Richard H. Turnage, MD: Mechanisms of Neutrophil-Mediated Pulmonary Edema

And the area of my research interest has been microvascular physiology. We have explored this on a variety of different levels, starting with animal studies and then using a number of other techniques to try to understand how fluid and solutes move from the intravascular compartment to the interstitium of tissues.

[ Slide 01 ]   Acknowledgements

I would like to at the outset bring credit to the fellows that have worked in my lab, Keith Wright, Carl Schulman, Jason Schwartz and also my partner Fiemu Nwariaku who have done a lot of this work with me.

[ Slide 02 ]   "Mediators" of Acute Lung Injury

Over the past several decades numerous mediators have been incriminated in acute lung injury or ARDS. Most of the studies have concentrated on identifying the mediator, the cause of ARDS or the acute lung injury. The most common physiologic manifestation of the lung injury is a change in microvascular permeability. But what is very apparent from these studies is that how fluid and particularly macromolecules and protein move from the blood vessel lumen into the interstitium of the tissue is absolutely unknown.

[ Slide 03 ]   Pathophysiology of Edema Formation

The Starling equation relates the movement of fluid and solutes across the vasculature to hydrostatic pressure gradient, the oncotic gradient, and the permeability characteristics of the membrane, shown as the capillary filtration coefficient, Kf, and the reflection coefficient (which describes the permeability to solutes or protein). In vitro studies describe a paracellular, that is the movement of fluid and macromolecules between cells, and a transcellular route, for fluid moving across the microvasculature. These studies have been in isolated cell cultures and there is very little in vivo models and no information in clinically relevant models of inflammation. Most of the studies have been done with cytoskeletal toxins or superoxide radicals.

But when you look at these ... in vitro studies and you expose endothelial cells to a variety of these toxins what you see is a couple of things. You see these cells shrink and you see a change in the actin cytoskeleton with a reorganization and pretty profound changes in the actin cytoskeleton. This is one hypothesis that's thought to account for this cell shrinkage. Another mechanism that is actually fairly recent and very interesting has to do with the actual contraction of the cell with actin microfilaments moving along myosin globules within the cytosol and so an active contraction much like a smooth muscle cell would contract can also result in these changes in cell shape.

And then lastly there's been a lot of interest recently in the intracellular junction proteins. For the endothelial cells the junctions are termed adherens junctions. They're much like the tight junctions of epithelial cells and a particularly important protein in adherens junctions are VE-cadherins. We'll talk a little bit more about them.

[ Slide 04 ]   Burn Injury Model

For the last year or so we've been using a burn injury model in Sprague-Dawley rats and mice. And the beauty of these models is there is a well defined event that you can time to the second when it occurs. We use a 40 percent total body surface burn. It results in a very reproducible systemic inflammatory state and a modest pulmonary edema. The controls for all of these experiments are manipulated but uninjured animals.

[ Slide 05 ]   Hypothesis

Previous studies over the last year have suggested that neutrophils and TNF are involved in this injury and I would like to describe these to you. The first is that this burn injury induces neutrophil activation and that is evidenced by CD11b/CD18 expression and the generation of superoxide radical by these neutrophils. The lungs generate TNF and the endothelial cells express ICAM on their surface. Lastly the increased pulmonary permeability is dependent upon both the presence of TNF and ICAM. And so these data suggested that neutrophils and TNF are involved in this injury.

The study that I would like to go over with you today and where we're headed in our laboratory is looking at how these mediators affect endothelial cell actin cytoskeleton and how that might ultimately relate to the movement of fluid and macromolecules across the microvasculature.

[ Slide 06 ]   TNF-a & NP-Mediated Lung Injury - ex vivo

In this first set of studies we used an isolated perfused lung model in which we were hoping to reproduce the in vivo injury. And what we did was take an isolated normal rat lung and perfuse it with the Krebs buffer alone, TNF, neutrophils harvested from an uninjured animal or from an animal sustaining a burn and then pretreating the lungs with TNF for one hour and then exposing them either to sham neutrophils or burn neutrophils. And we're able to measure Kf, which is the permeability characteristics of this lung using a gravimetric technique in which a gain in weight over a certain pressure gradient reflects an increased permeability. And what you see is that TNF alone had no effect on lung permeability. We used doses or concentrations anywhere from the one nanomolar to a micromolar dose. If you add neutrophils from a noninjured animal or even neutrophils from a burned animal 10⁶ or 10⁷ neutrophils, again it had absolutely no effect on pulmonary microvascular permeability. If you pretreat the lungs with 10 ng/ml of TNF, and then add this number of sham or normal neutrophils again there's no change in Kf. However when you add the neutrophils from a burned animal it results in about a doubling of lung permeability. We performed the exact same experiment using PMA to activate these neutrophils and had exactly the same results.

This suggested to us that the mechanism that TNF works is one of two mechanisms. One, it's well known that it upregulates ICAM expression and we've demonstrated that if you treat cells with TNF for even one hour you get up regulation of ICAM expression. And then also Carl Nathan showed in the early '90s that if you add a neutrophil that's bound to endothelial cells in the presence of TNF it releases huge amounts of hydrogen peroxide, and so we hypothesized that perhaps that plays an important role.

