by W. Alex Foxworthy
A few years ago, my father was diagnosed with Parkinson’s disease. For some time before that, one could see that his capabilities were declining, although the changes could still be attributed to age, retirement, or the narrowing of his circumstances. He had been a successful entrepreneur and consultant who worked with large companies across the United States and internationally. Then, over several months, and finally over a matter of weeks, the decline accelerated. I became directly involved in arranging his care and managing parts of his life that he had always managed himself.
I knew Parkinson’s in outline. It involved dopamine, the substantia nigra, tremor, slowness, and a drug called levodopa. When I was in graduate school studying to become a neuroscientist, I was presented with a number of facts about neurodegenerative diseases, including Parkinson’s. But for me, those facts did not amount to an intuitively satisfying explanation. They did not tell me why the disease begins, why it destroys some neurons and largely spares their neighbors, why replacing dopamine can seem miraculous, or why its consequences reach far beyond movement.
Though I have a background in neuroscience, I am by no means a Parkinson’s specialist. What follows is my attempt, as both a scientist and a son, to construct a simple and hopefully true picture of what is going on. The picture has three parts. First, a normally useful protein can enter a self-templating aggregation pathway. Next, the resulting seeds can propagate through connected parts of the nervous system. Finally, they do not affect every cell equally: they take hold most destructively where the cell was already operating with the least reserve.
This framework probably describes much of typical, alpha-synuclein-associated Parkinson’s disease, but it may not describe every biological subtype capable of producing the clinical syndrome we call Parkinson’s. Recent seed-amplification studies detect misfolded alpha-synuclein in most people with typical sporadic disease, but not in everyone diagnosed with Parkinson’s and not with equal frequency across genetic forms. The account that follows is therefore a model of a large and important part of the disease, not a claim that every patient arrives by one molecular road.
The standing condition
Inside the synaptic terminals of our nerve cells is a small, abundant protein called alpha-synuclein. It is expressed broadly throughout the central and peripheral nervous systems and concentrated at many presynaptic terminals. When a neuron fires, tiny membrane vesicles must be gathered, docked, fused, and recycled so that they can release their chemical contents across the synaptic gap. Alpha-synuclein participates in the organization and traffic of those vesicles, although its full physiological role is still being worked out.
What makes the protein unusual is that, by itself, it is intrinsically disordered. It does not settle into one rigid three-dimensional structure. Instead, it occupies a repertoire of shifting conformations. When it encounters the curved membrane of a synaptic vesicle, part of the protein’s chain folds into an alpha helix; when it lets go, that structure relaxes again. Its usefulness, and its fragility, lie in this conditional flexibility. It can respond to a membrane, assist the release machinery, and then return to the soluble pool.
The protein’s useful life is therefore not one durable shape but a metastable range of states: soluble and disordered in one context, membrane-bound and helical in another. A metastable state can persist for a long time even though other arrangements may become more stable once formed, because the molecule must cross an energetic barrier to reach them. Under unfavorable conditions, some alpha-synuclein molecules cross into an aggregation pathway and begin to associate with one another.
The mature fibrillar form of amyloid is a highly ordered outcome of that pathway. Protein chains stack into a repeating cross-beta architecture, forming fibrils that are exceptionally stable and difficult to reverse. The crucial event is nucleation. Before a seed exists, the transition is rare because several molecules must come together in the right abnormal arrangement. Once a seed has formed, its surface can stabilize a similar arrangement in the next alpha-synuclein molecule that binds to it. The assembly grows, and the process can become self-sustaining. The amyloid pathway is therefore a deep kinetic trap: difficult to enter, but much harder to escape once established.
The mature fibril is not necessarily the only, or even the most immediately damaging, species. Smaller oligomeric intermediates can disrupt membranes, mitochondria, vesicle traffic, and protein-clearance systems. Mature fibrils are not inert, however, and the toxicity of any assembly depends on its structure, location, mobility, and cellular context. In some circumstances, packaging smaller assemblies into a larger inclusion may reduce their immediate reactivity; in others, fibrils can continue to seed, fragment, or generate new toxic surfaces. The injury may arise not from one privileged form but from the prolonged, costly struggle to clear, contain, and reorganize what has formed.
