0:23
It's my pleasure today to present bYlok Technology from Lonza.
0:29
Today I will be discussing bYlok bispecific pairing technology to solve heavy-light chain mispairing in bispecific antibodies.
0:44
To set the scene, the number of bispecifics entering the clinic is rapidly growing.
0:52
Since 2011, over 400 bispecific molecules have successfully entered clinical trials.
1:02
In 2023 alone, we saw an introduction of 120 new molecules.
1:12
The benefits of bispecific antibodies are pretty clear.
1:16
The development of bispecific antibodies, the trend is increasing.
1:22
However, the production of bispecific antibodies still can be challenging.
1:31
For example, if we look at this classic four-chain bispecific antibodies instead of correcting assembled products.
[MISSING]
1:54
This will reduce the yield of correctly assembled product and increase the cost of your goods.
2:07
Fortunately, there is a technology to address this challenge effectively.
2:13
This technology called knob-into-hole technology.
2:19
This technology introduces mutations to CH3 domains of heavy chains to create a knob into hole structure.
2:30
This knob into hole technology is great for driving the correct pairing of heavy chain to heavy chain of bispecific antibody production.
2:42
Now the knob into hole technology has been considered as gold standard and well known technology when designing bispecific molecules.
2:55
However, even if you have correctly assembled heavy chain/heavy chain product, light chain/heavy chain mispairing can still occur.
3:07
In Lonza, we develop this technology bYlok to tackle this particular challenge.
3:17
So this technology bYlok involves some engineering steps to the antibody.
3:23
It involves elimination of the native disulfide bond between the heavy and light chain in the constant domain in one Fab arm: CH1-CL.
3:36
And generation of an artificial disulfide bond between cysteines at the interface of the variable region VH and VL domains.
3:50
This minor modification favours the formation of correct heavy chain/light chain pairing.
3:58
So when you use bYlok technology together with knob into hole we can significantly increase the yield of correctly assembled product.
4:09
I will demonstrate this in next few slides.
4:13
So firstly let me explain how we apply it.
4:17
Here we have two standard monoclonal antibodies.
4:22
Then we combine both monoclonal antibodies to create our bYlok bispecific molecules.
4:31
So for all the data referred to here, we applied bYlok together with knob into hole.
4:39
But there is no reason why bYlok doesn't work with other pairing solutions.
4:51
As you can see, we have four different versions of bYlok bispecific molecules here.
4:57
So the first one, we introduce the mutation on one of the Fab arms on the same side of the hole.
5:06
And then the one on the right, we introduce mutations on one of the Fab arms on the same side of the knob.
5:15
And then also we can try different orientation.
5:18
So we end up with four different versions.
5:21
What we typically do, we test four different versions, and we find out which one works for your molecule.
5:31
So in the next few slides, I will run through some experimental case studies to demonstrate the benefit of our bYlok technology.
5:46
So we selected parental monoclonal antibodies and generated bispecific antibodies.
5:55
We cloned them into our expression vector and expressed them in our chosen cell line.
6:04
We used two different scales.
6:06
We used 150mL small scale also scaled up to 5L.
6:13
Then we evaluated the assay titer and product recovery at both scales.
6:20
We performed some functional characterization of expressed bispecific antibodies.
6:29
So in terms of the monoclonal antibodies selected for this case study, we use some well-known monoclonal antibodies.
6:39
Here we have trastuzumab targeting Her2 coloured in blue.
6:46
We have panitumumab coloured in pink targeting EGFR.
6:51
We also have alemtuzumab colour in green targeting CD52.
6:58
On the left side of the screen.
7:00
You can see we combine the trastuzumab together with panitumumab.
7:06
We create four different version of bispecific molecules.
7:11
We have KIH 3 and KIH 10.
7:14
Those are the bispecific molecules without bYlok technology applied serving as controls.
7:23
We also have a two bYlok molecules bYlok 3 and bYlok 10.
7:29
On the right side of the screen we apply the same approach.
7:33
We combine the green chains together with pink chains.
7:36
We create four different bispecific molecules to control and then to bYlok bispecific molecules.
7:49
Then we performed the characterization of expressed bispecific molecules.
7:55
We did this by following the workflow as below.
8:00
We identify the correct assembly ratio by LC-MS.
