Tag: Academic research

Posted on

ASGCT 2026: Timothy Yu Wins Jerry Mendell Award For N-of-1 Therapies

BOSTON – Timothy Yu, MD, PhD, a leading neurogeneticist at Boston Children’s Hospital and Harvard Medical School, was awarded one of the top honors by the American Society of Gene and Cell Therapy (ASGCT), the Jerry Mendell Award.

Yu was recognized for his trailblazing work over the past decade in devising bespoke oligonucleotide therapies for patients with ultra-rare genetic disorders. He entitled his talk: “The paradox of N-of-1: Scaling the logic of genetic intervention.” A suitable sub-title, Yu said, could be: “A neurogeneticist’s accidental injection into the field of gene therapy.”

Timothy Yu
Timothy Yu, MD, PhD

Yu’s research has progressed from gene discovery to clinical applications of N-of-1 therapies. The long tail challenge of developing gene therapies for patients with rare diseases is actually immense. There are some 400 million individuals worldwide who suffer one of some 8,000 monogenic disorders. Three out of ten affected children do not see their fifth birthday, while most lack any kind of medical treatment.

But progress over the past 10-20 years has provided hope in the form of various molecular therapies—antisense, mRNA, gene therapy, siRNAs, CRISPR, and newer flavors of gene editing. The concept of “therapeutic programmability is very important,” Yu said.

Yu’s injection into the field began with a single patient, Mila Makovec, a young girl from Colorado, whom he met in 2018. This index patient was eventually diagnosed with a form of Batten disease, CLN7—a rare subtype that was progressive and fatal.

Mila’s gene mutation was private but correctable, an insertion sitting deep within an intron. This offered Yu’s team hope that they could block abnormal splicing. Clearly, no company could progress a therapy for a single patient, Yu recalled.

Allele-specific oligonucleotides are simple to manufacture and can boost gene expression, following the model of Spinraza for spinal muscular atrophy.

Yu’s team developed a customized ASO therapy in about a year, which was published in 2019. The therapy brought about a reduction in Mila’s seizures, but not in time to result in a cure. In May 2019, Yu’s team met with the FDA to establish a path forward. The first guidances were published in 2021.

Moving on

Yu recounted several other therapies designed for other patients with different genetic disorders. The second program, working with Jennifer Puck, MD, and colleagues, was for ataxia telangiectasia (A-T). One A-T mutation created a new splice site that appeared to be reversible with a custom ASO.

The pilot clinical study was initiated in 2018, when the child was two years old. It is, Yu said, the longest running N-of-1 trial. The child is now nine years old and shows no worsening of clinical symptoms. Various measurements and assays confirm there has been no clinical progression. “We have converted a classic case of A-T to a milder form,” Yu said. The trial is expanding in Europe, including ten additional children in Turkey.

A member of Yu’s lab, Claudia Lentucci, PhD, is among the team leading a third program treating infants with neonatal epilepsy (KCNT1-related epileptic encephalopathy). One patient had seizures halted but developed ventricular enlargement. The team has since modified the protocol to use intracerebroventricular injection, which reduces seizures by 60-70%.

A fourth example presented by Yu was a treatment for Grace, a 15-year-old girl with a rare form of retinitis pigmentosa. She presented seven years ago with vision loss and pain insensitivity. Launched in August 2023, the therapy corrects a deep intronic mutation. It has been well tolerated and resulted in a stabilization of her vision.

Yu summarized similar therapies developed for patients with Zellweger syndrome, Niemann Pick Type C, and Batten disease. In total, 35 N-of-1 oligonucleotides have been administered to more than 80 patients.

On hearing the news of Baby KJ last year, “we all stood up and took notice,” Yu said.

“Our motivations are to help patients without other options. We have expanded from a sick child in Colorado to generate pilot learnings for childhood neurologic diseases.” His team’s work is not only offering hope in a compassionate sense but is also leading the exploration of new delivery models for precision medicine.

Yu has built a large network of collaborators, including clinicians, regulators, and industry professionals. The N=1 Collaborative has grown to more than 2,000 members worldwide and will be holding its third conference this October in Denver.

A new approach

From what was once called “interventional genetics,” Yu said, “the data is inviting us to take a new approach” that he called “genetic surgery.”

N-of-1 therapies are “more akin to a complex surgery, [using] customized tools and procedures for therapeutic benefit akin to organ transplant or cardiac surgery,” he said.

