What BDNF Actually Does. Part Two: One Letter in the Gene

Abstract

Part One established that BDNF is processed into forms with opposing actions. This part concerns a single-letter difference in the gene that changes how that processing works — and which a large share of the human population carries.

The Val66Met polymorphism substitutes methionine for valine at position 66 of the prodomain. It has been linked to episodic memory, hippocampal volume and risk of mood disorders, and it is one of the most studied variants in psychiatric genetics.

It is also a case study in how a real molecular effect can generate a literature that repeatedly fails to replicate. We present both halves: what the variant demonstrably does in cells, and why the human association studies should be read with considerable caution.

1.A Position Conserved Across Seventy Species

The BDNF prodomain is highly conserved, carrying a valine at or near position 66 in more than seventy species examined to date. Conservation on that scale usually indicates that a position matters — evolution does not tend to preserve details that make no difference.

In humans, a common variant replaces that valine with methionine. Allele frequency varies enormously between populations, and we will return to that number because it turns out to be central to interpreting the entire literature.

The originally proposed mechanism was indirect. The Met66 version of proBDNF shows decreased binding to sortilin, altered intracellular trafficking, and reduced activity-dependent secretion of mature BDNF. On that account, carriers release less mature BDNF when neurons are active, and downstream consequences follow from that shortfall.

2.The Prodomain Does Something By Itself

A 2013 study by Anastasia and colleagues in Nature Communications complicated that picture in an interesting way, by asking what the cleaved prodomain does once it is on its own.

They produced both human recombinant prodomains and tested them on hippocampal neurons in a growth cone retraction assay — growth cones being the exploratory tips of growing neuronal processes, whose extension and retraction shape developing networks.

The result was asymmetric in a way that is hard to explain as a simple deficiency. The Met66 prodomain induced rapid growth cone retraction over twenty minutes. The Val66 prodomain was inactive — and remained inactive even at higher concentrations. The effect was confined to cells expressing p75NTR, the same receptor implicated in the weakening arm described in Part One.

Structurally, the substitution appears to change the shape of the prodomain, which is otherwise predicted to be largely disordered. So the variant is not only producing less of the strengthening signal. It is producing a fragment that actively does something the common version does not do.

The PNAS work discussed in Part One reported a parallel asymmetry: the Val-type pro-peptide facilitates low-frequency-stimulation-induced hippocampal LTD, whereas the Met-type attenuates it. Two independent lines converging on the conclusion that this is a change of function rather than a loss of quantity.

3.Where the Human Literature Gets Difficult

A clean molecular story of this kind creates an expectation that the human association studies will line up behind it. They do not, and a 2018 review by Tsai in Frontiers in Molecular Neuroscience sets out why.

The overall pattern is stated plainly: although many reports have demonstrated possible genetic effects of this polymorphism on disease or brain function, others have failed to replicate the findings, and attempts to replicate specific results have produced inconsistent outcomes.

This is not a case of one contested finding. It is the normal condition of the field, and understanding why is more valuable than any individual result.

4.The Number That Explains Most of It

Here is the detail we found most striking. Considerable BDNF allele and haplotype diversity exists between populations globally, and the frequency of the Met allele ranges from zero to seventy-two per cent across populations.

Read that again, because it does a great deal of work. In some populations the variant is essentially absent; in others a substantial majority carries it. The Met allele is generally reported as frequent in Asian populations and infrequent in Caucasian, Central and South American, and African populations.

Two consequences follow. First, a study conducted where the variant is rare may simply lack the carriers to detect anything, and the review notes that low prevalence can leave studies underpowered when attempting to validate associations found in earlier low-powered work. Second, several meta-analyses have found that positive associations for this polymorphism are dependent on ethnicity — a subgroup analysis found increased risk in Asian populations under one genetic model and in Caucasian populations under another.

When an effect appears under different genetic models in different populations, the most parsimonious reading is not that the biology differs by ancestry. It is that the field has been fitting models to noise in samples too small for the question.

5.What Does Hold Up

Not everything dissolves. The review notes that findings on the interaction between this polymorphism and life stress in depression were replicated by a meta-analysis of thirty-one studies involving 21,060 participants.

That is a large body of evidence and it points to something more interesting than a main effect. The proposal is not that carrying the variant causes depression, but that it changes sensitivity to circumstances — a gene-by-environment interaction, where genotype matters only in the presence of stress.

Interaction findings carry their own methodological hazards and have a poor replication record in psychiatric genetics generally. But twenty-one thousand participants is not a small study, and it is the strongest human evidence in this area.

6.The Mouse Caveat

A knock-in mouse expressing the Met66 variant recapitulates many phenotypic properties of the human polymorphism, which is genuinely useful — it means the variant can be studied experimentally rather than only observationally.

The review attaches an explicit warning to this: findings demonstrated in Met/Met mice have not all been consistently found in human studies. The model reproduces the variant, not the person carrying it, and the gap between those two is where most of the inconsistency lives.

The listed sources of inconsistency are worth repeating because they apply well beyond this gene: age, sex, environmental factors, ethnicity, the genetic model chosen for analysis, and gene-gene interaction. Any of these can turn a real effect into an unreplicable one.

7.How We Would Read a Val66Met Result

Direct-to-consumer genetic testing has made this variant reportable to individuals, which makes the question practical rather than academic.

Our reading is that the molecular finding is solid and the individual prediction is not. The variant demonstrably alters prodomain structure and produces a fragment with activity the common version lacks. What that means for a particular person's memory, mood or response to circumstances is not established, and the association literature is inconsistent enough that no confident personal inference is available.

This is a recurring shape in genetics and worth recognising in general: a mechanism can be real at the level of cells and close to useless at the level of advice to one individual.

Part Three turns to the question that connects all of this to anything a reader might actually be offered — whether BDNF can be measured in blood, and what the number means if it can.

Editorial Comment

MindHeaven® makes no claim that any product interacts with BDNF genotype, and we do not offer or recommend genetic testing. Nothing in this literature would support tailoring a supplement to this variant, and the population data above are the reason: an effect that appears under different genetic models in different populations is not a foundation for personalised recommendations.

We publish this because Val66Met is reported by consumer genetic services and is beginning to appear in marketing. A reader who understands why the association literature is inconsistent is well equipped to evaluate whatever is built on it.