Monday, May 01, 2017

Oxford Nanopore's Enigmatic Patent Litigation

Oxford Nanopore has launched lawsuits in the UK and Germany against Pacific Biosciences, alleging infringement of a European patent licensed from Daniel Branton's lab at Harvard, EP1192453, which is apparently exclusively licensed to Oxford.  When I wrote about Pacific Biosciences first lawsuit against Oxford Nanopore late last year I titled it "PacBio's Quixotic Patent Litigation", as it appeared the Oxford could easily dodge the lawsuit by abandoning the 2D sequencing technology, which Oxford is in the process of doing.  I've swapped in "enigmatic" for this title, as I'm not even sure what aspect of PacBio is allegedly infringing the patent.
I will first reiterate that I don't relish digging into these patent lawsuits, but apparently now I've built an expectation that I will, and so each time it happens I get public and private messages goading me into it.  The private ones I could just ignore; I am wondering if sending a dunning notice towards Robin Hood's old bailiwick might help suppress the public suggestions.


Okay, I've dug this hole and there's no easy getting out of it.  Enough self-pity -- for the moment.  Let's look at the patent.  Titled "Molecular and atomic scale evaluation of biopolymers", the patent has a filing date of June 22nd, 2000 and publication date of April 3rd, 2002.  This is one advantage of European patents in that they are supposed to publish 18 months after filing (no, I can't explain why this is a bit longer).  So some of the submarine tricks possible in the U.S. system aren't as easy to pull off.  Note that the on-line text is from OCR, which has a few errors -- I've corrected some obvious ones such as a propensity to convert "li" to "h" as in "hnearly" and "sohd"

Claims

The patent has 70 listed claims. Reading them carefully might be expected to shed some light on the matter, right? Claim 1, as you might expect, lays out the general scheme of the device
1. An apparatus for use in evaluating a linear polymer molecule, comprising: a vessel for holding a liquid containing a candidate polymer molecule; a solid-state membrane containing an aperture therein, wherein the aperture includes an entry port and an exit port defining a channel there between, and wherein the membrane is positioned to be contactable with a liquid in the vessel; means for causing a candidate polymer molecule to hnearly traverse the aperture; and a detector for detecting time-dependent or monomer-dependent interactions of a candidate molecule with the aperture.
Claims 2 through 8 refer back to this, adding details.  In particular, claim 5 introduces the idea of "a constraining diameter" and then this is further constrained to <20 1-10="" 1-5="" 12="" 13="" 14="" 15="" 20="" 9-11="" 9="" a="" about="" and="" aperture.="" aperture="" are="" around="" art="" back="" claim="" claims="" ctually="" detector="" due="" enabling="" even="" example="" familiar="" for="" forth="" give="" his="" if="" in="" is="" laims="" less="" literature.="" loss="" meaning="" more="" nbsp="" nm.="" nm="" obvious="" of="" one="" or="" p="" patent="" placement="" prior="" quantitative="" question="" raises="" ranges="" refer="" relative="" respectively.="" set="" shape="" size="" so="" stepwise="" survival="" than="" the="" there="" to="" two="" were="" which="" with="" wording="">
Claims 16 and 17 describe possible solid state membrane compositions, with 16 covering "inorganic compounds, organic and inorganic polymers and glasses" whereas 17 claiming specific materials of "silicon nitrides, silica and alumina".  Claims 18 and 19 describe the thickness of the membrane, covering 10nm-1mm and 50nm-100nm respectively.  Claim 20 describes the capacitance of the membrane as "less than about 0.1pF" (that "less than about" phrasing again!)

Claim 21 starts discussing electrical detection, using two electrodes adjacent to aperture.  Claim 22 places the electrodes on the same side and Claim 23 on opposite sides.  Claim 24 applies to creating the detector electrodes by depositing a metal layer on the solid state membrane.  Claim 25 brings in the idea of the detector itself being the constraining diameter of the aperture and Claim 26 covers monitoring via ammeter or electrometer.

Claim 27 introduces the idea of driving a biomolecule through the aperture using either (or in combination) voltage gradient or a biomolecular motor.  Claims 28 and 29 seem a bit out-of-order, going back to describing the placement of an insulating layer adjacent to the two electrodes (28) and and the idea that the solid-state membrane would have a supporting substrate.

Claim 30 gets back to biomolecules, claiming a "polymer replicating catalyst" (i.e. a polymerase) in contact with the aperture and Claim 31 claims the polymerase if it is adjacent to, above, below or within the membrane aperture (i.e. anywhere in the vicinity).  Claim 32 introduces the idea of the polymerase itself being the constraining diameter.

