Therapeutic strategies for SMA

Upregulation of SMN2 through drug treatment

Aim:
To increase the levels of SMN protein in spinal cord by upregulating the SMN2 gene present in all SMA patients.

Background:
Patients with ‘mild’ forms of the disease usually have more SMN2 copies than patients with more severe SMA. This reinforces the idea that SMA is the consequence of low SMN levels and that by increasing these levels we may decrease the severity of the disease. This approach takes advantage of the presence of SMN2 genes in patients and the possibility to administer oral drugs. The mechanisms of action involves a preferential increase of SMN transcription and /or inclusion of exon 7. By increasing transcription, the total amount of mRNA and protein is enhanced whereas by inclusion of exon 7, the amount of complete transcript and protein increases. In both situations the goal is to compensate for the lower amount of SMN protein in the spinal cord in SMA patients.

Compounds chosen for this purpose particularly hinders proteins involved in transcription. This process produces mRNA using DNA as a template.

Histones are proteins that “dress” the DNA. DNA becomes “undressed” (by a chemical reaction, acetylation) to replicate and transcribe to mRNA. Leaving the DNA “undressed” for a longer time with inhibitors, helps DNA to increase transcription of mRNA and consequently protein. Two well known examples of such compounds are phenylbutyrate (PBA) and valproic acid (VPA).

Other compounds increase transcription by influencing transcription factors, proteins that help the DNA for this task.
Hydroxyurea (HU) and Salbutamol are examples of compounds that increase expression and inclusion of exon 7 of the SMN2.
All these compounds cross the blood-brain barrier (BBB) and are currently being evaluated in clinical trials.

Other potential compounds in this category of upregulators of SMN2 include Aclaurobicin, SAHA, trichostatin, 2,4-diaminoquinazolines and LBH589. However, the possible toxicity of some of these drugs precludes their use in SMA and the safety of others is still under investigation.

Upregulation of SMN2 through drug treatment – continued

Challenge 1:
In laboratory studies investigators have been able to increase SMN in blood and skin cells. We need to find out if the compounds can affect the SMN2 genes of the spinal cord motor neurons.

Solution:
The current availability of protocols to reprogram skin cells in culture (induced pluripotent stem-cells or iPS) has led to the generation of neurons and motor neurons from SMA patients. In the future, this will allow these compounds to be tested in the primary cells affected by the disease.

---

Challenge 2:
Individual SMA patients may respond differently to some of these compounds.

Solution:
Patients need to be stratified according to their in vitro or in vivo response. This may help to homogenize treatment groups.

---

Challenge 3:
These compounds are likely to affect not only the SMN genes but 2 to 10% of other genes. We do not yet know which genes may be affected and whether the modification of these genes could influence SMA or even provoke other problems.

Solution:
It would be ideal to find a compound that specifically upregulated SMN2 without affecting other genes or causing other problems.

---

Challenge 4:
Clinical outcome measures and biological markers are needed to monitor changes in muscle function, levels of messenger RNA and protein produced by the SMN2 in blood.

Solution:
International clinical research groups are presently validating functional scales. Blood analysis methods are also being validated to measure the increase of SMA mRNA and protein.

Increase the inclusion of exon 7 in SMN2 transcripts

Aim:
To increase the levels of SMN protein in the spinal cord by specifically including exon 7 in the transcripts. Short nucleotide sequences or synthetic compounds may favour the recognition of exon 7 to produce a complete protein.

Background:
The exclusion of exon 7 in the RNA from SMN2 genes (the only SMN gene that is present in SMA patients) is due to a unique nucleotide change in position 6 of this exon (cytosine in SMN1 and by thymine in SMN2). The change makes exon 7 less recognizable as an exon in the splicing process (see Figure). The incomplete protein is less stable and apparently rapidly degraded. A matching short sequence (known as antisense oligonucleotides or AONs) may be directed to exon 7 favouring the recognition of this exon in the mRNA. The result is the production of a complete transcript and predictably, a more functional protein. Alternatively, a tetracycline-like compound (such as PTK-SMA1) may be effective to directly stimulate splicing of exon 7.

Challenge:
To date most experiments have been performed in cells in culture (in vitro) and in animal models. It is not known if the compounds may be effectively delivered to the SMN2 genes of the spinal cord motor neurons.

Solution:
Different vias for administration are being developed. These include vectors and specific CSF pumps that may deliver AONs to motor neurons.

Modulation of translation

Aim:
This approach focuses on increasing the length of the SMN delta 7 protein. The strategy is to avoid the stop of the transcript and protein present in exon 8, promoting a stop-codon read-through.

Background:
DNA of patients (SMN2 gene) is transcribed to RNA (SMN delta 7 transcript) and translated to protein (SMN delta 7 protein). The SMN delta 7 protein is shorter and less stable than a complete SMN protein.

In some circumstances, the addition of aminoacids that are not usually present in the protein may be useful to stabilize and increase the functionality of a given protein. Compounds, such as aminoglycosides antibiotics or TC007, that promote the reading through stop codons, may increase the stability and functionality of the SMN delta 7 protein.

Challenge:
Experiments have been performed in cultured cells and in animal models, but it is not yet knows whether read-through based therapeutic strategies will be suitable for SMA therapy.

Aim:
Compounds that favour neuroprotection such as Riluzol, or Trophos, although they are not specific to target SMN mechanisms, may enhance the survival of MNs.

