0:00
Start with Abby Cutchin Director and Product Management of Element Biosciences where she is now focusing on the commercial strategy for AVITI24.
0:10
Previously, Abby was the oncology lead at 10X Genomics, also engaged in partnership and collaboration and today she will talk about how we can accelerate drug discovery with high dimensional multiomics.
0:29
Thank you so much for the introduction.
0:31
My name is Abby Cutchin and as mentioned, I'm on the product management team at Element Biosciences.
0:37
And if you've heard of Element, you know we're a company based in San Diego.
0:41
You might think next generation sequencing.
0:44
We came to the market about two years ago with a bench top sequencing platform, but here today, my job is to really convince you that we are really focused on so much more than just sequencing.
0:56
And we're developing a set of multiomic tools on our bench top sequencer that we see is really being enabling in drug development.
1:08
So if we think about biology and really understanding the drug discovery process, I think it's so clear that biology is incredibly complex.
1:17
There are many different molecules, RNA, protein, DNA, phosphoproteins, metabolites that are all acting in the functional unit of biology, which is the cell.
1:28
And all of these cells are organised and spatial niches and complex tissues.
1:33
And all of this complex biology is really happening in a dynamic and temporal way.
1:39
And so when we think about the suite of multiomic tools that have existed over the last 10 or 20 years, there's really been this prolific ecosystem that's been developed up for upstream of next generation sequencing.
1:53
So assays like single cell RNA seq, spatial transcriptomics, proteomics, DNA sequencing, all of these tools have really been developed to barcode different molecules and bulk samples and cells and tissue, transform them into sequencing libraries and then really harness the scale, the throughput, the efficiency of next generation sequencing as this ultimate way to read out molecules at scale.
2:18
But of course, we have a lot of complexity and all of this upfront library prep, multiple pieces of equipment, multiple assays, long workflows.
2:28
And then I think most importantly, on the back end, often times you need multiple assays in order to measure multiple aspects of your biological system, which requires you to split your samples and you're making these often times unpaired measurements across your samples, which makes it a challenge to really integrate all of this data together.
2:48
And so really our foundational insight of Element and something that we've been working on actually for the last seven years as we've developed our bench top platform.
2:58
And you'll see this in every aspect of the way that we constructed the system.
3:02
You know, our flow cell design, our chemistry, our optical system.
3:06
You know, our fundamental insight was what if it were actually as simple as adhering a biological sample on a sequencing flow cell?
3:14
Whether you're culturing cells directly on your flow cell, your centrifuging cell suspensions down on your sample, on your flow cell, or even adhering tissue sections.
3:24
And what if we could directly deposit those on our flow cell, load them onto the sequencer and then directly perform our multiomic library prep on board a sequencing system.
3:35
And I would say, this is the Holy Grail of sequencing and something that we're very excited to deliver for the first time on AVITI24, which we just launched late last year.
3:47
And so to put this into perspective, I mentioned at the beginning of the talk that we're really focused on measuring multiple aspects of biology.
3:56
And so with our assays, we're able to measure cell morphology through our element cell painting assay, which captures a lot of the classical high content cell imaging markers, everything from cell membrane, nucleus, actin, Golgi, mitochondria, ER, and simultaneously from the same run in the same sample, we also look at protein expression and RNA expression.
4:20
Now the most important thing I think is that all of this data is actually processed through our imaging system.
4:26
So for all of our protein and RNA molecules, not only do we get counts per cell, but we also get the spatial coordinates of where those are being expressed in cells.
4:36
And again, with our optical system and wide field of view imaging system, we can also capture images of the cells so that you can look at how cells are interacting.
4:45
And especially when you think about deploying this technology with tissues, really look at large sections of tissue to understand a tissue architecture.
4:54
So for our cell based assays, one thing that we also have really focused on is actually creating slide kit consumables that allow you to effectively partition the surface area of your flow cell.
5:05
So we have different formats from 12 well to 48 well to one well.
5:10
And that means that you can actually look at different time courses, different concentrations of drug, different combinations of drug.
5:17
And I think this piece is really important.
5:19
And I'll show an example here in a minute.
5:21
But when we think about high dimensional assays, often times, they're so cumbersome, they're so expensive that you know, have pretty reduced sampling where you might get a pretreatment time point and a post treatment time point.
