Harnessing Antigen-Specific T Cells and Tonsil Organoids for Advanced Therapeutic Insights

0:13 

Thank you. 

 
0:27 
So I have two topics to cover today that both are around working with complex immune assays. 

 
0:34 
That's antigen specific T cells. 

 
0:37 
It's for targeted immunogenicity and efficacy testing as well as tonsil organoids that enable the ability to test for therapeutic manipulation, global adaptive immune responses. 

 
0:49 
Do you need to speak louder? 

 
0:50 
Can you hear me OK? 

 
0:57 
So both of these assay types can be used in a range of therapeutic areas, which I'll get into in more detail. 

 
1:05 
So first off, the antigen specific T cell, they're key contributors in the recognition of threats such as pathogens and cancers, but they're also important in autoimmunity where auto antigen specific T cells are key drivers of disease. 

 
1:23 
In tumour vaccine development, the correlation between the high frequency of tumour specific T cells with improved patient outcomes is leveraged, but preclinical research can be quite challenging. 

 
1:35 
So there's often a low precursor frequency of the antigen specific T cells, meaning that it's hard to isolate a high number of cells. 

 
1:45 
And also there can be low expression of the HLA molecules which bind the immunogenic epitopes. 

 
1:51 
There's a clear need for in vitro models to test antigen specific mechanism. 

 
1:55 
So we are having these standardised. 

 
1:59 
So first assay that we run is an HLA stabilisation assay. 

 
2:05 
So in this assay, T2 cells are loaded in this case with peptide and the presence of the peptide connecting to the MHC Class 1 molecule allows to stabilise on the surface. 

 
2:21 
So T2 cells are specifically useful for this because they're TAP deficient. 

 
2:25 
So they're not loaded with peptide inside the cell. 

 
2:29 
So most of the peptide that is added exogenously will be the specific type of peptide that is given. 

 
2:35 
So here we've loaded MART-1 or a negative control peptide and then looked over time at 6, 24 and 48 hours to look at both the binding strength by giving different doses and also looking at the stability over time. 

 
2:51 
So you can see here that the MART-1 really peaks at 24 hours and then the stability reduces by a 48 hour time point. 

 
3:01 
So this obviously isn't just used for MART-1 but allows for a testing of a range of peptides and really can be a first selection to show which peptides might be most suitable because both their strength and their stability over time will help in generating an immunogenic response. 

 
3:21 
So then after that immunogenicity assessment can be performed on the selected peptides. 

 
3:27 
And for that we use HLA or bloods that have the right HLA type. 

 
3:33 
We isolate monocytes and T cells, differentiate the monocytes into dendritic cells and then peptide load the dendritic cells with the peptides of interest and Co culture them with T cells for approximately 2 weeks. 

 
3:48 
Then when we do restimulation at the endpoint by ELISpot, we can see whether the peptide expanded cells show a good immunogenic response or not, which in the case of MART-1 is clearly the case. 

 
4:03 
So we can see that when peptide expanded cells are restimulated with MART-1, there's a clear increase in presence of intra and gamma producing T cells with quite low background in the control expanded and non-restimulated conditions. 

 
4:18 
So this assay can be used to predict the immunogenicity of novel vaccine candidates or new antigens and it really shows their ability to drive expansion of rare antigen specific T cells as well. 

 
4:31 
And obviously that expansion is needed to have good numbers of cells present at the end point to do further testing with. 

 
4:39 
And this can really help to further generate a hierarchy of the different vaccine candidates, which might be useful to take forward into further assays. 

 
4:51 
So the next step that can then be taken is with the selected targets of interest, polyclonal expansion assays can be set up to generate larger numbers of cells. 

 
5:02 
So to do that we can use multimers, so pentamers, dextrimers, all kinds of multimers that are loaded with the peptide of interest that is specifically labelled. 

 
5:14 
We then stain bloods of the correct HLA type with those multimers and FACS sort them. 

 
5:21 
So we can get really pure populations of these antigen specific cells and the purity of the population that allows us to do polyclonal expansion for a two week period. 

 
5:32 
And still getting quite high purity of antigen specific cells out at the end. 

 
5:36 
So in the case of MART-1, that's 91% and as you can see on the right, there's quite a nice expansion of those CD8 over the two weeks. 

 
5:48 
Once we have a good number of antigen specific T cells, we can then do some initial functional assessment just to ensure that those cells are still functional, which as you can see they are. 

 
6:01 
And then we go into our killing assays. 

 
6:03 
So  we run our killing assays in a number of different ways. 

 
6:09 
Perhaps the simplest way is using again the T2 cell line that I showed previously. 

 
6:13 
So they could be peptide loaded exogenously. 

 
6:17 
And then in this case, they are Co cultured with the antigen specific T cells and different E to T ratios. (Effector: Target ratio) 

 
6:24 
And after 24 hours, we assess whether there is the killing of the target cells, which as you can see at the bottom, hopefully it's readable even at a 0.25 to 1 E to T ratio, there's still quite significant killing present with these cells. 

