Skip to main content

Development of a cell-based system for high-throughput screening of antifibrotics

Periodic Reporting for period 1 - CollBioImag (Development of a cell-based system for high-throughput screening of antifibrotics)

Reporting period: 2016-03-01 to 2018-02-28

Fibrosis is a common outcome of many chronic diseases of the kidney (diabetic nephropathy), lungs (idiopathic pulmonary fibrosis), liver (cirrhosis) and heart (heart failure). In pathologies where fibrosis is a feature, there is an increased deposition of extracellular matrix proteins, including collagen. Due to the increasing morbidity and mortality of fibrotic disorders there is an urgent need to develop, test and monitor antifibrotic treatments. Due to its high abundance in collagen, proline has been widely explored as a marker of collagen synthesis by measuring incorporation of either radioactive or non-radioactive isotopic-labelled proline (18F-Pro, 3H/13C/15N-Pro) directly in tissue. The goal of this project was to develop a bioorthogonal imaging approach for visualization of the dynamic processes of fibrogenesis/fibrolysis in cells using collagen as a target. The methodology first involves the metabolic incorporation of a proline derivative equipped with an alkene tag into collagen using collagen-producing cells (Figure 1). Subsequent reaction with fluorogenic tetrazines would allow rapid, efficient and highly specific labelling of collagen (Figure 1). The ultimate goal is to develop a high-throughput screening method for identification of new antifibrotic drugs as collagen is highly expressed in fibrotic tissues.
This project started by testing a bioorthogonal labelling reaction designed to enable the selective imaging of collagen in mammalian cells. The chosen labelling reaction (inverse electron-demand Diels-Alder reaction) was first tested on proteins bearing an alkene handle that was shown to be reactive against tetrazine-based fluorescent probes (results published at Angew Chem Int Ed Engl 2016, 55, 14683–14687). These preliminary results clearly indicate that unstrained, small S-Allyl handles installed on complex structures such as proteins can be further used as a chemical partner for ligation with tetrazines. We further demonstrated the utility of this minimal handle for the efficient labeling of apoptotic cells by labeling apoptotic markers using a pretargeting approach (Angew Chem Int Ed Engl 2016, 55, 14683–14687). During the course of the project we envisage that electron rich alkenes like vinyl ethers might react faster with tetrazines. Using model compounds (amino acids, a monosaccharide and a fluorophore bearing a vinyl ethers) it was observed that upon reaction with tetrazines the reaction results in the traceless release of alcohols (Angew Chem Int Ed Engl 2017, 56, 243-247). This new reaction was used for drug activation of a vinyl ether derivative of duocarmycin (prodrug), which was successfully decaged in live cells to reinstate cytotoxicity (Angew Chem Int Ed Engl 2017, 56, 243-247). The small size of the alkene handle and selective reactivity towards tetrazines demonstrated with previous described work suggest that this approach can be readily extendable to other proteins and biomolecules such as collagen, which could facilitate their labeling within live cells. With these results in hands we then decided to install an allyl handle into collagen by metabolically incorporating a modified proline residue, O-allyl-proline (Figure 1), into foetal ovine osteoblast cells, a collagen producing cell line. As proline is highly abundant in collagen, this labelling strategy should allow selective labelling of collagen over other proteins. The O-allyl-proline was synthesized following published protocols. The reaction between O-allyl-proline and tetrazine was thoroughly validated using kinetics studies. O-allyl-proline was found not to be cytotoxic to foetal ovine osteoblast cells. An over two-fold fluorescence increase was observed by microscopy upon reaction of treated cells with 200 μM tetrazine-rhodamine B and, most notably, minimal unspecific reaction of tetrazine with untreated cells was visible. Control studies with a cell line that do not produce collagen also showed no labeling after incubation with the O-allyl-proline derivative followed by labeling with a fluorescent tetrazine. These very promising results were, however, not corroborated by other analysis performed (e.g. western blot, amino acid analysis after collagen digestion) to detect labelled collagen from cell lysate, extracellular matrix and media enriched with externalised collagen. The results obtained may be interpreted as a result of low incorporation of O-allyl-proline by prolyl tRNA-synthetase. Incorporation could not be identified also by MS/MS analysis (searching for peptide fragments with O-allyl-proline). However, only a small amount of collagen was identified in tested samples of osteoblast extracellular matrix, which were shown to be very insoluble using different extracting buffers. Further work is needed to test samples containing a larger amount of collagen in order to draw final conclusions regarding the incorporation of O-allyl-proline and the labelling of collagen protein. For these studies we are considering using media enriched with externalised collagen obtained from growing FSO cells, which after protein purification by size exclusion, should result in pure collagen samples with high protein content. The low incorporation might be also explained by the large size of the O-allyl handle or because of its low reactivity against tetrazines. To test this hypothesis other proline derivatives with alkene, cyclopropene and norbornene handles were synthesized. Incorporation of these derivatives into collagen is currently being tested.
With this project we tested a bioorthogonal reaction based on the IEDDA ligation for the specific labeling of collagen in collagen-producing cells. Although the fluorescent studies performed with collagen producing cells are suggestive that collagen is the labeled protein (please note that control with cancer cells could not be labeled using same conditions) further experimental data (western blot, MS/MS analysis of labeled collagen samples, amino acid analysis of digested collagen with O-allyl-proline, etc) are still under evaluation. It is important to mention that the fluorescent labeling of collagen production by cells was recently reported (Tissue Eng Part C Methods. 2017, 23, 228-236) using conditions previously established by our group (Chem Commun 2015, 51, 5250-5252). Therefore, further development of this bioorthogonal reaction to be used as a high-throughput screening method for testing antifibrotic drugs will also be a priority in the future. Although funding from the fellowship is no longer available we are committed to continuing investigating this approach that if successfully applied has potential for the discovery of antifibrotic drugs.