Cupriavidus necator, an alternative source for isotopic enrichment of proteins expressed in insect cells for NMR investigations

Abstract

Isotopic enrichment of pharmacologically relevant protein targets is crucial for structural studies by nuclear magnetic resonance (NMR) and plays a key role in advancing structure-guided drug discovery. Many clinically important drug targets require expression in eukaryotic systems—such as mammalian, yeast, or insect cells—rather than prokaryotic hosts. This requirement limits the feasibility of high-throughput isotopic labeling and poses challenges for obtaining uniformly isotope-labeled proteins suitable for NMR analysis. While several enrichment strategies have been developed, no broadly applicable enrichment platform has emerged for eukaryotic expression systems. In this study, we introduce Cupriavidus necator as an alternative biological source for ¹⁵N and ¹³C isotopic enrichment to support protein production in eukaryotic systems. To evaluate this approach, we selected the kinase domain of EPHA2, a receptor tyrosine kinase implicated in colorectal cancer progression and an important target for therapeutic inhibitor development. Isotopic incorporation was quantified using liquid chromatography-mass spectrometry (LC-MS), revealing enrichment levels of 79% for ¹⁵N and 69% for ¹³C. These results demonstrate that Cupriavidus necator can serve as a robust and flexible platform for generating isotopically enriched biomolecules compatible with eukaryotic protein expression, thereby enabling NMR investigations of disease-relevant protein targets.

Introduction

Driven by rapid advancements in structural biology and the increasing integration of multimodal methodologies, the field is progressively unraveling the architectures of complex biomolecular assemblies across a continuum of spatial resolutions: from atomic-scale detail to cellular context (Schwalbe et al. 2024). This progress has driven a substantial increase in scientific interest aimed at deciphering the structural mechanisms underlying the functional dynamics of biomacromolecules. Preparing biomacromolecular samples is a fundamental requirement for any structural analysis. This process often involves either isolating biomolecules from natural sources, which can pose significant limitations, or employing recombinant expression technologies. Recombinant protein expression predominantly utilizes prokaryotic systems, particularly Escherichia coli (E. coli), due to its numerous advantages, including genomic manipulation, ease of cultivation, minimal equipment and time requirements, and cost-effectiveness. Furthermore, E. coli allows for the incorporation of selenomethionine residues for X-ray crystallography and mass spectrometry, as well as selective or uniform (such as ¹⁵N, ¹³C and ²H) labeling of proteins, which is crucial for multidimensional nuclear magnetic resonance (NMR) and quantitative mass spectrometry (MS) analyses. However, many human proteins of pharmaceutical interest, including membrane proteins, those requiring post-translational modifications, or large macromolecular complexes, are often unsuitable for expression in bacterial systems. For drug discovery research, it is critical to obtain correctly folded proteins that accurately represent their physiologically active conformation. Consequently, targets that are large in size or require complex post-translational modifications, such as phosphorylation, glycosylation, or the formation of multiple disulfide bonds, are typically expressed in eukaryotic expression systems. These systems include yeast (e.g., Pichia pastoris, Saccharomyces cerevisiae), mammalian cells (e.g., CHO, HEK), or insect cells (e.g., Sf9), which are better suited for producing functionally relevant proteins.

The incorporation of stable isotopes (SI) into target biomolecules is essential for the structural and functional investigation of macromolecules via NMR and quantitative mass spectrometry. However, achieving uniform SI enrichment in eukaryotic expression systems continues to present technical hurdles, primarily stemming from metabolic complexity and relatively modest incorporation efficiencies, which typically range from 75 to 80%. While foundational research established numerous protocols designed for broad applicability (Brüggert et al. 2003; Egorova-Zachernyuk et al. 2009, 2011; Gossert et al. 2011; Gossert and Jahnke 2012; Saxena et al. 2012; Skora et al. 2015; Sitarska et al. 2015; Opitz et al. 2015; Zhang et al. 2017; Franke et al. 2018), recent efforts have shifted toward more specialized and economically viable strategies. For instance, the move toward using mammalian cells for disease-relevant proteins (Rößler et al. 2024) and the development of targeted side-chain labeling (Rosati et al. 2024) have improved the feasibility of studying complex proteins in native-like environments. Additionally, sparse labeling techniques using ¹³C-glucose have facilitated the study of glycoproteins (Rogals et al. 2021), while metabolic studies in insect cells have identified glutamine as a critical factor for optimizing both yield and enrichment (Wu et al. 2023). Despite these advancements, several of the existing protocols for uniform labeling still rely on the total replacement of the nitrogen and carbon ingredients of the media with their labeled counterparts. To overcome the technical demands of this methodology, alternative sources must be explored to streamline efficiency and broaden accessibility.

The use of ¹³C-labeled CO₂ serves as an economically viable carbon source for isotopic enrichment. In this process, the facultative autotrophic bacterium Cupriavidus necator (C. necator, previously named as Ralstonia eutropha) utilizes redox-driven chemical energy to fix CO₂ via the Calvin cycle, facilitating the incorporation of the ¹³C isotope into its metabolic products (Yabuuchi et al. 1995; Shimizu et al. 2015). This microbial platform has been utilized (e.g., by Silantes) to produce stable isotope-labeled biomass enriched with ¹³C and ¹⁵N, which has subsequently served as a nutrient source for E. coli and yeast expression system (Krüger and Heumann 2014). Such biomass-derived media have proven effective in proteomic applications, particularly for protein tracing in mouse models via ¹⁵N-labeled mouse diets and mass spectrometry (Frank et al. 2009; Dethloff et al. 2018). Despite these developments, the efficacy of C. necator-derived media for producing uniformly labeled proteins in higher eukaryotic systems has yet to be fully investigated.

To evaluate stable isotope (SI)-labeled protein expression using C. necator hydrolysate, we employed the baculovirus-insect cell (Spodoptera frugiperda, Sf9) system, a well-established platform for structural characterization of proteins via NMR and X-ray crystallography. Our group has previously demonstrated successful uniform and selective ¹⁵N labeling of multiple proteins in insect cells(Saxena et al. 2012). For this study, we selected the intracellular kinase domain of the ephrin type-A receptor 2 (here after abbreviated as EPHA2) tyrosine kinase as a model protein to assess the performance of the Cupriavidus-based medium. EPHA2 plays a critical role in the development and progression of various cancers, including colorectal cancer, making it a compelling target for detailed structural investigations to aid drug discovery efforts.