 
6:41 
Another option is using native endogenous expression in the target cells. 

 
6:46 
So here the target cells are expressing MelanA endogenously. 

 
6:52 
That is perhaps a better model to use if you really want to be close to the endogenous levels of expression that you would see in vivo. 

 
7:03 
But obviously the amount of MART-1 in this case that will be expressed on the surface of the cells is lower than when you exogenously low MART-1 peptide. 

 
7:12 
So to compare the two, we've shown both here. 

 
7:16 
So when endogenous MelanA target cells are used with effectors you can see in the black bar here that there's clear killing of the endogenous MelanA cells. 

 
7:33 
But when the cells are also loaded with MART-1 peptide as we'd expect, we'd get higher killing capacity. 

 
7:41 
And the third option is using transduced overexpression target cells. 

 
7:48 
So this can be used when either there isn't a suitable cell line that endogenously expresses the cell or perhaps when multiple targets need to be expressed on the same cell type for the asays that are needed. 

 
8:00 
So here what we did was it was we used a wild type cell line that did not express MelanA, transduced it efficiently and then we did an Incucyte assay here where we tracked tumour cell growth over time. 

 
8:14 
And when you compare tumour cell growth in the presence of antigen specific T cells, but the tumour cells don't express MelanA, there's a clear increase in growth of the tumour cells compared to the tumour cells that are expressing MelanA, indicating that there is killing of the cells. 

 
8:31 
And that is also shown by apoptosis here at the bottom, there's clear apoptosis of the target cells that are transduced with MelanA. 

 
8:41 
And again when the tumour cells are loaded exogenously with MART-1, we see that there's further killing. 

 
8:47 
So that obviously is stronger killing as we would expect. 

 
8:53 
So then alongside the 2D killing assays, we also are able to do 3D killing. 

 
9:00 
So here we use the same MelanA transduced cell lines with the wild type control cell lines. 

 
9:06 
We also have the MART-1 pulse condition again in green. 

 
9:11 
And as an additional condition here, we also have therapeutic antibody on the right to see if we can further enhance the MelanA transduced effect. 

 
9:22 
So these videos, yes, are working. 

 
9:26 
So what you can see here in red are the NLR cells. 

 
9:29 
And when there's killing present, you can see a black ring forming around the 3D structure. 

 
9:36 
And you can see that there's a reduction in the number of red cells. 

 
9:40 
So that's also summarised at the bottom here on the left you can see that when wild type is compared to wild type plus MART-1 pulsed cells, there's a clear killing window. 

 
9:52 
And we see a clear killing window as well in the MelanA transduced cell lines compared to the MelanA without CD8 T cells added. 

 
10:04 
And that window is further enhanced with the therapeutic antibody. 

 
10:10 
As I mentioned, the antigen specific T cells are not only important in tumour assays and IO, they're also important in other areas such as viral, auto antigen and allergen specific assays. 

 
10:26 
So I have a brief overview of that here where these different peptide pools of viral auto antigen and allergen stimulations were done over time and the cells were expanded in antigen specific manner. 

 
10:40 
Then we performed restimulation at the end point and looked at flow phenotyping and cytokine release to determine whether there were antigen specific responses present. 

 
10:49 
So we can see that there's an increase in CD25 indicating T cell activation with all three of the pools that we've tested and similarly cytokine release. 

 
10:59 
So interferon gamma also showed that there was activation of the T cells present and that was most clearly the case in the viral simulation in blue. 

 
11:11 
In addition to the assays that I've shown, there are also other tools available that can be applied as well. 

 
11:16 
So there are inventories available with readily expanded antigen specific T cells. 

 
11:23 
They're also PDX and CDX patient tumour cell lines that can be used in 2D or 3D formats in these killing assays. 

 
11:33 
And as I showed, gene editing can be performed so that the targets are expressing what is required for the assay to work. 

 
11:43 
So in summary of this part, there are a wide range of in vitro antigen specific assays that can be run to support antigen specific T cell programmes. 

 
11:53 
The first assay was peptide binding capability and stabilisation and that can then be followed by immunogenicity. 

 
12:01 
We can do expansion polyclonally as well, followed by 2D and 3D killing assays. 

 
12:12 
Depending on the target there are different options for running these assays, so some may already have readily available stocks. 

 
12:20 
They are optimised protocols that can be used to expand these fresh or depending on the target they could need further optimization. 

 
12:32 
Moving on to the second part, tonsil organoids, as you can see here, there is poor translatability of preclinical models at the moment in vaccine development. 

 
12:49 
So even though some vaccines can be quite promising preclinically, they don't always show effectiveness in the clinic, sorry, with some either failing, showing poor efficacy, limited efficacy or even enhanced risk. 

 
13:05 
So there's a need here to develop further assays that can bridge the gap between the preclinical studies that we already have in place ahead of going into the clinic. 

 
13:17 
So an option for that is the use of tonsil organoids. 

 
13:20 
So they have complex 3D architecture and support germinal centre formation for therapeutic testing. 

