0:00
And we'll kick it off with the presentation of Steven Watt from he's the managing director of A&M Stabtest.
0:10
So he studied biology and earned a PhD in molecular biology and biotechnology.
0:19
And he was then had a position as a trainer and consultant for mass spectrometry for pharma development and QC analysis before joining A&M Stabtest in 2010.
0:36
So please, Steven, the floor is yours.
0:41
OK, thank you for the kind introduction.
0:50
So today I'm going to talk about QC and formulation accompanying analysis of inhalable mRNAs and yeah, oligonucleotides.
0:58
And this is an example where we looked at an antisense oligonucleotide.
1:03
So just a very few words about our companies or we are a QC companies or testing organisation dedicated to stability testing, release testing, EU release testing of pharmaceuticals from small molecules over to large molecules, mRNAs, nucleotides, but also gene cell therapies.
1:24
And yeah, we do basically everything that you need to get a drug to the market and obviously also released.
1:33
So one of our clients gave us the task to identify an inhalable mRNA formulation.
1:40
And so we had to look for workflow to test many different formulations and also set up an RNA which we can analyse quite easily.
1:54
So model RNA and also set up a lab scale production of LMP's and establish different kinds of analytical techniques to get a good overview which of these formulations work quite well in a formulated setting and which don't.
2:10
And yeah, so what we've done here since it's the easiest way to get a drug let's say into the lungs is using a nebulizer because you can easily fill it with any kind of drug that you can put into a liquid.
2:23
The only problem here is that this nebulization process itself, either using a vibrating mesh or jet nebulization stresses the sample or in this case the drug quite strongly.
2:36
So you get disruptions of the LNP.
2:38
So there's at least what you can expect.
2:41
And to a lesser extent, even sheer force is so strong that they would disrupt the mRNA.
2:46
So we had to identify different analytical methods to basically easily show us which kind of formulation is able to withstand the stresses of the nebulization process.
2:59
And all of you are involved in mRNA development probably know this USP draught guideline for analytical procedures of mRNA quality for mRNA vaccines.
3:11
So this is a list of different kind of methods that the USP recommends for development, but also for release of DS and DP.
3:22
And so we had a look at this method panel and identified those methods where we think that we need to implement so that we can quite surely say if the product has the capability of being inhaled.
3:36
So obviously content, this is quite interesting.
3:40
So here we have implemented the rival green I think, so that we get the RNA content of course and the encapsulation efficiency which gives you an overall quality of the formulation itself.
3:50
Another quality parameter which gives you the integrity of the LNPs is DLS analysis, which gives you the size distribution and polydispersity.
3:59
So knowing exactly how homogeneous the particles are.
4:06
And lastly, and this is the most important point, is potency.
4:09
So of course you can have nice particles, but if these particles are unable to transfect the cell to give you protein expression, then obviously it's not useful for anything.
4:21
So here we set up a quite easy cell based assay.
4:25
So considerations to identify a model mRNA that we would able to use.
4:29
We first of all thought of yeah, let's look on the protein level.
4:34
So we have this is something that we do for a few of our clients is to have a cell based bio assay where the readout is an MS readout.
4:42
So where we just drop the cells, extract the proteins to a digest and then measure the amounts of proteins which are actually transcribed.
4:49
But this is just too time consuming.
4:51
You need stable isotope labelled standard peptides and these kind of things.
4:56
So we dismissed this quite early on.
4:58
The other thing is green fluorescence protein.
5:01
It's also quite nice tool to use because you don't need any reagents.
5:04
But nonetheless, it's you need quite high expression to get enough fluorescence to get over the, let's say autofluorescence of the cells.
5:12
So in the end we got stuck with the luciferase assay, which is actually the easiest assay that you can do.
5:17
The only downside is it uses quite expensive reagents, but it's easy to use and it's very sensitive.
5:23
So even if you have low levels of expression, you get a quite nice signal that you can see over the background.
5:31
Then we had to choose a device for lab scale, R&D, LNP formulation to make these.
5:39
And there's a small, there's a startup, a French startup called Inside Therapeutics.
5:43
They make a device which is called the TAMARA, and the TAMARA is a really nice mixing device, easy to use, has the great advantage that has no void volume.
5:53
So all your expensive reagents that you would have in an R&D stage, there's none lost.
5:59
So everything that you put into the instrument comes out as a formulation and it uses different kinds of pre-manufactured chips.
6:06
So if you want different kind of mixing, so here we use the Herringbone mixer, but you can have jet mixers and whatever you want insert into this.
6:15
So we've thought of this workflow.
6:17
So obviously we would start with the formulation using the TAMARA system and then do a first QC check to see if we get LNPs which are of good quality.
6:28
So here the easiest thing to do is just to do normal DLS to see if you get particles and if you get particles in a correct size and in size range.
