|
Return to Past Workshop
Reports main page
Workshop Reports
Hereditary Disease Foundation
Genetic Models of Treatment and Cure for
Huntington's Disease
January 11 and 12, 1997
Santa Monica, California
Prepared by Mary Elizabeth Bach
Genetic Models of Treatment and Cure for Huntington's Disease
January 11 and 12, 1997
Santa Monica, California
Participants
Mary Elizabeth Bach
Columbia University
Gillian Bates
Guy's and St. Thomas's Hospital
M. Flint Beal
Massachusetts General Hospital
Marie-Fran‡oise Chesselet
University of California Los Angeles
Don Cleveland
University of California San Diego
Robert C. Collins
University of California Los Angeles
Stephen Davies
University College London
Peter Detloff
University of Alabama Birmingham
Allison Doupe
University of California San Francisco
Stephen B. Dunnett
University of Cambridge
Robert H. Edwards
University of California San Francisco
Argiris Efstratiadis
Columbia University
Kenneth H. Fischbeck
University of Pennsylvania
Susan Hockfield
Yale University
J. Graeme Hodgson
University of British Columbia
Carlo Iannicola
Stanford University
Greg E. Lemke
The Salk Institute
Jose Lucas
Columbia University
Marcy E. MacDonald
Massachusetts General Hospital
John C. Mazziotta
University of California Los Angeles
Edward R.B. McCabe
University of California Los Angeles
Paul H. Patterson
California Institute of Technology
Christopher Ross
Johns Hopkins University
Ira Shoulson
University of Rochester
Ethan Signer
Massachusetts Institute of Technology
Larry W. Swanson
University of Southern California
Danilo Tagle
National Institutes of Health
Allan Tobin
University of California Los Angeles
Nancy Wexler
Columbia University
Jacqueline K. White
Massachusetts General Hospital
Charles Wilson
University of Tennessee Memphis
Anne B. Young
Massachusetts General Hospital
Scott O. Zeitlin
Columbia University
S. Lawrence Zipursky
University of California Los Angeles
Genetic Models of Treatment and Cure for Huntington's Disease
January 11 and 12, 1997
Santa Monica, California
Transplant Therapy
The workshop opened with the introduction of two people with Huntington's
disease (HD) and a focus on fetal cell transplant surgery therapy. One person had
undergone the transplant surgery whereas the other person was considering it.
Both the individual who had had the surgery and his wife agreed that the
behavioral symptoms of HD had been ameliorated by the surgery although there
was no effect on his motor symptoms. The person's wife pointed out though that
the behavioral improvement had begun to occur prior to the surgery during pre-
surgery testing suggestive of a placebo effect. The family was wrestling with
whether a modest quality of life improvement was worth the cost and potential risk
of the surgery. The surgery is expensive and not covered by insurance. There was
also discussion of the critical necessity of pre and post implantation assessments
according to the HD-CAPIT internationally agreed upon guidelines. Regular follow-
up examinations to assess short and long term effects are critical to compare
before and after functions.
Many participants wrote in their workshop summaries that meeting the people
and their families inspired them and kept them aware of the big picture and the
implications of their own work in the search for a cure.
Overview of Genetic Mouse Models of HD
The genetic mouse models of HD were discussed next. A wide variety of
genetic approaches are being employed to generate the mouse models including
knock-in and transgenic strategies. The importance of phenotype reproducibility
was discussed at great length as reproducibility across these models may reveal
molecular mechanisms that will be vital in the search for a cure. In the knock-in
models, one issue that might affect phenotype reproducibility is the loxP sites that
are left behind. To date, in all the models studied, the loxP site does not appear to
affect expression. Several issues that might affect phenotype reproducibility in the
transgenic models were discussed. In the transgenic models both cDNA and YAC
constructs under the control of different promoters (CMV, Hdh or HPRT) have been
employed. The site of integration varies across models and this may affect
reproducibility. Second, the models differ in terms of whether a full length or
truncated transgene was inserted. Third, the promoter controlling transgene
expression has varied across the models. Lastly, an issue that might affect
phenotype reproducibility in both knock-in and transgenic models is strain
background.
