Mois : décembre 2024

Scientists from the Longevity Research Institute (LRI), which was formed by the merger of SENS Research Foundation and Lifespan.io, have achieved expression of an essential mitochondrial gene in the nucleus and proper functioning of the protein. This could pave the way for curing diseases caused by mitochondrial mutations [1].

The fragile mitochondrial DNA

The prevailing scientific consensus is that mitochondria were once independent microorganisms that entered a symbiotic relationship with larger cells. This duo gave rise to eukaryotic cells: the building blocks of all multicellular life. Without that fateful “marriage,” complex life would not exist, as mitochondria provide cells with essential energy via oxidative phosphorylation.

Over the millennia, mitochondria have retained their own DNA. However, this mitochondrial DNA (mtDNA) has several vulnerabilities: it lacks the protective proteins that nuclear DNA is wrapped around (histones), has fewer repair mechanisms compared to nuclear DNA, and exists in a harsh environment of oxidative stress generated by its own metabolic activity.

The fragility of mtDNA might have contributed to the relocation of most of its genes to nuclear DNA. Proteins encoded by those genes are synthesized in the cytosol and transported across the cell into mitochondria via a highly regulated process. However, 13 essential proteins involved in oxidative phosphorylation remain encoded by mtDNA and still suffer from the same vulnerabilities. This makes mtDNA prone to mutations, particularly as we age.

Mutations in mtDNA contribute to a range of diseases, such as Leber hereditary optic neuropathy (LHON), and are linked to a wide range of age-related pathologies, including sarcopenia and Alzheimer’s disease [2]. Addressing problems caused by mtDNA mutations is a major challenge in biomedical research. However, in this study, LRI researchers have achieved a significant breakthrough by successfully relocating a mitochondrial gene to the nucleus in vivo.

Overcoming the challenges

Previously, the same team had achieved promising results in vitro [3], but finding a suitable animal model proved difficult: mtDNA genes are so essential that mutations in them usually render mice non-viable. However, a particular strand of mice exists that harbors a relatively benign mutation in ATP8, a gene encoding a subunit of the ATP synthase complex, which causes only a mildly pathologic phenotype. Alongside those mutants, wild-type mice were used as controls.

The team synthesized a nuclear-compatible version of ATP8 and inserted it into the ROSA26 locus, a well-characterized “safe harbor” site in the mouse genome. This locus is widely used in genetic engineering because it allows stable organism-wide expression of inserted genes without interfering with other essential genomic functions.

The researchers had to overcome significant technical challenges to achieve nuclear expression of a gene that is normally expressed in mitochondria (allotopic expression) and to make the protein transferrable to mitochondria. For instance, they found that efficient allotopic expression requires codon optimization: altering the DNA sequence of a gene using codons that are more efficiently translated by ribosomes.

Efficient, persistent, non-immunogenic

Eventually, their efforts paid off: allotopic ATP8 was able to compete with mitochondrial ATP8 even in wild-type mice and outperformed the mutant ATP8. The allotopic gene was expressed in all the tissues that the researchers tested, and the protein successfully integrated into the mitochondrial machinery.

“The key question was ‘How well can an allotopic protein compete with pre-existing protein?’” said Dr. Amutha Boominathan, Assistant Professor and Principal Investigator at LRI and the study’s leading author. “One fundamental concept in the field is that mitochondrial DNA exists because proteins need to be synthesized on demand for easier incorporation into their respective complexes.”

“For allotopic expression to succeed,” she explained, “you must demonstrate that protein coming from the nuclear side can be incorporated with similar efficiency. In wild-type mice, we see equal efficiency between endogenous and exogenous proteins. In our mutant model, we see increasing incorporation over time, suggesting the nuclear protein actually outcompetes the mutated one from a stability perspective.”

