|
Return to Past Workshop
Reports main page
Workshop Reports
Hereditary Disease Foundation
Advances in Huntington's Disease Genetics
October 29, 1996
San Francisco, California
Prepared by Leslie Thompson
Advances in Huntington's Disease Genetics
October 29, 1996
San Francisco, California
Participants
Wojtek Auerbach
New York University
Gillian Bates
UMDS-Guy's Hospital, UK
Brian Brooks
University of Pennsylvania
P. Michael Conneally
Indiana Universityr>
Didier Devys
Institute of Genetic Biology
Peter Faber
Massachusetts General Hospital
Michael Hayden
University of British Columbia
Graeme Hodgson
University of British Columbia
David Housman
Massachusetts Institute of Technology
Carlo Iannicola
Stanford University
George Jackson
University of California Los Angeles
Lesley Jones
University of Wales
Christopher Karlovich
Stanford University
Esther Leeflang
University of Southern Californiar>
Marcy MacDonald
Massachusetts General Hospital
Edward McCabe
University of California Los Angeles
Diane Merry
University of Pennsylvania
Richard Myers
Stanford University
Jamal Nasir
University of British Columbia
Harry Orr
University of Minnesota
Hemachandra Reddy
National Institutes of Health
Christopher Ross
Johns Hopkins University
Russell Snell
University of Wales
Danilo Tagle
National Institutes of Health
Leslie Thompson
University of California Irvine
Allan Tobin
University of California Los Angeles
Martine Usdin
Stanford University
Gert-Jan Van Ommen
Leiden Universityr>
Nancy S. Wexler
Columbia University
Huda Zoghbi
Baylor College of Medicine
Advances In Huntington's Disease Genetics
October 29, 1996
San Francisco, California
Animal models were the major focus of this meeting with one of the highlights of the day being
the exciting presentation by Dr. Gillian Bates. A wide variety of transgenic mice are being
produced, utilizing cDNA and YAC based constructs as well as targeted "knock-in"
approaches. One of the themes that emerged was the importance of developing a wide
range of animal models containing both truncated and full length receptors concurrently to truly
understand the process of HD neuropathogenesis.
This report is divided into three sections: 1) an overview of the mouse models and general
topics discussed, 2) a detailed description of meeting presentations and discussions and 3) a
composite chart containing descriptions of all mouse models presented.
OVERVIEW
Many transgenic mice have been produced containing different portions of the HD gene with
varying size expanded repeats driven by a variety of promoters. Using a genomic clone
encompassing promoter elements and exon 1 of the HD gene and carrying very large CAG
repeats (115-155), Gill Bates has been able to produce mice with a severe neurological
phenotype resembling HD. Mice manifest a pronounced, resting tremor and a choreic type
movement disorder as well as handling induced seizures and unusual grooming abnormalities.
Fertility is compromised by the expanded repeats and neuropathology reveals smaller brain
size in transgenic mice as well as pathologic changes associated with early HD. Rick Myers
lab (Carlo Iannicola) has also generated cDNA transgenic mice carrying 39 exons (coding for
1676 amino acids) and 67 CAG repeats expressed under either a CMV or HD promoter. A
phenotype similar to that observed by Gill Bates becomes manifested; an explosive twitching
movement disorder and excessive scratching and grooming is observed in these mice.
Michael Haydens lab described a variety of mice generated using full length and truncated
(1955 bp, intermediate size between Dr. Bates and Dr. Myers transgenic constructs) cDNA
constructs with different numbers of repeats driven by CMV, HPRT or the HD promoter. Mice
carrying 44 repeats had recently become available and by 6 months, there is no obvious
phenotype. All three labs have looked at protein expression and find widespread expression in
the tissues tested. Transgene expression levels have been compared to endogenous levels
and found to be similar (Hayden), slightly reduced (Bates) or not yet tested (Myers). Several
other labs are also creating transgenic mice containing expanded repeats (Dan Tagle, Didier
Devys, and Hamachandra Reddy); evaluation of these mice are ongoing, but no overt
phenotype has been observed as yet. Transgenic mice containing normal range repeats have
also been produced by Gill Bates, Dan Tagle and Michael Haydens lab; no phenotype has
been observed in these mice.
Several groups (Gill Bates, Graeme Hodgson and Carlo Iannicola) described YAC transgenics
carrying expanded CAG repeats within full length or truncated genomic DNA fragments, but no
obvious neurological phenotype has yet been observed in these animals. Where protein
expression studies have been done, protein has been detected. In addition, YAC transgenics
carrying a normal size repeat do not manifest any phenotype.
Targeted replacement of the murine endogenous CAG repeats with larger repeats
corresponding in size to the disease range has been done by several groups. Rick Myers
described such a "knock-in" approach to produce mice with 70 CAG repeats. These mice
became hyperaggressive, hyperactive and hypersexual beginning at around 3-4 months of age
and homozygotes are also noticeably smaller in size. These transgenics express protein at
only 1/3 endogenous levels; new animals have been produced which produce mutant protein
at similar levels to the wild type product. The same overall strategy is being used in the
laboratories of Michael Hayden (Jamal Nasir) and Marcy MacDonald (Wojtek Auerbach) to
produce knock-in mice; no overt phenotypes have yet been observed. Protein expression
varies from normal endogenous levels to less than normal levels.
Transgenic mice models for other expanded CAG repeat disorders were also presented and
proved extremely informative. Harry Orr and Huda Zoghbi described SCA1 transgenic mice
containing full length ataxin1 cDNA with normal range and expanded repeats with expression
targeted to Purkinje cells. Transgenics containing 82 repeats develop both behavioral and
movement (ataxia) deficits. While RNA was overexpressed compared to endogenous levels,
no protein containing the 82 repeat was detected. Diane Merry also presented data from
SBMA transgenic mice containing 24 repeats or 65 repeats driven by different promoters and a
movement phenotype has just begun to develop in one of the oldest mice. Expression at the
RNA level ranges from endogenous levels to 5 times endogenous and, as described for SCA1
transgenics, the expanded repeat containing protein has not been detected.
Additional animal models were also discussed as were novel interacting proteins under
investigation.
MEETING NOTES:
MOUSE MODELS
GILL BATES: Exon 1 transgenics
Gill described the severe neurological phenotype suggestive of HD in transgenic mice
produced using only exon 1 of the HD gene carrying a large expanded CAG repeat (published
in November 1, 1996 volume of Cell).