Behavioral phenotype reproducibility will also be important in the search for a
cure. To date in many of the models, both transgenic and knock-in, behavioral
phenotypes have been observed. In many, movement disorders were exhibited
(tremors, head twitches, and stereotypic grooming) although differences between
phenotypes have also been observed (breeding difficulties, seizures, both weight
loss and gain and both an increase and decrease in reactivity and activity). There
is currently no consensus in terms of which aspects of the phenotypes should be
targeted for therapeutic approaches. This may become clear with the emergence
of ubiquitous phenotypes in and across current and future models and a more in
depth analysis of these models.
Lastly, many of the researchers generating the mouse models commented that
the cost of maintaining their colonies is large and hinders their progress in a search
for a cure. Assistance from the Foundation was requested.
Specific Models
Carlo Iannicola representing Rick Myers laboratory
Rick Myers' laboratory has generated transgenic models carrying 39 out of 67
exons and 67 CAG repeats under the control of either the CMV or Hdh promoter.
The protein expression levels of the transgene were analyzed in kidney, liver and
brain tissues and were found to vary. Further, expression under the CMV promoter
was found to be higher than under the Hdh promoter. Dr. Iannicola employed
Jean-Louis Mandel's antibody that recognizes proteins with long stretches of
glutamines.
A behavioral phenotype was observed at 5 months in lines generated with both
promoters (CMV:1 line, Hdh:2/4 lines). Specifically, "explosive" jerking and
twitching was observed in addition to excessive grooming and scratching. No
weight loss was observed in either the founders or F1s and both feed normally.
According to Dr. Iannicola, at 8 months, a MRI examination (which has been
found to be accurate in mice) was done at Good Samaritan Hospital and revealed
in mutants an increase in lateral ventricle volume (5-10 fold). Dr. Iannicola pointed
out that this is interesting given that people with HD also exhibit an increase in
ventricle volume. This increase was observed in mutants generated with either the
Hdh and CMV promoter. Dr. Iannicola said further MRI studies are necessary to
validate these findings.
Dr. Iannicola's laboratory has also generated YAC transgenics with 67 CAG
repeats. This work is in the early stages and no other information was available.
Lastly, Dr. Iannicola discussed the progress of knock-in mice currently being
generated in the laboratory by Peggy Shelbourne. A human exon containing 70-80
repeats replaced mouse exon 1. A behavioral phenotype emerged at 6 months
Specifically, hyperactivity and movement disorders were exhibited. In a subset of
the knock-in mice sudden "back flips" were observed although this may be
consistent with vestibular damage.
Jaqueline White and Marcy MacDonald
Drs. White and MacDonald have generated a knock-in mouse model with 20,
48, 89 and >120 repeats. The first 17 amino acids of exon 1 from the mouse were
retained, while the remainder of the exon consists of human sequence with the
expanded CAG repeat. A neo cassette, flanked by bacteriophage P1 loxP site-
specific recombination sequences, was inserted 950bp upstream of the ATG. In
mice with the neo cassette present, no expression was detected. Knock-in mice
were also generated in which the neo cassette was excised in the presence of cre
recombinase by site-specific recombinase. To remove the neo cassette the knock-
in mice were mated with a cre transgenic mouse that expresses the recombinase
early in development (from Gail Martin's laboratory). Following the removal of the
neo cassette, normal amounts of protein containing the expanded repeat were
obtained.
A behavioral and neuropathological examination of the mice with 48 repeats has
been done. The oldest are 8.5 months and exhibit no behavioral phenotype and
feed and breed normally. The neuropathological examination was done (at 4.5
months) and no abnormalties were revealed.
Scott Zeitlin and Argiris Efstratiadis
Drs. Zeitlin and Efstratiadis have generated knock-in models with 55, 71 and 94
CAG repeats inserted in exon 1 by replacing a portion of the mouse exon
(extending from a conserved Xmm I restriction site to the end of the Exon) with
human sequence from a person with juvenile onset of HD containing the expanded
CAG and polyproline repeats. The 5' splice site of the first intron was re-created
with linker sequence and approximately 100 bp of the intron was deleted up to an
internal Kpn I site.
In lines with 94 CAG repeats, the neo cassette, flanked by loxP sites, was
inserted 600bp upstream of the transcription initiation site. Five ES lines were
generated (3 using CCE 33ES cells and 2 using W9.5ES cells). In lines with either
55 or 71 CAG repeats the neo cassette was inserted 1.3kb upstream of the cap
site. The targeted allele expression in the line with 94 CAG repeats was
approximately 33-50% of the wild-type allele.
