Combinazione di approcci fenotipici e basati sul target per l’identificazione di nuovi agenti antivirali ad ampio spettro

The ongoing COVID-19 pandemic caused by the emergence of SARS-CoV-2 in 2019, along with the growing concern about old re-emerging viruses (such as Dengue virus, Zika virus, Ebola virus, etc.) and new potentially pandemic viruses, strongly highlight the need of new antivirals to fight future outbreaks of unknown origin or for which no drugs are available yet. The limited “one-drug, one virus” strategy should be therefore substituted by a less selective "one drug, multiple viruses" approach that should exploit highly conserved viral targets or host cell factors for the design of pan-viral drugs. These broad-spectrum antiviral agents (BSAAs) can prevent the replication of a variety of viruses and may be used to treat a variety of viral illnesses. The two main drug discovery strategies can be employed for the development of new BSAAs: i) a target-based approach, which has the advantage of allowing an easy optimization of the drug candidates via structure-based refinement; and ii) a phenotypic approach, which has the advantage of quickly identifying new drugs with the desired functional effect in a complex system independently from their mechanism of action. The main goal of this Ph.D. thesis was to exploit both approaches to successfully speed up the identification of new antiviral agents able to inhibit the replication of multiple viruses. In the first part of this Ph.D. thesis, an evolution-inspired phenotypic approach based on the modification of the formamide prebiotic model was used to identify novel antiviral heterobases with potential broad-spectrum antiviral activity. This approach stems from the idea that, from the origin of life on our planet, ancestral viruses, already obligated parasites that rely exclusively on the host to survive and replicate themselves, have co-evolved with the Last Universal Common Ancestor (LUCA) by sharing and exploiting the same chemical toolset. Thus, a modification of the prebiotic chemical conditions that originated common precursors to viruses and LUCA may result in the generation of new molecules unable to sustain the replication of viruses but still tolerated by eukaryotic cells thanks to their defence mechanism. Next, a system-oriented optimization of 2,6-diaminopurine heterobases is presented, demonstrating how phenotypic studies allow the identification of new promising BSAAs independently from the exact knowledge of their mechanism of action. Finally, the conversion of the most promising 2,6-diaminopurine antivirals into their corresponding nucleoside derivatives was a first attempt to create dual-acting drugs that could act against viruses both directly through the nucleoside itself and indirectly through the antiviral nucleobase that might be formed after metabolic cleavage of the glycosidic bond. The second part of this thesis was focused on the target-based approach to new BSAAs. SARS-CoV-2 helicase nsp13, which has a highly conserved structure with common features among highly pathogenic human coronaviruses (HCoVs), was chosen as a promising target to identify pan-coronavirus inhibitors. The inhibition of the host helicase DDX3 by the same target compounds was also taken into account, since DDX3 is involved in the replication of multiple viruses and less prone to drug resistance selection. Finally, new inhibitors of the host kinase PI4KIIIß, which is exploited by several viruses for their replication, were also developed and evaluated against multiple viruses belonging to different families.