Two papers: a sialic-acid-mimicking antibody crosses flu neuraminidases, and gut-derived brain T cells shape feeding behaviour.
Using immunohistochemistry in mice, they looked for CD4 T cells sitting in tissue away from blood vessels. These were relatively rare, but the subfornical organ stood out by numbers, with the choroid plexus second. Whole-brain clearing, 3D imaging and a CD4 fate-map tdTomato mouse supported the same conclusion.
Answered around 43:22The novelty-suppressed feeding test measures how long a hungry mouse waits before eating when the food sits in the middle of an open box, balancing hunger against anxiety. Mice lacking alpha-beta T cells, interferon gamma, TBET or CXCR6 showed delayed feeding, while IL-17-deficient mice did not.
Answered around 1:18:36Interferon gamma and IL-17 producing cells appeared in the brain right after weaning, and germ-free mice lacked them. Monocolonizing with segmented filamentous bacteria increased them, antibiotics depleted them, and blocking the trafficking receptors VLA4, LFA1 and the S1P receptor also kept these T cells out of the brain.
Answered around 1:02:08HB420 bound the H5N1 neuraminidase but none of the other group one NAs, and it did bind a subset of N2 NAs. It inhibited neuraminidase enzyme activity and blocked growth of an H1N1 virus, PR8, and an H3N2 virus, but not H5N1.
Answered around 15:24They solved the structure of HB420 bound to an H3N2 neuraminidase. The CDR H3 loop inserts directly into the enzyme's active site, which is how the antibody blocks activity. That site's conformation is similar in H3N2 and H5N1 neuraminidases, consistent with the antibody's cross-reactivity.
Answered around 17:18Mice got HB420 by IP injection either four hours before or 24 hours after intranasal infection. Against H3N2 it gave complete protection: 100% survival, minimal weight loss, reduced lung titers. Against a virus bearing the N1 of H5N1, protection was partial, 40% survival and a two-log titer reduction.
Answered around 29:03Reverting every somatic hypermutation left an antibody that still bound N2 neuraminidase but no longer H5N1. So the germline is already set up for N2, and the mutations added by affinity maturation, mainly in the heavy chain and outside the paratope, confer the N1 cross-reactivity.
Answered around 23:49Since 2016 the circulating human H3N2 neuraminidase has carried an N-linked glycosylation site at amino acid 245. The structure shows the attached sugar blocking antibody binding, so HB420 does not bind recent H3N2 strains. Escape came from glycosylation occluding the epitope, not from amino acid changes in it.
Answered around 19:15Most people had neuraminidase antibodies against seasonal H1N1 and H3N2, and levels were higher in the younger age groups. The same held for antibodies against the H5N1 NA. They attribute this to imprinting by H1N1 infection in younger individuals, whose first exposure was around 2009.
Answered around 12:39HB420 does not neutralize in cell culture, so they suspected Fc-mediated effector functions. They made a LALA variant, which cannot bind Fc receptors or complement, and it no longer protected against viruses bearing the H5N1 neuraminidase. Effector functions are therefore probably the basis of the protection.
Answered around 30:02“So the interesting part of this paper is they solve the structure of this antibody, HB420, to an H3N2NA, right?”
“So they're painting this picture that in the gut, these T-cells can migrate to the brain.”
“So are you convinced that there are a specific subset of FTC cells in the parenchyma?”
“I'm not convinced that the cells driving the behavior later in the paper are in that organ.”
“Influenza NA Cross-Reactive Antibodies.” https://traffic.libsyn.com/secure/immune/Immune098.mp3. Transcript summary by WhipScribe, https://whipscribe.com/transcript-pages/influenza-na-cross-reactive-antibodies.html.