protein (Fc-RAE-1), which was prepared from the supernatant of transfected HEK 293T cells using standard purification procedures. Additionally, binding capacity for Fc-RAE-1 was determined on anti-RAE-1 column to be approximately 5 mg of fusion protein per mL of chromatographic support. The result confirmed the possibility of purifying at least 200 µg of antigen using 50 µL column format in a single chromatographic step, what should be sufficient amount of protein needed for N-glycosylation studies.
BIA-SEP, together with MEDRI, isolated a small quantity of antigen from lysates of RAE-1 transfected cells, as confirmed by western blot analysis and HPLC-MS/MS analysis. Thus, the method was successful for analysis of mammalian membrane proteins. Next, our efforts were focused to the optimisation of sample preparation as well as chromatographic protocol for immunoaffinity isolation of larger (micro to milligram) quantities of membrane proteins from eukaryotic cells. One of the main issues was prevention of column clogging during sample application. Due to the high complexity of a lysate sample, an aggregation of biological molecules occurred within the pores of monoliths, negatively influencing the monolith permeability. A working solution was the addition of an acidification step inducing precipitation of impurities following by filtration through a 0.22 um filter, what drastically improved column permeability during the sample loading as a first chromatographic step. (D.1.4.).
We then performed 24 successful purification rounds of RAE-1 protein from infected and uninfected cells, both primary MEF mouse embryonic fibroblasts and NIH 3T3 cell line. There were no detectable unspecific bands in elution from immunoaffinity column when we used a lysate of cells that do not express RAE-1 (negative control). Finally, after protocol optimisation, we succeeded in isolation of approximately 1 mg of transduced RAE-1 and approximately 200 ug of transfected RAE-1 (Deliverable D.3.4.). The RAE-1 protein generated by transduction was isolated from cells infected with a previously constructed recombinant mouse cytomegalovirus in which the m152 viral gene, otherwise responsible for RAE-1 downregulation, was exchanged for the RAE-1 gene (the virus was constructed by replacing the m152 ORF, in the BAC-cloned MCMV genome, with a cassette comprising RAE-1 ORF under the control of HCMV immediate-early (IE) promoter). Therefore, the isolated RAE-1 represents the RAE-1 form present in infected cells that can be compared to the RAE-1 protein from infected cells, where exactly the same RAE-1 ORF (GenBank accession no. AAI32033) was cloned into the pB45-Neo plasmid and used to transfect NIH 3T3 (ATCC CRL-1658) cells. Flow rate of 40 column volumes per minute was used to pump more than 100 mL of lysate through 50 µL column (see Figure 8), what is impressive result, enabling high-throughput purification of a low-abundant protein in micro to milligram scale in every biotechnological laboratory.
N-glycosylation analysis was performed for transduced RAE-1 (MEF cells) for the first time (D.1.4.). Coomassie blue-stained gel (the elution fraction from one of chromatographic purifications) was sent to partner MPI. After confirming the identity of the SDS-PAGE band to be RAE-1 by HPLC-MS/MS analysis, an in-gel deglycosilation was performed using PNGase F, followed by extraction of N-glycans from gel. The analysis of MULT1 as well as RAE-1 glycans could be performed only on an analytical scale due to low amounts of this glycoprotein obtained during isolation. The foreseen high-throughput analysis of these membrane glycoproteins was postponed due to low amounts of the protein obtained during isolation that were adequate only for the analytical-experiments. Therefore, optimization of the high-throughput purification protocol (UHPLC) coupled to MALDI-TOF MS glycan analysis was done by use of the model protein, namely yeast invertase, which is a yeast cell-surface protein that may be easily isolated in large amounts. In particular, we developed a chromatographic procedure that combines step elution with sample displacement chromatography on anion exchange CIM monolith for separation of glycoforms. The analysis of the model-enzyme invertase that contains more than 50% oligosaccharides by weight showed 13 N-glycosylation sites in the invertase amino acid sequence. It is known that high mannose N-glycans may locally contain an outer shell composed of oligo-mannose composed of up to 60 mannose units and glycoforms of enzyme bearing different amounts of negative charges attached to glyco-components, thus, exhibit different stability. For industrial application, only glycoforms with highest stability should be selected to increase the overall efficiency of the catalytic industrial process. The chromatographic procedure developed was therefore used for separation of invertase glycoforms that carry different negative charges densities from a natural matrix, namely S. cerevisiae. These invertase glycoforms were used in the course of the project for development of glycan hydrolyses protocols based on glycosydases immobilized at monolithic supports.
