BEHAVIOURAL NEUROGENETICS
Genes may
ultimately control both simple and complex behaviour, but it is less clear how
this happens. Therefore if we consider a mechanistic pathway:
gene > biochemical mechanisms >
behaviour
The biochemical
mechanisms which are the integrators of behaviour are a mixture of primary arid
secondary effects of the causal genes. Therefore biochemical mechanism on their
own or biochemical abnormalities where behavioural abnormalities are present
may not give clear indications of cause and effect.
In contrast,
fairly complex sets of behaviour may be analysed and understood if the genetic
basis is understood, since characterisation of the genetic cause will elucidate
primary biochemical mechanisms.
Here we will deal with some of the more clearly
characerised genetic mechanisms which underlie simple behaviour patterns. One
classical system for studying these has been the fly Drosophila.
Several paradigms
have been used for the study of memory and learning in higher organisms, one
being associative conditioning. In Drosophila,
an example of this paradigm is the linking of an attractant odour with a
negative stimulus (an electric shock if the fly comes towards the odour). The
fly will learn to associate the two and avoid the odour.
Several single
gene mutations which disturb memory function have been detected using these
paradigms (dunce, amnesiac, cabbage, turnip, rutabaga). Interestingly, other
behaviour which may be assumed to be memory related is also affected in these
mutants such as that associated with courtship. Courtship conditioning is a
behaviour whereby males avoid mating again for about four hours after mating
with a female. This is thought to reduce the number of unproductive matings in
the population. The mutant cabbage, however,
immediately seeks another mate, as if he has forgotten what has just occurred.
Similarly, acoustic sensitisation, whereby a female is made receptive to males
after hearring their courtship song, is also significantly reduced in memory
mutants.
The biochemical
systems associated with these mutations have been studied, and an important one
appears to be the cyclic AMP second messenger system. Thus, dunce mutations are
associated with reduced levels of cAMP phosphodiesterase., and rutabaga with
adenyl cyclase, both important enzymes in the control of levels of cAMP.
ATP
à Protein Kinase à AMP
cAMP Dunce
cAMP itself is
part of the cells second messenger system, which is presumed to act as an
integrator between membrane signalling events and longer term changes within
the cell via protein phosphorylation. It seerns, therefore that cAMP is an
important pathway in the translation of short term events into longer term
memory. Less clear, however, is how two mutations with opposite effects on cAMP
levels are able to produce the same effect on learning. Physiological experiments
on muscle junctions in normal and mutant flies has shown, however, that cAMP appears
to regulate at the properties least one type of K+ channel directly, and may
also modulate synaptic behaviour through an interaction with Ca++ dependent
mechanisms. There appears therefore to be two ways in which neuronal plasticity
may be altered; firstly, through the normal intracellular signalling pathway
and, secondly, through the disruption of normal modulation of synaptic
function.
Circadian
rhythms, a daily cycle of activity, are exhibited by nearly all higher animals.
They comprise an intrinsic biological clock, which is also fine-tuned by
entrainment mechanisms related to light-dark cycles etc. Rhythms are easily
followed by plotting activity-time graphs. In Drosophila, whilst the normal
circadian cycle is 24 hours, there are penod
mutants which have either altered or missing circadian
activity.
STRAIN RHYTHM (h)
wild-type 24
dunce 24
per' 19.5
per’ 28
per' arrhythmic
The ease by which
drosophila genetics can be studied has made it a popular system to identify genes
from causal mutations. In particular, the existence of giant polytene
chromosomes (many copies of one chromosome aligned together) in the salivary
gland enables firstly a very accurate means of visualising deletion/translocation
mutants, and secondly provides material to microdissect the relevant area of a
particular chromosome to produce a genomic library of that region. Very elegant
work by Hall and colleagues eventually cloned the per gene after microdissection of the per region into large cosmid clones. The relevant regions were
further defined by subcloning portions of thee cones and attempting ~element mediated
germ-line transformation of per' and per flies. It as found that constructs
containing a particular region from which a 4.5 kb mRNA was transcribed
appeared to be all that was necessary for rescue of mutant flies to full
rhythmicity. This region was also within the boundaries of known deletions
which produce per' flies- i.e. this
defines the practical limits of the region where deletions disable the per gene.
The gene
producing the 4.Skb message has therefore been accepted as the per gene, but a number of questions were
subsequently asked about its function:
Does the gene
simply define structural circuitry or does it represent genuine dynamic
control? (i.e. is it necessary to build a circadian mechanism), This question
was answered very elegantly by the use of germ-line transformation of per' flies by a functional per+ gene, but putting this gene under a
conditional promotor (the heat-shock promotor). Expression of per in these flies is therefore
controllable. Transformed flies were then subjected to heat-shock at various
stages in development to induce expression. Rescue of circadian activity as
obtained even if expression was only induced in the adult. Thus the gene
appears to play a purely "physiological" role, and not a
developmental one.
Does the level of
per cycle? Early work suggested that,
surprisingly, the level of 4.5kb transcript was constant throughout the cycle.
This was later fond to be due to a technical error, and later studies demonstrated
that the level of the 4.5kb transcript did, in fact, fluctuate in a cyclical
manner, maximal levels being at the end of the light part of the day, and
minimum levels at the end of the dark portion. Moreover, per' flies had a transcriptional cycle shorter than and per flies
longer than 24 h.