The cell pays continuously to keep alpha-synuclein and thousands of other proteins within functional regimes. Molecular chaperones help proteins avoid damaging interactions. The ubiquitin-proteasome system breaks down some damaged proteins. Autophagy and lysosomes engulf and digest larger assemblies and worn cellular components. Mitochondria supply the ATP that powers this work, while mitophagy removes mitochondria that have become liabilities. None of these systems is emergency maintenance performed only after an accident. Rather, they are part of the ordinary cost of remaining alive. A cell’s order is active rather than passive: proteins must be repaired or removed, damaged mitochondria recycled, chemical gradients restored, and errors continually exported. In other words, the cell has to keep paying to go on being itself.
The underlying risk is therefore broad rather than confined to one fragile corner of the brain. Alpha-synuclein expression, cellular architecture, proteostatic capacity, and local inflammatory conditions all vary from place to place, so the risk is not uniform. Even so, Parkinson’s is not explained simply by asking where alpha-synuclein is present. It is distributed widely throughout the central and peripheral nervous systems. The harder questions are where aggregation first becomes self-sustaining, how the resulting seeds appear in other regions, and why some cells fail while others endure.
Those questions can be organized around three verbs: seed, propagate, and catch. A seed is an abnormal assembly capable of reproducing its structure. To propagate is for that structure to appear along connected parts of the nervous system. Catch is an ordinary word borrowed from fire. A spark can land on many surfaces without starting a blaze; it catches only where the material can sustain combustion. In the same way, a seed may reach many neurons, but it becomes a persistent and destructive burden only where the cell cannot clear or contain it. Seed, propagate, catch: three processes with three different logics.
Where the fire is lit
Most Parkinson’s is not inherited in any simple way. The great majority is sporadic: it appears in people with no clear family history and no single identifiable cause. But genetic forms and genetic risk syndromes provide unusually clean clues about what can drive the disease.
One clue points directly to alpha-synuclein. Some families carry extra copies of SNCA, the gene that encodes it. The protein itself is not necessarily abnormal; there is simply more of it. More substrate means more opportunities for a rare nucleus to form, and the clinical pattern follows the dose: additional copies tend to bring earlier and more severe disease. That makes alpha-synuclein difficult to dismiss as a harmless by-product of degeneration.
Other genetic findings point to the systems that handle the burden. Variants in GBA1, an important but incompletely penetrant risk factor, implicate lysosomal function. LRRK2 implicates membrane trafficking, endolysosomal biology, and immune regulation. Recessive mutations in PINK1 and PRKN, the gene encoding Parkin, implicate mitochondrial quality control and mitophagy. These genes do not perform one interchangeable task, and some genetic forms do not show the typical alpha-synuclein pathology. Still, they repeatedly lead us back to the same terrain: a metastable protein; the machinery that traffics, clears, and contains cellular material; and the energy supply that allows that machinery to work.
The genetic cases are therefore not a complete explanation of the much more common sporadic disease. They are closer to a Rosetta stone. Where causation is unusually visible, it repeatedly points toward alpha-synuclein itself or toward the systems that keep protein handling, mitochondrial function, and cellular cleanup within workable bounds. Extra alpha-synuclein provides more material from which an aggregate can form. Impaired lysosomal, mitochondrial, or protein-quality-control systems make an early aggregate less likely to be removed. Both increase the odds that a transient molecular error will survive long enough to become a seed.
What produces that first consequential seed in most people is still uncertain. Some abnormal assemblies may arise from ordinary molecular fluctuations and disappear without consequence. Aging reduces the efficiency with which cells correct such errors. Oxidative stress can chemically modify alpha-synuclein and increase its tendency to assemble. Mitochondrial injury can reduce ATP while increasing reactive chemistry. Inflammation can alter the local environment and strain clearance systems. A seed may therefore arise because an aggregation-prone structure forms more often, because it is cleared less effectively, or both.
