0:00
Thank you for the kind introduction and hello everyone.
0:09
Well, what are the current challenges and AAV titer determination? So accuracy first of all.
0:17
So you really have to determine the titer accurately, but along with this comes precision and that sensitivity.
0:26
So you really need highly precise and sensitive quantification methods.
0:33
Then comparability is a big issue, particularly regarding the methods and results which are in line with the high numbers of orthogonal methods that are on the market right now.
0:45
And of course in some settings also increasing sample numbers can be challenging and this comes along with high time pressure and hands on time which you really would like to reduce.
0:58
So what would be the ideal method then?
1:00
What features should this method have?
1:04
So it should be accurate and reliable.
1:07
Of course, with regard to assay variances, of course, precise and sensitive.
1:13
It should be suitable from R&D to clinical settings.
1:17
It has to be easy to use and establish in every lab and as I said, comparable and in settings with increasing sample numbers it should be automated.
1:28
And today I'd like to address the last two points today, comparability and I will show you some data on orthogonal methods in titer determination.
1:39
And I will show also some data on the automation of our ELISAs where you can then use them in high throughput settings. Now to the ELISA, the oldie but Goldie.
1:54
So why this is still a popular choice?
1:57
So it's accurate.
1:59
So in the kits, the kits come with a standard which is comprehensively characterised.
2:04
And in case of AAV2 and AAV8, we calibrate the kits to the International reference standard materials of the ATCC.
2:13
And in case of the other serotypes, we implemented in house a very thorough characterisation process which is according to what’s used for the ATCC standard material.
2:29
The assays are reliable.
2:31
So they have low assay variances.
2:33
With regard to inter and intra essay variance, they are comparable.
2:37
So and that was also shown in the study of the ATCC material.
2:41
It has very low interlab variances.
2:44
It's suitable.
2:45
So you can use it from crude lysate up to purified material.
2:51
It is available which means that you have no equipment and sample costs and you can use it in every lab.
2:59
So with the ELISA you have a precise and reproducible quantification of your intact AAVs, a little bit on the history of the AAV ELISA from past to present.
3:10
So earlier it was used for the characterization of the ATCC standard material and they also could demonstrate the comparability.
3:21
So very low lab to lab variance.
3:27
It's still a trusted method in AAV development and manufacturing and very often used in these process steps there.
3:35
It's present in small academic labs in startups and big pharma companies.
3:39
And it's also the method of choice for validation.
3:42
And I just chose two examples here for SEC-MALS and BLI.
3:48
We can see very good correlation for both methods with the ELISA.
3:56
So now I'm coming to the first part which I mentioned the automation.
4:00
So if you are in a high throughput screening setting, then you really are suffering maybe from the number of samples, but you would like to reduce the hands-on time, maintain at least reproducibility and accuracy of your test system and you would like to reduce the time pressure.
4:20
And I’d like to talk about two potential solutions, one of which is the Simple Plex kit from Protein Simple, which is based on our AAV ELISAs.
4:29
And then I will show you also some data on our ELISA and how we transferred it to a device from Dynex.
4:37
But also, of course, one could use other instruments like Hamilton or other devices. To the Simple Plex, which was developed in cooperation with Protein Simple, which is a Bio-techne brand.
4:51
So they developed the cartridges which are used on the other system that's actually also a sandwich immunoassay which takes place in these glass nanoreactors.
5:03
So what you do is simply load the samples onto the cartridges and then you save a lot of hands on time. As they have used in the development our antibodies and standards.
5:19
You can see on the right-hand side there's a very strong correlation between both assays, the Simple Plex and the PROGEN ELISA.
5:30
Now I'm coming to the automation.
5:32
So here we tried to really transfer our AAV9 manual protocol to a fully automated solution from Dynex.
5:43
We did it on the DSX which is a four plate handler, but a customer of us also did it for the transfer to the agility, which is a plate 12 handler.
5:54
And the instruments are fully automated as I said.
5:57
So you'd simply load all reagents to the instrument plate and the reagents and the instrument does the rest for you.
6:04
The dilutions and everything takes place in the instrument.
6:10
So we have analysed homogeneity, intra and inter assay variance and the recovery and then compared it to the manual processing.
6:22
So first of all, the data for the homogeneity.
6:24
And as you can see, so we have measured a sample across the 96 well plates and six plates and you can see the range of the CVs from around 5% up to 8%.
6:37
So the mean CV here was around 7% compared to the manual ELISA which was 2%.
6:45
So we had a significant increase here.
6:47
However, it still met our acceptance criteria.
6:51
So we were fine with the homogeneity here.
7:02
To the intra assay and recovery. Also here we measured 3 samples in different replicates and at the bottom line you can see the CVs and again they range around 6 to 8% compared to the manual ELISA where it ranges from 3 to 4%.
7:20
Also here very slight increase of the CVs and that was the beauty.
7:26
The recovery still was at around 100% and very well comparable to the manual ELISA.
7:40
To the interesting variants, here we also tested several samples on different plates and different days.
7:48
And again you see a very slight increase of the CVs ranging from around 4 to 6% for the automated ELISA and from 3 up to 4% for the manual ELISA.
8:05
So actually we could successfully transfer the manual protocol of the ELISA to a fully automated system without changing anything on the protocol and we could show a low intra and inter assay variance.
8:18
Although we had observed slightly higher CV's and OD values which you can also see in the sample curves.
8:25
There we really maintained accurate recovery which was very good.
8:33
So the accuracy and reproducibility could be maintained.
8:38
And with the system we then can enhance the efficiency in settings where you have increasing sample numbers.
8:48
OK, then coming to the second part, the orthogonal methods for capsid titer determinations.
