July 27, 2021

I, Science

The science magazine of Imperial College

Mind control is an unsettling prospect. Its psychology is far less obfuscated than its biology, which is fantastic for advertising companies and frustrating for scientists. Brains contribute about 2% of our body weight and require 20% of its energy requirement. We recognise early that brains are the seat of our humanity, autonomy and ‘self’ and a basic education engenders the concept of brain cells. Documentaries, public interest articles and ‘brain scans’ have even made most aware that there is functional segregation in parts of the brain. “Emotional processing centres”, “visual cortex” and “memory areas” are certainly phrases not unfamiliar in the public domain.

Unfortunately times have moved on from the Phrenology era, and much of our knowledge about functional areas has been learned through association, which is NOT proof of cause. Two examples: 1) most smokers have stained teeth, and have an increased risk of lung cancer; 2) when someone listens to periodic sounds, certain regions of the brain ‘light up’, and they hear music. In the first example it is silly to assume stained teeth cause lung cancer – although it is not silly to remark upon the association, since both are sequelae caused by smoking. In the second example, people tend to assume that because that a brain region is active whilst harmonies are played, they are forming our perception of music. This is a correlation, not a cause. Some areas might be active because the sound is familiar, or evokes a memory, germinating an emotional response … picturing the face of someone you danced with activates part of the visual system.

Scientists have found various ingenious ways round these artefacts such as using novel sounds, subtracting ‘resting responses’, improving the imaging technique so that the temporal resolution betters matches context … but ultimately these techniques are still associative. This has its uses, for example, chronic pain. Twenty years ago people with no obvious cause of pain or injury were thought to be lying. Now, when you compare ‘pain areas’ of healthy controls and chronic pain patients, there tends to be constitutive hyperactivity in the pain patient. When you give effective treatment, the pain areas stop glowing so brightly. The patient also reports less pain. Therefore, they are probably not lying and clinicians are taking chronic pain more seriously. Thus this has been useful … but it is still associative.

The best proofs come from experiments where a variable is demonstrated to be causal. For example, back in the 1950’s, long before the advent of neuroimaging, Arthur Penfield decided to perform lobectomies to cure intractable epilepsy. To assess precisely which brain regions were too vital to remove, he stimulated various brain regions with an electrode. He was able to show that stimulation of the somatosensory cortex could evoke sensation in different parts of the body (so accurately, he could ‘map’ the body across the parietal lobe). The equivalent in motor regions would produce movement. Stimulation of A1 (primary auditory cortex) produced buzzing, the superior temporal gyrus germinated voices, the occipital cortex (visual system) induced flashing.

If we perform invasive neurosurgery, we must have detailed knowledge of anatomy. This is reflected most poignantly in the case of patient H.M. To treat severe epilepsy, surgeons excised considerable portions of his temporal lobes and underlying structures like the hippocampus. Before this no-one knew what this small, coiled region was responsible for. They could now tell you – memory. H.M retained his past memories but could not form new ones, every three minutes his brain resets and he is stuck in a time warp. His anterograde amnesia is devastating. He does not know he has aged, that his father has died, who his doctors are and cannot recognise his grown children. This will continue until the day he dies.

So for detailed functionality assessment we have become very concerned with how the activity of a particular brain cell relates to perception or behaviour. Arthur Penfield had the right idea, but there are problems with his technique. We have 100 billion brain cells. They each form, on average, 1000 connections. Some of these are to neighbours, some are to distant relatives, some are to managerial ‘hub cells’. Brain cells are not uniform. Some excite other brain cells, some inhibit them. Some brain cells respond to messages that arrive at the same time, while some respond to tardy messages if they are all from neighbouring inputs. The brain epitomises organised chaos.

It is this complex electrophysiology that underlies the neural code: aka the “language of the brain”. Therefore knowing what one brain cell is saying is vital to our understanding of how the brain puts together all these electrical signals into thought, perception, emotion, hearing, seeing and feeling. Electrical stimulation is crude and unwieldy. Imagine a surgeon using a meat cleaver rather than a scalpel. Electrodes are large, excite more than one cell at a time, interrupt intersecting brain cells, and it is not possible to tell by eye whether the cell you are exciting is one of the particular types mentioned above.

But nature has been kind. We have a genetic code. This makes proteins. Proteins run everything inside a cell. Cell function can be carefully and specifically deduced from the types of protein it contains and sticks on its surface like an identifying flag. Suppressive cells like to make a chemical called GABA and contain the right enzymes. Excitatory cells will make Glutamate. Cells in the motor cortex extending to the spine make parvalbumin. In addition to the other problems electrodes are evidently highly invasive; they may allow infection, swelling, neuronal death, irreversible brain damage on implantation, and can’t access deep brain areas without accidentally stimulating others.

Is there a way to electrically excite one genetically specific type of cell without getting inside the skull? YES. Enter the remote control brain. This is where the details get tricky. A brain cell can be activated by causing calcium to enter from the fluid outside of the cell. Cell membranes are lipid and repel ions; the only way for them to enter is through protein channel highways inserted in the membrane. Some ion channels are heat sensitive (TRP family). Step 1) take your genetic marker, a gene specific to the brain cell type you wish to investigate – the protein it makes should be found on the cell surface. Modify this gene so the final protein has a biotin tag. Step 2) make a magnetic nanoparticle (manganese ferrite) and attach it to streptavidin. Step 3) Add the magnetic nanoparticle/streptavidin molecule to the correct area of the brain (injection) and the streptavidin will recognise all the biotin tags. Now all your desired brain cells are magnetically reactive. Step 4) use magnetic fields (e.g. MRI) across the subject’s skull. The nanoparticle will heat up, opening all those heat sensitive ion channels and exciting only the brain cells you have genetically engineered. You can then ask the subject what they are ‘feeling’.

This technique is a long way off in humans – we are not sufficiently adept at gene therapy, which uses viral vectors, to safely engineer human cells, but it holds promise for the future. Quite apart from the experimental potential – showing related cause and function – brain stimulation has notable medical prospects. Parkinson’s patients can undergo deep brain stimulation in order to activate areas which are tonically inactive due to the degeneration of neurons in the striatum. Patients with depression often do well when areas like the reward centre, the ventral tegmentum is electrically excited. In fact electroconvulsive therapy is the MOST effective therapy in severely depressed and suicidal patients.

So we are a long way from mind control, but to really understand how our brains work, we must move on from simply observing patterns of behaviour, towards learning how to cause them. Associative experimentation is no bad thing. Associations are vital to survival and it’s why we have memories. “It is night. There are predators.” “She’s smiling. It’s safe to come out now.” “There is smoke. There is fire.” Relating this last to smoking, a high proportion of lung cancer patients smoked. It was later proved that carcinogens in cigarettes were capable of mutating DNA. Hence the now national campaign for the abolition of smoking. In an effort to point out the dangers of coincidence, it should be said that no-one has suggested, quite rightly, that lung cancer would be reduced by cleaner teeth (the first example earlier in this article). Meanwhile tooth brushing has been associated with heart disease …