How Spinraza (nusinersen) and Evrysdi treat Spinal Muscular Atrophy (SMA) through splice-switching

Nusinersen (trade name “Spinraza”) has really done wonders for patients with the debilitating genetic neuromuscular disorder Spinal Muscular Atrophy (SMA), the most common genetic cause of infant death in the US. Nusinersen isn’t your typical type of drug - it isn’t “just” a “small molecule” compound like a little chemical. Instead, it’s a strand of modified RNA letters called an antisense oligonucleotide (ASO) that helps patients make functional protein from a “backup” version of the problematic gene. The science behind it was largely worked out and developed by Dr. Adrian Krainer at CSHL. I got to see him pretty frequently (he was on my thesis committee and I even rotated in his lab).
This might be the first some people have heard about the drug, but I’ve been fascinated with it for several years - and even wrote a post about it…. And now I've updated it with info about a couple other SMA drugs that have since come on the market - Zolgenzma and Evrysdi.
blog form : bit.ly/spinrazasma
Proteins are like “molecular robots” that can perform specialized tasks in your cells. In SMA, the instruction manuals for making a “robot” called “Survival of Motor Neuron” (SMN1) protein are defective. SMN is important for the nerve cells that talk to your muscles (motor neurons), so patients with SMA, unable to make functional SMN1, have progressive muscle weakness & respiratory failure.
Nusinersen (Spinraza™️) was the 1st FDA-approved drug to treat SMA & it works by “overriding” cells’ natural “auto-correct”. Biochemically-speaking, it’s an AntiSense Oligonucleotide (ASO) that alters the RNA splicing pattern of a “backup” version of the SMN gene to produce functional protein to make up for a protein SMA patients lack. It’s important to note that ASOs do NOT affect the DNA, so it’s NOT gene editing - instead, it works at the RNA level. Let me explain…
Proteins (our molecular robots) are made of building blocks called amino acids & the “instruction manuals” for putting them together are called genes. These genetic manuals are written in the language of DeoxyriboNucleic Acid (DNA). Lots of manuals are collected together in long, coiled-up pieces of DNA called chromosomes & housed in a membrane-bound compartment in your cells called the nucleus.
When you need to make a robot, you 1st have to make a copy of its gene - transcribe it into the similar RiboNucleic Acid (RNA) language, then take this messenger RNA (mRNA) out of the nucleus & into the cytoplasm (general cellular interior) where it can be translated into a protein (the robot is born!)
BUT before you can take the RNA copy out of the nucleus, you have to edit it. Genes are often broken up into parts; some parts (exons) have instructions for making different parts of the protein (e.g. 1 exon might code for a robot’s leg & another for its arm). In between these EXpressed exons are INTerrupting introns. Introns don’t have “building instructions” (e.g. put a screw here) instead they’re like “margin notes” that provide regulatory info like when to transcribe a copy. These introns get edited out through a process called RNA splicing, which turns pre-mRNA (which has exons & introns) into mature mRNA (exons only)(to fully mature mRNA has to get a 5’ methyl-G cap & 3’ polyA tail that tell the nucleus it’s okay to let it out and help it get translated once it gets out).
Because exons are spaced apart, they can be spliced together in different ways (e.g. you can splice out (exclude) some exons together w/introns) to make different mRNAs & therefore different proteins from the same gene. Such alternative splicing can be really useful. e.g. maybe your robot doesn’t need x-ray vision for this task, so don’t waste resources including it this time. BUT sometimes it can “make mistakes” & accidentally cut out important exons &/or fail to cut out introns. more on splicing here: bit.ly/altsplicing
Normally, you have 2 copies of each volume of manuals (chromosome), one inherited from each biological parent, so you have 2 copies of instructions for each robot. SMA is a recessive disease meaning that BOTH the instruction manuals for a robot are defective - there are mutations in both of the SMN1 genes.
Sometimes in the course of evolution, genes get duplicated so that you have multiple DNA copies of an instruction manual. These duplicated versions are called paralogs & they can get altered subtly or dramatically to make new proteins. SMN1 has a paralog called SMN2. It’s almost identical, and if a full-length version of it’s made, it can compensate for missing SMN1. BUT a single nucleotide substitution (a swap of 1 letter in the instruction manual) causes one of SMN2’s 8 exons (exon 7 (E7)) to be spliced out most (~90%) of the time (even in healthy people). This gives you a truncated version of mRNA that, when translated, leads to a truncated protein (SMN2Δ7) that’s recognized as defective & degraded by the cell’s “quality control.”
Finished in comments

