Transcript for:
Understanding Nephrotic Syndrome in Children

I was taught that it's given by Marina Livanovi from Rome too on multi-trakulosis dengue-phobic syndrome. Marina trained in PEDS at the University of Pavia in Italy. She completed the research fellowship at Harvard, Children's Hospital in Boston for the neural reseal function and since 2006 she's in the Division of Pediatric Podiatry at the University of Pavia in Rome and she's worked at the new from the Department of Tensational Research in Pediatric and Immunomediated Venal diseases and the public learning of chronic symptoms.

This is forward. Okay, so thank you very much for the kind invitation. First of all, I'd like to thank Franz.

I'm very excited to be part of this absolutely staggeringly interesting meeting. And so we're going to completely switch gears now. And move to diseases in which there is not a monogenic cause and the pathophysiology is by and large unknown. So we have a completely different approach to finding treatments for these diseases.

So first of all, I'd like to sort of talk a little bit more in detail about exactly which types of patients we're going to be trying to target and treat. So we all know that idiopathic... nephrotic syndrome in children is a world, extremely heterogeneous disease. A large percentage of these children are steroid sensitive. Luckily, there is a percentage of children who are steroid resistant, about 10 to 25%.

But of these, about half, sometimes even a little bit more, respond to second line immunosuppression, so calcineurin inhibitors. And both these patients and the initially steroid sensitive nephrotic syndrome patients then progress to have a disease which can last for decades with multiple relapses and in a certain percentage of cases also last into adulthood. The multi-drug resistant nephrotic syndrome patients, the patients who don't respond to steroids and don't respond to a second line which usually is a calcineurin inhibitor, are called multi-drug resistant nephrotic syndrome.

And of these we know that a part, a large part, are genetic forms and these will never respond to immunosuppression because they have a monogenic disease of the podocyte. which will be probably addressed by the speaker after me, Moin Salim. And these diseases invariably lead to end-stage renal disease.

However, they do not recur in post-transplant. Instead, the multidrug-resistant nephrotic syndrome forms have an immune-mediated pathophysiology, which we don't understand. So they're particularly tough to treat. And not only can they lead to end-stage renal disease, but they also tend to recur following renal transplantation.

So the kids that I'm going to address as our problem of the current talk today are these. So these are the patients that have a protracted prolonged course of nephrotic syndrome with the need for immunosuppressive therapy for a long time. Some of them will not reach end of stage renal disease, some of them will, and the ones who do tend to recur post-renal transplantation.

So I'm going to go through some clinical scenarios just to drive this concept a little bit closer. And I will then try to go back to these clinical scenarios at the end of my talk and give some solutions. So one of the scenarios is the primary multidrug resistant child.

So this can be, for example, a child who has an onset of nephrotic syndrome quite late. So when you do not suspect a genetic cause around nine years of age, for example, he doesn't respond to the six weeks of oral prednisone and to three pulses of ibuprofen nizolone. So. And so he's defined as steroid resistant. He also doesn't respond to the second line, a calcineurin inhibitor.

So he becomes defined as a multidrug resistant nephrotic syndrome child. And he can go through, for example, two cycles of rituximab. He has a negative genetic panel for all the genes coding for podocyte proteins.

And histologically, classically, the first renal biopsy, which according to guidelines is usually performed at four weeks, will show often minimal change disease. However, if you repeat the biopsy... Further down the line, even to see how much margin you have to continue treating this patient, you almost invariably will find that he has progressed, he or she has progressed to focal segmental glomerulosclerosis.

These children often have a rapid progression. And for example, in the child that I chose to describe, 15 months after disease onset, he proceeded to end stage renal disease. And we know that this child, if he doesn't have a genetic mutation, or a variant, a pathogenic variant, has a substantial risk, in this case around 30 to 50 percent of recurring following renal transplantation.

The second scenario is a completely different scenario where we don't get to end-stage renal disease. These are the children who maintain steroid sensitivity. So for example, a child with a nephrotic syndrome onset at four years, response in three weeks, so he's a steroid sensitive nephrotic syndrome because he's responded within four weeks.

He then proceeds to have multiple relapses over the years. Some of them are serrate resistant because they take more than four weeks to respond and or because they require IV methylprednisolone. In this case that I chose to describe, this patient, which I then saw for the first time when he was 16 and a half, had never achieved stemple remission for at least 12 months, despite having used for prolonged periods of time CNIs, MMF, mycophenolamophetyl.

multiple infusions of anti-CD20 monoclonal antibodies, both ritaxamab and olfagilumab. He had stunted growth, osteopenia due to prolonged use of glucocorticoids, and signs of a CNI toxicity. So yes, he had a preserved renal function, but he had a very severe load of toxicity and morbidity associated with his primary disease. Another clinical scenario, and this is the last one, is so-called secondary steroid-resistant nephrotic syndrome. So, for example, a child with an onset at seven years of age with a late responder, so he responds in the window between four and six weeks of oral prednisone.

Subsequently, though, he did achieve response, and then he was continued to treat with three pulses of methoprotein, salone, and CNI. He had subsequently multiple relapses, some of them steroid resistant, taking more than four weeks to respond. He had use of CNIs, MMS, multiple infusions of anti-CD20 monoclonal antibodies, so stunted growth, osteopenia due to prolonged use of glucocorticoids. However, to make it worse, in addition to what I presented in the previous patient, this child, two years following disease onset.

had a multi-drug resistant relapse. So he relapsed, nothing that we can do for him, including plasma braces, induced remission. So this child progressed to end-stage renal disease.

