FightSMA
2005 Annual Conference
April 19 and 20, 2005
The Melrose Hotel, Washington, D.C.
FightSMA’s 2005 Annual conference was held in Washington, D.C for the second year in a row. Researchers from around the world were present. Families, friends and other members of the SMA Coalition joined together for SMA Day on the Hill. FightSMA families and members of the SMA Coalition traveled to the offices of Representatives and Senators to seek their support. Congressman Cantor has graciously hosted an evening reception on Capitol Hill.
 |
 |
 |
The meetings in April opened the door to a deeper understanding of the
disease and strengthened relationships within the SMA community. The following
lay research summaries were prepared by Chris Lorson, PhD, FightSMA’s
Science Director.
Adrian Krainer (Cold Spring Harbor Laboratory,
NY)
Arthur Burghes (The Ohio State University)
Chris Lorson (University of Missouri)
Darryl DeVivo (Columbia University)
Kathy Swoboda (University of Utah)
Nick Boulis (The Cleveland Clinic)
Adrian
Krainer (Cold Spring Harbor Laboratory, NY)
The SMN2 gene is an excellent therapeutic target because
it has the potential to produce normal SMN protein. However, due to a
process called RNA splicing, a very small, but very important, piece of
the SMN2 gene is not made into protein. Briefly, the DNA that encodes
the SMN2 gene produces a pre-mRNA transcript. This is essentially an exact
duplicate of the DNA but has now been copied into a slightly different
molecular form called RNA. RNA is what is used to make protein (DNA cannot
be used for this process), however, only a small percentage of the actual
pre-mRNA contains the information that is used in the production of a
protein. As an example, this pre-mRNA could be equated to a book that
has 20 chapters, however, the information that is important for a particular
recipe is found in chapters 1, 4, 9, and 20, the remaining chapters are
simply junk and can be discarded. This is essentially what happens in
RNA splicing. The important regions that are the instructions for making
a particular protein are encoded in “exons” and the sequences
in between are called “introns.” The cell can find the exons,
bring them together, and remove the intron sequences, forming an RNA that
is ready to make a protein. Unfortunately, in the case of SMN2, there
is a mistake that tells that cell to throw out the final “chapter”
or specifically, exon 7. Dr. Krainer has previously very elegantly shown
that the basis for this error in assembling the SMN RNA is due to a disruption
in a binding site for a protein that tells the cell’s machinery
to include exon 7 in the final RNA.
The protein that is responsible for telling the cell that SMN exon 7 is
an actual exon (and important “chapter”) is called SF2/ASF.
This protein has two parts: 1) a large region that binds a specific RNA
sequence such as the one present in SMN1 exon 7; and 2) a smaller section
that is an “activator” domain. The activator domain is the
portion of the protein that is directly involved making sure that SMN
exon 7 is included in the final spliced RNA. The SF2/ASF protein and others
are essential for producing the full-length SMN protein, however, it is
highly unlikely that a therapy for SMA (or any other disorder) would involve
treating a cell with a full-sized splicing factor such as SF2/ASF. To
this end, Dr. Krainer and colleagues have developed small novel synthetic
molecules that are designed to stimulate SMN2 expression. These molecules
are not drugs per se, rather they are rationally designed molecules that
are conceptually similar to SF2/ASF. The molecules are comprised of an
RNA-binding domain and a potent activation domain, however these domains
are significantly smaller than the similar domains found in SF2/ASF.
A critical distinction between the normal SF2/ASF protein in the novel
small molecules is that the “RNA binding domain” of the small
molecules is an anti-sense molecule. Anti-sense technology is based upon
the biochemical nature of nucleic acids. Briefly, DNA and RNA are comprised
of small building blocks called bases. There are only four bases (G, A,
T, C; U substitutes for T in RNA sequences) and there are rules for how
the bases interact with one another. For example, in double-stranded DNA,
G forms a bond with C and T (or U in RNA) forms a bond with A. Based upon
these principles, Dr. Krainer and colleagues have develop short nucleic
acid molecules that are the “complement” to various regions
throughout SMN exon 7 and they have systematically analyzed anti-sense
molecules that bind to SMN exon 7 to identify the ones that result in
the highest levels of full-length (good) SMN. This has been performed
in collaboration with ISIS pharmaceuticals, a company that specializes
in anti-sense technology. This approach has identified 3 anti-sense molecules
that appear to change the splicing pattern of SMN2, resulting in high
levels of full-length SMN from the SMN2 gene.
