CME Slides Forum - E. Rondeau

A joint Congress by ERA-EDTA and ISN

EARLY EMT IN RENAL GRAFTS IS CORRELATED WITH OUTCOME

Eric Rondeau, Paris, France

Chair: Michael J. Mihatsch, Basel, Switzerland

Mai Ots, Tartu, Estonia

rondeau

Prof Eric Rondeau
Service de Nephrologie A
Hopital Tenon
Paris, France

Thank you very much Doctor Ots and Doctor Mihatsch. I’m very pleased to be here today to present our results and I thank the organisers of the meeting for the invitation.
So, in the next 15 minutes I would like to show you the results we obtained studying the epithelial to mesenchymal transition markers in renal graft and show you whether or not it could be correlated with the outcome of the graft.

As you know and as it has been said by the previous speakers, interstitial fibrosis and tubular atrophy progressively develop during the graft’s lifespan. This is responsible for a significant loss of graft with time after transplantation and this is probably the second cause of graft loss after the death of the patient with a functioning graft.

The mechanisms and the risk factors for interstitial fibrosis and tubular atrophy are all well known and have been reviewed before and as you know, we used to differentiate immune mechanisms and especially acute cellular or humoral rejection with the various risk factors which have been determined by clinical studies. Also the non-immune mechanisms which are the cold ischemia time, acute tubular necrosis and delayed graft function after transplantation, the use of calcineurin inhibitors but also all the non-specific factors which are involved in the progression of chronic kidney disease, whether the patient is transplanted or not and this is a very important point. As you can see, the mechanism may be immune or non-immune but all contribute to the development of interstitial fibrosis and tubular atrophy. So the question is how can this occur and what are the mechanisms of tissue fibrosis in the kidney?

The potential mechanisms of kidney fibrosis are listed here. The main thing is the production of interstitial extracellular matrix protein by interstitial matrix producing cells and these cells may arise from either the resident fibroblasts which in normal conditions are in very low numbers, it could derive from periadventitial cells or pericytes. The role of mesenchymal stem cells is also debated and one mechanism could be the epithelial to mesenchymal transition which is described here. That a tubular epithelial cell that loses its epithelial phenotype and gains a mesenchymal phenotype is able to digest the tubular basement membrane, to move into the interstitium and to differentiate into fibroblasts or myofibroblasts producing extracellular matrix protein.
I have to say that this mechanism which is very well known in embryologic development or in cancer has been debated in the kidney because people think that at least we can demonstrate that tubular epithelial cells can change their phenotype. There is a question whether they can move into the interstitium and can be true fibroblasts. These things have been well demonstrated in animal models and especially in the unilateral ureteral obstruction but not in other conditions and especially in kidney biopsies in men.
So, in the next part of my talk I will talk about epithelial phenotypic change or epithelial to mesenchymal transition and I will use both words similarly.

The molecular --- the EMT process is quite complicated but the main thing is that epithelial cells lose their characteristics and especially this molecule which is called cadherin and which is involved in homotypic ligation between two epithelial cells and this is very important for the polarisation of the cells and the differentiation of these epithelial cells.
Cadherin is linked in the intracellular compartment to α and β catenin which are described here to the actin microfilament so the cytoskeleton of the epithelial cells. During the process of EMT the disappearance of e-cadherin leads to the release into the cytoplasm of β catenin which can bind to another protein called LEF-1 and this complex migrates to the nucleus being like a transcription factor which is able to induce several genes and especially mesenchymal genes which result for example, in the expression of Vimentin in what were previously epithelial cells.

The molecular signature of epithelial to mesenchymal transition is quite well known now and there is an acquired or sustained expression of various transcription factors and especially Snail. It’s interesting to know that the specific over expression of Snail in tubular cells in the kidney is responsible for very rapid interstitial fibrosis and tubular atrophy. It’s also associated with an increased expression of FSP-1 which has been described by the group of Erik Nielsen. It’s very specific for these fibroblasts and in men it corresponds to the molecule S100A4. Also Vimentin of course, alpha smooth muscle actin in our hands it’s very late and it’s very difficult to demonstrate that epithelial cells express alpha smooth muscle actin. At the same time there is a decrease or complete loss of epithelial markers in these epithelial cells for example e-cadherin, desmoplakin, Muc-1, ZO-1 and so on.

In a previous study made in collaboration with Novartis we were able to produce this data showing that during the various grade I, grade II and grade III what was called at that time chronic allograft nephropathy we were able to see that there was a significant increase in several genes here shown in red and that the EMT transcription was in these red genes. At the same time according to the intensity of the fibrosis there was a decrease in these blue genes which include all the epithelial genes.