[@thebumblingbiochemist]

It’s kinda like 2 pages of the instruction manual got stuck together so you think you’re only tearing out the margin notes BUT you’re really also taking out important instructions. So now when ribosomes go to put the robot together, the robot’s missing an important piece and thus it can’t function (it’s “missing a leg”) Most of the time, the pages are stuck together (so no okay mRNA). But occasionally, ”the pages come apart” and the exon doesn’t get spliced out, so you get full-length mRNA so you get full-length, functional SMN2 protein that can compensate for the missing SMN1.⠀
⠀
People can have multiple copies of SMN2 & the more copies of SMN2 that an SMA patient has, the less severe the disease usually is because, even though most of the time each copy produces nonfunctional product, sometimes it produces full-length, functional product and with more copies you get more of these “sometimes.”⠀
⠀
What determines whether E7 gets included? RNA splicing is carried out by a protein/RNA team called the spliceosome, which is like an editor & it gets help from your cell’s “autocorrect” helpers. These helpers are stretches of the pre-mRNA (like words) which can act as silencers or enhancers by making it fold in ways that hide or “unhide” splice sites (cis-regulation) and/or recruiting additional protein helpers (trans-regulation). Activators recruit the spliceosome (hey, over here!) & repressors help “hide” the site so the spliceosome ignores it (nothing to see here, move along… )⠀
⠀
The combination of these positive and negative cis & trans factors tells the spliceosome where (and where not) to cut. BUT as you well know if you’ve ever had “solvation” “corrected” to “salvation” or “assays” “corrected” to “asses” (that one was awkward…) autocorrect isn’t always correct…
⠀
You can think of that single letter swap in SMN2 as a “typo” in an enhancer site that your cells “incorrectly autocorrect” by removing an exon, autocorrecting SMN2 (full-length) to SMN2Δ7 (truncated). Sometimes autocorrect “misses” finding the typo, so full-length protein gets made, but this doesn’t happen often enough to keep SMA patients healthy. Nusinersen to the rescue! Nusinersen “overrides” the autocorrect so that the typo is “accepted” this time and full-length SMN2 is made, which prevents disease progression.⠀
⠀
What’s going on at the molecular level? Nusinersen is an antisense oligonucleotide (ASO) - it’s a synthetic, single-stranded nucleic acid (like RNA or DNA but chemically modified to last longer). Its sequence is designed such that it binds to an intronic splicing silencer (ISS) site named ISS-N1 located in Intron 7 (In7) just next to the “end” of E7. This alters RNA folding (cis-regulatory effect) & prevents a repressor from binding (trans-regulatory effect) -> splice site is now recognized -> exon & intron are split up -> when intron gets removed, exon stays put.⠀
⠀
Since the drug binds an intron, it’s removed when the intron’s removed and thus doesn’t get in the way of exporting the mRNA into the cytoplasm or making the protein. And since its binding is sequence-specific, you don’t get much in terms of off-target effects. ⠀
⠀
BUT you’re not adding the spelling to your autocorrect’s dictionary - you’re not clicking “learn spelling,” you’re just saying “ignore it THIS TIME.” So each time you want to make a copy you have to click “ignore spelling,” and the drug therefore needs to last a long time. RNA’s not very stable and your cells have mechanisms to recognize & destroy “foreign” RNA to protect you from things like viruses, so some chemical modifications have to be made to the drug to make it last longer…⠀
🔹The backbone of RNA & DNA differ at the 2’ site of the sugar ring, where RNA has an -OH & DNA just has an -H (hence “deoxy”)⠀
🔹🔹 Nusinersen has neither. Instead it has an -O(methoxyethyl)(MOE) group(-O-CH₂-CH₂-O-CH₃), which protects it from nuclease degradation (keeps it from getting chewed up by proteins called nucleases)⠀
🔹Further protection is provided by a swap at the backbone’s 5’ site. Nucleic acid letters (nucleotides) usually join together by sharing a phosphate (PO₄⁻) group. In nusinersen, one oxygen (O) is replaced with sulfur (S) in a protective phosphorothioate (PS) linkage⠀
⠀
Nusinersen is delivered intrathecally (through spinal injection), because it can’t pass blood-brain-barrier. And even w/these protective modifications, it can’t last forever, so patients still have to get injections every few months.⠀

