MACROPHAGES IN SENSING AND MODULATING INJURY

Jeremy Hughes, Edinburgh, United Kingdom

Chair: Detlef Schlöndorff, Munich, Germany

Andrew Neil Turner, Edinburgh, United Kingdom

Prof J. Hughes Wellcome Trust Senior Research Fellow in Clinical Science MRC Centre for Inflammation Research, University of Edinburgh Edinburgh, United Kingdom

---

Slide 1

Thank you very much. I’d like to thank the organisers for giving me the opportunity to come and talk about our research and also for booking the superb Mediterranean weather, which is very welcome. Now what I want to do over the next 20-25 minutes or so is really run through a couple of aspects of macrophage biology and I find toll receptors and chemokines incredibly complicated, far too difficult, so I’ll try and stick just at looking at one cell which helps me think about things.

Slide 2

What I’m going to deal with is basically our investigation looking at the role of macrophages sensing injury, so sitting there in tissues actually waiting for perturbation and then acting accordingly. To do that we’ve actually used two models of experimental peritonitis, these are thioglycollate and also pleurisy model using carrageenan which comes from a seaweed. Then I’m going to move on to the kidney, which is probably more interesting to all of you and we’ve been looking at the role of macrophages in the progression of nephrotoxic nephritis in the mouse and also ureteric obstruction, which is a well-known model of disease of interstitium.

Slide 3

Now these slides have been made very straight forward by a wonderful mouse invented or device by Richard Lang in New York. What he’s done, he’s actually taken the human diphtheria toxin receptor, which binds diphtheria toxin much more potently than the murine receptor. He’s expressed it in a transgenic mouse under the control of the CD11b promoter. This really directs it to the monocyte macrophage lineage. When you administer tiny quantities of DT, nanogram quantities, you can actually ablate by inducing apoptosis in the monocytes in the blood and also macrophages in a variety of organs and you knock out 90% of monocytes in the blood and you can taper almost 98% of macrophages in some of these areas here.

Slide 4

So it’s a very useful tool. I think it’s allowed us to make some insight observations although as I said it needs more work. This shows the power of the technique. So here we have ---- from a peritoneal --- of a mouse we’ve given thioglycollate to. At day 4 you have a macrophage rich infiltrate and here is the control animal with lots of large mononuclear cells and after you’ve given them 20 ng/g body weight, you basically see nothing, you’ve killed everything there. This is actually taken 24 hours after you give the DT to the mouse. You could do it 6 hours later because they’ve all gone by 6 hours-12 hours. So it’s a very powerful technique for killing macrophages and monocytes.

Slide 5

Now I should press that all these studies by the fact that we’ve looked very carefully at neutrophyls. It doesn’t affect neutrophyls, it doesn’t kill them, the numbers in the blood are comparable, it doesn’t kill them in the peritoneum or in the pleural cavity and also doesn’t affect the migration to chemokines. So it actually really is acting very much on the macrophages and monocytes.

Slide 6

So we started off by asking a very simple question. In the peritoneum there are different kinds of cells. You’ve got macrophages, mast cells, mesothelial cells, and also lymphocytes and the question is now what is more important in terms of initiating or recognising perturbation and then initiating acute inflammation? The published data was actually very controversial, there were papers saying that the mast cells were important, papers saying that it was dependent upon the model and the macrophage could be important, it might be unimportant.

Slide 7

So we tried actually to try and define which cell is key and we used initially the thioglycollate model of peritonitis, a very well established model and this is just to show you in the normal peritoneum of the mouse, this is a normal mouse, you get a veering number of between 1 and 2.5 million macrophages in a normal mouse. When you give DT you take out 97% of them and actually they get taken out for quite some time. So certainly during our investigations there were no macrophages present in the peritoneum. All our control experiments are performed in the same strain, which is non-transgenic, and we give them as DT as well to control for any other confounding effects of DT in the system.

