Optimization of Morpholino Antisense Oligonucleotides Targeting the Intronic Repressor Element1 in Spinal Muscular Atrophy

Abstract

Loss of Survival Motor Neuron-1 (SMN1) causes Spinal Muscular Atrophy, a devastating neurodegenerative disease. SMN2 is a nearly identical copy gene; however SMN2 cannot prevent disease development in the absence of SMN1 since the majority of SMN2-derived transcripts are alternatively spliced, encoding a truncated, unstable protein lacking exon 7. Nevertheless, SMN2 retains the ability to produce low levels of functional protein. Previously we have described a splice-switching Morpholino antisense oligonucleotide (ASO) sequence that targets a potent intronic repressor, Element1 (E1), located upstream of SMN2 exon 7. In this study, we have assessed a novel panel of Morpholino ASOs with the goal of optimizing E1 ASO activity. Screening for efficacy in the SMNΔ7 mouse model, a single ASO variant was more active in vivo compared with the original E1(MO)-ASO. Sequence variant eleven (E1(MOv11)) consistently showed greater efficacy by increasing the lifespan of severe Spinal Muscular Atrophy mice after a single intracerebroventricular injection in the central nervous system, exhibited a strong dose-response across an order of magnitude, and demonstrated excellent target engagement by partially reversing the pathogenic SMN2 splicing event. We conclude that Morpholino modified ASOs are effective in modifying SMN2 splicing and have the potential for future Spinal Muscular Atrophy clinical applications.

Figure 1

Designing and identification of lead ASO variants targeting the intronic repressor Element1. (a) Schematic representation of the design of various Morpholino-modified ASO targeting the intronic repressor Element1 (E1^MOv01 through E1^MOv12ASO). The design of the previously published E1^MO is illustrated above the E1 repressor region (blue). (b) Severe SMNΔ7 mice showed a range of longevity after injections with various E1^MO oligonucleotides. Kaplan–Meier survival curves were constructed from the various treatment groups following ICV delivery on P1 of each ASO (2 µl of a 40 nmol/l solution). Log-rank (Mantel-Cox) statistics were applied for comparisons between lead candidates (E1^MOv10 and E1^MOv11) where P < 0.0001 compared with SMA noninjected animals (top). Bar graph showing the mean survival for each treatment group where (*) indicates P < 0.05 (bottom). (c) Weight gain of SMNΔ7 mice injected with ASO variants (E1^MOv01 – E1^MOv12). Total body weight was measured daily for all animal groups postinjection. (d) Scatter plot of time-to-right (TTR) performance of mice injected with all the E1^MO variants. A late-stage in disease progression, TTR values are shown for P12. ASO, antisense oligonucleotide; ICV, intracerebroventricular; SMA, spinal muscular atrophy; SMN, survival motor neuron.

Figure 2

Increase in full-length SMN transcript and SMN protein levels after E1^MOv11 treatment. (a) RT-PCR image of endogenous SMN1/2-derived transcripts from HEK293T cells and (b) in spinal cord tissues from three SMNΔ7 mice. SMN2 and SMN1 transcripts are clearly distinguishable after reaction products of the cell culture extracts were digested with restriction enzyme DdeI. Plasmids containing the full-length or SMN▵7 products, pCI-▵7 and pCI-FL, respectively, were used in PCR reactions to generate size standards. Injection of the E1^MOv11 variant targeting the Element1 repressor increases total SMN protein in brain (c) and spinal cord (d) tissues of SMNΔ7 mouse model. Western blots (n = 5) for each treatment group were performed on tissues harvested at P7. (e) Western blot quantification. Western blot (n = 3) from brain and spinal cord tissue of five (5) ICV injected animals with E1^MO and E1^MOv11. Bar graph showing significant percent increase in SMN protein induction compared to the non-injected SMA control group. Significance was calculated using Student's t-test, where ns = nonsignificance, ****P ≤ 0.01, ***P ≤ 0.001. RT-PCR, reverse transcriptase-polymerase chain reaction; ICV, intracerebroventricular; SMA, spinal muscular atrophy; SMN, survival motor neuron; HEK, human embryonic kidney.

