|  It
is difficult to overstate the likely impact of recent advances in mammalian
genetic engineering.
The ability to engineer the mammalian genome will
eventually rank along with splitting the atom, developing the computer,
and inventing the printing press in its impact on the human condition.
If taken to its limits, genetic engineering could allow us to restructure
ourselves at the molecular level—a thought both ethically sobering
and medically promising.
Emerging genetic engineering technologies allow us to delete, mutate, and
overexpress mammalian genes in a variety of animal models. It is also possible
to introduce genes from one species into another, thereby adding to its
cellular milieu a new protein not previously available to that species.
This has been used to confer disease resistance from one plant to another.
Plant genes can also be introduced into the mammalian genome for use as
unique cell tracing markers. Developing genetic engineering technologies
will soon allow us to remove a gene at different times during development:
from conception up through adulthood. This will help investigators to determine
the function of specific genes. The development of gene deletion technologies
specific to particular tissues, organ systems, brain regions, and cell
types will allow us to implement sophisticated and highly restricted gene
product excisions in particular subdomains of the body.
At present, most of these technologies are being developed in mice. Genetic
engineering first became a practical reality in mice largely because of
their convenient size, short generation time, and long history in experiments
in mammalian genetics (many varieties of mouse mutants are commercially
available and the mouse genome is relatively well understood). Mouse embryonic
stem cells are particularly robust when applied to homologous gene
recombination experiments. In this process, an altered gene introduced into the mouse
genome replaces the normal, unaltered variant, a first step in the production
of what is commonly referred to as knock-out mice. For these reasons, the
laboratory mouse currently represents the state of the art in genetic engineering
technology.
Understanding and treating human medical disorders is an important application
of these technologies. In this column and the next, we address the contribution
of genetic engineering technology to our understanding of the molecular
basis of human mental retardation syndromes. These advances have been based
on recent in vitro studies as well as the use of genetically engineered
mice. A unified model is beginning to emerge that helps explain the complexities
of the molecular signal transduction pathways that underlie mammalian learning
and memory. This, by extension, is improving our understanding of cognitive
disorders related to these processes. The present column describes the
major signaling pathways involved in normal learning and memory, and the
next column reviews disruptions in these pathways and their contribution
to abnormal cognitive development.
The most sophisticated approaches to genetic engineering have been developed
in laboratories focused on understanding the molecular basis of learning
and memory. Two Nobel laureates, Susumu Tonegawa and Eric Kandel, along
with their students and colleagues, have pioneered the use of genetically
engineered mouse lines to investigate specific modification of the mouse
genome. These models of cognition are advantageously positioned to lead
us to further insights into the molecular underpinnings of higher-order
CNS function. These investigators have discovered that learning activates
specific signaling pathways that lead to the phosphorylation of downstream
target proteins associated with the formation of memory. First we review
how phosphorylation leads to protein activation, and then we turn to the
molecular pathways and proteins involved in learning and memory.
After a cell assembles a specific protein from a string of amino acids
(under the direction of a specific gene), many proteins undergo additional
changes before they can function. This is called posttranslational
modification.
One of the most common posttranslational modifications is the addition
of one or more phosphate groups. When these bulky and highly charged moieties
are added, the protein undergoes a change in structure: a conformation
change. For example, the addition of a phosphate group may cause a conformation
change that brings a previously hidden amino acid sequence from the central
core of a protein to the surface, where it can interact with target proteins.
Thus the addition or removal of phosphate groups can alter a protein’s
enzymatic activity or its ability to interact with other proteins. This
process is common to many organisms and appears to have been an early event
in the evolution of eukaryotic cells. Several families of enzymes have
evolved over time that either add or remove phosphate groups. The class
of enzymes that adds phosphates are kinases and the enzymes that remove
them are phosphatases.
Typically, kinases or phosphatases recognize specific amino acid sequences
on target proteins called consensus sequences. Rapid phosphorylation or
dephosphorylation of an amino acid occurs within (or adjacent to) the consensus
sequence. Most kinases phosphorylate their substrates on either serine
or threonine and less often on the amino acid tyrosine. For this reason,
kinases are called either serine/threonine kinases or tyrosine kinases.
Similarly, phosphatases are divided into either serine/threonine phosphatases
or tyrosine phosphatases.
Kinases and phosphatases themselves may be regulated by phosphorylation.
As with other proteins, phosphorylation of specific amino acid residues
within these enzymes leads to a conformational change that changes their
enzymatic activity. For example, the arrival of a chemical signal at the
outer membrane of a cell often leads to the phosphorylation of a kinase
on the inner side of the membrane. Once phosphorylated, the activated kinase
phosphorylates another kinase, which targets its own set of proteins, potentially
additional kinases. In this way, the original signal at the outer cell
membrane initiates an intracellular cascade that amplifies the initial
signal many times, i.e., many more of the final molecules in the pathway
are activated than would be without such amplification. An important example
of this type of signal amplification is the mitogen-activated protein
kinases (MAPK) in neurons involved in learning and memory.
Kinases and phosphatases are involved in the biological functions of most
cells of the body, including the CNS. The birth of neurons, their migration
to specific target regions, the growth of axons and dendritic arbors, and
the formation of synaptic connections are all governed by distinct phosphorylation
pathways. Signals generated at the surface of a neuron find their way through
the cell membrane and may either activate a cytoplasmic protein (such as
an enzyme) or lead to gene transcription in the nucleus. The process of
turning an extracellular signal into a intracellular signal is called signal
transduction. This process may be initiated by an action potential, neurotransmitters,
growth factors, or other chemicals that bind to specific receptors. When
the end result of such a pathway is the phosphorylation of specific subsets
of transcription factors, specific genes are transcribed. The protein products
of the genes being expressed are those needed for the cell to function
properly at that moment. In this way, different signal transduction cascades
may instruct progenitor cells to divide or differentiate, or instruct a
mature neuron to strengthen particular synapses through synaptic modifications
that accompany aspects of learning and memory.
