101 Theory Drive
A Neuroscientist's Quest for Memory
Pantheon: 272 pp., $25.95
There's a cartoon on my office wall captioned, "How our
recalls things." It shows an old galoot (overalls, baseball cap) in a stockroom, leaning on the drawer of a filing cabinet, one hand draped across the folders, the other holding up a sheet of paper. A phone receiver is tucked under his chin, and he seems to be relaying the extracted information to someone upstairs.
It's a satisfying metaphor for a process that neuroscientists have struggled to pin down for decades. In "101 Theory Drive," Terry McDermott gets us a lot closer to the problem of how the brain records experience. The intrepid McDermott, a former national reporter for The Times with no background in neuroscience, does this by embedding himself in the lab of Gary Lynch, a leading memory researcher and one of the field's most radical practitioners. "101 Theory Drive" is the lab's address at
Hard-drinking, cigar-chomping and potty-mouthed, Lynch — described by one colleague as "the hippie of neurobiology" — is nothing if not good copy. In December 2004, when McDermott met him, he was 61 and "had a reputation for being exceptionally, even stupidly, pugnacious, but also … uncannily right about a lot of things." Lynch had spent half his life pursuing the physical manifestation of memory in the brain, phantoms that must nevertheless exist, since, as McDermott aptly states, "If memory left no mark, then there could be no such thing as memory, no such thing as a personal past, no learning, no store of … knowledge." To a great extent, he writes, "our memories constitute our selves."
How is a memory — a single event, among countless such events in any one day (or in any one five minutes, for that matter) — captured within the human brain's 100 billion neurons, each with dozens of dendrites (branches) and communicating via synapses, electrochemical exchanges numbering "anywhere from 100 trillion to many quadrillion"? The collection of brain neurons you're born with is, after some pruning, by and large with you for life; therefore, McDermott notes, "[i]f new information does not come in new cells, the old ones have to change."
For Lynch, the change — the memory trace — is effected by a process called long-term potentiation (LTP), which strengthens interneuronal communication. When he began pursuing this notion in the 1970s, he was scoffed at; nowadays, LTP's role is widely accepted. Experimenters in his Irvine lab have found the resulting shape changes in neurons — shape changes encoding a memory in neuronal networks. Of course, this is putting it simply: You're looking for changes at the postsynaptic sites, the spines of dendrites; each of the dozens of dendrites on a single neuron has thousands of spines. You're working with a wee slice of rat hippocampus (the "you" here is not Lynch himself, who is allergic to rats), measuring actions on millisecond time scales. The rat brain has something like a trillion synapses. McDermott notes that its use in neuroscience labs (and also the use of fruit flies, worms and sea slugs) is "a daily expression of absolute trust in evolution as a fact of human history.... Even as debates raged in the broader society over the idea that human beings are descended from apes, the routine use of animals to model human beings in biology labs around the world was an affirmation … that human antecedents go back way past the apes to the flies and beyond."
McDermott is good at detailing the tedious and slippery bench work that the search for memory entails. Sometimes you're attacking not sliced rat brain but a single dissociated neuron, puncturing its membrane with an electrode, "creating a tiny electrical circuit." Once you've managed that, you "introduce changes to the chemical composition of the cell's interior" and measure the results. People in Lynch's lab, McDermott reports, spend "hours at a sitting, days or weeks in succession — staring at graphical renderings of those results on computer screens … unmoving except for the occasional precise note written in a lab journal." The lab, he says, is "quiet."
He is also good on the sociology of neuroscience: He contrasts the "wild man" Lynch, fond of sniping at the "pygmies" in his department, with his gentlemanly counterpart, Columbia's Eric Kandel, who proposed protein synthesis as the biochemical basis of long-term memory. "Lynch thought the emphasis was wrongheaded, but he could do little to overcome it," McDermott writes. "Kandel, for his part, had nothing but nice things to say about Lynch." Kandel probably typifies the neuroscientific establishment; Lynch, at least while his LTP research was being ignored, "largely absented himself from the numerous academic conferences and symposia at which neuroscience findings are presented and debated and, not insignificantly, reputations are made and maintained."
In 1978, Lynch co-founded a biotech company called Cortex to capitalize on discoveries made in his lab (a not uncommon practice in the neurosciences; Kandel established one called
), specifically to develop ampakines, drugs that enhance communication between neurons. The supreme ambition of all such work is to wipe out Alzheimer's disease, an outcome currently only a blip on the horizon. Biology is not an exact science. McDermott nicely characterizes the human organism as "a magnificent contraption, a sackful of accidents." He concludes at the end of his long stint at 101 Theory Drive: "We are compromised, full of repurposed parts, always building on top of what was already there. The human brain, by any definition, is a kluge. It's amazing it works at all."