We optimized miniaturized supports with immobilized glycosidases for high-throughput glycosylation analysis by use of mass spectrometry. The immobilized glycosidases activity was tested by characterization of separated glycoforms and confirmation of the protein amino acid sequence by mass spectrometry. In addition, optimized protocols for mass spectrometry analysis were tested and proved to be efficient to yield evidence on differences between separated forms of invertase that are due to the variety of identified glycan structures. Analysis of N-glycan invertase structure was done in a high-throughput fashion. Results showed that outer shell oligo-mannose units are highly complex and that ionization efficiency obtained so far was expectedly rather low. Tested protocols proved efficient to yield evidence on differences between separated forms of invertase that are due to the variety of identified glycan structures. This means that different glycoforms can be successfully separated from mixtures (including those from natural sources) by developed UHPLC protocol and mapped by use of MALDI-TOF MS. Moreover, we delivered protocols and procedures for purification and separation of plasma IgG and IgM and their subsequent high-throughput glycomic analysis by use of mass spectrometry-based methods. Developed chromatographic procedures rely on combined step elution and displacement chromatography on monolithic matrices for separation of protein glycoforms (highly glycosylated enzyme invertase was utilized as a model protein). Major portion of results in this field provided evidence on N-glycoforms that carry different charge due to presence of charged glycans and that impact stability of glycoproteins.
Furthermore, in the frame of WP3, we performed N-glycan analysis using multiplexed capillary gel electrophoresis with laser-induced fluorescence (xCGE-LIF). The main goal was to annotate N-glycan structures released from mouse IgG. This was performed using exoglycosidase digestions and xCGE-LIF analysis. Additionally, N-glycans released from mouse plasma were analyzed and the major structures were annotated using the same method. Structural assignment of glycans has been reached through fractionation of glycan peaks and exoglycosidase digestions approach. About 48% of all N-glycans are of the type high-mannose, while really high percentage of the complex N-glycans contain no terminal galactose. Obtained results will help to expand knowledge about mouse IgG glycome which is still lagging behind human IgG glycome. Developed methods and gained know-how will allow us to understand the role of individual protein glycosylation pattern in health and disease.
Furthermore, we have prepared new samples of purified RAE-1, namely RAE-1 from transduced NIH cells, as well as RAE-1 from transfected NIH cells (for sample preparation scheme, cf. Figure 9 that is submitted separately). Unfortunately, due to the low amount of sample obtained form transfected cells not all exoglycosidase digests needed to completely characterize the differences between uninfected and infected samples have been performed (see Figures 10 and 11).
Optimisation of protein G, protein A and protein L immobilisation on CIM monoliths has been performed and was used for the construction of optimised protein G 96-well plates. In addition, 96-well plates for isolation of immunoglobulin G (IgG) and immunoglobulin M (IgM) with mounted anion-exchange (DEAE) mini-disk (100µL) followed by mini-disk with immobilized amino acid tryptophan were developed. Here, the scheme that was earlier developed in the project leader’s laboratory for both preparative and high-throughput analytical work was adopted and optimized for high-throughput and semi-preparative isolation of low-abundance proteins clotting factor VIII (FVIII) and von Willebrand (vWF) factor from human plasma by use of ion-exchange and affinity chromatography on monoliths. Also, protein L immobilized on monolithic discs was tested as an alternative capture agent to proteins A and G in protocols for isolation of IgG and IgM from serum. High-throughput laboratory analysis (TECAN robot), and mass spectrometers MALDI-ToF/ToF and LC-ESI-MS/MS were used for high-throughput isolation and analysis of glycoproteins. We have succeeded to prepare CIM monoliths with high utilisation of protein L or G and with dynamic binding capacities for IgG around 10 mg/mL. We have successfully developed a scheme for fast and simultaneous affinity purification of α (A), γ (G) and µ (M) immunoglobulins from human serum through protein A, G and L affinity monolith chromatography. A novel 96-well plate with immobilised protein G or protein L enables more than 50 purification cycles of plasma sample without losing its activity.