In various experiments or in clinical studies in the literature you can see some evidence that the EMT or epithelial phenotypic change can occur in kidney disease in almost all the cases with interstitial fibrosis and tubular atrophy.
Especially in the experimental models of tissue fibrosis such as the unilateral ureteral obstruction and that was a very similar paper, a very important paper from the group of Erik Nielsen and also for example, in nephrotoxic serum nephritis from the group of Raghu Kalluri.
In human kidney disease Maria Pia Rastaldi from this lovely county described in different kinds of human diseases the expression of mesenchymal markers by tubular epithelial cells and that was correlated with the degree of tissue fibrosis. Concerning the human kidney allograft several groups reported that epithelial cells could change their phenotype during acute cellular rejection or also during chronic allograft nephropathy.

So with this data we made the hypothesis that if the EMT process is important and is instrumental and responsible for further development of fibrosis in the kidney the EMT process should start before we are able to describe or to observe the tissue fibrosis. So it should appear at an early time point after transplantation.
Second if this is true, the appearance of EMT markers in early biopsy in kidney graft should predict the long term development of tissue fibrosis and a decrease in renal function.

So to answer this question we made two main studies which are summarised in the next slide. The first one was the collaboration with our hospital and Hopital Saint Louis in Paris the group of Denis Glotz and we chose to analyse 56 patients receiving cadaveric kidneys and with a stable graft function and who had a protocol biopsy at 3 months. In these patients the renal function was stable at the time of transplantation and as we know that the interstitial fibrosis and tubular atrophy occurs very early as you can see from this very important study from Nankivell, we chose to analyse the biopsy at an early time point that is at 3 months.

We use several markers I have no time to go into detail but I can tell you that we demonstrated Snail expression as 104 vimentin and the loss of e-cadherin and so on. But we chose to quantify this phenomenon especially two molecular markers that are the β catenin and the Vimentin. This is an implantation biopsy to show you the normal pattern

and you can see that β catenin is just localised at the basolateral side of the cell whereas there is no Vimentin expression in any epithelial cell in the normal human kidney. In what we call an EMT positive graft we can see that there is a mixed picture and as you can see here this is the normal pattern for β catenin and this is pathological pattern with β catenin which is released into the cytoplasm of each epithelial cell.

In the corresponding session you can see that when the β catenin is normal here, you have no Vimentin expression in tubular epithelial cells whereas when the β catenin is released in the cytoplasm you have a strong expression of Vimentin in this tubular. So we use these two markers, they’re very strong, very reliable, very reproductive and we are in the process of confirming, cross-validating these results with the group of Sandrine Florquin in Amsterdam.

We made a semi-quantitative score from 0 to 4 and as described here according to the percentage of tubular sections which were positive and we defined the EMT positive graft as more than 10% of tubular sections with the phenotypic changes since, we did not observe any implantation biopsy with more than 10%.

This is the description of the 56 patients who were involved in this study, most of them were on CNI treatment. The mean age was 45 years, the mean cold ischemia time was 21 hours. In this routine biopsy while the creatinine level was still stable and normal, we found 10 patients who had a subclinical acute rejection and maybe this is interesting for the next presentation.

What we showed is that at 3 months 41% of the grafts were considered with our score as EMT + in protocol biopsies. The statistical analysis showed that the risk factor to get positive EMT markers at 3 months were first the cold ischemia time and we were very surprised to see this but this makes sense with the previous studies and also the acute cellular rejection we could observe.

I want to show you one nice picture exemplifying that acute cellular rejection may play a role. This is a picture of tubulitis with two small nuclei of lymphocytes coming into contact with the epithelial cells. This is the pattern of β catenin in these epithelial cells where the others in the neighbourhood are still normal. So this is a demonstration of the role of probably the direct contact between T lymphocytes and epithelial cells.
So yes we can demonstrate the EMT process in the early phases in a stable graft at 3 months after transplantation.

So that the second question was does it predict outcome? So to do this we took an independent cohort in collaboration again with the Saint Louis hospital and the Necker hospital with the group of Christophe Legendre and we took in these patients we had 3 protocol biopsies at 3 months and at 12 months and we defined the progressor that is the patient in which IF/TA score that is interstitial fibrosis/ tubular atrophy score progressed of 1 or more in our scale.

As you can see, we could involve 83 patients in this study. Most of them received a cadaveric kidney donor. The cold ischemia time was similar to the first study, the mean creatinine was stable between 3 months and 12 months we had an acute rejection rate of 22% in these patients. Here is the summary of the results

and it is very interesting to see that the graft with phenotypic change at 3 months that was 21 out of 83. This disclosed a higher interstitial fibrosis tubular atrophy score at 12 months here as compared to those who did not have any epithelial phenotypic change at 3 months. Moreover more interesting those who had and an EPC score or EMT markers at 3 months had a greater progression of the IF/TA score than those who were negative at 3 months.