[@thebumblingbiochemist]

It’s easy for biomedical “breakthroughs” to get “overhyped.” BUT this drug really is incredible (& I’m not just saying that because I rotated in Dr. Adrian Krainer’s lab here at Cold Spring Harbor Laboratory (which developed the drug in collaboration w/Ionis pharmaceuticals) & he’s on my thesis committee!) The FDA approved Nusinersen for SMA treatment in December 2016. It’s been shown to dramatically slow disease progression in infants & young children. ⠀
⠀
ASO & other splicing-targeting therapies have potential for other diseases as well, & hopefully further research will lead to more treatments, lower costs, & greater access to care for patients with a wide range of conditions!⠀
⠀
Another splicing modulator for SMA, a small molecule drug called Evrysdi (risdiplam), was approved in 2020 for SMA patients 2 months and older. It can be taken orally, which expands its accessibility, but has to be taken every day. Similarly to nusinersen, it works by promoting the inclusion of exon 7 and thus the production of full-length SMN2 by binding to the SMN2 pre-mRNA. However, it does so in a different way. A study (of a close analog) found that it binds to a sequence in exon 7 to trigger a slight conformational change (shape shift) in the pre-mRNA and promotes the binding of splicing activators - by binding both the pre-mRNA and the activator, it helps get the activator stuck where you want it so that splicing occurs at that junction. www.pnas.org/content/115/20/E4604 It may also have additional mechanisms of action journals.sagepub.com/doi/pdf/10.1177/2633105520973985 ⠀
⠀
A third SMA treatment that has come to market recently (it was actually 2nd to market) is Zolgensma (Onasemnogene abeparvovec). Instead of an ASO, Zolgensma is a gene therapy that uses a harmless virus (in this case an adeno-associated virus called AAV9) as a vector, or vehicle, for getting a normal copy of SMN1 into patient cells. It’s approved for SMA patients under 2 years of age and is given as a single (very expensive) IV infusion. dss.mo.gov/mhd/cs/advisory/rdac/files/evrysdi-new-drug-fact-blast.pdf ⠀
⠀
here’s an article with some more info: www.biopharmadive.com/news/spinal-muscular-atrophy-drugs-hard-choices/586393/ ⠀
⠀
note: I am NOT a medical doctor (I’m the PhD kind), and this post should NOT be construed in any way as medical advice. If you or a loved one have SMA & are interested in learning more about any of these drugs, please talk to a doctor⠀
_⠀
more on topics mentioned (& others) #365DaysOfScience All (with topics listed) 👉 bit.ly/2OllAB0

[@mateuszcielas3362]

in case of other gene therapies, how it is possible to work well when mRNA that is provided is degraded rather quickly? In case of Zolgensma it is circular form, are there any other methods?

[@thebumblingbiochemist]

​@@mateuszcielas3362 Zolgensma is delivering a DNA version, which helps with the stability issue.

[@mateuszcielas3362]

@@thebumblingbiochemist yeah i just read, but i read than in other therapies they use RNA, do you know how do they stabilize it?

[@thebumblingbiochemist]

there are various modifications they can make, but they still would then need to be administered on a regular basis