McSwiggen Patents not Fundamental

Alnylam recently issued a press release regarding the upholding of a European patent (the ‘McSwiggen’ patent EP 1423406) that it had inherited from Merck as a result of the $175M+ January deal.  Given that Alnylam called the patent ‘critical for the development and commercialization of all RNAi therapeutics’, I thought it would be worthwhile pointing out that while valuable, it is by no means to be considered a fundamental RNAi trigger patent. 

The reason is that the patent claims contain a number of important limitations.  Remember that for an RNAi trigger to infringe, it would need to fulfill each of the various specifications.  Here is the main claim with my annotations in red highlight:

1. A chemically modified short interfering nucleic acid (siNA) molecule that down-regulates

expression of a target gene by RNA interference (RNAi), wherein:

a. the siNA comprises a sense strand and a separate antisense strand wherein said

antisense strand comprises a sequence that is complementary to RNA of the target gene

and wherein the sense strand comprises a nucleotide sequence that is complementary

to the antisense strand,

b. each strand of the nucleic acid molecule is independently 18 to 24 nucleotides in length

and the siNA duplex comprises 17 to 23 base pairs;

This encompasses most canonical (Tuschl-type) RNAi triggers, but not Dicer-substrates for example (means it already does not apply to Dicerna’s and Arrowhead’s Dicer-substrate RNAi triggers)

c. 10 or more pyrimidine nucleotides of the sense and/or antisense siNA strand are

chemically modified with 2’-deoxy, 2’-O-methyl or 2’-deoxy-2’fluro nucleotides,

10 or more modifications (on pyrimidines at that) is a lot, but not unprecedented.  Tekmira’s use of RNAi triggers utilizes fewer modifications since they do not rely so much on them for stabilization purposes; however, conjugate approaches such as DPCs and GalNAc-siRNAs are usually heavily modified

with one or more phosphorothiate internucleotide linkages and/or a terminal cap molecule

at the 3’-end, the 5’-end, or both of the 3’ and 5’-ends, being present in the same or

different strand.

A single, terminal phosphorothioate linkage is often employed in RNAi triggers for apparent stability reasons, although I have not actually seen a study that this is actually beneficial; self-delivering RNAi triggers often use more extensive phosphorothioates in addition to an already heavy use of modifications.  ‘Caps’ are used for stability reasons and to increase ‘guide’ over ‘passenger’ strand loading specificity. This is a nice tool, but not essential. For example a UNA (usiRNA) modification at the 5' terminus would do the same.

In sum, the upheld McSwiggen patent has just too many modifications so as to present Alnylam’s competition headaches.  Most affected will be self-delivering RNAi triggers and RNAi trigger conjugates.  In fact, Alnylam may have needed the patent the most and I’m wondering whether it also needed it to do the deal with Genzyme-Sanofi.  Remember both the Alnylam-Genzyme and Alnylam-Merck deals were basically announced on the same day.

And no, this is no evidence that Alnylam's patent estate covers 'usiRNAs' as the company went to pains to point out in the press release.