Environmental exposures may help create these conditions in at least some susceptible people. The nose and gut are credible candidate initiation sites because they are highly exposed neural-immune interfaces. Inhaled or ingested chemicals, microbial products, local inflammation, and mitochondrial stress can converge there on neurons and supporting cells. The exposure need not bind alpha-synuclein and misfold it directly. It may act indirectly by damaging mitochondria, increasing oxidative stress, changing the protein chemically, raising its local concentration, or weakening lysosomal clearance. Experimental work has demonstrated several pieces of this pathway, and epidemiology implicates some environmental toxicants. What has not yet been demonstrated in humans is the entire causal history: a measured exposure, followed by local mucosal seeding, neural propagation, and eventual Parkinson’s disease.
Once a small aggregated assembly exists, it can act as a template. Its surface stabilizes a similar abnormal arrangement in soluble alpha-synuclein molecules that bind to it, causing the assembly to grow. If pieces break away, each fragment can become a new seed. In experimental systems, seeds can be released from one neuron, enter another connected cell, and recruit the receiving cell’s own alpha-synuclein into the growing structure. Through this process, a rare molecular event acquires a way to reproduce its form.
The technical term for this is prion-like templating. The word prion here refers to conformational templating, not to contagion between people. Parkinson’s is not passed across a dinner table. What may spread is a molecular arrangement within one nervous system, through release and uptake between connected cells and perhaps through several cellular routes. The exact vehicles, and the quantitative importance of this process in naturally occurring human disease, remain uncertain. What is firmly established is the biological possibility of templated seeding and the presence of seeding-competent alpha-synuclein in most people with typical Parkinson’s.
Where does the first important seed appear? The current picture includes at least two proposed broad patterns. In a body-first pattern, the earliest detectable abnormalities appear in the peripheral and autonomic nervous systems, especially around the gut, before they become prominent in the brain. People on this route may have years of constipation, unstable blood pressure, or REM sleep behavior disorder, in which they physically act out dreams, before the classic motor syndrome appears. Animal models show that alpha-synuclein seeds placed in the gut can ascend toward the brainstem along the vagus nerve. At present, the human evidence for this exact route is suggestive rather than conclusive.
In a brain-first pattern, early pathology appears to arise centrally, often involving olfactory or limbic structures before the autonomic nervous system is heavily affected. Loss of smell may precede motor disease by years. Exposure may contribute to some brain-first cases through the olfactory system, but some apparent starting points are not directly exposed surfaces. The body-first and brain-first account is therefore a map of common phenotypes and trajectories, not a direct observation of the first seed, a complete theory of first causation, or an obligatory route for every patient.
Everything in this section concerns ignition: what raises the odds that aggregation begins, where a consequential seed first survives, and how a molecular structure can reproduce. But none of it yet explains why the deepest early loss falls on one small population of cells whose disappearance has made my father slow, rigid, and unable to rise from a chair. Here we move from where the fire starts to where it burns hottest.
Why it comes for him
Once seeded, alpha-synuclein pathology can appear in one connected region after another: this is the second verb in the model, propagate. In a body-first pattern it may ascend from the periphery toward the brainstem; within the brain, experimental and modeling work suggests movement along neuronal connections rather than simple diffusion through nearby tissue. But the pattern permits two distinct and likely overlapping readings. One is transmission: seeds are handed from affected cells to connected cells. The other is exposed weakness: regions appear in sequence because they are the next populations to cross their own limits, like low ground disappearing as water rises. The evidence establishes that templated spread is biologically possible and patterned. It does not establish that cell-to-cell transmission is the sole, or even always the primary, engine of sporadic human Parkinson’s. Both the route taken by the pathology and the condition of the receiving cell appear to matter.
That brings us to the third verb: catch. A seed may arrive in many places without producing the same outcome. Some neurons accumulate Lewy pathology without dying. Other neurons die with little obvious Lewy pathology. A seed catches when the receiving cell cannot clear it, contain it, or absorb the additional burden, allowing aggregation and cellular injury to become self-sustaining. The decisive question is therefore not only whether a seed reaches a population, but how much energetic, proteostatic, and redox reserve that population has left.
