What BDNF Actually Does. Part One: A Protein With Two Opposite Faces

Abstract

BDNF is probably the most name-dropped molecule in popular writing about the brain. It appears as a kind of fertiliser — more of it is better, and various activities and substances are said to raise it. Almost nothing about that framing survives contact with the literature.

This first part follows a 2018 review by Kowiański and colleagues in Cellular and Molecular Neurobiology, together with a 2015 PNAS paper by Mizui and colleagues, to establish something the popular account omits entirely: BDNF is made as a precursor whose fragments do opposite things at the synapse.

One form strengthens connections. Another weakens them. Both are released by neurons, and which one dominates is a matter of processing rather than of quantity. That single fact makes 'raise your BDNF' an incoherent goal — and it is where any honest discussion of this protein has to start.

1.A Molecule Built in Stages

BDNF is not synthesised in its final form. It is produced through a multistage process that generates several precursor isoforms, and the intermediate stages are not inert waiting rooms — they are biologically active in their own right.

After translation and transport to the Golgi apparatus, a signal sequence is cleaved to produce the precursor known as proBDNF. This molecule consists of two parts: a prodomain of roughly 120 amino acids at one end, and the mature domain at the other. Further cleavage separates them, releasing mature BDNF and, as a distinct product, the BDNF pro-peptide.

So the cell can put at least three things into the extracellular space: the uncleaved precursor, the mature protein, and the leftover fragment. Each has its own receptor preferences, and this is where the story stops being about quantity.

2.Two Receptors, Two Directions

Mature BDNF binds the TrkB receptor, a receptor tyrosine kinase. Binding triggers dimerisation and autophosphorylation of intracellular tyrosine residues, producing the phosphorylated receptor that initiates the downstream cascades associated with growth and strengthening. The review notes an additional requirement: the complex must translocate toward cholesterol-rich lipid raft microdomains in the cell membrane for its stimulatory effect to proceed.

The physiological effect described for mature BDNF is maintenance of synaptic strength together with decreased excitability of hippocampal GABAergic interneurons — an effect that enhances long-term potentiation, the cellular process most often invoked as a substrate of memory. Numerous studies have shown that mature BDNF expressed after high-frequency stimulation enhanced LTP.

The precursor takes a different route. ProBDNF interacts with p75NTR and sortilin, and this complex initiates its own signalling cascades — c-Jun N-terminal kinase, RhoA, and nuclear factor kappa B. Work cited in the review reports that proBDNF acting through p75NTR facilitates hippocampal long-term depression, the weakening of synaptic connections, and that in some conditions proBDNF induces neuronal apoptosis through the same receptor complex.

Strengthening and weakening, growth and retraction, from the same gene product at different stages of processing. This is the central fact about BDNF and it is systematically absent from the way the molecule is discussed outside the literature.

3.The Leftover Fragment Is Not Leftover

The 2015 PNAS study pushes this further, and its finding is the more surprising of the two. The pro-peptide — the fragment cleaved off to liberate mature BDNF — turns out to be a signalling molecule in its own right.

The authors report that it is highly expressed compared with proBDNF in the adult brain, that its expression rises during postnatal development and plateaus in adult mice, and that it is released from neurons in an activity-dependent manner. They measured it directly, estimating around 1.63 picograms per microgram of total protein in hippocampal tissue from three-week-old animals.

Functionally, applying the pro-peptide to hippocampal slices facilitated long-term depression induced by low-frequency stimulation. Blocking NMDA receptors abolished the effect, as did ifenprodil, an antagonist selective for the GluN2B subunit — placing the mechanism within a well-characterised pathway rather than leaving it as an unexplained observation.

The authors' own summary is the sentence worth carrying away: BDNF and its pro-peptide may exert opposing roles on the synaptic mechanisms underlying LTD and on the trafficking of AMPA receptors. One cleavage event produces two molecules that push in opposite directions.

4.Why Activity-Dependence Changes the Question

Both the mature protein and the pro-peptide are released in response to neuronal activity — BDNF after depolarisation, as an activity-dependent ligand. One hypothesis discussed in the review links this to stimulation of glutamatergic NMDA receptors and the subsequent influx of calcium.

This matters for how the molecule should be thought about. BDNF is not a background nutrient present at some concentration that could be topped up. It is released where and when circuits are active, and its effects are local and temporally structured.

That reframing dissolves the popular question. Asking how to raise your BDNF is a bit like asking how to raise your muscle contractions: the quantity is a consequence of activity, not an input that can be supplied independently of it. Where the balance between the strengthening and weakening arms falls depends on processing and context, not on total amount.

5.What This Evidence Is, and What It Is Not

We rate this material as moderate rather than strong, and the reason is not the quality of the work — it is the species.

The receptor biology, the cleavage chemistry and the electrophysiology come from cell cultures, hippocampal slices and rodents. The PNAS measurements were made in mice. This is normal and unavoidable: the experiments require tissue, and the manipulations are not performable in people.

The consequence is that the mechanism is well established in animal preparations and inferred in humans. Wherever a claim about BDNF in a person is made, the chain runs through an extrapolation that the underlying papers do not make.

There is one further gap, and it becomes the subject of Part Three. Everything described here happens inside the brain, at synapses, in tissue nobody is sampling. The BDNF that appears in consumer contexts is measured in blood, and whether those two things are the same quantity is a question with its own substantial literature and a less comfortable answer than usually assumed.

6.What Survives

Three things, in our reading, and they are enough to change how the word should be used.

BDNF exists in forms with opposing actions. The precursor and the leftover fragment push synapses toward weakening through p75NTR; the mature protein pushes toward strengthening through TrkB. A single number describing total quantity cannot capture which of these is happening.

Release is activity-dependent, local and structured in time — a consequence of circuits doing work, not a reservoir to be filled.

And the entire mechanistic account rests on animal and cell preparations. It is good evidence about how the system works; it is not evidence that anything a person does or takes moves it in a particular direction. Part Two examines the one place where human genetics does bear on this directly — a single-letter variation carried by a substantial fraction of the population.

Editorial Comment

MindHeaven® makes no claim that any product affects BDNF, its processing, or any receptor described here. No compound in our formulations has been shown to influence this system in humans, and given that the mature and precursor forms act in opposite directions, we would treat a product marketed on raising BDNF as making a claim its own biology does not support.

We publish this because BDNF appears constantly in the marketing of this category, almost always as a single quantity that goes up or down. It is not one thing, and knowing why is a reasonable defence against a great deal of confident nonsense.