Slide 8

When you give thioglycollate to the mouse, in the wild-type or the non-transgenic, you actually see a very brisk infiltrate with neutrophyls 7-8 million, this is at the 8-hour time point, which then comes down over the next sort of day or two. When you ablate the resident macrophage, it’s tremendously blunted and also it actually still falls off over the next 24-72 hours.

Slide 9

So obviously, macrophages are playing a key role in this process and we went on to try and not just knock them out but also put them back. So this study is where you’ve taken, this is a non-transgenic control which has DT, thioglycollate, this is the 6 hour time point and there are 4 million neutrophyls there on average. This is our sort of positive control. It’s transgenic, it’s been ablated with DT, it gets given thioglycollate and you have a large reduction in the inflow of neutrophyls.

Slide 10

Here we’ve actually taken the same kind of mouse, it’s ablated but 4 hours before we give that thioglycollate we actually put cells back into the peritoneum from a syngeneic non-transgenic mouse, these are macrophage rich cells and we can restore the infiltrate of neutrophyls.

Slide 11

If you do the same experiment but simply take the macrophages out of the population by adhesion or by using immunomolenic depletion, using beads again you find you do not respond and you do not draw in as many neutrophyls.

Slide 12

We did many controls, here is one of them. This just shows that if you put those cells into a normal mouse, which has been ablated, the actual transfer cells per se is not inherently inflammatory. So again it backs up by knocking them out, putting them back. It appears that a macrophage in the peritoneum is very important to orchestrate this initial influx of neutrophyls.

Slide 13

Now, I’m not going to show lots of chemokines but certainly, if you look at some of them and this is an IL-8 homologue in a MIP-2, you certainly blunt it dramatically when you take out the macrophage implying that in this system the resident macrophage is a key source to this particular chemokine. If you look at KC, it’s a different picture you get a half way house is knocked down by 3 % and we didn’t look at MPT-1 because again we can’t study monocyte recruitment when we’ve actually taken away all the monocytes in the blood. So certainly chemokines are significantly affected when you take out macrophages and I should say as well that when you take out TNF and IL-6 almost completely, 90% of reduction in those levels. So we’re trying to compare and contrast between the peritoneum and another cavity, so we looked at the pleural cavity asking the same kind of very simple question and in this system we give carrageenan into the pleural space. There is slightly a difference, certainly at the early time points where you have 5 million neutrophyls recruited, you see almost nothing in the pleural cavity. But unlike thioglycollate where at 72 hours there’s still nothing, it’s actually slowly building up, so a slightly different kinetics to the curve in this particular model. Again we see some differences with the chemokines. If you look at MIP-2, we found that was completely ablated in the peritoneum. Here we’ve only blunted it and shifted it. So we have less at the early time points but again by 6 hours when we don’t have any macrophages in the pleural cavity, we’re actually seeing significant levels of MIP-2. KC we are seeing some reduction but really I don’t think we’ve changed that very much. What that’s telling us is that in this scenario I think that local cells are obviously very important and I would imagine that these chemokines are being produced by other cells such as maybe mesothelial cells and other cells in the pleural cavity. So there are some subtle differences in different biological systems.

Slide 14

Again carrageenan comes from seaweed, so the reviewer said well, who gets seaweed in their pleural cavity? Not many people. So we actually did some preliminary experiments with S. aureus because obviously MRA S. aureus is important in thoracic surgery. Again we see a very similar system whereby if you take out macrophages here the virus controls and you put in fluorescently labelled, kill S. aureus, you actually found that you markedly abrogate that influx of neutrophyls, which is obviously what you need, if you are dealing with infection possibly.

Slide 15

Then when you actually look at the cytosines from the pleural cavity, what you see in a normal mouse is that the resident macrophage in these large cells here are very good at hoovering up large numbers of staphylococchi and actually the neutrophyls that come in there are actually very few left over for them to eat. In contrast, when you take out the macrophages, what you see is there are almost no macrophages there and it’s now down to the neutrophyls to ingest and degrade the Staphylococchi as best they can.