Importantly, the allotopic gene functioned well in the genetically modified mice’s offspring for at least four generations, with no adverse effects on fertility. While mtDNA can sometimes trigger an immune reaction when released into the cytoplasm, this gene was also well-tolerated by the immune system, as confirmed by cytokine array analysis.

A blueprint for the future

In the paper, the researchers note that their successful proof of concept does not necessarily apply to all mtDNA genes, and many challenges lie ahead. However, Boominathan is optimistic: “This provides a platform for testing other genes. With appropriate engineering we can overcome all the challenges. We’ve proven it for one protein and have promising data for others. What we’ve demonstrated here is the feasibility of expressing mitochondrial genes in a whole-body context. The inheritance patterns and lack of immune response are particularly encouraging for therapeutic applications.”

There are over 250 mitochondrial DNA diseases that could potentially benefit from this approach, according to Boominathan. “If we can achieve allotopic expression for all 13 genes,” she said, “we’d have a pathway to treat many of these rare diseases.”

The aging connection

LHON, one of the diseases that the researchers are after, “is actually an aging disorder,” Boominathan explained. “While the mutation is inherited, it specifically affects males over 40. These mutations amplify with age, particularly in tissues with high oxidative phosphorylation demands, like retinal ganglion cells. Symptoms only appear when the mutation load reaches a certain threshold.”

This is particularly relevant to post-mitotic cells that form the brain, skeletal muscle, and cardiac tissue since those cells cannot dilute mutations through cell division. “While internal recycling mechanisms like mitophagy exist, they decline with age,” said Boominathan. “If you inherit a mutation or acquire one early in life, it amplifies over time as mitophagy decreases, and these mutations often have an advantage that helps them take over.”

A three-pronged approach

“This work represents the culmination of more than a decade’s worth of effort to provide a genetic backup system for mitochondrial DNA in mammals, for which inherited mutations cause disease in nearly 1 in 200 people,” said Dr. E. Lillian Fishman, Director of Research and Education at LRI, about the study. “I am proud of Dr. Boominathan and her team’s persistence to rise to meet this technically challenging proof-of-concept that paves the way for the treatment of debilitating mitochondrial diseases like Leigh’s syndrome and progressive diseases of aging.”

This study was done as part of MitoSENS, a wider LRI project that includes a three-pronged approach to mitochondrial dysfunction. In addition to allogenic mtDNA expression, the researchers pursue boosting mitophagy with small molecules and de novo synthesis of healthy mtDNA for transfer into exogenous mitochondria, followed by introduction into cells.

Literature

[1] Begelman, D. V., Dixit, B., Truong, C., King, C. D., Watson, M. A., Schilling, B., … & Boominathan, A. (2024). Exogenous Expression of ATP8, a Mitochondrial Encoded Protein, From the Nucleus In Vivo. Molecular Therapy Methods & Clinical Development.

[2] Zhunina, O. A., Yabbarov, N. G., Grechko, A. V., Yet, S. F., Sobenin, I. A., & Orekhov, A. N. (2020). Neurodegenerative diseases associated with mitochondrial DNA mutations. Current Pharmaceutical Design, 26(1), 103-109.

[3] Boominathan, A., Vanhoozer, S., Basisty, N., Powers, K., Crampton, A. L., Wang, X., … & O’Connor, M. S. (2016). Stable nuclear expression of ATP8 and ATP6 genes rescues a mtDNA Complex V null mutant. Nucleic acids research, 44(19), 9342-9357.

In a study published in Aging Cell, researchers have outlined a relationship between mitochondrial fragmentation in skeletal muscle and the loss of strength with age.

Broken power plants

As its authors point out, this is far from the first study to link mitochondrial dysfunction and aging in muscle [1], nor is it the first to connect exercise habits, aging, mitochondria, and the loss of physical function [2].

There has also been significant prior work showing how the mitochondria in muscle tissue behave. Mitochondria in muscle are not equal in their behaviors: the mitochondria closest to the blood-filled capillaries (subsarcolemmal mitochondria) bring energy to more centrally located ones (intermyofibrillar mitochondria) through an intracellular network [3]. Fragmentation of this network destroys this energy transfer but may also offer protection against damage being transferred as well [4].

