Timothy Billiar, MD: iNOS: From Cloning to Therapeutic Applications

Instead of focusing on a specific disease process, I'd like to take you through the series of events in our lab that have occurred over the last decade which have led up to our current focus. I adhere to the mindset of my mentor, Dick Simmons, which is that science is a smorgasbord and you should really treat it that way. If an interesting observation comes up even if it's not in your particular area it might be worth pursuing. Certainly surgeons shouldn't be limited to any specific area. We touch on so many disease processes that you should never, in my opinion, restrict yourself to a particular area of science but instead let the science take you in whatever direction that is most productive and interesting to you.

[ Slide 01 ]   From Clinical Observations to Molecular Biology

Our own work started over a decade ago when I was a resident in Dick Simmons lab, looking at hepatocellular dysfunction and injury that occurred as part of sepsis and multiple organ failure. My predecessors in the lab had established an in-vitro model of macrophage-induced hepatocyte dysfunction. A great deal of effort was exerted on understanding the mechanisms of how macrophages, or Kupfer cells, could induce dysfunction in hepatocytes. We then demonstrated that this dysfunction was caused by the production of nitric oxide by hepatocytes in the late '80s. This was very early at the time which the nitric oxide pathway was first being elucidated. And at that time it wasn't clear that humans could even express the pathway of the inducible NO synthase and this is where I think our role as surgeon-scientists really paid off, because we had access to human tissue and we knew how to use it. We were able to isolate some human hepatocytes and show that human hepatocytes produced nitric oxide. These cells expressed iNOS and then a very talented fellow in the lab, Dave Geller, went on to clone the human iNOS cDNA gene and we were able to be the first lab to do this, and answer to a very important question at the time, a question that seems passe now, but at the time it was a major question in the field.

Well this spawned a great deal of work in the lab. We now were the first to have the cDNA gene and the question is what do you do next? Do you then get into the complex molecular biology that's necessary to characterize the expression and take advantage of the cDNA or do you go on carrying out more typical physiology? We moved heavily into molecular biology. And this spawned, I think, successful careers for some other investigators. Dave Geller then went on to study the molecular regulation of the iNOS gene and is now an NIH dependent funded investigator. Edith Tzeng went on to develop iNOS gene therapy and she's now an NIH funded investigator. It's these two topics that I'd like to touch on just briefly.

[ Slide 02 ]   Incucible NO Synthase (iNOS) Gene Expression

Well the observation that Dave had made was that the up regulation of the inducible NO synthase, or iNOS, was dependent on a number of external signals. These signals act synergistically to transcriptionally activate the iNOS gene. iNOS is then expressed and NO was produced when arginine is converted to NO. Dave became very interested in the signals involved in the transcriptional activation of the iNOS gene and how this synergy takes place. This had relevance to a number of important areas of medicine. Certainly iNOS is up regulated during sepsis; it's known to be up regulated in colon adenomas that progress to cancer. So understanding this regulation has a number of important implications in surgical diseases.

[ Slide 03 ]   Deletional Analysis of the Human iNOS Promoter

After you establish that the gene is regulated at the transcriptional level, one way to study the up regulation is to study the promoter, which is the regulatory region for transcription. This is done in standard promoter-reporter gene construct assays in which regions of the 5' upstream regulatory aspect of the gene is placed in front of a marker gene, in this case firefly luciferase, such that when these constructs are transfected into a target cell, and we're using some human epithelial cell lines, the gene is expressed, you can measure light by luciferase assay. The more light, the more transcriptional activation there is. What Dave did was clone progressively longer regions of the promoter in front of luciferase and then measure light activity. You can see as you go from 4.7 to 5.8 to 7.2, there is a greater transcriptional activation of the gene, such that the regulatory regions are clearly in a region between 5.8 and 7.2.

[ Slide 04 ]   NF-kappaB Mutations Decrease Cytokine-Induced Human iNOS Promoter Activity

He then narrowed this down even further by chopping the promoter up into ever smaller fragments. Here we have 5.8, 6.1, 5.2, 5.5 and you can see that there's [a loss of activity, or] a gain of activity as progressively longer units are used. In fact he identified a number of NF-kappaB transcriptional factor binding sites within these regions and began to work up these binding sites.

[ Slide 05 ]   Gel Shift Assay for NF-kappaB DNA-binding Activity in Human AKN-1 Liver Cells

He then demonstrated that NFkappa-B is activated by the treatment of these cells with the various cytokines. This is an electromobility shift assay in which radio-labeled oligonucleotides are mixed with nuclear extract and if there are proteins binding to these oligonucleotides, they are retarded on a gel and will appear as bands. TNF, IL-1 both activate NFkappa-B in this assay. A combination of cytokines is also effective. He was able to prove that in fact this was specific by using cold competition, in other words probes that are not radioactively labeled.

