Artificial peptides could be a promising weapon against antibiotic-resistant bacteria
Artificial peptides could be a promising weapon against bacteria resistant to antibiotics, suggest these researchers from the Massachusetts Institute of Technology (MIT) who, using a computer algorithm, develop peptides more powerful than those found in nature. This innovative strategy, presented in Nature Communications, which imitates the natural process of evolution with the algorithm, has already led to the discovery of a promising drug candidate capable of eliminating bacteria in mice.
This approach thus constitutes a possible response to the need for new antimicrobials, given the emergence of many bacterial strains resistant to existing antibiotics, and the extremely small number of new antibiotics in development. The research on antimicrobial peptides, natural peptides found in most organisms is part of the main lines of research, but most of these natural peptides do not prove to be powerful enough to fight infections in humans. The approach of MIT scientists is therefore to create more powerful synthetic versions. The strategy is based on a computer algorithm that identifies new antimicrobial peptide sequences, says researcher César de la Fuente-Nunez from MIT, lead author of the study.
An algorithm capable of testing thousands of peptides:
Researchers point out that antimicrobial peptides kill microbes in different ways:
they enter microbial cells by damaging their membranes,
once inside, they can disrupt several types of cellular targets such as DNA, RNA and proteins.
In this race for more potent artificial antimicrobial peptides, scientists will have to synthesize hundreds of new variants and then test them against different types of bacteria. But the computer algorithm is there to simplify the work, it starts from a peptide sequence then generates thousands of variants and tests them on precise objectives, specified by the researchers. The algorithm allows you to explore many more peptides.
A textbook case: the researchers started with this study, an antimicrobial peptide found in guava seeds. This peptide, known as Pg-AMP1, has only weak antimicrobial activity. But by programming the algorithm so that it offers, from Pg-AMP1, peptide sequences having 2 characteristics allowing the peptide to penetrate the bacterial membranes, the algorithm generates and evaluates tens of thousands of peptide sequences , then selects a hundred that the researchers synthesize. The peptide obtained ultimately the most efficient, known as guavanine 2, is very different from the original peptide Pg-AMP1, it is much more powerful.
A promising synthetic peptide: this first synthetic peptide is indeed effective against Gram-negative bacteria, responsible for the most common nosocomial infections, including pneumonia and urinary tract infections. When researchers test guavanine 2 in mice with Gram-negative bacterial skin infection (Pseudomonas aeruginosa), they find that the peptide guavanine 2 eliminates the infection much more effectively than the original peptide Pg-AMP1.
This innovative approach – because there is currently no artificial antimicrobial peptide approved for human use – can therefore lead to peptides with the properties necessary to be good antibiotics. Guavanine 2 is still in the pipeline for possible use in humans, and the algorithm “spins” to identify other potent antimicrobial peptides.
Source: Nature Communications 16 April 2018 doi: 10.1038 / s41467-018-03746-3 In silico optimization of a guava antimicrobial peptide enables combinatorial exploration for peptide design