This issue compiles 43 papers, grouped by receptor/ion channel, neuroscience, drug discovery, protein prediction, viral genome, transcription, cryo-ET, and other topics. Most are structural biology, mechanistic, or animal model studies, covering ion channels, GPCRs, psilocybin, ALK degraders, and more. Most report preclinical or short-term findings and cannot be directly extrapolated to clinical efficacy. Summaries were compiled by AI from the original texts.
This issue covers the 43 papers added to a personal Zotero library between September 7 and September 13, 2026. Dates and times are Beijing time, based on when each item was added. Papers are grouped by topic and listed in reverse chronological order within each group, with the miscellaneous items at the end.
The summaries below were compiled by AI from abstracts, with missing abstracts filled in from the original sources; short communications and news briefs are marked separately and were not read in full. One paper had no verifiable abstract and is listed by metadata only. Publication and volume/issue dates are taken from the entry records; they may be later than the date the paper was added and do not indicate publication this week. For methods and conclusions, please consult the original articles.
Receptors, Ion Channels, and Membrane Protein Recognition (11 papers)
01 · Gating crosstalk in potassium channels
Proceedings of the National Academy of Sciences · Publication/issue date: 2026-09-15
Authors: Lyubin Hu, Bert L. de Groot, Ruo-Xu Gu
How the central cavity and selectivity filter of potassium channels open and close in concert is a key question for understanding ion conduction. The authors ran molecular dynamics simulations of six potassium channels spanning multiple subfamilies and found that opening and closing of the central cavity simultaneously changes the degree of cavity hydration and the width of the filter entrance. The former affects the free energy of potassium ions entering the cavity, while the latter regulates ion occupancy in the filter, with the fastest conduction occurring at an intermediate entrance size. Hydrophobic residues on the inner transmembrane helices link these two conformational changes. State differences seen in existing experimental structures also match the simulations, supporting a gating crosstalk model in which hydration and entrance geometry jointly regulate conduction. This consistency across channels hints at a potentially general principle, but it does not mean that all potassium channels follow exactly the same mechanism.
Highlight · The central cavity influences potassium conduction through two routes: the degree of hydration and the width of the filter entrance.
The human creatine transporter hCRT handles cellular creatine uptake; when it goes wrong, it is linked to creatine deficiency syndromes, and it has also drawn attention in tumor metabolism research. The study solved cryo-EM structures of the transporter with no ligand, bound to creatine, and bound to the inhibitor RGX-202, showing how this twelve-transmembrane-helix transporter recognizes its substrate. The creatine-binding pocket is formed mainly by several transmembrane segments, and the inhibitor occupies the same region, offering a structural explanation for competitive inhibition. Comparing different states links the binding site to transport conformations and provides a reference for understanding how genetic variants impair function. These results mainly address the molecular basis of substrate uptake and blockade; they do not directly demonstrate the efficacy or safety of inhibiting this transporter in patients.
Highlight · Creatine and RGX-202 share a binding pocket, providing a structural basis for substrate recognition and competitive inhibition of hCRT.
Key points from the original short communication: this entry has no standalone abstract, so the following is based on the readable portions of the publisher's text. The authors compared structures of the human creatine transporter with no ligand, bound to creatine, and bound to RGX-202, linking inward-open and inward-closed conformational states. Structural and functional analyses together clarify how substrate, ions, and inhibitor are recognized, and support the idea that RGX-202 blocks transport by competing for the substrate-binding region. The paper also uses the structural information to discuss pathogenic variants. This work addresses questions similar to those in another hCRT paper from the same group but has a different DOI, so both are retained separately; it provides evidence at the level of transport mechanism and does not mean that therapeutic benefit in disease has been clinically validated.
Highlight · Different ligand states help connect creatine recognition, conformational change, and transport inhibition.
Authors: Thomas P. Newton, Muhammed Aktolun, Maria V. Yelshanskaya, Alexey A. Alekseev, Laura Y. Yen, Shanti Pal Gangwar, Ivan A. Sobolevsky, Maria G. Kurnikova, Alexander I. Sobolevsky
AMPA receptors pass through several short-lived states between ligand binding and ion channel opening. Using time-resolved cryo-EM, the study compared conformations of the GluA2–γ2 complex under full and partial agonists, covering the ligand-free, pre-activated, open, and desensitized stages. The degree of clamshell closure in the ligand-binding domain correlates with the downstream gating pathway: smaller closure can correspond to a pre-activated intermediate, whereas greater closure is associated with open or desensitized states. Single-channel recordings and molecular dynamics simulations complemented the structural interpretation. The results help explain how ligands of differing efficacy select different state distributions, and they caution against reducing open probability to whether ligand binding has occurred.
Highlight · Full and partial agonists shape the gating pathway of AMPA receptors by altering domain closure and the distribution of intermediate states.
Authors: Pinqi Wang, Xuan Zhang, Abdul-Akim Guseinov, Laura Jenkins, Carl von Hallerstein, Jonathan D. Colburn, Rowan Ives, Vincent B. Luscombe, Sara Marsango, Listiana Oktavia, Arun Raja, David R. Greaves, Philip C. Biggin, Graeme Milligan, Cheng Zhang, Irina G. Tikhonova, Angela J. Russell
GPR84 takes part in immunometabolic signaling, and different ligands can produce varying degrees of bias between G protein and β-arrestin pathways. The authors compared three matched ligands: they had similar Gi activity but differed in β-arrestin response. Cryo-EM, molecular dynamics, and mutagenesis showed that steric changes in the binding pocket can be relayed through local residues, indirectly perturbing the receptor's internal network of polar interactions. In this way, a ligand can shift signaling preference without directly touching every downstream structural element. The study offers an atomic-level mechanistic clue to biased agonism, but whether structural bias translates into better disease intervention still needs independent validation in the relevant biological setting.
Highlight · Local steric effects from a ligand can be amplified through the receptor's internal network to shift GPR84's signaling preference.
Authors: William G. Ludlam, Chu-Ting Chang, Kristina Cechova, Brandon W. Liauw, Hwa-Jin Cho, Safoura Salar, Anjelique Sawh-Gopal, Afroza Parvin, Simrat K. Dhaliwal, Simran K. Dhaliwal, Tina Izard, Huan Bao, Anne M. Brown, Henry A. Dunn, Reza Vafabakhsh, Kirill A. Martemyanov
Group III metabotropic glutamate receptors not only recognize neurotransmitters but are also modulated by transsynaptic interactions. The study solved structures of complexes involving ELFN adhesion proteins, showing how cell-to-cell contact connects the receptor's ligand-binding region to its internal allosteric network. The ELFN binding site and its effect on receptor conformation explain previously observed functional modulation. The abstract also notes that receptor activation can strengthen association with ELFN, suggesting feedback regulation between the two, and uses the structural information to interpret variants linked to neurological disease. The work highlights how important the synaptic environment is for receptor behavior, but the interactions seen in structures cannot be equated directly with net effects in an intact neural circuit.
Highlight · Transsynaptic adhesion interactions can modulate group III mGlu receptors through an allosteric network.
07 · The molecular basis of force selectivity in PIEZO2
Nature · Publication/issue date: 2026-05
Authors: Eric M. Mulhall, Oleg Yarishkin, Rose Z. Hill, Anna K. Koster, Ardem Patapoutian
Both PIEZO1 and PIEZO2 respond to mechanical stimuli, but they prefer different types of force. Combining MINFLUX imaging in intact cells with electrophysiology, the authors compared the conformations and gating behavior of the two channels. PIEZO2 proved more rigid, and different mechanical stimuli could drive conformational changes in different directions, with those changes tied to opening behavior. The authors further found that filamin B–mediated actin linkage helps give PIEZO2 its selectivity for indentation stimuli, and they observed the relevant spatial proximity at sensory endings. The study connects the channel's own structural features with cytoskeletal coupling, suggesting that mechanical selectivity must be understood in the cellular context rather than explained by the shape of an isolated protein alone.
Highlight · PIEZO2's force selectivity arises from a combination of the channel's conformational features and its actin linkage.
The oxytocin receptor is an important target for regulating uterine contractions, and achieving selectivity among closely related receptors is a persistent challenge in drug research. The authors used AI-assisted screening to look for antagonist nanobodies and obtained candidates that act on the oxytocin receptor. According to the abstract, one candidate's antagonism was comparable to that of atosiban while showing better receptor subtype selectivity and working through a different mechanism; the study also observed effects consistent with suppressed uterine contractions in mice. The work offers a route to modulating this receptor with antibodies and illustrates how computational screening can be combined with functional validation. The current results say nothing directly about effectiveness in human pregnancy, fetal safety, or the conditions for long-term use.
Highlight · Antagonist nanobodies offer a candidate route to more subtype-selective modulation of the oxytocin receptor.
Authors: Dagan C. Marx, Alberto J. Gonzalez-Hernandez, Joon Lee, Dirk Siepe, Willem F. Weber, Kevin Huynh, Pamela N. Gallo, Sheida Sharghi Moshtaghin, Anisul Arefin, Johannes Broichhagen, Marian Kalocsay, David Eliezer, Joshua Levitz
Metabotropic glutamate receptors do not adopt a single configuration when binding β-arrestin. Using single-molecule pull-down assays, the study examined binding strength and stoichiometry across different receptor complexes, observing arrangements such as 2:1 or 2:2. Both the receptor tail and the core region can participate in coupling, and mGlu8 homo- and heterocomplexes display different cis and trans modes of action. The abstract further describes complexes that bind two types of β-arrestin at once, or that accommodate β-arrestin together with a G protein. The results broaden our understanding of receptor signaling platforms and suggest that summarizing function simply as whether β-arrestin is recruited may miss important differences, though the physiological contribution of each configuration still needs to be determined separately.
Highlight · mGlu receptors can organize β-arrestin signaling complexes through a variety of stoichiometries and contact modes.
Authors: Min Zhang, Xin Li, Qing-ning Yuan, Wen Hu, H. Eric Xu, Li-hua Zhao
PAC1R can couple to different signaling proteins, and this study solves the cryo-EM structure of the receptor bound to β-arrestin 1 and compares it with the Gs-coupled state. When β-arrestin binds, the receptor shows a rearrangement of transmembrane helix 5 and an inward shift of helix 6, creating a contact environment distinct from that of the other transducer. The comparison also shows that PTH1R, which belongs to the same receptor class, does not use exactly the same β-arrestin coupling mode; both the receptor's own structure and the finger loop of β-arrestin contribute to selectivity. The takeaway is that no single receptor's pattern can stand in for signaling bias across the whole family — the specific receptor–transducer interface is a key level at which to understand functional differences.
Highlight · PAC1R selects its transducer through specific transmembrane conformations and interfaces, and β-arrestin coupling differs from receptor to receptor.
11 · How cholesterol interacts with membrane proteins: an exploration of cholesterol-binding sites including CRAC, CARC, and tilted domains
Frontiers in Physiology · Published/issue date: 2013-02-28
Authors: Jacques Fantini, Francisco J. Barrantes
This review examines how cholesterol interacts with membrane proteins, focusing on structural features that may mediate recognition, including CRAC, CARC, and tilted transmembrane domains. Starting from cholesterol's rigid sterol ring, flexible side chain, and asymmetric surface, it maps out its compatibility with different amino acid environments and explains why both local sequence and transmembrane geometry can influence binding. It supplies the basic concepts for reading studies on cholesterol's regulation of membrane proteins and helps connect sequence patterns with spatial contacts. One caveat to keep in mind: spotting a candidate sequence motif is not by itself proof of actual binding or a functional effect — structure, membrane context, and experimental evidence all still have to line up.
Highlight · Cholesterol recognition depends on the local chemical environment and transmembrane geometry; a candidate motif alone is not sufficient evidence of binding.