Slide 16

So I think that kind of work really shows that in serosal cavities peritoneum pleura, it’s really the macrophages that is one of the key sensors of the micro environment and when they sense perturbation either by inflammatory mediators or pathogens, they actually kick start things by generating chemokines. They’re not the whole story of chemokines and cytokines and by doing that they appear to be a key regulator of neutrophyl influx in these situations.

Slide 17

Now moving on to the kidney. I’m sure you’re all aware that macrophages are important in kidney disease. We see them often in biopsies, certainly in our murine model of obstruction we get lots of F4/80 deposited macrophages. Here’s a rat model of NTN. Again ED1 positive macrophages.

Slide 18

This happens to be a biopsy from a patient with acute rejection in the transplant and again lots of CD68 positive macrophages in that biopsy. I could have showed you other conditions as well from human disease but they are important and they are there.

Slide 19

So we asked a very simple question, if you take a murine model of nephrotoxic nephritis and you let it get established, what happens if you take out macrophages for a short time period? This work has really been led by Jeremy Duffield in the lab so we got some serum from ---tipping, induced nephrotoxic nephritis, waited until day 15 and then some mice either had PBS or DT for 5 days only to give a short period when you actually ablate the macrophages and then did some standard readings. As I showed you before, if you quantify either by computer image analysis or by counting, you certainly reduce the number of macrophages in the interstitium and also in the glomerular compartment compared to various controls.

Slide 20

It’s always gratifying when you look in the microscope and you can see a difference and certainly if you look at the glomerular compartment of the control animals or if you look at the interstitium, there’s a significant difference between the ablated animals and animals that have been left to develop disease normally.

Slide 21

When you quantify these parameters, you preserve function obviously very important. So serum creatinine at day 20 is down following the treatment with DT compared to the controls.

Slide 22

Proteinuria was also down and also you have less injury in the tubular cells, so less atrophy and also you have less tubular cell apoptosis compared to control.

Slide 23

Now there’s a whole host of other things we measured. Essentially you have less crescents, you have less tubular injury, you have less fibrosis and that’s probably because macrophages in this scenario are actually supporting a population of active proliferating myofibroblasts both around the glomeruli but also in the interstitium and as a result of that you actually knock down the level of scarring whether you measure collagen III, collagen I or whatever. Also I haven’t mentioned here, you actually also knock down the number of CD 4 infiltrating T cells. So again macrophages maybe involved in bringing in T-cells to the inflamed kidney.

Slide 24

Now we’ve had interesting cell interactions in vitro because we know that macrophages are certainly capable of killing cells and they’ve killed tumour cells, mesangial cells and they can kill them directly or indirectly. I can’t really mention that today for lack of time but I think that’s a very important mechanism because obviously if macrophages are profibrosis, that will actually have an effect on cell survival, they can actually regulate virus cytokines and levels of VEGF in tubular cells that may have an effect on the endothelium etc. but we’ve been focusing on this sort of direct cell interaction.

Slide 25

There’s a whole host of things in the matrix that could be toxic. iNOS derived nitric oxide, TNF-alpha, Fas L, reactive oxygen species, these can combine to form peroxynitrite, MIA and obviously others but they certainly can make a whole variety of potentially toxic mediators.

Slide 26

We’ve used a very robust and very simple assay where you simply take your cells of interest, your macrophages and we used primary murine bone marrow derived macrophages and you take your other cell, it could be a tubular cell or whatever, you can label them with different pheochromes, you pop them into a dish, let them settle down for a while and you can activate this system with various cytokines, we used IFN-gamma and LPS and they sit for a while in an incubator and you come back 24 hours later or at varying time points and you simply fix the whole lot in situ. So you spot weld all of the cells to the bottom of the dish even the cells that have undergone programmed cell death apoptosis. You stain it up with Hoechst and you can actually quantify the number of cells. You can quantify cell number, mitosis, apoptosis and also phagocytosis.