In the lines with either 55 or 71 CAG repeats, expression levels from the
targeted allele were similar to the wild-type allele. The neo cassette was removed
by transient expression of Cre recombinase in the ES cell (targeted ES cells were
transfected with a supercoiled form of 1:50). Expression levels of the targeted
alleles (measured in ES cells) were unaffected by removal of the neo cassette. Dr.
Zeitlin noted that the discrepancies in the expression levels reported by various
laboratories generating knock-in mice may be related to the different positions of
the neo cassettes. Although homology comparisons suggest that the core
promoter is within 250bp of the cap site, the promoter is not mapped functionally
and additional promoter elements may be located at or near some of the neo
insertion sites.
Dr. Zeitlin is currently generating mice from ES clones in which the neo cassette
has been deleted by cre/lox recombination (Several participants at the meeting
requested more detailed information on the cre/loxP system which Dr. Zeitlin
provided. It can be found at the end of this report). Analysis of knock-in mice
lacking the neo cassette should clarify the discrepancies in the expression levels
observed in different laboratories and perhaps the differences in behavioral
phenotypes.
A behavioral phenotype was observed in the lines with 94 CAG repeats and was
found to correlate with the degree of chimerism. In mice that were close to 100%
chimeric, stillborns were observed.
In mice that were 70-50% chimeric, tremors and weight loss emerged at about 1
week of age. These mice had difficulty moving and rigidity in their limbs. Organ
atrophy was also observed and death occurred around 2 weeks of age. Marie
Fran‡oise Chesselet did the neuropathological examination and found general
brain atrophy although no other neuropathology was observed. Dr. Chesselet
pointed out that she did not have the appropriate controls due to the spontaneous
deaths of the knock in mice.
In mice that were 50-30% chimeric, slight tremors and weight loss were initially
observed, though by 3 weeks of age the mice appear to recover and have since
gained weight and exhibit no tremors.
Gillian Bates
Dr. Bates and her group have generated transgenic mice using exon 1 of the
HD gene carrying an expanded CAG repeat (approximately 115 to 150). From a
single founder with 5 integration events, 4 separate lines were generated. Single
copy integrants were obtained in 2 lines (R6/0-142 repeats; R6/1-116 repeats).
One line was found to have 3 copy integrants (R6/2- 145 repeats) whereas the last
line had 6 (R6/5-130-156 repeats). RT-PCR analysis revealed that the transgene
was expressed ubiquitously. Expression levels were variable but in some lines
similar to endogenous expression.
A behavioral phenotype characterized by tremors, seizures and weight loss was
exhibited in various degrees in lines R6/2, R6/1 and R6/5. The mice may also be
incontinent although an alternative interpretation is that they are drinking a greater
volume of water. A discussion ensued regarding whether the weight loss observed
in these mice was due to a failure to gain or weight loss. The questions
surrounding the weight and incontinence issues may be answered by metabolic
studies which are currently being done. Dr. Wexler noted that metabolic studies
were done in people with HD. The calorie intake of the people was found to
exceed that of their matched spouse. Further, the extra intake was found to be
due to the increase in motor activity observed in HD and not a metabolic change.
Spontaneous deaths have also been observed and may be due to seizures.
The mice were analyzed for myelin defects given that transgenic models with this
defect have been found to have both tremors and seizures. The CNS myelin
looked normal. Cardiac defects were also suggested as a cause of seizures. No
cardiac examination has been done to date.
Fertility is compromised in the transgenics as they are capable of breeding for
only a short period of time. Consistent with this, reproductive organ atrophy
occurs. Breeding difficulties may have been compounded in these mice by the fact
the genetic background was C57 and Dr. Bates and other laboratories have found
this strain ceases to breed across generations.
Stephen Davies
Dr. Davies has done the neuropathological examination of Dr. Bates' mice with
the bulk of the analysis occurring on the R/2 line. The transgenics brains were
found to be 16-20% smaller. By contrast their body weight is 60-70% of wild-types.
Only two significant differences were observed between the transgenics and wild-
types. First, shrinkage of cerebellum purkinje cell bodies was detected in
transgenics. Second, in a subset (25%) of the transgenics that Dr. Bates
categorized as "end-state," focal astrocytosis was observed in the dorsal striatum.