We conducted comparative studies using the following commercially available isotope-labeling media: (i) Cortecnet ¹⁵N-labeled yeast extract, and (ii) Merck ISOGRO Algal extract. These media were evaluated against the newly developed C. necator biomass and hydrolysate. Key cellular parameters, including growth behavior and cell morphology, were systematically assessed in insect cells, which demonstrated significant sensitivity to changes in media composition. Under optimized conditions, cells transitioned from standard growth media to Cupriavidus-based hydrolysate medium were tested for expression following viral infection. The intracellular kinase domain of EPHA2 was successfully expressed, with the resulting isotopic enrichment profile analyzed via NMR spectroscopy and LC-MS. Evaluation of the ¹⁵N and ¹³C enrichment yielded 79% and 69%, respectively. High-resolution NMR analysis of the ¹⁵N-labeled EPHA2 confirmed a native-like fold, serving as a key indicator of the protein’s conformational stability when produced using this method. These results highlight the efficacy of Cupriavidus-derived hydrolysate as a robust medium for isotopic labeling in eukaryotic expression systems.

Materials and methods

Plasmid and construct design

This study used the human receptor tyrosine kinase EPHA2’s catalytic kinase domain as a model protein to assess the stable isotope incorporation efficiency. The design of the gene construct was previously described (Gande et al. 2016). In short, EPHA2 kinase domain coding gene (NP_004422, amino acids D596 to G900) was optimized for expression in insect cells and synthesized at GenScript Inc., USA. Furthermore, they sub-cloned it into a modified pTriEx 1.1 transfer vector (Novagen), which provided an N-terminal Flag-His fusion tag and TEV protease recognition sequence to the human EPHA2 kinase domain.

Insect cells and generation of recombinant baculovirus and amplification

In this study we used Sf9 cells from Invitrogen adapted to Sf900II or III medium and Sf9 cells adapted to ESF 921 medium from oxford expression technologies (OET Sf9 cells). The recombinant baculovirus containing the target gene EPHA2 was generated by co-transfecting the insect cells in a 6-well plate with midiprep purified pTriEx 1.1-EPHA2 transfer vector and BacMagic™ -3 DNA (Novagen) according to the user protocol TB459 (and modifications as described earlier (Gande et al. 2016). After one week of transfection, the initial seed stock EPHA2 recombinant baculovirus P₀ was harvested. Further, amplification of the recombinant virus was performed by infecting fresh Sf9 cells in suspension culture with P₀ recombinant baculovirus and incubating the cells in the shaker at 28 °C for a week. When the infected cells appeared larger, granular, or showed partial cell lysis, the medium was harvested by centrifugation at 1000 x g for 20 min at 4 °C to collect the P1 recombinant baculovirus. Similarly, P2 recombinant virus was generated from P1 recombinant virus, which was then used for EPHA2 expression.

Media preparation and composition

Different growth media from various vendors were used in this study. They are SF900 III, Sf900 II from Invitrogen; ESF921, ESF921 w/o amino acids (ESF921Δaa) from Oxford expression systems; and SF4 baculo growth express, SF4 w/o aa & yeast-extract (SF4ΔaaΔYE), SF4 w/o aa, yeast-extract glucose, maltose & sucrose (SF4ΔaaΔYE ΔGSM) from Bioconcept. Sf9 cells were maintained in SF900 III and OET Sf9 cells in ESF921 complete growth media, supplemented with 5% FBS (or dialyzed FBS) and 0.5% Pen Strep (Gibco™ Penicillin-Streptomycin, 10000 U/ml). Protein expressions were performed in different depletion media prepared as shown in the table below.

Sf9 cell growth in 24 deep-well block

To perform a comparative study across various growth media, we performed a high-throughput expression screening conditions in 24 deep-well blocks. We monitored Sf9 cell growth kinetics in 24 deep-well block vs. Erlenmeyer flask, by seeding Sf9 cells with similar cell densities (0.5 × 10⁶ cells/ml), but in different volumes (3 ml and 20 ml, respectively). The block was placed on a mini shaker at 250–300 rpm in a static 28 °C incubator and the Erlenmeyer flask in a 28 °C incubator with 80 rpm. For the next five days, small aliquots of the cells were taken from both cultures and performed a cell count with the Neubauer counting chamber.

Protein expression tests with different extracts

Small-scale expression tests were performed either in 24 deep-well blocks or in Erlenmeyer conical flasks depending upon the volume of expression medium. Sf9 cells were centrifuged and seeded at high cell density (~ 6 million cells per ml) in the desired medium and infected with 2% P2 EPHA2 recombinant virus. After three days of post infection cells were observed under a microscope, harvested, and checked for the expression of the target protein EPHA2 by Western blotting. In the case of large-scale expression, cells were harvested and stored at -80 °C until processed for protein purification.

Western blots

Cells from the test expression were collected by centrifugation of the cells from 24 deep-well block. Cells were suspended in ice-cold PBS buffer with protease inhibitor and lysed with an ultrasonic water bath (20 s, 5 cycles, intermittent cooling on ice). Roti quant Bradford assay was performed to determine the total protein concentration in the cell lysates and equal amounts of the cell lysates were loaded on pre-casted NuPAGE 4–12% Bis-Tris Gels for immunoblottings (used MES SDS running buffer 50 mM MES, 50 mM Tris base, 0.1% SDS, 1 mM EDTA, pH 7.3). Total cell proteins resolved by SDS-PAGE were then transferred onto PVDF membrane activated by methanol with a semi-dry Western blot transfer device. Protein blots are probed with anti-His mouse monoclonal primary antibody (Thermo; dilution 1:3,000) and HRP conjugated Goat anti-mouse secondary antibody (Dianova/Jackson Immunoresearch; dilution 1:5,000). Immunoblots were incubated with the luminol-based chemiluminescent substrate to detect HRP-conjugated antibodies by a Chemiluminescence imaging system/Lumicapture.

Protein expression and purification

OET Sf9 cells were cultivated in ESF921 medium (Oxford Expression technology) to high cell densities (~ 6 to 10 × 10⁶ cells/ml, 1-liter) which were used directly by dilution (~ 6 to 10 × 10⁶ cells/ml, 1-liter) with fresh medium for unlabeled expression or pelleted down at 1,000 x g/5 min at room temperature for ¹⁵N/¹³C/¹⁵N¹³C isotope labeling. For the initial labeling experiments in ESF921 Δaa medium with the respective isotope extracts, pelleted cells were transferred into ESF921 Δaa depletion medium and starved for 2 h at 28 °C with shaking (80 rpm). The cells were then pelleted again and transferred into ESF921 Δaa medium containing the respective isotope extracts.

