0:00
And I would like to welcome our next speaker, Steve from Sciex.
0:07
He will be talking about analytical solutions for viral vectors from process to QC.
0:13
So welcome to this session.
0:15
Yeah, thank you for the kind invitation to present today.
0:19
My name's Steve Lock.
0:20
I work for Sciex now.
0:21
For over 25 years, I've been working with the industry to develop assays from the days of peptides or AstraZeneca, now to more viral vectors and more complicated assays.
0:31
I think this whole process of developing assets for viral vectors really kicked off after when the COVID pandemic started, when we were asked to develop new assays for this particular area.
0:42
I actively promote what I do.
0:44
I work all this day to today is actually being collected with customers.
0:50
So I work quite closely with collaborators to develop new methods and I can ask a favour.
0:55
First of all, can you raise your hands?
0:57
You want a copy of this deck because it does have links in to our external data.
1:02
So if you would like a copy of the deck, please raise your hands.
1:04
Also, if you'd like to link in with me, please raise your hands because that way I can share material.
1:08
Afterwards, Sarah will go round to scan your badges.
1:11
That's the best way for us to do it.
1:13
So thanks everyone for allowing us to scan your badges.
1:18
Sciex itself as a company has been around for over 50 years and we deliver develop analytic solutions for the industry.
1:24
We typically use two main techniques.
1:27
We use either mass spectrometry or more recently we use a capillary electrophoresis which is actually from Beckman Coulter.
1:35
So Sciex is part of the Danaher organisation and they purchased Beckman Coulter in the early 2010s and at that point then the CE systems with Beckman Coulter moved across to Sciex.
1:49
So some of the data here is originally being acquired on Beckman Coulter CE systems, but a lot of the data has now been acquired in system developed within Sciex itself.
2:00
I'm going to concentrate, I could speak for hours, but I'm going to concentrate specifically today on viral vectors and AAV.
2:06
Again, this is this sort of work has been driven by vaccines produced to deal with COVID and COVID pandemic, but also now these sort of type of medicines are moving into oncology and other therapeutic areas.
2:18
And they're a real good way to deliver DNA, single stranded DNA to modify cells.
2:25
So, but they also pose some challenges for the people who work in this analytical space.
2:31
You know, I've moved from peptides to proteins now to viral vectors.
2:34
Every time we go bigger, it gets more complicated.
2:37
So there's challenges, the challenges you have to tackle when dealing with these particular modalities.
2:43
And I'll discuss just two types of technology we've used to tackle some of the questions that have been asked by the industry and basically one of them.
2:51
The first half of this talk will cover a case study.
2:54
I've worked with a team in France that are CDMO, Yposkesi, the SK Pharmteco business and they actually manufacture viral vectors and they've used the technology called capillary electrophoresis to use to profile some of them and some of the critical quality attributes of the viral vectors.
3:13
This particular system is an 8-capillary system and allows us to run Multiplex samples, run 8 samples at a time.
3:20
And it means that we can high throughput screen samples.
3:24
And we've been designing methods now not to use analyse just quality control samples.
3:30
But the idea of the project I started with this company was to look at process samples.
3:35
We want to move this technology into a way it could help you to make quicker decisions on your process and not just use it for defining the quality of your final product.
3:46
And one of the benefits I said of using this system is an 8-capillary system which allows us to do some sort of QBD for the statement quality by design quite a few times from speakers because we can parallel process samples now we can look at ways to fine tune the sample prep needed for the methods.
4:03
And that's often what the team at Yposkesi did in developing the assay.
4:07
And this is not just looking at one serotype, this is looking at multiple serotypes for this particular technology.
4:13
And again, I would say we do multiple kits in this case.
4:17
I won't go through everything we do again, I don't have the time.
4:21
I'm focusing on 2 specific areas.
4:23
One's once, one we call genome integrity, that's looking at the payload inside the viral vector.
4:29
And the other one is SDS or CE-SDS with the automation of SDS-PAGE.
4:33
Again, that's looking at the proteins which make up the viral vector envelope.
4:37
Again, they can really affect the efficacy and safety of the drug or viral vector you produce.
4:47
What I'll try to do today is first of all start off with DNA and look at the single stranded DNA payload using capillary electrophoresis and then move into the proteins themselves, OK?
4:57
Again, it says I mentioned this is work done with Nadia's team in Paris and here we will try to use CE to look at the payloads within process samples coming out to the viral vector production lines.
5:13
And this is one of the things you basically want to understand, right?
5:16
You understand first of all, is my viral vector, does it contain any genome?
5:21
Doesn't contain any genome.