Immunocytochemistry revealed no difference between transgenic and wild-type
mice across neurotransmitter systems, receptor density or in the spinal cord. No
difference was observed in the transcription factors jun, cfos, or fos B. One
hinderance to these studies that Dr. Davis noted is that many of the commonly
used rodent biological cell markers may not work in mice. Also little is known
regarding possible mouse strain differences in these markers or neuropathology.
Dr. Beal pointed out that strain related differences in vulnerability to MPTP
neurotoxicity have been observed.
The striking contrast between the behavioral phenotype and the results of the
neuropathological examination is inconsistent with the neuropathology observed in
people with HD. A discussion ensued regarding whether the observed phenotype
correlated with cerebellum purkinje cell atrophy and not any functional or structural
striatial defects. Dr. Chesselet pointed out that this disparity may be a function of
the fact that the mice do not live long enough to develop the neuropathology
observed in late stage people with HD. Also little is known regarding the biological
markers during the early stage of HD. Lastly, the idea was entertained that
phenotypes observed in the mouse may not mimic those observed in HD.
Peter Detloff
Dr. Detloff's laboratory has generated several different knock-in models
(129/C57 background) in which CAG repeats of 39, 70, and 150 were inserted
within exon 3 of the HPRT gene. For controls, a construct with a stop codon
followed by 150 repeats was prepared. A targeting vector with CAG repeats
encoding Ala was also constructed. To date, the expression level of the mutated
HPRT protein is not known.
A behavioral phenotype has been observed at 24 wks in male mice with 150
CAG repeats (females are mosaic for the expanded repeat and show no
phenotype). The males exhibit handling induced seizures. EEG recordings in the
males are consistent with the observation of behavioral seizures. Tremors may
also be occurring but only in a small subset of the mutants. In contrast to Dr.
Bates' mice, Dr. Detloff's knock-in mice are heavier, weighting approximately 10g
more than wild-types. Motor activity and strength was assessed using an activity
monitor and a test of forearm strength. The activity monitor revealed that mutant
ambulatory movements were normal. By contrast, the mutants were found to rear
up (vertical movements) significantly less. Forearm strength was significantly
impaired in the mutants as demonstrated by their inability to hang from a bar.
Graeme Hodgson
Dr. Hodgson in Michael Hayden's laboratory has generated YAC transgenics
with 18 and 48 CAG repeats. Expression has been measured in both the brain
and testes. A behavioral phenotype was observed and correlates with the length
of the CAG repeats. No behavioral phenotype has been observed in the YAC
transgenics with 18 repeats. By contrast, the transgenics with 48 repeats were
observed to exhibit head twitches, reactivity, an increase in both activity and
excessive grooming. Breeding and weight appear to be normal.
Dr. Hodgson also discussed the transgenic mice generated in Michael Hayden's
laboratory that have 15, 44, 80 and 128 CAG repeats under control of the CMV,
HPRT or the endogenous Hdh promoter.
Lastly, Dr. Hodgson reported that Jamal Nasir is beginning work on a knock-in
model that has 48 CAG repeats.
Danilo Tagle
Dr. Tagle's laboratory has generated several mouse transgenic lines using full
length HD cDNA (all 67 exons) constructs expressed using the CMV promoter.
The parent full length plasmid containing 16 CAG repeats was initially expressed in
the baculovirus system. 48 and 89 repeats were each introduced into this
construct and recloned into pCDNA1.1. To date, 11 founders of various ages have
been produced; 5 from the construct with 16 repeats, 3 from 48 repeats and 3
from 89 repeats. The transgene has integrated at 1-2 sites although variable copy
numbers have been observed (as many as 10 copies in a couple of lines).
Expression seems to be ubiquitous in several tissues examined as per an antibody
from Rick Myer's laboratory (the HD1 antibody raised against the first 17 amino
acids of the HD protein). Currently there are no data regarding the relative amount
of transgene protein expressed versus the amount of mouse endogenous Hdh
expression.
A behavioral phenotype has been observed at 2-3 months. Specifically, head
twitching, tremors, stereotypic grooming, a decrease in exploring, activity and
startle threshold were all observed. The transgenics were also found to be
defective at hanging on to a wire mesh, a measure of fore arm strength. Weight
and feeding appears to be normal. No phenotype was observed in the line with 16
repeats. Lastly, seizures were not observed in any of the lines.