One central player in an emerging model of learning and memory is a signal
transduction pathway known as the ERK MAPK pathway (Fig.
1). The mitogen-activated
protein kinase (MAPK) cascades, as the name implies, are critical for the
regulation of cell division and differentiation during nervous system development.
The extracellular-signal regulated kinase (ERK) cascade is the prototypical
MAPK pathway. The ERK cascade is representative of a family of signaling
molecules that share a specific motif, three serially linked kinases regulating
each other by sequential phosphorylation. The two ERK MAPK isoforms, named
for their molecular weights as p44 MAPK and p42 MAPK, show a striking similarity
to one another at the amino acid level and are referred to as ERK-1 and
ERK-2 or ERK1/2.
The ERKs are abundantly expressed in neurons in the mature CNS, raising
the question of why the prototype molecular regulators of cell division
and differentiation are present in these nondividing, terminally differentiated
neurons. One theory that has emerged is that the ERK signaling system has
been co-opted in mature neurons to function in the synaptic plasticity
that is required for memory. Similar to their role in nonneuronal cells,
these molecules serve as biochemical signal integrators to coordinate responses
to extracellular signals. We will highlight a few of the essential details
of how the MAPK are regulated.
The MAPK were introduced previously when we discussed how growth factors
signal to cells. When nerve growth factor is released from a neuron, it
diffuses across the synaptic cleft to bind to a specific receptor on an
adjacent neuron. Growth factor binding induces a conformational change
in the receptor that activates the intracellular portion of the receptor.
This type of receptor, designated as R1 in the accompanying figure, is
a tyrosine kinase receptor. When growth factor binds to a tyrosine kinase
receptor, several tyrosine residues on the intracellular portion of the
receptor become phosphorylated.
The phosphorylation on tyrosine residues has an important effect. These
phosphorylated residues act like a magnet, pulling several proteins out
of the cytoplasm that initiate the MAPK signaling cascade. Grb, a small
adapter protein, is the first protein to join the complex. It binds to
the tyrosine phosphorylated residues, but has no intrinsic enzymatic activity
itself. Instead, it possesses several amino acid sequences, or motifs,
that function only to bind to other proteins and assemble them at the membrane.
One end of the Grb protein binds to phosphorylated tyrosine residues on
the growth factor receptor, and another portion binds to a second protein
called SOS.
SOS
activates ras, a very important signaling molecule located nearby on the
cell membrane. Ras belongs to a family of low-molecular-weight guanine
nucleotide binding proteins. This type of protein is activated by adding
phosphate groups. So far, when we’ve spoken of protein phosphorylation,
the phosphate groups were donated by an ATP and became attached to the
target protein through a covalent bond. In contrast, the ras molecule is
activated by binding to GTP, a molecule with properties similar to those
of ATP. GTP binding activates ras and its removal inactivates ras. The
addition of the GTP is facilitated by a group of enzymes called guanine
nucleotide exchange factors; the GTP removal is facilitated by another
family of proteins called GTPase activating proteins, or GAPs. The next
column describes how a mutation in a GAP protein, which is required to
return ras to its basal level of activity, causes neurofibromatosis. Under
these circumstances ras cannot release the GTP molecule and it remains
in a hyperactive state.
Normally, activated ras induces rapid, sequential phosphorylation of a
series of kinases, ending with phosphorylation and activation of ERK1/2.
ERKs, the main subject of this column, are involved in triggering long-term
synaptic changes in the mammalian CNS. The earliest studies in this area
showed that ERK activation plays a role in NMDA receptor-dependent long-term
potentiation (LTP) in hippocampal area CA1. Subsequent studies showed that
ERK activation is necessary for several types of long-term synaptic changes
in the CNS involved in learning and memory. These include LTP in the amygdala,
dentate gyrus, and regions of the cerebral cortex. LTP is believed to be
the physiological basis of many forms of learning and memory; a separate
column in this series will discuss LTP in greater detail. ERK activation
is not universally required for synaptic plasticity in the mammalian brain,
but the majority of long-term synaptic changes appear to depend on ERK
activation as a triggering event.
ERKs are involved in many forms of animal learning, which is not surprising
given the widespread involvement of ERKs in CNS synaptic plasticity. Associative
conditioning, nonassociative conditioning, and various forms of spatial
learning are all subject to disruption by inhibiting ERK activation. Some
examples include fear conditioning, spatial learning in the Morris water
maze, taste learning, and conditioned taste avoidance. These disparate
learning paradigms in a variety of species and several brain regions, such
as the amygdala, hippocampus, and insular cortex, suggest a highly conserved
role for this mechanism in learning processes through evolution.
How
did investigators learn that ERKs are involved in learning and memory?
This is the subject of the next column. The question, however, brings us
back to our opening discussion—the importance of genetic engineering
and the use of knock-out mice in investigations of cognitive function.
Identifying the downstream proteins phosphorylated by ERK-1 and ERK-2 makes
it possible to construct knock-out mice lacking one or another of these
proteins. The absence of various members of the ERK signaling cascade have
dramatic effects on the ability of knock-out mice to learn new tasks. As
we will review in the next column, some of these molecules appear to be
mutated in certain human developmental disorders.
top of page
|