Figure 3

E1^MOv11 transient transfection results in a significant improvement in full-length SMN expression in induced pluripotent stem cell (iPSC) derived motor neuron cultures. (a) An increase in full-length SMN transcript after E1^MOv11 treatment as measured by RT-PCR. (b) Quantification full-length SMN transcripts from iPSC derived motor neurons. Two concentrations of E1^MOv11 resulted in significant increase of full-length SMN mRNA expression and a decrease in SMN▵7 mRNA when compared with sham-transfected cultures. *P ≤ 0.05, **P ≤ 0.01, ***P ≤ 0.001. (c) Representative images of iPSC derived motor neuron cultures. Astrocytes within the cultures were stained for GFAP (red). Cytoplasmic expression and nuclear localization of SMN (green) was increased with E1^MOv11 treatment. Nuclei were labeled with Hoechst (blue). Scale bar = 10 μm. RT-PCR, reverse transcriptase-polymerase chain reaction; SMN, survival motor neuron; GFAP, glial fibrillary acidic protein.

Figure 4

Assessment of muscle pathology. (a) Representative pictures of gastrocnemius muscle cross-sections. Gastrocnemius muscles were dissected from P12 mice and cross-sections were stained with hematoxylin and eosin and analyzed by microscopy. (b) Mean fiber size was quantified using the Metamorph (Version 6.2r4) software. Data are presented as mean ± SEM. (c) Muscle fiber distribution was analyzed for each group. Both E1^MO and E1^MOv11 had significantly fewer small fibers and significantly higher percent of larger fibers than the SMA noninjected controls. Significance was calculated using Student's t-test, where ***denotes P ≤ 0.001. SMA, spinal muscular atrophy.

Figure 5

Improvement in neuromuscular junction (NMJ) pathology. The longissimus capitis (LC) and the splenius capitis (SC) muscles from ICV injected and control animals at P12 were immunostained for nerve terminals with antineurofilament/antisynaptophysin (Nerve/Syn) (in green) and motor endplates with α-bungarotoxin (in red). While the untreated SMNΔ7 mice displayed typical severe denervation, E1^MOv11 treatment substantially restored NMJ's pretzel-like structures. (a) Selected high-magnification images of NMJs in splenius muscles of control and SMNΔ7 mice are shown. Arrows indicate occupied endplates in both treated and unaffected animals contrasting the empty endplate seen in untreated controls. (b) Quantification of the number of functional NMJs from LC and SC muscles from healthy unaffected mice, SMNΔ7 noninjected controls and E1^MOv11-treated SMNΔ7 mice (n = 3 per group) where ∗∗∗indicates P ≤ 0.001). (c) Quantification of the percentage of fully and partially innervated NMJs in the LC and SC muscle groups (Unaffected n = 6; SMA noninjected n = 6; E1^MOv11-treated n = 6). ICV, intracerebroventricular; SMA, spinal muscular atrophy; SMN, survival motor neuron.

Figure 6

Dose dependent phenotypic improvement of SMNΔ7 mice injected with E1^MOv11. (a) The higher dose treated animals lived longer. All concentrations increased survival compared with the untreated mice. The Kaplan–Meier survival curve depicts the correlation between longevity and dose dependent E1^MOv11 treatment (top). A bar graph showing the dose dependent increase in average survival of animals treated with E1^MOv11 (bottom). Box within bar shows the average life span (days) per treatment group. (b) Significant increase in the average weight of SMNΔ7 animals after treatment with different concentrations of E1^MOv11. (c) Time-to-right test results shows that SMNΔ7 mice treated with E1^MOv11 at higher doses perform substantially better. Animals injected with suboptimal doses exhibit less muscle control and turn slower than mice injected with higher concentrations of E1^MOv11(Two-way analysis of variance P = 0.0122 for lowest dose and noninjected SMA groups; P < 0.0001 for highest dose and noninjected SMA groups). SMA, spinal muscular atrophy; SMN, survival motor neuron.