5:34
But with an assay like this, we're actually able to collect data on the order of minutes of, cells receiving drugs, for example.
5:44
And so I'll share today just a couple of examples of how our Teton assays really span the drug development pipeline.
5:50
Specifically, I'll focus on an application for target discovery where we've enabled optical pooled screening on AVITI24.
5:57
And then also an example where we use some of our Teton panels to really uncover mechanisms of response and resistance to tyrosine kinase inhibitors.
6:08
Before I go into my first example, I just want to briefly share some of the curated panels that we've launched as of late last year.
6:16
Later this year we'll have unbiased whole transcriptome assays, but we also have a series of fixed panels covering RNA, protein and cell paint content, the ability to add custom protein and then lots of different protein add-ons to cover, different applications from cell metabolism, neurodegeneration, cytokine signalling.
6:36
And we're really just getting started in terms of building out content on the platform.
6:41
So to dive into my first example, we were interested in using the multiomic capabilities of Teton to really understand mechanisms of response and resistance to tyrosine kinase inhibitors.
6:55
So how we performed this experiment is we took one of our 12 well slide kits, we took an EGFR mutated non-small cell lung cancer cell line and we looked at a first and a third generation drug over a multiple time points from 5 minutes all the way to 24 hours.
7:11
So this is not live cell imaging.
7:12
Essentially what we did was a serial fixation over time.
7:16
And once our time course experiment was done, the off instrument workflow, again, we've completely automated library prep.
7:22
So all you have to do is make sure your cells are fixed and then you do a short assembly step to basically transform your slide kit into a sequencing flow cell.
7:32
From there we just simply load our sequencing flow cell and our cartridge, which includes all of our panel reagents onto the system.
7:38
And the runtime is also incredibly fast because the system has been optimised for ultra-fast NGS runtimes.
7:45
So we get data back in just 24 hours on the system.
7:49
So here's just a snapshot of some of the underlying structure of the data.
7:53
So I mentioned that we're measuring cell paint features, protein, RNA, all of the vertical panes here are the same field of view of cells.
8:01
We're just toggling off and on different features in our cell system.
8:06
And again, this is not live cell imaging, but we can follow ensembles of cells over, this dense sampling of different time points to see how they're changing after we incubate with drug.
8:18
So what's really exciting about all of this data, and you might have seen it in the last slide, is we actually use our cell paint features to define the boundaries of our cells with cell segmentation.
8:30
So when we look at RNA and protein and even our cell morphology features, we're capturing all of this data in the context of single cells.
8:38
So we can look at it very similar to how you look at single cell RNA seq, here that UMAP plots where every single dot on this plot is a unique cell and we're clustering based on gene and protein expression profiles.
8:51
Of course, we can also append all of the morphology data.
8:55
And what's interesting about this assay is, as part of our onboard primary analysis, we actually extract 100 morphology features per cell paint feature.
9:05
So that's over 600 cell morphology features from intensity, granularity, size, shape, texture and we append all of that with our RNA and protein expression data.
9:14
So you are truly getting paired multiomic morphology data at single cell spatial scale across many cells in your sample.
9:24
But I think where all of this becomes incredibly interesting is that again, we're looking at all of these layers of biology simultaneously in the same assay.
9:34
And so what that affords us to do is the ability to really reconstruct a very complex signalling pathways.
9:41
And again, I mentioned this earlier, often times with high dimensional assays, you're capturing a pre-treatment and maybe 24 hour post treatment.
9:49
But something that was so clear to us, and I think this has been widely appreciated by lifestyle imaging folks for a long time now, is that drug activity, happens on the order of minutes.
9:58
And so we actually see these really interesting phosphorylation changes, on the order of 5 minutes post exposure to drug.
10:06
But I mentioned that our original question was, what's the difference in response between a first and a third generation tyrosine kinase inhibitor?
10:14
And what we found is that there's actually appeared to be this dysregulation of the MAP kinase pathway in the first generation drug that was allowing cells to escape apoptosis really mediated by the cyclin D1 activation.
10:29
So all of this data was just produced with two flow cells on the system, and it drove our next hypothesis, which was, what if we could actually rescue this part of the pathway that was becoming dysregulated with a combination therapy, specifically a CD4 inhibitor?