[ Slide 06 ]   Cytokines Induce a Nuclear Multi-Protein DNA Complex Consisting of NF-kappaB and STAT1

An important observation was to identify which specific factors are binding to these oligonucleotides. This is done by using a super shift analysis in which antibodies are added. When antibodies bind to the protein, they will modify the protein such that it will no longer bind to the oligonucleotides or they'll change the migration, and that's shown here, antibodies to p65, you can see resulted in a loss of binding activity but interesting an antibody to another transcriptional factor, STAT1, caused a shift in the complex suggesting that STAT was also interacting at this site.

[ Slide 07 ]   Summary of Cytokine Synergy

If I can summarize this work, what Dave has now been able to show is that the cytokines TNF and IL-1 activate NFkappa-B which bind to the human iNOS promoter, but some of the synergy takes place because interferon is also activating STAT1, STAT1 somehow cooperates with NFkappa-B to bind to the human iNOS promoter and activates the gene. Understanding the cytokine synergy in inflammatory responses is a very important area and this represents a significant advance.

[ Slide 08 ]   Vascular Properties of Nitric Oxide

Moving into the other area of our interest with the iNOS gene is gene therapy. One of the targets we have looked at is cardiovascular disease. We are looking specifically at vasoocclusive complications such as restenosis and intimal hyperplasia. We knew from the literature that NO played a very important role in the endothelium. It plays a vasoprotective role. Here NO produced by the endothelium by a constitutive NO synthase, this is a different NO synthase than the inducible I have been talking about. This NO regulates a vascular function and protects the endothelium through a number of mechanisms. It protects the endothelium by interfering with platelet adherence and aggregation, it interferes with leukocyte adhesion, and it also suppresses the proliferation of underlying smooth muscle cells, in addition to acting as a endogenous vasorelaxing agent.

[ Slide 09 ]   Nitric Oxide in Vascular Injury

Following vascular injury due to trauma, or surgery or angioplasty, the endothelium is lost. It's well known that platelets will adhere, leukocytes come in. These release growth factors. The smooth muscle cells themselves release growth factors. This results in a progressive migration and proliferation of the underlying smooth muscle, resulting in neointimal formation.

[ Slide 10 ]   iNOS Gene Transfer in Vascular Injury

Our initial hypothesis was that we could replace that deficient NO, by simply transferring the iNOS gene into the site of injury. That the small number of cells infected with the gene would produce enough of this diffusible agent to impact on all of the factors that can result in intimal hyperplasia, platelet deposition, leukocyte accumulation, and directly suppress the proliferation of underlying smooth muscle. I should point out that NO is diffusible locally but not systemically. It won't go systemically because it's inactivated by hemoglobin. So the gene therapy approach here had much appeal because a systemic administration of NO donors would result in hypotension.

[ Slide 11 ]   Viral Vectors

Of course, how do you get the gene into the site? You need a vector. And we certainly believed viral vectors are the way to go. They're simply much more efficient than nonviral vectors. This gives you an example of two viral vectors that we made. Viral vectors differ considerably in their characteristics. You can almost custom-make your vectors to meet your needs based on their duration of expression and other factors. Retroviral vectors have low efficiency, for example, they will transduce only proliferating cells but they do result in long-term expression. This vector simply is not very useful for in vivo gene transfer. Therefore we turned to an adenoviral vector. These vectors have high efficiency. They'll transduce non-proliferating cells and they have short-term expression so if your goal is a short-term transient overexpression, this vector is extremely effective.

[ Slide 12 ]   AdNos Gene Transfer

And so we generated an adenoviral vector.

[missing video]

[ Slide 13 ]   iNOS Gene Transfer Inhibits Neointimal Thickening in the Rat Carotid Artery Injury Model

The control-transduced vessels have a significant buildup of neointima. If we transduce the vessels with iNOS, no intimal hyperplasia whatsoever. And we can reverse that if we simply infuse an inhibitor of the iNOS enzyme, showing the effect was due to enzyme activity.

[ Slide 14 ]   iNOS Gene Transfer Inhibits Neointimal Thickening in the Pig Iliac Artery Injury Model

Pig models, porcine models, are thought to be a little more representative of human disease, and so we carried out the same experiments in pig iliac arteries that were injured. Shown here is the intimal hyperplasia that's observed three weeks after injury. If we've transduced with iNOS we get about a 60 percent reduction in the intimal hyperplasia.

