快猫短视频

IN ONE of Jorge Luis Borges鈥檚 bleaker short stories, a young man called Funes
is cursed with perfect recall. Remembering every wrinkle in every face he sees
and the shape of every cloud he looks at, Funes sinks into a pit of madness and
despair as his mind fills up with garbage.

Gary Lynch鈥檚 rodents suffer no such problem. Like lab rats everywhere, they
have minds like sieves. Sure, when Lynch, a neuropharmacologist at the
University of California at Irvine, puts the creatures in one of his
mazes鈥攁 set of identical tracks fanning out from a central pen鈥攖hey
soon learn to navigate it, taking their cues from landmarks in the room beyond
the maze. But whatever is learnt is quickly forgotten. Eight hours after their
last visit, even the most experienced rats struggle like novices to get their
bearings in the same maze.

Enter Lynch with one of the pills he and his colleague Gary Rogers have spent
the past few years developing. Now the rats seem to soak up information better.
Eight hours after the training session, they still remember the layout of the
room. Up and down the tracks they scuttle, retrieving food rewards twice as
efficiently as normal. 鈥淭here is no question,鈥� claims Lynch, 鈥渢hat their memory
is much stronger.鈥� Furthermore, the improvements are greatest in middle-aged
animals.

Might the same pills help forgetful 鈥渇iftysomethings鈥� remember where they
parked their car after a trip to the theatre? Sceptics abound. In humans, 鈥渢he
idea that you can take a pill and fix something as complicated as learning and
memory is unlikely鈥�, says Charles Stevens of the Salk Institute in San Diego, an
expert on the cellular basis of learning.

But in the Lynch camp these days there is only optimism about the chemical
compounds they have christened 鈥渁mpakines鈥�. Drugs, like Valium, that dampen
electrical activity in the brain have been around for decades. Now, for the
first time, memory researchers have compounds that amplify the chemical signals
that are the brain鈥檚 main means of conveying electrical impulses from neuron to
neuron. A clinical trial involving Alzheimer鈥檚 patients has just begun at the US
National Institutes of Health. And in preliminary tests, Lynch and his
colleagues saw few signs of side effects when they gave the mildest of their
ampakines鈥攖hey have synthesised more than a hundred in recent
years鈥攖o men with normal memories (This Week, 23 November 1996, p 14).

There was even a hint of efficacy. In a study to be published in
Experimental Neurology in May, for instance, young men performed up to 20
per cent better in some standard tests of short-term recall and learning. In a
second study, published in the same journal, men in their 60s and 70s doubled
their scores in a simple test of short-term recall.

Whiz kids

For Lynch, this is all very gratifying, because ampakines are designed to
make neurons鈥攐r rather the synapses that connect them鈥攎ore
responsive to natural chemical signals that help trigger learning mechanisms in
the brain. The signals are carried across synapses by glutamate molecules.
Ampakines amplify them by stimulating some of the receptors whose job it is to
respond to this glutamate.

But don鈥檛 hold your breath: Wall Street is not about to be invaded by
70-year-old whiz kids on ampakines just yet, and nobody is about to leap to
conclusions from two preliminary studies鈥�well, almost nobody. When Lynch
announced the human findings for the first time at an international meeting for
neuroscientists last November, Cortex Pharmaceuticals, the Irvine-based company
he helped to found in the late 1980s, instantly doubled its stock value.

The dark whispers about commercial hype that followed weren鈥檛 wholly
surprising. Time and again, labs have come up with so-called 鈥渃ognition
enhancers鈥� that work wonders on rats but flop in controlled clinical trials.
Many turn out to be no more beneficial than a stiff espresso. Either that, or
they have side effects that limit their use to the severely ill. And one basic
question continues to divide researchers: just how manipulable is the brain
chemistry of learning? Everyone agrees it is possible (in theory at least) to do
something for Alzheimer鈥檚 sufferers, where there is a clear loss of synapses and
chemicals involved in learning. What is contentious is whether science has any
chance of remedying the subtler effects of normal ageing in the brain.

But belief in this possibility is steadily gaining ground, thanks to the
giant strides now being made in understanding how synapses work.