Yu also highlighted the FDA’s Plausible Mechanism Pathway, announced last February, which offers opportunities to approve medicines on the basis of very small numbers of patients. But the guidance emphasized the importance of data sharing—an issue that required the community’s full attention.

“You can’t build a modular system in a silo,” Yu said. “If you want cures that are greater than the sum of their parts, you must share the data to see how the pieces fit together.”

The post ASGCT 2026: Timothy Yu Wins Jerry Mendell Award For N-of-1 Therapies appeared first on Inside Precision Medicine.

Posted on

Visualizing Receptor Transport Within Neurons via Transcytosis

As spindly, elongated cells, neurons must be able to transport proteins and receptors between distant sites in their cell bodies and axons to function properly. A new imaging study by researchers at Johns Hopkins University has now visualized the ebb and flow of the nerve growth factor receptor TrkA within neurons, via an unusual process known as transcytosis. Their study also explains how this phenomenon supports neuronal function and connectivity in mice.

Senior and corresponding author Rejji Kuruvilla, PhD, at Johns Hopkins University Department of Biology, and colleagues reported on their findings in Science Signaling, in a paper titled “Transcytosis-mediated anterograde transport of the receptor TrkA mediates the formation of presynaptic sites in sympathetic neurons.” In their paper, the authors concluded, “These findings provide mechanistic insight into an atypical mode of receptor trafficking and demonstrate its physiological relevance in sympathetic neuron connectivity in mice … Our study suggests that transcytosis might be a more general mechanism than now appreciated for the targeted transport of trophic and guidance receptors, adhesion and synaptic proteins, as well as ion channels.”

The axons of neurons are extremely long compared to their main cell bodies, with axon terminals sometimes residing a long distance from the cell nucleus. “Axon terminals can be meters away from cell bodies where many axonal membrane proteins with critical functions in regulating axon guidance and growth, neuronal survival, presynaptic organization, and synaptic transmission are made,” the authors wrote.

Neurons need to be able to transport these proteins efficiently across these relatively vast distances. They do this by either directly sending the protein through a secretory pathway or via an indirect mechanism called transcytosis. The latter occurs when the central cell body takes in newly synthesized proteins or surface receptors, after which they move to axons through the cell cytoplasm. “Transcytosis is an atypical endocytosis-based mechanism, where newly synthesized proteins are first inserted on cell body surfaces, internalized, and anterogradely transported to axons,” the team continued.

Transcytosis is still relatively obscure and enigmatic compared with the direct secretion method, and questions remain about how exactly it sustains the function and connectivity of neurons. “In contrast to the considerable progress made in understanding the direct secretory pathway, there is limited knowledge about transcytosis, specifically the underlying transport kinetics and organelles involved, whether it occurs in vivo, and its contributions to neuronal connectivity and function,” the investigators noted.

Seeking answers, first author Kuruvilla, together with first author Guillermo Moya-Alvarado, PhD, and colleagues, used live cell imaging and electron microscopy to peer at the movement of receptors across compartments within mouse neurons.

They visualized the trafficking dynamics and transcytosis of a receptor named TrkA. “The family of tropomyosin-related kinase (Trk) receptors provides a prominent example of membrane proteins that undergo long-distance axonal trafficking to control neuronal survival, axon growth, and synaptic transmission,” the scientists explained.

Through their study, the authors noted various shifts in speed and direction as vesicles carried TrkA from the soma to axons. Using labeled TrkA proteins, the scientists also confirmed that transcytosis occurred within nerve terminals of living mice. “Live imaging and electron microscopy of compartmentalized cultures revealed that soma surface–derived TrkA proteins underwent dynamic transport within axons, with changes in speed, direction, and the vesicular organelles that carried them as they moved from proximal to distal axon compartments,” they stated. “In mice, soma surface–labeled TrkA proteins were observed in sympathetic nerve terminals, demonstrating that transcytosis occurs in vivo.”