8:06
Then we assessed the byproduct and product quality also bispecificity affinity using a wide range of analytical techniques.
8:18
We also tested stability and immunogenicity.
8:25
This slide shows the correct assembly ratio for our bispecific molecules for both scales, 150mL and 5L.
8:39
The results are consistent as you can see.
8:43
So in this result we coloured the bYlok molecule in dark blue and a knob into hole control in light blue.
8:54
For the knob into hole control only, we observed around 30 to 35% of mispairing product for bispecific generated with bYlok consistently display greater than 95% correct heavy chain and light chain pairing.
9:17
Specific yield is increased by using bYlok over knob into hole technology alone.
9:28
Then we moved on to functional assay - dual antigen binding.
9:33
We use the method ELISA based on biotin and streptavidin.
9:39
So we have antigen 1 coloured in blue here was immobilised onto the plate.
9:47
We also have antigen 2 which is coloured in green which was conjugated to biotin but not immobilised to the plate.
9:58
We also have the HRP conjugated streptavidin to indicate the binding.
10:07
On the right side of the screen, from the ELISA result, we can see a nice correlation between the antibody concentration to the absorbance indicating successful binding.
10:19
So the takeaway message here is that the bispecifics generated using Lonza's bYlok technology returns antigen binding properties of the input parental monoclonal antibodies.
10:36
Then we wanted to have a deep understanding about affinities of our bispecific molecules.
10:44
So focusing on bYlok 10.
10:48
Here we tested bYlok 10, the knob into hole control, and both parental monoclonal antibodies. And from the result demonstrated here we can see the bispecific generated using bYlok have a comparable affinity and binding kinetics to knob into hole only controls.
11:15
We also tested the thermal stability using nanoDSF.
11:19
The overlay data shows the thermal stability of the bYlok bispecific is similar to the knob into hole only controls.
11:33
Immunogenicity is a critical factor to consider when assessing clinical efficiency.
11:43
In Lonza, we use a platform called Epibase in silico HLA Class II binding T Cell epitope profiling platform to assess the risk of immunogenicity.
12:00
Essentially what this method does is provide you a DRB1 score.
12:06
This score provides you a measure of how likely this will induce immune response in your body.
12:16
So we tested both parental monoclonal antibodies.
12:20
We also tested our bYlok molecule and knob into hole controls.
12:27
So we also broke down the scores to individual light chains and also a heavy chain knob and heavy chain hole and full antibody.
12:39
So the takeaway message here is bYlok modifications have a little to no contribution to the overall score of the different chains.
12:50
The addition of the bYlok to the knob into hole does not increase predicted immunogenicity.
13:02
Then we wanted to test the feasibility of downstream process.
13:06
So we purified our bYlok bispecific antibodies using standard downstream methods that we use internally at Lonza.
13:16
Then we assessed our purified molecules based on standard metrics such as aggregation, fragment purity, and HCP.
13:28
The conclusion is antibodies generated with bYlok technology can be purified using the standard downstream process.
13:40
In the next few slides I would like to demonstrate how bYlok compares to the other pairing solutions.
13:50
We compared bYlok to two other pairing solutions currently on the market: DM-1 and DM-2.
13:58
DM-1 involves the mutation of the light chain in the constant region.
14:07
DM-2 involves more significant changes to the antibody sequence.
14:16
All of them involved engineering steps to the antibody.
14:20
We evaluate the result from transient expression, stable pool, and scaled down cell line construction.
14:32
As I mentioned, we can create four different versions.
14:36
So we tested four different versions per scaffold.
14:41
The best format per scaffold was selected based on transient expression read out, including percentage of correct pairing, functionality, titer, aggregation, and impurities.
15:01
Once we select the best format per scaffold, we went ahead with a stable pool expression.
15:11
This result demonstrates the stable pool evaluation using our GS piggyBac transposes technology.
15:20
As you can see the result here bYlok shows comparable stable pool titers to alternative scaffolds.
15:33
And then we proceeded with head to head cell line construction (CLC) evaluation.
15:40
Established CLC workflow was used here to test scaffold format head to head.
15:48
We use the Beacon platform for our cloning process and also productivity assessment.
15:56
Clones were selected based on the Beacon on-chip productivity and then we ranked our clones based on product concentration on day 4 during the shake flask stage and then we assessed for product concentration, QP, and metabolite analysis.