Because protein G has different selectivity towards different groups of antibodies, we developed and optimised an immobilisation of complementary affinity ligand – protein L. So far we have succeeded to prepare CIM monoliths with high utilisation of protein L and with binding capacities for IgG around 10 mg/mL, what is comparable with CIM protein G monoliths. Using immobilised protein L on monolithic support in combination with protein A or protein G monoliths enables simple fractionation of antibody pool. Additionally, protein L enables isolation of rat antibodies, which do not interact with protein A or G. The protein L-based monolithic columns and 96-well plates were tested and successfully utilised for high-throughput isolation of these antibodies within WP3. Additionally, this method was adapted for the application of laboratory robotics.
Within WP3, we determined N-glycan structure of rat IgG isolated using protein L by different analytical approaches (HILIC-UPLC and mass spectrometry). N-glycosylation characterization of rat IgG will enable further research on the impact of different factors (e.g. different drugs and diseases) on IgG glycosylation in animal models. Next to that, we characterized N-glycans in brain tissue samples from different brain regions in four different species: human, chimpanzee, macaque and rat. Obtained results will enable analysis of brain tissue glycosylation patterns specific to brain regions, species and stages of development.
In the frame of WP2, development of high-throughput glycoproteomic analysis, analytical approaches for glycosylation analysis of the purified proteins by UPLC were developed. Optimal separation conditions were established, and the exact composition of glycan structures within each UPLC peak was determined for IgG and IgM. We optimized protocols for purification of three individual glycoproteins from human plasma (fibrinogen, transferrin and haptoglobin) for which new affinity monolithic columns were developed within WP1. Different column batches in combination with purification conditions (buffer, volume of buffer, flow rate, quantity and quality of starting material, etc.) were tested in order to achieve successful and reproducible purification of each glycoprotein. 8-well (out of 96-well) plate for fibrinogen isolation was used for successful isolation and glycan analysis of fibrinogen from approximately 100 plasma samples. Variation of glycan composition of fibrinogen standards was shown to be smaller than variation of glycan composition of fibrinogen isolated from ten individuals, which implies notable natural variability of fibrinogen glycome. Order of glycoprotein isolation (IgG, fibrinogen, transferrin and haptoglobin) from the same plasma sample was optimised, and composition of glycan structures in each UPLC peak was determined. By development and optimization of purification and hydrophilic interaction liquid chromatography (HILIC) separation of N-glycans released from human fibrinogen, transferrin and haptoglobin, methods for glycoprotein analysis have been realized in full.
N- and O-glycans can be analysed via „Porous Graphitized Carbon“[PGC]-Liquid Chromatography [LC]. In order to establish this technique a sample preparation method prior mass spectrometry analysis for N-and O-glycans was introduced using standards. A Quadrupole-Time-Of-Flight (Q-TOF) hybrid mass spectrometer was used for the analysis of the self-produced standards, first, an intensive parameter optimization was necessary. The instrument settings had then to be further optimized for the nano-LC setup of the PGC-LC separation that was coupled online to the mass spectrometer. This included the determination and optimization of the complete set of MS and MS/MS parameters for N-and O-glycan analysis for PGC nano-LC tandem MS runs. After establishment, initial analyses of samples were successfully run. However, due to problems with the spray stability of this Q-TOF instrument (low acetonitrile concentrations in the separation gradient; negative-ion mode) the PGC nano-LC set up was also introduced to a second Q-TOF instrument from another vendor.