Assessing TrkA receptors transcytosis from cell bodies to nerve terminals in vivo. Superior cervical ganglion (SCG) in Ntrk1Flag mice, at postnatal day 2 to day 3 were injected in one of each paired ganglia per animal with the contralateral ganglion and target tissues (noninjected side) serving as internal controls to assess any systemic leakage of injected label. Representative image of the injected side. Flag (green) and sympathetic neurons (Tuj1, red) immunofluorescence in the superior cervical ganglia. DAPI is shown in blue. Scale bars, 50 μm. [All images and movies were generated by Guillermo Moya Alvarado]
Assessing TrkA receptors transcytosis from cell bodies to nerve terminals in vivo. Superior cervical ganglion (SCG) in Ntrk1Flag mice, at postnatal day 2 to day 3, were injected in one of each paired ganglia per animal with the contralateral ganglion and target tissues (noninjected side) serving as internal controls to assess any systemic leakage of injected label. Representative image of the injected side. Flag (green) and sympathetic neurons (Tuj1, red) immunofluorescence in the superior cervical ganglia. DAPI is shown in blue. Scale bars, 50 μm. [All images and movies were generated by Guillermo Moya Alvarado]

They also found that disrupting its transcytosis by introducing a point mutation into TrkA reduced the number and size of presynaptic sites and decreased synaptic transmission in culture and in rodents in vivo, confirming the importance of the process for neuronal physiology.  “These findings provide mechanistic insight into an atypical mode of receptor trafficking and demonstrate its physiological relevance in sympathetic neuron connectivity in mice,” the team concluded “Uncovering mechanisms of axon delivery has implications that extend beyond the healthy nervous system to understanding cell biological pathways that contribute to nerve repair after injury or neurodegeneration, because the correct complement of membrane proteins must be accurately targeted to regenerating axons to ensure functional recovery.”

The post Visualizing Receptor Transport Within Neurons via Transcytosis appeared first on GEN – Genetic Engineering and Biotechnology News.

Posted on

A Digital Assistive System for Maintaining Nutrition and Mobility in Older Adults: Usability and Feasibility Findings From a Pilot Study

Background: Due to demographic changes, the number of older people is increasing, often accompanied by limitations in mobility, nutrition, and independence. Preventive monitoring is rare, as care systems struggle with staff shortages and limited resources. Technical assistance systems can support older people in self-assessing their health and maintaining independence. We developed the AS-Tra system, which combines an application with a measurement and training station (MuTS), to enable early detection of nutrition and mobility-related deficits and risks. Objective: This paper presents the pilot study of the AS-Tra system with the aim of evaluating its usability and testing the feasibility of collecting health-related data from older adults (≥70 y) with early/mild deficiencies in nutritional state and mobility in preparation for a future randomized controlled trial. Methods: The system used in this 4-week pilot study was developed as a complex intervention in accordance with the Medical Research Council framework. Participants (target n=10) were recruited through a participant registry. They completed standardized mobility assessments (grip strength, Timed “Up and Go,” and 5-Time Chair Rise) at baseline and after 1, 2, and 4 weeks (T0, T1, and T2, respectively). Mini Nutritional Assessment—Short Form and short physical performance battery were recorded at baseline and at T2. Participants received a tablet app for regularly documenting nutrition and an activity sensor for 7 days of physical activity monitoring and performed weekly training starting at T0. At T2, the System Usability Scale (SUS) and feedback questionnaires (Evaluation Overall System [EOS] questionnaire—the evaluation of all subcomponents on a scale of 1-5, weekly Experience Report) were additionally collected. Data were analyzed descriptively using IBM SPSS Statistics, in which data were shown as total numbers, percentages, and means with SDs, and data from the activity sensor were displayed and analyzed using Python. Results: A total of 9 older adults, with 1 dropout (mean 80, SD 5 y, 50% female), participated in this study. The SUS score was good (mean 79, SD 13.4 points). The MuTS devices had minor technical problems (in <17% of MuTS sessions), while 57% (17/30) of the users experienced instability issues with the food diary in the tablet app. The average overall system ratings were positive, with an EOS score of 2.01 (SD 0.99). Conclusions: The usability of the technical assistance system used in this study was rated as good. The data collection using questionnaires, sensors, and automated assessments proved feasible. The biggest challenge was the tablet-based food diary, which still needs improvement before the effectiveness of the AS-Tra system regarding mobility and nutritional status can be evaluated in a randomized controlled trial.
Posted on

Targeted Ultrasound Could Offer Alternative to Chronic Pain Medication

A new study has shown that targeting ultrasound stimulation to brain regions involved in processing pain can induce long-lasting changes in brain activity, significantly reducing pain perception. Published in Nature Communications, these findings point at a novel non-invasive strategy to treat chronic pain. 