10:45
And so as you can see, and again, these are all images that are collected on our sequencing system, which is I think pretty remarkable.
10:51
We were able, with this combination, able to rescue cell death.
10:55
And this was all very exciting to us.
10:57
You know, we're not drug developers, we're technologists.
11:00
And we did this in a very naive way.
11:01
You know, one scientist working in the lab, 4 flow cells, one week of experimental time.
11:07
But what became really exciting to us was when we actually saw this paper published in Nature where they tried this exact combination therapy just a few months before we naively stumbled upon it with our mechanistic studies.
11:22
And they found it to be an effective combination in patients.
11:25
And so we were able to in an unbiased way, really uncover in a deep mechanistic hypothesis of why this therapy would be effective.
11:34
And I really invite all of you to take a look at the preprint that we just published if you're interested in learning more.
11:40
And of course, please find us at the booth and we're happy to talk more about the study.
11:44
All right.
11:45
So in the second part of my talk, I want to next turn towards a new set of assays that we're releasing starting as early as August.
11:53
So what I've shown so far are our Teton office and on the RNA side, we're really leveraging what we call our barcoding chemistries.
12:01
So effectively, we're basically amplifying and sequencing barcodes and then mapping them back to a target on our panel.
12:08
But what I'm very excited to show today are our direct and sample sequencing chemistries, which we're launching on Teton Atlas, which for the first time on a commercial platform actually allows you to directly sequence a native transcript inside of a cell with no library prep.
12:26
You know, people have been trying to do this if you're familiar with optical pooled screening by cracking open Illumina reagents and doing manual cycling by hand on a microscope.
12:35
But we can actually do direct in sample sequencing, up to 300 cycles, 100 base pair reads, directly on AVITI24 in less than 48 hours.
12:45
And importantly, all of this is compatible with our cell painting chemistries and our protein barcoding chemistries.
12:54
So now we have to really add another dimension of our biology where again, we're not just getting counts per cell of gene expression, we're also getting a FASTQ file in our cells.
13:04
So we're getting the native RNA sequence that we can align and append back to our cell table, which is really exciting and opens up all of these new applications.
13:13
You know, the ability to sequence things like guide, RNA's, express SNPs, fusions, you name it.
13:22
There's a lot of value that comes in actually looking at the native RNA sequence.
13:28
So we have two flavours of our direct and sample sequencing chemistry.
13:31
So we'll have a three prime assay that makes use of poly-T probes to capture polyadenylated mRNA.
13:37
And of course we have targeted RNA approaches where you can design custom probes for any region of interest.
13:44
And very briefly, to describe how this works, we create a cDNA copy of our target, we circularise this, we amplify and sequence by our core sequencing chemistry.
13:54
This is the same sequencing chemistry that we use for whole axiom sequencing, whole genome sequencing.
13:59
And all of this biochemistry is happening automatically on the flow cell inside of your cells.
14:07
And here's just a snapshot of some of the data that we're able to develop with this chemistry.
14:11
You know, for our targeted RNA approaches, we've demonstrated doing, 20 base pair reads, looking at guide RNAs, we've targeted specific genes up to 100 base pairs.
14:22
And we're able to demonstrate, finding express SNPs or exon splicing.
14:26
And then of course, for an unbiased discovery lens, we have our Poly A base three prime transcriptome assays, which allow you to really do whole transcriptome inside of cells.
14:38
What's really interesting about all of this is that with these assays, we're really focused on making AVITI24 really the first fully programmable sequencing system where you know, we're moving library prep from the bench to software.
14:53
And you can design your own custom multiomics workflow and load in your cartridge with any custom primers or custom panels loaded onto the system.
15:01
And the instrument software actually completely executes your multiomics workflow in less than 48 hours.
15:08
So now I want to give my second brief example focused on using this chemistry for optical pooled screening.
15:15
So it's just a brief background.
15:17
You know, for optical pooled screening, we're looking at cells that have been edited with CRISPR.
15:23
In this case, we take a pooled cell library that was transfected with over 2000 guide RNAs.
15:28
We cultured the cells directly on our flow cell, loaded them onto the system and have all of these optical phenotyping and multiomic readout.