5:22
It's not going to be active, all right?
5:24
And it will also cause immunogenetic response to the body.
5:26
So you don't want to give empty viral vectors to people because then you're going to give them a couple of times before you get an immune response.
5:32
So that's pointless.
5:34
The next thing you want to understand is it does it contain truncated versions of the gene, OK?
5:39
Because again, therefore the information is not complete and the activity of the particular biotherapeutic will be down.
5:47
So there's a couple of things you can do here whereby CE to measure this, we use a technique called capillary gel electrophoresis.
5:55
Basically it's a capillary.
5:57
These 8 capillaries are filled with gel.
5:58
We inject the material and then the gel access a filter.
6:03
It separates out the single stranded DNA based on its size and it gives you a peak for your genome in the viral vector or no peak at all or and it also picks up size based impurities like truncated genomes just to show 1 slide.
6:20
This is just a reproducibility slide.
6:21
I have to show this because people say is it reproducible?
6:24
Yes, it is.
6:25
And this is just some looking at the capillary array to show how reproducible it is and multiple assays down the array and customers get this data.
6:35
All right.
6:35
This is taken from an application that we've done, but this is, I've seen this data routinely at a customer sites, not just at Nadia's team in Paris.
6:44
So I said this is what we were trying to do here.
6:46
We actually do have to do some prep on the samples.
6:49
OK?
6:50
Nadia's team used KingFisher Flex magnetic bees to capture the genetic material is they adapt the process.
6:58
This is all coming out an application note.
7:00
This is why I say raise your hands.
7:01
The application notes not available yet.
7:03
It's currently in our legal review for both companies.
7:06
It will come out in the next couple of weeks, so that's why I want to share.
7:10
That's all.
7:11
Has all the method information in it.
7:13
I can't have time to go for the method information today, but that's one of the reasons why I asked you to raise your hands.
7:18
But this just shows you where we're looking at now.
7:21
This is the first time CE has been used in processing samples.
7:24
That's why I thought it was quite important to show today.
7:26
And this is looking at how you can use it to track your level of your genome as it moves across from harvest to clarify to affinity elements right way through to your final DSP sample.
7:37
All right, it shows you the peaks increasing and also how you can remove the impurity genetic impurities present in your sample.
7:44
And this is a way to track your process.
7:46
All right.
7:47
So now they're starting to use CE now to check to actually track the efficiency of the process, to clean up right the viral vector samples as well as look at genome titer, the amount of active material present in it.
8:01
And again, you multiplex since you run eight samples at a time.
8:04
The typical runtime here is the order of 20-25 minutes.
8:07
So the throughput's quite high.
8:08
And This is why the stance is now applies to actually process samples.
8:14
And again, this is going just to compare to the bottom of full and empty sample, empty, no peak for quite clear peak.
8:20
So it's a really good way to measure your amount of material.
8:24
And you can also then do sizing.
8:26
Now the key thing here for sizing is we use a calibration standard or a ladder to base our sizing on.
8:33
Now, if you're going to use, if you can do LNPs, for example, which has messenger RNA using use basically a RNA standard, right?
8:43
If in this case you're using with viral vectors like AAV, single stranded DNA, please use a single stranded DNA ladder, OK?
8:51
Because you get far more accurate sizing.
8:53
This is something we tested with the group here.
8:55
And this shows you that by moving to the single stranded DNA ladder, your size becomes far more accurate.
9:01
It's actually a bigger impact as the viral vectors get bigger.
9:05
OK, we were, we're not sure yet where the ITR loop is affecting the sizing because there's a loop present here.
9:11
So we're doing a bit of work now to take the loop off and to see if the sizing gets more accurate and the larger single stranded DNA, but that's a that's say that's work ongoing by Yolande.
9:20
They're still doing this work in [unclear].
9:22
So we're hoping to get a bit more information on this space.
9:25
So that just looks at genome integrity.
9:28
The next assay I'm going to cover here is looking at viral vectors and this is looking at the proteins of viral vectors.
9:34
Now again, we're using the same system, essentially the same capillary.
9:39
What we're doing is it filling it with another a different gel, OK.
9:43
This case we're filling it with an SDS gel.
9:46
So we're automating SDS-PAGE in this case and we're still using the detection technique called laser reduced fluorescence.
9:53
We have to use this technique because it's about 100 to 500 times more sensitive than classical UV, OK.
10:01
And again, we've applied this here to a series of different viral vectors.
10:06
And here you can see the classical profile of VP3, 2 and 1. OK, you can see some pre and post peaks around the VP 3.
10:14
This is VP3 prime peaks.
