Irresponsible Musings on Lamarckian Evolution
Markus Meister — Caltech neuroscientist and formerly Harvard biology chair — delivers a 2018 Janelia conference talk in transcript form. Flagging it as 'utter speculation', he asks whether the brain's enormous capacity for learning from experience during a lifetime leaves any trace in the genome for future generations. He sketches a theoretical mechanism: RNA synthesized at newly formed synapses might travel via exosomes through the blood-brain barrier to the testes, where, potentially guided by endogenous CRISPR-like machinery, they could bias future neural wiring. He cites CRISPR in bacteria as proof that acquired immunity can be written into DNA.
Opens on markusmeister.com · Curated by GlobeRead
GlobeRead's Take
Lamarckian inheritance — the idea that acquired traits can be passed to offspring — was expelled from mainstream biology so thoroughly after the Weismann barrier and the Modern Synthesis that even raising it seriously requires a disclaimer. Meister opens with a 'trigger warning' for audiences at American universities, and then proceeds anyway. The essay is interesting precisely because it is so careful to flag its own speculative status while still making the underlying question feel genuinely alive.nnThe first key move is the real existence of Lamarckian mechanisms in biology. CRISPR did not arrive as a gene-editing tool; it evolved as a bacterial immune system that inserts sequences from viral infections into the bacterial genome, giving descendants molecular memory of threats the parent encountered. This is, by definition, acquired inheritance — the environment writes itself into the genome. Meister's question is simple: if evolution found this mechanism once, in a relatively simple single-celled organism, could it have found analogous mechanisms in more complex animals where the challenge (communicating from plastic brain circuits to germ cells) is harder?nnThe second and more specific move is the proposed mechanism. A newly formed synapse between cell types A and B, Meister hypothesises, could produce a unique small RNA signal — about 10 bases, which is enough to encode a million possible synaptic event types. Extracellular vesicles (exosomes) are already known to transport RNA, cross the blood-brain barrier, and be taken up in the testes. Endogenous RNA-directed DNA editing mechanisms have been reported. Each link in the chain is individually plausible; the chain as a whole is highly speculative. But the speculation is grounded enough to be testable, and Meister proposes an experiment: chemically label all RNA synthesised in the brain and look for specific sequences arriving in germ cells.nnWe picked this because it is one of the rarest forms of scientific writing — a senior researcher publicly entertaining a heterodox idea under their own name, with full acknowledgment that they might be completely wrong. The question it leaves you with is what other ideas are sitting at conferences on napkins waiting for someone to be irresponsible enough to say them out loud.
We curate the internet's most interesting articles — science, psychology, history, philosophy and more. Always free, no algorithm.