For the later labeling experiments in SF4 depletion media, cells grown in ESF921 complete medium were pelleted and washed twice with the appropriate depletion medium (SF4 Δaa, SF4 Δaa ΔYE, or SF4 Δaa ΔYE ΔGSM). The cells were then transferred into the corresponding SF4 depletion medium containing the respective isotope extracts. In some cases, cells were first starved for 1 h in SF4 Δaa ΔYE ΔGSM before being transferred into SF4 Δaa ΔYE ΔGSM supplemented with ¹⁵N, ¹³C Cupriavidus protein hydrolysate.

Once OET Sf9 cells were in the appropriate expression medium, they were infected with P2 recombinant baculovirus (2% v/v) to express the target protein, EPHA2. The cells were harvested after 3 days of post-infection by centrifugation at 2,500 x g/15 min at 4 °C. Cell pellets were stored in -80 °C or processed immediately for protein purification.

For protein purification cells were suspended in lysis buffer (50 mM Tris pH 8, 500 mM NaCl, 5 mM MgCl₂, 5 mM DTT, 5 mM imidazole) with protease inhibitor cocktail tablet (Complete EDTA free, Roche) and lysed by passing through microfluidizer with 15,000 PSI, 5 cycles. Cell lysates from the microfluidizer were centrifuged at first at 43,000 x g with JLA 16.250 for 45 min at 4 °C and subsequently, the partially cleared supernatant was centrifuged further with ultracentrifuge Ti45 at 100,000 x g for 60 min at 4 °C. Finally, the clarified supernatant from the ultracentrifugation was filtered with (0.45 µM) and loaded onto 5 ml HisTrap FF or HP column equilibrated with Ni buffer A (20 mM Tris pH 8, 200 mM NaCl, 5 mM MgCl₂, 5 mM DTT, 35 mM imidazole) on an ÄKTA purifier system. Non-specifically bound proteins were washed away with 5% Ni B buffer (i.e. 35 mM imidazole) and eluted with a gradient elution from 5% to 100% Ni buffer B (20 mM Tris pH 8, 200 mM NaCl, 5 mM MgCl₂, 5 mM DTT, 500 mM imidazole). Fractions from affinity chromatography were analyzed on SDS-PAGE by Coomassie Brilliant Blue staining and fractions containing His-Flag-Tev-EPHA2 protein were pooled and treated with TEV protease while dialyzing against Ni A buffer without imidazole. Next, we performed inverse NiNTA purification to separate the His-Flag- fusion tag from the target protein EPHA2, and as a final polishing of the target protein we carried out size-exclusion chromatography (Sect.  75) with the buffer containing 20 mM Tris pH 8, 200 mM NaCl, 5 mM MgCl₂, 3 mM TCEP. Finally, the fractions of EPHA2 (34 kDa) were concentrated using Viva spin (GE) concentrators with 10 kDa cut-off and stored at -80 °C.

NMR spectroscopy

Spectra were acquired at 298 K on a Bruker 800 MHz spectrometer equipped with a cryo TCI ¹H [¹³C, ¹⁵N] probe using the pulse sequence hsqcfpf3gpphwg. A 200 µL protein sample containing 130 µM EPHA2 kinase domain was prepared in 20 mM Tris-HCl (pH 8.0), 200 mM NaCl, 5 mM MgCl₂, and 3 mM TCEP, with 10% (v/v) D₂O included for field-frequency locking and measured in a 3-mm NMR tube. Spectra were recorded with 40 scans per increment, 2048 complex points in the direct dimension (F2), and 224 increments in the indirect dimension (F1). Chemical shifts were referenced to 1 mM 3-(trimethylsilyl)-2,2,3,3-tetradeuteropropionic acid (TSP-d₄) used as an internal standard. NMR data were processed and analyzed using TopSpin version 4.5.0 (Bruker BioSpin).

Mass spectrometry (LC-MS)

Digestion

In-gel

Gel bands were digested with the In-Gel Tryptic Digestion kit (Thermo Fisher Scientific) with minor adaptations. The excisions were destained, reduced and alkylated according to the manufacturer’s protocol and digested overnight with trypsin (SERVA) at room temperature. The digested peptides were transferred to a clean tube and the solvent was evaporated in a speed vac (Eppendorf). The dried peptides were stored at -20 °C until LC-MS analysis.

In-solution

About 50 µg protein were loaded per filter unit (Microcon-10, Merck Millipore). After addition of 200 µL UA solution (6.5 m Urea in 0.1 m Tris/HCl pH 8.5) to the filter the sample was centrifuged at 14 000 × g for 20 min. This step was repeated twice and the flow through was discarded carefully. The sample was incubated at room temperature with 100 µL UA solution containing 10 mm dithiothreitol (DTT) using a thermomixer for 30 min in the dark. The filter unit was centrifuged at 14 000 × g for 20 min and the flow through was discarded carefully. Subsequently 100 µL of UA solution containing 95 mm Iodoacetamide (IAA) was added and incubated at room temperature using a thermomixer for 30 min in the dark. The filter unit was then centrifuged at 14 000 × g for 20 min and the flow through was discarded. The filter was washed three times with 100 µL of UB solution (6.5 m Urea in 0.1 m Tris/HCl, pH 8) and centrifuged at 14 000 × g for 20 min. The filter unit was incubated with 40 µL of UB containing Lys-C in a ratio of 1:50 in the dark at room temperature overnight. The unit was transferred to a new collection tube and 100 µL of ABS solution containing trypsin in a ratio 1:100 were added and the unit was incubated in the dark at room temperature overnight. The unit was centrifuged at 14 000 × g for 15 min. Then 50 µL of 0.5 m NaCl was added and the filter unit was centrifuged at 14 000 × g for 15 min. The filtrate was acidified with 0.1% TFA and was desalted by C-18 ZipTips (Merck Millipore) according to the manufacturers protocols and dried in a SpeedVac. The dried peptides were stored at -20 °C until LC-MS analysis.

