Slide 1

I’m grateful for the opportunity to briefly discuss the existing markers of endothelial dysfunction and offer to you some most recent investigations which might contribute to future diagnostic and pathophysiological understanding of this syndrome.

Slide 2

Nephrological literature is awash with data on validity of various markers of endothelial dysfunction. These constitute first functional markers which can be studied using macrovascular or microvascular interrogation. These include unusually large von Willebrand multimers, markers of oxidative stress, circulating levels of adhesion molecules like selectins, VCAM and ICAM, microalbuminuria, endothelial microparticles, circulating endothelial cells and endothelial progenitor cells together with several cytokines that participate in endothelial dysfunction.

Slide 3

Along these diagnostic lines, let me offer some insights into something that we are pusruing, namely microvascular interrogation, and remind to you that, although the cardiologists are placing a significant diagnostic role on the post-occlusive reactive hyperaemia, in a large group of hemodialysis patients there was no major difference in the reactive hyperaemic responses (dialysis patients shown in grey) compared to controls.

Slide 4

However, when laser Doppler imaging was performed, we were able to show that there is impairment in capillary recruitment in response to reactive hyperaemia. Specifically, we’ve demonstrated that patients with ESRD and coronary artery disease have not only reduced numbers of these “hot spots” which correspond to the active or open capillaries in the skin, but they have reduced post-occlusion recruitment of these capillaries.

Slide 5

Thermal hyperaemic response showed much more promise (grey area again demonstrates the group of ESRD patients.)

Slide 6

There is growing interest in the spectral analysis of oscillations and Anita Stefanska has provided strong evidence that there exist different harmonics in these oscillations, some corresponding to heart activity, myogenic, neurogenic activity, with the slow component being reflective of endothelial function. Indeed, patients with ESRD and previously diagnosed coronary disease have severely impaired slow component of oscillations.

Slide 7

Moreover, thermal hyperaemic response has co-clustered with the high cardiovascular risk based on Framingham or CardioRisk system (the latter developed by Patrick Valance’s group.)

Slide 8

It also co-clustered with a mortality of these patients: higher mortality was associated with a diminished amplitude of the thermal peaks.

Slide 9

In collaboration with Carmine Zoccali’s group, Francesco Addabbo has demonstrated through screening 21 different cytokines that pro-BNP, IL-8, VEGF, MMP9, PAI-1 and IL-6 each correlated with a degree of renal dysfunction. In patients with stages 3-4 of CKD all these biomarkers combined contributed to the existing model of atherosclerotic burden in the carotid artery of these patients.

Slide 10

Let me now discuss a phenomenon that we and others described recently, namely endothelial-mesenchymal transformation, as a mechanism explaining microvascular rarefaction in CKD.

Slide 11

In previous studies employing a cDNA microarray screen of endothelial cells exposed to NOS-inhibitor we found that collagen XVIII gene was upregulated.

Slide 12

Now, this is an atypical collagen produced by endothelial cells. Moreover, this type of collagen has been described recently as a precursor for the powerful anti-angiogenic substance endostatin, its C-terminal fragment. Edmund O’Riordan in the lab has demonstrated that endothelial cells subjected to ADMA or L-NAME showed an increased expression of collagen XVIII at the message and protein level. Using subpressor doses of NOS inhibitor to induce an indolent endothelial dysfunction in mice, he demonstrated that these mice expressed much higher levels of collagen XVIII, an increased level of endostatin.

Slide 13

The endothelial cells obtained from kidneys of Tie2/GFP mice treated with L-NAME showed that there was a higher proportion of endothelial cells co-expressing alpha-smooth muscle actin or calponin - both markers of smooth muscle cells or myofibroblasts.

Slide 14

Moreover, not only ADMA resulted in this transformation of endothelial cells into mesenchymal cells but also the products that are downstream of ADMA signalling: endostatin and TGF-beta. All three resulted in the diminished expression eNOS and an increase in the expression of calponin and fibronectin.

Slide 15

Most recently in collaboration with Sebastian Bachman and Adelina Stoessel we performed studies of these mice. Sebastian has demonstrated that microvasculature in control animals was absolutely intact, as expected. There were multiple areas of microvascular denudation, with the flaps of endothelial cells protruding into the lumen in experimental L-NMMA-treated mice; areas of leukocyte infiltration of vascular wall; multiple areas of platelet adhesion to the sometimes denuded or sometimes non-denuded endothelial layer.

Slide 16

So I believe that endothelial-mesenchymal transformation is one of the additional contributors to endothelial and  microvascular dysfunction initiated by ADMA.

Slide 17

The proposed working hypothesis shown here predicts that endothelial-mesenchymal transformation as a result of accumulation of ADMA results in the microvascular dropout which in turn leads to the development of tissue hypoxia.

Slide 18

Let me now describe studies performed in the lab by Francesco Addabbo: proteomic studies of isolated vessels from the kidneys of the mice treated with non-pressor doses of NOS-inhibitor.

Slide 19

Francesco performed 2-dimensional electrophoresis which identified more than 2000 different bands in these isolated microvessels, but comparison of control and L-NMMA treated mice demonstrated the existence of 13 non-redundant proteins that showed significant difference from control.