Claim 31 is another long one, bringing this all together into a design

33. An apparatus for use in evaluating a linear polymer molecule, comprising: a first vessel having a first inlet therein; a second vessel having a second inlet therein an elongated cylinder having first and second ends, each end in sealing communication with the respective inlets of the first and second vessels; a solid-state membrane containing an electrically insulating aperture therein disposed in the first end of the elongated cylinder, wherein the aperture includes an entry port and an exit port defining a channel there between, and the membrane is positioned to be contactable with a liquid containing a candidate polymer molecule in the first vessel; means for causing a candidate polymer molecule to linearly traverse the aperture; and a detector for detecting time-dependent or monomer-dependent interactions of a candidate molecule with the aperture.

Claim 34 is another long one, describing the use of the design
34. A method for evaluating a polymer molecule, the polymer molecule including linearly connected monomer residues, comprising: providing a polymer molecule in a liquid; contacting the liquid with an insulating solid-state substrate having a detector capable of detecting polymer molecule characteristics; causing the polymer molecule to traverse a limited volume on the solid- state substrate so that monomers of the polymer molecule traverse the limit volume in sequential order, whereby the polymer molecule interacts linearly with the detector and data suitable to determine polymer molecule characteristics are obtained.
Back to concise claims.  Claim 35 describes using Claim 34's scheme using electrode detectors and detecting current.  Claim 36 describes setting this up in a groove on a surface, with the polymer molecule traversing the length of the groove. (Shades of BioNano Genomics!).  Claim 37 specifically calls out DNA and RNA and other polynucleic acids as analytes.

Claim 38 is another long one; Claim 39 simply extends this by saying the liquid solution "includes reagents necessary to replicate the polymer molecule" (well, duh! How else would the polymerase be able to function?)
38. The method of claim 35, wherein the detector is a metal electrode located on the substrate surface, and further comprising: a polymer replicating catalyst attached to the solid-state surface adjacent to the detector, whereby the polymer molecule is acted upon by the polymer replicating catalyst, so that the polymer molecules interacts linearly with the detector as it advances through the polymer replicating catalyst.
Claim 40 is another big one
40. A method for evaluating a polymer molecule, the polymer molecule including linearly connected monomer residues, comprising: providing a candidate polymer molecule in a hquid; contacting the liquid with an insulating solid-state membrane having an aperture therein, wherein the aperture includes an entry port and an exit port defining a channel there between; causing the candidate polymer molecule to traverse the aperture of the membrane, whereby the polymer molecule interacts linearly with the aperture and data suitable to determine polymer molecule characteristics are obtained.
Claim 41 covers solid state membranes that are electrically insulating. Claim 42 describes a membrane coated with an insulating or passivity layer.  Claim 43 describes the polymer molecule interacting with the constraining diameter of the aperture.  Claims 44-46 are again claiming size ranges for the constraining diameter of  "less than about" 20nm,  5nm  or in the range "about 1-2nm" respectively.  Claims 47 and 48 cover placement of the constraining diameter as either an an entry or exit port (#47) or within (#48) the channel of the aperture. Claims 49 and 50 specify the length of the constraining diameter as 1-10 nm or 0.1 to 5nm (overlap!).

Claim 51 is back to solid-state substrates, though specifically mentioning Claim 43, made from "inorganic compounds, organic and inorganic polymers and glasses". Claims 52 and 53 discuss membrane thicknesses again, 10nm to 1mm and 50nm to 100nm respectively.  Claim 54 talks about capacitance of the membrane of "less than about 0.1pF"

Claim 55 describes applying Claim 40 using electrical detection:
55. The method of claim 40, wherein polymer molecule interactions with the aperture are detected as electronic currents at first and second electrodes adjacent to the aperture and in electrical communication with said channel.
Claims 56 and 57 specify the type of current detected, either translational current (#55) or along the length of the channel (#57). Claim 58 applies claim 40 to detecting ionic conductance and Claim 59 to the use of a biomotor.

Claim 60 refines claims 35, 55 and 58 to detecting sequential monomer identities, the glorious process more commonly known as sequencing, using the amplitude of duration of conductance measurements. Claim 61 covers counting the number of monomers by counting the number of changes in conductance.  Claim 62 uses the duration of events to measure the number of monomers.  Claim 63 describes using the system to determine the size distribution of a population of monomers.

Claim 64 is back to a polymerase in contact with the aperture whereas Claim 65 places the polymerase "adjacent to, above, below or within" the aperture.