LC-MS data acquisition

The dried peptide fractions were dissolved in 5% acetonitrile with 0.1% formic acid, and subsequently loaded on reverse phase columns (trapping cartridge: 5 μm C18-beads, L = 5 mm, inner diameter = 100 μm; Thermo Fisher Scientific, Bremen); analytical column: ReproSil-Pur C18-AQ 1.8 μm beads, 75 μm inner diameter, L = 50 cm (PepSep, Marslev, Denmark) using a nano-HPLC (Bruker nanoElute). Eluted peptides were separated over a 90 min gradient of water (buffer A: water with 0.1% formic acid) and acetonitrile (buffer B: acetonitrile with 0.1% formic acid). All LC-MS-grade solvents were purchased from Fluka. Gradients were ramped from 2% to 35% B in 90 min at flowrates of 300 nL/min. Peptides eluting from the analytical column were ionized online using a Bruker CaptiveSpray ESI-source and analyzed in a Bruker Impact-II mass spectrometer. Mass spectra were acquired over the mass range 150–2200 m/z, and sequence information was acquired by a computer-controlled, dynamic method with a fixed cycle time of 3 s and an intensity-dependent acquisition speed for MS/MS-spectra between 8 and 20 Hz (instant expertise-mode) of the candidate ions.

Database searches

The resulting files were recalibrated, and database searches were performed using the MaxQuant search engine (1.5.3.8) against a combined database containing the host cell proteome (Spodoptera, tax_id:7106, downloaded 17-08-2017), the construct sequence and common contaminants. For database searches, precursor mass tolerance was set to 20 ppm in the primary analysis and 4.5 ppm in the full search, with fully tryptic cleavages with up to two missed cleavages. Oxidation of methionines residues (+ 15.995) and acetylation of protein N-terminal (+ 42.011) were set as variable modifications and carboxyamidomethylation (+ 57.021) as a fixed modification. Peptide and protein FDR was set to 0.01.

The mass spectrometry proteomics data have been deposited to the ProteomeXchange Consortium (http://proteomecentral.proteomexchange.org) via the PRIDE partner repository (Perez-Riverol et al. 2025). Anonymous reviewer access is available upon request.

Calculation of isotopic labeling efficiency (¹⁵N, ¹³C and ¹⁵N+¹³C labeling)

First, target-protein derived peptides were selected in unlabeled samples based on high-quality, unambiguous fragment spectra for identification, peptide intensity and chromatographic elution. In addition, all peptides were manually curated to ensure no interfering other peaks were present within the expected isotopic envelope. For quality control. all spectra were curated manually, confirming that the observed isotopic patterns corresponded to the expected charge states and no interfering ions were present, ensuring the reliability of isotope quantification. Isotope incorporation was then calculated using intensity-weighted centroids comparing the observed isotopic distributions to the respective uniformly-labeled theoretical masses, as outlined below.

A list of 20 high-quality candidate peptides were manually selected for further processing. The incorporation of ¹⁵N and ¹³C was determined by comparing the measured isotopic envelope distributions to the respective theoretical mass of the unlabeled and heavy isotope-labeled peptide, which is a common strategy in HDX-MS (Masson et al. 2019). In brief, extracted ion chromatograms were created for high quality peptide candidates (peptide identification and protein FDR 0.01, intensity > 10^e4, no interfering isotopic envelopes by other compounds in control samples). Then, peak areas were integrated for all assigned isotopic peaks for each peptide based on the maximum possible incorporation for ¹⁵N and ¹³C, and intensity-weighted centroids were calculated. All assigned ion spectra were manually inspected and revised, and candidates with interfering isotopic envelopes from other compounds in the labelled samples were discarded. ¹⁵N/¹³C incorporation was then calculated by comparing the difference in intensity-weighted centroids compared to the unlabeled control and the fully labeled theoretical mass. At least 5 unique peptides were used for calculation of the ¹⁵N/¹³C uptake. The full code is available on PRIDE together with the raw data.

Results

Optimization of Sf9 insect cell culture using deep-well blocks and C. necator-derived media supplements

To enable high-throughput optimization of insect cell culture conditions for protein expression, we adapted a miniaturized Sf9 culture system using 24-deep-well blocks (Bahia et al. 2005) and validated its performance against conventional 125 ml Erlenmeyer flask cultures (3 ml vs. 15 ml culture volumes respectively) with Sf-900 III medium supplemented with 5% FBS and 0.5% Penicillin/Streptomycin. Growth kinetics were comparable between the two formats, confirming that the scaled-down system maintains equivalent cell proliferation and viability (Fig. 1A). This miniaturized platform significantly reduces reagent usage, labor, and cell input while enabling the parallel testing of multiple conditions, making it well-suited for systematic optimization of baculovirus-mediated protein expression workflows.

Fig. 1.

Sf9 cell growth and proliferation in various growth media. (A) Sf9 cell growth performance in 125 ml Erlenmeyer flask cultures (15 ml volume) versus 24-deep-well blocks (3 ml culture volume) in Sf-900 III medium supplemented with 5% FBS and 0.5% Pen/Strep. (B to I) Microscopic images of Sf9 insect cells cultured in media with different supplement conditions in a deep-well plate. The image shows cells with typical rounded morphology and uniform distribution, indicative of healthy proliferation. Cell density and morphology were assessed to determine the optimal supplementation strategy for isotopic labeling experiments in biomolecular NMR applications. Image captured at 100x magnification

Using this format, we initially evaluated the efficacy of C. necator-derived protein hydrolysates (CPH) and biomass extracts as alternative media supplements by performing a comparative growth analysis with established supplements, including algal extract (ISOGRO, Merck), YE, Spirulina extract (Silantes), and complete media (Sf-900 III, ESF921). For this analysis, ESF921 Δaa medium, a basal medium devoid of amino acids, was supplemented with 1% of either algal, bacterial or YE and further enriched with tryptophan, NH₄Cl, glucose, & YE, following prior studies on the acid hydrolysis-based preparation of algal extract for improved growth outcomes (Sitarska et al. 2015). This analysis revealed robust proliferation occurred in complete media, while cells grown in C. necator-supplemented media achieved growth levels comparable to those in algal or YE enriched formulations.

However, cell proliferation in these 1% supplemented media (Fig. 1D-I) remained suboptimal compared to that in complete media (Fig. 1B, C). This reduced performance may stem from a lack of cellular adaptation to the unique composition of the bacterial extracts, which appeared to exert cytostatic rather than cytotoxic effects. Notably, 1% C. necator extract supplementation consistently supported cell viability and moderate growth without adverse effects, aligning with effective concentrations used in isotope-labeling protocols involving algal or yeast hydrolysates. These findings underscore the value of the deep-well block platform for scalable, resource-efficient optimization of insect cell culture conditions and highlight C. necator extracts as a promising, non-toxic supplement and may further enhance their utility for high-yield recombinant protein production and stable isotope labeling.