The initial experiment had been to create YAC transgenics for the purpose of assessing
meiotic instability of CAG repeats. To this end, a 130 CAG repeat had been cloned into a
human HD containing YAC, resulting in extreme instability in the YAC context. From these
studies, a 1.9 kb human genomic fragment was generated containing 1 kb upstream sequence
(potential promoter element) + exon 1 + 260 bp intronic sequence which was used to
generated the transgenics. The repeat within exon 1 was heterogeneous within a given DNA
prep (unstable when propogated in E. coli), but averaged about 130 repeats. This genomic
subclone was not part of an expression vector and therefore was missing polyadenylation
signals normally considered important for mRNA stability. Within the transgenic population,
there were many neonatal deaths, (possibly significant). From a single founder mouse (R6)
with 5 (!) separate integration events, 5 separate lines were established (range of repeats:
115-155 CAGs in original transgene).
3 lines (R6/0:142 repeats) single copy integrants
(R6/1:116 repeats)
(R6/T: truncated)
1 line (R6/2:144 repeats) 3 copy integrant
2 flanking copies around 1 complete copy highly deleted;
acts as a single copy integrant
1 line (R6/5) 6 copies integrated
4 copies intact
2 deleted/rearranged (complex)
Original founder was determined to be a germ line chimera with one set germ cells containing
R6/0 and R6/T and the other containing R6/1, R6/2 and R6/5.
Expression studies: performed on the three lines with a phenotype (R6/2, R6/1, R6/5). In
these three lines, ubiquitous expression at both mRNA level and protein level was observed.
When compared to endogenous mouse hd gene expression, R6/2 is 75% of normal, R6/1 is
31% of normal and R6/5 is 77% of normal.
RTPCR: transgene expressed ubiquitously (tissue level)
Northern: used intronic portion of transgene as probe, therefore will detect mRNA
with intron only; implies at least some of message is unspliced and that
there is no splicing to mouse exonic sequences.
Western: Due to stop codon present at beginning of intron, can predict size of protein
being made. (Antibody epitope used was expanded polyglutamine repeat). In 2 lines (R6/2
and R6/5), protein is detected by Western, in one line (R6/0), protein cannot be detected
(possibly not large enough repeat to react with antibody). The observed protein is larger than
the expected 23 kDa, but is consistent with aberrant migration patterns observed for expanded
polygln containing proteins when compared to normal.
Most of the work described has been conducted on the R6/2 line (3 copy integrant) containing
a 144 CAG repeat. This transgene was found to be unstable in first F1s and not as many F1s
as would have been expected were born. These F1s did manifest a progressive neurological
disorder with the following phenotype. From weaning, these mice are indistinguishable from
others. At approximately 2 months of age, they begin to manifest symptoms and a handful
show sudden death. When held by the tail, mice clasp legs together (indicative of a neurologic
disorder). A pronounced, resting tremor develops which is less noticeable when quiet or
asleep and there are gait changes, that is, the mice suddenly will dart around. A movement
disorder also occurs which is sudden and irregular; mice will suddenly begin rapidly shaking
and twitching. (Veterinarians who have witnessed this movement, have told Gill that it is
similar to chorea described for dogs). There are unusual stereotypic grooming habits as well
as handling induced epileptic seizures (grand mal). One mouse died during such a seizure. A
dosage effect is associated with the phenotype: transgenic mice with >1 transgene display a
more severe phenotype. F1s die suddenly with age of death between 10-13 weeks. Mice
demonstrate a weight plateau, are constantly hungry and become emaciated. They lose
muscle bulk and develop different facial features.
Transgenic mice are also reproductively compromised: F6/2 females are sterile and males
have greatly reduced fertility. In the last 2-3 months, the line has become well established, but
it was difficult at first. No breeding problems with untranslated transgenes were found,
therefore polygln has to be expressed.
Upon autopsy, mice often have a small thymus. The liver has polyploid nuclei and a reduction
in cytoplasmic mass (often associated with aging). A consistent finding at 11-12 weeks is that
the transgenic mice brains are approximately 20% smaller than those of their littermates
(normal density throughout). This shrinkage was maintained across all structures. Gliosis is
not apparant. This neuropathology is consistent with early stage HD and suggests that
shrinkage of the brain occurs early in the disease process.
Note: Marcy MacDonald commented that they found brain size (80% N) normal relative
to total body weight (60-70% N) in their transgenics and Michael Hayden commented
that while they see ubiquitous protein expression, and therefore cannot localize by
immuno data, they saw no specific neuronal cell loss, either.
Gill then described experiments by Steve Davis, who uses GFAP staining with immunoprobes
for every brain region/spinal cord region. By GFAP staining at low power (R6/2 line at 2
months of age), the region most affected are the glial proliferating striasomes; clusters around
the ventricular edge of the striatum are not present in control. More recently, it has been
possible to look at older mice (around 13.5 weeks). GFAP staining has shown patches of
reactive astrogliosis in the caudate, close to the lateral ventricular edge. This is very similar to
the GFAP staining observed by John Hedreen to correlate with the position of stiosomes in the
brains of early HD cases. This is extremely exciting as both of these changes suggest that the
mice are developing the neuropathology associated with early HD.
Gill also discussed repeat instability, which will be described later in report.
The concern was raised that with so many integration sites (5) in the original founder, it could
be integration events in the genome that are causing the observed effects versus the
expanded polyglutamine tracts. However, with the various F1 lines generated, hemizygous for
each and in different combinations, you basically have transgenics for each integration event.
Also, the possibility of chimeric proteins was ruled out because, even though disruption of
other genes could occur, as far as could be determined, no fusion proteins were detected by
Western immunoblot analysis.
RICK MYERS: Transgenic mice
Transgenic mice have been developed using three different techniques:
1. Knockins: (Peggy Shelbourne)
Transgenic mice with an average of 70 repeats in germline (changes repeat size during
passage in E. coli) were produced. Three founders yielded a large number of offspring
and a large number of homozygotes.
A behavioral phenotype has developed in these mice. At this point, 9 months is the oldest
heterozygote. At 3-4 months, mice become extremely aggressive (males more than
females, but both much more aggressive than normal), hyperactive, and hypersexual (all
pairings). Aggressiveness is so pronounced that some mice are killed by biting, etc; it is
not 100% of the time, but so increased that mice need to be separated ($$$). Rick's group
will be collaborating with Larry Tecott at UCSF (Psychiatry) to quantitatively measure
aggressive behavior.
Recently, homozygotes have yielded new information. Whereas, heterozygotes are of normal
size (weight), homozygotes are noticeably smaller in size with no corresponding change in
diet. From birth, there is a growth delay. Very preliminary data presented is that they are
beginning to see a movement phenotype, i.e. dragging hind legs, in a few mice (at ò 5
months). No pathological findings have been observed as yet, but they have only looked at
younger animals and have obtained equivocal results from a few autopsies.
