Slide 1

So, good afternoon. I want to thank the organisers for giving me the opportunity of presenting my data to this court of experts in the field I end up having to work but in which I’m not an expert but actually what I know more is about this enzyme called PI3 kinase which is translocated to the cell membrane when growth factors are activated as is envisioned there.

Slide 2

So when these growth factors are activated, this enzyme induces the generation of these lipids which are normally at very low levels in the cell membrane and what’s important about the generation of these new lipids called phosphoinositides is that they are able to activate a number of intracellular cascades which I’m not going to enter in too much detail and end up inducing gene expression and these pathways are capable of inducing cell responses such as invasion survival, cell division, cell growth and activation of cells in the immune system. Because of this power of these pathways it has been involved recently in cancer and in immune response pathologies. So, I will be talking today about the involvement of one of the PI3 kinase isoforms that is called PI3K γ or PI3K 1B in lupus and I work in animal models so it would be murine systemic lupus.

Slide 3

Life is not as simple as we would like and the enzyme is not a single one but 4. 3 of them are activated via Tyrosine kinase and they are grouped in the 1A subfamily as indicated there and 1B which is induced by the protein-coupled receptors. They are heterodimers composed of a catalytic and a regulatory subunit and so the enzyme itself is the P110 indicated there catalytic subunit, alpha and beta are ubiquitous and these are the ones where research in cancer is focused because they are the ones that regulate cell division, whereas γ and δ are mainly expressed in cells of the haematopoietic system and in fact, mouse models with gene defects in these molecules have defective immune responses.

Slide 4

So our contribution to this topic which is our area of research is to investigate the specific function of each isoform for one of the models of the study which is inflammation and in particular T cell activation and contribution to inflammatory processes or cancer.

Slide 5

As you know the T cells are in the periphery. Maturity cells are two main types, the CD4s and the CD8s, they are both activated by antigen presenting cells and antigen and when they do so they proliferate and they recognise the cells that have the viral or bacterial peptide on their surface to kill these cells. But when we finally finish the immune response and we have very low titres of antigen or bacteria or virus, these cells have to calm down to clear and they do by apoptosis or by other mechanisms that I do not have time to enter into. But this is what should happen. Only a very few of these cells are kept for the entire life and they are called memory cells.

Slide 6

So to find out the function of these two main families the 1B gamma and the 1A gamma interstitial activation, the first thing we did was prepare a mouse which has the whole 1A family activated and we sort of predicted that this animal was going to develop tumours and that the hypothesis. When these animals are very young and we focus on the T cell maturation in the thymus, we see that it is enhanced a little bit, they migrate better and they accumulate in the periphery. But as months go through the life of these animals you start to have accumulation of these memory cells in the periphery and they end up developing lupus which is the reason why these animals die more or less in their mid age, when they are like a year old.

Slide 7

So here I’ll just show some of this data you don’t have to focus if it’s not your main interest but just to show that this is really based on data we obtain. As for the lupus the way we see this disease is that it could be originated and this would be the case here by an effective T cell response that doesn’t go downward as it should at the end of the immune response and therefore induces in a continuous manner the activation of B cells and macrophages. These two things increase the content of γ immunoglobulins in the blood which end up being deposited in the kidney and contribute to the inflammation generated there causing in the end glomerulonephritis. This is exactly what we see in these mice when we have enhanced PI3 kinase because we have --- enhances the enzyme activity, the negative regulatory mechanism in this pathway is PTEN, you may have heard about the involvement of PTEN deficiency in cancer. All the PTEN deficient mice also develop lupus. So it’s confirmatory to our studies.

Slide 8

So the conclusion would be that if you have an enhancement in this route that normally induces survival, we have lupus because the memory cells have an enhanced survival and don’t die during the end of the immune response. In this case, in the case of chronic inflammation the antigen would be autoantigen and this is the reason why chronic inflammatory responses based on T cell function don’t really get cured throughout your life because you have the cells there.

Slide 9

When we finished this study, it was published that the γ isoform so the one we saw causes lupus, the 1A family, the 1B isoform. Deficiency of this enzyme results in a production of T cell numbers and lower T cell activation. So we wonder whether they could have complementary or synergistic actions and to investigate this we performed a detailed study on the phenotype of the γ isoform knockout mice and we saw that the phenotype that these animals have is kind of complementary to the excess of 1A activity. When we cross both mice, so we don’t have the 1B isoform and we have excess of 1A what happens? Do we still develop lupus? Do we still develop tumours? These mice also have enhanced tumour formation, they have an enhanced invasion of T cells. What of these functions when we increase the amount of PIP3 of the enzyme activation requires the gamma activity?