“Our study represents an important first step in understanding how this technology can non-invasively stimulate deep brain regions involved in pain processing,” said Sam Hughes, PhD, senior lecturer in pain neuroscience at the University of Exeter. “We found that targeting a specific brain region involved in pain processing can alter how pain is perceived and change how this area communicates with other parts of the brain’s pain network. The next stage of our research will be to test whether this approach can help people living with chronic pain.”

Hughes and colleagues used transcranial ultrasound stimulation (TUS), a low-intensity neuromodulation technique, to target the dorsal anterior cingulate cortex (dACC), a brain region implicated in chronic pain. The study recruited a total of 32 healthy volunteers, who were treated either with TUS or a sham while putting their right hand in a cold gel to trigger pain due to the low temperature. All participants were asked to rate the severity of the pain they were feeling and underwent MRI and MRS scans to monitor the physiological changes caused by the treatment. 

Results showed that, while TUS had no immediate effect on pain intensity, participants reported a significant reduction in pain from 28 to 55 minutes after the stimulation, suggesting it can trigger a delayed analgesic effect. At the physiological level, TUS was found to disrupt the relationship between temperature and pain intensity, increasing the connectivity between the dACC and other brain regions involved in pain modulation and changing the concentration of the GABA neurotransmitter within the dCC. 

“The study aimed to characterize how transcranial ultrasound stimulation interacts with—and potentially also alters—the brain’s processing of pain,” said Sophie Clarke, PhD, postdoctoral research fellow at the University of Plymouth and lead author of the study. “Understanding these mechanisms will be very important to support the next steps in understanding whether the stimulation can be effective in helping patients with chronic pain.”

Previous research at the University of Plymouth had shown the potential benefits of TUS for psychiatric conditions including anxiety, depression, and addiction. This study shows these benefits could extend beyond neurological disorders and one day offer a non-invasive treatment option for those experiencing chronic pain due to conditions such as fibromyalgia, back pain, and arthritis, or recovering after cancer treatment.  

“Having shown the use of ultrasound can yield positive results for people with a variety of neurological conditions, we wanted to explore what it could mean for those living with chronic pain,” said Elsa Fouragnan, PhD, director of the University of Plymouth’s Brain Research and Imaging Centre (BRIC) and Centre for Therapeutic Ultrasound (CENTUS). “Most of us know someone experiencing chronic pain, and there are very few treatments that deliver any form of long-term benefit. The findings of this new work are really promising, and we are already building on it to assess whether TUS could be a beneficial and non-invasive therapeutic treatment.”

The post Targeted Ultrasound Could Offer Alternative to Chronic Pain Medication appeared first on Inside Precision Medicine.

Posted on

CSF Platform Enables Near Real-Time Monitoring of Multiple Biomarkers

Scientists have developed a sensor platform that can monitor cerebrospinal fluid (CSF) in intensive care unit patients, overcoming major delays in diagnosis associated with current testing methods. A study published today in Science Translational Medicine reports that the NeuroSense platform can provide near real-time readings of four key biomarkers every 27 minutes, with results accurately reflecting standard clinical measurements. 

In neurological intensive care units, external ventricular drainage (EVD) systems are routinely used to temporarily assist patients with drainage of excess CSF, manage postoperative complications and monitor intracranial pressure. However, the use of these devices carries a high infection risk, with rates reaching up to 20% of patients. 

Delayed diagnosis of these infections can lead to severe meningitis, neural damage, cognitive impairment, permanent disability, or even death. However, current testing methods are labor-intensive and require sending samples to external laboratories for biomarker analysis and manual inspection. This limits testing to every one to two days, significantly delaying clinical decisions that can be critical for preventing severe complications. 

“To address these limitations, we developed NeuroSense, a multiplexed sensing platform that integrates with standard external ventricular drainage systems to enable near real-time monitoring of key CSF biomarkers, including glucose, lactate, pH, and flow rate, that are essential for detecting infection and drain dysfunction,” write the study authors. 

The NeuroSense platform employs aptamer-based biosensors to detect glucose and lactate levels in CSF, which are key markers of bacterial infections. These types of biosensors are more stable and have a longer shelf life than conventional enzymatic biosensors, ensuring the platform can consistently and accurately track these markers for the entire time EVD systems remain in place, typically between five to 10 days. 

Furthermore, an impedance-based sensor measures CSF flow rate to monitor for potential catheter obstructions or incorrect EVD settings, while a polydopamine sensor keeps track of pH changes, which can indicate acidosis, hemorrhage, infection, or a disrupted blood-brain barrier.