The transgene is expressed in ES cells and in brain and blood of knock-in mice at
approximately 1/3 N intensity, due to transcriptional interference of the selectable markers.
For these experiments, two polyclonal antibodies (published) raised against the first 130
amino acids (peptides) were used. More recently, the selectable marker (neo), which had
been inserted in the first intron, has been removed by cre-mediated excision (Gail Martin
technique) and animals have been derived which produce mutant protein at similar levels
to the wild type product.
Discussion developed regarding constructs in ES cells. Marcy mentioned that they have a
double knockout in ES cells (in collaboration with Scott Zeitlin, et al) and that ES cells are
viable, displaying no phenotype at this point. They do not yet have a double knock-in in ES
cells, but would like to develop this in order to manipulate levels of abnormal protein. No
breakdown product observed yet. Didier also mentioned manipulating repeats in cell
culture; he has HD cDNA in a neuroblastoma cell line and has observed increased cell
death (20-30% more) when the mutant protein is expressed.
The next two approaches from the Myers lab were described by Carlo Iannicola:
2) YAC transgenic:
A YAC transgenic has been produced containing 850 kb of genomic DNA (lost 100-200kb
DNA), although it has been difficult establishing a stable line. The transgene contains 67
CAGs (grew from 60 to 67 repeats during course of experiment) and two stable
integrations occurred. It is not yet known if compete integration of YAC was obtained. No
phenotype has developed yet, however, mice are still very young.
3) cDNA injection:
Two lines of mice in B6CBF1 background were produced which are transgenic for a truncated
portion of the HD gene. Each contains 39 out of 67 exons for a total of 1676 amino acids
(N-terminal end) or 5800 nucleotides with 67 CAG repeats. Each construct also contains
an epitope tag from the human c-myc protooncogene at their C-terminus.
Two different promoters drive the transgenes, a CMV promoter or an HD promoter containing a
portion of the 5' UTR from the original HD exon 1 which extends back to a SmaI site
approx. 350 nucleotides 5' of translation initiation site.
7 total founders of variable age have been produced: 4 containing transgenes driven by the
HD promoter and 3 driven by the CMV promoter (1 older/2 younger). Transgenes have
integrated at 1- 2 sites, however copy number varies. Mice containing the HD promoter
have 1, 2-3, or >10 copies while CMV mice have 3-4 (youngest 2, not known yet).
Expression has been tested in some, but not all lines. RT-PCR and immunohistochemistry
(anti-cmyc human antibody) have yielded preliminary data. In one HD founder, widespread
expression (of transcript and protein) is observed in brain (only tissue checked so far).
CMV founder also shows widespread expression in brain at slightly higher levels. There is
no information yet about relative amounts of endogenous vs transgenic HD protein
expression.
Two of the oldest founders with the HD promoter and one with the CMV promoter are
beginning to show variable degrees of movement abnormalities and Carlo is in the process
of quantifying and defining abnormality systematically. Preliminary results for one female
founder in particular (CMV) displays two features in common with Gill's mice: 1) intense,
sudden, explosive twitching movement where the mice will twist body at a 90o angle. A
stereotypical movement of the head and trunk is also observed. These movements are
episodic, lasting 2-3 minutes, while in the middle of doing other things, then the mouse will
quiet. They do not occur during sleep periods. Abnormality is not related to voluntary
movement, but occurs during active time of the day. 2) excessive scratching and grooming
occurs during episode. Other mice look the same, but are younger. CMV founder showed
symptoms earlier than did the others and symptoms worsen over time with increasing
intensity and frequency. There is no correlation of severity with copy number.
Animals are not smaller in size, but are rather robust. No data relating to neuropathological
changes has been obtained so far, however no gross changes have been observed with
F1s. Carlo would like to attempt MRI on mice, as a noninvasive approach to study
neuropathology without having to sacrifice mice before potential phenotypic changes
occur, and is setting up system with a collaborator.
Discussion: How does one know if these symptoms are not seizures rather than a
movement disorder? There appears to be a big difference in Gill's mice between seizures
and these episodes. Discussion also focused on need to better control for differences in
mouse genetic backgrounds. For instance, behavioral tests of animals used by the
genetics community showed that for 2 different 129 strains, different responses on some
tests were obtained, while similar responses were obtained on others.
Information on transgenics with normal size repeats was also discussed. Both Gill Bates and
Graeme Hodgson have created transgenic mice containing 18 repeats; at 8 months of age
(Gill) or by 1 year of age (Graeme) no phenotype is observed.
Christopher Ross noted that it is important to separate out the various phenotypes. A case in
point is the NO synthase knockout mouse (S. Snyder with Ted Dawson) is hyperactive,
hypersexual and hyperaggressive (including biting, especially in males).
MICHAEL HAYDEN: cDNA transgenic mice
Described transgenic mice with 18, 44, 80, and 128 CAGs contained within full length and
truncated forms of the HD cDNA.
Constructs presented were either full length (10,366 nucleotides) or truncated (1955
nucleotides, just proximal to apopain cleavage site) cDNAs containing 15, 44, 80 or 128
repeats. Two different promoters, either CMV or HPRT, were used as drivers for each
cDNA. Transgenic lines are at different ages and stage of analysis.
CMV full length (5 lines) or truncated transgene carrying mice express the HD protein at levels
similar to endogenous levels with some variability. Widespread expression is observed for
the 44 repeat transgene (CMV promoter), however, patterns are different than
endogenous. Also see evidence of smaller fragments which are either breakdown
products or cross-reactive species.
By 6 months, the mice carrying a full length cDNA with 44 repeats (CMV promoter) show no
obvious behavioral changes. No gross neurophathological abnormalities have been
observed, however see increased TUNEL staining in cortex and striatum. EM studies show
morphological changes characteristic of apoptosis, such as condensation of cytoplasm. No
evidence of necrosis has been observed.
Analysis is also in progress for the full length cDNA containing 44 repeats with the HPRT
promoter. In one line tested so far, by 4 months no behavioral phenotype and no changes
in body size have been observed. Expression under this promoter is basal ganglia
specific; protein is not found in the cortex or cerebellum.
GRAEME HODGSON (Hayden lab) : YAC transgenic mice
Graeme (see also Dec 1996 Human Molecular Genetics) next described a YAC transgenic
approach for producing an HD animal model. Two YACs [YGA2 (250 kb 5' sequence) and
353G6 (25 kb 5' sequence)] containing the entire HD gene with 18 CAG repeats and all
endogenous regulatory elements were used to create several transgenic lines using the
FVB/N mouse strain. Many of the lines are now greater than one year old and show no
evidence of a neurological phenotype.
