An international team, led by researchers at the Champalimaud Foundation (CF), has shown – for the first time in a realistic way – that it may be possible to diagnose Parkinson's disease (PD) years before it becomes untreatable, by scanning people's brains with functional magnetic resonance imaging (fMRI).
Their results were published today (May 8th, 2025) in the Journal of Cerebral Blood Flow and Metabolism. The study was funded by the €200,000 Mantero Belard Award, a neuroscience prize attributed by the Portuguese charity Santa Casa da Misericórdia de Lisboa.
It has been known for some time that otherwise symptom-free people who are slowly developing PD complain about the loss of their sense of smell. This can happen five to 10 years before they get really sick and experience the full symptoms of the disease: slowness of movement, resting tremor, rigidity, and postural instability.
The significance of this type of sensory impairment in PD has not been the object of much research. Moreover, although a lot of people report loss of the sense of smell, only some of them will develop PD, which means smell impairment by itself is not a specific biomarker for the disease. But what also happens is that people developing PD and related disorders may also suffer from visual deficits and even hallucinations – and it is here that there may be room for more reliable biomarkers.
Now, for the first time, Noam Shemesh (leader of the Preclinical MRI lab at CF) and his team – together with Tiago Outeiro, a neuroscientist and Parkinson's specialist at the University Medical Center Gottingen – have joined efforts and demonstrated that evaluating these two (and maybe other) sensory impairments in the brain, at the same time, could provide robust biomarker for early Parkinson's. And the earlier the diagnosis, the better the chances of developing effective treatments for PD patients.
Ultra-high resolution functional MRI
Using an ultra-high field experimental MRI scanner installed in Shemesh's lab, the researchers submitted a mouse model of Parkinson's to a technique called functional magnetic resonance imaging (fMRI). To give an idea of the power of the experimental machine, it generates a magnetic field of 9.4 Tesla (whereas medical machines typically only go up to 3 Tesla). This substantially improves the images and enables a clear view of brain structures in the small mouse brain.
The transgenic mice used here were widely used in Outeiro's Lab, and carried increased levels of a human protein called alpha-synuclein. This protein is thought to play a major role in disease, as it tends to accumulate and form inclusions, also in the substantia nigra – the brain region that produces dopamine and whose progressive degeneration is responsible for the motor impairments in PD. "The aggregates then spread to other regions in the brain, and affect motor areas", explains Ruxanda Lungu, co-first author of the study.
"This mouse model is very useful", says Outeiro, "because it produces the human type of alpha-synuclein." Moreover, the behaviour of the mice denotes an impaired sense of smell – and it is also thought that these animals experience visual impairment.
The vast majority of fMRI studies in animal models focus on a single sense. We analyzed both visual and olfactory sensory modalities. That's pretty rare in fMRI experiments."
Noam Shemesh, leader of the Preclinical MRI lab at CF
Functional MRI is used to see which areas of the animal (or human) brain are activated in certain conditions – in this case, when they are exposed to odours or to visual stimuli. In the whole-brain images obtained, areas "light up" in response to stimulation due to changes in blood flow and oxygenation, which are driven by the neural activity.
The researchers started by comparing, using fMRI scans, activity in the brains of living mice that produced tangles of alpha-synuclein to those of siblings that did not. The mice were around nine months old, analogous to an intermediate stage of development of PD.
And indeed, the main analyses, conducted by Francisca Fernandes (co-first author on the study), showed that the control mice had normal activity in the corresponding brain areas, while in the Parkinsonian mice, there was much less activity.
Untangling the neural from the vascular
However, the problem with fMRI is that "it doesn't detect neural activity per se", says Shemesh. "Since it relies on interactions between ongoing neural activity and vascular properties, it detects a complicated combination of both effects." And in the present study, it was paramount to "untangle" these two components to visualize the purely neural effects of the disease. "It is very, very difficult to do that with fMRI", Shemesh points out.
So they had to use other approaches as well. Co-author Sara Monteiro assessed the vascular properties with a method called "cerebral blood flow mapping" and showed that the vascular effects were indeed weaker in the PD mice when compared to the controls.
In parallel, Lungu measured neural contributions using a protein called C-FOS, which is released when a neuron is activated. And when she quantified how much of this protein was present in the PD mice's brain (post-mortem), she found that the reduction of neuronal activity was even more pronounced than that of vascular flow. "We concluded that while both the neuronal and the vascular effects exist, the changes we saw in the fMRI scans were mainly driven by neuronal effects," says Shemesh. "The mice neurons were just firing less."
Early biomarkers for PD?
"To our knowledge, this is the first observation of a combined visual and olfactory sensory aberration in the brain activity of PD rodent models in general and the alpha-synuclein model in particular", the authors write in their paper. "This provides an opportunity for future studies to interrogate how sensory deficits progress along the disease, and perhaps lead to early imaging biomarkers [of the disease] (...)."
"Assuming that the effects of alpha-synuclein in the mouse brain and in the human brain are similar, which I think is a reasonable assumption, one of the things we could now do would be to check fMRI signals in the brain of people who are reporting some anosmia, as well as their visual responses", Shemesh explains. "And if we saw something weird in both sensory modalities, we could potentially say that there is something more global happening in their neural circuits, and that we need to follow up on that."
"The big advantage of such a method would be that it is truly non-invasive and easy to perform", says Outeiro. "This could add to the toolbox for diagnosing and classifying PD, something that is urgently needed", he adds.
"I think this work is a nice first demonstration of the fact that we can detect multisensory impairments in the brain fairly robustly", Shemesh concludes. "And it gives us some hope that, with future studies, there will be more things we can look at that hint at early development stages of PD and also determine which treatments might help if they're given early on."
ZIP4 protein identified as driver of glioblastoma aggression