Neuroscience and Neuropsychiatric Interventions (9 papers)
12 · To sleep and dream: unraveling narcolepsy
Proceedings of the National Academy of Sciences · Published/issue date: 2026-09-15
Authors: Jeffrey M. Friedman
This piece traces how narcolepsy research moved from disease phenotype toward molecular mechanism. Mignot's team started from inherited narcolepsy in dogs and used forward genetics to uncover defects related to the orexin receptor; Yanagisawa's team approached it from the orexin ligand and knockout mice, producing evidence that dovetails with the first. Together the two lines of work established the orexin system's central role in maintaining wakefulness and stabilizing the switch between sleep and waking, and laid the groundwork for later treatment strategies. The article is really about the process of discovery and its conceptual impact, making it a good way to grasp how genetics, animal models, and neuropeptide signaling fit together in sleep research. It is a historical and mechanistic review, and should not be read as a new clinical drug trial.
Highlight · Canine genetics and reverse genetics in mice together revealed the orexin system's key role in maintaining wakefulness.
13 · Psilocybin-assisted therapy for treatment-resistant major depressive disorder in a public healthcare setting: a randomized controlled trial
Nature Medicine · Published/issue date: 2026-08-06
Authors: James J. Rucker, Tim Mantingh, Jess Kerr-Gaffney, Catherine Bird, Petrina Chu, Nadav Liam Modlin, Kete Campbell-Coker, Sadie Hambleton, Paige Seath, Rebecca Hignett, Diede Fennema, Elliot Hampsey, Dimosthenis Tsapekos, Rebecca J. Thomas, Joseph Cattell, Hassan Jafari, Camilla Day, Anya Borissova, Miranda Lloyd, Theo Boardman-Pretty, Mathieu Seynaeve, Famia Askari, Michael Creed, Luke Baxter, Raphael Rifkin-Zybutz, Matt Butler, Luke A. Jelen, Ben Carter, Allan H. Young
This single-center, randomized, controlled feasibility trial in a public healthcare setting enrolled sixty people with treatment-resistant major depressive disorder, comparing a single dose of psilocybin with placebo; both groups also received psychological support. Fifty-nine participants completed the study, with primary observation over six weeks. The abstract reports a between-group difference in MADRS score of −10.41 at week three, with improvement sustained through week six; at the same time, the drug group reported more non-serious adverse events than the placebo group. The results support moving ahead with larger confirmatory trials within real healthcare systems, and they also point to psychological support and adverse-event assessment as essential parts of the protocol. Given the sample size, follow-up duration, and the study's feasibility design, these findings are not enough to establish long-term efficacy or broad applicability.
Highlight · A small randomized trial in a public healthcare setting showed short-term signs of improvement, but larger and longer-term validation is still needed.
Authors: Eric K. Brengel, Bryce Axe, Ashwath N. Maheswari, Muhammad I. Abeer, Richard J. Ortiz, Taylor J. Woodward, Reagan Walhof, Wenwen Du, Rachel Utama, Brandon Rodrigue, Noah J. Cavallaro, Anna Brettler, Courtney Sawada, Anna Moore, Shreyas Balaji, Praveen P. Kulkarni, Argel Aguilar-Valles, Heather B. Bradshaw, Michael A. Gitcho, Craig F. Ferris
This study looked at how psilocybin affects imaging and molecular readouts in a model of repetitive mild head injury in adult female rats. The abstract reports reduced vasogenic edema, restored vascular reactivity, and altered resting-state functional connectivity, along with lower phosphorylated tau and changes in BDNF/TrkB signaling and lipid-related pathways. Together these readouts suggest the intervention may act on cerebrovascular state, network activity, and post-injury molecular responses at once. Changes at different levels offer leads for further work, but the causal chain between them cannot be pinned down from correlational measures alone. Both the subjects and the evidence come from animal models, so this is no basis for recommending psilocybin for concussion or repeated head injury in people.
Highlight · Imaging and molecular changes in an animal model point to a possible injury-intervention mechanism, but there is still a wide gap before judging efficacy in humans.
Authors: Xiaodong Zhang, Chunmei Lan, Peng Qing, Xiaolei Xu, Yuanshu Chen, Juan Kou, Lei Xu, Xinqi Zhou, Dezhong Yao, Benjamin Becker, Keith M. Kendrick, Weihua Zhao
The authors studied depression-related changes in white matter functional connectivity tensors and asked whether oxytocin can shift these dynamic measures. The clinical sample included major depression, generalized anxiety, and healthy controls; the relevant metric in the genu of the corpus callosum was reduced in the depression group, and the authors then examined its associations in a subclinical sample. In people high in depressive traits, oral oxytocin coincided with higher local metrics, reconfiguration of genu-to-prefrontal connectivity, and changes in how negative social information was processed. The work links white matter functional representations to circuit-level intervention responses and points to candidate markers. That said, changes in imaging metrics are not the same as symptom benefit, and they are not yet usable as a diagnostic or treatment standard in practice.
Highlight · White matter connectivity dynamics may reflect depression-related circuit states and show measurable changes in response to oxytocin.
Authors: Mario Heles, Lilach Pasvolsky, Hinduja Sathishkumar, Shorook Naara, Yen Vu, Caitlyn L. Stewart, Tongxin Xie, Frederico O. Gleber-Netto, William McCarthy, Dan Yaniv, Shashank S. Kamal, Hajira Elahi, Pengyu Zhu, Megan L. Uhelski, Jordan Chatwin, Andrew Lara, Danielle L. Stolley, Michael R. Migden, Z-Hye Lee, Shiyanth Thevasagayampillai, Preethi H. Gunaratne, Kaoutar Ait-Ahmad, M. J. Kuykendall, Sebnem Ece Eksi, Lorenzo Cohen, Jeremy C. Borniger, Eyal Gottlieb, Gregory H. Jones, Patrick M. Dougherty, Moran Amit
This study examined whether psilocybin prevents peripheral neuropathy in animal models of platinum- and taxane-based chemotherapy. The abstract reports that it reduced neuropathy across repeated chemotherapy cycles without weakening the antitumor effects in the models tested. Mechanistic analysis tied the protective effect to preserved mitochondrial trafficking within axons, involving TrkB downstream signaling, motor proteins, and regulation of mitochondrial anchoring, with related changes in prefrontal activity also observed. The findings suggest that redistributing mitochondria along axons may be one route to neuroprotection. Even so, this evidence is still preclinical: it cannot tell us whether combining psilocybin with the various chemotherapy regimens used in people is safe, nor should it become a reason to try preventing neurotoxicity on one's own.
Highlight · Preclinical results link chemotherapy neuroprotection to preserved mitochondrial trafficking in axons.
This supplementary abstract comes from the author abstract indexed in PubMed. Working in mice, the researchers identified a descending glycinergic circuit running from the rostral ventromedial medulla to the spinal dorsal horn that can drive mechanical allodynia insensitive to morphine. The pathway connects GlyT2-positive neurons to somatostatin-positive spinal neurons, and craniofacial pain can recruit it through the relevant inputs. Mechanistically, the released glycine acts on GluN3A-containing NMDA receptors, producing an unconventional excitatory effect at this site and bypassing opioid inhibition. The results offer a circuit- and receptor-level explanation for hard-to-treat mechanical pain and suggest potential targets; the evidence comes from mice, so no conclusions can yet be drawn about analgesic efficacy in humans.
Highlight · Glycine can drive opioid-insensitive mechanical pain in mice through a GluN3A-related excitatory mechanism.
18 · Architecture of the serotonergic projectome in the mouse brain
Cell · Published/issue date: 2026-09-10
Authors: Jun Ho Song, Drew Friedmann, Yunming Wu, Ashley Moses, Jalal Kenji Baruni, Yuan Yuan, Isabella Rana, Kelly Zhao, Tom Hindmarsh Sten, Qian Wang, Shuyun Alina Xiao, Xiaoke Chen, Scott W. Linderman, Liqun Luo
Abstract unavailable: after checking the publisher page, DOI registration records, and searches, no verifiable abstract could be obtained for this formal entry. This issue lists only the title, authors, journal, date, and link; a preprint on a similar topic but with a different title is not substituted for its content.
Highlight · The abstract for the formal entry is pending; no research conclusions are inferred for now.
Authors: Fabio V. Caixeta, Stina Lundberg, Akilandeswari Balasubramanian, Charlotte van Gelder, Ella Mercer, Guofeng Lou, Jörgen Jonsson, Åsa Konradsson-Geuken, Klas Kullander
To separate developmental effects from adult function, the authors conditionally removed SLC10A4 from serotonergic neurons in adult mice and examined the consequences for transmitter levels and antidepressant response. According to the abstract, extracellular serotonin fell in the prefrontal cortex, fluoxetine-induced changes were blunted, and the mice showed anxiety-like behavior and weaker antidepressant-like responses in some behavioral tests, while general locomotor ability showed no clear impairment in the same direction. The study supports a role for SLC10A4 in regulating the adult serotonin system and suggests that transmitter-related transport control can alter drug responses. Behavioral readouts come from a mouse model and cannot be mapped directly onto a human depression diagnosis or used to predict how a given patient will respond to medication.
Highlight · SLC10A4 in adult serotonergic neurons shapes extracellular serotonin levels and the mouse response to fluoxetine.
Proceedings of the National Academy of Sciences · Publication/issue date: 2026-09-22
Authors: Andrian G. Basargin, Andras Domokosa, Joseph J. Hennessey, Isak K. Aarrestad, Rohini Sambyal, Yara A. Khatib, Johanna Krüger, Lee E. Dunlap, Samuel J. Carter, Isabella A. Rebek, John L. McKee, Serena S. Schalk, Min Liu, James C. Fettinger, Monica A. Gonzalez, Abhay Potluri, Dean J. Tantillo, Oliver Fiehn, John D. McCorvy, David E. Olson
By comparing a set of simplified LSD-related analogs, the study asks which structural features track with pharmacological activity at different serotonin receptors. The nine analogs span 5-HT2A, 5-HT2B, and 5-HT2C, with the goal of teasing apart the functional contributions buried inside a complex molecule rather than preserving overall activity alone. The abstract notes that UCD0094 and UCD0076 show favorable preclinical safety profiles, with UCD0076 biased toward 5-HT2C and producing antipsychotic-like effects in mice. These results point toward candidates with more differentiated activity, but receptor pharmacology and animal behavior cannot predict clinical efficacy, and the safety gains described hold only within the preclinical comparisons performed so far.
Highlight · Breaking down LSD-related structures by function helps distinguish serotonin receptor activities and identify more selective candidates.
21 · TRI-611, a selective, brain-penetrant molecular glue degrader of ALK
Nature · Publication/issue date: 2026-09-09
Authors: Andrew R. Conery, Daniel S. La, Artyom A. Alekseyenko, David Marcoux, Aaron G. Bart, Matt L. Harlow, Patrick R. Arsenault, Nico R. Cantone, Rebecca L. Casaubon, Alexander Constan, Hari B. Kamadurai, Aravind Prasad Medikonda, Duncan E. Nunes, Hannah Szeto, Tim J. Wigle, Maolin Yu, Aleksandra Zagulyaeva, Christine M. Zarate, Lauren Highfield, Nobuyuki Kondo, Aaron N. Hata, Kenneth Ngo, Jisu Lee, Prafulla C. Gokhale, Caroline Germa, Kathleen I. Seyb, Patrick Trojer, Vito J. Palombella
TRI-611 is a molecular glue degrader aimed at ALK, designed for brain exposure and activity against resistant variants. The molecule promotes ALK–CRBN interaction through an interface outside the kinase active site, thereby inducing degradation of the target protein. Proteomic profiling supports its selectivity, and cell and patient-derived tumor models show activity against wild-type ALK fusion proteins as well as some that are resistant to tyrosine kinase inhibitors. Tumor regression in subcutaneous and intracranial models, together with synergy in combination with kinase inhibitors, forms the preclinical rationale for further development. The candidate has entered clinical development, but the animal and cell results in the abstract still do not translate into proven patient benefit.
Highlight · Degrading ALK through a molecular glue interface outside the active site offers another route for tackling resistant and intracranial tumor models.