The cells most famously lost in Parkinson’s are dopaminergic neurons in the substantia nigra pars compacta, a darkly pigmented strip of the midbrain. Their axons project densely into the striatum, where dopamine modulates basal-ganglia circuits that help select actions and determine how readily and vigorously they are carried out. A relatively small number of nigral neurons must distribute a broad, continuous signal across an enormous territory and maintain it without waiting for a command to arrive. Their systems-level job requires vast axonal arbors, autonomous activity, and continual dopamine production and release. The features that make the signal dependable are the same features that make the cells expensive and fragile.
The first cost is the arbor. Reconstructions in rodents, together with extrapolations to the much larger human striatum, suggest that a single nigral neuron may sustain hundreds of thousands of release sites. In humans, the number may exceed a million, distributed across a branching axonal tree whose total length would extend for meters if uncoiled. Membranes, proteins, vesicles, and mitochondria must be manufactured, transported, repaired, and replaced throughout that territory for decades. This has helped motivate a dying-back model in which synaptic terminals and distal axons begin to fail before the cell body is finally lost.
The second cost is autonomous pacemaking. These neurons generate a slow rhythm even without a command arriving from elsewhere. Sodium conductances participate, but L-type calcium channels also contribute to the pacemaking phenotype of vulnerable populations. Calcium is useful precisely because cells keep its resting concentration extremely low: a small influx can therefore carry a powerful signal. Restoring that low concentration requires pumps and exchangers, and mitochondria take up some of the incoming calcium and use it as a signal to increase ATP production, matching energy supply to activity. In moderation this is useful. Repeated calcium loading, however, pushes mitochondrial respiration harder and increases the chance that electrons leak from the respiratory chain and generate reactive oxygen species. Mitochondrial damage can therefore produce the worst combination: less ATP and more oxidative stress.
A third cost is limited buffering. Compared with more resistant dopamine populations in other brain areas, many of the most vulnerable nigral neurons express relatively little calbindin and related calcium-binding capacity. Calbindin is a marker of resistance, not a proven master switch, and the developmental or functional reason for this difference is not settled. Still, the contrast matters: these cells repeatedly admit calcium while carrying less molecular capacity to smooth the peaks.
A fourth cost comes from the transmitter itself. Dopamine is chemically useful and chemically unruly. When it escapes secure storage inside synaptic vesicles, it can oxidize into reactive products that damage proteins, mitochondria, and lysosomes and can modify alpha-synuclein in ways that stabilize toxic intermediate assemblies. The neuron is therefore maintaining a vast arbor, firing continuously, handling calcium, and manufacturing a cargo that can itself become a source of oxidative stress.
These burdens reinforce one another. Mitochondrial strain reduces ATP and increases reactive chemistry. Reduced ATP weakens protein clearance. Persistent alpha-synuclein assemblies interfere with mitochondria, vesicle trafficking, and lysosomes. Misfolded protein activates microglia and astrocytes, adding inflammatory and oxidative pressure. The cell does not encounter just one poison, instead it enters a feedback loop in which each failing system raises the load on the others.
Put the pieces together and the substantia nigra pars compacta looks like a population operating with unusually little slack. Its cells can sustain this bargain for decades. But when aging, genetic liability, mitochondrial stress, inflammation, environmental injury, and alpha-synuclein aggregation add a new surcharge, the most expensive cells are among the first unable to keep paying.
The neighboring ventral tegmental area supplies a natural comparison. Its dopamine neurons carry the same transmitter and the same alpha-synuclein, and many are also spontaneously active. Yet they are relatively spared. On average, the more resistant populations have smaller or differently organized axonal arbors, rely less heavily on the calcium-linked pacemaking phenotype, and express more calbindin. Other molecular and circuit differences also exist. The contrast does not prove that one feature causes survival, but it supports the broader claim that cellular phenotype and metabolic margin help select the victims.
