By David N. Leff

Now try this brain teaser: A family with two children, ages 12 and 18, is transferred to a foreign country. Which of the siblings will learn the strange language of his new home faster and with more fluency?

Princeton University neurobiologist Joe Tsien answers. "If you come to a different country when you're a 12-year-old," he told BioWorld Today, "you're able to learn the language without any accent. But if you come at 18 years of age, then you're going to be stuck with some accent.

"Based on our own life experiences," Tsien pointed out, "our memory and learning abilities decrease with age, especially once you reach sexual maturity. I'm not talking about late-stage memory loss, when you get to 80 or so," he went on. "But during the period from juvenile to adulthood is a challenging time. So grown-ups have more difficulty learning new languages compared to, say, 12-year-old kids.

"This dramatic decrease in memory and learning ability," Tsien went on, "has long been postulated as due to the age-dependent decrease in the brain's plasticity - that is, the neurons' ability to modify the synaptic efficacy of their connections."

That postulate was put forward exactly half a century ago by a now-celebrated Canadian psychologist named Donald O. Hebb, Tsien recalled. "In 1949 he tried to explain how learning and memory take place. Hebb's one major hypothesis," Tsien continued, "is that when the two neurons can fire together, they are able to strengthen their synaptic connections. Hebb's rule, as it's now widely known, has become a central doctrine in neurobiology. It's also important in communication networking and artificial intelligence.

"What Hebb's rule really says," Tsien pointed out, "is that almost all forms of learning and memory involve the association of different events. For example, associate your voice with your name, or the name with a face; a particular event with a particular place. You need to have these two events coming together at almost the same time, which we think is probably strengthening that synaptic transmission efficacy in your brain."

Ion Channel Gatekeeper, NR2B Controls Memory

At the molecular level of nerve cells in the forebrain is a key neurotransmitter called N-methyl-D-aspartate (NMDA). Its receptor, NR2B, officiates at the nuts and bolts of learning and memory, mainly in the hippocampus and cortex - regions essential for learning, memory and high cognition. Specifically, NR2B oversees the opening and closing of calcium channels across the cell membrane.

"NMDA acts as a coincidence detector," Tsien pointed out, "which means it's only active when the two cells are firing together. So the NMDA receptor serves that function, controlling the ion channel's opening time. That open-channel duration decreases gradually from juvenile to adulthood age."

Putting that one and one together, Tsien arrived at the hypothesis that "maybe this is due to the decrease of NMDA activity, shortening of the channel duration, which leads to a decrease of plasticity, thus, less learning and memory ability, correlated with influx of calcium ions."

The clinching coefficient in this equation, the Princeton scientist went on, "is the down-regulation of NMDA's NR2B protein. We knew from in vitro experiments that if we add NR2B to this NMDA complex, the channel is able to stay open longer. That's become our central hypothesis," Tsien recounted. "So we asked: 'If there is such a linear relationship, then what if we boost up the amount of protein in the adult brain? Would we see that the channel stayed open longer, retaining or even enhancing the juvenile NMDA channel property, making the brain more plastic, more flexible, with better learning and memory?' All this is exactly what we did see," he recalled, "when we increased NR2B protein in the forebrain neurons of transgenic animals."

A research paper in today's Nature, dated Sept. 2, 1999 - of which Tsien is senior author - reports this feat under the title: "Genetic enhancement of learning and memory in mice." He and his co-authors created these prospectively precocious rodents by "injecting extra DNA sequences of the NR2B gene into the pronuclei of embryonic mice. The transgene construct allowed for the gradual onset during postnatal development, reaching high levels during adulthood, while retaining the forebrain specificity."

Mice Graduate Cum Laude From Memory College

Then the team subjected its transgenic progeny, at various time points of age, to a four-paradigm curriculum of Learning and Memory 202, as vividly described in the Nature article.

¿ Object recognition: Days after exploring two objects for five minutes, the transgenic subjects remembered objects four to five times longer than did their normal control counterparts.

¿ Emotional memory: Placed in a chamber, the animals with overexpressed NR2B in their neurons received mild shocks to their paws. Returned to that memorable chamber of horrors one hour, one day or 10 days later, they evinced a more pronounced fear response than did controls.

¿ Learning improvement: Returned to the shock chamber, but this time without shocks, the transgenic animals quickly learned there was nothing more to fear, while their controls continued to cringe.

¿ Spatial learning: Placed in a swimming pool with a hidden, submerged escape platform, the transgenics learned to find the platform after three sessions; the control mice required six.

As for translating their mouse model's masterful mentation to the human condition, Tsien is cautiously bullish. He made the point, "Learning and memory are not unique to humans. It's evolutionarily conserved throughout the animal kingdom. Also, from what we know of the NMDA receptor, it's nearly identical in different species - from songbirds to mice, rats, cats, monkeys, humans. So from this point of view we think that a critical role for NMDA in humans is very likely.

"Conceptually," he suggested, "the first thing to do is look for a good orally available drug target. You may be able to develop a pill that stimulates or modifies the channel in such a way that it mimics high-level NR2B, so the channel stayS open longer. This, of course, is a pharmaceutical thing."

The clinical scenario he proposed "is gene therapy. Although gene therapy itself is at an infant stage, it does represent an alternative. That means you can deliver those NR2B genes to the brain, and maybe restore certain functions. Patients with cognitive difficulties would be good ones to start with."

"Princeton," he noted, "has a use patent pending, covering the screening of compounds to enhance learning and memory, treat cognitive disorders, or apply these genes in therapy. The university," he observed, "is actively looking for partners to commercialize the technology."

Tsien then alluded to a potential application that he termed "not scientific but ethical and social. It's a more hypothetical situation, which I am very reluctant to get into, but - yes, go on to manipulate human embryos for enhanced intelligence. So if you just talk about whether it's scientifically feasible, the transgenic technology is here, having been developed for over 10 years. There's no technical barrier, really, to applying this to humans. The ethical and social issues, and all such things," he concluded, "are the real difficulties."