CGE protocol for analysis of APTS-labled glycans was optimized for a high-throughput format (96-well plates). Optimization of sample preparation included adjustment of amount of released glycans and ratio of dye to glycans. For enrichment and clean-up of the labelled glycans, HILIC and size-exclusion protocols were tested. The procedure of CGE run was optimized as well with regard to the total time of the run, length of capillaries used and injection conditions. Analysis of glycosylation of antigen-specific antibodies using CGE and LC-MS methods was performed. Capillary flow high performance liquid chromatography (with porous graphitized carbon as stationary phase) coupled via online electrospray ionization to tandem iontrap mass spectrometry with collision induced dissociation (capPGC-HPLC-ESI-IT-CID-MS(/MS)) based separation and subsequent fragmentation of released glycans was implemented. On close inspection with standards such as fetuin, few problems were noticed with the workflow. This resulted in a low resolution of the chromatogram of the glycan peaks that were analyzed which, in turn resulted in poor recovery of a population of highly sialylated fetuin glycans. Furthermore, analysis of the fetuin glycome revealed an underrepresentation of other sialylated glycans, which pointed to an over-acidification of glycans. After troubleshooting and introduction of a nano-flow PGC-HPLC-column with the appropriate modifications to the chromatography setup including installation of a new to the UPLC system, the installed system was able to detect with high sensitivity the representative glycans of various standards. The data was confirmed with xCGE-LIF.
Furthermore, a new MS-based fragmentation mode was introduced. It has been observed that using high-energy regimes (high normalized collision energy HCD) was sufficient not only to show sugar fragmentation, and thus confirm the presence of glycopeptides, but also the fragment data can be also be used to determine the sequence on which the glycan resides. The intention of this work was to apply the optimized glycomic and glycoproteomic platform to the analysis of influenza virus A/PR/8/34 that is derived from MDCK adhesion and suspension cell lines. Qualitative glycomic analysis of influenza A/PR/8/34.MDCKADH and influenza A/PR/8/34.MDCKSUS were performed. The intention was to create a glycan library from which to make the assignment of their residence on the influenza protein easier to assign.
All planned secondments, both from industry to academia and vice versa, were realized within WP4 and WP5. Following researches have been recruited within WP 6: Dr. Uros Andjelkovic (UNIRI), Dr. Blaž Nemec (BIA-SEP), Dr. Terry Nguyen-Khuong (MPI), Dr. Olga Schvechuk (MEDRI), Dr. Olga Zaitseva (GENOS). In the frame of WP 7 and in accordance with the project plan, training in various research methods occurred in every project partner institution as a part of the research process, both by home staff training visiting researchers, and by visiting MERs and ERs transferring knowledge to the receiving institution. Within WP 8, all researchers from the HTP-GlycoMet network were involved in various teaching and training activities, such as laboratory demonstrating, tutorials/seminars; undergraduate project supervision and PhD project supervision; supervision of technical staff, in accordance with the project plan. Career planning workshop was not organized as a separate event, but as a part of the partner meetings.
Different trainings in scientific and technical methods applied for isolation of proteins from human and animal plasma, as well as hydrophobic membrane proteins were organized during three scientific meetings (Monolith Summer Schools and Monolith Symposia 2014 and 2015 and Glyco23 International Symposium), and researchers that were recruited by UNIRI and Genos in the frame of this project were participating and also teaching there. Our collaborative project was presented to the publics by different interview and presentations given by the project coordinator and group leaders (WP9), and project management administrative support and foreseen meetings were organized and reports were submitted on time (WP10).
The potential impact and the main dissemination activities and exploitation of results
The project deals with a very a very specific research field confined to glycoproteins and glycan structures in general. This research field is in close relationship with advanced in analitical methods and development of dedicated databases for glycan structures assignements. With inovative tools, the field would immediately boost the biomedical research area. Indeed, it is widely recognized that disease, including viral infection, cancer and chronic disease of aging population have been witnessing a steady stagnation in the field of new therapeutic approaches since years. Glycoproteins are major biomarkers of these groups of disease and advancements in their mapping in general might provide major breakthroughs in the field of improved diagnostics, prognostics and development of current and new therapies.