>>
back to top
Arthur Burghes (The Ohio State University)
Dr. Burghes and colleagues have developed a number of
different SMA mouse models that have been instrumental in understanding
SMA development, including mice that represent severe, intermediate, and
“mild” SMA. As in humans, the severity of the disease in mice
depends upon the number of functional SMN2 copies (the mice lack their
own Smn gene and have 2, 4, or 8 copies of human SMN2).
Recently, the Burghes lab has developed additional murine models including
a mouse that carries a specific mutation called A2G. The rationale behind
this mouse was that it was milder than the severe mouse and might be more
useful for therapy treatments since the mice lived longer. Additionally,
a recent mouse model was developed using the SMND7 gene (the protein primarily
produced from SMN2) which demonstrated that even though SMND7 does in
fact provide some protection from SMA – it is just not as effective
as the normal SMN protein. This is important because it answers the question
whether the SMND7 is detrimental to cell survival. Since many compounds
such as the HDAC inhibitors are envisioned as potential therapeutic agents
for SMA, these compounds not only elevate the level of normal SMN but
SMND7 is also significantly elevated.
Other mouse models have shown that unnaturally high levels of normal SMN
protein do not hurt the development and survival rates of mice. This is
important because some therapies such as the Lenti-SMN gene therapy strategy
could actually achieve levels of SMN protein that are even higher normal
SMN1 levels. Other mouse studies suggest that SMN may be required very
early in development and that even high levels of SMN, if delivered too
late, are not sufficient to prevent SMA development.
The severe mice and the mild mice are also being systematically tested
in drug studies to identify compounds that rescue the SMA phenotype. Many
of these studies are ongoing and will be reported shortly, however, an
interesting note was observed in valproic treated mice. The treated severe
mice did live several days longer, however, there was no detectable change
in SMN protein levels or in SMN2 RNA splicing patterns. This was interesting
because it raises the possibility that drugs such as valproic acid which
provide some protection do not elevate SMN, therefore if this drug were
combined with other drugs that did efficiently elevate SMN there might
be an even greater degree of protection. >>
back to top
Charlotte
Sumner (NIH/NINDS)
The goal of a number of potential therapeutic approaches
for spinal muscular atrophy (SMA) has focused upon increasing expression
from the SMN2 gene. The SMN2 gene is nearly identical to the SMN1 gene,
and has been envisioned as an ideal target for therapies since it encodes
an identical protein – however, most of the SMN2 protein is truncated.
There is, however, a small amount of the “good” protein made
from SMN2, unfortunately this dramatically reduced level (~10% compared
to SMN1 levels) cannot protect from SMA. To identify compounds/molecules
that stimulate higher levels of “good” protein from SMN2,
Dr. Sumner has analyzed a family of compounds called HDAC (histone deacytylase)
inhibitors. This group of drugs has previously been shown to function
by modifying the structure of a cell’s DNA. Generally, genes that
are active and are being used to make lots of protein in regions of DNA
that are not packed tightly. The degree to which DNA is compressed and
tightly packed is controlled (partially) by a family of proteins called
histones. Histones wrap DNA tightly or loosely based upon the amount of
“acetylation.” This chemical process effectively allows regions
of DNA to be turned up or down. “Deacytylation” is the process
that turns gene expression down, therefore, HDAC inhibitors would be a
means to prevent gene expression from being turned down.
Through a series of rigorous experiments that were designed to identify
the degree to which the SMN gene was bound by the histones, Dr. Sumner
identified several HDAC inhibitors that increased expression from the
SMN gene. While there are certainly going to be a number of other genes
expression patterns that will be altered by HDAC inhibitors, Dr. Sumner
was able to demonstrate some specificity for the SMN gene with the HDAC
inhibitor compounds and began to dissect the molecular mechanisms that
were actually behind the increase in SMN gene expression. Specific histone
proteins were found to be preferentially bound to the SMN gene. This is
important because different HDAC inhibitors target specific HDACs. Since
a considerable amount of data is already known regarding the specificities
of the HDAC inhibitors, this would allow Dr. Sumner and other researchers
to focus their efforts on the compounds that targeted the specific histones
associated with the SMN gene. In essence, understanding what proteins
block expression of the SMN gene, researchers can better and more specifically
target these factors to stimulate higher levels of SMN2 expression.