So these children, the secondary cerebrosis and nephrotic syndrome, as Wayne Salim's group has showed very convincingly, are the ones who have the truly highest chance of recurring post-renal transplantation. So secondary cerebrosis and nephrotic syndrome. leads to end-stage renal disease with a negative genetic panel. This child was mononephrectomized, put on peritoneal dialysis for about six months, and then received renal transplantation from a living donor. And as was predicted by his chances, given his clinical history, he recurred almost immediately following renal transplant.

And this was also histologically documented. So at this point, also here, he was treated with intensive immunosuppression. and also plasmapheresis. And I'll go back to his treatment further on at the end of my discussion.

So we know that idiopathic nephrotic syndrome in children can be a long and bumpy road, extremely heterogeneous, and to define the best treatment strategy, we would need to understand the physiopathology. Unfortunately, we don't. And I think this is, you know, in stark contrast to all the different diseases who have been presented throughout the course of this day, unfortunately.

So we know that there's something triggering dysregulation of the immune system, but other than knowing that they trigger events who lead to a disease relapse, we don't have a very clear understanding of what's going on within the immune system to go to proceed to damage the protozoites and induce proteinuria. So there have been different ways and approaches to enhance our understanding of this. One has been to look at genome-wide association studies.

And there have been different studies that have been performed over the last few years. Here I present results from the three main ones that have been published in the last 10 years. And these have given us a very important response because they have showed us that there are different genetic variants predisposing to developing steroids or including nephrotic syndrome that are in the HLA locus.

So that SNS is associated. with the HLA region and in proteins encoding for MHC class II molecules. So clearly an involvement in the interaction between B cells and T cells and T cell post-stimulation, further reinforcing the fact that these are diseases that are driven by an immune dysregulation.

This is a nice review that I'm just showing to you. I'm not going to go through the results of all the different GWAS studies that have been performed up to now. It was published this year in Pediatric Nephrology, and it shows both the HLE SNPs and also SNPs that have been found outside of the HLE region, most interestingly one, for example, in nephrin.

And I think that this approach is an approach that will allow us moving forward, not only to identify potential mechanisms and therefore potential therapeutic targets, but also to cluster this extremely heterogeneous group of patients. into patients with a more similar phenotype in terms of time of presentation, modality of prognosis. So I do have the fortune of being able to very briefly show you also unpublished results from the most recent and most important up to now in terms of numbers, genome-wide association study to which many of us also have been able to participate.

which has been led by Matt Sampson at Children's Hospital Boston and has been able to collect the 2440 cases from all over the world and he has presented these results at IPNA so I'm not going to go through them in detail but he has been able to reproduce some of the previously identified SNPs and also to add a few new ones and I think again that this type of an approach will be able to help us define our targets and define our sub-populations moving forward. within this extremely heterogeneous disease. So moving on from lessons learned from GWAS, lessons learned from clinical efficacy, and so from the therapeutics that we use for this disease. So the most useful indirect evidence that has helped us move forward in treating these children has come from the efficacy, completely unexpected, of rituximab. So the anti-CD20 miracle antibody that depletes CD20-positive B cells.

And this is the definitive trial published by Ijeoma and the Lancet. And this is just something showing from a quarter of 40 patients followed in our hospital, how this is the only therapeutic agent up to now who can be disease modifying. So there have been children treated with rituximab, even with a single infusion, but never relapsed again, even looking at them five years down the line.

So this clearly indicates that we're doing something right when we're giving this thing up a drug, but we don't know why. So rituximab and also opatumumab, the humanized form of an anti-CD20, has an effect on depleting B cells. And by depleting B cells, it also depletes, because it's precursors, short-lived plasma cells.

This is shown clearly here, where we had the opportunity to compare rituximab and opatumumab, and we found that the effect on all the B cell subpopulations was similar at one month. and also on CD20 negative plasma blasts, as I showed in the previous slide. So we just reconfirmed this, and we didn't find a difference between opatumumab and rituximab in terms of B-cell depletion.

We also looked at the different subpopulations with these two drugs, and what we found was that these drugs are able to very effectively, and for a prolonged period of time, deplete memory B-cells. And why do we think that depletion of memory B-cells is particularly important? We think that this is important because we and others have shown that memory B cells are able to predict relapse following anti-CD20 treatment. And this was initially shown by our group in a small cohort in 2016 and has been where we showed that this percentage of memory B cells was highly predictive of relapse.

And this was subsequently reconfirmed in a different cohort. So with a completely different. ethnicity in the Indians by Arvind Baga's group and also by Paolo Cavetti at Columbia. We also have unpublished data now showing that memory B-cell levels at baseline, so before giving anti-CD20, can predict relapse and that children having higher levels of memory B-cells are much more likely to relapse. So this can be something that can allow us to fingerprint our patients moving forward and to choose who to treat and when.

So we know that B-cell depleting treatment has a large variety of effects on B and T cells. And I'm going to show you more in depth and more in detail the evidence that we have for a role of B cells in the pathogenesis of nephronic syndrome. So this is just an overview figure.

I'm going to go through the different parts of it. So in terms of B cell phenotype in children, moving backwards. So based on our observation of the effect of rituximab.

in these children. We tried to look at a B-cell phenotype at disease onset in these kids. And what we confirmed was that even at disease onset, total B-cells are higher in children with serosensitive infarct syndrome before receiving any other form of treatment. And also the different subclasses of B-cells are higher.

But particularly what we saw is that with treatment, all the different subclasses go down. So you see that transitional, mature, and total B-cells go down. The only one that instead goes up are memory B cells.

And we think that really this is, again, an indirect sign of the efficacy and the importance in the pathophysiology of this subgroup of cells that instead remain lower in remission. So in adults, this is also the case. So there aren't many studies in adults, and we're trying to address this in minimal change disease in adults.