Claim 66 is another big claim
66. A method for evaluating a polymer molecule, the polymer molecule including linearly connected monomer residues, comprising: providing a candidate hybridized polynucleotide molecule in a liquid; contacting the liquid with an insulating solid-state membrane having an aperture therein, said aperture having a diameter insufficient to permit traversal of the hybridized molecule of the aperture; causing the candidate polymer molecule to traverse the aperture of the membrane, whereby the hybridized polymer molecule is denatured and the single-stranded polymer interacts linearly with the aperture and data suitable to determine polymer molecule characteristics are obtained.
Claim 67 describes a system in which double-stranded molecules cannot enter the aperture, but denaturing enables one strand to enter it.  Claim 68 describes performing the denaturation stepwise, allowing the molecule to advance through the channel by a monomer with each step.  Claim 69 describes using varied applied potential to control movement through the channel. Finally, Claim 70 describes slowing molecule movement through the channel being slowed by an order of magnitude if the molecule is double-stranded as opposed to single stranded.

Ooooooofff, that's a lot of claims.  And so many seem repetitive, though I'm sure that's by design.  It's too bad patent law isn't more like a programming language in which one could create operators and functions once and then apply them ("Claim 41, the operator defined in claim 2 applied to claims 10-12"). Now on to two possible interpretations as to how Oxford might be claiming infringement by PacBio, neither of which I find convincing.

Interpretation One:  ZMWs Are Inherently Infringing

The first interpretation, which a few people on Twitter have suggested (I'll leave them off here to spare them notoriety, but if they'd like credit I'm happy to give it) is that Oxford is claiming that the Zero Mode Waveguides (ZMWs) of Pacific Biosciences platform are the root of the complaint. A ZMW is a very small aperture, and PacBio's ZMWs have polymerases sitting in them. That's how they might run afoul of this patent.

But ZMWs are using optical detection, which isn't discussed in the claims of the patent. Nor, to my understanding, does the DNA or RNA molecule traverse the ZMW, going in one side and out the other.  The longer description does mention
 Aside from ionic and electronic sensing, it is within the scope of the invention to exploit the mechanical, optical, induced charging or other properties of the polymer molecule/aperture system to obtain the desired sequential record of the molecular structure as it passes through or past the aperture.
So if something is in the description but not in a specific claim, can it be enforced?  Given my complete lack of training in patent law, I have no clue!  But if this is the case, then this could be the fulcrum Oxford is trying to leverage.

Interpretation Two:  Nanopore-based ZMW Loading

The second interpretation, suggested by a private correspondent, is that PacBio has, or might be planning to, commercialize work performed with Meni Wanunu at Northeastern University in using silicon nitride nanopores to guide DNA onto the ZMWs so that high loading efficiencies are obtained. This does involve voltage-potential driven DNA traversal of nanopores made of a material described in the patent.  For PacBio, the system offers the potential of both faster loading and obtaining a higher density of productive ZMWs.

But, the nanopore isn't actually being used to measure the DNA, though perhaps that is an option.  The GenomeWeb article linked above does describe reversing the current to eject loaded molecules, so the system could potentially be used to eject small DNA from being loaded.  It might also be possible to use this system to detach DNA-polymerase complexes from ZMWs once done sequencing and reload the ZMW with a new molecule.

Implications of The Interpretations

Suppose one of these interpretations is correct: how bad is it for PacBio?  Well, that all depends on what the patent courts find, which seems to me a bit of a crapshoot.  And please, before you get angry at me for my interpretation remember I have no say in this nor is there any suggestion what I write here could influence those proceedings (I would certainly be appalled if I somehow could!).

A patent decision ruling interpretation 1 valid would clearly be a disaster for PacBio; Oxford would utterly control whether PacBio instruments could be sold in the affected jurisdictions.  So either Oxford blocks PacBio or charges a significant royalty.  But, this would seem to be a remarkably broad interpretation of the patent.  The line in the abstract about optical detection seems like a throwaway and hardly teaches anything about how to design a ZMW, and in any case ZMWs don't technically have a detector present.

Interpretation 2 is less devastating, but if this was a key part of PacBio's plan to increase throughput then that would be a serious crimp.  Again, Oxford could either block or charge a royalty.  While the nanopore-guided loading seems much closer to the Harvard patent, it is still quite different, particularly if not measuring the molecule.

I haven't made any sort of attempt to locate prior art; if I was getting paid to do this that would be the next onerous step.  There were a lot of ideas for sequencing swirling around at the fin-de-si├Ęcle; finding some others that mention apertures wouldn't be shocking.  Doing that properly means dredging for conference abstracts, patent applications (even abandoned ones) as well as literature articles.  Quite a slog, but potentially valuable if you're company's future is on the line.

