This is a genuine translational project in which we aim to develop a combinatorial therapy for SMA. It is based on our newly discovered protective modifier CHP1; we have shown that genetically induced Chp1 downregulation together with low dose SMN-ASOs ameliorates SMA deficits in a severe SMA mouse model and in zebrafish (Janzen et al 2018).

Here we aim to use the most efficient and least toxic Chp1-ASOs in a preclinical randomized and blinded studiy in SMA mice. Moreover, human CHP1-ASOs and NCALD-ASOs will be tested in motoneurons differentiated from IPSCs paving the ground potential therapies in humans. 

Spinal muscular atrophy (SMA) is the leading genetic cause of early childhood lethality. Mutations in the SMN1 gene cause functional loss of alpha-motoneurons in the spinal cord mainly disturbing development and maturation of neuromuscular junctions. Impaired synaptic transmission causes muscle weakness and atrophy of proximal voluntary muscles (Wirth et al 2020 and Wirth 2021).

The disease severity inversely correlates with the number of SMN2 copy genes –usually varying between 1-4 copies– each producing only 10% correctly spliced RNA and protein. The SMN2 copies represent the main target for SMA therapy. Three therapies, two of which (nusinersen and risdiplam) restore correctly spliced SMN2 transcripts and one gene replacement therapy (onasemnogene abeparvovec-xioi) have recently been FDA- and EMA-approved for SMA therapy. While all three therapies show impressive results, for the majority of patients, who carry only two SMN2 copies, the SMN elevation seems to be insufficient to fully counteract motoneuron dysfunction even if administered presymptomatically (Wirth 2021).

In the past few years, we identified two SMA protective modifiers, plastin 3 (PLS3) and neurocalcine delta (NCALD), in asymptomatic SMN1-deleted individuals and corroborated the protective potency in various genetically modified SMA animal models (Oprea et al 2008, Ackermann et al 2013, Hosseinibarkooie et al 2016, Riessland et al 2017, Torres-Benito et al 2019). Moreover, we identified two further SMA protective modifier, Coronin 1C (CORO1C) and Calcineurin-like EF-hand protein 1 (CHP1) that interacts with PLS3 (Hosseinibarkooie et al 2016, Janzen et al 2018). While PLS3 and CORO1C are F-actin binding and bundling proteins and act SMA protective upon overexpression, CHP1 and NCALD are calcium sensors and act protective by downregulation (Oprea et al 2008, Ackermann et al 2013, Hosseinibarkooie et al 2016, Riessland et al 2017, Torres-Benito et al 2019). Moreover, all four modifiers restore impaired endocytosis in SMA (Hosseinibarkooie et al 2016, Riessland et al 2017, Janzen et al 2018).

Our recent work based on yeast-two-hybrid screens, identified calcineurin-like EF-hand protein 1 (CHP1) as a novel interacting partner of PLS3. Independent co-immunoprecipitation and pull-down experiments proved a direct interaction between CHP1 and PLS3. CHP1 is ubiquitously expressed, but particularly abundant in the neuronal tissue. CHP1 is a negative regulator of calcineurin, the most important phosphatase dephosphorylating the dephosphins involved in endocytosis (Janzen et al 2018).

Most importantly, we show that low-dose SMN-ASO-treated SMA mice and reduced CHP1 levels due to a heterozygous splice site mutation in Chp1^vac/wt (vacillator variant), ameliorates the SMA phenotype as compared to only low-dose SMN-ASO-treated SMA mice (Janzen et al 2018). Thus, CHP1 is a highly exciting novel protective modifier, which not only interacts with PLS3 but is a calcium sensor, similar to NCALD, and acts SMA protective by downregulation, which can be achieved by specific ASO treatment. Caveat: biallelic variants in murine or human CHP1 cause cerebellar ataxia (Liu et al 2013; Mendoza-Ferreira et al 2018) and therefore CHP1 reduction below 50% is detrimental.

Novelty and clinical relevance: This study aims to test a combinatorial therapy in SMA by using the power of protective genetic SMA modifiers. Here, we aim to restore motor neuron function using Chp1-ASOs (SMN-independent acting pathway) in combination with SMN-dependent ASOs (nusinersen) in a severely affected SMA mouse model (8). Moreover, the development and testing of human CHP1-ASOs will be a next step towards developing a combinatorial therapy in humans.

The applicant has recently received the Innovationspreis NRW 2019, for her ground-breaking work in SMA and the discovery of NCALD and CHP1 and their potential use for development of an SMA therapy.

In the present study, we evaluated the therapeutic effect of a SMA genetic modifier, Chp1, by its pharmacological downregulation using a Chp1-ASO4 in combination with suboptimal dose of SMN-ASO, the latter used to produce an intermediately affected SMA mouse model and to reduce multi-organ impairment (Munios-Bühl et al. 2022).

Chp1 modulation is a major challenge since its reduction to ~50% ameliorates SMA pathology (Janzen et al 2018), while the downregulation below those levels causes cerebellar ataxia (this work, and see previous work). A full battery of Chp1 ASOs have been developed and first tested in a cellular system and wild-type adult mice for efficiency and toxicity by IONIS Pharmaceuticals. Different concentrations of the top four leads where further tested in neonatal mice by ICV injection by us. Efficacy and tolerability studies determined that a single injection of 30 µg Chp1-ASO4 in the CNS is a safe dosage that significantly reduced CHP1 levels to ~50% at postnatal day (PND)14. A double blinded preclinical study was carried out in the Taiwanese SMA mouse model (Riessland et al 2010) using either Chp1-ASO4 or a Ctrl-ASO together with low amount of SMN-ASOs. Unfortunately, neither electrophysiological predictors such as compound muscle action potential (CMAP) or motor unit number estimation (MUNE) nor histological properties of neuromuscular junction (NMJ), spinal cord or muscle were ameliorated in SMA mice treated with Chp1-ASO4 compared to Ctrl-ASO at PND21. Surprisingly, CHP1 levels were almost at control level 4-weeks post injection, indicating a rather short-term effect of the ASO. Therefore, we re-administrated Chp1-ASO4 by i.c.v. bolus injection at PND28 and retested all pathological features in 2-month-old mice. However, no significant improvement of SMA hallmarks were seen at 2 month-of-age.

In conclusion, in contrast to the protective effect of genetically induced Chp1 reduction on SMA (Janzen et al 2018), combinatorial therapy with Chp1- and SMN-ASOs failed to significantly ameliorate the SMA pathology (Muinos-Bühl et al 2022). Chp1-ASOs compared to SMN-ASO proved to have rather short-term effect and even reinjection had no significant impact on SMA progression, suggesting that further optimization of the ASO may be required to fully explore the combination. Furthermore, our study shows that ASO-mediated CHP1 reduction below 50% induces cerebellar ataxia, similar to biallelic Chp1 mutation in mice or CHP1 in patients (Liu et al 2013, Mendoza-Ferreira et al 2018). Therefore, the development of specific Chp1-ASOs that induce cerebellar ataxia, could be of great interest to test different therapeutic agents against this pathology. In addition, it could help to better understand the function of Purkinje cells and the dysregulated network that underlies disease pathology.

Moreover, we designed and tested human CHP1 and NCALD ASOs and studied these in iPSC- derived motor neurons from SMA patients and controls and identified a highly significant increase in growth cones in NCALD-ASO treated motor neurons and an interesting underlying cellular pathway that is currently further investigated.