While the HDAC inhibitors have a well described role in changing gene
expression patterns, it was quite fortuitous that not only do some HDAC
inhibitors stimulate SMN2 expression, but there is the added advantage
that the RNA splicing is altered such that more functional SMN is produced
by a second mechanism. It is currently unclear why there is a change in
the splicing pattern for the SMN2 gene and is not directly related to
the molecular dissection of the HDAC inhibitor functions. >>
back to top
Chris
Lorson (University of Missouri)
The SMN2 gene primarily produces a protein that is slightly
truncated compared to the normal full-length SMN protein. This protein,
called SMND7, is missing an important piece of the gene called exon 7.
The exon 7 region has been shown to be involved in locating the SMN protein
to the proper places within a cell. Work from the Lorson lab now shows
that exon 7 is not specifically required. In fact, many seemingly random
pieces of DNA can compensate for the exon 7 sequences. These results raised
the possibility that if a non-specific “tail” could be included
at the end of the SMND7 protein, this would stabilize and restore functional
to the SMND7 protein.
A well characterized class of antibiotics known as aminoglycosides has
a previously described property that forces the cells protein production
machinery to continue even after it encounters or reads a signal to stop
in the gene. In the case of SMN, “stop signal suppression”
could result in a slightly longer SMND7 protein. A number of aminoglycosides
were examined and two were identified that could significantly stimulate
higher levels of SMN protein in SMA patient fibroblasts. Tobramycin and
amikacin increased SMN protein production in SMA patient fibroblasts.
This affect was detected several days after the removal of the drugs,
although there was a decrease in the levels of SMN protein at the later
time points.
A novel panel of aminoglycosides synthesized by a collaborator at Utah
State University (Dr. Tom Chang) were used in similar studies. 6 of the
22 novel aminoglycosides resulted in significantly increased levels of
SMN protein in SMA patient cells. It is currently unclear what domain
of the drugs is the active domain, however experiments are underway to
identify the “active” domain. Additionally, it is important
to note that the “read-through” product has not been detected
in these assays, and while it is a logical hypothesis that the SMN stop
codon is not being efficiently recognized following aminoglycoside treatment,
it remains to be formally demonstrated.
These experiments raise the possibility that there are other therapeutic
angles that can be explored for SMA. While the majority of therapeutic
efforts have focused upon stimulating the SMN2 promoter or altering the
splicing of SMN2, these results suggest that function can be restored
to the protein produced from SMN2. This raises the possibility that multiple
drug treatments that target separate mechanisms for SMN induction may
prove especially helpful in the treatment of SMA. >>
back to top
Darryl
DeVivo (Columbia University)
Dr. DeVivo presented an overview of SMA and how the
understanding of the clinical manifestations of the disease should be
incorporated into the development and implementation of clinical trials.
SMA results in proximal weakness which exceeds distal weakness. Clinically,
SMA presents in a broad clinical spectrum and due to its difficult to
diagnosis in the past, the incidence of likely underestimated (US SMA
birth incidence is 577/year – prevalence = ~20,000). Interestingly,
the eye muscles are spared in SMA. As the disease progress, tendon reflexes
appear to be absent and the function of the diaphragm becomes impaired.
There was some debate as to the validity that SMA patients are more intelligent
that unaffected children. The scientific documentation for this assertion
was partly based upon a comparison of SMA children and Duchene’s
patients (who normally have slightly lower intelligence scores), however,
even with that taken into account, SMA patients appeared to have higher
intelligence scores with an average IQ of 108. The basis for this difference
in intelligence is not understood.
The subacute phase of the disease was predicted to be the most likely
window of opportunity regarding therapeutic intervention. In other words,
Dr. DeVivo predicted that before SMA patients show clinical signs of the
disease, but the underlying defects have initiated, that this would be
the most promising time to establish a therapy. However, for some of the
most severe SMA cases, this pre-clinical window would likely only be detected
by prenatal screening. Currently the U.S. does not support prenatal screening
for SMA. From the work in animal models, it is not clear when SMN protein
is required and at what levels. This is an outstanding question from previous
years and remains a high priority going forward.
Dr. DeVivo highlighted a number of challenges as clinical trials are established.
For example, the phenotypic heterogeneity; fragile state of type 1 patients;
outcome measures – reliability and validity; natural history of
disease, and the clinical progression and basis for functional loss were
mentioned as issues that needed to be addressed during an SMA drug trial.