10:16
Again, this sort of ratio is really important because it affects how the viral vectors deliver their cargo.
10:22
So this sort of VP3, 2 and 1 ratio is quite an important attribute to measure for looking out ability for that particle to deliver its payload into the target cells.
10:33
By the way, the bottom link here is a Technology Networks webinar which is run by myself and the team back in September, which goes through into a bit more of the data.
10:41
There's only some of the data shown in this webinar and you can see here there's a reproducibility there we're getting on sites in the BioPhase system in the samples, RSD is all below 5%.
10:52
So we know again, it shows you how reproducible this type of assay is.
10:57
Again, we want to then now start using this on process samples.
11:01
So again, we actually apply this not to quality control.
11:04
Again, it's the first study we're using this now on process samples.
11:08
And again, we've used a new clean up and B clean up.
11:12
This again will be in the tech note which is coming out.
11:15
But it shows you how you can apply this viral protein profiling now to process samples and give you an idea if the what the protein level is, what the viral particle level is, not just the genome level is.
11:28
And this can also give you an empty full ratio.
11:31
Then you think about it, for the total product you run 2 assays, two to give you lots of data, it gives you protein profile ratios, it gives you amount of protein of viral particles present.
11:43
The genome gives you amount of genomic material present.
11:46
From that to calculations you can get empty full ratios, OK.
11:52
And again, this is it gives you again really good signals for VP3 even in the process samples to go for this quite quickly.
12:00
But what the other thing they wanted to do was do sizing again, you can do sizing with genome integrity, right?
12:07
The question is everyone says to me or you can't get accurate sizing when you do SDS, CE-SDS compared to SDS-PAGE.
12:13
Actually you can just need to choose the right standard to calibrate your system.
12:19
And now Progen is a company actually produced viral protein standards.
12:22
So this is what we did.
12:24
This is a classical ladder which we used to map sizes.
12:29
But when you use that ladder, you tend to find because these are quite large proteins, around about 80 to 90 odd kilodaltons, that you can get slightly inaccurate sizing.
12:42
OK, because you're using the wrong proteins in your ladder.
12:45
If you then move to a different ladder and use this Progen standard, your sizes are far more accurate.
12:54
Yeah, just change your ladder and now there's commercial standards available.
12:58
So you're actually measuring against accurate standards and I can do correct sizing of your proteins.
13:05
Again, gives you more accuracy in your results.
13:07
Simple thing, we tested, it works OK.
13:10
Again, this will come out in an application where you have more data behind this.
13:14
So that just shown you where we're using CE in this particular area.
13:18
My final 5 or 10 minutes, I'll discuss using mass spec now because this looks at the overall profiling of proteins, right?
13:25
CE-SDS.
13:26
But what happens you want to look down to PTM levels, like post translational modification levels.
13:32
You're only getting large changes in sizes.
13:34
When you're looking at CEA, you're not looking at where a modification which happens, which is a small mass change could be phosphorylation plus 79, for example.
13:42
You have to then go down to mass spec.
13:44
And this is the other sort of set of analytical tools that we use at Sciex to analyse viral factors.
13:52
And we use our 7600+ QTOF system.
13:55
And you can have a couple of different ways to do this is intact profiling of the viral proteins, where you use a classical C4 separation bit of temperature to separate your viral proteins.
14:07
And then you look at the intact mass.
14:09
But also you could do what was called classical bottom up where you basically digest it and look up where the peptides which form the proteins and then look for where the sites are on those peptides.
14:22
This is some data, some real data from on done with the team in the US.
14:26
This is by with GeneLeap Bio.
14:28
This is the my colleague Kirsten Paul worked with GeneLeap Bio to generate this data.
14:33
And here you can see look at the intact lab.
14:34
We can pick up things like phosphorylation modifications of viral vectors.
14:38
Now these modifications are quite important, right, because they can affect the charge of the viral particle.
14:45
You start affecting charge, you start affecting that's ability to deliver the payload again.
14:50
So they sort of again, this is well known thing you need to look for is charge variance of viral proteins because they can affect how it delivers the payload.
14:58
This is looking at the intact workflow again, it does, it tells you got things like Acetyl modification, the one of the ends or you've got a phosphorylation site modification doesn't tell you where it is, just says it's there on the proteins.
15:14
So to tell you what it is, then we have to do the tryptic digest, OK.
15:18
So what we're doing now is taking the protein, we're chopping it up, we're generating peptides, and then we're looking at where that modifications happened on the peptide.
15:29
Now the beauty with this instrument is that you've got alternative fragmentation techniques, OK?