Slide 10

So, we looked at adult mice and we saw that if we don’t have γ because we genetically delete it, we still have invasion. So invasion when we have it in a tumour is because of the activation of the 1A forms α and β and maybe δ. We still have a larger number of T lymphocytes in this case because we are targeting the T cells in particular.

Slide 11

So again cell accumulation probably relies on the activity of 1A enzymes. But however, we start to see as these mice age that despite having elevated levels of PIP3 because of the whole 1A family enhancement if they don’t have the γ isoform, they don’t accumulate immunoglobulins in their blood, in their serum.

Slide 12

They don’t present proteinuria and in fact we have a reduction in the immunoglobulin deposits and the glomeruli according to the pathologist look much better and these mice actually reach the normal age for the end of their life which is about 2 years.

Slide 13

This represents the life span of these animals, when we erase from them the 1B isoform.

Slide 14

So the reason behind this we investigated the different population of cells is in fact that you have less memory cells. So this is part of the reason we think this is an important contribution in the maintenance of the disease.

Slide 15

This is a model, the model I just presented, in which we either induce a subclass family and delete one isotype.

Slide 16

It’s a little bad as a model of human SLE, so we decided to investigate another mouse model which is more similar to the human disease in that the disease is caused by several alleles, by several different mutations that increase their susceptibility to develop during life this disease. This model is the MRLMPR and because it is extremely difficult to cross this with knockout mice because we have to select in the crosses up to 8 alleles, what we decided was to use for this study a selective inhibitor of 1B isoform. First of all we wanted to confirm that really in these animals we have activation of this pathway, it this abnormally activated and that we have an abnormal high amount of memory T cells. This is what is illustrated here. The mice strains are indicated at the top and the PKB, you see phospho-PKB there is one of the enzymes that I showed you on the cascade whose activation depends on the activity of PI3 kinase. It is easier to follow than the enzyme itself. So we see that in fact, these animals have a higher activation of this enzyme and also that these animals have a higher proportion of memory cells. There are other mice strains that do not develop lupus because they do not have the alleles of susceptibility.

Slide 17

So, for the drug we cooperated with Serono and in the frame of a European grant and they developed the isoform specific inhibitor which is a very difficult task because the 4 catalytic domains that I showed you are certainly similar but they did find a compound that if it is given to purified enzymes here I have put together the 31As or the 1B and you see selective inhibition on the 1B and not on the 1A isoforms and this is looked into cells not the purified enzyme. So we give this to cells. Then in the other scheme on the right is looking at the mice. Cells in the blood of the mice have decreased this pathway. So the enzyme is selective and the enzyme works in vivo so we prepared an experiment in which we gave this compound every 12 hours during 3 months because these mice developed the disease at about 5 months of age and every 14 days we took samples of the blood and the urine to investigate how the mice id doing with regards to glomerulonephritis.

Slide 18

So we see that inhibiting the γ isoform of PI3K results in less hyperplasty. These are mice present a higher number of lymphocytes, this is a little bit different in humans because mice don’t have anti lymphocytes, antibodies in their disease.

Slide 19

They had lower levels of alpha auto-antibodies and proteinuria, significantly reduced compared to the lupus prone mice that were used as controls.

Slide 20

Their glomeruli you can see on the very left how bad it is when the mice develop the disease and it looks much better when you give these compounds and you have a lower accumulation of immunoglobulins in the kidney.

Slide 21

For that reason, I’m sorry the slide got a little bit spoilt, you don’t see early death and we call early death at the time we finished the experiment which is when they are 5 months old.

Slide 22

So you only have 40% of the animals non-treated already dead and the rest of them have very advanced glomerulonephritis.

Slide 23

With dexamethasone and we give maybe a little bit too high dose you don’t have glomerulonephritis either but you have a lot of secondary effects. With this inhibitor what we see is that you don’t have much of early death and you don’t observe advanced glomerulonephritis and more around 90% of the animals in different doses saw a better shape in their glomeruli function. Again, the reason for this when we investigated the different populations in these mice was the production in memory T cell numbers. So together what we think is that γ could be a promising drug for this chronic inflammation and maybe other models of chronic inflammation based on excessive T cell function because it actually reduces the survival of the memory cells which are the cellular component that is causing the maintenance of the disease.

