Mostly Microbes and Infection - by David Ojcius

GLP‑1 analogs rely on structural features that enhance stability, receptor affinity, and half‑life. N‑terminal protection, C‑terminal and backbone stabilization, and lipidation drive potency and duration, while structural variation shapes signaling bias. Multi‑receptor agonists and computational design now guide next‑generation incretin therapies.

A cross‑sectional study of SARS‑CoV‑2 patients reveals reduced gut microbial diversity, lower vitamin D, increased Enterococcus faecalis, and loss of key commensals. Functional assays show Bacteroides uniformis can modulate inflammatory responses without altering viral entry. Findings highlight gut microbes’ potential to influence COVID‑19 outcomes and future preparedness.

Rice uses the circadian regulator Nhd1 to time root exudation and coordinate rhizosphere microbial nitrogen cycling. Day‑peaking quercetin 3‑gentiobioside and baseline baimaside activate microbial N‑cycling genes, boosting ammonium availability when plant demand is highest. Engineering microbes to match this rhythm increased grain yield and NUE in field trials.

Corynebacterium glucuronolyticum‑like bacteria are enriched in calcium oxalate stone patients and promote crystal nucleation and aggregation by degrading the anti‑crystallization metabolite hippuric acid and forming biofilms. Their multidrug resistance limits antibiotic efficacy, while hippuric acid supplementation or cranberry diets reduce crystallization in models.

KRAS‑mutant colorectal tumors reshape their microbiome, enriching Bacteroides acidifaciens, which predicts poor prognosis. Using KRAS‑mutant cells and mice, the bacterium drives malignancy via SusF‑mediated stabilization of ARHGEF2, boosting KRAS–RAF–MEK–ERK signaling. Loss of SusF or ARHGEF2 blocks these effects, revealing therapeutic targets for KRAS‑driven colorectal cancer.

Deep learning reveals and redesigns the multifunctional targeting landscape of the T7 phage receptor‑binding protein. Comparing four model architectures, the study engineers phages optimized for 26 tasks, showing that complex specificities can be tuned with few mutations. Multiobjective models uncover principles of phage biology and offer a general strategy for optimizing multifunctional proteins.