6:37
Obviously.
6:38
Then we did a dialysis to get rid of the ethanol.
6:41
And then again we did a panel of different kinds of analysis DLS again to see if the particles shrink of the polydispersity is improved.
6:50
Then encapsulation efficiency because you want to see how much mRNA is actually encapsulated within the lipid particles.
6:57
And then finally the first potency assay.
7:00
So here we included the potency assay because we were testing loads of different kinds of formulations and of course we wanted to see if it's any good in transfecting cells.
7:08
And here we use HEK cells for the transfection.
7:12
Then we took these samples which had good quality here and went on to do the aerosolization.
7:19
So we tried different kind of nebulizers throughout this study just to see if there's an impact.
7:26
Don't have the data here because that would be too much to see if they have an impact on the quality of the product that comes out of there.
7:33
And then of course after the aerosolization, we did the same kind panel of analysis.
7:38
Again, these are a few of the formulations that we've tried to just skip over this.
7:43
You can see this in detail.
7:45
If you want, I can mail you the presentation.
7:47
So after the formulation, the first thing that we've done is we've looked at the size and the pulley dispersity index.
7:54
So pulley dispersity index.
7:56
So you can see this here at 0.2 below 0.2 is good.
8:04
So most of them have been quite good when it comes to the pulley dispersity index.
8:09
So this is this light blue bar and from the size range we were looking for anything from roughly 80 to 100, a little bit above as well.
8:21
This is before dialysis was after dialysis they shrink a little bit.
8:25
So these all were more or less OK.
8:27
We discarded the F15 formulation because the polydispersity index was too high.
8:33
So this formulation didn't really formulate very well to lipid nanoparticles.
8:39
Then what we've also done is the F1 formulation, we've done 3 batches just to see how well the TAMARA system works in the setting.
8:47
So if we can get reproducible batches, so small batch to batch consistency studies.
8:53
So where you can see in light blue the encapsulation efficiency.
8:56
So here we tested the encapsulation efficiency after dialysis and the size.
9:02
So you can also if you would compare this to the others, you will see that the size decreases slightly and this is just due to the fact that the ethanol leaves the lipid nanoparticles and they start to shrink.
9:13
We had to exclude one of the formulations after the dialysis that was the formulation of 14 because it started to precipitate.
9:20
So one of the lipids that was used didn't tolerate the high aqueous environment.
9:28
So then we went to the, so the potency assay.
9:32
So we transfected the HEK cells with these different kinds of formulations.
9:36
So also you can see that these different F1, these three batches of the F1 formulation more or less had similar transfection efficiency.
9:45
We had good transaction with F9, F12 and not so good transaction efficiency with the F13 formulation.
9:54
So we discarded F10, F11, F14 we've discarded earlier on.
10:01
Then we went for aerosolization.
10:04
So in this we took these formulations which worked quite well, put them into, I think in this case it was a jet nebulizer and nebulized them.
10:13
And the fog that you get out of the nebulizer we catch with a test tube and then repeated some of the analysis again.
10:21
So here again, you can see in light blue the encapsulation efficiency.
10:25
You can see that this went down really fast.
10:28
So the encapsulation efficiency for these formulations before we did the aerosolization was roughly around 100%, so between 90 and 100%.
10:38
And what you can also see is that the size of the particles increased for most of the formulation of formulation F13 was actually quite OK.
10:47
And the encapsulation efficiency of F9 was also more or less the same.
10:56
So here these bars that have included were the values before aerosolization.
11:00
So you can see that the aerosolization process has a very strong impact on the quality of the lipid nanoparticles.
11:08
We also did, we looked at the mRNA.
11:11
So what you could also see is that the there was a decrease in mRNA concentration and also quite a lot of fragmented mRNA.
11:18
So we looked at this with CE analysis, but I didn't include this here.
11:23
So finally we went to the bioassay.
11:28
In this case, you wouldn't need to do the bioassay because already the data shows that with these very simple types of analysis that we've done here, that the LNPs obviously were disrupted and also that the encapsulation efficiency went down.
11:42
The size didn't match anymore.
11:44
But nonetheless, we did it just to show this, to prove the case.
11:48
So this is the transfection efficiency that we got prior to aerosolization and in dark blue and you can't hardly see any dark blue that was after aerosolization.
11:58
So it proves the point that these formulations that we've tested here and one of those are really good for aerosolization, so to make them accessible by the lung.
12:10
So how could we adapt this approach for antisense oligonucleotides?
12:16
So for antisense oligonucleotide, we were also first thinking, OK, we need some sort of potency assay as well, but in this case it's a little bit difficult to get generic potency assay.
12:27
So what we're currently working on is a cell that constantly expresses GFP at a quite high level and making an antisense GFP, antisense oligonucleotide so that we can do this bioassay.