 
13:28 
And they are generated by dissociating tonsil tissue into single cell suspensions and then setting up high density transfer cultures. 

 
13:39 
They can then be stimulated in the presence of antigen for several days or weeks, depending on the antigen to form these tonsil organoids. 

 
13:47 
So for illustrative purposes, you can see that these tonsil organoids are considered to have germinal centres even though they're obviously in vitro generated. 

 
13:58 
So they retain that 3D architecture in addition because if this is a single cell suspension of all of the tonsil, there are a lot of different cell types present. 

 
14:12 
So this is quite key that a lot more cell types are present then you would see in a typical in vitro assay. 

 
14:22 
And yeah, that really is a kind of key point that helps in this translatability. 

 
14:30 
So what we did here was look at flu vaccine, we generated the tonsil cultures and did a simulation for seven to 10 days. 

 
14:41 
And then we looked at the presence of antigen secreting cells as well as antigen specific antibodies to determine whether the tonsils were actually functional in our hands. 

 
14:52 
So we're looking at antibody secreting cells and ELISpot was done to look at the number of responsive cells and we can see that in all donors tested here, there's a clear difference between the unstimulated number of antigen secreting cells and those in the presence of the vaccine. 

 
15:09 
And in a lot of cases, there's actually more of an increase than in the IL2-SAC control. 

 
15:16 
We also looked at flu IgG presence and we can see that it's clearly detected in several donors in the presence of the vaccine and it's clearly increased compared to the unstimulated condition as well. 

 
15:29 
So now that we have this model and we're happy that this is functioning as it should, we can look at specific immune cell subsets in these structures and we can also assess immunomodulators in this influenza model to determine if they have an impact on the populations present. 

 
15:53 
So first off, there are a number of T cell subsets present in tonsil organoid structures. 

 
16:00 
So just briefly showing here the presence of three different populations of T cells that we see here. 

 
16:10 
So we have the CXCR5 negative, PD1 negative, non T follicular helper memory cells, the intermediate pre GCT follicular helper cells and then the CXCR 5 high PD1 high germinal centre T follicular helper cells that cannot be distinguished. 

 
16:28 
And then as you can see in the gating strategy, there is also obviously quite a significant number of B cells present in these tonsils, which is what we looked into in more depth. 

 
16:38 
So we tracked the B cell response to maturation and the GC differentiation over times over a two week period. 

 
16:47 
And the different populations are identified here. 

 
16:50 
So hopefully that's readable. 

 
16:52 
So these are the naive B cells and the pre GCs the germinal centre B cells and plasmablasts with the memory B cells in black. 

 
17:00 
And so over time you can see these, this is the unstimulated and these are the influenza simulated cells. 

 
17:06 
You can see that there is an increase in the pre-GC and germinal centre cells, but we also see a clear increase in the plasmablast population in the top right over time. 

 
17:17 
And this is specifically the case in the influenza vaccine treated cells. 

 
17:21 
So that's summarised here. 

 
17:23 
And you can see that there are clear differences between the unsimulated and inactivated influenza vaccine condition. 

 
17:31 
So we then decided to select a single time point to look at more closely. 

 
17:36 
We decided to go with day 11 because there are nice clear differences present between the unstimulated and stimulated conditions and we can clearly see the differentiation in the different populations. 

 
17:50 
So we set up a new assay for that. 

 
17:53 
So again we can see the nice difference between the unstimulated and influenza stimulated cells with good populations of pre GC germinal centre and plasmablasts present here. 

 
18:06 
So then when we added different immunomodulators, we could look at the differences between the influenza alone and the influenza with those immunomodulators. 

 
18:16 
So first off, rapamycin that was added for the 11 day culture. 

 
18:22 
And when that was done, we could see that very specifically the plasma blast population was depleted and was lower than that in the unstimulated condition, whereas all the other populations really remained present in the same frequency as we see in the influenza treated cells. 

 
18:42 
And interestingly, in the tofacitinib condition, we see the same decrease in the plasma blast differentiation which almost completely disappears. 

 
18:52 
But we also see that the pre GC and the germinal centre cells are almost completely gone, which is in line with literature. 

 
19:01 
So that was really good to see. 

 
19:03 
And so that's summarised on the bottom left here. 

 
19:08 
So you can really only see the naive and memory populations here in the tofacitinib treated cells. 

 
19:15 
We also looked at the IgG and IgA presence on these different cell populations. 

 
19:25 
And we can see summarised on the right here that both IgA and IgG are clearly present in good cell numbers in the influenza simulated cells and those disappear in the presence of rapamycin and tofacitinib. 

 
19:45 
So just to briefly summarise this last section, I've really only have scratched the surface of what can be done with tonsil organoids. 

 
19:53 
It's really a platform that can be used across drug discovery, not only to investigate novel vaccine targets as I've shown, but also to predict immunogenicity, biologics and determine gene therapy, neutralisation and AAV capsids. 

 
20:08 
So more to come. Thank you.