16:17
Then top six clones were selected for product quality analysis.
16:26
This result shows the product assembly ratio in top six clones for bYlok and DM-1.
16:36
The method we used here was SEC-MS.
16:41
As you can see, for the product assembly using DM-1 technology, we can see a big variation among top six clones.
16:53
However, the top six clones expressing the bYlok scaffold demonstrated more consistent full product assembly than the comparable disulfide modification alternative.
17:10
Again, this slide demonstrates the result for the product assembly in the top six clones using bYlok and DM-2.
17:21
Again, we see the similar result for the product assembly using DM-2 technology.
17:28
We observe a big variation among top six clones.
17:33
The top six clones expressing the bYlok scaffold demonstrated more consistent full product assembly than the established domain modification alternative.
17:48
Then we went ahead.
17:49
We select our lead clone.
17:51
So we selected our lead clone based on correct assembly functionality, titer, aggregation, and impurities.
18:01
The result shows from the selected lead clone expressing each bispecific antibody scaffold.
18:09
bYlok demonstrated consistent performance across all key metrics of titer, assembly, and purity.
18:21
A short summary for this part.
18:23
The bYlok technology provides an easy to implement solution for heavy chain and light chain mispairing and for the manufacturing of IgG like bispecific antibodies.
18:36
It has minimal modification to the monoclonal antibody design.
18:40
It fits into standard a manufacturing workflow.
18:44
It generates more than 95% correct heavy/light chain pairing.
18:50
It's portable and versatile platform.
18:54
The data I demonstrated today shows immunogenicity, thermal stability, and binding properties of the bispecific antibodies are not affected.
19:07
Combining with our GS Discovery and GS piggyBac, we can help boost bYlok bispecific antibodies’ titers in both transient and stable pool modes.
19:22
In my last few slides, I would like to run through some experimental case studies to show you how we can boost bYlok bispecific antibodies expression with our GS toolbox.
19:38
At Lonza, we have the integrated GS Gene Expression System, making expression easy from discovery through to commercial production.
19:50
We have host cells, vector, innovative technologies, and system.
19:58
We also provide extensive know how, technical process, and regulatory support throughout your product journey.
20:09
Last year, we launched a few new offerings coloured in green.
20:15
Here we launched our GS Effex host cell line for the development of antibodies with enhanced potency.
20:26
We also launched our own thorough TheraPRO CHO media and feed system version 9.6.
20:34
We also launched a new offering GS Discovery transient expression system last year.
20:42
This system is great for generating small amount of material in a short time span.
20:49
I will mention a bit more later.
20:55
GS piggyBac transposon-based GSquad vectors are one of the key components of our Lonza GS system.
21:06
This technology was launched in 2019.
21:11
How does this system work?
21:12
So this system has two parts, has GS piggyBac vector, and GS piggyBac transposase.
21:21
In the vector you have your gene of interest and also the ITR - inverted terminal repeated sequences.
21:32
So we transfect both vector and transposase into the host cell.
21:39
The transposase can recognise ITR and cut and paste it to open regions of the genome associated with highly expressed genes which help boost titer.
21:57
This slide demonstrates the result for our GS Discovery transient expression of our bYlok bispecific molecules.
22:07
As I mentioned, GS Discovery is a great technology for generating small amount of material in a short time span.
22:21
So here what we did, we combined the pool selection and pool expansion in the same stage and harvest the product on day 14.
22:35
The right side of the screen we show our titer result with legacy electroporation process in light blue and our GS piggyBac transient process in dark blue.
22:50
We observed a 24 and 12 fold titer increase respectively for both of our bYlok molecules.
23:02
So the key takeaway message here is GS Discovery transient system offers a newly optimised transient process built around the GS piggyBac that achieves substantial titer increase of bYlok molecules in comparison to legacy transient process.
23:27
This is my last slide.
23:28
So on my last slide we show the result of the stable pool expression for our monoclonal antibodies and also bYlok molecules with the GS piggyBac in dark blue and random integration in light blue.
23:45
The result demonstrates GS piggyBac stable pools expressing bYlok bispecific antibodies generate higher titers than random integration controls.
23:59
Thank you very much, I hope you enjoyed my talk.