Background:
The basic mechanism of loss and degeneration of motor neurons in SMA is still under investigation. To increase survival and performance of the remaining MNs, protection from stress and toxicity and promotion of axonal sprouting (the ability of axons to form new connections in muscle fibres) can be approached. Therapeutic agents focusing on this angle are known as neuroprotectors. For example, Riluzol is a neuroprotective agent that promotes neuronal survival through glutaminergic antagonism, blocking the increase of glutamate involved in activation transmission of neurons. Trials in animal models and in type I SMA patients are promising although not yet categorical. Olesoxime (TRO19622) is an agent that prevents neuropathy in anticancer therapy and it is also under investigation in motor neurone diseases. Carnitine deficiency has been reported in SMA although it is not known yet whether this is a primary or secondary phenomenon. Given that carnitine is a natural compound present in all tissues, it has been used in both children and adults with a very favourable safety and toxicity profile. Moreover, carnitine may act as a neuroprotector and is used as a complement in children under treatment with valproic acid (given that valproic acid may cause carnitine deficiency).

Challenge:
The mechanism of protection is non-specific and it is not expected to modify the cause of the disease.

Solution:
It is possible that the SMA phenotype might be changed by SMN-independent mechanisms. Once the efficacy of these compounds is clearly demonstrated early intervention and long-term treatment protocols may be implemented. However, this therapy should not be considered alone but in combination with some other agents acting on the SMN2.

Aim:
In addition to physiotherapy, this approach envisages drugs that may improve the performance of weak SMA muscles. Alternatively, research is conducted to potentially identify new therapeutic targets to prevent or slow down the muscle wasting process.

Background:
Whilst it is clear that lower motor neurons are important disease targets in SMA, the role that muscle plays in the onset and progression of the disease remains controversial. Several lines of research support the idea that muscle is partially responsible for the disease and not only a victim of denervation (loss of nerve supply).

Although restoring SMN levels only in muscle may not be not sufficient to ameliorate the disease, this does not rule out a role for muscle in SMA. The ability to reverse muscle atrophy has therefore been highlighted as a potential therapeutic target in SMA. Follistatin for example is being studied for its role in regulation of muscle growth and has shown an increase in muscle mass and strength. Follistatin is the antagonist of myostatin that decreases muscle growth in muscle cells. Interestingly, physical exercise decreases the levels of myostatin and increases levels of follistatin in humans and rodents. We need to investigate further whether moderate physical exercise may be a good treatment option.

Aim:
The principle of this approach is to rescue the SMA phenotype by transfer of the normal version of the SMN1.

Background:
Given that the transfer of DNA alone to cells is ineffective, a vector (usually a virus) is used as a vehicle in this type of experimental work. There are currently no effective results in SMA gene transfer approaches, partly because the Blood Brain Barrier (BBB) prevents gene therapy vectors from reaching the brain and spinal cord after systemic delivery. Interest in gene therapy has recently been renewed following the development of a vector, called self complementary adeno-associated virus or scAAV9. This vector can pass through the BBB after systemic administration. Indeed, promising results of longer survival and improved motor function have been found in SMA mice after scAAV9-SMN injection in postnatal day 1.

Gene therapy is a promising approach for SMA treatment not only because of the monogenic nature of the disease (transfer of the non-mutated SMN gene) but also because it provides an opportunity to transfer non-SMN-related genes and to change the disease progression.

Challenge 1:
Efficacy of the scAAV9-SMN in severe SMA mice appears to be related to early injection, at postnatal day one, whereas it appears to be negative at later periods. However, more experimental work and protocols should be investigated to determine the efficacy of this therapy at later stages. If this therapy proves to be suitable, early treatment in humans (perhaps at presymptomatic stages) could be considered.

Solution:
Presymptomatic-neonatal screening could identify affected patients prior to the manifestations of the disease.

Challenge 2:
It is not known whether or not these AAV9 vectors could be immunogenic, in which case their effect may be neutralized by host antibodies.  Other safety issues should be addressed.

Solution:
More experimental work employing larger animal models should be provided.

Challenge 3:
It is important to define the large amount of viral genomes that are needed to treat SMA infants.

Solution:
Encouraging progress has been made over recent years by academic groups and companies in the production of viral genomes.

Aim:
This alternative considers stem cell therapy to replace MNs and to influence their surroundings and environment in the spinal cord.

Background:
Human embryonic stem cells (hESCs) offer great promise for cellular replacement strategies due to their ability to generate every cell in the body, and their seemingly unlimited ability to replicate. Human motor neurons (HMNs) have been developed from hESCs. High purity preparations of HMNs (such as MotorGraftTM) are being tested in animal models to determine whether they can work in a living system, and whether the treatment is safe. Future clinical trials in type I SMA patients and amyotrophic lateral sclerosis (ALS) patients are envisaged.

Challenge 1:
While stem cells appear to be suitable to treat localized spinal cord injury, their efficacy in a complete spinal cord full of motor neurons (MNs) that are missing or degenerated is still unknown.

Solution:
HMNs can be delivered in small animal models either by local injection or via cerebro-spinal fluid (CSF). These and other routes should be investigated in larger animal models and subsequently in humans.

Challenge 2:
It is possible that these hESCs derived cells are tumorigenic in the short or long term.

Solution:
More experimental work in animal models is needed to define the issues of safety.