Comparison to the endogenous murine gene reveals an identical tissue expression pattern
using a human specific monoclonal antibody. Specifically, the HD gene was expressed in
all tissue types tested with highest levels observed in brain and testes. Subcellular
distribution was determined by immunohistochemical analysis done on sections derived
from multiple regions of the brain, and revealed that both human and mouse proteins
localize primarily within the cytosol of neurons with little to no immunostaining in the
nucleus. Western blot analysis following differential centrifugation of cortical tissue
showed that the protein localizes primarily to the soluble cytoplasmic fraction and to a
limited extent with the synaptosomal and plasma membrane fractions.
Appropriate developmental expression of the YAC is presumed since the embryonic lethal
phenotype in homozygous null HD knock-out mice can be rescued by breeding the YAC
into the knock-out background, indicating that the human gene was expressed at the
appropriate time and location in early embryonic development. Note: In contrast, the full
length gene under the CMV promoter cannot rescue targeted disrupted mice, possibly
because the CMV promoter is not expressing the gene at appropriate time.
Graeme has also introduced 48 CAG repeats in YGA2 YAC via homologous recombination in
yeast. Several lines have now been created using this 48 repeat YAC; all are
heterozygotes so far. Again, protein expression patterns identical to the mouse
endogenous gene is observed, with high levels expressed in cortex, cerebellum and brain
stem and low levels expressed in peripheral tissue. The founder is approximately 6 months
old and shows no gross behavioral abnormalities, although he described the mice as
becoming progressively more difficult to handle and somewhat hyperactive.
A third construct is now in progress. Larger repeat sizes (72 CAGs) have been introduced into
both YACs via the same homologous recombination based approach in order to create
more YAC transgenic lines and to determine if a more severe phenotype manifested earlier
in life will be a function of increased repeat size.
DANILO TAGLE: cDNA transgenic mice
Dan has expressed a full length cDNA HD clone containing 16 repeats, pHDFL16, in a
baculovirus system. A protein at the expected 350 kDa was detected. This construct was
modified in vitro to change the repeat size from 16 to that derived from patient genomic
DNAs of an adult onset case with 48 repeats (pHDFL48) and a juvenile onset case with 89
repeats (pHDFL89).
Each of the above plasmid clones were recloned into pCDNA 1.1 under the control of a strong
CMV promoter that shows high expression levels in different tissue types. Additional
constructs are being made using the hypoxanthine-phosphoribosyl transferase (HPRT)
promoter that directs a more specific expression to the basal ganglia, and using the HD
gene endogenous promoter sequence. Founder transgenic mice are being characterized
(have found 1-2 integration sites) and phenotypes of the founders, F1s and F2s are being
analyzed. One death in a founder with 48 repeats (from being attacked by a stud male)
showed no overt neurophathology other than an observation of possibly some ectopic
cortical neurons.
YAC transgenic:
A 350 kb YAC (353G6) containing the 180 kb HD locus with 16 CAGs and 20 kb of 5' and 140
kb of 3' flanking sequence was retrofitted with 48 repeats. Both normal and expanded
purified YAC DNAs are being used for microinjection to generate founders.
DIEDIER DEVYS: Transgenic mice
Briefly described a transgenic mouse with an expanded repeat created in a Black 6
background. Founder is 2 years old with no abnormal movements, but a behavior
phenotype characterized by the mouse wanting to jump back in cage whenever removed
has developed. Low expression of protein under CMV promoter control is observed.
WOJTEK AUERBACK (in collaboration with Marcy MacDonald): Knock-in mice
A knock-in approach using a mouse-human hybrid of exon 1 containing different length
repeats and the neo gene 900 bp upstream of the ATG surrounded by lox P sites was
described. In ES cells with neo marker present, 1/3 normal expression was observed. In
the knock-in mouse, following excision of neo by microinjections of CMV-Cre plasmid, both
18 repeat and 48 repeat containing proteins are expressed at indistinguishable levels and
are present at similar levels as normal (90-100% expression). Recently, transgenic
animals have been created containing 89 and approximately 120 repeats. Homozygous
and hemizygous (crossed with knock-out) mice are the same size as heterozygous and
wild type. The oldest heterozygotes with 48 CAGs are over 6 months old and none of these
show any obvious behavioral changes.
JAMAL NASIR: Knock-in mice
Knock-in mice using a construct containing 51 repeats and a neo casette within intron 1
flanked by cre-lox sites is in progress. Presently, he has ES cells containing the knock-in
mediated by homologous recombination.
HEMACHANDRA REDDY: Transgenic mice
Has just generated F1 from transgenic mouse containing 48 repeats; phenotypic analysis is in
progress.
At this point, the discussion turned to mice transgenic for constructs containing genes for other
triplet repeat disorders.
HARRY ORR (AND HUDA ZOGHBI): SCA1 transgenic mice.
Under the control of a Pcp-2 promoter (Purkinje cell specific), transgenics containing full length
ataxin1 cDNA with 30 (normal, interrupted repeat) and 82 repeats (expanded,
uninterrupted repeat) have been created. Upon Northern analysis, find expression
patterns similar to endogenous targeted to Purkinje cells, cortex, brainstem and
cerebellum, regions which are subject to pathological changes in the disorder. Levels of
RNA varied from 10x to 100x the levels of endogenous mouse Sca1. Interestingly,
immunoblot analysis revealed protein expression in the 30 repeat lines, but no protein
could be detected in the 82 repeat lines.
For transgenics containing 82 repeats, mice develop both behavioral and movement disorders.
By 5 weeks, there is a deficit in motor learning (rotorod learning) and by 12 weeks, ataxia
develops. Upon neuropathologic exam, Purkinje cell dendritic pruning, ectopic Purkinje
cells and Purkinje cell loss is observed. Now have transgenics with 52 repeats that are
approximately 3 months of age. Also, have NSE promoter driving 82 repeat ataxin-1
(approximates endogenous ataxin-1 levels) and at 1 year, see morphological changes in
Purkinje cells.
Huda also described SCA1 knockout mouse deleted for most of coding region. Homozygous
mice appear fine, however, there are behavioral abnormalities. When placed in a new
cage, they appear anxious and stay very still and show deficits in cerebellar learning
(spatial learning, associative learning and open field tests) as compared to wild type mice.
No clinical ataxia is observed. (Note: This learning deficit varies depending upon the
genetic background: in a mixed C57BL/6x129SV/EV background, null mice can learn but
are impaired. In a pure genetic background 129SV/Ev, the learning disability becomes
more pronounced in Sca1 null mice.)