22 · These six advances could change how drugs are delivered
Nature · Publication/issue date: 2026-09-09
Authors: Nic Fleming
News roundup: this piece gathers six advances that could change how drugs are delivered, with the intro citing examples such as circadian-rhythm-based dosing and DNA self-assembly. The material collected here does not lay out the methods, results, or validation stage of each technology, so only the general direction is noted; no full list of the six technologies or efficacy conclusions has been added.
Highlight · The report focuses on when, where, and in what form drugs reach their site of action.
23 · Special delivery: bacterial couriers are smuggling drugs into cancers
Nature · Publication/issue date: 2026-09-09
Authors: Liam Drew
News roundup: the article looks at research using bacteria to ferry drugs into tumors and discusses how this delivery approach might combine with chemotherapy, radiotherapy, and immunotherapy. The available intro is not enough to judge the targeting efficiency, applicable cancer types, or clinical maturity of specific platforms, so this approach is not described here as an already usable treatment option.
Highlight · Bacterial carriers are being explored for delivering drugs locally into tumors, but the intro provides no comparable efficacy data.
24 · Drug-delivering particles aim to breach the blood–brain barrier
Nature · Publication/issue date: 2026-09-09
Authors: Simon Makin
News roundup: the blood–brain barrier keeps many drugs out of brain tissue, and this report covers efforts to cross it with carriers such as nanoparticles and extracellular vesicles. The current material only indicates a direction for delivery; it gives no specific diseases, drug exposure levels, or clinical outcomes, so it cannot support the claim that any particular particle type has achieved reliable brain therapy.
Highlight · Particle carriers are one exploratory route to brain drug delivery, but crossing the barrier and improving treatment outcomes still have to be demonstrated separately.
25 · Dose once, treat forever: drugs are being engineered to last much longer
Nature · Publication/issue date: 2026-09-09
Authors: Ben Deighton
News roundup: the article covers engineering strategies to extend how long drugs act, aiming to stretch dosing intervals to months or longer, and discusses the potential value for regions with limited medical resources. The intro does not list the duration, indications, or safety results of each technology; the long-term treatment vision in the headline should not be read as a single dose delivering a permanent cure.
Highlight · Long-acting drugs could cut dosing frequency, but how long they last and what they can treat depend on the specific technology and validation results.
Protein Structure Prediction and Simulation (4 papers)
26 · Multi-state structure prediction of G protein-coupled receptor proteins via prompting on AlphaFold
Computational and Structural Biotechnology Journal · Publication/issue date: 2026-09-10
Authors: Zhigang Sun, Tao Zhang, Kexin Zhang, Anqi Pang, Jiale Yu, Liting Zeng, Sibei Yang, Suwen Zhao, Jie Zheng
A single GPCR can adopt multiple functional conformations, and a single predicted structure may not capture the states that matter most in drug research. PromptGPCR tries to steer AlphaFold-Multimer and AlphaFold 3 toward activated or inactive receptor conformations using biological sequence prompts with structural-biology meaning. The key idea is to feed state-related information into the prediction so that the model's output aligns better with the target functional state. The abstract reports that this strategy improves multi-state prediction over baselines and examines how the resulting models perform in molecular docking. It offers a way to add state control to existing prediction tools, but the abstract gives no single comparable effect size, and the predictions still need validation against experimental structures or functional data.
Highlight · Biological sequence prompts can guide AlphaFold toward GPCR conformations closer to the target functional state.
Authors: Yu Liu, Boming Kang, Peng Li, Qinghua Cui
iFold tries to distinguish how much individual residues matter in protein structure prediction, rather than letting every position influence the model the same way. The method uses representations from a protein language model to apply dynamic positional weights during AlphaFold 2 training, so predictions focus more on key residues. The abstract reports that across a benchmark of 3,599 targets, average error dropped by about 0.3 Å and the success rate improved, with further testing on 167 independent targets. Gains appeared for ribosomes, membrane proteins, and proteins lacking typical homology information, and the extra computational cost was low. The results suggest that residue-level prioritization can complement global structure learning, though the size of the improvement should be read in light of the benchmark composition and success criteria.
Highlight · Residue-importance signals from a language model can refocus what structure prediction learns.
28 · Rapid simulation of glycoprotein structures by grafting and steric exclusion of glycan conformer libraries
Cell · Publication/issue date: 2024-02-29
Authors: Yu-Xi Tsai, Ning-En Chang, Klaus Reuter, Hao-Ting Chang, Tzu-Jing Yang, Sören von Bülow, Vidhi Sehrawat, Noémie Zerrouki, Matthieu Tuffery, Michael Gecht, Isabell Louise Grothaus, Lucio Colombi Ciacchi, Yong-Sheng Wang, Min-Feng Hsu, Kay-Hooi Khoo, Gerhard Hummer, Shang-Te Danny Hsu, Cyril Hanus, Mateusz Sikora
The conformational diversity of glycans makes glycoprotein structure analysis computationally expensive. GlycoSHIELD draws on prebuilt libraries of glycan conformers, grafts them onto protein structures, and then filters out incompatible conformations by steric exclusion, giving a fast estimate of the dynamic shield the glycans form. The authors validated the approach with molecular dynamics, small-angle scattering, cryo-EM, and mass spectrometry data and demonstrated its use across several glycoprotein systems. The abstract emphasizes that the analysis runs in a short time even on an ordinary personal computer, making it well suited to exploring how glycosylation affects surface accessibility. That speed comes from approximating the problem, so the resulting conformational ensembles should not be treated as a substitute for full protein–glycan coupled dynamics.
Highlight · Grafting from conformer libraries plus steric filtering can quickly map the dynamic shielding glycans impose on a protein surface.
29 · Transfer learning to leverage larger datasets for improved prediction of protein stability changes
Proceedings of the National Academy of Sciences · Publication/issue date: 2024-02-06
Authors: Henry Dieckhaus, Michael Brocidiacono, Nicholas Z. Randolph, Brian Kuhlman
Predicting how mutations alter protein stability requires striking a balance between data scale and model generalization. ThermoMPNN uses transfer learning to repurpose the structure-related representations learned by ProteinMPNN for stability-change prediction, training further on larger experimental datasets. The idea is to reuse an existing model's understanding of local protein environments and avoid relearning every feature from limited stability data. The abstract reports that the model improves on relevant benchmarks while keeping the prediction pipeline lightweight enough to scale. This supports a complementary role for large datasets and pretrained representations, but reliability for any specific mutation still depends on training coverage, structure quality, and independent experimental validation.
Highlight · Transferring structure-model representations to a stability task makes large-scale mutational data easier to exploit.
Viral Genomes and Replication Mechanisms (2 papers)
30 · Generative design of bacteriophages with genome language models
Science · Publication/issue date: 2026-08-06
Authors: Samuel H. King, Claudia L. Driscoll, David B. Li, Daniel Guo, Aditi T. Merchant, Garyk Brixi, Max E. Wilkinson, Brian L. Hie
This study asks whether genome language models can produce designs that actually work biologically at the scale of a whole phage. The authors report 16 viable designer phages with varying fitness, and combine structural analysis with an examination of how genomic changes relate to viral particle features. Under laboratory conditions, some combinations also acted against phage-tolerant E. coli. The results show that generative model output can go beyond looking like a plausible sequence and reach a testable functional level — but they also show that biological performance across designs is uneven. These findings are limited to the phages and bacterial systems tested; they cannot be generalized into universal designability, clinical efficacy, or predictable behavior against arbitrary hosts.
Highlight · Phages offer an experimental system for testing whether genome language model output can translate into full biological function.
31 · Structure and operating principles of a monkeypox virus replisome
Nature · Published/issue date: 2026-09-02
Authors: Zishuo Yu, Pradeep Sathyanarayana, Joel M. J. Tan, Side Hu, Xiaoyi Fan, Angela Gao, Philip J. Kranzusch, Joseph J. Loparo, Jonathan Abraham
This work uses cryo-EM to analyze how the monkeypox virus replication machinery is put together and how it operates, focusing on how the polymerase complex and the E5 helicase coordinate. The different conformations provide a structural basis for understanding how template handling, unwinding, and nucleic acid synthesis are coordinated, extending our picture of replication from isolated proteins to a more complete assembly. Its main contribution is revealing the spatial relationships between components and the possible principles behind their coordinated movement, giving later mechanistic studies a reference point. Structural observation alone does not mean an effective drug has been found, nor does it directly establish the full share and switching order of these conformations during infection. This entry covers the basic mechanism only and does not extend to building or engineering a replication system.
Highlight · The structure of the replication assembly reveals the spatial basis for how helicase and polymerase work in concert.
Transcription initiation and chromatin regulation (2 papers)
32 · Structural basis of a SWI/SNF-associated transcription pre-initiation complex
Cell · Published/issue date: 2026-09-08
Authors: Shigeki Nagai, Chun Yang, Dong-Hua Chen, Michael J. Trnka, David A. Bushnell, Alma L. Burlingame, Kenji Murakami, Roger D. Kornberg
Supplemental abstract from the publisher. SWI/SNF is usually understood as clearing the way for transcription by moving or removing nucleosomes; this study goes further, finding that it can integrate directly into the RNA polymerase II transcription initiation machinery. Combining cryo-EM, crosslinking, and mass spectrometry, the authors resolved a complex containing the PIC, Mediator, SWI/SNF, and a nucleosome, in which the entire assembly can dimerize. The structure suggests that SWI/SNF both stabilizes the organization of the initiation complex and helps remodel the downstream +1 nucleosome. The model links two DNA translocases — Snf2 in SWI/SNF and Ssl2 in TFIIH — offering an explanation for how chromatin remodeling and transcription start are synchronized; the physiological state of the assembly still needs more in-situ evidence.
Highlight · SWI/SNF can be a component of the initiation complex, directly linking nucleosome remodeling to the transcription machinery.
This review lays out the molecular basis of RNA polymerase II transcription initiation, discussing general transcription factors, Mediator, and the chromatin environment within a single framework. It covers TFIID-mediated promoter recognition, assembly of the pre-initiation complex, the influence of the +1 nucleosome on start site position and direction, and DNA opening, promoter escape, and the transition into elongation. Phosphorylation of the polymerase C-terminal domain and the orderly exchange of regulatory factors are key threads tying these steps together. The piece is well suited to clarifying the logical relationships between different structural snapshots and to understanding how initiation is dynamically regulated. It synthesizes existing progress in the field, so one should not mistake the integrated model of these steps for a complete process directly observed in a single new experiment.
Highlight · Transcription initiation depends on coordinated changes and an orderly hand-off among chromatin, initiation factors, and polymerase state.
34 · MemBrain v2: an end-to-end tool for analyzing membranes in cryo-electron tomography
Nature Methods · Published/issue date: 2026-09-08
Authors: Lorenz Lamm, Simon Zufferey, Hanyi Zhang, Ricardo D. Righetto, Florent Waltz, Wojciech Wietrzynski, Kevin A. Yamauchi, Alister Burt, Ye Liu, Antonio Martinez-Sanchez, Sebastian Ziegler, Fabian Isensee, Julia A. Schnabel, Benjamin D. Engel, Tingying Peng
Supplemental abstract based on the abstract registered by the publisher. Membrane analysis in cryo-ET is hampered by low signal-to-noise ratios, the missing wedge, and the complexity of membrane-associated particles. MemBrain v2 brings membrane segmentation, particle localization, and spatial statistics into one workflow: MemBrain-seg uses diverse training data to improve segmentation generalization across different tomographic conditions; MemBrain-pick combines geometric constraints with deep learning to reduce the heavy manual annotation burden when localizing membrane proteins; and MemBrain-stats computes spatial metrics for particle distribution and intra-membrane organization. Its contribution is connecting analysis steps that are usually scattered, making them easier to apply across many datasets. The abstract gives no single accuracy figure, so in practice users should still check segmentation and localization quality on their own samples.