But from the results that we have, it's not like that. Memory B cells do not seem to be important in adults, where instead there's an increase in plasma blasts. And if you think about the way in which adults respond to prednisone at disease onset, minimal change disease, it really does seem to be a different disease. So we think there's something very different between the pathophysiology of this disease in children and in adults.

And the initial data in B cells seems to confirm this. So when we look at pathogenic antibodies, there's been a lot of evidence looking at different types of pathogenic antibodies in nephrotic syndrome. This obviously brings more value to the hypothesis that B cells are important in nephrotic syndrome. So one of the experiments that has been done and has actually done by our group and looked at hyposialylated IgM and its role in nephrotic syndrome. And this stemmed from an initial observation in which we saw that a subset of children with INS at onset, so before receiving any immunosuppressive therapy, had very high levels of IgM on the surface of their T cells.

And these were the children who did more poorly, who relapsed sooner. So they seem to identify a subgroup of steroid-dependent patients. When we looked at patients in different moments of the disease, we confirmed this. The steroid-dependent nephrotic syndrome children have higher levels of hyposialylated IgM on the surface of their T cells. So why is this important?

This took us quite a lot of time to try to understand. So pantomeric IgM is sialylated in a whole variety of spots. The sialylation of pantomeric IgM helps it bind, and these are natural... IgM to the surface of T cells and actually inhibits the T cell proliferation. If the IgM are desialylated, and these patients do have desialylated IgM as we showed here, these have a reduced capacity of proliferation, of inhibiting proliferation of T cells.

And this we did performing experiments using commercially available IgM. And we showed it here where we show that the desialylated IgM are proliferate just as much as T cells that do not have IgM on the surface, whereas when they have sialylated IgM, they proliferate much less. And this is also connected to the loss of inhibitory capacity of dexamethasone.

These are in vitro results, of course. But in vitro, these cells are less susceptible to inhibition of dexamethasone proliferation if they have the sialylated IgM on their cell surface. So other pathogenic antibodies have been identified. There's been a flurry of interesting experiments and results in the last few years.

And the three main antigens that have been identified are CD40, UCHL1, and nephrin. So these are the three papers that have been published. In nephrin, anti-CD40 antibodies were shown in adults in a cohort of relapsing FSGS patients by Delville. And this was also shown, these antibodies were shown to be pathogenic in mice. In children, Jorde-Shen's group found the existence of anti-UCHL1 antibodies, and again showed that these were pathogenic in a mouse model.

And much more recently, both in children and in adults, Astrid Wein's group at MGH showed the presence of antinephrine antibodies, which were particularly interesting. and also quite fascinating given also the GWAS results in minimal change disease, both in adults and in children, both in the circulation and in renal biopsies. Another aspect of podocytes in which they can act as antigen-presenting cells, which is hotly debated, but I thought I would mention it because it does have therapeutic implications, is the fact that in some situations, injured podocytes appear to be able to act as antigen-presenting cells. by expressing B7-1 or CD80. So this is hotly debated because this was first shown by Jochen Reiser's group in 2004, where he showed that myriam-potocyte upregulation of B7-1 induced proteinuria.

However, when almost 10 years later, this was brought into the clinic with the Abathacept paper that was published in the New England Journal, Abathacept was shown to be effective in five of SGS patients. And Abatacept, excuse me for not mentioning this, is a fusion protein that is composed of unifying IgG1 and CTLA-4 and binds CD80. And this monoclonal antibody appeared to be effective in a small group of patients, even though they were patients that had received a lot of stuff at the same time. So it was not very easy to dissect out the positive effect of Abatacept. More importantly, these patients seem to express B7-1.

in their renal biopsies. However, this was not reconfirmed subsequently. On our side of the pond, in a European cohort, B7-1 was not expressed in patients with Fsgs and this was also looked at by Giuseppe Rimossi's group in Bergamo and also not confirmed. And also prospective trial of nine patients with Fsgs with a relapsed post-transplant treated with Abatacept did not show any response. However, more recently, again, the Americans have showed another positive cohort response in which they treated 12 patients, many of them children, nine out of 12 were children.

And they showed that in the cohort of patients who did express B7-1, there was a clinical response to Avastastept. Of these nine patients, six had positive B7-1 in their renal biopsies. The ones who did not either had not had a biopsy performed.

or had a biopsy performed before reperfusion. So it wasn't informative because they didn't have their relapsed of disease anyhow. So I'm not sure that this is data that will be able to be reconfirmed.

But I think that given the absence of alternative therapeutic options, it's something that we need to keep in mind. So the fourth and last thing that I wanted to discuss about is antibody-independent effects on B-cells that might also have an effect on protocytes. And on this, I just briefly wanted to mention BAF. BAF is a B-cell survival cytokine.

And we thought that BAF might be useful because there is a monoclonal antibody targeting it called Belimumab. There are also others. But we thought that given the fact that this had been shown to be useful also in membranous nephropathy, it might be useful in frequently relapsing nephrotic syndrome.

This is a very small pilot study that was performed and that did not show a significant effect. However, I do not think that Belimumab should be ruled out just because of this initial observation. Thank you. I think that if we think about the importance that B cells seem to have in these forms, difficult forms of nephrotic syndrome, and if we think that when we're giving a depleting agent, such as anti-CD20, we're then getting an increase in BAF to reconstitute the B cells, at that point, giving BAF might be useful. This has been done in lupus nephritis.