15:38
So for those of you that are familiar with optical pool screening, people have really been trying to do this with other in situ platforms, but you know, there's really a lot of difficulty and complexity when you're trying to use barcoding approaches for these applications, mostly because you have to design a unique barcode for every unique perturbation.
15:59
But with our direct and sample sequencing chemistry, we really leverage, a dual primer design where and sequence the spacer region.
16:06
So with just two primers, we can look it up to 80,000 unique perturbations in a single run.
16:12
And here's just some examples of some data in MCF7 and U2OS cells where we're looking at a range of guide numbers.
16:22
And I want to focus just very quickly on some data that we developed in collaboration with Genentech, where we used AVITI24 to collect all of this optical phenotyping and multi omic data as well as sequence guides in just under 48 hours.
16:39
So, here's just a snapshot of what some of the actual data looked like.
16:42
We're in the middle pane where we're looking at a single guide RNA that's being expressed in the nucleus. And again, in the same cells, we can overlay that with all of our cell painting data and then all of our RNA and protein data.
16:55
Whereas before a lot of these data sets are collected independently, either through, classical methods like PerturbView or with, single cell methods like PerturbSeq or CROP-seq.
17:05
But for the first time ever in a single assay, we're able to marry these data types together in a high throughput manner.
17:11
And so when we zoom in on this specific guide RNA that's being expressed, you can see here that we do see these RNA and protein expression changes that are associated with the perturbation.
17:22
And if we dig deeper and again, here's a heat map of all of the guide RNA that we're knocking out and all of our RNA, protein, and cell paint features.
17:30
And here if we again focus on the specific guide RNA, we can look very deeply at the different RNA and protein markers that are regulated with this guide RNA.
17:40
What was also really interesting to us is when we actually looked at anyone feature alone versus the combination of all of our different modalities, we found that.
17:49
And probably unsurprisingly, we found that cell paint RNA and protein together provide a richer view of the phenotypes that are impacted by these guide RNA modifications.
18:02
So that was a very brief example.
18:03
And again, very happy to talk to anyone after the talk.
18:07
And then in my last couple of minutes, I just want to mention a few things that are upcoming from us at Element.
18:13
So what I've described so far are really guide RNA sequencing applications.
18:17
But as I mentioned, we're also focused on enabling 3 prime whole transcriptome in situ with paired morphology and protein readouts and that will really come with our high output kit that's coming in Q4 of this year.
18:34
And so I think, there are a lot of people in the multi omics space that are working on whole transcriptome readouts at spatial resolution, both in terms of, probe based approaches, in situ based approaches.
18:47
But we I think have a very unique approach for three prime transcriptome and the fact that we're able to use our sequencing primer schemes to actually turn on different biomolecules at different batches, which helps us overcome things like optical crowding.
19:01
So even when we're trying to measure a whole transcriptome scale, we can get high sensitivity in the assay.
19:09
And to show just a very brief example, we did a study where we basically looked at four different cell types and very simply thought, can we actually get clean clustering of these cell types?
19:19
We're just using 33 batches of our three prime whole transcriptome assay, which is just, not even the full capacity of the system.
19:27
And you can see here, even with just three batches, we're able to get really clean clustering.
19:31
And so we feel that, with the Teton assays, there's a really good opportunity to significantly scale up single cell RNA-seq studies where now you have, no library prep, no PCR amplification, high scale in terms of the number of cells that you can look at per run and you get all of your paired morphology and protein data.
19:50
So we're very excited to bring this capability later in the year.
19:54
And then on my very last slide, I just want to give everyone a very quick preview of things that we are working on.
20:01
We are porting all of our direct and sample sequencing chemistries over into tissue.
20:06
And so here's just a little teaser of some three prime transcriptome data that we've developed in fresh frozen liver tissue.
20:13
And we plan to launch FFPE and fresh frozen capabilities early 2026.
20:21
And I, of course, I didn't talk at all about our sequencing capabilities, but I'm just here to remind everyone that you get all of these multiomic capabilities, but AVITI24 is also a bench top sequencer where you get 3 billion reads of output for low cost and high quality.
20:36
So really packs a big punch for a benchtop platform.
20:41
And we're so excited to talk about more applications that you might be interested in.
20:45
So please be sure to visit our booth to learn more.
20:50
And with that, thank you so much for your attention.