Discussion focused on possiblities of potentially rescuable cells that could relieve cell death
and improve prognosis for these disorders. In SCA1 transgenic mice with 82 repeats, see
ataxia and other symptoms well before cell death; i.e. have gait ataxia even when >80%
Purkinje cells are present. Therefore, cell death is not the underlying deficiency. Early
phase might be loss of function or functional changes in the neurons of mice which cause
the phenotype, then, gait ataxia, etc manifestation, THEN motor neuron loss. This is very
important for understanding the mechanisms for pathogenesis; it may be that CAG
expansion is not simply causing cell death with the phenotype resulting from neuronal loss.
Discussion continued with clinical aspects of HD. Ann Young has looked at neuropathology in
an individual at risk for HD who committed suicide. This patient was presymptomatic, but
neurochemical changes and early clinical pathology was found. Other studies have
shown that for HD, using an expected age of onset, the farther from the predicted age of
onset, the more normal the clinical pathology (by volumetric MRI), the closer to predicted
age of onset, begin to see caudate atrophy, although not to the same extent as in
symptomatic HD. Marcy reported that PET studies have shown no changes at
presymptomatic stages. Several individuals agreed that for several of these disorders, it
appears that the phenotype is due to neuronal dysfunction before onset of pathologic
changes.
Harry also mentioned that he had crossed bcl-2 mice into the SCA1 transgenics, with no
change in phenotype. (I believe he also mentioned that neurons were rescued, but not the
neuropathy.) If the Purkinje cells were saved, but the phenotype of ataxia remained, this
would be additional compelling evidence that the phenotype is caused by functional
changes and not by cell death. He also mentioned that bcl-2 does not block the effect of
expanded CAGs, but that the protein containing the repeat is highly expressed.
DIANE MERRY: Spinal Bulbar Muscular Atrophy Transgenics
Transgenic mice created with NSE (neuron-specific enolase; 2 lines with 24 repeats, 3 lines
with 65 repeats) or NFL (neurofilament light chain - shows high expression in motor
neurons: 1 line with 24 repeats, 2 lines with 65 repeats) promoters driving full length
androgen receptor containing 24 or 65 repeats. At 18 months, there is no overt phenotype
in any of the mice. Not mentioned at the meeting, one mouse (20 months) has developed
a hindlimb motor deficit, this mouse's symptoms have progressed over 6-8 weeks, showing
hindlimb proximal muscle atrophy and gait abnormalities.
NSE (65) expression ranges from less than endogenous to 2 times endogenous expression.
NFL expression ranges from 2-5 times endogenous expression in the motor neurons. So
far, have been able to detect expression of the 65 repeat by RTPCR, but have not been
able to detect protein containing the expanded repeat by Western analysis.
Using grip strength measurements of forelimbs on a mesh grid when mice are held by the tail,
overexpression of normal repeat (24) androgen receptor, driven by the NFL promoter
results in increased grip strength (no loss of motor neurons). When driven by the NSE
promoter, a 65 repeat containing AR (highest expression) also have increased grip
strength, although it is lower than NFL promoter mice. The physiological basis of this effect
is not yet known.
Hyperaggressive behavior, especially with the NSE promoter was observed in mice and found
to be more pronounced in males than females. Mice are also very hyperactive and circle
constantly. Do not see this behavior to the same degree with N repeat transgene. No
hypersexual activity has been observed and no pathology has yet developed.
Note: Experiments in yeast using the amino-terminus of the androgen receptor shows that the
receptor is translated and functions without the polyglutamine tract.
Diane also commented on experiment by Kakizuka (Nature Genetics, June, 1996) who showed
that there is a differential effect on Purkinje cell loss depending upon whether full length or
truncated ataxin3 is used: when full length ataxin3 (MJD1) driven by a Pcp-2 promoter
(targets expression to Purkinje cells) with expanded or normal repeats is used, there is no
phenotype, whereas when a truncated form of ataxin3 containing expanded or normal
repeats is used, Purkinje cell loss is observed. (Interestingly, Purkinje cells are not normally
affected in Machado-Joseph diseases as they are in SCA1). This effect (toxicity of the
polyglutamine tract alone) was observed both in cell culture and in transgenic mice (severe
ataxia), raising the issue of whether the expanded polyglutamine tract is toxic by itself, but
loses cell type specificity which is dependent upon the full-length protein. In the absence
of information regarding transgene copy number, integration sites or expression levels,
however, it is difficult to come to a firm conclusion from these experiments. There was
much discussion regarding this idea; most groups are following up experimentally on these
results by comparing the effects of truncated and full-length cDNAs.
Discussion then turned to a description of the putative apopain cleavage site in the HD protein.
JAMAL NASIR: Apopain Results
Apopain is a protease involved in programmed cell death and is the human counterpart to C.
elegans CED-3. It normally works as a cell death effector when apoptosis is set up and is
part of the normal physiologic process. Within the HD protein, (at approx nucleotide 1900),
there exists a specific substrate cleavage site and amino-terminal fragments that could be
representative of a truncated protein have been observed by Western blot. Paul Goldburg
(Nature Genetics) has shown that cleavage is related to the length of the CAG repeat; the
greater the repeat length, the greater the amount of apopain mediated cleavage. The
theory presented is that the cleaved product enters the nucleus and acts as a possible
transcription factor (similar to BRCA1), however, cleaved products may also enter other
compartments. There is some support from the lab for an increase in apoptosis (TUNEL
staining reveals evidence of 3-8% cell death in mice containing 44 repeats compared to
none in control), but no evidence as yet in vivo.
How does process start? Questions were raised addressing the fact that apopain is a
downstream signal for apoptosis (downstream of CED9 in yeast and bcl(2)), and that it is
difficult to understand how apopain mediated proteolysis of the HD protein can be a trigger
of the process as opposed to a possible result.
Discussion then centered around aggregation of the HD protein as a possible model for
pathogenesis of expanded repeats as recently described by Max Perutz (PNAS, July,
1995). Henry Paulson has seen aggregates with truncated forms of SCA1 in COS cells,
but no aggregates have been observed yet in brain (1 brain checked). Various
experiments were described in E. coli and other systems addressing this issue. For
instance, expression of 200 amino acids (20 kDa representing N-terminus of HD protein)
with and without a long glutamine repeat resulted in dimer formation of the expanded
repeat protein that could not be taken apart with DTT (Rick Myers) . In contrast,
Christopher Ross has looked at N-terminal pieces of the HD protein and has not seen any
evidence of aggregation. Aggregation is also found that is not resolvable to a single
species with SDS, DTT or boiling using a truncated form of ataxin 3 (40 aas ouside of
polyglutamine tract) containing normal or expanded repeats. Three different groups
(Martine Usdin, Christopher Ross and Peter Ross) reported that no interactions
representing potential dimerization of the HD protein were observed using the yeast two
hybrid system with any combination (normal or expanded) of repeats.