Highlight · MemBrain v2 links membrane segmentation, membrane protein localization, and spatial statistics into a complete cryo-ET analysis pipeline.
35 · Streamlined montage cryo-electron tomography for exploring the ultrastructure of cells and tissues
Structure · Published/issue date: 2026-09-03
Authors: Ryan Hylton, Micaela Boiero Sanders, Adriana Prajica, Gavin Rice, Stefan Raunser
Supplemental abstract from the publisher. Conventional cryo-ET can resolve molecular-scale structure but only over a small field of view; montage tomography can widen that field, yet it often requires complicated dose allocation or hardware modifications. The MOSAIC workflow stays as close as possible to standard automated acquisition, using MontageMaker to merge overlapping tilt series. According to the abstract, the resulting montage tomograms retain nanometer-scale detail despite uneven cumulative dose and work across a range of biological samples. Its main value lies in lowering the barrier to wider fields of view, making it easier for labs that already have automated cryo-ET acquisition set up to examine cell and tissue ultrastructure. The specific resolution, dose effects, and montage quality still need to be checked against the sample and data at hand.
Highlight · MOSAIC and MontageMaker extend the cryo-ET field of view with minimal changes to acquisition.
36 · Cryo-EM structure of TGFBIp fibrils driven by a corneal dystrophy–linked mutation enables design of peptide inhibitors of aggregation
Proceedings of the National Academy of Sciences · Published/issue date: 2026-09-15
Authors: Yi Xiao Jiang, Lukasz Salwinski, Michael R. Sawaya, Peng Ge, Liisa Lutter, Xinyi Cheng, Carolyn J. Hu, Hillary Hernandez, David R. Boyer, Conrad Wang, Filipe A. Melo, Duilio Cascio, David S. Eisenberg
Corneal dystrophy–linked TGFBIp variants can promote amyloid fibril formation. Using cryo-EM, the authors studied the FAS1-4 segment carrying the V624M variant and resolved a fibril structure built from symmetric protofilaments and an internal channel. Its core region contains a segment found in patient deposits, linking the experimental structure to disease material. On that basis the team designed peptide inhibitors: G1 and H4 suppressed aggregation and fibril extension in vitro in a concentration-dependent manner, with support from fluorescence, fractionation, and electron microscopy. The work demonstrates a path from a pathology-relevant fibril structure to the design of aggregation inhibitors, but the evidence remains at the in vitro level and cannot be directly extrapolated into therapeutic benefit in corneal disease.
Highlight · The pathology-relevant TGFBIp fibril structure provides a direct template for designing peptides that block aggregation.
Nature Chemical Biology · Published/issue date: 2026-09-08
Authors: Maja Bencun, Etienne Boileau, Christoph Dieterich
Quick take: this piece centers on TransCODE's analysis of the non-canonical human proteome, looking at translation products beyond standard annotation. The available abstract offers only this research direction — no numbers on what was identified, what fraction was validated, or any specific functional conclusions — so this issue does not treat the "dark proteome" as fully mapped.
Highlight · Non-canonical translation products are expanding our sense of where the human proteome ends.
Quick take: this entry discusses the role of graphical abstracts as visual summaries in science communication. The available teaser is very short, with no experimental comparisons, design guidelines, or data on communication outcomes, so this issue simply notes the topic and does not add production advice that the source does not support.
Highlight · Graphical abstracts present research highlights visually, but the current material offers no evaluation of their effect.
Outer membrane protein assembly in Gram-negative bacteria requires the chaperone SurA to work with the BAM complex, but how the two dynamically hand off substrates has remained unclear. The authors resolved multiple cryo-EM conformations of the E. coli SurA–BAM complex and observed that different SurA domains help regulate the relationship between the core region and the assembly machinery. The flexibility of the P1 and P2 regions, along with the interaction between P2 and BamE, provides a structural basis for binding and substrate hand-off. Comparing multiple states supports a model in which conformational changes drive substrate transfer forward. The work fills in a dynamic step in outer membrane biogenesis, though the pathway suggested by structural snapshots should still be read alongside functional experiments; a single conformation cannot pin down the full sequence of events.
Highlight · SurA's domain movements and its contacts with BAM together support the dynamic hand-off of outer membrane proteins.
40 · Structural insights into how the metallochaperones UreE and UreG interact to deliver a toxic metal to urease
Proceedings of the National Academy of Sciences · Published/issue date: 2026/09/08
Authors: Chun-Long Chan, Longson Tsz Hin Pang, Tung Choi, Ka-Chun Chan, Ka Lung Tsang, Ka Ming Lee, Kam-Bo Wong
Supplemental abstract based on publisher registration information. Urease maturation requires nickel ions, but free nickel is cytotoxic, so controlled transfer between metallochaperones is essential. The authors resolved the structure of the complex formed by UreE and UreG and found that the two make contact asymmetrically; GTP binding drives changes that help the complex assemble and bring the nickel-binding regions of both partners into adjacent positions. Together with in vitro analysis, the structural comparison suggests that nickel transfer then weakens the interaction between the chaperones, allowing metal-loaded UreG to be released. From this the study proposes a hand-off model allosterically regulated by both nucleotide and metal, explaining how cells promote enzyme maturation while limiting exposure to a toxic metal; this is not a direct validation of therapeutic effects against infection.
Highlight · GTP and nickel binding jointly control the assembly and dissociation of UreE–UreG, tying metal hand-off to the chaperone cycle.
Authors: Ke Hu, Yin-Hu Wang, Anthony Y. Tao, Maxwell McDermott, Miki Jishage, Li Zhong, Thomas Boehringer, Marcus J. Hines, Sergei B. Koralov, Ebbe Boedtkjer, Stefan Feske
Antibody-secreting cells have heavy metabolic demands during differentiation, and this study ties that process to intracellular acid–base balance. Loss of SLC4A7/NBCn1 caused intracellular acidification, changes in the lysosomal environment, and reduced mTORC1 activity, all of which impaired differentiation into antibody-secreting cells; forcing mTOR signaling back on in vitro partially restored the process. In mouse immunization and influenza infection experiments, antibody-secreting cells and antibody responses were also reduced. The effects of extracellular acidification lend further support to pH as an environmental factor in this regulation. The work shows that ion transport does more than maintain general homeostasis — it may also shape immune cell fate — though these findings do not mean that simply shifting external acid–base conditions would improve immune function in people.
Highlight · SLC4A7 sustains intracellular pH to support mTORC1 activity and antibody-secreting cell differentiation.
42 · TM184C is a GPCR-like regulator of intercellular exchange and autophagy
Nature · Published/issue date: 2026-09-09
Authors: Kyutae D. Lee, Samuel Taylor, Jennifer Arcuri, Shraddha Chandthakuri, Jacinda Pujols, Bruno Colon, Qiang Wang, Chenzhou Wu, Zhipeng Meng, Sebastian J. Thompson-Ceccato, Jonathan Mitchell, Defne Bayik, Alexandra Carbone, Vladlen Slepak, Tatiana I. Slepak, Scott M. Welford, Michael E. Ivan, Dazhi Wang, Blake O. Goldberg, Warner Baringer, Elvis Maradzike, Daniel G. Isom
The authors used homology relationships among large-scale predicted structures to find GPCR-like proteins that sequence comparison alone tends to miss, then went on to study the cellular function of TM184C. Their results link the protein to intracellular vesicles, microtubule-associated distribution, and the exchange of material through intercellular bridges, and reveal regulatory features involving β-arrestin and GRK. Functional analysis also suggests it restrains LC3B lipidation and autophagic flux, while cross-species findings point to a relatively ancient origin for these functions. The study shows how structural similarity can help uncover under-annotated regulators of signaling and transport. Still, a GPCR-like structure does not automatically mean a classic ligand has been identified or that the signaling mechanism is identical.
Highlight · Structural homology analysis links TM184C to intercellular exchange and autophagy regulation.
This study looked at how semaglutide affects ageing-related phenotypes and lifespan in aged female mice. Treatment starting at around twenty months of age improved several physiological measures and ageing-related markers, and continued treatment was accompanied by longer lifespan. The authors included a matched caloric restriction comparison and found that the two shared some effects, but not all — memory, exploratory behavior, and glucose regulation differed — suggesting the drug's effects may go beyond simply eating less. The results provide animal evidence that starting a metabolic intervention late in life may still help, and make a case for further work to separate reduced food intake from other mechanisms. Because the study is limited to one sex, strain, and set of conditions, it cannot be used to conclude that the drug extends lifespan in humans.
Highlight · Improvements in aged female mice point to the potential of late-life metabolic intervention; some effects cannot be explained by matched caloric restriction alone.
This issue covers the 43 papers added to a personal Zotero library between September 7 and September 13, 2026. Dates and times are Beijing time, based on when each item was added. Papers are grouped by topic and listed in reverse chronological order within each group, with the miscellaneous items at the end.
The summaries below were compiled by AI from abstracts, with missing abstracts filled in from the original sources; short communications and news briefs are marked separately and were not read in full. One paper had no verifiable abstract and is listed by metadata only. Publication and volume/issue dates are taken from the entry records; they may be later than the date the paper was added and do not indicate publication this week. For methods and conclusions, please consult the original articles.
Receptors, Ion Channels, and Membrane Protein Recognition (11 papers)
01 · Gating crosstalk in potassium channels
Proceedings of the National Academy of Sciences · Publication/issue date: 2026-09-15
Authors: Lyubin Hu, Bert L. de Groot, Ruo-Xu Gu
How the central cavity and selectivity filter of potassium channels open and close in concert is a key question for understanding ion conduction. The authors ran molecular dynamics simulations of six potassium channels spanning multiple subfamilies and found that opening and closing of the central cavity simultaneously changes the degree of cavity hydration and the width of the filter entrance. The former affects the free energy of potassium ions entering the cavity, while the latter regulates ion occupancy in the filter, with the fastest conduction occurring at an intermediate entrance size. Hydrophobic residues on the inner transmembrane helices link these two conformational changes. State differences seen in existing experimental structures also match the simulations, supporting a gating crosstalk model in which hydration and entrance geometry jointly regulate conduction. This consistency across channels hints at a potentially general principle, but it does not mean that all potassium channels follow exactly the same mechanism.
Highlight · The central cavity influences potassium conduction through two routes: the degree of hydration and the width of the filter entrance.
DOI: 10.1073/pnas.2605103123
02 · Structural insights into substrate uptake and inhibition by the human creatine transporter (hCRT)
Proceedings of the National Academy of Sciences · Publication/issue date: 2025-09-09
Authors: Xinyi Yuan, Jian Yin, Chang Liu, Xudong Chen, Meiying Chen, Yixue Wang, Zi Yang, Yue Wang, Li Jiang, Niyun Zhou, Xiaojuan Wang, Botong Liu, Zhaoqi Ma, Kaiyan Wang, Hongen Li, Sensen Zhang, Yongfeng Shang, Maojun Yang
The human creatine transporter hCRT handles cellular creatine uptake; when it goes wrong, it is linked to creatine deficiency syndromes, and it has also drawn attention in tumor metabolism research. The study solved cryo-EM structures of the transporter with no ligand, bound to creatine, and bound to the inhibitor RGX-202, showing how this twelve-transmembrane-helix transporter recognizes its substrate. The creatine-binding pocket is formed mainly by several transmembrane segments, and the inhibitor occupies the same region, offering a structural explanation for competitive inhibition. Comparing different states links the binding site to transport conformations and provides a reference for understanding how genetic variants impair function. These results mainly address the molecular basis of substrate uptake and blockade; they do not directly demonstrate the efficacy or safety of inhibiting this transporter in patients.
Highlight · Creatine and RGX-202 share a binding pocket, providing a structural basis for substrate recognition and competitive inhibition of hCRT.