The supremacy of the synapse in theories of learning is nothing new. For
years, neurobiologists have suspected that the brain soaks up new information
simply by adjusting the strengths of the links between its neurons. Now the
evidence for this idea is looking indisputable. Using miniature electrode
devices, researchers can eavesdrop on brain circuits as lab animals lay down new
memories. And drugs and genetic techniques which can raise or lower the learning
thresholds of synapses with pinpoint precision are emerging. Almost at will,
researchers can manipulate synapses to make them super-supple or as rigid as
wood.

The effects on the navigation skills of rats and mice are telling. Einstein
or dunce, it all depends which synaptic receptor or enzyme you stimulate, block
or remove. 鈥淵ou just have to hit the right spot,鈥� says Matthew Wilson, a memory
researcher at the Massachusetts Institute of Technology in Cambridge.

Rodent dunces

Wilson should know. The genetically manipulated mice that inhabit his MIT lab
are as synaptically challenged as rodents in a maze can be鈥攁nd he has the
technology to prove it. A few years ago, working with Bruce McNaughton at the
University of Arizona in Tucson, Wilson began developing a new type of miniature
electrode system. Existing microelectrodes were good enough for eavesdropping on
the electrical activities of single neurons in rat brains. But by this stage it
was clear that memories, even the simple spatial ones acquired by rats in a
maze, are not solo works, created by single neurons adjusting the strengths of
single synapses. They involve whole choruses of neurons, whole networks of
synapses. To catch a rat鈥檚 brain soaking up information, reasoned Wilson and
McNaughton, one would have to eavesdrop on entire networks while the animal
scampered around a maze.

By 1993, the researchers had an answer to the problem鈥攁 miniature array
of super-fine electrodes. Powered by a microchip, the device could be implanted
into the brains of rats without restricting their ability to move. And once in
place, it could eavesdrop on upwards of 30 neurons at a time. A new era in
memory research could begin, yet still some key experiments stayed out of reach.
The genetic engineers who had begun to study learning mechanisms, at places like
MIT, weren鈥檛 interested in lab rats. They worked instead with mice, whose brains
are ten times smaller. For these animals, the electrode device was too big.

But last year Wilson succeeded in shrinking it further. Now down to a few
grams, the device would at last be used to keep close tabs on the brain circuits
of lab mice as they learn鈥攐r in some cases, don鈥檛 learn.

Take those rodent dunces in Wilson鈥檚 lab. Until now, manipulating genes in
mouse brains has been messy, producing results that are difficult to interpret.
But over the past few years, MIT researchers, led by Nobel prizewinner Susumu
Tonegawa, have been trying to remedy this problem by perfecting a brand new
technique, which allows them to switch off specific genes in specific groups of
brain cells. The dunces are their first success.

As far as anyone can tell, the brains of these mice are normal except in one
small area. Here, in a structure known as the hippocampus, that is essential for
laying down memories, cells have been robbed of a gene they need to make a type
of glutamate receptor called the NMDA receptor. Lacking these receptors, the
mice are hopeless in the maze. 鈥淭hey wander all over the place,鈥� says Wilson.
鈥淵ou simply can鈥檛 train them to remember spatial cues.鈥�

Nor can you do much with their hippocampal synapses. In normal mice,
artificially strengthening such synapses is easy. You simply stimulate neurons
with high frequency pulses of electricity. But try this in the dunce mice and
nothing happens. The synapses continue to transmit impulses with lacklustre
efficiency; or as neurobiologists put it, the synapses fail to show long-term
potentiation, or LTP.

Such woodenheaded behaviour confirms what researchers have long suspected
about NMDA receptors: they are vital to the suppleness of synapses in the
hippocampus. Without them, synapses cannot set off the changes in enzyme and
gene activity that build long-lasting strength. But for the first time, Wilson
and his colleagues have been able to take this insight further. Armed with their
eavesdropping device, they have been able to investigate why NMDA receptors and
supple synapses are so important for learning.