12:43
But the cell line is not finished, and this is still in progress.
12:46
So what we would see here is if we do the transfection that we see a decrease, if it, if the transfection works, a decrease of the fluorescence caused by the GFP fluorescence.
12:59
But instead what we've done here in this case is because it was much easier to do is to get a greater overview of the oligonucleotide is to do LCMS.
13:08
So obviously these antisense oligonucleotides, they're rather small.
13:12
So they're easily accessible to LC-MS.
13:14
You can see in the LC side, and we'll show you next a few examples in the next slide.
13:19
So that actually works quite well.
13:21
So in this case, we've skipped the potency assay and just had a look at the antisense oligonucleotides if they are looking good.
13:29
But to spare you with all the different data that we've got here, just going to show you the LC-MS part here.
13:34
So we set up the method with an oligo T standard.
13:38
So this is just a strand, a single strand RNA which is just loads of Ts in behind it.
13:45
So from T23 to T27.
13:49
And we, here you can see the deconvoluted mass spectrum of the T25 or you go.
13:54
And what you can see here firstly, of course, is that you can easily with a chromatography separate different chain lengths.
14:03
But also via mass spectrometry, you get a really nice mass spectrum.
14:06
So here there aren't any impurities and these sodium adducts.
14:10
This is what you get from the ion source.
14:12
You can clean them out as good as possible.
14:14
So you can even further reduce doesn't that.
14:16
But that was the first shot.
14:17
It actually looks quite good.
14:20
Then we went for a real antisense oligonucleotide that is also marketed currently.
14:27
And I cut the chromatogram because it was too long, but you can believe me, it was just the one peak that you could see.
14:35
So it was actually, if you just looked at the LC chromatogram, LC UV chromatogram, we would say it's 100% pure.
14:42
But interestingly, and most of you already know this because you have done this for yourself probably is if you look at the mass spectrum is accumulated mass spectrum under the peak, you will see quite a lot of impurities that you're unable to resolve by LCO, which gets quite hard to have resolved by LC.
15:00
So this is the real spectrum.
15:03
So the un-deconvoluted spectrum.
15:04
So this is very hard to work with.
15:06
So you have to do some computer wizardry to get the mono isotopic masses of these ions.
15:14
And this is what you can see here.
15:15
And then you can quite easily see the different kind of impurities that you get.
15:19
Most of them have been in exchange from oxygen for sulphur.
15:24
So and what you can see here, this is actually the molecular ion you are looking for the intact or unmodified product.
15:32
And what you can see here is that we've at these of this size, these are a few other impurities, loads of impurities.
15:39
So when you just look at the LC chromatogram, you would say, yeah, nicely good work, 100% purity.
15:45
If you look at the LC-MS spectrum, you can see loads of different kinds of impurities.
15:50
Most of them are the same though, which you can't really resolve via the chromatogram.
15:57
Another that nice thing you can do at least in the size range of the oligonucleotide is a complete identity test by fragmentation and sequencing of the oligonucleotide sequence.
16:09
So that's also works quite nice for identity testing.
16:12
You can also see let's say exchanges for different nucleotides at different positions.
16:17
So that's also quite nice.
16:20
So just a brief conclusion.
16:21
So the TAMARA system, so if anyone of you wants to have quite high throughput, easy to use the instrument is also not very expensive system to do research grade lipid nanoparticle formulations or any kind of nanoparticle formulation.
16:39
The TAMARA system really works very nicely.
16:42
Also the guys at Inside Therapeutics, they are very accessible and they're easy to work with.
16:46
They helped us a lot.
16:48
Although the system is really easy to use.
16:50
You just need pressurised air to run the system.
16:54
Yeah.
16:54
So next time when we do this, we will go for the buffer exchange for diol filtration and not for dialysis because dialysis just takes too long.
17:04
And yeah, so another thing that we're trying to implement together with a partner is to get away from the ribo-green assay.
17:12
The ribo-green assay works quite nicely, but it's quite time consuming.
17:16
You need, well, 200 micro litres of sample.
17:19
It's not too much depending on the concentration of course, the RNA concentration.
17:24
So we're going for a different technique here.
17:25
It's a photometric technique and they are together with Repugene and the application chemist within do some AI wizardry to somehow deconvolute the spectrum to get the encapsulation efficiency.
17:43
And yeah, for oligonucleotides, LC-MS worked actually quite nice.
17:47
And for a generic cell based assay, we are setting up the system with constitutively GFP expressing cell and the anti-GFP antisense nucleotides so that we have a generic test model also for the potency assay.
18:07
And that's it.
18:08
So thank you very much for your attention and coming after lunch and the lunch coma.
18:14
And if you want to discuss more or learn more about this, you can also visit us at booth.
18:19
So thank you.