OTHER ANIMAL MODELS
Drosophila
George Jackson (Zipursky lab in collaboration with Marcy MacDonald and Peter Faber)is in the
process of developing transgenic flies with a truncated portion of the HD gene and using a
dominant suppressor screen (via crosses with flies carrying EMS-induced mutations) for
genes that interact downstream in relevant pathways. Constructs containing 2 or 58
repeats in the N-terminal portion of the HD gene (extending into exon 4) have been used in
an effort to create transgenic flies with high levels of expression in the eye alone (using a P
element, GMR, that targets expression to photoreceptor neurons). In this way, a cell death
phenotype may be is localized to the eye alone in flies that remain viable, permitting
crosses in order to screen for enhancer and suppressor mutations.
Leslie Thompson (in collaboration with Dr. Larry Marsh at UC Irvine), is also using Drosophila
to set up a genetic screening system aimed at distinguishing between cellular toxicity
caused directly by the expanded polyglutamine tract versus pathology caused by a gain of
function or inappropriate function of the HD protein brought about by the incorporation of
the expanded polyglutamine repeats. To distinguish between these possibilities, the
naturally occurring polyglutamine tract of a phenotypically assayable Drosophila gene,
dishevelled (dsh), is being modified to contain normal range and expanded repeats (0, 27,
95 and several hundred) and the effect of these repeats on protein function and cell
viability will be assessed. Free peptides are also being expressed containing the
equivalent variable numbers of polyglutamines and their effect on cell function is also being
physically and functionally assesed. Preliminary data (not presented at meeting) indicates
that dsh is very sensitive to the presence of the repeat, implying that they are necessary for
full function.
HD homologs:
Drosophila: Low stringency hybridization using the C-terminal end of the HD gene as a probe
(region conserved across species (human, mouse, and fugu) has only yielded aspartate
aminotransferase after several attempts (Marcy MacDonald). Several other groups have
tried as well, with no success.
Zebrafish (Christopher Karlovich): Chris has isolated the entire zebrafish cDNA, except for
what corresponds to about the N-terminal 7 amino acids and 5' untranslated region. This
homolog shows 84% identity to the human HD gene so far. The zebrafish gene is also
similar to puffer fish (Fugu) sequence (both have 4 glutamines). A goal of his research will
be to delineate functional domains of the protein using developmental expression studies.
The advantage of using zebrafish is that development can be so well studied. These
studies will utilize injection experiments (not antisense as in Xenopus) and overexpression
of domains, including the apopain cleavage product. Marcy mentioned that a fellow in her
lab had achieved expression by RNA injection of a truncated portion of the zebrafish gene.
C. elegans (Jamal Nasir): no homolog identified as yet.
Yeast: nothing significant in yeast database that is homologous.
Large animal: Russell Snell would like to see a sheep model.
Other homologs:
Puffer fish (fugu): Several Sca1 homologs have been identified (Huda) containing regions with
varying degrees of homology (40%-60%).
BIOCHEMICAL INTERACTIONS
LESLEY JONES:
Lesley discussed experiments that show that Huntingtin is associated with microtubules
displaying a cytoskeletal pattern in cells. Following rounds of polymerization and
depolymerization, find continuous MAP2 association with microtubules, whereas Huntingtin
associates for 2 rounds, then dissociates. She also noted that interestingly, GAPDH, which
has been shown to interact with huntingtin, is a microtubule bundling enzyme. Cross-
linking experiments with Huntingtin showed no association with MAP2, GAPDH or Tau
protein. Will move on to a molecular system, such as yeast two hybrid, to more clearly
define potential interactions. Mandel has also seen a micotubule association with
Huntingtin by EM.
Lesley then looked at diseases with known microtubule involvement, including Alzhemer's and
Pick's diseases. Huntingtin was represented in plaques, tangles and dystrophic neurons,
in Alzheimer's brains and in neurofibrillary inclusions in Pick's disease brains.
When she examined HD brains, commenting on the fact that in the UK there can be a long
interval post-mortem, she finds staining in reactive astrocytes in caudate (also see in
Alzheimer's brain), with more gliosis. The pattern is best observed in formalin fixed parafin
embedded tissue where there is a clear difference between HD and N caudate. Question
is: what is HD doing in astrocytes? She also described evidence for apoptosis in
astrocytes (4-8%) in HD brain observed using TUNEL staining, mentioning that one can get
TUNEL positive cells for apoptosis due to ds breaks, but false negatives are also possibile,
therefore; experiments need to be well controlled.
RUSSELL SNELL:
Russell has recently begun working on HD again and his hypothesis is that the expanded
polyglutamine is the toxic element. He is in the process of examining repeats using NMR
and CD (with industry collaborators) as well as electroporating repeat containing DNA into
cells in order to see if there is any response. He is also looking at an innovative treatment
approach which utilizes RNA/DNA oligos that bind to the tertiary structure of a protein, thus
disrupting potential interactions of that region of the protein with other molecules. One can
get a high degree of specificity with this technique and he would like to try the oligos on
animal models as a viable treatment protocol.
INTERACTING PROTEINS:
HUDA ZOGHBI: GAPDH
Based on the need to test for interaction between GAPDH and Huntingtin in vivo, Huda is
making GAPDH transgenic mice to cross with the pcp2 promoter ataxin-1 mice that show
the pheotype of ataxia. Her hypothesis is that GAPDH overexpression might protect
Purkinje cells. However, based on papers showing there is an increase in GAPDH in
granule cells undergoing apoptosis and that GAPDH oligos block apoptosis, is GAPDH
involved in apoptosis and how might it fit into the picture of CAG repeat diseases.
Experiments were also described where an inducible androgen receptor was expressed in
CHO cell lines and the expanded repeat found to kill cells. In cells negative for the
receptor versus induced or uninduced cells, no change in the function or expression of
GAPDH was found.