The platform’s performance was evaluated in a small-scale study that recruited six patients with EVDs hospitalized in the intensive care unit. Every four hours, readings from the NeuroSense platform were compared with those from standard testing methods, revealing a strong correlation between the sensor platform and clinical reference measurements.

A survey of the healthcare providers and clinicians involved in the study further showed that most participants found the platform easy to use, as it integrates with standard EVD systems routinely used in the intensive care setting.

Going forward, the researchers plan on further improving the performance of the pH sensor and continue developing the platform to comply with regulatory requirements for running large-scale clinical studies and eventually making the platform available to healthcare providers. 

“Beyond infection detection and EVD assessment, NeuroSense enables higher temporal-resolution tracking of CSF biomarkers and flow dynamics, supporting earlier recognition of evolving trends that may be missed with intermittent sampling,” write the researchers. “Although the current system measures glucose, lactate, pH, and flow, the platform is modular and can accommodate additional sensors in future iterations. By providing near-bedside, actionable insights into patients’ neurological health, NeuroSense has strong potential to enhance clinical decision-making and improve patient care.”

The post CSF Platform Enables Near Real-Time Monitoring of Multiple Biomarkers appeared first on Inside Precision Medicine.

Posted on

Synthesizing a Natural Sunscreen and Antioxidant in E. coli

Researchers at Jiangnan University have engineered microbial “cell factories” to sustainably produce the UV-protective compound gadusol, a compound that could eventually serve as a sunscreen ingredient and an antioxidant additive.

Found in the eggs of various fish and other marine organisms, gadusol helps protect against ultraviolet damage. However, “it is scarce in nature, and extracting it is inefficient and can carry environmental costs,” said research lead Ping Zhang, PhD. “We want to find a scalable and greener way to produce gadusol.” Previous efforts to acquire gadusol have relied mainly on extraction from fish eggs or tissues at specific developmental timepoints. For their reported study, rather than harvest the molecule from nature, the team turned microbes into mini chemical factories, increasing yield more than 90 times.

The researchers reported on their developments in Trends in Biotechnology, in a paper titled “Multidimensionally engineered Escherichia coli for efficient gadusol biosynthesis with high-throughput quantitative analysis, in which concluded “This work provides valuable insights into gadusol biosynthesis and establishes a solid basis for future investigations into the feasibility of its industrial applications.”

How do fish survive sunlight in the open waters without getting burned? They make their own natural sunscreen. Gadusol is “a naturally occurring cyclohexenone compound that has attracted significant attention due to its potent UV-absorbing properties and broad biological functionalities,” the authors wrote. “Gadusol, as a cyclohexenone compound with potent antioxidant and biological properties, has shown significant potential for applications in pharmaceuticals, cosmetics, and nutraceuticals.”

The compound has been identified in the eggs of various fish species, including zebrafish, salmon, and sturgeon, as well as in coral reef ecosystems, where it shields marine embryos and coral from UV radiation. Moreover, the researchers explained, “Beyond its photoprotective role, gadusol also demonstrates pronounced antioxidant activity, with extracts from fish eggs exhibiting free radical scavenging capacity comparable to that of ascorbic acid under physiological pH and even superior chain-breaking activity against peroxyl radicals.”

However, the team pointed out, extracting gadusol from fish eggs is labor intensive and costly, and results in low yields.  For their newly reported study the team rebuilt a zebrafish’s pathway for making gadusol inside the bacterium Escherichia coli, then tweaked the microbes’ genetics and growing conditions. “In this study, we constructed the gadusol biosynthetic pathway in E. coli and established an integrated engineering strategy to enable its efficient production,” they noted.

These modifications increased gadusol yield by nearly 93 times, from 45.2 milligrams per liter to 4.2 grams per liter. The compound’s antioxidant property inspired a useful shortcut, in the form of a color-based screening test. In the test, a purple chemical signal turns yellow when gadusol neutralizes free radicals. This color shift allows researchers to quickly identify bacterial strains that produce more of the compound. “Compared with traditional chemical analysis, this approach is more convenient, efficient, and economical,” said senior author Ruirui Xu,” PhD.  “In addition, an efficient downstream process for gadusol extraction and purification was established, strengthening the translational potential of this platform,” the authors noted.