So you give rituximab, and then when the BAF goes up, you give the limumab and prolong the effect of the anti-CD20. This is something that could probably quite safely be used also for nephrotic syndrome. So going back to the clinical scenarios, I'm going to go back to the severe serodependent child that I had showed you before. We have all encountered in our clinical practice this type of a child with all these relapses, all this toxicity, normal renal function, but a really, you know, impaired quality of life.

What do we do in these patients? So one approach which has been done by Claire Dossier in Paris, which I think is innovative and interesting, but quite, you know, with significant toxicity, is to give Opinotuzumab, which is a new anti-CD21-formal antibody, a class 2, which has the ability to induce a more prolonged and protracted B-cell depletion, followed by Barachumumab, which is an anti-CD38, so plasma cell depleting agent. And they treated 14 patients in which they really had, you know, no other thing to do.

These children had tried everything under the sun and they had not improved. And in these children, I think I remember correctly, nine out of 14, they had a very significant clinical response. However, they did also find the prolonged use need for IVIG for about one, one and a half years depending on the patients.

But this is something that is an object and this are we moving in the right direction. So what about trying something that is a little bit less toxic? So we thought about trying mesenchymal stromal cells quite a few years ago following a pilot study that we had performed, which we thought, we know that mesenchymal stromal cells are immunomodulatory. They have zero side effects.

So let's try to see if we can purify them and reduce them from these patients and use them in a safe fashion. So we had performed a pilot study in which we took bone marrow when we were doing biopsies, renal biopsies from five children. really severe forms of either steroid-dependent or steroid-resistant but responsive to immunosuppression children.

We took a little sample of their bone marrow, we expanded it to form mesenchymal stromal cells, and we looked at their phenotype and function. And we saw that their phenotype in terms of capacity to differentiate into the different tissues that mesenchymal stromal cells produce was absolutely normal. We also looked at their immunomodulatory function in vivo, I mean in vitro, sorry. compared to healthy controls, so to mesenchymal stem cells from healthy controls, and we saw that they were also completely normal.

So based on this, we designed a phase one study, called the METHNAP study, in which we chose to try prospectively to use autologous mesenchymal stem cells in children with very severe forms of steroid-dependent nephrotic syndrome. They would receive two MSC infusions at day zero and day eight. And from a month after the first infusion, they would start to bring all the immunosuppression that they were on.

And we would look at them for 12 months. Obviously, as this is a phase one study, the primary outcome was safety and feasibility. And this is the demographics of the study of the children and adults that we treated. Overall, we treated 16, 10 adults, I mean, sorry, 10 children and 6 adults with a very long disease duration and also having received a ton of immunosuppression previously. And very briefly, these are the results.

Just look at very, very simply at the number of relapses in the 12 months before and the 12 months following treatment and the number of immunosuppressive medications before and after. And what we saw was that there is a pretty striking effect in the children and absolutely no effect in the adults. So I think this is interesting and probably points to the fact that there is really a difference in these diseases as they move forward. So the last clinical scenario that I wanted to go back to is the nightmare, the secondary steroid-resistant nephrotic syndrome child who then tends to relapse. So going back to the child that I had described before, this child promptly relapsed following his renal transplant.

He actually relapsed on the operating table. As soon as his diuresis restarted, he had proteinuria. And we confirmed this histologically. followed the classic protocol of rituximab followed by intensive plasmapheresis. However, after four weeks of rituximab and 12 sessions of plasmapheresis, we decided to give him one dose of ibuprofen instead of rituximab.

And he actually continued plasmapheresis but achieved complete remission. So we had no idea whether the remission would have arrived even if we had given rituximab in all frankness. However, what we were surprised by, and we usually do not see with rituximab, is the prolonged depletion.

So at six months from OB, his lethal depletion is still complete. He had managed to taper his blood narcofasces to two sessions per week and he had an abnormally renal function in UPCR, borderline normal both of them, but satisfactory given the severity of the phenotype. And at nine months, he's accused of doing blood narcofasces now once a week and he has a satisfactory UPCR of 0.48 and a stable EGFR.

This is also instead a small pilot study that we are hoping to be able to implement for these particularly difficult patients within the ERG treat grant that has been submitted. It's going to be called the Focal Step Study, and it addresses FSGS relapses post-renal transplantation for patients with severe multidrug-resistant necrotic syndrome. And in these patients, we would... We're interested in following a stepwise approach in which we try one approach after six months.

If it hasn't worked, we try another approach and so on. And the three approaches that we're going to look at are Abotacept only in B7-1 positive patients. We're actually curious in doing this just to see if B7-1 is actually found in our patient cohort or not.

And in those who are Abotacept negative, Obinutuzumab for six months. If they have a positive response, we continue. Otherwise, we move on to DARA2-MOM.

So in conclusion, I think for this group of diseases, which are particularly challenging, we are better off now than 50 years ago, but we are far from meeting the clinical need. There is a need for collaborative efforts, both for pathogenesis and for prospective randomized control studies. Pilot studies with innovative approaches, sampling pre-and post-treatment are crucial. And I think there is much more information to be gained by large unbiased genomic and transcriptomic efforts.

I'd like to thank all my collaborators, our funding agents, the patients, their families, and you for your attention. Thanks. Thank you very much.

That was a high word. In the room. Okay, okay.

Maybe I start. Looking at the data with memory based cells. Would you consider that maybe, you know, looking at the initial data already, you would sort of take the vision on the intensity of anti-CD20 treatment and would that be a path to publish? Yeah, that's a really good question.