8:56
So currently they're still increasing numbers of new methods available, they all have their pros and cons and that ELISA is often used for validation.
9:08
And we also were then interested in using these orthogonal methods and comparing them to each other with our material.
9:17
So therefore we initiated a very small study with partners.
9:21
We sent out AAV2, 5, 8 and 9 standard material and our partners blindly measured them with their orthogonal methods.
9:30
So for immunoassays, we included capillary electrophoresis on the Simple Western device from Bio-techne and also biolayer interferometry from Gator.
9:43
And with regard to biophysical assays, we used segments and dynamic light scattering from Waters Wyatt Technology.
9:54
We then always compared the results to our ELISA as a reference method and we measured the recovery and we classified the results.
10:05
So + or -20% discrepancy of the ELISA was accepted around 20.
10:13
Up to 30% discrepancy was conditionally accepted and results which were higher or lower than 30% were rejected.
10:28
To the results of the biolayer interferometry which was done by Gator on the Pro device.
10:35
So BLI measures a change in wavelength of the reflective beam due to capsule binding in comparison to a standard.
10:44
And in the study here they used our protein empty capsid standards.
10:48
So actually the same as we include in our kits.
10:52
Then they draw the standard curve and then determined from the standard curve the titer.
10:57
And you can see on the right-hand side of the table the results looking first at the CVs which were ranging from around 5% up to 10%.
11:09
And as you can see for AAV2, 8 and 9 we had very good correlation.
11:14
However, AAV5 was a little bit underestimated by the BLI.
11:23
The second method was the capillary electrophoresis on the Simple Western Jess Device by a protein sample.
11:30
Here you detect the VP protein amounts and compared to the reference standard material.
11:36
And in this case, they have used material from a third party as a reference standard material.
11:44
And please now have a look at the table on the right-hand side for the results here.
11:49
The CVs ranged from around 6% up to 14%.
11:54
And with regard to the titer recovery, we had some good correlation for AAV9.
12:01
However, AAV8 was overestimated while AAV2 and AAV5 were underestimated.
12:08
And this discrepancy might be due to the to the use of different standard materials in both immunoassays.
12:23
This is the result for the Sigma's from Waters Wyatt Technology.
12:28
So here you first separate your prep by its size and then followed by multiple detectors to measure the AAV quality attributes simultaneously.
12:37
As this is a biophysical method, there is no standard material required to determine the titer.
12:44
And as you can see here, we have quite low CVs ranging from 1 to 3%.
12:51
And with regard to the titer recovery, we had a very good correlation for AAV8.
12:59
However, very slight overestimation of AAV5 and an underestimation for AAV2 and AAV9.
13:11
And last but not least, the DLS data done on a DynaPro Nanostat II from Wyatt Technology.
13:19
Here you determine the size, size distribution and physical titer.
13:23
And again, no standard material is required.
13:28
Here we saw quite high CV values ranging from 10 up to more than 30%.
13:35
And that's why I do not show only the recovery, but also the recovery range, which is quite broad in any cases.
13:43
And as you can see for AAV5, we had quite some good correlation.
13:48
However, the titer range was very broad.
13:52
And for AAV2, 8, and 9, the titer was highly underestimated.
14:03
OK, that graph actually sums the results up again.
14:09
And as you can see also from the graph, there was no systematic over or underestimation of the title, meaning that several factors might influence the measurements.
14:20
That might be serotype specific or also method dependent factors.
14:25
However, if you have a closer look again at the AAV2, you'll see that three out of the four orthogonal methods were out of the range.
14:35
So when we know that AAV2 has a strong tendency to form aggregates, aggregations, and that might be a factor that could influence several methods here.
14:49
And of course, these serotype specific and also method dependent factors need to be analysed further if we again have a closer look to the immune assays.
14:58
So the ELISA and blue, dark blue and BLI and light blue over there, these dots you see very they are very closely together.
15:09
So here we use the same standard material.
15:11
Whereas with the other immunoassay the CE, the displayed in the grey dots over there, they are mainly off out of the range.
15:20
And here we use the different standard material.
15:23
So whenever one is using methods which are based on standard materials, you should use the same one.
15:31
And that should be comprehensively characterised.
15:38
And as I've talked so much about the standards, I just want to give you a brief introduction.
15:43
Our standards materials that we provide in our internal quality control process.
15:50
So we determine the title, the capsule titer with our PROGEN ELISA and in terms of the full capsules we do the genomic titer by PCR.
16:03
We also measure full/empty ratio by UV/Vis and CD-MS.
16:08
However, in the recent lots we also included other orthogonal methods.
16:13
So we also have a poster on that.
16:16
So if you are more interested also in the full/empty ratio analysed by orthogonal methods, come by our booth over there and we can talk a little bit more in detail about this study.
16:29
We also measure purity and silver staining.
16:33
We test endotoxin levels by spectrophotometric measurements, and we analyse the aggregation by dynamic light scattering.
16:46
With this comprehensive protocol, then we end up with the following quality attributes for our standard materials.
16:51
So the full capsids come with a viral genome load of higher 1 E12, higher than 70% full.
17:00
But this is really a serotype depending.
17:03
So most serotypes really are more than 90% full.
17:09
They have a high purity of higher than 95% low endotoxin levels that's actually lower than the detection level 1.
17:19
And we provide these full capsids for the most commonly used serotypes for the EGFP gene and the CMV promoter.
17:29
And the empty capsids come with higher 5E 12 capsids per mil, over 90% empty.
17:36
And again, a high purity, low endotoxin levels.
17:40
And also here we provide the most commonly used serotypes.
17:43
And AAV 3 is coming soon.
17:44
I've heard that they were released this week.
17:47
So all are now off the shelf and right now available after ordering.