[ad_1]
Their algorithm devoured the entire human proteome and spat out a preliminary list of about 43,000 peptides. Torres narrows it to 2,603, which come from proteins known to be secreted by cells. Some were full of small proteins and hormones. Others were just fragments, encrypted chains in a much larger complex. None of them have ever been described as antibiotics before.
To test whether their AI is on the right track, Torres synthesized 55 of the most promising candidates. He tested each of them in liquid samples against “who’s who” of drug-resistant microbes: Pseudomonas aeruginosa, a known gross contagion of the lungs; Acinetobacter baumannii, which is known to spread endlessly in hospitals; Staphylococcus aureus, the embryo behind the dangerous staphylococcal infections – plus others, a total of eight. Of the 55, the majority managed to prevent the bacteria from multiplying.
Several peptides stood out, including SCUB1-SKE25 and SCUB3-MLP22. These peptides live in regions called “CUB domains”, which exist in proteins involved in a long list of functions such as fertilization, the creation of new blood vessels and the suppression of tumors. SCUBs are just parts of the whole. But by themselves, they seemed shockingly adept at killing germs. So Torres promoted these two SCUBs in mouse experiments.
Torres tested whether either SCUB, or a combination of both, could eliminate infections in mice with infections under their skin or in their thigh muscle (a model for a more systemic disease). In all cases, the populations of bacteria taken from these tissues stopped growing. And in some cases, as Torres observed on his warm agar, the number of bacteria drops sharply.
Torres also tested how easily bacteria could develop resistance to peptides, compared to an existing antibiotic called polymyxin B. After 30 days of exposure, the bacteria could tolerate doses of polymyxin B that were 256 times higher than the initial amount, but SCUBs remained effective at the same dose. (Many genetic changes are needed for bacteria to adapt to membrane damage.) Of course, this does not mean that they will never adapt, especially at longer intervals. “Nothing will ever be resistant to resistance,” de la Fuente said. “Because bacteria are the biggest evolving people we know.”
As systematic as the team’s plan was, Torres was still a little puzzled. “We thought we were going to have a lot of hits,” he said of the peptides discovered by AI. But to his surprise, the peptides come from all over the body. They were from proteins in the eyes, nervous system and cardiovascular system, not just the immune system. “They’re literally everywhere,” Torres said.
The team believes that life evolved in this way to bring the biggest blow to the genome. “One gene encodes a protein, but that protein has many functions,” says de la Fuente. “I think it’s a really smart way for evolution to just keep genomic information to a minimum.
This is the first time scientists have found antibiotic peptides in proteins that are not linked to the immune response. The idea was “really creative,” said John Stokes, a biochemist at McMaster University in Canada who did not participate in the study but is preparing his lab to include AI in the search for small-molecule antibiotics. “The trick for me is: Start looking for non-obvious places for antibiotics.
[ad_2]
Source link