Too much fragmentation and fission, however, causes muscle wasting in mice [5]; the opposite, mitochondrial fusion, causes muscles to grow in these animals [6]. The researchers’ previous work on healthy volunteers demonstrated that fragmentation begins to occur at day 6 of bed rest, while functional impairments were found to occur on day 55 [7]. However, that work did not prove one way or another whether mitochondrial fragmentation is a useful biomarker or warning sign for age-related muscle decline.

Decline begins before retirement

Wanting to avoid physical inactivity as a confounder and suspecting that this process may not be the same as actual sarcopenia, the researchers recruited a dozen young (average age 27) and ten middle-aged (average age 55) volunteers rather than significantly older people. The older group was slightly more overweight than the younger group.

Unsurprisingly, the younger people’s muscles used more oxygen to generate more power than the older people’s, according to multiple metrics of respiration and energy use. This was not linked to blood flow; instead, it was linked to how the muscles pull oxygen from the blood.

The researchers then examined the mitochondria more closely in biopsied muscle tissue. The total density of the intermyofibrillar mitochondria was the same between younger and older people; however, the older people had more, smaller mitochondria. While their shapes did not differ, markers of mitochondrial fragmentation were greater in this area in the older group.

In the subsarcolemmal area, however, the older people had approximately as many mitochondria as the younger people, which led to a reduction of density with age as these mitochondria were also smaller. Here, too, they were found to be significantly more fragmented. This fragmentation in both areas was associated with the accumulation of fat (lipid) droplets.

Looking ever closer

There were also differences involving the tiny folds inside mitochondria (cristae). Younger people’s mitochondria had regular and dense cristae, while those of older people were less regular, with some areas having no cristae at all. This, the researchers hold, represents “age-associated deterioration at the level of the individual mitochondrion.” Interestingly, however, further data suggests that the increased number of smaller mitochondria may have made up for this, restoring some of the lost function.

The authors then pivoted to the key thrust of their research: the connection between fragmentation and loss of capacity. Fragmentation in the intermyofibrillar mitochondria and a reduction in the cristae was found to be responsible for nearly all of the changes in the well-known metric of VO2max. Unsurprisingly, the density of the subsarcolemmal mitochondria was found to be associated with the muscles’ ability to extract oxygen from blood.

The researchers believe that their findings explain the basic reasons why people lose strength with age, even in the absence of defined sarcopenia. They also warn that this mitochondrial dysfunction only gets worse with aging. Furthermore, they hold that their findings “reflect an early ageing phenotype, making the mitochondrial changes observed herein strong candidates for intervention studies aiming to slow the progression of the effects of ageing on physical function.”

As exercise is associated with mitochondrial fusion [8] and one study had suggested that six months of endurance training can compensate for 30 years of aging [9], the authors suggest that further research on exercise in older people should be done with a close examination into mitochondrial changes.

Literature

[1] Gouspillou, G., Bourdel‐Marchasson, I., Rouland, R., Calmettes, G., Biran, M., Deschodt‐Arsac, V., … & Diolez, P. (2014). Mitochondrial energetics is impaired in vivo in aged skeletal muscle. Aging cell, 13(1), 39-48.

[2] Grevendonk, L., Connell, N. J., McCrum, C., Fealy, C. E., Bilet, L., Bruls, Y. M., … & Hoeks, J. (2021). Impact of aging and exercise on skeletal muscle mitochondrial capacity, energy metabolism, and physical function. Nature communications, 12(1), 4773.

[3] Glancy, B., Hartnell, L. M., Malide, D., Yu, Z. X., Combs, C. A., Connelly, P. S., … & Balaban, R. S. (2015). Mitochondrial reticulum for cellular energy distribution in muscle. Nature, 523(7562), 617-620.