So essentially, that's exactly the direction that our research is trying to move in and the whole idea is to do tailored treatment. So I think everyone in their experience has, you know, tried rituximab in a lot of these different patients and there's some patients who respond really really well and others who really don't and we really don't have a very clear idea of why that is and that's really always been for me sort of the crux of the issue also because rituximab is expensive that also has quite a few side effects and you know we're not treating cancer here we're treating something that has other therapeutic options so trying to figure out which patients will respond is definitely where we're trying to go And I didn't have time to show you all the data that we have, but we have clear indications now that age plays an important role and that older children will respond better to rituximab. So it's probably quite appropriate to wait and get it when they're a little bit older.

They have a much higher chance of responding. And we also think that, yes, the levels of memory B cells at baseline are going to influence how well they respond. So we probably need to find something else to pull down their memory B cells.

or say okay maybe for this child we need to choose a different therapeutic approach. Thank you. Yes, so I noticed that the GWAS low side is showing the beginning for the steroid-resistant metabolic syndrome with NPH as one, whose mutations cause a... Not for steroid-resistant, for steroid-sensitive.

Yeah, that's what I mean. Okay. Whose mutations cause a classic form of steroid-resistant metabolic syndrome. So how do you interpret this? Is this a steroid-resistant modifier effect or is this related to the fact that the antibodies are actually binding to nephrin?

Yeah, so I think we were all super perplexed and really surprised when nephrin started to come out in the first cohort as one of the SNPs that predicted steroid-sensitive nephrotic syndrome, but it's been confirmed in a really large cohort now. And in parallel, having aspirin binds showing the anti-nephrin antibodies sort of brought it home, I think that's by far the most likely explanation to that finding. Any questions on that? No.

What's the question? I don't know. Marina, there is some interesting experience in the use of lipoproteins to treat both METI, but also post-trial plant Fsgs.

And do you think that lipids could be a potential target for this condition? Yes, I do. I think that is also something that we would probably need to acquire more experience with. putting together a large group of patients, but it's just very difficult.

Even there, the impression is that it doesn't work for all patients. And another thing that I have discussed also with Anya Guzil, who treated quite a few patients at the Evelina in London, was whether the LDL-feresis makes the patient more susceptible to the CNI, because achieving a good and consistently high CNI therapeutic level when you have such a important, this lipidemia is very difficult, and this may be a co-cause of the response that we see in these patients. But this is just something that we need to look at more in detail.

We don't really know. All right. Thank you very much, Marina. And then we go as well as a leading national and international registrates on nephropic syndrome.

the founder and chief scientific advisor of PureSpring, focusing on the genocide for renal gene therapies. And apart from all of these qualifications, he's a great football fan and he's already given his opinion on GetDates events. Thank you for your extreme hospitality of Germany in particular.

It's always. As an England football fan, it's always difficult to come to Germany during a major football tournament. But this time I feel much more relaxed.

Nevertheless, you're going to knock us out in penalties in the knockout stages. So it's too early. So thank you. It's a fantastic meeting.

I've really enjoyed it so far. I hope that I can continue with the theme. Marina's very nicely introduced necrotic syndromes for us.

So I'm going to talk to you predominantly about... the genetic forms of nephrotic syndrome and clearly there's a huge unmet need here. There's more than 80 genes that have been described that if mutated causes nephrotic syndrome and the most common of these is NPHS2 which encodes prodosin so at least 20% of the monogenic causes in Caucasian populations are caused by NPHS2 mutations. There's no current treatment. And what I want to talk to you about today is a couple of studies that have been done by very talented PhD students in my lab over the last few years that are looking at ways of addressing that.

So the first, and as I say, so most of the genes, well, genes that we all know about, they're expressed in the podocyte. The genes in particular in PHS2 are expressed in the slit diaphragm and causes the most severe, the slit diaphragm. mutations tend to cause the most severe phenotypic disease that we see in the earliest onset.

So we wanted to look at podocin. Podocin was our gene of interest for these studies. Podocin, biologically, it's a podocyte-specific membrane, integral membrane protein. And mutations, as I say, the commonest form of genetic nephotic syndrome.

And the most common, the most frequent mutation is R138q. And this is a... mis-boulding mutation that causes mist-trafficking of ptosin, the ptosin gets stuck in the endoplasmic metabolism and the question we wanted to ask was can we discover small molecules that will correct this trafficking defect?

So that was the hypothesis. That's what we wanted to do, kind of correct the defect. And then once you've corrected that trafficking, does it still function? So we started with our tools, our in vitro tools. So we had mutant pidosin cell lines from a patient with a Monterey accumulation and compared that to our wild type cells.

What's the best way of using pointers? Yeah. So what you can see in these cells is that the pidosin is mislocalised. and it actually co-localized with calnexin which is an ER marker so that's in the mutant cells you can see that you've got mistrafficking to the endoplasmic reticulum.

So the idea came about through a discussion with a collaborator in Paris, Alexander Edelman, and he works on cystic fibrosis so the CFTR molecule and he found that this interaction with the mutated CFTR, the common mutation, the F508 deletion, interacts with cytokeratinase, so a housekeeping molecule. And if you disrupt that interaction, that corrects the functional defect in that misfolded protein of CFTR. So we were talking about this and we thought, well, could the same occur for pidosin?

So could it be that that misfolded pidosin also has an interaction with keratinase and that's responsible for its mistrafficking? So we... We used our tools, our cell lines, to have a look at this.

So first of all, we looked to see whether keratinase associates with the misfolded 138-cupidosin, and indeed it does, by using a couple of methods here. This is a proximity ligation assay, which uses the rest of the tagged oligonucleotides, which you probably can't see very well on the screen, but what you're seeing is if the molecules are close together, you get more red dots, and that's quantified here. So you can see in the mutant cells, you get much more association of keratinase with the mutant pidosin. And you can do the same by more standard means by immunosecretation, for example, we can show that keratinase associates with the mutant pidosin, so that pulls down, keratinase pulls down the pidosin and doesn't pull down wild type pidosin.

