16th European Conference on Fungal Genetics (ECFG16)), Innsbruck, Avusturya, 5 - 08 Mart 2023, ss.539-540, (Özet Bildiri)
KOMATAGAELLA PHAFFII: A MODERN CELL FACTORY FOR
PRODUCING CYSTEINE-RICH ANTIMICROBIAL FUNGAL
PROTEINS AND AI-PREDICTED CYSTEINE RICH PEPTIDES
Busel Ozcan1, Huseyin Okan Soykam2, Betul Gundogdu1,
Fatma Hande Osmanağaoğlu1, Seyma Colakoglu Ozkaya3, Serbulent Unsal4,
Tunca Dogan5, Meltem Muftuoglu1, Günseli Bayram Akçapınar1
1Acibadem University, Institute of Health Sciences, Department of Medical
Biotechnology, Istanbul, Turkey, 2Acibadem University, Institute of Health
Sciences, Department of Translational Medicine, Istanbul, Turkey, 3Marmara
University, Institute of Health Sciences, Department of Medical Biology , Istanbul,
Turkey, 4Karadeniz Technical University, Institute of Health Sciences, Department
of Biostatistics and Computer Sciences , Trabzon, Turkey, 5Hacettepe University,
Faculty of Engineering, Department of Computer Engineering, Ankara, Turkey
Antimicrobial resistance (AMR) is a serious threat to human health worldwide. The COVID-19 era has once again highlighted the need for novel, improved, and cost-effective antimicrobials due to the continual emergence of pathogenic and/or resistant organisms, the unbalanced use of antimicrobials, and the nature of resistance evolution. Due to their abundance in nature, antimicrobial peptides (AMPs) and antimicrobial proteins may provide novel alternatives to existing antimicrobials. Komagataella phaffii (aka Pichia pastoris) is a methylotrophic yeast capable of secreted protein expression in high titers along with PTMs required for eukaryotic proteins. In this study, we used K. phaffii microbial cells due to their superior properties as modern cell factories to produce two families of antimicrobials: Cysteine-Rich Peptides(CRPs) and fungal defensins. CRPs are small AMPs of 10-40 aminoacids, contain several cysteine residues with ability to form “Cysteine stabilized alpha-helical” structure.
Defensins are small, secreted and cysteine-rich proteins with antimicrobial activity. Both are products of the innate immunity in a pleuthera of organisms from mammalians to plants and fungi, and are potent therapeutic agents against pathogenic bacteria and fungi.
In an attempt to find novel antimicrobials, we trained several deep learning architectures (CNN, LSTM, CNN-biLSTM etc.) on a diverse set of AMPs/non-AMPs curated from databases and datasets. Novel AMPs were predicted from a library of 20 aminoacid long protein fragments compiled from Uniprot. Alternatively, a homology-based methodology was used to screen for eurocin-like fungal defensins. We formed a selection strategy based on the structural, physico-chemical properties. Rationally selected five novel defensins and three CRPs were expressed in K. phaffi using the AOX promoter and α-factor secretion signal. A high-throughput screening strategy was employed for the selection of high-producers. Purification and antimicrobial screening of the CRPs and defensins indeed verified K. phaffii as a suitable host for high yield, cost-effective production of active antimicrobials.