Back in the maze, it turns out that normal mice triumph by creating internal
maps of their surroundings. Forget A-to-Zs and road atlases. Fluid and
schematic, the mouse maps are based on the behaviour of a network of neurons in
the animal鈥檚 hippocampus. Like pensioners in retirement homes who become wedded
to favourite chairs, each neuron in the network grows 鈥渁ttached鈥� to a particular
patch of the maze, firing when鈥攁nd only when鈥攖he animal wanders into
it. What鈥檚 more, each neuron鈥檚 鈥減atch鈥� tends to overlap other patches. So
wherever the mouse wanders in the maze, two or more of these 鈥減lace sensitive鈥�
neurons will always be firing in concert. But is this orchestrated firing really
the essence of a memory in the making? Does it actually remind the animal where
it is?

For the answer, Wilson and his colleagues went back to the dunces. This time,
when they used their eavesdropping device, they detected none of the usual
discipline. The 鈥渁ttachments鈥� of the neurons to particular areas of the maze
were fickle and unfocused. The network, says Wilson, behaved like a collection
of randomly connected neurons. Stevens, who has studied the results, comes to a
similar conclusion. In normal mice, he says, the neurons tend to fire in
synchrony, but in the mutant animals the synchrony is lost. 鈥淚n lots of cases,
when neuronal firing is really significant for something, it tends to be
蝉测苍肠丑谤辞苍颈蝉别诲.鈥�

For rodents in mazes, then, the rules of learning seem clear: no NMDA
receptors means no tuning of synapses to a particular landmark, which means no
synchronised electrical activity in the hippocampus鈥�which means no
learning.

Perk me up

If this seems no great revelation, think again. Brain researchers have spent
years trying to establish solid links between things like LTP and NMDA receptors
on the one hand, and the behaviour of living animals and their brain circuits on
the other. What鈥檚 more, it鈥檚 unlikely that mice are the only creatures to use
synapses in this way. Wilson believes we humans use similar mechanisms to lay
down memories, and not just of where we parked the car. It seems likely that
recalling names, faces, and information in general, requires the brain to
鈥渞ecreate鈥� particular patterns of synchronised electrical activity associated
with the information. If so, then learning the information in the first place is
simply about making those patterns more likely. Tuning synapses might be the
brain鈥檚 main way of achieving this.

For Wilson, the research on the mutant mice is important for another reason.
鈥淚t shows,鈥� he says, 鈥渢hat extremely subtle manipulations of synapses can
produce profound cognitive effects.鈥� And if we can shut down learning mechanisms
so smartly, maybe we can perk them up as well. 鈥淭he idea doesn鈥檛 have to be
restricted to severe pathologies,鈥� says Wilson. 鈥淭here鈥檚 a good possibility that
pharmacological treatments could compensate for general memory decline during
补驳别颈苍驳.鈥�

A couple of decades or more ago, such optimism might have seemed rash. Then,
researchers thought the ageing process destroys whole brain cells. Increasingly,
however, evidence from postmortem studies suggests that if there is a consistent
pattern of change in normal brains鈥攁nd some researchers still dispute
this鈥攊t involves nerve endings shrinking in certain areas. Synapses, not
cells, are what seem to disappear. And those synapses that remain, suspect
researchers like Wilson and Lynch, may also lose some of their suppleness.

One answer might be to stimulate NMDA receptors. Squeezing more activity out
of these receptors as people enter their old age might help to keep their
synapses supple for longer. The problem for researchers has been to find
effective compounds that don鈥檛 produce side effects. But hope springs eternal,
and from spider venoms to the toxins of tropical snails, the search
continues.

Jogging cousins

Lynch, meanwhile, is taking a different tack. His ampakines target another
type of glutamate receptor with a vital role in learning. AMPA receptors handle
virtually all routine transmission of electrical impulses in the brain. NMDA
receptors might throw the switch for learning, but it is their AMPA-type cousins
that jog them into action, by triggering, or 鈥渆xciting鈥�, electrical activity in
neurons. 鈥淥ne of the singular advances in the last few years,鈥� says Lynch, 鈥渉as
been the identification of the AMPA receptor as the main agent of excitatory
communication in the brain.鈥�

The bad news is that these communication links seem to wane as we age.
鈥淎lmost from our 20s onwards we鈥檙e losing glutamate synapses in the neocortex,鈥�
says Lynch. 鈥淏y the time you鈥檙e 70 or so you鈥檝e lost 20 per cent of these
synapses. That can鈥檛 be good.鈥� Nor is the news much better from the hippocampus.
Here, more than 90 per cent of synapses are powered by glutamate and some
estimates of the number lost during ageing run as high as 40 per cent. Might
stimulating AMPA receptors help? Could it increase the efficiency and suppleness
of the synapses that remain?