Model protein (EPHA2) expression in various media

To evaluate the suitability of Cupriavidus-derived supplements for insect cell culture and their potential future use in isotope labeling workflows, expression of the model protein EPHA2 was analyzed in Spodoptera frugiperda (Sf9) cells using various media conditions (Fig. 2). EPHA2 carried a N-terminal Flag-His-Tev fusion tag and expression was assessed by Western blotting using an anti-His antibody.

Fig. 2.

Western blot analysis of EPHA2 expression in Sf9 insect cells using anti-His antibody. Expression of His-tagged EPHA2 was evaluated under various supplementation conditions in Sf9 cells. All samples were prepared identically from 3 ml of expression culture by resuspending a 500 µl cell pellet in 200 µl lysis buffer with protease inhibitor and cell-lysis with ultra sonification. Loaded 100 µg (Top left, Invitrogen Sf9 cell lysate) or 50 µg (Top left & Bottom) total protein from each lysate on SDS-PAGE. Western blot detection was performed using an anti-His antibody to assess recombinant protein expression levels. (Top Left): EPHA2 expression in Sf9 cells (Invitrogen) cultured with different media supplements, including Spirulina biomass, SPH, Cupriavidus biomass, CPH, YE, algal extract, Sf900III, and ESF921. (Top Right): EPHA2 expression in Sf9 cells (OET) under the same supplementation conditions as in the left panel. (Bottom Left): Comparison of EPHA2 expression in OET Sf9 cells in the absence (−) and presence (+) of fetal bovine serum (FBS) for each supplement condition. (Bottom Right): EPHA2 expression in complete media

Initial small-scale expression experiments were performed using in-house Sf9 cells (Invitrogen) in a 24-well format (3 ml culture volume). Cells were transferred into amino acid-depleted ESF921 (ESF921 Δaa) medium supplemented with different biomass- or extract-based supplements, including Spirulina biomass, Spirulina protein hydrolysate (SPH), Cupriavidus biomass, Cupriavidus protein hydrolysate (CPH), yeast extract (YE), algal extract, or complete commercial media (Sf900III or ESF921). For expression of EPHA2 protein, cells were infected with EPHA2 P2 recombinant virus. After three days of infection, equal amounts of total protein were loaded for each condition, and EPHA2 expression was detected by immunoblotting.

As shown in Fig. 2 (top left), EPHA2 expression was detectable across all supplementation conditions, indicating that each supplement provided at least partial nutritional support for recombinant protein expression. However, EPHA2 levels observed with Cupriavidus-based supplements (both biomass and CPH) were consistently weaker compared to fully supplemented media (Sf900III or ESF921). Among the various supplements, CPH supported detectable EPHA2 expression, with signal intensities comparable to SPH or algal extract, but lower than Spirulina biomass, YE and markedly lower than complete media. These data demonstrate that Cupriavidus-derived supplements are bioavailable to Sf9 cells, although initial expression levels were reduced under these conditions.

Because the in-house Sf9 cells were routinely maintained in nutrient-rich Sf900III medium prior to transfer into ESF921 Δaa medium, limited adaptation to nutrient-restricted conditions may have contributed to reduced expression levels. To address this, expression experiments were repeated using Sf9 cells obtained from Oxford Expression Technologies (OET), which are pre-adapted to ESF921 medium.

When OET Sf9 cells were used, EPHA2 expression levels increased across several supplementation conditions (Fig. 2, top right). Notably, expression supported by CPH was substantially improved compared to Cupriavidus biomass, indicating that the hydrolyzed form is more readily utilized by the cells. Under these conditions, EPHA2 expression in CPH-supplemented cultures was comparable to that observed with Spirulina biomass or SPH, although still lower than that achieved with complete commercial media.

To further examine whether additional factors could enhance protein expression, OET Sf9 cells were cultured with the same supplements in the absence (−) or presence (+) of fetal bovine serum (FBS). As shown in Fig. 2 (bottom left), inclusion of FBS led to a general increase in EPHA2 expression across most supplement conditions, including those containing CPH. Expression in complete media (Sf900III and ESF921) also remained robust under these conditions (Fig. 2, bottom right).

Together, these results indicate that Cupriavidus protein hydrolysate is taken up by Sf9 cells and can support recombinant protein expression, with performance improving significantly when cells are adapted to the basal medium and when additional supplements such as FBS are included.

Isotope enrichment of EPHA2 for NMR and Mass spectrometry labeling efficiency analysis

Isotope enrichment of EPHA2 for NMR and mass spectrometry analysis was achieved using stable isotope labeling with ¹⁵N and/or ¹³C. Following the optimization of EPHA2 expression in pre-adapted OET Sf9 cells, these cells were employed for isotope labeling experiments. Initial work utilized regular fetal bovine serum (FBS) with ESF921 Δaa medium due to its benefits in enhancing protein yield and cell growth.

For ¹⁵N labeling, OET Sf9 cells were initially grown to high densities in fully supplemented ESF921 medium, to be used for expression. After centrifugation of the desired number (~ 6 × 10⁶ per ml) of OET Sf9 cells, the cells were transferred to ESF921 Δaa medium and were allowed to adapt for (two hours) nutrient-starved conditions. In the next step, the starved cells were transferred to ESF921 Δaa amino acid-free medium which was supplemented with 5% FBS, 0.5% penicillin/streptomycin (Pen/Strep), 1% unlabeled glucose, 1% ¹⁵N-labeled CPH, 20 mg/L ¹⁵N₂-tryptophan, and 250 mg/L ¹⁵NH₄Cl. Followed by infection with 2% EPHA2 P2 recombinant baculovirus. Cells were harvested after three days of post-infection. Uniformly ¹⁵N-labeled EPHA2 was purified from the cells with different purification steps(Gande et al. 2016) and used for NMR. The folding of the protein was assessed using ¹H-1D and ¹H, ¹⁵N HSQC 2D NMR spectra. High signal dispersion and uniform peak intensities confirmed that the EPHA2 protein was well-folded and homogeneous (Fig. 3).

Fig. 3.