Given the unpredictable nature of intellectual property litigation, it might be useful to look at this through a game theory lens.  Plaintiffs in these suits have a somewhat known downside risk -- the cost of litigation, which they could withdraw if it becomes too painful.  The upside is potentially huge, getting a big award or hobbling a competitor or rarely knocking a competitor out completely. For the defendant, that's all reversed in an ugly way: no upside and the potential for a disastrous outcome.

So if you are sued, then the obvious thing to do is examine who sued you for a possible retaliatory suit.   If you can launch one, then both of you are now locked in a fight which could prove to be nirvana or disaster or either one of you.  In such a situation, it can be very rational to settle, particularly if you need to raise money from investors to stay afloat.  By my utterly amateur analysis, PacBio is probably going to need to raise money very soon.  In their brand new Q1 2017 quarterly statement, PacBio reported a net loss of $23.9 million, reducing their cash and cash equivalents to $56.1M.  Two quarters of cash is not a happy situation (ugh, I've sunk to reading quarterly statements, a rung below reading patents!).  Without either impressive sales growth or ugly cost cutting, PacBio needs to raise money this year via either the markets or a major partnership, and neither investors nor potential partners will love the casino-like uncertainty of an open patent suit.

More broadly, this patent has obvious dangers to other players in the field.  Two Pore Guys and Genia are obvious candidates for falling under this patent.  In the case of Genia, they might wiggle free because the biopolymer doesn't traverse their pore, but otherwise their system is very close to what is described in the patent -- solid state membrane, polymerase sitting over/near a channel, electrical detection.    Two Pore Guys adds on the idea of tandem nanopores, but again we're looking at driving biopolymers through solid state pores.  Perhaps TPG's scheme for immunoassays could slip by, but they've also made noises about DNA sizing and sequencing that would seem to be square in the crosshairs of this patent. I noted also in the claims section that broad interpretation of the claims in this patent might also cast a shadow on BioNano Genomics and any other companies using nanopatterned surfaces for biomolecule analysis  -- in the extreme this could extend to Ion Torrent and Illumina.   Yikes! So for Oxford this probably isn't just about trying to even the score with PacBio, but also putting multiple competitors on notice that ONT has a weapon they're not shy to wield.  Should PacBio settle, these other companies get nearly no information on whether the ONT claims would hold up in a lawsuit. 

Well, that's my rather feeble attempt to make sense of this.  I'd much rather speculate about sequencing science than patent law, so expect much more of the former in the near future (and a dunning notice if you goad me towards the latter!).

4 comments:

Anonymous said...

All the cases don't really seem to apply to what each company are actually doing themselves. It is infuriating neither company is protecting their tech (directly I mean) just using patents as weapons to try to block the other.

Stop squabbling and why not just let us decide who we want to use!

Anonymous said...

I hate to break it to you but you have analysed the wrong set of claims. Those 70 were of the original application. The 7 granted claims of the patent described on ONT's website are found here:

https://www.google.com/patents/EP1192453B1?cl=en

Keith Robison said...

Anonymous: thanks I think that disqualifies me from any further analyses of this type :-)
(Oxford made a change in their release but I accidentally went out to the original electronic announcements).

Another disqualifying bit: I'm clueless as to how ONT effectively submarined a European Patent (filing date Jun 22, 2000 - publication date Feb 15, 2012), as I thought their process prevented this.

The seven claims are strongly related to the claims covered above, though not the same (and shorter, though digging through all the merged-in patents leads to an even bigger morass)

Claim 1 describes evaluating a polymer molecule with linear connected monomers in a liquid, the liquid in contact with "an insulating solid-state substrate" and causing the polymer to traverse a limited volume of the substrate with the polymer's monomers traversing "the limited volume in sequential order" enabling the determination of "polymer molecules characteracteristics".

Claim 2 claims an electrode as detector and current at that electrode as the detection metho. Claim 3 covers claims 1 and 2 if the limited volume is a groove in a solid-state substrate and the detector is in the groove.

Claim 4 specifies claims 1,2 and 3 applied to DNA or RNA. Claim 5 covers all these claims for a detector which is "a metal electrode located on the substrate surface". Claim 7 places a polymerase adjacent to the detector, whereby the polymerase is used to replicate the molecule being analyzed. Claim 7 gives the polymerase reagents to replicate the molecule.

So this would very clearly be targeting PacBio's ZMWs. At the point of publication, PacBio's ZMW sequencing technology was well established for multiple years and being sold, which means someone (someone else!) must sort through the history of this thing to figure out how the priority date was established.




Anonymous said...

Rookie mistake. Everyone fancies themselves a patent lawyer these days...