The HTP-GlycoMet project is a collaborative effort of three academic and two SME partners with expertise in complementary fields of science and technology. Individual partners have in the recent past been involved in several bilateral collaboration projects, but these have been of limited scope. The project implemented a high-level research plan aimed at bringing together the scientific expertise and infrastructural capacities of the partners and thus result in three immediate sets of outcomes. The first benefit of the project is an increased knowledge of complementary scientific fields: for example, Genos will obtain from UNI-RI and MEDRI further insights into the role of protein glycosylation in health research, with an emphasis on viral infections. On the other hand, from SMEs involved in the project UNI-RI and MEDRI will acquire reliable and tailor-made methods for specialised chromatographic tools (BIA-SEP) and high-throughput glycomic analysis (Genos). In addition, new ultra-sensitive xCGE-LIF methods for glycoproteomic analysis were developed by joint efforts of BIA-SEP, Genos and MPI. Thus, an important benefit is the increased research and business performance of the partners involved. Second, the realization of the HTP-GlycoMet project resulted in an extensive exchange of personnel between the commercial and academic partners as well as the recruitment of excellent experienced researchers and consequently, significantly strengthen the human resource capacity of all partners. Third, the transfer of knowledge between academic and industrial partners yielded in further development of innovative products and services. This, together with the actions like joint applications to additional EU and other calls for proposals, will ensure the continuation of collaboration beyond the lifetime of the project, and the establishment of new inter-sectoral collaboration projects. This consortium intends to gradually expand to other academic and industrial organisations involved in the research of various pathophysiological processes with the aim of providing glyco-analytical tools to the biomedical community. Resources which are generated through this project will provide proof of principle for the application of the developed glyco-analytical technologies for the study of immunity and infection. This will be continued through joint applications of HTP-GlycoMet partners to new calls expected in Horizon 2020 and other funding schemes. In the wider sense, collaboration with industrial partners from the Zagreb region (Croatia) and Vipava Valley region (Slovenia) will considerably improve the utilisation of technologies available at UNI-RI and MEDRI (Rijeka). This is relevant taking into account the lack of biotech SMEs into which these technologies could be successfully transferred in the immediate surroundings of the city of Rijeka. The implementation of this project could also boost knowledge transfer activities from UNI-RI and MEDRI to the business sector in the future. Owing to the extensive engagement of two research-intensive SMEs in this project, a new line of products and services is now in development. By a large part these is a collaborative effort which will require continuous future collaboration. The relevance of the project for the enhancement of innovation potential of the European Research Area is closely tied to the fact that after completion of the human genome project, and the glycoproteomics has become one of the next large challenges for biomedical science. It is now evident that genome is only a basic database of life and that several layers of complex omics integrate to form life in its full complexity. The same is true for complex human diseases. Glycoproteomic analysis is still globally deficient and is more and more required for research, but also in quality control and development of new biological drugs. Over 200,000 individuals were included in large genomic studies in the past 5 years, and many of these cohorts also have plasma samples available. In addition, many new cohorts are currently being collected and they are all highly interested in further phenotypisation of their patients, what is a great research and market potential. Moreover, newly developed technologies will be applied to study function of several proteins which are important in the process of viral infection (WP3). The expertise in antibody and fusion protein immobilisation was successfully deployed to extract the immunologically relevant membrane glycoproteins. This will greatly improve our current understanding of the interaction of the viral proteins and their ligands as well as the importance of their glycosylation profile. The impact on the innovation potential of the European Research Area is significant because of the identification of possible targets for diagnostics and treatment of viral infections and immunological processes in general. Second, UNI-RI and MEDRI will benefit from secondments to BIA-SEP, since these will be aimed at acquiring specific know-how on potential avenues for commercial exploitation of the existing unique collection of antibodies generated at MEDRI. These activities are expected to result in improved capabilities for commercialization of university-generated technologies and industry-oriented R&D at UNIRI and MEDRI, which currently have no spin-off companies in the field of biotechnology. A better exploitation of the strong biomedical science base available at these institutions will contribute to strengthening of the innovation potential of the European Research Area. Apart from the training in state-of-the-art laboratory techniques, training in complementary