¹H-¹⁵N HSQC spectra of EPHA2. Two-dimensional ¹-¹⁵N HSQC NMR spectrum of ¹⁵N-labeled EPHA2 kinase domain expressed and purified from a 250-ml culture grown in ESF921 Δaa medium supplemented with ¹⁵N Cupriavidus protein hydrolysate (condition III; Table 2). Spectra were acquired at 298 K on a Bruker 800 MHz spectrometer equipped with a cryo TCI ¹H [¹³C, ¹⁵N] probe and measured in a 3-mm NMR tube. A 200 µl protein sample containing 130 µM EPHA2 kinase domain was prepared in 20 mM Tris-HCl (pH 8.0), 200 mM NaCl, 5 mM MgCl₂, and 3 mM TCEP, with 10% (v/v) D₂O included for field-frequency locking. Spectra were recorded with 40 scans per increment, 2048 complex points in the direct dimension (F2), and 224 increments in the indirect dimension (F1). Chemical shifts were referenced to 1 mM 3-(trimethylsilyl)-2,2,3,3-tetradeuteropropionic acid (TSP-d₄) used as an internal standard

For ¹³C and combined ¹³C/ ¹⁵N labeling of EPHA2, ESF921 Δaa medium was supplemented with 5% FBS, 0.5% Pen/Strep, 1% ¹³C-glucose, 1% ¹³C-labeled or ¹³C/ ¹⁵N-labeled CPH, 20 mg/L ¹⁵N2-tryptophan, and 250 mg/L NH₄Cl (unlabeled for ¹³C labeling; ¹⁵NH₄Cl for ¹³C/ ¹⁵N labeling). The labeling protocol mirrored that of ¹⁵N incorporation, with cells adapted to amino acid-depleted conditions before being transferred to the isotope-enriched medium. After three days of infection with 2% EPHA2 P2 recombinant virus, the cells were harvested, and the isotopically labeled respective proteins were purified for further analysis by LC-MS measurements.

Determination of isotope enrichment by mass spectrometry (LC-MS)

The use of liquid chromatography-tandem mass spectrometry (LC-MS) for determining ¹⁵N and ¹³C isotope enrichment in proteins derived from insect or mammalian cells is crucial for metabolic studies and quantitative proteomics. This method allows precise detection of isotopically labeled amino acids incorporated into proteins such as actin, tubulin, or GAPDH, enabling insights into protein turnover and synthesis rates. The workflow typically involves culturing cells in media enriched with ¹⁵N - or ¹³C-labeled amino acids, followed by protein extraction, enzymatic digestion (e.g., trypsin digestion), and LC-MS analysis to quantify the isotope incorporation by measuring mass shifts in peptide fragments. This approach is widely used in stable isotope labeling by amino acids in cell culture (SILAC) and metabolic flux analysis. While mammalian cell cultures, such as HEK293 or CHO cells, allow controlled isotope labeling, insect cell systems like Sf9 or Sf21 are also effective for recombinant protein expression with isotope enrichment. This methodology provides high sensitivity and specificity, enabling differentiation between naturally occurring isotopes and experimental enrichment levels.

In this study, unlabeled (¹²C/¹⁴N) EPHA2 protein samples were first analyzed to establish a baseline peptide profile. Peptide fragments generated from tryptic digestion were subjected to LC-MS (Fig. 4A), and the resulting spectra were matched against a peptide database (Perez-Riverol et al. 2025). A total of 33 peptides were identified, yielding 30 unique peptide spectra and providing 79% sequence coverage of the EPHA2 protein, as indicated by bold regions in the sequence diagram (Fig. 4A). Retention times and m/z (mass-to-charge) envelopes were defined for each identified peptide, forming the foundation for subsequent isotopic incorporation analysis.

Fig. 4.

LC-MS-based determination of ¹⁵N/¹³C incorporation. (A) Sequence coverage of the target construct in a control, unlabeled LC-MS run used for peak assignment (detected segments in bold; sequence coverage of 79%, using 30 unique peptides detected with 514 PSMs; linker and TEV cleavage site not shown) (B) Total ion chromatogram and extracted ion chromatograms of selected peptides of a representative tryptic digest of EPHA2 used for isotopic envelope calculation. (C) Isotopic envelopes of selected peptides in unlabeled, ¹⁵N-, ¹³C- and ¹⁵N+¹³C-labeled samples. (D) Detected incorporation rates for indicated peptides (SF4Δaa ΔYE, replicate 1; 19 curated peptides, peptide values in grey, average in black). (E) Summed incorporation rates for EPHA2 for all quantified peptides per sample (dark grey bars: ESF921; light grey bars: SF4Δaa ΔYE/ SF4Δaa ΔYE ΔGSM)

Using this reference dataset, the uptake of ¹³C and ¹⁵N was calculated by aligning spectra from labeled samples to the unlabeled controls based on retention time and m/z envelopes. For example, the peptide KEVPVAIK (C₄₁H₇₄N₁₀O₁₁; [M + 2 H]²⁺ = 442.515 m/z) was identified in the unlabeled sample with a retention time of 17.9 min (Fig. 4B). This information enabled detection of labeled isotopologues in subsequent runs, including ¹³C-labeled, ¹⁵N-labeled, and dual-labeled (¹³C + ¹⁵N) forms. Accurate frame alignment allowed the extraction of isotopic signals from these labeled peptides (Fig. 4C).

Several medium compositions were tested to assess the efficiency of isotope enrichment using different nitrogen and carbon sources (Table 1). Isotope incorporation into recombinant EPHA2 was quantified by LC-MS, while protein yields were determined independently after purification using UV absorbance (NanoDrop). All yields were normalized to milligrams of purified protein per 100 ml culture to allow direct comparison across conditions (Table 2). Figure 4D and E presents the isotope incorporation levels of ¹⁵N, ¹³C, and combined ¹⁵N-¹³C enrichment in the EPHA2 protein (D596-G900) in OET Sf9 cells, determined through LC-MS.

Table 1.