DOI: 10.1073/pnas.2426135122
03 · Transport and inhibition mechanisms of the human creatine transporter
Cell Discovery · Publication/issue date: 2025-05-05
Authors: Jiahui Chen, Yimin Zhang, Nanhao Chen, Jingpeng Ge, Jie Yu
Key points from the original short communication: this entry has no standalone abstract, so the following is based on the readable portions of the publisher's text. The authors compared structures of the human creatine transporter with no ligand, bound to creatine, and bound to RGX-202, linking inward-open and inward-closed conformational states. Structural and functional analyses together clarify how substrate, ions, and inhibitor are recognized, and support the idea that RGX-202 blocks transport by competing for the substrate-binding region. The paper also uses the structural information to discuss pathogenic variants. This work addresses questions similar to those in another hCRT paper from the same group but has a different DOI, so both are retained separately; it provides evidence at the level of transport mechanism and does not mean that therapeutic benefit in disease has been clinically validated.
Highlight · Different ligand states help connect creatine recognition, conformational change, and transport inhibition.
DOI: 10.1038/s41421-025-00801-4
04 · Pre-activation and gating pathway of AMPA receptors revealed by full and partial agonists
Nature Structural & Molecular Biology · Publication/issue date: 2026-09-09
Authors: Thomas P. Newton, Muhammed Aktolun, Maria V. Yelshanskaya, Alexey A. Alekseev, Laura Y. Yen, Shanti Pal Gangwar, Ivan A. Sobolevsky, Maria G. Kurnikova, Alexander I. Sobolevsky
AMPA receptors pass through several short-lived states between ligand binding and ion channel opening. Using time-resolved cryo-EM, the study compared conformations of the GluA2–γ2 complex under full and partial agonists, covering the ligand-free, pre-activated, open, and desensitized stages. The degree of clamshell closure in the ligand-binding domain correlates with the downstream gating pathway: smaller closure can correspond to a pre-activated intermediate, whereas greater closure is associated with open or desensitized states. Single-channel recordings and molecular dynamics simulations complemented the structural interpretation. The results help explain how ligands of differing efficacy select different state distributions, and they caution against reducing open probability to whether ligand binding has occurred.
Highlight · Full and partial agonists shape the gating pathway of AMPA receptors by altering domain closure and the distribution of intermediate states.
DOI: 10.1038/s41594-026-01882-9
05 · Steric control of signaling bias in the immunometabolic receptor GPR84
Nature Communications · Publication/issue date: 2026-09-05
Authors: Pinqi Wang, Xuan Zhang, Abdul-Akim Guseinov, Laura Jenkins, Carl von Hallerstein, Jonathan D. Colburn, Rowan Ives, Vincent B. Luscombe, Sara Marsango, Listiana Oktavia, Arun Raja, David R. Greaves, Philip C. Biggin, Graeme Milligan, Cheng Zhang, Irina G. Tikhonova, Angela J. Russell
GPR84 takes part in immunometabolic signaling, and different ligands can produce varying degrees of bias between G protein and β-arrestin pathways. The authors compared three matched ligands: they had similar Gi activity but differed in β-arrestin response. Cryo-EM, molecular dynamics, and mutagenesis showed that steric changes in the binding pocket can be relayed through local residues, indirectly perturbing the receptor's internal network of polar interactions. In this way, a ligand can shift signaling preference without directly touching every downstream structural element. The study offers an atomic-level mechanistic clue to biased agonism, but whether structural bias translates into better disease intervention still needs independent validation in the relevant biological setting.
Highlight · Local steric effects from a ligand can be amplified through the receptor's internal network to shift GPR84's signaling preference.
DOI: 10.1038/s41467-026-77312-7
06 · Structural basis for modulation of group III mGlu receptors by transsynaptic interactions
Science Advances · Publication/issue date: 2026-09-04
Authors: William G. Ludlam, Chu-Ting Chang, Kristina Cechova, Brandon W. Liauw, Hwa-Jin Cho, Safoura Salar, Anjelique Sawh-Gopal, Afroza Parvin, Simrat K. Dhaliwal, Simran K. Dhaliwal, Tina Izard, Huan Bao, Anne M. Brown, Henry A. Dunn, Reza Vafabakhsh, Kirill A. Martemyanov
Group III metabotropic glutamate receptors not only recognize neurotransmitters but are also modulated by transsynaptic interactions. The study solved structures of complexes involving ELFN adhesion proteins, showing how cell-to-cell contact connects the receptor's ligand-binding region to its internal allosteric network. The ELFN binding site and its effect on receptor conformation explain previously observed functional modulation. The abstract also notes that receptor activation can strengthen association with ELFN, suggesting feedback regulation between the two, and uses the structural information to interpret variants linked to neurological disease. The work highlights how important the synaptic environment is for receptor behavior, but the interactions seen in structures cannot be equated directly with net effects in an intact neural circuit.
Highlight · Transsynaptic adhesion interactions can modulate group III mGlu receptors through an allosteric network.
DOI: 10.1126/sciadv.aed3793
07 · The molecular basis of force selectivity in PIEZO2
Nature · Publication/issue date: 2026-05
Authors: Eric M. Mulhall, Oleg Yarishkin, Rose Z. Hill, Anna K. Koster, Ardem Patapoutian
Both PIEZO1 and PIEZO2 respond to mechanical stimuli, but they prefer different types of force. Combining MINFLUX imaging in intact cells with electrophysiology, the authors compared the conformations and gating behavior of the two channels. PIEZO2 proved more rigid, and different mechanical stimuli could drive conformational changes in different directions, with those changes tied to opening behavior. The authors further found that filamin B–mediated actin linkage helps give PIEZO2 its selectivity for indentation stimuli, and they observed the relevant spatial proximity at sensory endings. The study connects the channel's own structural features with cytoskeletal coupling, suggesting that mechanical selectivity must be understood in the cellular context rather than explained by the shape of an isolated protein alone.
Highlight · PIEZO2's force selectivity arises from a combination of the channel's conformational features and its actin linkage.
DOI: 10.1038/s41586-026-10182-7
08 · Discovery of an antagonist nanobody targeting the oxytocin receptor with tocolytic effect in mice
Nature Communications · Publication/issue date: 2026-09-11
Authors: Jia Nie, Kaixuan Gao, Xin Zhang, Hengyu Meng, Qi Li, Xiaowen Zhang, Yanying Wu, Hongtao Zeng, Waner Liu, Lijuan Liu, Lunbo Tan, Xiaoou Sun, Xiangyu Liu, Weishe Zhang
The oxytocin receptor is an important target for regulating uterine contractions, and achieving selectivity among closely related receptors is a persistent challenge in drug research. The authors used AI-assisted screening to look for antagonist nanobodies and obtained candidates that act on the oxytocin receptor. According to the abstract, one candidate's antagonism was comparable to that of atosiban while showing better receptor subtype selectivity and working through a different mechanism; the study also observed effects consistent with suppressed uterine contractions in mice. The work offers a route to modulating this receptor with antibodies and illustrates how computational screening can be combined with functional validation. The current results say nothing directly about effectiveness in human pregnancy, fetal safety, or the conditions for long-term use.
Highlight · Antagonist nanobodies offer a candidate route to more subtype-selective modulation of the oxytocin receptor.
DOI: 10.1038/s41467-026-77680-0
09 · Configurational diversity of metabotropic glutamate receptor complexes with beta-arrestins
Nature Communications · Publication/issue date: 2026-09-10
Authors: Dagan C. Marx, Alberto J. Gonzalez-Hernandez, Joon Lee, Dirk Siepe, Willem F. Weber, Kevin Huynh, Pamela N. Gallo, Sheida Sharghi Moshtaghin, Anisul Arefin, Johannes Broichhagen, Marian Kalocsay, David Eliezer, Joshua Levitz
Metabotropic glutamate receptors do not adopt a single configuration when binding β-arrestin. Using single-molecule pull-down assays, the study examined binding strength and stoichiometry across different receptor complexes, observing arrangements such as 2:1 or 2:2. Both the receptor tail and the core region can participate in coupling, and mGlu8 homo- and heterocomplexes display different cis and trans modes of action. The abstract further describes complexes that bind two types of β-arrestin at once, or that accommodate β-arrestin together with a G protein. The results broaden our understanding of receptor signaling platforms and suggest that summarizing function simply as whether β-arrestin is recruited may miss important differences, though the physiological contribution of each configuration still needs to be determined separately.
Highlight · mGlu receptors can organize β-arrestin signaling complexes through a variety of stoichiometries and contact modes.
DOI: 10.1038/s41467-026-77555-4
10 · Structural basis of β-arrestin coupling and transducer selectivity in PAC1R
Acta Pharmacologica Sinica · Publication/issue date: 2026-09-08
Authors: Min Zhang, Xin Li, Qing-ning Yuan, Wen Hu, H. Eric Xu, Li-hua Zhao
PAC1R can couple to different signaling proteins, and this study solves the cryo-EM structure of the receptor bound to β-arrestin 1 and compares it with the Gs-coupled state. When β-arrestin binds, the receptor shows a rearrangement of transmembrane helix 5 and an inward shift of helix 6, creating a contact environment distinct from that of the other transducer. The comparison also shows that PTH1R, which belongs to the same receptor class, does not use exactly the same β-arrestin coupling mode; both the receptor's own structure and the finger loop of β-arrestin contribute to selectivity. The takeaway is that no single receptor's pattern can stand in for signaling bias across the whole family — the specific receptor–transducer interface is a key level at which to understand functional differences.
Highlight · PAC1R selects its transducer through specific transmembrane conformations and interfaces, and β-arrestin coupling differs from receptor to receptor.
DOI: 10.1038/s41401-026-01924-4
11 · How cholesterol interacts with membrane proteins: an exploration of cholesterol-binding sites including CRAC, CARC, and tilted domains
Frontiers in Physiology · Published/issue date: 2013-02-28
Authors: Jacques Fantini, Francisco J. Barrantes
This review examines how cholesterol interacts with membrane proteins, focusing on structural features that may mediate recognition, including CRAC, CARC, and tilted transmembrane domains. Starting from cholesterol's rigid sterol ring, flexible side chain, and asymmetric surface, it maps out its compatibility with different amino acid environments and explains why both local sequence and transmembrane geometry can influence binding. It supplies the basic concepts for reading studies on cholesterol's regulation of membrane proteins and helps connect sequence patterns with spatial contacts. One caveat to keep in mind: spotting a candidate sequence motif is not by itself proof of actual binding or a functional effect — structure, membrane context, and experimental evidence all still have to line up.
Highlight · Cholesterol recognition depends on the local chemical environment and transmembrane geometry; a candidate motif alone is not sufficient evidence of binding.
DOI: 10.3389/fphys.2013.00031
Neuroscience and Neuropsychiatric Interventions (9 papers)
12 · To sleep and dream: unraveling narcolepsy
Proceedings of the National Academy of Sciences · Published/issue date: 2026-09-15
Authors: Jeffrey M. Friedman
This piece traces how narcolepsy research moved from disease phenotype toward molecular mechanism. Mignot's team started from inherited narcolepsy in dogs and used forward genetics to uncover defects related to the orexin receptor; Yanagisawa's team approached it from the orexin ligand and knockout mice, producing evidence that dovetails with the first. Together the two lines of work established the orexin system's central role in maintaining wakefulness and stabilizing the switch between sleep and waking, and laid the groundwork for later treatment strategies. The article is really about the process of discovery and its conceptual impact, making it a good way to grasp how genetics, animal models, and neuropeptide signaling fit together in sleep research. It is a historical and mechanistic review, and should not be read as a new clinical drug trial.
Highlight · Canine genetics and reverse genetics in mice together revealed the orexin system's key role in maintaining wakefulness.