skills, such as presentation, proposals writing, entrepreneurship, business planning and leadership skills will also be organised. This will enhance the competences and career prospects of the recruited and seconded researchers. This has become increasingly important in the present science system, which demands from researchers the continuous transfer of basic research results into the commercial sector, or understanding how they can transform ideas into innovative products and services that address European and global health challenges. In addition to secondments, workshops on basic CIM technology and high-throughput affinity purification that will be organised in frame of the project will significantly facilitate the sharing of knowledge between the participants and external lecturers and participating researchers. B.5.2. Plans for exploitation of results and Dissemination strategy Exploitation of results Participation of two companies which are already global leaders in their respective fields warrants the successful exploitation of the research results. New technologies developed in the HTPGlycoMet network will enable Genos and BIA-SEP (in collaboration with UNI-RI, MEDRI and MPI) to preserve their current positions of market leaders in this rapidly advancing field by securing enabling IPR for the whole series of products and services. BIA-SEP is successfully selling a wide range of monolithic chromatographic support tools in EU, US and Asia. The development of new affinity 96- well products for rapid high-throughput purification of glycoproteins will enable them to enter the fields of high-throughput research and monitoring of production of recombinant proteins (therapeutic antibodies, other biological drugs, enzymes for cosmetic industry, etc.). Both of these fields desperately require methods for rapid high-throughput analysis of protein glycosylation and are promising markets, which will only grow in the future. Genos Ltd is currently the largest commercial provider of highthroughput glycomic analysis with sales in Europe, US, Australia and China. The majority of analyses are being done for researchers conducting a large population or patient stratification studies that need better characterisation of their patients. The demand for glycomic analysis is increasing and to preserve its leading position, Genos needs to develop new innovative products, which cannot be easily copied by low-cost companies in Asia. The third potential line of exploitation of research results stems from the experiments planned in WP3. In this work package important physiological processes during viral infection will be studied from a new angle enabled by technological development in WP1 and WP2. Contrary to WPs 1 and 2, where planned deliverables have clear potential for commercial exploitation, in WP3 it is not possible to predict whether the outcome will have some commercial potential or not. However, the results of this work package will be closely monitored and all potentially exploitable IP will be properly protected. The development of potential new therapeutic targets for combating viral infection exceeds the exploitation potential of our SME partners. Dissemination Among network participants. Scientific and training progress will be presented at the annual network meetings. In addition, the project web site will have a confidential section, which will be restricted to project participants and will contain confidential information relevant for the realisation of the project. To the scientific community / biotechnology industry. The dissemination of the results during and after the project will be through publications in high impact peer-reviewed journals and in contributions to scientific conferences (including three workshops on high-throughput technology that will be organized jointly with The Monoliths conferences in 2014 and 2016 and IGO congress in 2015 as detailed in WP7 and WP9). Partnerships with industry and European networks will be fostered, in particular at meetings and other events within the partnership’s own projects. To the general public. A dedicated web site will be set up for the project by IT services of UNI-RI that will contain all relevant and up-to-date information about the project progress, public research reports and relevant publications. Relevant media will be informed about the important project achievements and will be invited to interview project team members. Preparation of articles for the popularisation of science and participation of recruited and seconded researchers and other team members in different events, such are summer schools, festivals of science, and open door events, will help promote and better understand science and technology, in particular in the student population (from elementary schools, high schools and universities). Each recruited fellow will contribute to at least one outreach activity per year. As required by Annex II of the grant agreement, we will ensure that all publications and presentations by members of the project consortium - including all funded fellows - acknowledge the EU financial support received. This acknowledgement
will specifically refer to the Marie Curie Industry-Academia Pathways and Partnerships action, as well as the project number and acronym.
International conferences/events open to external researchers:
Monolith Summer School (2014), Portorož, Slovenia - BIA-SEP and UniRI
International Symposium on Glycoconjugates, (2015), Split, Croatia - GENOS
Monolith Summer School (2016), Portorož, Slovenia - BIA-SEP and UniRi
GlycoBiotech (2017), Berlin, Germany – MPI
Address of the project public website: www.glycomet.uniri.hr/en
.