What effect do
absolute levels of the transcript have? Since per' flies often have deletions of the whole gene, it may be
assumed that activity of the gene is necessary for rhythmicity, but the
question of whether more subtle changes in the gene activity can influence
cycle times was open to question. This was approached by further transfection
experiments. Rescue of arrhythmicity by transfection of the germ line produces
both varying cycle times and varying degrees of expression of per mRNA. Levels of per mRNA were shown to be inversely proportional to the cycle time
in transformed flies. These data suggest that it is activity of the per protein that defines cycle times,
and that point mutations of the per protein
which result in cycle ranges probably arise due to resulting hyper- (per) and the activity of the protein.
All these
observations suggest that there is a daily feedback loop in which per itself affects the oscillations of
its own mRNA.
What is the
mechanism of cycling?
Is the level of per regulated by transcription or
breakdown of per mRNA? This is an
important question, since they would entail entirely different mechanisms:
This has been
answered very elegantly using germ-line transfection of flies with constructs
containing the per prornotor region and a reporter gene (Choline acetyl
transferase, CAT). Essentially, reporter gene levels fluctuated cyclically
(mirroring changes in normal per fluctuations)
in response to sequences 1310 bp downstream of the normal per gene. Thus per trnscription
probably fluctuates in a cyclical manner in response to these elements. This
appears to be all that is necessary for oscillations of the mRNA level, since
breakdown rates of the mRNA are constantly high.
The love song of
Drosophila males is produced by wing vibration and consists of two components,
courtship hums and a series of pulses with interpulse intervals (IPIs). The
mean IPI of Drosophila Melanogaster strains falls between 30 and 40ms, whereas
strains of Drosophila Simulans have a mean IPI of 40-80ms. The IPI itself
varies cyclically. This cyclical fluctuation is under genetic control with a
period of about 55s in D. Melanogaster and about 355 in D. Simulans. Moreover,
the song cycle also appears to be controlled by per, per', strains having short song cycles, per' having long cycles, and per'
having at best weak cycles. The effect of per on IPIs therefore mirrors its effects on circadian cycles, and
the courtship song has therefore been used as another measure of per function.
Although the per gene has been cloned and sequenced
for some time, not much is known about its function in the cell. It is a
proteoglycan protein, but this gives us no clues as to its function. A large
number of tissues express the per protein.
The intron/exon structure is know, as well as the position of the mutations
affecting cycle times. One interesting feature of the protein is the existence
of a block of threonine/glycine (T-G) repeats within exon 5. One interesting
feature of the protein is the existence of a block of threonine/glycine (T-G)
repeats within exon 5. There is a homologous region in the frq gene of the
fungus Neurospora, which also exhibits a daily rhythm. Mutations in frq are
also able to cause long and short rhythms as well as arrhythinicity. The region
of T-G repeats is also implicated as important by the observation that it is
one of the regions which is conserved between different species of Drosophila.
The T-G region does, however, exhibit length polymorphism in both different
strains of Drosophila Melanogaster as well as different species such as D.
Simulans.
Further analysis
of this region has been achieved by further transformation experiments, which
are summarised below.
transfecton with per+
per >
24hr circadian rhythm,
60s song rhythm
transfecton with per +
per > 24hr circadian rhythm,
with T-G region removed 30s song rhythm
The T-G region
therefore appears to be important for song- rhythm, abut not for circadian
rhythm.
Further
experiments have used hybrid genes from different species in transformation
experiments with per flies. Thus construct have been made with either the D.
Simulins per containing D. Melanogaster T-G region, or the the D. Melanogaster
per gene containing D. Simulins T-G region, the results of which are shown
below:
The T-G region
therefore also appears to specify song rhythms between species. The mechanism
of this, despite some very elegant molecular biological experiments, remains to
be clarified.
Per does not
appear to regulate development and metabolism (it is not a member of the
oncogene family!). As such, it acts as a prototype molecule of a molecule whose
function appears to be to regulate behaviour solely). Proteins which are
related to per include DNA binding proteins, though their DNA binding motif is
not shared. However, it is clear that per can affect its own transciption level
at least indirectly.
Circadian rhythms
in higher animals also appear to be due to the same mechanism of an intrinsic
(genetic) clock combined with an entrainment mechanism. In the hamster there is
a defined stain (T) with a shortened
intrinsic activity period (r/+ =22hr, T/T
--20hr).
The biological
site of the pacemaker region has been shown to be the Suprachiasmatic nucleus
of the hypothalamus by elegant experiments, whereby ablation of the SCN
abolishes circadian activity.. That this region is also the site of action of
the intrinsic clock is indicated by the observation that hamsers with ablated
SCN region can have their circadian rhythm restored by trasplantation of fresh
SCN cells, but that the rhythm depends on the strain of hamster from which the
donor transplant was taken. This establishes conclusively that that the donor
tissue contains its own intrinsic pacemaker, and does not act simply by
recoupling existing oscillators. Is a mammalian homologue of per involved in this control? There may
be a mammalian counterpart to the Drosophila gene, but it has proved refractory
to analyse so far . The identification of such mammalian genes may be important
in psychiatry, since disorders such as manic-depression have been shown to
disrupt the normal activity cycle of the patient,