[ Slide 15 ]   Efficacy of Vascular Gene Therapies

This gives you some idea of where iNOS sits in the hierarchy of gene therapy approaches and experimental models. A number of approaches have been taken, but cytotoxic approaches such as herpes thymidine kinase has been used, there's about a 60 percent reduction in the pig.

Cell cycle inhibitor approaches have also been used, with about a 30 or 40 percent efficiency. The eNOS enzyme, or this constitutive enzyme, have also been used, but you can see here that our results with iNOS are much better, and we are using a much lower titer of the adenovirus. In some cases several logs lower and having a much greater efficiency and we believe that's because the iNOS enzyme is simply a much more active enzyme than the eNOS in terms of Vmax and enzyme activity rates.

[ Slide 16 ]   Nonocclusive Gen Delivery Device

One of the roles that I think surgeons can play in gene therapy is through understanding the structural limitations and technical limitations there are getting genes into sites. I think surgeons function as excellent bioengineers in this regard. A cardiac surgeon in our group from Japan wanted to determine whether or not he could devise ways to place genes on the outside of small vessels without occluding the vessels. So he developed this prototype of a device which is just a Silastic piece of tubing with some Teflon on the end with holes which will allow him to infuse the vector on the outside of the vessel. The advantage here is that you wouldn't have to occlude the vessel for even a short period of time. You could simply put the gene on the outside and if it was as effective as the inside, that technical limitation would be overcome.

[ Slide 17 ]   Perivascular Gene Delivery

Clearly in some vessels that's important. Here he's placed this device on the outside of a rat carotid and he's infusing the gene around the vessel here.

[ Slide 18 ]   The Efficiency of Periadventitial Gene Transfer

At least for these small rat carotids, the inhibition of intimal hyperplasia was almost as effective as placing the gene on the lumen. Here's the control-transduced vessel, you can see the injury induced intimal hyperplasia and almost no intimal hyperplasia when the gene is simply placed on the outside.

[ Slide 19 ]   Would Overexpression of Inducible Nitric Oxide Systhase Inhibit Transplant Arteriosclerosis?

We can extend this to other disease processes. Transplant atherosclerosis looks very much like intimal hyperplasia. It's the number one cause of allograft loss for cardiac transplantation.

[ Slide 20 ]   Allograft Aortic Transplant Model

The question is if we transduce vessels that are allograft vessels, will we see the same inhibition of intimal hyperplasia. Larry Shears turned to an aortic allograft model in rats in which the thoracic aorta is transplanted into the abdominal position of a recipient rat. We transduced the vessels ex-vivo for 30 minutes and then at four weeks later looked at the degree of intimal hyperplasia.

[ Slide 21 ]   Allograft Aortic Transplant Intimal Thickening

Isografts have no intimal hyperplasia. This is intimal thickness. Allografts had the expected increase in intimal hyperplasia. Transducing with Lac-z had no effect. iNOS completely blocked this at four weeks. In our hands, cyclosporine in-vivo actually accelerates the intimal hyperplasia and the adenoviral vector also completely blocked this, raising the possibility that this may be useful for suppressing those few patients that develop allograft vasculopathy early.

[ Slide 22 ]   The Mechanism of NO Inhibition of Intimal Hyperplasia

Just a few seconds about mechanism, and that's where most of the work has gone, is trying to understand how NO inhibits intimal hyperplasia and Melina and Edith are now focused on the cell cycle. You've heard a little about p21, it's a cell cycle inhibitor. It blocks various steps of the cell cycle. We wanted to know whether NO was up regulating p21.

[ Slide 23 ]   iNOS Gene Transfer Upregulates p21 Expression In Vivo

And in fact it is. This is a injured vessel transduced with Lac-z. The green is just fluorescence from the elastin. An iNOS transduced vessel has a marked up regulation of p21. The reddish orange color is p21 up regulation. Although we haven't demonstrated an association between p21 and the inhibition of proliferation, a direct connection, there is this strong association, suggesting a mechanistic link.

[ Slide 24 ]   Mechanism for the Up Regulation of p21

And a number of experiments have now been done to characterize the mechanism for the up regulation of p21, and we know that it involves the map kinase system and that it is cyclic GMP ...

[ Slide 25 ]   From Clinical Observations to Molecular Biology

[missing video]

...independent observations intensely into molecular biology of a number of diseases relative to surgery.

Thank you.