[4] Glancy, B., Hartnell, L. M., Combs, C. A., Femnou, A., Sun, J., Murphy, E., … & Balaban, R. S. (2017). Power grid protection of the muscle mitochondrial reticulum. Cell reports, 19(3), 487-496.

[5] Romanello, V., Guadagnin, E., Gomes, L., Roder, I., Sandri, C., Petersen, Y., … & Sandri, M. (2010). Mitochondrial fission and remodelling contributes to muscle atrophy. The EMBO journal, 29(10), 1774-1785.

[6] Cefis, M., Dargegen, M., Marcangeli, V., Taherkhani, S., Dulac, M., Leduc‐Gaudet, J. P., … & Gouspillou, G. (2024). MFN2 overexpression in skeletal muscles of young and old mice causes a mild hypertrophy without altering mitochondrial respiration and H2O2 emission. Acta Physiologica, 240(5), e14119.

[7] Eggelbusch, M., Charlton, B. T., Bosutti, A., Ganse, B., Giakoumaki, I., Grootemaat, A. E., … & Wüst, R. C. (2024). The impact of bed rest on human skeletal muscle metabolism. Cell Reports Medicine, 5(1).

[8] Huertas, J. R., Ruiz‐Ojeda, F. J., Plaza‐Díaz, J., Nordsborg, N. B., Martín‐Albo, J., Rueda‐Robles, A., & Casuso, R. A. (2019). Human muscular mitochondrial fusion in athletes during exercise. The FASEB Journal, 33(11), 12087-12098.

[9] McGuire, D. K., Levine, B. D., Williamson, J. W., Snell, P. G., Blomqvist, C. G., Saltin, B., & Mitchell, J. H. (2001). A 30-year follow-up of the Dallas Bed Rest and Training Study: II. Effect of age on cardiovascular adaptation to exercise training. Circulation, 104(12), 1358-1366.

In a new study, ChatGPT 4.0 achieved significantly better diagnostic scores when evaluating complex cases than either unassisted human physicians or physicians who consulted the chatbot [1].

Bad news for human doctors?

For millions of people, chatbots powered by large language models (LLMs) have quickly become an indispensable source of information on everything from finances to relationships. These digital aids often come across as more knowledgeable, polite, patient, and compassionate than human experts.

It has been questioned, however, if it is really a smart idea to turn to a robot for medical advice. In what could be a troubling sign for general practitioners, chatbots have shown they can outperform humans in this area too. A study from May of last year found that the earlier version of ChatGPT, 3.5, handily outclassed human health professionals in answering patients’ questions. Responses from both the bot and verified physicians were graded by a panel of health experts, and the gap was striking: for instance, 27% of human answers were deemed “unacceptable” compared to just 2.6% of machine-generated ones.

That study had relied on doctor responses pulled from Reddit, but a more recent study went further. Earlier this year, researchers at Google developed a dedicated model called Articulated Medical Intelligence Explorer (AMIE) and tested it against human primary care practitioners. Wide-ranging health scenarios were distributed at random, with actors playing the roles of patients who discussed their cases with either the chatbot or a human physician without knowing who was who. According to expert evaluators, AMIE outperformed its human counterparts in 24 of 26 categories, including empathy.

“Meet my assistant, ChatGPT”

In a new study published in JAMA Network Open, Stanford researchers stripped AI of its perceived edge in empathy and bedside manner. They eliminated the patient interaction element entirely, tasking either ChatGPT 4.0 or 50 human physicians (26 attendings and 24 residents) with diagnosing six carefully selected cases. These cases had never been published before, ensuring that the LLM could not have encountered them during training.

Here’s the twist: half of the doctors were allowed to consult ChatGPT. The aim was to gauge whether physicians would embrace AI as an assistant and whether doing so would improve their diagnostic reasoning. All participants could also use conventional resources like medical manuals.