Producing gadusol through a microbial cell factory for sun protection. [Science Center for Future Foods, Jiangnan University]
Producing gadusol through a microbial cell factory for sun protection. [Science Center for Future Foods, Jiangnan University]

The lab-made compound showed promise in preliminary UV-protection tests. “Achieving this level of production in the lab is very promising,” says Zhang. “It suggests that we may be able to meet future demand for natural sunscreen ingredients through microbial production.” The compound may offer more than just sun protection. In experiments, gadusol showed antioxidant activity comparable to that of vitamin C, suggesting it may help neutralize cell-damaging free radicals from UV exposure.

In their paper the authors suggested that “Beyond laboratory-scale optimization, this work demonstrates a practical and scalable route for manufacturing gadusol, a natural, nontoxic UV-protective, and antioxidant cyclohexenone.” The research comes amid growing interest in alternatives to some conventional sunscreen ingredients, which can irritate sensitive skin, harm marine organisms, or rely on petrochemicals. Gadusol’s combination of UV protection and antioxidant activity could make it an attractive ingredient for sunscreens and skin-care products. The team in addition pointed out, “By enabling sustainable microbial production independent of marine extraction, this strategy provides a renewable alternative to synthetic UV filters and antioxidants that may raise environmental or health concerns.”

But gadusol won’t make it into our beach bags just yet. The study did not compare gadusol head-to-head with commercial sunscreens, nor assess long-term safety or large-scale manufacturing. Regulatory approval would also be needed. Still, Xu believes that the research provides a starting point for moving gadusol toward practical applications. Based on current technology, he expects to start seeing some products appear on the market within two years.

“For small molecules with application potential, we hope people look beyond traditional extraction methods,” Zhang commented. “Microbial cell factories are emerging as a greener and more sustainable way to bring laboratory discoveries into real-world use.”

The post Synthesizing a Natural Sunscreen and Antioxidant in <i>E. coli</i> appeared first on GEN – Genetic Engineering and Biotechnology News.

Posted on

TNBC Ecotypes Reveal Molecular Signatures Tied to Chemotherapy Response

Researchers at The University of Texas MD Anderson Cancer Center have identified immune cell and tumor-specific features in triple-negative breast cancer (TNBC) that may help predict which patients are most likely to respond to chemotherapy before treatment begins, according to a study published in Nature. Using single-cell and spatial transcriptomic analyses of pretreatment tumor samples, the team identified specific macrophage subtypes and cancer-cell gene expression programs associated with response to neoadjuvant chemotherapy (NAC). The team also developed a 13-gene panel and a machine learning model that could help classify tumors according to their likelihood of responding to chemotherapy.

“This study provides novel insights into the gene-expression programs and the different cell states of the tumor microenvironment in patients with triple-negative breast cancer,” said Nicholas Navin, PhD, chair of systems biology at MD Anderson. “Importantly, we’ve identified certain programs and macrophage subtypes that are associated with good responses to neoadjuvant chemotherapy, which has tremendous potential to improve patient outcomes.”

TNBC accounts for between 10% and 20% of breast cancer cases. Because it lacks estrogen, progesterone, and HER2 receptors, treatment options are limited, resulting in a higher rate of recurrence compared with other form of breast cancer. Chemotherapy is the main treatment approach, particularly in early-stage disease, where neoadjuvant chemotherapy can achieve pathological complete response in 40% to 50% of patients. However, treatment outcomes vary widely from patient to patient, and researchers have been looking for ways that can better predict response before therapy begins.

For this study, the researchers analyzed pretreatment core biopsy samples from treatment-naive patients with early-stage TNBC. They performed single-cell RNA sequencing on 427,857 cells collected from 101 patients and spatial transcriptomic profiling on tumors from 44 patients. The findings also were compared with normal breast tissue data from the Human Breast Cell Atlas.

Based on their testing the researchers classified TNBC tumors into four patient-level “archetypes” based on cancer-cell gene expression patterns. They also identified 13 metaprograms that reflected heterogeneity within tumors at the single-cell level.

The tumor microenvironment consisted of 49 immune and stromal cell states organized into eight cellular communities, or ecotypes, defined by the co-occurrence of cancer cells and surrounding immune cell populations. Researchers found these cellular neighborhoods were associated both with tumor archetypes and chemotherapy response.

The study homed in on macrophages, a type of immune cell that has received less attention in TNBC research than T cells. The investigators said that seven of eight macrophage cell states were significantly associated with treatment response, while none of the 14 T-cell and natural killer-cell states showed significant associations with NAC response.