The sensitivity of outcome measures such as DEXA scanning and MUNE measures
were addressed. Additionally, the drug trial design should consider variables
such as patient population, study duration, and biological markers. >>
back to top
Doug
Kerr (Johns Hoplins)
Dr. Kerr is a pioneer in understanding how embryonic
stem (ES) cells can be used in spinal cord injuries and neurodegenerative
disorders such as SMA. Work from his lab has previously shown that mouse-derived
ES cells can be used to develop motor neurons that look and act as if
they were normal, mature motor neurons. This can be done in a dish in
the lab or in an animal. Remarkably, the ES-derived motor neurons can
restore function to a paralyzed rat. This model of spinal cord injury
is induced by a specific virus that attacks motor neurons, however, it
was not a specific model for SMA. Following injection of the ES cells,
the paralyzed rats that dragged their hind legs were able to stand and
walk. Interestingly, it was less clear whether the improvement was due
to ES-derived motor neuron function or whether the transplanted ES cells
provided support and an improved cellular environment for the existing
rat-derived motor neurons.
Additionally, transplantation techniques are being improved that allow
the mouse ES cells to grow better and fully differentiate into functional
motor neurons. For example, there is a molecule in the spinal cord called
myelin. This molecule inhibits axons from extending into portions of the
spinal cord and therefore myelin inhibits ES-derived motor neurons from
growing into the appropriate places within the nervous system. Dr. Kerr
has developed a technique that requires a series of drug treatments to
1) inhibit myelin, and 2) stimulate axon growth. If all of the drug treatments
are included, the results are dramatic, resulting in the presence of significantly
increased motor neurons, increased functionality of motor neurons, and
increased mobility in paralyzed rats.
Dr. Kerr is developing a technique to develop SMA motor neurons from SMA
ES cells. This would be an important cellular context to analyze candidate
drugs and other molecular therapies. Since animal models, including mice,
are difficult and expensive to use to analyze a large number of compounds,
the development of SMA motor neuron cultures would represent a more relevant
context for testing potential SMA therapies. Currently the SMA motor neuron
cultures are able to be generated in the lab, however, there are some
genetic abnormalities that were not expected. A number of experiments
are being performed to determine the nature of these genetic abnormalities
and whether this is specific to SMA ES cells or motor neuron populations.
>>
back to top
Jill
Heemskerk (NIH/NINDS)
Dr. Heemskerk’s presentation outlined the structure
and goals of the SMA Project, an initiative from the National Institutes
of Health. The goal of the SMA Project is to result in new interventions
for SMA clinical trials by September 2007. The SMA Project has developed
a “business-sensibility” that is designed to streamline the
time from the identification of potential compounds in the laboratory
to the testing of these compounds in clinical trials. In essence, this
collection of scientists from the pharmaceutical industry, academic research,
nonprofit organizations, and the NIH are attempting to systematically
optimize leads for trial development. To facilitate this, new discovery
tools are being developed or requested from the research community, including
motor neuron cellular models, an SMN ELISA (a rapid means to accurately
measure SMN protein levels), an inducible SMN mouse model, and an improved
mouse model that is more amenable to drug testing.
In May there will be an NIH/NINDS-contracted collection of facilities
that allow for chemical optimization of lead compounds from academia and
industry and for centralized testing of these compounds in in vitro (cells)
and in vivo (animal/mouse) SMA models. Currently, this infrastructure
is under development, but several “requests for proposals”
(RFPs) have been circulated within the scientific community (not simply
within the SMA community) and contracts are near completion.
Currently, the Project has identified two lead compounds: Phenyl butyrate
(upregulates the SMN2 promoter and SMN2 splicing), and Indoprofen (an
unknown affect upon the post-transcriptional increase in SMN2). These
and other compounds will undergo a systematic development, including medicinal
chemistry, a primary screen, a secondary screen and an additional round
of medicinal chemistry. This effort is hoped to increase the potency and
reduce the side effects of these drugs to make them suitable for testing
as treatments for SMA.
The Project website is at www.smaproject.org.