The calcium story has also met an important negative result. Isradipine, at the dose and formulation patients could tolerate, failed to slow early Parkinson’s in a large trial. That argues against a simple model in which clinically achievable blockade of these channels is sufficient to alter the disease, while leaving calcium handling as one burden among several.
This account makes a testable prediction. If a vulnerability profile built in advance from arbor size, pacemaking phenotype, calcium buffering, oxidative burden, proteostatic capacity, and inflammatory context predicts neuronal loss better than connectivity alone, then metabolic margin is doing real explanatory work. If it cannot, something else is choosing the victims.
Whatever the final weighting of these causes, the central clinical fact is not in dispute. One small population on which the motor system depends is progressively stripped away. For a long time the person losing those cells still looks like himself, only slower. Eventually he cannot rise from a chair. But how does the loss of cells in a small dark band of the brain lead to such dramatic changes?
The slowness that is not weakness
An easy mistake about Parkinson’s is to picture it as weakness, as though the muscles were giving out, or as a creeping paralysis in which motor commands disappear. It is neither. My father’s musculature was not initially weak. The cortical and brainstem circuits that knew how to stand, walk, reach, and lift a fork remained largely intact long after performing those actions had become difficult. Something had changed instead in the way action was selected, initiated, and scaled.
The dopamine lost from the nigrostriatal pathway is often described as fuel for movement, but that is misleading. The motor programs are built elsewhere, and dopamine does not specify the detailed pattern of a reach or a step. Within the basal ganglia, dopamine adjusts the competition among possible actions and the gain with which a selected action is released. It influences how readily movement begins, how large it becomes, and how much vigor the system is prepared to invest. Dopamine in this circuit is closer to a throttle than an engine.
A striking phenomenon makes the distinction visible. Some people with Parkinson’s who can barely initiate walking will briefly move much more fluidly when given the right external cue. A staircase, a metronome, or transverse lines taped across the floor can provide structure for each step; in rare cases, a sudden emergency can call forth paradoxical movement. The underlying movement program has not vanished. A vivid sensory structure can sometimes supply what the depleted internal system no longer generates reliably.
Why should the brain contain a dial for vigor? Movement costs energy and time. An animal should act quickly when the expected return is high and conserve effort when the environment is poor. One influential account treats tonic dopamine as part of the brain’s estimate of average reward rate: a background signal that helps determine how vigorously action is worth pursuing. On that view, depleted nigrostriatal dopamine does not inform the person, consciously, that nothing matters. It changes the machinery that prices action. Movements that remain physically possible are assigned too little gain and too high an apparent cost, and so they are not initiated with ordinary speed or amplitude.
Dopamine also operates across multiple timescales. Rapid bursts and pauses participate in reward-prediction-error learning: they help update expectations when outcomes are better or worse than predicted. Slower changes in dopaminergic tone influence excitability, action selection, and vigor. The distinction between phasic and tonic signaling is useful rather than absolute, but it captures something important. Faster signaling helps revise what the system has learned; slower signaling helps set the gain on acting from that learning.
This is why levodopa can seem miraculous. The brain converts levodopa into dopamine, partially restoring the depleted signal in the striatum. A person who could barely rise may stand and walk. Rigidity loosens; movements grow larger and easier to initiate. For years, often, dopamine replacement gives back a substantial part of ordinary life. That is one of the genuine triumphs of twentieth-century medicine.
But levodopa replaces a transmitter; it does not halt the underlying disease. As nigrostriatal terminals are lost, the striatum becomes less able to buffer fluctuations in drug-derived dopamine. The short-lived medication signal increasingly produces alternating periods of inadequate and excessive stimulation, while downstream plasticity contributes to wearing-off and dyskinesias. The therapeutic window narrows: too little benefit and the slowness returns; more medication, delivered in the wrong amount or pattern, can produce involuntary movement. More decisively, levodopa can help only where dopamine loss is the central problem.