Various protein expression conditions and their media composition

Isotope	Requirement	Media	Extract 1% (w/v)	15NH4Cl 0.025% (w/v)	15N Tryptophan 0.002% (w/v)	15N YE 0.06% (w/v)	Glucose 1% (w/v)	FBS or Dia FBS* 5% (v/v)	Pen Strep 0.5% (v/v)
Unlabeled complete medium		Sf900 III or ESF921 or SF4		-	-	-	-	✓	✓
15N	Δ amino acids (aa)	ESF921 Δ aa	YE	✓	-	-	-	✓	✓
			Cupriavidus protein hydrolysate (CPH) or biomass	✓	✓		✓
			Spirulina protein hydrolysate (SPH) or biomass	✓	✓		✓
	Δ aa Δ yeast extract, Cortecnet (YE)	SF4 Δaa ΔYE	YE	✓	-	-	-	✓*	✓
			CPH or biomass	✓	✓	✓
			SPH or biomass	✓	✓	✓
13C	Δ aa	ESF921 Δ aa	13C CPH	✓ (unlabeled)	✓ (unlabeled)	-	-	✓	✓
	Δ aa Δ YE Δ Sugars (Glucose, Sucrose, Maltose- GSM)	SF4 Δaa ΔYE ΔGSM	13C CPH	✓ (unlabeled)	✓ (unlabeled)	✓ (unlabeled)	13C	✓*
15N, 13C	Δ aa	ESF921 Δ aa	15N 13C CPH	✓	✓ (unlabeled)	-	-	✓	✓
	Δ aa Δ YE Δ GSM	SF4 Δaa ΔYE ΔGSM	15N 13C CPH	✓	✓	✓	13C	✓*

Table 2.

Isotope incorporation levels of ¹⁵N, ¹³C, and combined ¹⁵N-¹³C enrichment in the EPHA2 protein in various growth media

Sample ID	Isotope	Various medium composition for isotope enrichment of EPHA2 protein in OET Sf9 cells	% of isoptope incorporation	Std.dev of incorporation	Mass spec IDs	Protein Yield mg/100 ml of Culture
I (c)	15N	ESF921 Δaa medium: supplemented with 15N YE, 15NH4Cl	71.2	2.85	1387	3.38*
II	13C	ESF921 Δaa medium: supplemented with 13C CPH, NH4Cl, Trp, 13C Glucose	57.0	4.57	1381	1.47**
III	15N	ESF921 Δaa medium: supplemented with 15N CPH, 15NH4Cl, Trp, Glucose	63.1	1.92	1383	1.90**
IV	13C-15N	ESF921 Δaa medium: supplemented with 13C 15N CPH, 15NH4Cl, Trp, 13C Glucose	50.0	4.04	1385	1.23**
V (c)	15N	SF4 Δaa ΔYE medium: supplemented with 15N YE, 15NH4Cl	78.5	1.97	1687_3	0.74
VI (c)	15N	SF4 Δaa ΔYE medium: supplemented with 15N SPH, 15NH4Cl, 15N Trp, 15N YE	75.2	2.39	1687_1	0.60
VII	15N	SF4 Δaa ΔYE medium: supplemented with 15N CPH, 15NH4Cl, 15N Trp, unlabeled YE	75.7	3.87	1687_2b	2.48
VIII	15N	SF4 Δaa ΔYE medium: supplemented with 15N CPH, 15NH4Cl, 15N Trp, 15N YE	79.0	3.58	1687_2a	2.84
IX	13C	SF4 Δaa ΔYE ΔGSM medium: supplemented with 13C CPH, 13C Glucose, Unlabeled: NH4Cl, Trp, YE	69.0	4.72	1687_6	0.34
X	13C-15 N	SF4 Δaa ΔYE ΔGSM medium: supplemented with 13C 15N CPH, 15NH4Cl, 15N Trp, 15N YE	70.6	8.19	1734_02	0.25
XI	13C-15N	SF4 Δaa ΔYE ΔGSM medium: supplemented with 13C 15N CPH, 15NH4Cl, 15N Trp, 15N YE after 1 h starvation	75.1	5.69	1734_03	0.04**

Actual culture volume: * 1000 ml ** 250 ml

In ESF921 Δaa medium supplemented with ¹⁵N YE and 0.027% ¹⁵NH₄Cl, ¹⁵N incorporation reached 71.2% with a normalized protein yield of 3.38 mg/100 ml (Table 2, Sample I(c)). This condition provided a balanced reference, combining moderate isotope incorporation with relatively high protein yield.

Removal of YE from the basal medium (SF4 Δaa ΔYE), while supplying ¹⁵N YE and 0.027% ¹⁵NH₄Cl, as complex nitrogen source, increased ¹⁵N incorporation to 78.5% (Table 2, Sample V(c)), but resulted in a pronounced reduction in protein yield (0.74 mg/100 ml), indicating that complete YE removal negatively impacts expression efficiency despite improved labeling.

To compare alternative complex nitrogen sources, SPH and CPH were evaluated under comparable conditions in SF4 Δaa ΔYE medium. Supplementation with ¹⁵N SPH resulted in 75.2% ¹⁵N incorporation (Table 2, Sample VI(c)), but protein yield remained low (0.74 mg/100 ml), similar to the ΔYE condition. This indicates that SPH supports efficient isotope incorporation but provides limited support for high-level protein expression.

In contrast, substitution of SPH with ¹⁵N CPH under the same basal conditions resulted in 75.7% ¹⁵N incorporation (Table 2, Sample VII) while increasing protein yield to 2.48 mg/100 ml. Thus, although SPH and CPH achieved comparable isotope incorporation, CPH supported substantially higher protein yields, indicating improved metabolic support for sustained protein synthesis.

The highest ¹⁵N incorporation (79.0%) was observed when ¹⁵N CPH was combined with ¹⁵NH₄Cl, ¹⁵N tryptophan, and ¹⁵N YE in SF4 Δaa ΔYE medium (Table 2, Sample VIII). Importantly, this condition also maintained a moderate protein yield, demonstrating that CPH can support both high isotope incorporation and acceptable expression levels when background unlabeled nitrogen is minimized but essential nutrients are retained.

When ¹⁵N CPH was used in ESF921 Δaa medium (Table 2, Sample III), ¹⁵N incorporation decreased to 63.1%, while protein yield remained moderate (~ 1.9 mg/100 ml). This reduction in enrichment reflects isotopic dilution from background nitrogen sources in the richer basal medium rather than inefficient utilization of CPH.

For carbon labeling, use of ¹³C CPH in ESF921 Δaa medium (Table 2, Sample II) resulted in lower ¹³C incorporation and moderate protein yield, consistent with dilution by unlabeled carbon sources present in the basal medium.

In contrast, SF4 Δaa ΔYE ΔGSM medium supplemented with ¹³C-labeled CPH and ¹³C glucose (1%) achieved 69.0% ¹³C incorporation (Table 2, Sample IX). However, this condition resulted in a markedly reduced protein yield (0.34 mg/100 ml), indicating that extensive removal of background carbon sources strongly limits protein expression despite reasonable isotope incorporation. These results show that while carbon derived from CPH is readily incorporated into recombinant protein, carbon restriction imposes a stronger penalty on yield than nitrogen restriction.