So doing some structural modeling with Richard Sessions in Bristol, he showed that that mutation replaces charged arginine with a neutral residue and that essentially turns, creates a hydrophobic patch and that hydrophobic patch is able to then again structurally predict and combine keratin. So keratinase then binds to that misfolded protein and directs it to a different degradation pathway rather than trafficking normally. And using the molecule that Alexander had used to correct the defect in CFTR. He predicted that that molecule sits in that pocket and therefore will disrupt the binding of keratinase to the misfolded protein.

So we wanted to test whether that molecule can rescue the trafficking of pidosin. So this is some of the first experiments that were done. This is using turf microscopy that's only looking at the cell surface. So you can see pidosin nicely at the cell.

You can see it producing nicely at the cell surface. It might be better to use this. So I'll use this so I'm not accidentally moving the slides on. So you can see it nicely at the cell surface in the wild-type cells, in the mutant cells, they've been producing at the cell surface. And you add this compound and it comes back to the cell surface.

So you're seeing a nice relocalization of producing with the... with the molecule and that also rescues the function of these cells in vitro so what we see is normally in wild type cells compared to mutant cells, mutant cells are less adhesive so they have this cytoskeleton defect which causes them to be less adhesive and if you add the flora7 compound that significantly rescues the adhesion property so what it's showing is that you can both rescue the localization of the protein and also rescues the function so that even though the protein is mutated it still has function once it gets to the cell surface which is important. So we wanted to test that in vivo, so we created a mouse model and this is starting with a mouse that Karin Antonyak's lab kindly gave us and what we effectively did without going through the details is created a mouse which had the 140q which is the mouse equivalent 138q on one allele and a knockout an inducible knockout on the other allele so that gives severe disease we then want to oops We then wanted to treat the mice with Foro7, so we used a mini pump and these mini pumps were implanted under the skin and they delivered the drug for a period of four weeks.

And what we saw very nicely was actually a very very significant reversal of proteinuria. So this is over time, you can see up to week four and by week four in the treated mice, I can't even see my own pointer now, in the treated mice you're getting proteinuria right back down to baseline levels. in this mouse model.

And then we can see localization, so the correction of the plasma localization of pidosin, these are mice that haven't been induced, so you should see pidosin nicely on the cell surface there. In the mice with induced disease, that cell surface localization of pidosin disappears, and then you add the drug for four weeks and you've got the plasma membrane localization of pidosin again reappearing. And the histology was remarkable, actually.

It's a complete reversal, complete rescue of histology after four weeks. And the ultrastructures and polycyclic processes was also completely restored. So we saw effectively a complete rescue of this mouse model using the drug. So that's sort of as far as we've got with that particular study.

So we think that keratinase binds misfolded pedosin in the ER and traffics it for... of protosomal degradation, if you interrupt that with a small molecule, that rescues the localization and you get full function back in vitro, but much more convincingly in vivo. So we've done some work towards a medicinal, so this is effectively what's called a tool compound. And what you need is a medicinal chemistry program to optimize the drug-like properties of that compound and make sure that you can get, for example, oral bioavailability. and make it into a compound that has minimal up-target effects, etc.

So that's an ongoing programme of work. But what I think is quite nice with this is that this has the potential to treat other genetic misfolding diseases. So lots of mutations cause misfolding of proteins.

Clearly, this works for the CFTR protein. It works for pedosin. Alexander's group has shown it works for alpha-1 antitrypsin. So one of the diseases we want to try is cystinosis, which also has misfolding mutations. So I think it has potential to look at a number of different misfolding proteins by working on that particular mistrafficking biology.

So I'm going to switch tack now and talk more about gene therapy. So gene therapy is clearly the other approach that people have thought about in kidney disease. But so far, it's not been achieved. And it's been very successful in a number of other diseases where. you can target the cell types or you can target the organs perhaps more easily than the kidney and I won't go through them all but you'll know about for example treatments for haemophilia B and there was an approval yesterday for a drug for haemophilia in the eye you clearly you can get to the eye much more easily than you can to other organs with direct injections so there's been several successes in treating eye disease and spinal muscular atrophy and traumatic effects in SMA.

So that's sort of giving you a flavour of where gene therapy is at the moment in medicine but where's the kidney disease? That's a question a lot of people have asked. So we started a project, so as I say so far it's been successes mostly in monogenic diseases in other organs and the vector is usually adeno-associated virus.

So we wanted to ask the question, can we do the same for a monogenic kidney disease? And what are the major hurdles? So the hurdles include that you need to target the correct cell type. And there are at least 32 different cell types in the kidney. And the cargo capacity of AAV, the virus that's used, is relatively limited.

So is the gene of a suitable size that it will fit in the payload? And then will it last? Does it stay long enough?

And then what kind of dose do you need and what kind of off-target toxicity is that going to give you? And the approaches that we wanted to use to try and address some of those hurdles include the use of cell-specific promoters. And that's a real advantage of targeting the podocyte because proteins such as nephrin and prodosin are pretty much only expressed in the podocyte. So we have really good cell-specific promoters. Can we get a capsid, an AAV capsid that has tropism for the right cell type?

So different AAV serotypes have different tropisms for the different cells, different species even, so that's a little bit of a trial and error. Can you get local delivery to the kidney rather than using systemic approaches and can that gene last for the duration? So this was a PhD project initially for Wen, who was one of my clinical PhD students.

And the hypothesis was that podocyte is effectively an ideal target for gene therapy. A, because we have those promoters. B, because it's a terminally differentiated cell that once you get your gene in, it should stay there for life.