PETER FABER: yeast two hybrid (MacDonald lab)
Peter emphasized the difficulty in showing interactions with Huntingtin in a reproducible,
consistent manner, especially with the yeast two hybrid system. He used two different
GAL4 based systems resulting in low expression levels or high expression levels of a
truncated construct containing either 20 or 60 repeats. He screened several libraries (no
interactors with Clonetech system, luciferase system yielded 5-6 interactors). Roger
Brent's lexA based system using a slightly shorter HD bait yielded 5 interactors. Constructs
containing 2, 23, or 62 glutamines, produced a differential interaction pattern with one
interactor; the 62 repeat HD gene interacted while the 23 repeat gene did not. These
represent novel cDNAs, ubiquitously expressed; some do not even have ESTs in the
database. To attempt to establish interactions outside of yeast systems, Peter will try to
co- immunoprecipitate interacting proteins with Huntingtin.
MARTINE USDIN: yeast two hybrid (Hayden lab)
Martine discussed trying to use the yeast two-hybrid system of Elledge (gal4 based vectors) to
find proteins which may interact with Huntingtin and to reproduce results previously
obtained. Bait was constructed which contained the first 1676 amino acids, but this did not
appear to be stable (no expression by Western blot). A piece containing the first 1516
amino acids (and 17 glutamines) did produce a positive blue color when expressed with the
hap1 clone given to them by Christopher Ross. A high-efficiency screen yielding about
150 positives, 80 of those passing the first tests of specificity. However, none of these 80
turned blue when retransformed with the original bait, implying that these putative
interactors are in fact false positives. Others, including Christopher Ross and Peter Harper
mentioned that their labs experience similar difficulties.
CHRISTOPHER ROSS: HAP1
Preliminary results at the yeast level show interaction partners with HAP1. Using the amino-
terminus of HAP1, one interacting protein contains spectrin like domains and GEF
domains, which could connect Huntingtin to some known pathways. They now have some
better antibodies with which to study HAP1. The HAP1-Huntingtin interaction is
reproducible in either orientation in yeast (i.e. Huntingtin as bait or HAP1 as bait), but is
strongest in the original orientation. A potential therapeutic (pharmaceutical) application
would be to screen for small molecules that could interfere with this interaction.
REPEAT INSTABILITY STUDIES
ESTHER P. LEEFLANG: Mathematical model using Single Sperm typing
The goal of this research is to understand the dynamic mutation process by which an increase
in triplet repeat size corresponds to the mutation process of HD and other similar disorders.
The increase in repeat sizes during transmission of the HD allele explains the observed
phenomenon of genetic anticipation. In affected individuals at the HD locus,
the majority of somatic samples show no heterogeneity in repeat size, whereas semen
samples show marked heterogeneity. Although it has been observed through family data
that transmission of the HD allele via paternal transmissions show not only a more frequent
change, but a larger increase in repeat number, when compared to maternal transmissions,
little is known about the mutation process.
Family transmission data is limited not only by the number of offspring available for study, but
is also not controlled for variables which could affect triplet repeat instability, possibly
including, but not limited to, chromosomal haplotype (or cis-acting factors), genotype,
environmental effects, and age. In contrast, single sperm typing affords an unlimited
number of meioses. Also, single sperm typing resolves problems associated with bulk
sperm typing that can mask rare events, such as PCR stutter and small allele bias. In
order to detect any effects of age, the ideal situation would be to study sperm samples
from one individual over a course of time. Since these resources will not be available for
several years, individuals with identical allele sizes but of different ages from the
Venezuelan pedigree were investigated, thus reducing the contributing genotypes and
limiting environmental effects.
Mutation spectra were generated from an analysis of over 4000 HD single sperm for 27
samples, where mutations are scored relative to somatic lymphoblastoid cell control sizes.
The samples include several sets of brothers, cousins, second cousins, third cousins, and
a father-son pair. The samples range in age from 17 to 65, and in repeat size from 35 to
62. The mean change, percent expansion and percent contraction range from -.86 to 32.5,
13% to 99%, and 0% to 41%, respectively. In general, mutation frequency and mean
change increase with allele size.
Comparison of 5 individuals with 44 repeats with ages 23, 32, 44, 54 and 55 show a general
increase in mean change and increase in the upper limit of repeat sizes observed. In
addition, comparison of the mutation profiles in a 39 repeat allele in a 65 year old with a 40
repeat allele in a 20 year old demonstrates that the mutation frequency and mean change
increase with age (an effect also seen by Gill with her transgenic mice).
Interestingly, the 55 year old sample showed a vastly broader profile than his nearest age
match of 54 years old (upper limit 156 repeats versus 69 repeats), and despite not being
the oldest sample or the one with the highest somatic allele size, was clearly a unique
profile when compared to the data set as a whole. His son, an 18 year old with 62 repeats
also showed a "more mutated" profile when compared to his age and size match.
Sequence analysis showed no interruptions of the CAG or CCG tracts, nor any nucleotide
substitutions in the bases separating these two tracts. These more highly mutated profiles
suggest that possibly an additional genetic factor(s) or modifier is contributing to the
mutations observed.
In an effort to understand the mutation process, collaborators in the Department of
Mathematics developed a stochastic model which can simulate the observed mutation
spectra. The model incorporates known biological mechanisms and currently held
theories, i.e. allows for the changes of the addition or loss of one trinucleotide repeat, P1,
(as observed in microsatellites) vs a geometric addition of repeat number, P2 (increases of
integer values). An additional parameter, Q, is used to describe the size of the P2
additions, or the shape of the geometric distribution. Other considerations include
leading/lagging strand model of DNA replication and age of the sample (based on 23 cell
divisions possible per year after puberty in spermatogonial stem cells).
The modeled data fits well; there is good agreement between the modeled and actual data as
to which allele size is the most frequent, and the lower and upper boundaries of allele
sizes. Models allowing only P1 and not P2 mechanisms are mathematically rejected using
the one step model, that is, P1 alone is not sufficient to explain the data. This implies that
the mutation process is not a simple stepwise accumulation of small one repeat
expansions. Plotting P1 and P2 values versus triplet repeat length shows low but common
levels of expansion in individuals with less than 49 repeats. However, individuals with
greater than 49 repeats show a huge jump in P1 and P2 values. That the P1 and P2
mutation rates should suddenly increase for alleles greater than 49 repeats is a revelation
that has biological consequences which merit further investigation.
Since investigations of samples with high repeat number and advanced age are precluded due
to disease progression, the model has been used to simulate such a scenario. When a 49
repeat sample was simulated to age 70, the modal value was shifted from approximately
52 to 130 repeats, and the highest repeat size increased dramatically. In a similar
simulation, the mutation spectrum of the father in the father-son pair was simulated to the
age at his son's conception. In contrast to the tested sample of the father 18 years after
his son's birth, the simulation showed a higher frequency of the transmitted allele.