15:33
This is why we use the 7600 to do this work, Classical fragmentation techniques called CID or collision induced dissociation.
15:41
That's basically like driving your car into a brick wall.
15:44
It shatters everything to pieces instantly.
15:47
There's another fragmentation technique now which you have available.
15:49
This is called electron activated dissociation, and this is basically giving your car a bit of energy, so you're shaking it OK, and the bits fall off.
15:58
It's effectively what you're doing.
16:00
And this is a tuneable technique, so you can adjust it depending what you're looking at.
16:04
You can apply it to the intact proteins, but in this case, we're applying it to peptides, and this gives us different fragment ions because there's different way of breaking up the peptides which are diagnostic to some of these modification sites.
16:18
A good example of this is just looking at, and this is not a viral, by the way, a bit of note of caution here.
16:23
This is not a viral peptide, from a viral protein, it's from a monoclonal antibody, but it works for viral proteins as well.
16:31
This is just looking at CID, a classical fragmentation compared with EAD and you can see you get more data with this new fragmentation technique which allows you in this case to find out where the N glycans are on that peptide backbone.
16:45
OK, the location, normally with CID, they kicked off straight away.
16:49
You don't know where they were.
16:51
OK.
16:51
Now we can locate it.
16:54
And again this is another example of the Fc site.
16:56
Again, this is Etanercept [unclear] data, but it's not viral proteins.
17:00
But when I have got viral protein data is looking at phosphorylation site map.
17:04
And as I mentioned before in the first example on the intact VPs we found phosphorylation sites.
17:11
OK.
17:11
So again where are they?
17:14
Can you locate them?
17:15
And this shows an example here where we've used PTM mapping.
17:19
I would generate the peptides themselves and look for where there’s phosphor groups on the serines.
17:25
And you can see here we'll get different peaks coming out, different series of ions, which highlights the presence of these locations.
17:35
OK, another really important area as well as looking at a summarization and deamidation sites of viral vectors.
17:42
OK, if you anyone knows in proteins, right, If you go, if your pH changes and goes up, right, you can get potential deamidation happening on your backbone.
17:54
And this is again something which is a charge variance.
17:57
And that deamidation site, it's not like a phosphorylation site which you got a + 79 or something in the change of mass.
18:04
Here's a plus one, right?
18:05
It's difficult to pick up because the mass change is so much smaller.
18:09
But you can again use this technique called EAD to specifically dig down deeper and produces alternate fragment ions that typically c and z ions is what we see with this technique.
18:21
If you're into mass spectrometry, rather than the b and y ions you see classically with the other fragmentation techniques, CID, it allows you, these ions will pop up and suddenly now you can look at isomerization, which is Asp, IsoAsp, which is again a charge variant, which happens OK, and you get again.
18:40
So this is an important way to pick up this sort of fragmentation sort of PTM sites, which can really affect the charge of the protein and therefore affect his ability to load up that sort of viral vector payload into this target cells.
18:57
I'm sort of on time.
18:58
I've time for 20 minutes to do this.
19:00
Again, a lot of this work has been done.
19:03
It's a very quick overview of how you can use this technology.
19:05
I haven't covered everything we can do, but I want to thank basically the team of Nadia Avenier in Paris.
19:12
Her team of really worked me really well over the last year to develop some of this process testing, which is the first time we've done it on process samples.
19:19
And now they're implementing this technique in their manufacturing.
19:23
So specifically Juliet and Yolande, they actually are the speakers in the webinar and the technology webinar.
19:29
So Juliet did the work on the CE-SDS assay on process samples and Yolande did the work on the genome integrity.
19:38
OK.
19:38
They worked together with my colleague Marcus Heinl.
19:41
We worked as a team to develop the new assays with them.
19:45
I also want to thank my team in North America, Kirstin and Jane and Elliot's group for working with the company in North America to look at intact viral particles by MS.
19:56
We're currently looking at methods to look at charge het of viral proteins as well by CE.
20:03
And we're also potentially looking at using the same CE methods to couple that to mass spec, to look at intact profiling charge het of viral vectors.
20:13
We're playing with the sensitivity at the moment.
20:15
It works in quality control samples, but we haven't got yet to work on process samples because the concentration is a lot lower.
20:21
Watch the space.
20:23
We'll hopefully get some of that data out in the next year, but it's to it to show you.
20:27
So I haven't included in this presentation, but that's again an area we're working on look is to actually do charge het a viral vetted by CE and then couple that CE charge het to mass spec.
20:39
So that's again the next step in this evolution of technology in this particular space.
20:44
And thank you for your attention.
20:45
I'll look forward to any questions you have now.