So that's a picture of Wen. So we use AAV, as I've described. It's a small virus.

It's non-pathogenic. It doesn't integrate as episomal, which is good. It has low immunogenicity, but as I say, it has a relatively small packaging capacity. So what we wanted to do was use AAV as the Trojan horse, as the vector to get transduced the podocyte and express a wild type version of NPHS2, the podocin gene that encodes our protein.

And so Wen made lots of different constructs that fit into the AAV virus. The promoters we used were either human or mouse nephrin promoters. We put mouse podocin or human podocin, depending on whether we're using human cells or the in vivo mouse experiments, and also GFP tags as a marker protein. And various other constructs are used. in the field to try and maximise your ability to express your protein in the cell.

And initially we did in vitro experiments. So the in vitro experiments were transduction of our cell lines. So we use podocytes and as controls we use endothelial cells and proximal tubular cells. And busy slide, but what you can see is that with a CMV promoter, so a generalised promoter, you get nice expression in all the cells.

but with a nephrin promoter you get expression only in the podocytes and not in the other cell types so we saw nice specificity of expression with the promoter in vitro so that was a good start and then we wanted to to put the wild type podocin into the mutant podocin cells so the R138q cells that are already mentioned so we did the same thing we transduced our podocytes we could see nice expression of the HA tag and a pidosin in our cells and it goes to the right place and also it's functional so using the same adhesion as they have mentioned before you get normal adhesion in wild type cells impaired adhesion less than 50% adhesion in the mutant cells and then the cells that have been transduced wild type pidosin you get a pretty good rescue of adhesion which is statistically significant so we we're demonstrating we can get it into the cells and it's functional, what about in vivo? So again we used a pidosin knockout mouse, this time it's an inducible knockout of the whole gene. We used either a human or a mouse promoter, so we tried both promoters just in case there's a difference and put mouse pidosin as our gene of interest into the construct.

And we injected the virus via tail vein into these mice. We injected it before induction of disease, so two weeks later we induced disease, and then followed them through life up to 16 weeks. And what we see is that you get very significant reduction of proteinuria over that time period. So this is the saline-treated mice, and this is quite a big scale, so they get massive proteinuria.

And you get very significant attenuation of proteinuria. The two colours just indicate the mouse and the human promoters. And actually very nicely, you get long term efficacy here.

So the mice that have saline, they're all dead by about 100 days. And the mice that have been treated, about two thirds of them had survival well beyond a year. So it appears that the effect is long lived and we're getting really good efficacy of the drug. And the effects on real function, this is biochemistry.

So, for example, cholesterol comes down very nicely, the albumin. improves, not yet significant in that small initial group, urea comes down and creatinine comes down. So biochemically they're also improving and histologically we can see nice improvement as well.

It's not completely normal, but we're getting mostly sort of back to normal structure and normal histology in the treated mice. And that was assessed blindly by pathologists. So we can see that the number of sclerosis gloves, again, the two bars indicate the human promoter or mouse promoter there was no difference between the two but we're getting nice improvement in sclerosis and less interstitial fibrosis as well And we want to understand why is there that little bit of variability? Why are some mice doing better than others? And could that be due to the delivery, the route of delivery and the fact that it's actually technically can be difficult sometimes to get your tell when injection just right, etc, etc.

So what we did was firstly look at the level of podocyte transduction in these injected mice. And what we see is. there is some variability in transduction and we're getting about between say 15 and 30 percent transduction of podocytes using the HA tag so there's variability in transduction in individual mice and does that correlate to proteinuria and it seems to so what we did was we correlated the urinary credit for this very interesting case these cases are always so difficult to to deal with because of different implications and the choosing direction. I'll carry on with that brief pause. So we looked at urinary albumin creatinine ratio and correlated that with the copies of single-stranded DNA, the virus is single-stranded, so this is by PCR. And even in this, again, relatively small numbers in the first experiment, but we're seeing a significant correlation of proteinuria.

with the load of virus that you get in the renal cortex. So that seems to suggest that at least some of that variability could be due to the mode of delivery and the success of delivery. So overall then we've got a sort of nice principle.

The next experiment we did was that was effectively a prophylactic dose because you're giving your virus before the mice gets proteinuria. So next we did an intervention experiment where we induced the proteinuria, this time again this is going back to the R140q mouse which I described in the first study, and these get very severe early onset proteinuria, so it's a severe model of disease. So we waited two weeks till they get proteinuria, then we gave the intervention and we followed them through life.

So this is the saline mice, what they get is very significant proteinuria. We followed them for a bit longer and eventually they get renal failure and that proteinuria curve comes down. And then with the intervention, we see that initially you get an increase in uptake in proteinuria, flattens out and actually comes down much better.

I think we're getting better at the IV injections, same dose of virus and it comes right down. So your proteinuria comes right down and the albumin level has improved significantly. So that looks very encouraging.

Just looking at the histology again, we can see that in the glomerulus we're getting, using nephrin as a marker, we're getting pretty good transduction efficiency of our podocin into the glomerulus. And this is the histology at the end of treatment. So the untreated mice and sclerosis from the treated mice have relatively normal histology.

So I think that's a really nice critical concept that we can successfully target the podocyte now. with a specific promoter we can get our protein of interest expressed and we can get long-term rescue of the disease phenotype so that that's a sort of nice concept that's now led to the spin-out of a company called Pure Spring so we were very lucky to get venture capital funding from a company called Syncona that has already spun out four other gene therapy companies so they've got lots of experience And it's really good to present this in this audience here, actually, because our plan now is to move forward towards first in man treatment by 2025. So the timeline that we're coming to, and we started to have discussions already with France and other CKOs in the field about recruiting patients and getting a registry of patients that will, if you think about patients who are going to go into the trial in 2025. Actually, those patients, given the natural history of the disease, are probably going to present in the next two or three years. And those are the patients we need to identify.