Macrophage subtypes linked to interferon signaling and complement activity, identified as Mac-IFN and Mac-lip-C1Q, were more abundant in patients who achieved pathological complete response. By comparison, two macrophages associated with angiogenesis and extracellular matrix remodeling, called Mac-angio and Mac-ECM, were enriched in patients with residual disease after chemotherapy.

The team also found that tumors linked to good response to NAC showed increased interferon signaling and elevated expression of human leukocyte antigen class II genes. Researchers said these findings indicate that cancer cells themselves may actively participate in modulating immune signaling related to chemotherapy response.

As part of their work, the researchers developed a 13-gene transcriptional signature panel developed from the single-cell analyses that can be used as a predictive model for chemotherapy response. Researchers said the model’s predictions correlated with chemotherapy response and overall survival across multiple public TNBC cohorts.

These new findings have the potential to influence how patients with TNBC are treated in the future by helping clinicians identify which patients are more likely to benefit from standard chemotherapy and which patients may need alternative therapeutic strategies earlier.

In addition, “these findings suggest that targeting specific macrophage subtypes could potentially provide new therapeutic opportunities in TNBC,” the researchers wrote.

The MD Anderson team noted that the study is one of the first large-scale single-cell genomic studies of TNBC integrating cancer cells, immune cells and treatment-response data. Earlier research exploring tumor heterogeneity has often lacked therapy response information, focused only on cancer cells or immune cells separately, or included relatively small patient cohorts.

Whether single-cell RNA seq could eventually become a basis for predictive diagnostics remains an open question. Today, the method is still expensive and technically challenging, two hindrances to it wider adoption. The researchers noted, however, that advances in sample multiplexing and other methods compatible with formalin-fixed paraffin-embedded tissue could make it feasible in the future.

Clinton Yam, MD, an associate professor of breast medical oncology at MD Anderson, said the findings could support more individualized approaches to TNBC care.

“These insights provide an important foundation for improving our understanding of why different TNBC tumors respond differently to chemotherapy, and the findings have strong potential to inform future strategies aimed at better predicting treatment response and guiding more individualized care for patients with triple-negative breast cancer.”

Future research will focus on validating the predictive models in prospective patient cohorts and evaluating TNBC treated with chemo-immunotherapy, which has become the standard of care when TNBC is detected early. The researchers also plan to study longitudinal tumor samples collected before, during, and after treatment to better understand how cancer cells and the tumor microenvironment evolve over time and how those changes relate to chemotherapy response and survival.

The post TNBC Ecotypes Reveal Molecular Signatures Tied to Chemotherapy Response appeared first on Inside Precision Medicine.

Posted on

Automatic Speech Recognition and Large Language Models for Multilingual Pathology Report Generation: Proof-of-Concept Study

Background: Accurate transcription of pathology gross examination dictation is important for clinical documentation, but multilingual dictation remains challenging in settings where clinicians mix Chinese and English while final pathology reports are written in English. Objective: This study aimed to evaluate whether a Whisper-based automatic speech recognition (ASR) pipeline guided by contextual system messages and combined with open-source large language models (LLMs; Qwen2:72b, Llama3.1:70b, Gemma2:27b) could improve multilingual (Chinese-English) pathology dictation transcription accuracy and generate clinically appropriate English gross description reports. Methods: We conducted a controlled proof-of-concept study using 125 simulated mixed Chinese-English pathology gross examination audio recordings created by physicians or pathologists. Audio recordings were transcribed using Whisper ASR with and without a contextual system message. The ASR transcripts were then converted into English gross description reports using 3 open-source LLMs: Qwen2:72b, Llama3.1:70b, and Gemma2:27b. Outcomes included character error rate, Bilingual Evaluation Understudy, Recall-Oriented Understudy for Gisting Evaluation (ROUGE)-1, ROUGE-2, ROUGE-L, Metric for Evaluation of Translation with Explicit Ordering, pathologist Win-Tie-Lose rankings, report-level error categories, inference time, and interrater agreement. Results: The ASR contextual system message reduced the mean character error rate from 0.344 (SD 0.176; 95% CI 0.313‐0.375) to 0.066 (SD 0.100; 95% CI 0.048‐0.084; <.001). Qwen2:72b achieved the highest automated metric scores, including a Bilingual Evaluation Understudy of 0.644 (SD 0.307), ROUGE-1 of 0.866 (SD 0.163), ROUGE-2 of 0.771 (SD 0.235), ROUGE-L of 0.842 (SD 0.178), and Metric for Evaluation of Translation with Explicit Ordering of 0.805 (SD 0.214). Pathologist-coded total error rates were 16.8% (21/125) for Qwen2:72b, 45.6% (57/125) for Llama3.1:70b, and 92.8% (116/125) for Gemma2:27b. The exact agreement between the 2 pathologists across full ranking categories was 76.8% (96/125; Cohen κ=0.668), and agreement on the top-ranked model or tied top group was 81.6% (102/125; Cohen κ=0.722). Conclusions: In this proof-of-concept evaluation, contextual prompting improved ASR transcription accuracy, and Qwen2:72b generated the most accurate English pathology reports among the evaluated LLMs. However, the study used simulated audio recordings, a local vocabulary prompt, and report-level rather than term-level clinical annotation. LLM-generated reports should therefore be considered draft documentation requiring pathologist verification, and prospective validation in real clinical workflows is needed before clinical deployment.
Posted on