[ Slide 07 ]   TNF-a & NP-mediated Endothelial Injury - in vitro

We then took this one step further from the animal and using a cultured endothelial cell monolayer we are in the process of repeating this study, but I would like to show you some preliminary data. This is a endothelial cell monolayer of cells grown from the pulmonary artery of rats and the movement of protein across this monolayer is expressed by measuring the amount of protein that moves from one side of the monolayer to another side in a Ussing chamber type system. You can see with media alone, TNF alone, TNF added to the media (10 and 100 ng of TNF) results in no significant difference from media alone. However when you add TNF plus burn neutrophils (again we're pretreating those cells for two hours, and then adding the burn neutrophils), we see a doubling of the movement of protein across the monolayer. We certainly have a lot of work left to do in this particular set of experiments but I think it's interesting and it certainly goes along with what we've seen both in vivo, and in our isolated perfused lung model.

[ Slide 08 ]   Endothelial Cell Cytoskeletal Changes

Recently we have begun to examine the effect of these mediators on the endothelial cells and look at what's happening in the endothelial cytoskeleton. And this is the results of some of the first set of experiments. Again, these cells are pulmonary artery endothelial cells that are stained with rhodamine phalloidin and then looked at using a fluorescence confocal microscope. The rhodamine phalloidin stains actin filaments, actin microfilaments not the globular form of actin, and you can see in the media alone group the actin filaments pretty much span the entire length of the cell. Here they are nice thin fibers. In the TNF group you can see some disorientation of the filaments and some shortening. It even appears to be a little thickening in the fibers. Again you see it here, here and also some rounding of the cells.

In the cells that you treat with TNF and then add to them sham neutrophils, it looks similar to that of the TNF effect; that is there are some cells in which there is some shortening of these fibers but there's an awful lot of cells that look pretty good. When you treat this cell monolayer with burn neutrophils plus TNF, and again that's TNF for two hours and then add burn neutrophils actually for two hours, you see much more dramatic findings. And you can see in a lot of these cells dissolution of the actin microfilaments with this cloud of fluorescent material here. You can see it really pretty well here. Also you see a rounding of many of the cells here and here. We have utilized a variety of techniques to try to quantitate this better than just looking at these and what the software of our microscope allows us to do is actually look at cell surface area and we can see that the endothelial cell surface area of this particular group is about 40 percent of the media alone group. Also when you look at the intensity of fluorescence of individual cells, there is 50 percent reduction in this group compared to the other. This suggests that there is a reduction in the amount of actin microfilaments in this particular group.

[ Slide 09 ]   Adherens Junction Proteins & Relationship to Actin Cytoskeleton

The actin microfilaments are intimately associated with the intercellular adhesions of the adherens junctions and in fact they're connected through a variety of proteins, one of which is beta and alpha catenin and ultimately to VE-cadherin. As we began to study what's going on with the actin microfilaments, we were interested in knowing what's going on with the VE-cadherin and the adherence junction proteins as well.

[ Slide 10 ]   TNF & VE-Cadherin

Again I have some preliminary studies to show you. In these cells you can see, this is again an immunofluorescent study in which the VE-cadherin is stained with a fluorescently tagged monoclonal antibody and then looked at under confocal microscopy and you can see the fluorescent staining around the periphery of pretty much all of these cells. You treat the cells with TNF for 6 hours, you see really pretty much a loss of most of it, not all of it but most of it. When we did a Western blot to try to quantitate this better, you see here protein in the media alone group versus the amount of protein within the TNF treated cells. If you add to the media an inhibitor of TNF, which is a soluble TNF receptor, you can actually reverse this and again you see restoration of the fluorescent marker around the periphery of the cells.

[ Slide 11 ]   Conclusions

This suggests to us that burn activated neutrophils and TNF appear to enhance pulmonary microvasculature permeability and appear to be important in vivo as well as in an isolated perfused lung model. We also have a little preliminary data suggesting in our cell culture system. In vitro TNF alone and to a greater extent TNF and burn neutrophils induce morphologic changes in the endothelial cell cytoskeleton which we postulate may alter or enhance the movement of fluid and protein between the cells.

This brings to our mind a lot of questions. And it's interesting to ponder ... how does TNF actually have this effect on burn neutrophils, that is why is it that TNF plus burn neutrophils so much greater than the burn neutrophils alone, or the activated neutrophils alone enhance permeability. Also I think the relationship between changes in the actin cytoskeleton and the adherence junction integrity is going to be particularly important to look at. It's certainly related. In fact the expression of ICAM is also related via the actin cytoskeleton to the entire array of the cytoskeleton as well as the adherence junctions. And then what is the relationship between TNF and ICAM and then changes seen. And then also lastly the role of neutrophil derived oxygen free radicals in this whole process.

Again Chuck I'm delighted to have been asked to participate. Thank you.