Beyond dopamine
For a while, levodopa gives back much of what the motor syndrome has taken. But Parkinson’s was never necessarily confined to the dopamine system. In many people, autonomic, olfactory, sleep, and other nondopaminergic systems were affected before the first motor diagnosis. What changes with progression is that dysfunction outside the nigrostriatal circuit becomes increasingly prominent, disabling, and resistant to dopamine replacement.
The molecular cast does not necessarily change when the clinical picture broadens. Alpha-synuclein pathology, cellular vulnerability, mitochondrial strain, failed clearance, and inflammation recur in systems that use transmitters other than dopamine. These include autonomic networks that regulate the body, systems that manage alertness and mood, circuits that hold attention and memory, and, in some people, the cortex itself. Because these systems do not depend primarily on nigrostriatal dopamine, levodopa has little to offer them.
This is where the losses stop being mainly about movement and become harder to separate from the person. The autonomic machinery that regulates blood pressure and digestion falters. Sleep becomes disordered. Attention and memory thin; in some patients, their fluctuation becomes one of the disease’s signatures in a later phase: a person lucid in the morning and unreachable by evening, present and then, with fatigue, gone. My father, when he is tired, forgets where he is and who we are to him. It is not a steady erasure. It comes and goes.
Even here, the biology need not be singular. In the cognitive phases of Parkinson’s, alpha-synuclein pathology can interact with amyloid, tau, vascular injury, cholinergic degeneration, and other age-related burdens, as well as medication effects and sleep disruption. The final clinical picture may therefore arise from several processes converging on the same increasingly vulnerable networks. I cannot infer from my father’s fluctuations exactly which process is responsible. I can only describe the pattern by which he is sometimes present to us and sometimes, temporarily, beyond our reach.
The scientific picture can now be stated compactly. A useful protein enters a self-templating pathway. Its seeds can appear along connected networks, but the route of propagation does not by itself determine the victims. Pathology catches most destructively in cells already operating near their energetic and proteostatic limits. The loss of nigrostriatal dopamine lowers the gain on movement, and levodopa can temporarily restore that signal without stopping the process that removed it. The model is coherent. The life it describes is not.
How he meets it
The model is cleaner than the life it describes. Among the quieter cruelties of the disease is what it does to expression. The same reduction of movement amplitude that shortens the gait reaches the small muscles of the face, and the face goes still: the mask, clinicians call it, a loss of the automatic play of expression through which we read one another. My father’s smile still comes. But it comes late and slow, arriving a beat or two behind the thing that caused it, as though it had to travel a long way to reach the surface.
There may be more than one layer in that delay. One is motor: the same failure of gain that keeps him in the chair. Another may lie beyond the dopaminergic motor loop, in changes to attention, motivation, emotion, or cognition. From the outside I cannot cleanly separate them. What I can say is that the smile arrives: slow, delayed, but still his.
A small blessing is that my father, at least right now, does not seem to suffer as I had imagined he would. He sits most of the day, but he does not appear to spend his hours reckoning with how far he has fallen or mourning the man he was. I cannot know what his inward experience is, and I do not want to explain away either his suffering or his peace. Still, I have turned over, more than once, what to make of the quietness with which he seems to inhabit his narrowed world.
Perhaps some of the machinery that would register the size of the loss – the drive, the sharp sense of what matters, the capacity to be moved by things – has itself been thinned. On that reading, the absence of dramatic suffering could be part of the same subtraction: some of the capacity for that particular pain quietly diminished along with the capacities whose loss would otherwise be mourned. But this is only one possibility, and perhaps an ungenerous one. Peace is not always ignorance of loss. Sometimes it is also a way of meeting it.
My father has made friends in the facilities through which he has passed and readily expressed love and gratitude toward family members and anyone who shows him kindness. He has been quick to forgive those who have lost patience with him, and there have been many, both in his personal life and in the institutions through which his disease has carried him. I am sure there is confusion in him, and frustration, and stretches of something darker that I do not see. But the dominant note, for now, is not the one I braced for. He is meeting this narrowing world with a kind of grace and, in the process, teaching me something about accepting the conditions of my own life.
***
Enjoying the content on 3QD? Help keep us going by donating now.