DOI: 10.1073/pnas.2622458123
13 · Psilocybin-assisted therapy for treatment-resistant major depressive disorder in a public healthcare setting: a randomized controlled trial
Nature Medicine · Published/issue date: 2026-08-06
Authors: James J. Rucker, Tim Mantingh, Jess Kerr-Gaffney, Catherine Bird, Petrina Chu, Nadav Liam Modlin, Kete Campbell-Coker, Sadie Hambleton, Paige Seath, Rebecca Hignett, Diede Fennema, Elliot Hampsey, Dimosthenis Tsapekos, Rebecca J. Thomas, Joseph Cattell, Hassan Jafari, Camilla Day, Anya Borissova, Miranda Lloyd, Theo Boardman-Pretty, Mathieu Seynaeve, Famia Askari, Michael Creed, Luke Baxter, Raphael Rifkin-Zybutz, Matt Butler, Luke A. Jelen, Ben Carter, Allan H. Young
This single-center, randomized, controlled feasibility trial in a public healthcare setting enrolled sixty people with treatment-resistant major depressive disorder, comparing a single dose of psilocybin with placebo; both groups also received psychological support. Fifty-nine participants completed the study, with primary observation over six weeks. The abstract reports a between-group difference in MADRS score of −10.41 at week three, with improvement sustained through week six; at the same time, the drug group reported more non-serious adverse events than the placebo group. The results support moving ahead with larger confirmatory trials within real healthcare systems, and they also point to psychological support and adverse-event assessment as essential parts of the protocol. Given the sample size, follow-up duration, and the study's feasibility design, these findings are not enough to establish long-term efficacy or broad applicability.
Highlight · A small randomized trial in a public healthcare setting showed short-term signs of improvement, but larger and longer-term validation is still needed.
DOI: 10.1038/s41591-026-04541-0
14 · Psilocybin as a treatment for repetitive mild head injury: evidence from neuroradiology and molecular biology
Communications Biology · Published/issue date: 2026-09-08
Authors: Eric K. Brengel, Bryce Axe, Ashwath N. Maheswari, Muhammad I. Abeer, Richard J. Ortiz, Taylor J. Woodward, Reagan Walhof, Wenwen Du, Rachel Utama, Brandon Rodrigue, Noah J. Cavallaro, Anna Brettler, Courtney Sawada, Anna Moore, Shreyas Balaji, Praveen P. Kulkarni, Argel Aguilar-Valles, Heather B. Bradshaw, Michael A. Gitcho, Craig F. Ferris
This study looked at how psilocybin affects imaging and molecular readouts in a model of repetitive mild head injury in adult female rats. The abstract reports reduced vasogenic edema, restored vascular reactivity, and altered resting-state functional connectivity, along with lower phosphorylated tau and changes in BDNF/TrkB signaling and lipid-related pathways. Together these readouts suggest the intervention may act on cerebrovascular state, network activity, and post-injury molecular responses at once. Changes at different levels offer leads for further work, but the causal chain between them cannot be pinned down from correlational measures alone. Both the subjects and the evidence come from animal models, so this is no basis for recommending psilocybin for concussion or repeated head injury in people.
Highlight · Imaging and molecular changes in an animal model point to a possible injury-intervention mechanism, but there is still a wide gap before judging efficacy in humans.
DOI: 10.1038/s42003-026-10804-w
15 · Oxytocin modulates white matter functional connectivity tensor dynamics and circuit-targeted reconfiguration in depressive symptoms
Communications Biology · Published/issue date: 2026-09-07
Authors: Xiaodong Zhang, Chunmei Lan, Peng Qing, Xiaolei Xu, Yuanshu Chen, Juan Kou, Lei Xu, Xinqi Zhou, Dezhong Yao, Benjamin Becker, Keith M. Kendrick, Weihua Zhao
The authors studied depression-related changes in white matter functional connectivity tensors and asked whether oxytocin can shift these dynamic measures. The clinical sample included major depression, generalized anxiety, and healthy controls; the relevant metric in the genu of the corpus callosum was reduced in the depression group, and the authors then examined its associations in a subclinical sample. In people high in depressive traits, oral oxytocin coincided with higher local metrics, reconfiguration of genu-to-prefrontal connectivity, and changes in how negative social information was processed. The work links white matter functional representations to circuit-level intervention responses and points to candidate markers. That said, changes in imaging metrics are not the same as symptom benefit, and they are not yet usable as a diagnostic or treatment standard in practice.
Highlight · White matter connectivity dynamics may reflect depression-related circuit states and show measurable changes in response to oxytocin.
DOI: 10.1038/s42003-026-10914-5
16 · Psilocybin prevents chemotherapy-induced peripheral neuropathy by preserving mitochondrial trafficking
Science · Published/issue date: 2026-09-03
Authors: Mario Heles, Lilach Pasvolsky, Hinduja Sathishkumar, Shorook Naara, Yen Vu, Caitlyn L. Stewart, Tongxin Xie, Frederico O. Gleber-Netto, William McCarthy, Dan Yaniv, Shashank S. Kamal, Hajira Elahi, Pengyu Zhu, Megan L. Uhelski, Jordan Chatwin, Andrew Lara, Danielle L. Stolley, Michael R. Migden, Z-Hye Lee, Shiyanth Thevasagayampillai, Preethi H. Gunaratne, Kaoutar Ait-Ahmad, M. J. Kuykendall, Sebnem Ece Eksi, Lorenzo Cohen, Jeremy C. Borniger, Eyal Gottlieb, Gregory H. Jones, Patrick M. Dougherty, Moran Amit
This study examined whether psilocybin prevents peripheral neuropathy in animal models of platinum- and taxane-based chemotherapy. The abstract reports that it reduced neuropathy across repeated chemotherapy cycles without weakening the antitumor effects in the models tested. Mechanistic analysis tied the protective effect to preserved mitochondrial trafficking within axons, involving TrkB downstream signaling, motor proteins, and regulation of mitochondrial anchoring, with related changes in prefrontal activity also observed. The findings suggest that redistributing mitochondria along axons may be one route to neuroprotection. Even so, this evidence is still preclinical: it cannot tell us whether combining psilocybin with the various chemotherapy regimens used in people is safe, nor should it become a reason to try preventing neurotoxicity on one's own.
Highlight · Preclinical results link chemotherapy neuroprotection to preserved mitochondrial trafficking in axons.
DOI: 10.1126/science.aec6116
17 · A descending glycinergic circuit drives opioid-resistant mechanical pain via spinal GluN3A excitatory glycine receptors in mice
Neuron · Published/issue date: 2026-09-03
Authors: Dong Dong, Jiantao Huo, Guangjuan Yin, Kaifang Duan, Kaiyuan Hu, Dongmei Su, Xi Liu, Changyi Zhang, Shujia Zhu, Tanhan Xiang, Ruolin Wang, Changlin Li, Chao Guo, Minshi Chen, Longzhen Cheng
This supplementary abstract comes from the author abstract indexed in PubMed. Working in mice, the researchers identified a descending glycinergic circuit running from the rostral ventromedial medulla to the spinal dorsal horn that can drive mechanical allodynia insensitive to morphine. The pathway connects GlyT2-positive neurons to somatostatin-positive spinal neurons, and craniofacial pain can recruit it through the relevant inputs. Mechanistically, the released glycine acts on GluN3A-containing NMDA receptors, producing an unconventional excitatory effect at this site and bypassing opioid inhibition. The results offer a circuit- and receptor-level explanation for hard-to-treat mechanical pain and suggest potential targets; the evidence comes from mice, so no conclusions can yet be drawn about analgesic efficacy in humans.
Highlight · Glycine can drive opioid-insensitive mechanical pain in mice through a GluN3A-related excitatory mechanism.
DOI: 10.1016/j.neuron.2026.08.003
18 · Architecture of the serotonergic projectome in the mouse brain
Cell · Published/issue date: 2026-09-10
Authors: Jun Ho Song, Drew Friedmann, Yunming Wu, Ashley Moses, Jalal Kenji Baruni, Yuan Yuan, Isabella Rana, Kelly Zhao, Tom Hindmarsh Sten, Qian Wang, Shuyun Alina Xiao, Xiaoke Chen, Scott W. Linderman, Liqun Luo
Abstract unavailable: after checking the publisher page, DOI registration records, and searches, no verifiable abstract could be obtained for this formal entry. This issue lists only the title, authors, journal, date, and link; a preprint on a similar topic but with a different title is not substituted for its content.
Highlight · The abstract for the formal entry is pending; no research conclusions are inferred for now.
DOI: 10.1016/j.cell.2026.08.022
19 · Serotonergic SLC10A4 regulates extracellular serotonin levels and antidepressant responsiveness
Molecular Psychiatry · Published/issue date: 2026-10
Authors: Fabio V. Caixeta, Stina Lundberg, Akilandeswari Balasubramanian, Charlotte van Gelder, Ella Mercer, Guofeng Lou, Jörgen Jonsson, Åsa Konradsson-Geuken, Klas Kullander
To separate developmental effects from adult function, the authors conditionally removed SLC10A4 from serotonergic neurons in adult mice and examined the consequences for transmitter levels and antidepressant response. According to the abstract, extracellular serotonin fell in the prefrontal cortex, fluoxetine-induced changes were blunted, and the mice showed anxiety-like behavior and weaker antidepressant-like responses in some behavioral tests, while general locomotor ability showed no clear impairment in the same direction. The study supports a role for SLC10A4 in regulating the adult serotonin system and suggests that transmitter-related transport control can alter drug responses. Behavioral readouts come from a mouse model and cannot be mapped directly onto a human depression diagnosis or used to predict how a given patient will respond to medication.
Highlight · SLC10A4 in adult serotonergic neurons shapes extracellular serotonin levels and the mouse response to fluoxetine.
DOI: 10.1038/s41380-026-03697-y
20 · Deconstruction of lysergic acid diethylamide
Proceedings of the National Academy of Sciences · Publication/issue date: 2026-09-22
Authors: Andrian G. Basargin, Andras Domokosa, Joseph J. Hennessey, Isak K. Aarrestad, Rohini Sambyal, Yara A. Khatib, Johanna Krüger, Lee E. Dunlap, Samuel J. Carter, Isabella A. Rebek, John L. McKee, Serena S. Schalk, Min Liu, James C. Fettinger, Monica A. Gonzalez, Abhay Potluri, Dean J. Tantillo, Oliver Fiehn, John D. McCorvy, David E. Olson
By comparing a set of simplified LSD-related analogs, the study asks which structural features track with pharmacological activity at different serotonin receptors. The nine analogs span 5-HT2A, 5-HT2B, and 5-HT2C, with the goal of teasing apart the functional contributions buried inside a complex molecule rather than preserving overall activity alone. The abstract notes that UCD0094 and UCD0076 show favorable preclinical safety profiles, with UCD0076 biased toward 5-HT2C and producing antipsychotic-like effects in mice. These results point toward candidates with more differentiated activity, but receptor pharmacology and animal behavior cannot predict clinical efficacy, and the safety gains described hold only within the preclinical comparisons performed so far.
Highlight · Breaking down LSD-related structures by function helps distinguish serotonin receptor activities and identify more selective candidates.
DOI: 10.1073/pnas.2603412123
Drug Discovery and Delivery (5 papers)
21 · TRI-611, a selective, brain-penetrant molecular glue degrader of ALK
Nature · Publication/issue date: 2026-09-09
Authors: Andrew R. Conery, Daniel S. La, Artyom A. Alekseyenko, David Marcoux, Aaron G. Bart, Matt L. Harlow, Patrick R. Arsenault, Nico R. Cantone, Rebecca L. Casaubon, Alexander Constan, Hari B. Kamadurai, Aravind Prasad Medikonda, Duncan E. Nunes, Hannah Szeto, Tim J. Wigle, Maolin Yu, Aleksandra Zagulyaeva, Christine M. Zarate, Lauren Highfield, Nobuyuki Kondo, Aaron N. Hata, Kenneth Ngo, Jisu Lee, Prafulla C. Gokhale, Caroline Germa, Kathleen I. Seyb, Patrick Trojer, Vito J. Palombella
TRI-611 is a molecular glue degrader aimed at ALK, designed for brain exposure and activity against resistant variants. The molecule promotes ALK–CRBN interaction through an interface outside the kinase active site, thereby inducing degradation of the target protein. Proteomic profiling supports its selectivity, and cell and patient-derived tumor models show activity against wild-type ALK fusion proteins as well as some that are resistant to tyrosine kinase inhibitors. Tumor regression in subcutaneous and intracranial models, together with synergy in combination with kinase inhibitors, forms the preclinical rationale for further development. The candidate has entered clinical development, but the animal and cell results in the abstract still do not translate into proven patient benefit.