Kurt Fischbeck
Dr. Fischbeck first described transgenic mice generated in Dr. Mandel's
laboratory. In these mice a full length mouse cDNA with 73 CAG repeats was
expressed under control of the CMV promoter. The expression levels were found
to be low (1/10 of endogenous allele). A behavioral phenotype (agraphobia) was
initially observed but is no longer.
In Dr. Fischbeck's laboratory, Diane Merry has generated mice that contain
androgen receptor cDNA with expanded repeats. Two different promoters were
employed, the neurofilament light chain promoter and the neuron-specific enolase
promoter. A behavioral phenotype was observed. The mice were observed to be
stronger and more aggressive. At 20 months of age proximal weakness in the
limbs was observed. Consistent with this, atrophy of the muscles was detected.
Death occurs around 22 months.
Jose J. Lucas from Rene Hen's laboratory
Dr. Lucas is currently involved in generating a mouse model in which the
mutated Hdh gene will be regulated via the tetracycline transactivator system (tTA)
(see Mayford, Bach,..Kandel, 1996, Science, ). The first exon of the mouse Hdh
protein with a human 94 CAG repeat will be inserted and gene expression will be
limited to the basal forebrain (including the striatum). To achieve regulated
expression of the Hdh gene, two types of mice need to be generated. In the first
type of mouse, the tTA gene is expressed under the control of the CaMKIIa
promoter, which limits expression of the tTA transgene to neurons of the forebrain.
In the second type of mouse, the tTA-responsive tet-O promoter is linked to the
Hdh gene. The tTA gene expresses a eukaryotic transcription activator that binds
to and activates transcription from the tet-O promoter element; this transcription is
blocked by the tetracycline analog doxycycline. When both the tet-O and tTA
transgenes are introduced into the same mouse, the tet-O -linked Hdh gene will be
activated but only in those cells that express tTA (those in the forebrain). To
determine where in the brain, the expression of the expanded CAG repeat has
been transactivated Dr. Lucas is also planning to drive expression of a lacZ
reporter cassette with the tTA system. Dr. Lucas feels that using the tTA system
can be valuable as it will help elucidate whether expression of the CAG repeat is
enough to trigger the disease or if the degeneration seen in neurons in humans is
due to the ubiquitous low level of the mutated protein in the entire brain.
The second approach that Dr. Lucas's laboratory is undertaking involves genetic
and pharmacological manipulations of the serotonin 5-HT1B receptor. The genetic
manipulation involves using the tTA system to drive expression of either Cre
recombinase or the serontonergic 5-HT1B receptor. The pharmacological
manipulations of the 5-HT1B receptor will determine whether this approach is
useful to alleviate or postpone some of the motor and psychiatric manifestations of
HD. Dr. Lucas noted that the 5-HT1B receptor is reduced in HD. This hypothesis
will be tested in the 5-HT1B knock-out mouse generated by Rene Hen. The
receptor is expressed in the caudate putamen and the protein is enriched in the
axon terminals in the globus pallidus and the substantia nigra. In the substantia
nigra, the 5-HT1B receptor is known to modulate GABA release and therefore the
activity of dopaminergic neurons. Dr. Lucas noted that if his hypothesis is true, it is
possible that the knockout background might exacerbate a subtle phenotype in
transgenic lines with a low level of regionally restricted mutant protein expression.
Similarly, it is possible that administration of the few selective agonists that are
available for the 5-HT1B receptor might attenuate the phenotype of some mouse
models.
Behavioral Analysis
Dr. Bach stressed that since many of the transgenic and knock-in models are
exhibiting non-ambulatory motor disruptions (tremors, stereotype grooming, chorea
like movements) quantification of these behaviors is vital. Quantification will allow
progression of the phenotype to be documented. Also through quantification, a
baseline can be generated that can be used to assess the effect of genetic
(antisense, etc. ) or pharmacological treatment. Commercially available
transducers have been employed for many years to measure tremors in rats. Most
recently, Coulbourn Instruments has created a transducer that is sensitive enough
to measure tremors in mice. The benefits of using instrumentation to measure
non-ambulatory motor disruptions include objectivity and reproducibility. Objectivity
becomes especially important given that a blind study will be most difficult to
implement since both subtle and gross phenotypes have been observed.