So we're going to try and sort of, you know, global effort. But I think Europe's way, way ahead of the rest of the world in trying to identify these patients over the next few years with really good natural history, prospective and retrospective studies. So that we can really move rapidly into a first in man trial, good recruitment.

And recruitment is one of the biggest obstacles in rare disease research, as we all know. So, just a quick word on the delivery. So, clearing the mouse, we had to give systemic delivery, because it's very difficult to get into a mouse ring artery. But the advantages of local delivery, and we're working on the local delivery approach with doing large animal experiments, and I think this is going to be successful, is that you can...

I think I managed to regulate the volume. You can... and give those much lower doses but also and you can also commercially and this is important you can you can get much more virus out of the batches that you that you make and that affects the cost of foods hugely and as we all know the gene therapies that have been given so far have been very expensive and this is one of the one of the major drivers is how much virus you can make from the commercial batches so this is a very expensive process these are all things I'm learning as we go along So a local delivery will mean that you can give less dose, you get less off-target effects, usually viruses go to the liver preferentially, and you can make more of the product. So just to sort of finish off, then I think this opens a whole exciting new field, certainly for me, who's been working with podocytes for 20 plus years.

I think, you know, clearly I'm a believer in podocytes, but the podocyte... we describe often as a master regulator of glomerular function. And we know it's involved in lots of biological processes from regulation of filtration by isolate diaphragm proteins, maintenance of architecture by the collagens and the basement membrane proteins, regulation of the immune response and planetary response within the glomerulus, for example, by regulation of complement, and also metabolic effects. So Richard Coward, who's in Bristol, works very... has done some groundbreaking work on showing the polycytes and into the insensitive cell and as important in the initiation of diabetic nephropathy.

So that means that there's lots, if you can target the polycyte correctly you could go well beyond just the monogenic diseases. So you know the monogenic diseases are the obvious ones to start with, with what I've described, and diseases such as Alport syndrome, but you can also go into targeting inflammatory responses by a complement pathway or the insulin pathways as well for diabetic disease. So I think it opens a whole field of possibility if you can target the cell and target specific pathways. So hopefully I've shown you sort of a couple of ideas for new therapies in nephrotic syndrome, the genetic. monogenic diseases in particular by small molecule approaches for misfolding diseases but also showing that aav therapy is feasible and effective in rescuing podocyte genetic mutations and gives you long-term rescue um lots of people to thank and i hope i've probably missed people out but um lab in bristol particularly val and when who did the most of those studies i've described really important collaborators who've helped along the way.

Amit Nathwani who's done the filial trials really helped us with the capsids and various thunders and pure spring as well. So thank you for your attention. Good morning.

Is your friend still online? It was an interesting case though. Any questions from the room? Is there a question online?

Thank you for this very nice talk. Do mice develop an immune response antibody production against human or mouse transgenic NPHP protein? I suppose NPHS or the reporter proteins, HADP, for example.

Yeah, so the immune response is a really important aspect and that is true for all gene therapy approaches so far. And so far, we haven't seen it in the mass, so we haven't seen any. inflammatory response in so far and when we've looked in the mouse tissues but for humans you know in human trials you know haemophilia etc there have been some you get neutralizing antibodies to the to the virus and there has been some you know inflammatory responses as well seems and often patients are treated at least for some period of time by using immunosuppressants for those kind of things but those are all aspects of you know getting getting into man that we have to address and look for very carefully thank you i have two questions so the first one is about the first part um so did you actually show that the keratin is expressed in for the scent because my understanding was that it's more like a bacterial cell that has converted to mesenchymal cell that expresses then only mesenchymal intermediate filament would they actually actually express keratin yeah a really good question so so wild type podocytes actually express almost no keratinase but actually the mutant podocytes do so it seems they're written there must be some kind of feedback loop where where your keratinase is expressed when you've got a misfolded protein there's some kind of presumably a misfolded protein response but we did we did look at that quite carefully And the other thing, the other question is about this second part.

So I find it very intriguing that you can provide a promoter for a specific cell type then you get cell specific expression. So I guess the rationale is that you have transcription factors that are prototype specific that recognize the promoter but aren't there other cells that may have these transcription factors but the epigenetic changes will close the chromatin and then these potential factors will not access the promoters, but when you provide it virally, they might find it. So did you really not get any extra renal or extra cortiside expression?

Did you look? Yeah, so I can sort of address two aspects of that. One is that, you know, one of the aspects of the promoter, we kind of designed a slightly smaller nephrin promoter, but we've looked at what transcription factors may bind, and WT1 is one of the key ones. So we made sure we...

kept the WT1 binding site which clearly is relatively cell specific. In terms of off-target expression I didn't show the data but we have got data there is some off-target expression so it's not perfect we see some in the liver but IV you know these these mice get a big dose of virus to the liver we saw a little bit in the spleen but we looked at pancreas, cardiac muscle etc etc so none in most of other organs so there is some and and I think that's something that we have to look at. carefully at during you know as we move up to large animals and clinical trials what what the effect that might be in terms of off-target toxicity okay thanks thank you um one more in the turn um are there any techniques on the pipeline for the local kidney disease uh yeah there is so so we're working on it and i'm not allowed to tell this is you you find that working with companies that that you have to be slightly careful as to when you release data um but we've got yeah good enough data that i think it'll work all right thank you thank you very much we move on