Monday, October 12, 2020

10X Triples Down on Spatial Analysis

10X Genomics last week announced the purchase of ReadCoor, a company that unveiled its 3D spatial sequencing instrument back at AGBT, paying $350M to acquire the Cambridge MA company.  This follows quickly on the heels of 10X purchasing Swedish in situ sequencing company CartaNA for another $41M.  10X already had the Visium spatial transcriptomics product on the market.  So now 10X has three different technologies in the spatial profiling space.
Why is 10X pursuing more than one technology?  It wouldn't appear that the technologies are suitable for some sort of kit-bashing mix-and-match the technologies approach, though likely their intellectual property is complementary.  It would suggest that 10X thinks that spatial profiling has a big future and most likely that no technology is likely to cover all the applications and they'd prefer to juggle multiple technologies rather than allow some of those spaces to be taken by others. So what differentiates the three?

The Visium product requires no special hardware beyond a standard imaging microscope.  Specialized slides bear barcoded oligo-dT primers for reverse transcription, arranged in hexagonal pads with a barcode for each pad.  Imaging of the slide with standard stains enables recording the features for later comparison with the expression profiles.  After reverse transcription, the cDNA fragments are released and a somewhat typical Illumina library preparation process follows.  After sequencing, 10X's software enables overlaying transcriptional information onto the original slide images.  Back in my Infinity days, this would have been a no brainer to look for hedgehog pathway expression in the tumor and stroma of our xenograft models -- and since it is an untargeted technique one can get up an running very quickly.

CartaNA's technology is targeted.  Specific padlock probes are hybridized to the sample.  For those unfamiliar with padlock probes, these are giant oligos that when bound to the target will hybridize with the 3' and 5' ends of the oligo in close proximity, allowing ligation (sometimes after a single base extension, though CartaNA appears to just ligate).  Successfully ligated padlock probes are amplified in situ by Rolling Circle Amplification (RCA). A series of in situ hybridization reactions follows; the probes in these reactions hybridize to barcodes encoded by the padlock probes.  After multiple cycles of hybridization and imaging with different probes, the identity of each spot in terms of which gene was captured is revealed.  CartaNA website links to a paper describing using the system with off-the-shelf Nikon microscope, though they are thin on any details of what effort is required to do this.  Presumably the long-term plans will include a custom instrument based around a microscope. Other publications describe using the approach for DNA mutation analysis.

It's a bit surprising to see ReadCoor exit in this way rather than an IPO, given their major presence at AGBT including the unveiling of their large instrument and beta test program.  It's a targeted sequencing approach; ReadCoor's first solution in the space is aimed at the hot field of Immuno-Oncology and covers "up to 250 genes", but they also promise future expansion to cover RNA, DNA and protein. A key claim of ReadCoor is that their resolution can be subcellular, enabling the detection of RNA or protein localization. ReadCoor's FISSEQ technology is the only one of these that claims to image in three dimensions.  FISSEQ is roughly sequencing-by-synthesis (or ligation. the company is cagy about the exact chemistry) run on rolling circle amplified targets.

For any of these, the potential exists to also detect antibodies which carry barcoding oligos, hence allowing both nucleic acid and proteomics data to be acquired in the same reaction and associated with the underlying tissue morphology.

These deals also show the benefit of doing well.  10X Genomics now has a market capitalization of over $16B and their stock price since just over a year ago has been shockingly close to being on a steady upward slope to nearly triple in value since their debut (and no, I wasn't clever enough to take advantage of this).  That means they can either use their stock as currency or can be confident that cash can be raised if needed.  Market cap is real power!

The other main player in this space is NanoString with their GeoMx instrument.  NanoString has a market cap about 1/10th of 10X's, so they wouldn't be able to be as aggressive in the acquisitions space if they were so inclined.  There are certainly more startups out there -- some very early and interestingly many in Sweden (can you productively in situ profile a meatball?).  Lucidity Bio has a website, but that doesn't say more than 'in situ'.  VizGen announced this summer early access to their MERFISH technology, which like CartaNA is based on serial probing to reveal the identity of transcripts but promises to interrogate "hundreds to tens of thousands of RNA species".  Optical Biosystems claims sub-micron resolution of up to 30 targets.  There's also Fluidigm with their Imaging Mass Cytometry system for resolving up to 37 specific proteins using antibodies, which clearly overlaps the spatial protein profiling available on these other platforms with oligo-linked antibodies. ClearLight Biotechnologies seems to offer 3D profiling of proteins as a service.  And there's almost certainly others -- I haven't tried to dredge the web for this since earlier this year and some hits, such as SeqFISH, aren't clearly offering a product.

All of these companies must be hoping to climb the usual value ladder starting with academic research followed by biopharma translational applications finally leading to the big money in clinical diagnostic use.  I can certainly remember projects where this sort of spatial ability could have been very interesting.  Given the complexity of the final analysis, it may be challenging to push some of these into the clinic, but I'm sure clever minds will think of ways to package this sort of analysis into actionable clinical insights.  

In the shorter term, 10X seems to be hedging their bets or perhaps providing a ladder -- get customers hooked on the concept with the Visium technology and it's low cost of entry and then entice customers to move up to ReadCoor or CartaNA.   The convergence of so many companies using in situ probing for barcodes may lead to some patent court fireworks or IP horsetrading, so being flush and owning one company in this space puts 10X in a strong position.  Plus they now have two companies bagged that use Rolling Circle Amplification for in situ nucleic acid analysis, building a significant patent estate in that space.  

AGBT will be virtual this year but I still hope to hear from people and companies.  !0X buying ReadCoor consolidates two sponsors of last year's meeting.  What other news will this crazy year bring to the genomics space?







2 comments:

James@cancer said...

I think what will ultimately win out is ease of access to spatial technology. Illumina made NGS easy for anyone and 10X Genomics are on the same path for single-cell and spatialomics.

I also think this is where the 10X purchase of Spatial Transcriptomics, their first acquisition in this space, is going to win the day for them agains the competition you mention. You got it right at the start of your post "Visium...requires no special hardware". Pathologists are used to processing slides and Visium could be just another method for the path lab to deliver. It also means Visium is scalable in a way that Nanostring and Fluidigm can't hope to match.

Anonymous said...

I agree that ease of use will win the day. The reason 10X is as successful as they are is that they made Single Cell Analysis easy and robust. Now with Spatial Analysis, 10X will push the technologies they have acquired to be what the mainstream market wants. In no way am I a 10X partisan, but it does seem that they have a very good handle on the mainstream market as opposed to the much smaller early adopter part of the market.

Fluidigm and Nanostring can only fill a niche and some of the other technologies have major drawbacks. For instance, when Vizgen gets around to commercializing MerFISH they will see that an expensive, low resolution, low throughput, difficult to use instrument doesn't usually fly. Also, the IP landscape is treacherous and MerFISH/Vizgen may encounter FTO issues.

Best of luck to all of these companies as they push this exciting technology forward.