Highlight · Degrading ALK through a molecular glue interface outside the active site offers another route for tackling resistant and intracranial tumor models.
DOI: 10.1038/s41586-026-10998-3
22 · These six advances could change how drugs are delivered
Nature · Publication/issue date: 2026-09-09
Authors: Nic Fleming
News roundup: this piece gathers six advances that could change how drugs are delivered, with the intro citing examples such as circadian-rhythm-based dosing and DNA self-assembly. The material collected here does not lay out the methods, results, or validation stage of each technology, so only the general direction is noted; no full list of the six technologies or efficacy conclusions has been added.
Highlight · The report focuses on when, where, and in what form drugs reach their site of action.
DOI: 10.1038/d41586-026-02654-7
23 · Special delivery: bacterial couriers are smuggling drugs into cancers
Nature · Publication/issue date: 2026-09-09
Authors: Liam Drew
News roundup: the article looks at research using bacteria to ferry drugs into tumors and discusses how this delivery approach might combine with chemotherapy, radiotherapy, and immunotherapy. The available intro is not enough to judge the targeting efficiency, applicable cancer types, or clinical maturity of specific platforms, so this approach is not described here as an already usable treatment option.
Highlight · Bacterial carriers are being explored for delivering drugs locally into tumors, but the intro provides no comparable efficacy data.
DOI: 10.1038/d41586-026-02657-4
24 · Drug-delivering particles aim to breach the blood–brain barrier
Nature · Publication/issue date: 2026-09-09
Authors: Simon Makin
News roundup: the blood–brain barrier keeps many drugs out of brain tissue, and this report covers efforts to cross it with carriers such as nanoparticles and extracellular vesicles. The current material only indicates a direction for delivery; it gives no specific diseases, drug exposure levels, or clinical outcomes, so it cannot support the claim that any particular particle type has achieved reliable brain therapy.
Highlight · Particle carriers are one exploratory route to brain drug delivery, but crossing the barrier and improving treatment outcomes still have to be demonstrated separately.
DOI: 10.1038/d41586-026-02655-6
25 · Dose once, treat forever: drugs are being engineered to last much longer
Nature · Publication/issue date: 2026-09-09
Authors: Ben Deighton
News roundup: the article covers engineering strategies to extend how long drugs act, aiming to stretch dosing intervals to months or longer, and discusses the potential value for regions with limited medical resources. The intro does not list the duration, indications, or safety results of each technology; the long-term treatment vision in the headline should not be read as a single dose delivering a permanent cure.
Highlight · Long-acting drugs could cut dosing frequency, but how long they last and what they can treat depend on the specific technology and validation results.
DOI: 10.1038/d41586-026-02660-9
Protein Structure Prediction and Simulation (4 papers)
26 · Multi-state structure prediction of G protein-coupled receptor proteins via prompting on AlphaFold
Computational and Structural Biotechnology Journal · Publication/issue date: 2026-09-10
Authors: Zhigang Sun, Tao Zhang, Kexin Zhang, Anqi Pang, Jiale Yu, Liting Zeng, Sibei Yang, Suwen Zhao, Jie Zheng
A single GPCR can adopt multiple functional conformations, and a single predicted structure may not capture the states that matter most in drug research. PromptGPCR tries to steer AlphaFold-Multimer and AlphaFold 3 toward activated or inactive receptor conformations using biological sequence prompts with structural-biology meaning. The key idea is to feed state-related information into the prediction so that the model's output aligns better with the target functional state. The abstract reports that this strategy improves multi-state prediction over baselines and examines how the resulting models perform in molecular docking. It offers a way to add state control to existing prediction tools, but the abstract gives no single comparable effect size, and the predictions still need validation against experimental structures or functional data.
Highlight · Biological sequence prompts can guide AlphaFold toward GPCR conformations closer to the target functional state.
DOI: 10.34133/csbj.0179
27 · Enhancing protein structure prediction accuracy by prioritizing important residues using protein language models
Communications Biology · Publication/issue date: 2026-09-08
Authors: Yu Liu, Boming Kang, Peng Li, Qinghua Cui
iFold tries to distinguish how much individual residues matter in protein structure prediction, rather than letting every position influence the model the same way. The method uses representations from a protein language model to apply dynamic positional weights during AlphaFold 2 training, so predictions focus more on key residues. The abstract reports that across a benchmark of 3,599 targets, average error dropped by about 0.3 Å and the success rate improved, with further testing on 167 independent targets. Gains appeared for ribosomes, membrane proteins, and proteins lacking typical homology information, and the extra computational cost was low. The results suggest that residue-level prioritization can complement global structure learning, though the size of the improvement should be read in light of the benchmark composition and success criteria.
Highlight · Residue-importance signals from a language model can refocus what structure prediction learns.
DOI: 10.1038/s42003-026-10725-8
28 · Rapid simulation of glycoprotein structures by grafting and steric exclusion of glycan conformer libraries
Cell · Publication/issue date: 2024-02-29
Authors: Yu-Xi Tsai, Ning-En Chang, Klaus Reuter, Hao-Ting Chang, Tzu-Jing Yang, Sören von Bülow, Vidhi Sehrawat, Noémie Zerrouki, Matthieu Tuffery, Michael Gecht, Isabell Louise Grothaus, Lucio Colombi Ciacchi, Yong-Sheng Wang, Min-Feng Hsu, Kay-Hooi Khoo, Gerhard Hummer, Shang-Te Danny Hsu, Cyril Hanus, Mateusz Sikora
The conformational diversity of glycans makes glycoprotein structure analysis computationally expensive. GlycoSHIELD draws on prebuilt libraries of glycan conformers, grafts them onto protein structures, and then filters out incompatible conformations by steric exclusion, giving a fast estimate of the dynamic shield the glycans form. The authors validated the approach with molecular dynamics, small-angle scattering, cryo-EM, and mass spectrometry data and demonstrated its use across several glycoprotein systems. The abstract emphasizes that the analysis runs in a short time even on an ordinary personal computer, making it well suited to exploring how glycosylation affects surface accessibility. That speed comes from approximating the problem, so the resulting conformational ensembles should not be treated as a substitute for full protein–glycan coupled dynamics.
Highlight · Grafting from conformer libraries plus steric filtering can quickly map the dynamic shielding glycans impose on a protein surface.
DOI: 10.1016/j.cell.2024.01.034
29 · Transfer learning to leverage larger datasets for improved prediction of protein stability changes
Proceedings of the National Academy of Sciences · Publication/issue date: 2024-02-06
Authors: Henry Dieckhaus, Michael Brocidiacono, Nicholas Z. Randolph, Brian Kuhlman
Predicting how mutations alter protein stability requires striking a balance between data scale and model generalization. ThermoMPNN uses transfer learning to repurpose the structure-related representations learned by ProteinMPNN for stability-change prediction, training further on larger experimental datasets. The idea is to reuse an existing model's understanding of local protein environments and avoid relearning every feature from limited stability data. The abstract reports that the model improves on relevant benchmarks while keeping the prediction pipeline lightweight enough to scale. This supports a complementary role for large datasets and pretrained representations, but reliability for any specific mutation still depends on training coverage, structure quality, and independent experimental validation.
Highlight · Transferring structure-model representations to a stability task makes large-scale mutational data easier to exploit.
DOI: 10.1073/pnas.2314853121
Viral Genomes and Replication Mechanisms (2 papers)
30 · Generative design of bacteriophages with genome language models
Science · Publication/issue date: 2026-08-06
Authors: Samuel H. King, Claudia L. Driscoll, David B. Li, Daniel Guo, Aditi T. Merchant, Garyk Brixi, Max E. Wilkinson, Brian L. Hie
This study asks whether genome language models can produce designs that actually work biologically at the scale of a whole phage. The authors report 16 viable designer phages with varying fitness, and combine structural analysis with an examination of how genomic changes relate to viral particle features. Under laboratory conditions, some combinations also acted against phage-tolerant E. coli. The results show that generative model output can go beyond looking like a plausible sequence and reach a testable functional level — but they also show that biological performance across designs is uneven. These findings are limited to the phages and bacterial systems tested; they cannot be generalized into universal designability, clinical efficacy, or predictable behavior against arbitrary hosts.
Highlight · Phages offer an experimental system for testing whether genome language model output can translate into full biological function.
DOI: 10.1126/science.aec2657
31 · Structure and operating principles of a monkeypox virus replisome
Nature · Published/issue date: 2026-09-02
Authors: Zishuo Yu, Pradeep Sathyanarayana, Joel M. J. Tan, Side Hu, Xiaoyi Fan, Angela Gao, Philip J. Kranzusch, Joseph J. Loparo, Jonathan Abraham
This work uses cryo-EM to analyze how the monkeypox virus replication machinery is put together and how it operates, focusing on how the polymerase complex and the E5 helicase coordinate. The different conformations provide a structural basis for understanding how template handling, unwinding, and nucleic acid synthesis are coordinated, extending our picture of replication from isolated proteins to a more complete assembly. Its main contribution is revealing the spatial relationships between components and the possible principles behind their coordinated movement, giving later mechanistic studies a reference point. Structural observation alone does not mean an effective drug has been found, nor does it directly establish the full share and switching order of these conformations during infection. This entry covers the basic mechanism only and does not extend to building or engineering a replication system.
Highlight · The structure of the replication assembly reveals the spatial basis for how helicase and polymerase work in concert.
DOI: 10.1038/s41586-026-10937-2
Transcription initiation and chromatin regulation (2 papers)
32 · Structural basis of a SWI/SNF-associated transcription pre-initiation complex
Cell · Published/issue date: 2026-09-08
Authors: Shigeki Nagai, Chun Yang, Dong-Hua Chen, Michael J. Trnka, David A. Bushnell, Alma L. Burlingame, Kenji Murakami, Roger D. Kornberg
Supplemental abstract from the publisher. SWI/SNF is usually understood as clearing the way for transcription by moving or removing nucleosomes; this study goes further, finding that it can integrate directly into the RNA polymerase II transcription initiation machinery. Combining cryo-EM, crosslinking, and mass spectrometry, the authors resolved a complex containing the PIC, Mediator, SWI/SNF, and a nucleosome, in which the entire assembly can dimerize. The structure suggests that SWI/SNF both stabilizes the organization of the initiation complex and helps remodel the downstream +1 nucleosome. The model links two DNA translocases — Snf2 in SWI/SNF and Ssl2 in TFIIH — offering an explanation for how chromatin remodeling and transcription start are synchronized; the physiological state of the assembly still needs more in-situ evidence.
Highlight · SWI/SNF can be a component of the initiation complex, directly linking nucleosome remodeling to the transcription machinery.
DOI: 10.1016/j.cell.2026.08.020
33 · The molecular basis of transcription initiation by RNA polymerase II
Nature Reviews Molecular Cell Biology · Published/issue date: 2026-09-04
Authors: Xizi Chen, Yanhui Xu
This review lays out the molecular basis of RNA polymerase II transcription initiation, discussing general transcription factors, Mediator, and the chromatin environment within a single framework. It covers TFIID-mediated promoter recognition, assembly of the pre-initiation complex, the influence of the +1 nucleosome on start site position and direction, and DNA opening, promoter escape, and the transition into elongation. Phosphorylation of the polymerase C-terminal domain and the orderly exchange of regulatory factors are key threads tying these steps together. The piece is well suited to clarifying the logical relationships between different structural snapshots and to understanding how initiation is dynamically regulated. It synthesizes existing progress in the field, so one should not mistake the integrated model of these steps for a complete process directly observed in a single new experiment.