So yes, the EMT score at 3 months can predict the progression of fibrosis even in a small interval of time that is between 3 months and 12 months. There was a good correlation between the EPC score at 3 months and the IF/TA score at one year

and more interestingly a good correlation between the EPC score at 3 months and the progression of fibrosis between 3 months and 12 months.

In the multivariate analysis two factors appeared to be well correlated with the IF/TA score at 1 year that is the EPC score at 3 months and the interstitial score at 1 year. There was a value which was at the limit of significance for the donor age but for the progression of fibrosis between 3 months and 12 months the only factor which remained in the analysis as a significant factor was the EPC score at 3 months.

Interestingly the long term study showed that those who were EPC positive at 3 months had a greater creatinine plasma level as compared to those who were negative at 3 months and this lasted for at least 2 years after transplantation.

In addition, we recently repeated the study in a third independent cohort from the group of Marc Hazzan in Lille and we were able to show that in patients who were switched from calcineurin inhibitors at 3 months

and who just remained under MMF treatment without calcineurin there was no progression of the Vimentin score between 3 months and 12 months. No progression of the β catenin score.

However those who were on cyclosporine for one year were those who had a significant progression of the EMT score between 3 months and 12 months for both the markers.

When we excluded those who make an acute rejection of the CNI withdrawal, we showed that even under MMF treatment the score Vimentin or the score of β catenin could decrease significantly. In addition at 4 years after calcineurin inhibitor withdrawal we were able to see that the treatment in EMT positive patients at 3 months is associated with a lower renal function at 4 years.

So this gives us some indication that the chronic treatment with calcineurin inhibitors may be toxic and promote tissue fibrosis in those patients who were EMT positive at 3 months but not in those who were not EMT positive at 3 months and we are planning to start a randomised study based on the results of the renal biopsy at 3 months to see whether the calcineurin inhibitor withdrawal can be beneficial in such patients.

In summary, we showed that various factors are involved in tubular atrophy and interstitial fibrosis,. We suggest that the epithelial phenotypic changes/ epithelial to mesenchymal transition may be important in this process and that the epithelium could be the target of various injuries including the cold ischemia, the immune response, the calcineurin inhibitors, the pyelonephritis and we could add hypertension, hypercholesterolemia and so on. This is in addition to the initial condition of the kidney that is mainly the donor age, fibrosis which is already present at time 0 and which could be stable.

I would like to thank all the participants of this study and especially my group. Those who were very important to perform all these analyses Yi-Chum Xu Dubois and Alexandre Hertig played a very significant role and all the collaboration with the Hopital Necker and Hopital Saint Louis. Thank you very much for your attention.

Chairman: Doctor Rondeau thank you very much for your interesting presentation. There are probably questions. Yes please. Well this gentleman first.

Question: If I got you right the patients that are taken into your study are the cadaver transplants and the cadaver transplant graft is liable to acute tubular necrosis. I have a person with a live donor transplant. Acute tubular necrosis can inside the proinflammatory cytokines and will it influence the EMT parameters even at 3 months?

Prof. Rondeau: Sure, probably the EMT process may be also involved at the very early phase of transplantation for the tubular repair because the epithelial cells which are injured dedifferentiate, migrate on the tubular membrane and proliferate and gain a new differenciated epithelial phenotype. So may be at the initial phase during acute tubular necrosis EMT changes may occur also. I don’t think it’s the same process, it’s a repair process and it does not lead in any case to the rupture of the tubular membrane. Also it has been described in very severe forms of acute tubular necrosis.

Question: Do you think that visualising EMT could be clinically relevant and could influence our approach to treatment? I mean like ACE inhibitors is treatment towards and against TGF-β which is the main inducer of EMT. How dynamic is the process? Because if we biopsy and this is the moment and may be there is the reverse process of mesenchymal to epithelial transition which is also common in cells.

Prof. Rondeau: Yes you’re right. Actually it’s difficult to demonstrate the role of the ACE inhibitors at least in the clinical studies. We analyse subgroups to see whether those who were receiving ACE inhibitors or antagonists of the angiotensin II receptor were different but we have too small numbers to make significant analyses. What we know is that in vivo and in vitro too TGF β is very important to induce the EMT process and we know it’s also important to induce interstitial fibrosis. The idea that angiotensin II could promote EMT directly or through an increased expression of TGF β is possible. We still have to work on it and need a large number of patients to demonstrate it.

Question: Well doctor Rondeau a very short question. Have you ever seen a cell passing through the basement membrane?

Prof. Rondeau: Have you ever seen a red cell coming through a glomerular basement membrane?

Prof. Rondeau: We now have some evidence. That’s a good point we have some evidence that some cells are trying to escape and the group of Jeremy Chapman reported a paper, a very interesting paper in JASN showing that phyllopodia could be demonstrated going through or trying to go through the basement membrane.