The second point that Dr. Bach made was that electrophysiological and lesion
studies have elucidated that the dorsal striatum subserves a specific learning and
memory system (Graybiel et al; Dunnett et al; McDonald et al). The caudate area
of the striatum is required for learning to emit a specific response in the presence
of a specific cue. Striatum dependent learning and memory can be assessed
through various measures including simultaneous brightness discrimination, cued
go/no go olfactory discrimination, active avoidance and the cued versions of the
radial arm, Barnes and Morris mazes. The fact that poor performance on a given
task can reveal a defect in a specific anatomical location can prove valuable when
no gross pathology is detected.
If learning and memory studies are implemented, the appropriate controls need
to be used to assure that any impairment observed is not due to performance
defects (motor, sensory, attentional, motivational, etc). This is especially important
given the non-ambulatory motor disruptions observed in many of the mice. Lastly,
the effect of seizures has to be taken into account when interpreting the results of
learning and memory tasks. It is very clear that seizures profoundly impair
hippocampal-dependent learning and memory. Little is known regarding the effect
of seizures on striatum-dependent memory.
Drs. Bates and Detloff made the point that in depth analysis of the mutants was
not their object, rather they sought a general description. Both suggested that the
battery of sensorimotor tasks comprised by Irwin might be valuable to employ
(Irwin, S. , (1968). Psychoparmacolgia, 13, 222-257). Dr. Bates also noted that a
similar battery called SHIRPA is being employed by several groups in the U.K. and
will soon be published on the Harwell web site. Dr. Bach made the point, using the
example of the task which measures forearm strength, that many different
transgenic mice unrelated to the mouse models of HD have been observed to have
forearm weakness on this task as well as aged rodents. Therefore using tasks
subserved by an undefined anatomical locus may generate red herrings which
many not be ameliorated by either genetic or pharmacological treatment because
they are not directly mediated by the mechanism that causes HD.
A discussion ensued regarding which aspects of the phenotypes should be
targeted for therapeutic approaches. Dr. Beal noted that there are currently
several different compounds that can be tested for efficacy in the mouse models.
Dr. Shoulson suggested that in the case of Dr. Bates' mice the dramatic weight
loss observed would serve as an ideal index of phenotype onset and progression
especially given that people with HD also exhibit weight loss (none of the other
transgenic or knock-ins exhibit this loss). Dr. Efstratiadis noted that growth curves
can be generated for this purpose and may prove valuable in deciphering whether
these mice exhibit a failure to gain or weight loss. Dr. Beal mentioned that brain
weight and morphometry might prove valuable. Ultimately, no consensus was
reached in terms of what aspects of the phenotypes should be targeted for
therapeutic approaches. The idea that phenotypes observed in mouse models of
HD may not mimic those observed in HD was raised. It was decided that this may
be come clear with the emergence of future models and more in depth analysis of
current models.
The last point that Dr. Bach raised was the issue of strain-related effects.
Recent work in her laboratory revealed that transgenic mice on a CBA background
showed defects in hippocampal-dependent learning in the wild-type mice that
correlated with defects in LTP. The CBA strain it turns out is plagued with
problems ranging from retina degeneration to severe seizures. The laboratory is
now breeding back on to a C57 background given that this strain exhibits good
learning and memory. Little is known about striatum functioning across strains
other than strain related differences in vulnerability to MPTP neurotoxicity have
been observed. A discussion ensued regarding strain-related affects although no
consensus emerged regarding which strains would be best to employ within the
mouse models of HD.
Cell Biology, Assays and Neurodegeneration
Dr. Signer noted that the cell biology of HD gene function greatly lags behind
the mouse model research and that a cellular model of HD is needed to screen for
possible therapeutic compounds quickly and efficiently. E. Coli, yeast and tissue
culture models were all suggested as possible models of the neuro-degenerative
mechanisms mediated by the HD gene/protein. Dr. Hodgson noted that clues to
these neuro-degenerative mechanisms may be revealed by studying in vitro the
basic biochemistry of the protein and the nature of its interactions with other
proteins. Further, the use of a yeast cellular model may prove to be an extremely
valuable tool when studying some basic cellular functions of the HD protein. Dr.