Highlight · Transcription initiation depends on coordinated changes and an orderly hand-off among chromatin, initiation factors, and polymerase state.
DOI: 10.1038/s41580-026-01019-2
cryo-ET methods and membrane analysis (2 papers)
34 · MemBrain v2: an end-to-end tool for analyzing membranes in cryo-electron tomography
Nature Methods · Published/issue date: 2026-09-08
Authors: Lorenz Lamm, Simon Zufferey, Hanyi Zhang, Ricardo D. Righetto, Florent Waltz, Wojciech Wietrzynski, Kevin A. Yamauchi, Alister Burt, Ye Liu, Antonio Martinez-Sanchez, Sebastian Ziegler, Fabian Isensee, Julia A. Schnabel, Benjamin D. Engel, Tingying Peng
Supplemental abstract based on the abstract registered by the publisher. Membrane analysis in cryo-ET is hampered by low signal-to-noise ratios, the missing wedge, and the complexity of membrane-associated particles. MemBrain v2 brings membrane segmentation, particle localization, and spatial statistics into one workflow: MemBrain-seg uses diverse training data to improve segmentation generalization across different tomographic conditions; MemBrain-pick combines geometric constraints with deep learning to reduce the heavy manual annotation burden when localizing membrane proteins; and MemBrain-stats computes spatial metrics for particle distribution and intra-membrane organization. Its contribution is connecting analysis steps that are usually scattered, making them easier to apply across many datasets. The abstract gives no single accuracy figure, so in practice users should still check segmentation and localization quality on their own samples.
Highlight · MemBrain v2 links membrane segmentation, membrane protein localization, and spatial statistics into a complete cryo-ET analysis pipeline.
DOI: 10.1038/s41592-026-03178-8
35 · Streamlined montage cryo-electron tomography for exploring the ultrastructure of cells and tissues
Structure · Published/issue date: 2026-09-03
Authors: Ryan Hylton, Micaela Boiero Sanders, Adriana Prajica, Gavin Rice, Stefan Raunser
Supplemental abstract from the publisher. Conventional cryo-ET can resolve molecular-scale structure but only over a small field of view; montage tomography can widen that field, yet it often requires complicated dose allocation or hardware modifications. The MOSAIC workflow stays as close as possible to standard automated acquisition, using MontageMaker to merge overlapping tilt series. According to the abstract, the resulting montage tomograms retain nanometer-scale detail despite uneven cumulative dose and work across a range of biological samples. Its main value lies in lowering the barrier to wider fields of view, making it easier for labs that already have automated cryo-ET acquisition set up to examine cell and tissue ultrastructure. The specific resolution, dose effects, and montage quality still need to be checked against the sample and data at hand.
Highlight · MOSAIC and MontageMaker extend the cryo-ET field of view with minimal changes to acquisition.
DOI: 10.1016/j.str.2026.07.001
Other papers (8)
36 · Cryo-EM structure of TGFBIp fibrils driven by a corneal dystrophy–linked mutation enables design of peptide inhibitors of aggregation
Proceedings of the National Academy of Sciences · Published/issue date: 2026-09-15
Authors: Yi Xiao Jiang, Lukasz Salwinski, Michael R. Sawaya, Peng Ge, Liisa Lutter, Xinyi Cheng, Carolyn J. Hu, Hillary Hernandez, David R. Boyer, Conrad Wang, Filipe A. Melo, Duilio Cascio, David S. Eisenberg
Corneal dystrophy–linked TGFBIp variants can promote amyloid fibril formation. Using cryo-EM, the authors studied the FAS1-4 segment carrying the V624M variant and resolved a fibril structure built from symmetric protofilaments and an internal channel. Its core region contains a segment found in patient deposits, linking the experimental structure to disease material. On that basis the team designed peptide inhibitors: G1 and H4 suppressed aggregation and fibril extension in vitro in a concentration-dependent manner, with support from fluorescence, fractionation, and electron microscopy. The work demonstrates a path from a pathology-relevant fibril structure to the design of aggregation inhibitors, but the evidence remains at the in vitro level and cannot be directly extrapolated into therapeutic benefit in corneal disease.
Highlight · The pathology-relevant TGFBIp fibril structure provides a direct template for designing peptides that block aggregation.
DOI: 10.1073/pnas.2607937123
37 · Determining the dark proteome
Nature Chemical Biology · Published/issue date: 2026-09-08
Authors: Maja Bencun, Etienne Boileau, Christoph Dieterich
Quick take: this piece centers on TransCODE's analysis of the non-canonical human proteome, looking at translation products beyond standard annotation. The available abstract offers only this research direction — no numbers on what was identified, what fraction was validated, or any specific functional conclusions — so this issue does not treat the "dark proteome" as fully mapped.
Highlight · Non-canonical translation products are expanding our sense of where the human proteome ends.
DOI: 10.1038/s41589-026-02320-z
38 · The graphical abstract
Nature Methods · Published/issue date: 2026-09-01
Authors: Helena Klara Jambor, Simon L. Dürr
Quick take: this entry discusses the role of graphical abstracts as visual summaries in science communication. The available teaser is very short, with no experimental comparisons, design guidelines, or data on communication outcomes, so this issue simply notes the topic and does not add production advice that the source does not support.
Highlight · Graphical abstracts present research highlights visually, but the current material offers no evaluation of their effect.
DOI: 10.1038/s41592-026-03207-6
39 · Cryo-EM structures of the SurA–BAM complex reveal conformational changes in outer membrane protein assembly
Nature Communications · Published/issue date: 2026-09-04
Authors: Ryoji Miyazaki, Hidetaka Kohga, Nami Matsuoka, Yuki Maruno, Wataru Yoshimoto, Yutaro S. Takahashi, Dede Heri Yuli Yanto, Yudhi Nugraha, Hideki Shigematsu, Takuya Shiota, Tomoya Tsukazaki
Outer membrane protein assembly in Gram-negative bacteria requires the chaperone SurA to work with the BAM complex, but how the two dynamically hand off substrates has remained unclear. The authors resolved multiple cryo-EM conformations of the E. coli SurA–BAM complex and observed that different SurA domains help regulate the relationship between the core region and the assembly machinery. The flexibility of the P1 and P2 regions, along with the interaction between P2 and BamE, provides a structural basis for binding and substrate hand-off. Comparing multiple states supports a model in which conformational changes drive substrate transfer forward. The work fills in a dynamic step in outer membrane biogenesis, though the pathway suggested by structural snapshots should still be read alongside functional experiments; a single conformation cannot pin down the full sequence of events.
Highlight · SurA's domain movements and its contacts with BAM together support the dynamic hand-off of outer membrane proteins.
DOI: 10.1038/s41467-026-76843-3
40 · Structural insights into how the metallochaperones UreE and UreG interact to deliver a toxic metal to urease
Proceedings of the National Academy of Sciences · Published/issue date: 2026/09/08
Authors: Chun-Long Chan, Longson Tsz Hin Pang, Tung Choi, Ka-Chun Chan, Ka Lung Tsang, Ka Ming Lee, Kam-Bo Wong
Supplemental abstract based on publisher registration information. Urease maturation requires nickel ions, but free nickel is cytotoxic, so controlled transfer between metallochaperones is essential. The authors resolved the structure of the complex formed by UreE and UreG and found that the two make contact asymmetrically; GTP binding drives changes that help the complex assemble and bring the nickel-binding regions of both partners into adjacent positions. Together with in vitro analysis, the structural comparison suggests that nickel transfer then weakens the interaction between the chaperones, allowing metal-loaded UreG to be released. From this the study proposes a hand-off model allosterically regulated by both nucleotide and metal, explaining how cells promote enzyme maturation while limiting exposure to a toxic metal; this is not a direct validation of therapeutic effects against infection.
Highlight · GTP and nickel binding jointly control the assembly and dissociation of UreE–UreG, tying metal hand-off to the chaperone cycle.
DOI: 10.1073/pnas.2535662123
41 · SLC4A7/NBCn1 sustains intracellular pH to support mTORC1 activity and antibody-secreting cell differentiation
Nature Communications · Published/issue date: 2026-09-11
Authors: Ke Hu, Yin-Hu Wang, Anthony Y. Tao, Maxwell McDermott, Miki Jishage, Li Zhong, Thomas Boehringer, Marcus J. Hines, Sergei B. Koralov, Ebbe Boedtkjer, Stefan Feske
Antibody-secreting cells have heavy metabolic demands during differentiation, and this study ties that process to intracellular acid–base balance. Loss of SLC4A7/NBCn1 caused intracellular acidification, changes in the lysosomal environment, and reduced mTORC1 activity, all of which impaired differentiation into antibody-secreting cells; forcing mTOR signaling back on in vitro partially restored the process. In mouse immunization and influenza infection experiments, antibody-secreting cells and antibody responses were also reduced. The effects of extracellular acidification lend further support to pH as an environmental factor in this regulation. The work shows that ion transport does more than maintain general homeostasis — it may also shape immune cell fate — though these findings do not mean that simply shifting external acid–base conditions would improve immune function in people.
Highlight · SLC4A7 sustains intracellular pH to support mTORC1 activity and antibody-secreting cell differentiation.
DOI: 10.1038/s41467-026-77588-9
42 · TM184C is a GPCR-like regulator of intercellular exchange and autophagy
Nature · Published/issue date: 2026-09-09
Authors: Kyutae D. Lee, Samuel Taylor, Jennifer Arcuri, Shraddha Chandthakuri, Jacinda Pujols, Bruno Colon, Qiang Wang, Chenzhou Wu, Zhipeng Meng, Sebastian J. Thompson-Ceccato, Jonathan Mitchell, Defne Bayik, Alexandra Carbone, Vladlen Slepak, Tatiana I. Slepak, Scott M. Welford, Michael E. Ivan, Dazhi Wang, Blake O. Goldberg, Warner Baringer, Elvis Maradzike, Daniel G. Isom
The authors used homology relationships among large-scale predicted structures to find GPCR-like proteins that sequence comparison alone tends to miss, then went on to study the cellular function of TM184C. Their results link the protein to intracellular vesicles, microtubule-associated distribution, and the exchange of material through intercellular bridges, and reveal regulatory features involving β-arrestin and GRK. Functional analysis also suggests it restrains LC3B lipidation and autophagic flux, while cross-species findings point to a relatively ancient origin for these functions. The study shows how structural similarity can help uncover under-annotated regulators of signaling and transport. Still, a GPCR-like structure does not automatically mean a classic ligand has been identified or that the signaling mechanism is identical.
Highlight · Structural homology analysis links TM184C to intercellular exchange and autophagy regulation.
DOI: 10.1038/s41586-026-10993-8
43 · Late-life semaglutide treatment slows ageing and extends lifespan in female mice
Nature · Published/issue date: 2026-09-01
Authors: Yufan Feng, Marine Barthez, Yifei Wang, Yibing Chen, Huixian Qiu, Chih-Ling Wang, Kartoosh Heydari, Melaine Delcroix, Lene Juel Rasmussen, Vilhelm A. Bohr, Danica Chen
This study looked at how semaglutide affects ageing-related phenotypes and lifespan in aged female mice. Treatment starting at around twenty months of age improved several physiological measures and ageing-related markers, and continued treatment was accompanied by longer lifespan. The authors included a matched caloric restriction comparison and found that the two shared some effects, but not all — memory, exploratory behavior, and glucose regulation differed — suggesting the drug's effects may go beyond simply eating less. The results provide animal evidence that starting a metabolic intervention late in life may still help, and make a case for further work to separate reduced food intake from other mechanisms. Because the study is limited to one sex, strain, and set of conditions, it cannot be used to conclude that the drug extends lifespan in humans.
Highlight · Improvements in aged female mice point to the potential of late-life metabolic intervention; some effects cannot be explained by matched caloric restriction alone.
DOI: 10.1038/s41586-026-10940-7