This model does not distinguish between mutation and DNA repair, therefore, the mutation
rates derived are net rates which take into account any DNA repair which may have
occurred. This is the first time that triplet repeat mutation rates have been derived for the
HD locus, using mutation frequencies derived from single sperm mutation spectra.
Because of similarities in range of allele sizes and paternal germline instability, she feels
confident that this model can also be used to predict mutation rates at other CAG triplet
repeat loci.
Marcy MacDonald also reported on the PCR analysis of pooled sperm samples compared to
the constitutional blood allele (lymphoblast) samples from 360-370 individuals. Results
show that there is considerable variability between samples, but at the 45-47 repeat size,
the curve breaks precipitously; i.e. 40 repeats will expand to 48, 49 repeats in sperm,
whereas at sixty repeats, will find expansions in sperm up to 89 repeats. They are in the
process of analyzing the distribution of allele samples and have not yet done a comparison
with respect to age.
TRANSGENIC MICE: REPEAT INSTABILITY
Gill Bates analyzed repeat instability in transgenic mice containing exon 1 with different repeat
sizes and found both germ line and somatic instability. Male transmission was studied in
all four lines: R6/0 (CAG)142, R6/1 (CAG)115, R6/2 (CAG)150, R6/5(CAG)130-156. In lines R6/1,
R6/2 and R6/5, the repeats expand upon male transmission. Although the repeat size may
not change much as you follow each case through generations, within a litter see changes
(of 1-2 units), the largest change being 8-10 units in one generation. Female transmission
has only been studied with sufficient numbers in line R6/5, and the repeats show a
tendency to contract. The repeats were more stable on germ line transmission in line R6/0,
although changes of one or two repeat units could not be ruled out. It was possible to look
at the increase in instability with the age of the founder parent for lines R6/2, R6/1 and
R6/0. There was a statistically significant increase in the instability of the CAG repeat with
the age of the founder for line R6/2, but not for the other two lines.
Somatic instability was observed in lines R6/1, R6/2 and R6/5, but not in line R6/0. For each
of the lines, instability was first seen at around six weeks of age. 18 tissues were
examined and instability was first and most apparent in brain tissues, but liver and kidney
were also frequently affected. Expansions were primarily observed. However, by 42 weeks
(oldest mouse), the only tissue tested in an R5/5 mouse that did not show marked
instability was testis. Interestingly, no evidence of somatic instability was observed in line
R6/0, the line in which the transgene is not expressed.
Peggy Shelbourne (knockin mice/70 repeats):no repeat instability observed in mice so far
Carlo Iannicola (YAC transgenic with 67 repeats): repeat stable so far
Marcy MacDonald and Wojtek Auerback (knock-in): 1 contraction
Graeme Hodgson: in 100 meiosis of the YAC transgene (353G6), there was no effect on
stability; no instability has been observed with the YGA2 Yac either.
Other triplet repeat disorders:
Diane Merry: no intergenerational instability with SBMA transgenic mice.
Harry Orr: No instability observed in male transmissions in SCA1 mice, although contractions
were found upon maternal transmission. No somatic instability of the 82 repeat was found.
Other chromosomes do not appear to have an effect on instability based on the fact that
no change in instability is found when mice bred to homozygosity. In contrast, when the 82
CAGs are in YAC context (large genomic context) and replicated in yeast, see both
expansions and contractions.
Huda Zoghbi: Instability of the repeat is observed in sperm from SCA1 patients.
Note: David Houseman mentioned that myotonic dystrophy shows dramatic age effects in
repeat instability. The later the birth order, the worse the disease, correlating with
increasing repeat size.
CLOSING COMMENTS:
What have we learned and where do we go from here was addressed during the closing
thoughts from each attendee. Generally, scientists were very encouraged by the progress
being made and were enthusiastic about the meeting discussions. Several points were
raised. Many participants emphasized the importance of parallel experiments utilizing
truncated and full length forms of the HD protein. There was some concern that while
valuable information is being obtained from the truncated polyglutamine data that may be
relevant to the various disease states, a comparison to the full length protein is essential.
The question was raised as to whether we need to choose between neurological and
clinical presentation (Gill's Bates results) or look for a genetic paradigm such as the knock-
in mice that mimic the genetic situation of HD, but may not display the HD phenotype,
however it was generally agreed upon that both types of models are useful. The most
difficult aspect of the transgenic phenotypic results to understand is the cell specificity of
the neuropathology. Are regions in the protein other than the polyglutamine site
contributing to the cell specificity?
Since the different transgenic mouse models provide tremendous diversity in phenotype, it is
important to compare differences between the models for cause and effect, such as strain
of mice, expression levels and promoters, and sequences surrounding integration sites.
The possibility of posting the information gathered together (i.e. workshop report) or on a
Web page was presented as well. There appears to be an underlying need for rigorous,
uniform, phenotypic study and the importance of developing a variety of animal models to
compare, including large animals, was emphasized. The hope is that the use of various
animal models will help to explain what causes the disease in humans.
Other points raised included the need for better determination of cellular localization of the HD
as well as the SCA-3/ MJB proteins. Potential aggregation of full length Huntingtin in the
presence of expanded repeats should also be pursued, possibly by trying to expose the
epitope in cell culture to determine if it leads to aggregation. The question was also raised
as to whether there was any precedent for the possibility that the polyglutamine is normally
buried by surrounding sequence, but becomes toxic when it is exposed upon expansion or
if, as seems to be the most likely scenario in view of polyglutamine repeat containing
transactivating transcription factors, the expanded repeat could take on a different
conformation. Searching for potential genetic modifiers that may influence repeat
instability in families was also mentioned as an area for future research. Several
participants reemphasized the emerging information that there may be an altered function
of neurons preceeding neuropathology and cell death and that this should be investigated
further.
The meeting concluded with comments by Nancy Wexler and Allen Tobin. Nancy reiterated
the fact that we have many pieces of information and we need one composite of the
different approaches. Putting this information together is something that the foundation
can help with towards the goal of understanding what the whole animal model might look
like. It was emphasized that if researchers need something, to please ask the HDF, so
people do not have to sacrifice their research (a safe house for HD mice was mentioned
during the course of discussions). Nancy also commented on the videotape of Gill Bates's
transgenic mice containing exon 1 of the HD gene. With just a piece of the gene, it is
amazing to see mice having difficulty walking around and developing seizures; it has a
profound impact on a person watching it. From the position of HD research in 1968 to
now, there have been quantum leaps attained and she thanked everyone for their
contributions. Allen mentioned that a compendum of the data regarding all the models will
be forthcoming and that when we meet again we will revisit issues, such as the behavioral
phenotypes. He also encouraged people to send in grant applications to the HDF. |
|