Slide 20

Four of these non-redundant proteins are mitochondrial proteins and 2 of them, namely mitochondrial aconitase and enoyl-CoA hydratase, are mitochondrial proteins necessary for the normal functioning of the Krebs cycle. Just to remind you enoyl-CoA-hydratase provides acetyl-CoA to the Krebs cycle, whereas the aconitase-2 is responsible for the citrate-isocitrate conversion. The data demonstrated that both proteins are deficient in experimental vessels. Francesco confirmed it by western blotting and demonstrated that all treated animals had a significant reduction in the expression of both aconitase and enoyl-CoA-hydratase. In such a case, the entrance to the Krebs cycle in the vascular wall represents a bottleneck and glucose metabolism should be diverted from oxidative phosphorylation to glycolysis. Then one should expect to see abnormalities in the lactate production and this turned out to be indeed the case. The levels of lactate were twice as high in animals with mild endothelial dysfunction.

Slide 21

NO is critical for mitochondrial biogenesis and we asked the next question: Is it possible that NO deficiency could result in impaired mitochondrial biogenesis? Brian Ratliff in the lab performed intravital videomicroscopy in Tie2/GFP mice (endothelial cells labelled with green fluorescent protein). He labelled mitochondria using mito-tracker infusion into the renal artery and demonstrated that in mice chronically treated with NOS-inhibitor there is a diminished mitochondrial mass, consistent with impaired mitochondrial biogenesis.

Slide 22

So we believe that mitochondrial dysfunction manifesting in the suppressed Krebs cycle, conversion to glycolysis with accumulation of lactate (I would like to remind you that this is known as Warburg effect in tumor cells) observed in the renal  microvasculature and decreased mitochondrial biogenesis are all early markers of endothelial cell dysfunction induced by a sub-pressor dose of L-NMMA. Uncoupling of eNOS most probably plays a key role in the pathogenesis of these abnormalities.

Slide 23

These studies complemented the existing phenotypic characterization of endothelial cell dysfunction.

Slide 24

They demonstrated the development of mitochondriopathy, as an early sign of endothelial dysfunction, and added endothelial-mesenchymal transformation as a preamble to microvascular rarefaction.

Slide 25

I would like to leave you with is the idea of complexity of the syndrome of endothelial cell dysfunction (see the proposed classification). This complexity precludes the use of a single test or a single set of biomarkers to characterize completely this syndrome - it requires different biomarkers for different stages of endothelial cell dysfunction and therefore, I would like to offer as a first approximation this classification of endothelial cell dysfunction, which takes into consideration whether endothelial dysfunction is local or systemic, preclinical or clinically manifest, whether it’s reversible or irreversible, what is the leading pathogenic factor, etc.

Slide 26

In summary, molecular signatures of endothelial cell dysfunction are being developed. Endothelial cell dysfunction contributes to a progression of renal disease. Early pre-clinical diagnosis of endothelial dysfunction should have significant value.

Slide 27

These are the post-doctoral fellows and my collaborators – Chander, Nasjletti, Wolin, Gross, Zoccali and Bachmann. Thank you.

Slide 28

Chairman: Thank you very much Michael. This presentation is open for discussion so don’t hesitate to ask and to manifest yourself. I would like to point out two aspects that seem to me very important and that Michael just underlined, and which are too frequently ignored. He mentioned that one of the problems we are facing in endothelial dysfunction is rarefaction of the microvessels and the decreased recruitment - this is very important. Finally, the complexity of endothelial dysfunction.

Question:   Prof. Goligorsky, I enjoyed the nice pictures about the rarefaction that you could probably see in the skin, I assume with laser Doppler. What I was wondering is if this rarefaction has something to do with endothelial-mesenchymal transformation - I don’t know whether you looked at that? If this is somehow reversible after transplantation?

Chairman: You mean endothelial-mesenchymal transformation?

Questioner: Yes.

MSG: Yes, since it’s induced by inhibition of NOS due to accumulation of ADMA, one would expect that, with the development of new tools to suppress the synthesis of ADMA, it would be possible to suppress the downstream events such as activation of transforming growth factor-beta and subsequent to it endothelial-mesenchymal transformation.

Question: I have another question Michael. You have shown very clearly the difference between the classical test of post-ischemic vasodilation and thermal vasodilation, and thermal vasodilation of course takes a longer time. We know that according to the time of the stimulus applied with the long-term stimulus there is another player that comes into play, calcium activated potassium slow channel, the so-called endothelial-derived hyperpolarising factor. Do you have any idea about how this one works, if there is an imbalance between the classical NO system and this one? Yes but you see it takes a longer time, in seconds.

Prof. Goligorsky: There is data that demonstrated about a 40% decrease in thermal hyperaemic response after blockade of endothelium derived NO. Since there are no good tools to block EDHF I’m not aware of any data that would assess its role.

Chairman: It was tested in essential hypertension with apamine and tetra-ethyl ammonium, which blocks EDHF, and it has been shown that it plays, at least in essential hypertension, a role which is at least as important as NO and – but this is another question in ESRD. Other questions? If not, Michael, thank you very much and thank all speakers and the attendees for your patience and staying until the last minute.