ApexGO: AI-Driven Approach to Faster Antibiotic Discovery

Antibiotic resistance is on the rise around the world, creating an urgent need for faster and more dependable approaches to design antimicrobial candidates. While AI-driven methods have accelerated antimicrobial discovery, most have focused on screening fixed libraries or generating broad candidate sets.

Now, researchers at the University of Pennsylvania have developed ApexGO—a novel, AI-powered method that starts with a small number of candidates and improves them, using a predictive algorithm to evaluate each modification and guide the next.

“Antibiotic discovery is fundamentally a search problem across an enormous molecular space. ApexGO gives us a way to navigate that space with far more direction,” says César de la Fuente, PhD, presidential associate professor in the School of Engineering and Applied Science at UPenn.

This work is published Nature Machine Intelligence in the paper, “A generative artificial intelligence approach for peptide antibiotic optimization.”

“What is striking is that ApexGO’s predictions held up in the real world,” says Jacob R. Gardner, PhD, assistant professor in computer and information science (CIS) at UPenn. “ApexGO was optimizing against another computer model, so one concern was that it might find molecules that looked good to the model but failed in the lab. Instead, the majority of the molecules it designed actually worked.”

indeed, 85% of the AI-generated molecules halted bacterial growth, while 72% outperformed the peptides from which they were derived. In mice, two antimicrobial peptides created by ApexGO reduced bacterial counts at levels comparable to the antibiotic polymyxin B.

“This result points toward a future in which we can optimize molecules for a desired function in a fraction of the time,” adds de la Fuente, “using machines to guide discovery through chemical spaces too vast for humans to explore by trial and error.”

For years, the de la Fuente lab has looked for antibiotic candidates in unlikely places, from frog secretions to ancient microbes. Two years ago, the group released APEX, an AI model that predicts whether or not a given peptide is likely to have antimicrobial properties.

“APEX helped us find promising antibiotic candidates in enormous biological datasets,” says Marcelo Torres, PhD, research assistant professor of psychiatry in the Perelman School of Medicine. “ApexGO takes the next step: once we have a promising molecule, it helps us ask how to make it better.”

One part of ApexGO (short for APEX Generative Optimization) suggests molecular tweaks, while the previously published APEX model predicts whether those changes are likely to increase antimicrobial activity. ApexGO then uses those predictions to guide the next round of proposed edits.

While some of the molecules proposed by ApexGO showed promising antibiotic activity, the researchers emphasize that even the best-performing peptides are still early-stage candidates. Before any could be used to treat infections in humans, they would need to be further optimized for safety, stability, and how long they remain active in the body.

Still, the study suggests that AI can help researchers decide which molecules are worth making and testing in the first place. For de la Fuente, the approach could eventually extend beyond antibiotics. “In this case, we wanted to optimize peptides for antimicrobial activity,” he says. “But you could imagine applying the same idea to peptides with other biological functions, like modulating the immune system or targeting tumors.”

“ApexGO shows that AI can do more than predict which molecules might work: it can help us improve them,” adds de la Fuente. “At a time when antibiotic resistance is rising worldwide, we need technologies that help us move faster from an idea to a real therapeutic candidate. ApexGO is an important step toward that future.”

The post ApexGO: AI-Driven Approach to Faster Antibiotic Discovery appeared first on GEN – Genetic Engineering and Biotechnology News.