The primary outcome was a composite diagnostic reasoning score developed by the researchers, which measured accuracy in differential diagnosis, the appropriateness of supporting and opposing factors, and next diagnostic steps. Secondary outcomes included time spent per case and final diagnosis accuracy.

In the end of the day, the LLM dominated yet again, with a median score of 92% per case: 14 points higher than the non-LLM-assisted human group. It also achieved 1.4 times greater accuracy in the final diagnosis. Interestingly, the group of physicians consulting the chatbot didn’t fare much better than their non-assisted peers, scoring 76% versus 74%.

Why didn’t consultation work?

The researchers had anticipated that consulting the LLM would give physicians a marked advantage, but that wasn’t the case. “Our study shows that ChatGPT has potential as a powerful tool in medical diagnostics, so we were surprised to see its availability to physicians did not significantly improve clinical reasoning,” said study co-lead author Ethan Goh, a postdoctoral scholar in Stanford’s School of Medicine and research fellow at Stanford’s Clinical Excellence Research Center.

Why the lackluster collaboration? The authors suggest a few reasons. First, participants weren’t simply asked to provide a diagnosis. Instead, they had to demonstrate diagnostic reasoning by suggesting three possible diagnoses and explaining how they reached their final choice. The chatbot excelled at this aspect, while humans sometimes struggled to articulate their thought processes. This echoes longstanding challenges in modeling human diagnostic reasoning in computer systems before the advent of LLMs.

“What’s likely happening is that once a human feels confident about a diagnosis, they don’t ‘waste time or space’ on explaining their reasoning,” said Jonathan H. Chen, Stanford assistant professor at the School of Medicine and the paper’s senior author. “There’s also a real phenomenon where human experts can’t always articulate exactly why they made the right call.”

Another hurdle was that physicians often dismissed valid suggestions from their AI co-pilot, a sign that overcoming the natural sense of superiority toward machines may take time.

Finally, the researchers noted that the chatbot’s performance hinges on the quality of the prompts it receives. The research team crafted sophisticated prompts to get the most out of ChatGPT, while human participants often used it more like a search engine, asking short, direct questions instead of providing full case details. “The findings suggest there are opportunities for further improvement in physician-AI collaboration in clinical practice and health care more broadly,” Goh said.

One intriguing secondary finding was that doctor-LLM pairs completed cases slightly faster than doctors working solo. While, according to the paper, the difference of slightly more than a minute was negligible, Goh argues that even a small efficiency gains could help make doctors’ lives more efficient. “Those time savings alone could justify the use of large language models and could translate into less burnout for doctors in the long run,” he said. However, more rigorous studies are needed to fully understand this potential benefit.

AI will not replace doctors (until it will)

The authors of studies like this one have been careful to emphasize that AI is not a true substitute for a human health practitioner. “AI is not replacing doctors,” Goh reassures. “Only your doctor will prescribe medications, perform operations, or administer any other interventions.”

Still, it may only be a matter of time before AI demonstrates superiority over human physicians in nearly every aspect of care. Furthermore, vast regions of the world currently face limited access to healthcare, leaving many people without the option of consulting a human doctor at all. In such contexts, AI could fill a critical gap. Just as some countries skipped the landline phase entirely and adopted mobile phones, they might also be the first to transition to predominantly AI-driven healthcare, facing fewer entrenched bureaucratic barriers.

Building on this study, Stanford University, Beth Israel Deaconess Medical Center, the University of Virginia, and the University of Minnesota have joined forces to create AI Research and Science Evaluation (ARiSE), a network dedicated to evaluating generative AI outputs in healthcare.

Literature

[1] Goh, E., Gallo, R., Hom, J., Strong, E., Weng, Y., Kerman, H., … & Chen, J. H. (2024). Large language model influence on diagnostic reasoning: a randomized clinical trial. JAMA Network Open, 7(10), e2440969-e2440969.