Simultaneous ¹³C-¹⁵N labeling further accentuated the trade-off between isotope incorporation and protein yield. In ESF921 Δaa medium supplemented with ¹³C-¹⁵N CPH (Table 2, Sample IV), combined isotope incorporation remained low, with moderate protein yield (0.25 mg/100 ml), reflecting substantial dilution from unlabeled nutrients.

In SF4 Δaa ΔYE ΔGSM medium without starvation, combined ¹³C-¹⁵N labeling reached 70.6% incorporation (Table 2, Sample X), but protein yield was already reduced compared to single-isotope conditions. Introduction of a 1-hour starvation step prior to isotope supplementation further increased incorporation to 75.1% (Table 2, Sample XI) but resulted in an extreme reduction in protein yield (0.04 mg/100 ml).

In summary, CPH supported substantial isotope incorporation across ¹⁵N, ¹³C, and combined ¹³C-¹⁵N labeling conditions. For ¹⁵N labeling, CPH achieved isotope incorporation comparable to SPH and YE, while consistently supporting higher protein yields under matched conditions. Carbon and dual-isotope labeling using CPH were feasible but were associated with pronounced reductions in protein yield, particularly under stringent carbon restriction or starvation protocols. Overall, protein yield was more strongly influenced by basal medium composition and nutrient restriction than by the choice of hydrolysate, highlighting clear incorporation–yield trade-offs across labeling strategies.

Discussion and conclusions

This study demonstrates that C. necator–derived protein hydrolysates can serve as a biologically effective alternative source of labeled nitrogen and carbon for stable isotope (¹⁵N and ¹³C) incorporation in eukaryotic protein expression systems. Using Sf9 insect cells and the receptor tyrosine kinase EPHA2 as a model protein, we show that CPH is readily taken up by cells, supports recombinant protein expression, and enables substantial isotope enrichment when appropriately integrated into defined media formulations.

Western blot analyses established that Cupriavidus-based supplements are bioavailable and capable of supporting EPHA2 expression under amino acid–depleted conditions. Although expression levels supported by CPH were lower than those obtained in fully supplemented commercial media, they were comparable to established isotope-labeling supplements such as SPH and YE. Importantly, expression outcomes were strongly influenced by cell line adaptation. Sf9 cells pre-adapted to amino acid–restricted media exhibited markedly improved expression, particularly when supplemented with CPH, indicating that early limitations reflected metabolic stress rather than intrinsic deficiencies of the supplement. The consistently higher expression observed with CPH compared to Cupriavidus biomass further underscores the importance of hydrolysates in improving amino acid accessibility under restrictive growth conditions.

Quantitative LC-MS analyses revealed that CPH supports high levels of isotope incorporation. For ¹⁵N labeling, enrichment levels of up to ~ 79% were achieved when background unlabeled nitrogen sources were minimized and CPH was combined with defined labeled nitrogen supplements. These values are comparable to those reported for yeast- and algal-based hydrolysates in earlier studies of isotope labeling in eukaryotic systems (Sitarska et al. 2015). Notably, direct comparisons between Spirulina protein hydrolysate and CPH under matched conditions showed similar ¹⁵N incorporation efficiencies, but consistently higher protein yields with CPH. This indicates that the primary advantage of CPH lies not in superior isotope incorporation per se, but in its ability to sustain more robust protein expression while maintaining comparable enrichment. Carbon labeling imposed greater metabolic constraints than nitrogen labeling. Although ¹³C derived from CPH was efficiently incorporated into recombinant protein, extensive removal of background carbon sources led to pronounced reductions in protein yield. These limitations became even more evident during combined ¹³C -¹⁵N labeling, where intracellular metabolite buffering and regulatory constraints restricted both labeling efficiency and expression output. A short starvation step prior to isotope supplementation increased dual-isotope incorporation, consistent with enhanced uptake of labeled precursors, but at the cost of severe yield loss. These findings align with broader observations that metabolic stress can transiently favor isotope incorporation while compromising translational capacity (Saxena et al. 2012).

Collectively, the data emphasize that basal medium composition, cellular adaptation, and metabolic conditioning exert a stronger influence on labeling outcomes than the choice of hydrolysate alone. Within this context, CPH performs as established alternatives and, in several conditions, supports higher recoverable protein yields. This balance between isotope enrichment and protein productivity is critical for downstream applications such as NMR spectroscopy and mass spectrometry, where both labeling efficiency and material quantity are limiting factors.

Beyond uniform labeling, Cupriavidus-based hydrolysates offer potential for more advanced isotope-labeling strategies. Fractionated or selectively enriched hydrolysates could enable residue-specific or segmental labeling approaches, analogous to those previously demonstrated for methyl-specific labeling schemes that enhance spectral resolution in large protein systems (Sitarska et al. 2015; Rößler et al. 2024). Such strategies may be particularly valuable for multidomain proteins, dynamic receptors, or multimeric assemblies.

The utility of Cupriavidus-derived supplements may also extend to solid-state NMR applications, which often require both high isotope enrichment and substantial amounts of protein, particularly for membrane-associated or aggregation-prone targets (Saxena et al. 2012). In addition, because insect cells support post-translational modifications, including glycosylation, the ability to fine-tune isotope incorporation using defined hydrolysates could facilitate structural studies of glycoproteins, which are of increasing relevance in therapeutic development (Skelton et al. 2010; Rogals et al. 2021).

In conclusion, this work establishes C. necator protein hydrolysate as a robust and biologically compatible alternative supplement for stable isotope labeling in insect cell expression systems. While careful optimization of media composition and cell adaptation remains essential, CPH reliably supports recombinant protein expression and high levels of isotope incorporation. Future studies should explore its application in mammalian expression systems, assess compatibility with selective labeling strategies, and further dissect the metabolic basis of its performance. Together, these efforts will expand the toolkit available for isotopic labeling of complex eukaryotic proteins and enable more flexible structural and biophysical investigations.

Author contributions

S.L.G., S.S., F.R. conducted the research. S.L.G., S.S., J.D.L., H.S. wrote the manuscript.H.H., S.L.G., S.S. J.D.L., H.S. designed the project.H.H., H.S. acquired funding.All authors participated in writing the manuscript.

Funding

Open Access funding enabled and organized by Projekt DEAL.

Data availability

No datasets were generated or analysed during the current study.

Declarations

Competing interests

The authors declare no competing interests. Materials were provided by Silantes as a gift-in-kind for this research. H. Heumann is employee of the company Silantes.