Swanson noted that the regulation and function of genes and their products are
often different in different tissues and cell types and therefore it is important to gain
as much information about these issues in human striatum neurons. It was also
proposed that the establishment of striatal cultures from the transgenic and knock-
in mice may provide a useful cellular model. It was suggested that the various
laboratories generating these mice forward cells to a centralized laboratory that
would make hybrid system cultures that could be used to systematically screen
(with different markers) possible therapeutic compounds quickly and efficiently. Dr.
Detloff noted that a efficient high output screen will be vital given that
pharmaceutical firms may be less interested in HD given that it affects only a small
population of individuals. One caveat that was raised is that in Alzheimer's disease
much is known at the cellular level but this information has not lead to a cure.
The contribution to the search for a cure from information gained from
neuropathogical studies was discussed next. Dr. Tobin asked what can be learned
from neuropathogical studies of people with HD and what is the best approach. It
was suggested that one valuable approach would be to study people at risk for HD
as little is known regarding what happens early in HD. These studies might
reconcile differences observed between human and mouse phenotypes. Neuronal
cell counts was one method suggested. Given that much money was donated for
brain tissue from people with HD to be stored in a brain bank it was asked what
had been learned from the stored tissue? Another approach that was suggested
was doing biopsies on people undergoing fetal cell transplant therapy.
Genetic Therapy
Many of the participants felt that ultimately HD would not be cured until the
defective gene was replaced. Although this is currently not feasible other potential
gene therapies were noted. Dr. MacDonald suggested an approach that could be
tested in exsisting mouse models of HD involving antisense or complementation by
wild-type gene products. One specific hypothesis that could be tested involves
overexpressing the wild-type gene to determine whether a gene dose effect is
important for the phenotype observed in HD. To test whether the toxic gain of
function in HD can be reversed by overexpression of the wild-type gene a mutant
mouse can be mated to one that overexpresses the normal HD gene. It was also
suggested that further analysis of the mouse models could elucidate whether the
phenotype observed in HD is the result of gain or loss of function and that this
analysis could have important implication for therapeutic approaches. Lastly, Dr.
MacDonald noted that more biochemistry needs to be done to figure out what the
HD protein is doing.
Cre/loxP Summary (as per Dr. Zeitlin)
The generation of developmental and/or tissue-specific deletions in the mouse
genome requires an efficient site-specific recombination system. Many
laboratories have adopted the cre/loxP system derived from bacteriorphage P1
(see Sauer and Henderson, 1990). In the most common use of the system, two 34
bp loxP sequences in the same orientation
[5'-ATAACTTCGTATAATGTATGCTATACGAAGTTATTCGA-3'] are placed flanking
the DNA sequence to be deleted. In the presence of the Cre recombinase
enzyme, the DNA between the two loxP sites is looped out and recombination
occurs within the loxP sequence. Following recombination, one loxP site remains
in the genome and the deleted DNA (circular product) is eventually degraded.
For the purpose of generating conditional mutations, the Cre recombinase gene
is usually placed under the control of a tissue-specific promoter and introduced as
a transgene into the mouse. Standard gene-targeting methods are used to
generate a mouse with the target sequences flanked by loxP sites. Although not
required, many investigators will mate the conditional mutant mice with a mouse
already heterozygous for a null mutation in the target gene to generate progeny
that will have one loxP-modified allele and one null allele. In this case, the Cre
recombinase will only have to recombine one allele to generate the tissue/stage-
specific null mutation (Gu et al., 1994). The loxP-modified mice are now, in turn,
mated with the mice containing the Cre transgene. In the fraction of the progeny
that contain both the transgene and the loxP-modified allele, recombination will
take place in those tissues expressing the transgene.
For the knock-in model mice, the cre/loxP system was used to eliminate the neo
cassette from the targeted allele containing the expanded CAG repeat. Pop-out of
the loxP-flanked neo-cassette, leaving behind one integrated loxP site, can be
accomplished several different ways. First, by mating the mice to a Cre transgenic
mouse that expresses the recombinase early in development (from Gail Martin's
laboratory). The progeny have the neo cassette removed in all cells. Alternatively,
one can eliminate the neo cassette in cell culture by transient transfection of a Cre
expression plasmid in hte ES cell. Without selection, between 1-5% of the cells
undergo recombinations.
References:
Gu, H. et al., (1994). Science, 265, 103-106
Sauer, B. and Henderson, N., (1990). New Biologist, 2, 441-449.
Schwenk, F., et al., (1995). Nucleic Acids Res., 23, 5080-5081. |
|