Slide 24

Right now we are trying to examine how similar are the human SLE with the murine models and we have a number of collaborators we are have in large letters the ones that we are focusing now on in order to get enough numbers of blood samples from patients in different stages of SLE and test whether first of all in human lupus it is also true that you have an increased proportion of memory cells. 

Slide 25

The preliminary data I’m showing here doesn’t have a statistical significance, so it’s just an example of the first step that we are following in order to see if there is a parallelism between these small mammals and us. And there is, it is not exactly equal because in animals we only have the CD4 positive memory cells and in humans we see an increased number of memory cells of both the CD4 and the CD8 type. When we look at the activation of this pathway in the T cells in the blood there is also a frequent event to have this pathway induced and we don’t find that in normal individuals. But as I said we have to perform a larger cohort and well organised with a statistical significance, grade of the disease etc and this is in process.

Slide 26

So I hope I have convinced you that this target is interesting not only from the point of view of understanding better how physiology and pathology is regulated but it also may have some therapeutical interest.

Slide 27

This is the team that actually does the work I have in red the name of Domingo Barber and Carmen Hernandez for the main players in this project. I have been financed by all these institutions and we have counted on the help of the people at Serono in blue led by Doctor Rommel and the people from the animal facility in red and the pathologists which help to interpret what is in these tissue sections. And this is all thank you.

Slide 28

Chairman: Well, thank you very much for your beautiful work and very, very nicely presented. We have time for questions. Yes could you state your name and institution?

Question: Well, the data are very interesting but a few questions arise concerning clinical applicability. The question number one is if the number of memory cells is diminished, it creates a risk of impaired difference against infection? The second point...

Dr Carrera: Let me first answer this one because I am all activated and then I forget all the rest of the questions. Yes, these mice, the knockout mice do develop primary immune response. So the requirements for a primary immune response with regards to this enzyme which exists, it contributes it is also helped by the 1A subfamily. So we do see the development of a primary immune response which would be what allows you to fight against disease. Now memory cell survival, we still don’t know what’s going to happen. We are going to test that in culture because we don’t have any volunteers for taking the drug in vivo, alright? But I hope to find a window anyway the very abundant memory cells are expected to be the outer reactive memory cells for these patients.

Question: Ok we can argue in different ways but let me ask the second and the last point. You started your experimental treatment before lupus phenotype appeared in the patients, so this is very difficult to repeat in a clinical setting.

Dr Carrera: No, when we established the model for the 1A active that causes lupus and if you think about it, it is the first indication that deregulation of this enzyme could be a cause of disease, in that model when you cross it with the knockout mice, of course, they have their genetic alterations since they are fast fertilised. That’s forever. But in the MLR MPR model we did the experiment in both ways so we started treatment, in fact, if it were as you say we should have started when they were born and that was even more difficult than it was. So we started the treatment at 2 and a half months when you don’t have symptoms of the disease or at 3 and a half months when the proteinuria is already high.

Question: Ok thank you.

Chairman: Another question? I’m not very confident in lupus in mice but I’d like to ask a naïve question from a clinician. We treat lupus with corticosteroids and I’d like to ask you, maybe it’s a silly question, are there any relationships between the way your compound acts on the memory cells and the glucocorticoid receptor?

Dr Carrera: In principle no but we need more basic knowledge before answering this for sure because the fact that you don’t know something doesn’t mean it couldn’t happen in vivo and because of the diseases like estrogen receptor related breast cancer in which you see frequently activation of this pathway I wouldn’t be completely sure that there is not a way by which nuclear receptors activate this pathway. Now this is one thing and what you’re asking me is if the way that corticoids act in vivo resembles the way this inhibitors acts, I don’t think so and glucocorticoids would have an impact not only in reducing all the haematopoietic cell types but also in other cerebral control hormonal pathways. While γ, the fact that γ is only expressed in cells from the haematopoietic system makes and the fact that knockout mice in fact, only have defence in immune responses makes us think that it could be more helpful although you have to demonstrate that. So what you have in knockout mice is that primary response goes a little bit more slowly because macrophages and neutrophils arrive later to the site of infection and this reduces T cell effect. So I think it would be more selective.

Chairman: Certainly yes. Other questions? Well, thank you very much that was a very nice presentation really.