>>
back to top
Kathy
Swoboda (University of Utah)
Dr. Swoboda presented results from an ongoing natural
history study and two clinical pilot trials. The importance of establishing
reliable outcome measures was stressed as an essential component to the
success of a trial and Dr. Swoboda has previously presented her meticulous
work detailing the trial design and testing parameters. A “natural
history” study is ongoing and currently has more than 100 SMA patients
enrolled at the Utah site. This group has been studied to determine disease
progression and to develop techniques that can be used to reliably measure
function in SMA patients. Dr. Swoboda has implemented the following in
her study: motor unit number estimation (MUNE); functional motor scales;
biomarker assays; DEXA bone density, whole body lean and fat mass; and
a genotype-phenotype database for ongoing/future studies to identify modifiers.
One of the important findings from the natural history study was that
metabolic dysfunction was highly prevalent in SMA, even during periods
when SMA children were not ill. This spanned all SMA subtypes, but clearly
there is a relationship to disease severity. In a subset of patients,
carnitine (a substance required for transport of fatty acids for use as
an energy source within muscles) levels were shown to be significantly
depleted. This is important because some medications, including valproic
acid (VPA), can additionally deplete carnitine, which could substantially
increase the risk of serious toxicity using this medication. SMA patients
are clearly extremely vulnerable to worsening of their metabolic dysfunction
in the setting of fasting or illness.
In the ongoing pilot safety and tolerability studies with VPA, which began
in fall of 2003 and in which 40 patients have been enrolled, there have
been no toxic effects noted to date. Carnitine is monitored carefully,
and is indeed depleted at a much higher than usual rate. Two patients
out of 40 dropped out due to excessive weight gain. This remains a significant
concern to be addressed in future trials, since excessive weight gain
in a weak SMA patient could lead to functional decline rather than benefit.
Overall, however, VPA at the administered doses, with ongoing monitoring
and supplementation of carnitine, did not have toxic affects to date in
this initial pilot study. Functional assessments have been performed as
a part of this trial, and preliminary results will be available once the
study is concluded in December of 2005.
In ongoing pilot safety and tolerability studies with sodium phenylbutyrate
(PBA), approximately 18 patients between 0 to 2 years have now been enrolled.
No toxic effects have been noted to date when the medication is given
at the prescribed dosage. One infant was hospitalized briefly for lethargy
and decreased feeding when she was given a higher than recommended dose
of the medication. No children have dropped out of the study for adverse
effects. Ongoing recruitment of type 2 and 3 children less than 2 years
of age, and of presymptomatically diagnosed newborn infants is expected
over the next few months. Functional assessments have been performed as
part of this effort as well, and preliminary results in type 1 subjects
may be available as early as this summer.
Additional trials are actively planned for both valproic acid and sodium
phenylbutyrate (the latter in collaboration with the NPTUNE project).
The first of these trials may enroll patients as early as July, 2005.
More information will be forthcoming in the coming weeks to months. >>
back to top
Nick
Boulis (The Cleveland Clinic):
Dr. Boulis is a neurosurgeon with two primary projects
that relate to SMA. First, Dr. Boulis and others are attempting to develop
a gene therapy vector that has an increased specificity for neurons. While
the virus that he is interested in, adeno-associated virus (AAV) already
binds and enters neurons, a better and more specific interaction between
AAV and neurons is trying to be developed. Essentially the proteins that
make up the outer surface of the virus can be manipulated in the lab such
that an artificial protein is introduced onto the outside of the virus.
This non-viral protein (such as fragments derived from tentanus toxin
C-fragment, rabies G protein, or others) would not interfere with the
production of the virus but would allow AAV to achieve higher levels of
expression in neurons.
Another component of Dr. Boulis’ work involves the development of
a tool that will be used in spinal cord surgeries. Delivery of therapeutic
agents into the spinal cord such as gene therapy vectors or drugs may
prove a challenging obstacle, however, Dr. Boulis proposes to develop
a “scaffold” that will allow drugs/etc to be delivered directly
into the spinal column. The device affixes to a small section of the spinal
cord stabilizes a region of the spine. From this, a very thin cannula
(needle) can be inserted into the spine and drugs can be delivered. Importantly,
the relative position of the cannula can be determined by a series of
electrophysiological experiments which would allow the surgeon to fine
tune the position of the cannula.
Dr. Boulis also raised the possibility that additional animal models of
SMA and motor neuron degeneration would be beneficial. Since experimental
surgeries are technically challenging on small animals such as mice and
rats, a model such as pigs would allow surgical procedure and delivery
techniques to be optimized. >>
back to top