In the late 1980s, recalls Lynch, 鈥渢hat was just a dream because there was no
way to do it鈥�. Then in 1990, he received a copy of a paper from Isao Ito at the
Japanese drugs company Chugai Pharmaceuticals. By screening chemicals, Ito had
discovered the answer to Lynch鈥檚 dream鈥攁 compound that prolonged the
electrical currents produced by the AMPA receptor. 鈥淚 said, my God, this guy has
come up with a drug that is going to up-modulate one of the primary receptors of
the mammalian brain.鈥� That very day, Lynch went out and got some of the
compound. 鈥淚 came back to the lab with my research associate, Ursula Staubli, at
about 7.30 at night and we threw it into some brain slices鈥�I was
astounded at what we saw.鈥�

Since then, Lynch and his colleagues have synthesised scores of similar
compounds, testing each one in turn in lab animals and brain slices. They have
not been disappointed. MIT鈥檚 genetic engineers might excel at taking the
suppleness out of synapses; at nobbling the NMDA receptors so that the brain
circuits are useless at soaking up information. But the Lynch team has achieved
the opposite. By stimulating the AMPA receptors with drugs, they make it easier
to artificially strengthen glutamate synapses in LTP experiments in rats; easier
to detect strong firing of 鈥減lace sensitive鈥� neurons in rats in mazes; and
easier, too, to train rats to discriminate between odours. 鈥淎mpakines cut the
training period in half,鈥� says Lynch. 鈥淲hat they鈥檙e doing is enhancing
neurotransmission in the brain and thereby making it easier for things like LTP
to happen.鈥�

Still, not everyone thinks ampakines will produce such clear-cut results in
humans. Some sceptics say there is no good evidence that losing synapses weakens
memories in normal people. After all, they point out, the brain loses billions
of synapses early in life as part of its normal developmental process. Others
worry about the drugs鈥� specificity, or rather the lack of it. 鈥淭he idea of a
drug that increases synaptic transmission all over the brain and makes people
learn better is naive,鈥� says Stevens.

Certainly, stimulating glutamate receptors in the hippocampus can be
dangerous. High doses of ampakines, for example, cause brain seizures in rats.
Other side effects could arise from the drug鈥檚 action on brain cells that play
no part in learning.

Yet, at the doses used so far, Lynch and his colleagues have seen only one
side effect in humans鈥攁n increased tendency to swallow. Indeed, far from
being worried about ampakines鈥� power to stimulate synapses in many different
parts of the brain, Lynch sees it as a bonus. It may pave the way for ampakines
to treat mental conditions other than memory loss.

Photographic memory

鈥淧lanning for schizophrenia trials is well along,鈥� says Lynch. Researchers
are starting to suspect that schizophrenia is the result of several different
chemical imbalances in the brain, one of which is reduced communication at the
glutamate synapses in the front of the cortex. 鈥淎mpakines are drugs that could
boost this transmission, they鈥檙e an obvious therapy to try.鈥�

Even hardened sceptics would agree about one thing: we haven鈥檛 heard the last
of glutamate receptors and scientists鈥� attempts to manipulate them. Estimates of
the number of subtypes of these receptors in the brain run as high as a hundred.
But when it comes to manipulating learning and memory processes, even this
extended receptor family won鈥檛 have the last word. There are plenty of other
options emerging from labs, including sex hormones and nerve growth factors,
which seem to influence the suppleness of synapses, as well as new ways to
influence our ability to concentrate while learning new information.

Of course, the sceptics may be right. None of this may in the end produce
drugs that work safely in normal people. But what if it does? Photographic
memories, like the one that tormented the fictional Funes, are far-fetched. But
you don鈥檛 need an imagination the size of Borges鈥檚 to realise that pills which
help people soak up information wouldn鈥檛 be all good news. Problem number one:
would we ban them from schools?

Further reading can be found on 快猫短视频鈥檚 Planet Science
Web site at /ps/ns/reading/brainboost.html.
New readers will need to register at /