Slide 27

This just shows you some tubular cells in the dish. There are macrophages here that you can’t see and this is the typical appearance of apoptotic cells, they round up, they come off the plate slightly and they have condensed chromatin and you can count these very straight forwardly.

Slide 28

Now there’s a lot of data I’m not going to show you but essentially we found in this kind of study, we found that nitric oxide was a key mediator. So if you block it using various inhibitors like L-NIL or L-NAME or if you take macrophages from either an iNOS knockout or wild type animal you find that what happens is that when you have the co-culture sitting with no cytokines, the cells are pretty happy, you get no difference compared to cells alone. When you activate the system, you find that you get a significant level of death with the wild type macrophages but the knockout macrophages are unable to really give you a good killing signal. We’ve also shown if you block NO, you don’t affect TNF release with these mediators, you don’t effect Fas L expressions, so this appears to be purely because of the release of NO in this kind of co-culture assay.

Slide 29

Now in vitro it’s all very interesting, does it have any relevance in vivo?

Slide 30

So we have used the model of UUO, it’s a very simple model you simply tie off the urethra, clip it, cut it and you’ll get disease and during the disease process you get proliferation, you get cell death, lots of macrophages and also lots of scarring. So you get a variety of readouts to study.

Slide 31

If you look at the infiltrate that comes into the kidney in the model, here’s a normal mouse with just a couple of resident F4/80 positive macrophages at day 7 of the UUO, you actually get quite a number of macrophages infiltrating. If you do some staining either with iNOS or F480, you can co-localise them both to macrophages. So here you have a system where you do have a population of infiltrating iNOS positive cells in the kidney – disease.

Slide 32

So we initially did some experiments with iNOS knockout animals and it was complicated because they actually get differing levels of macrophage infiltration possibly because of a chemokine mechanism. So we had a much more simple approach, we simply allowed the mice to develop disease over 5 days and then for two days we either blocked iNOS with the inhibitor L-NIL given in the drinking water or we gave them the control D-NIL at the same dose and had a look at the kidney at day 7.

Slide 33

Apart from a variety of other things I could show you but essentially you do see a reduction in the level of tubular cell death and as you’ll see it’s not complete and that just shows that there are many other factors that actually cause tubular cell death in this kind of system that are macrophage independent such as stress hypoxia etc. but we were able to find a significant difference between the level of death in the controls compared to the iNOS inhibited animals. 

Slide 34

So I’m going to dare in the lab, you see the transplant surgeon, she’s very interested in macrophages in the field of transplantation and she’s taken the co-culture assay and performed similar experiments by using endothelial cells. We have a cell line from Justin Mason in London and these cells behave at 37° very similar to primary endothelial cells. She found very similar data to Tina Kipari in the lab whereby if you activate the co-culture you see significant levels of death. You can block it almost completely by blocking iNOS derived NO and D-NIL has no effect and again the controls, you don’t see large levels of death with this particular cell.

Slide 35

So she went back to that previous study and actually began to stain, it doesn’t actually project very well but she stained for CD31 and you can see these small capillaries that are CD31 positive in the L-NIL animal and if you quantify this and we used the method of Biohashi who showed previously in the rat UUO model he showed that if you actually obstruct a rat kidney, you actually lose a significant number of peritubular capillaries and we think that’s important because of the hypoxia progression ideas in renal disease.

Slide 36

So Hanne used the same methods to quantify it and if you actually look at the number of peritubular capillaries per hundred tubules during disease over 7 days, you actually lose a significant number of CD31 positive peritubular capillaries and again blockade of iNOS is actually only for 2 days. So it implies that this might be an acute loss over the past 48 hours. If you block iNOS with L-NIL, you actually see normal values and D-NIL is kind of slightly in between but still significantly different from the L-NIL values. So it appears that L-NIL and iNOS is important in a murine system not only in vitro but also playing a role in peritubular rarefaction in the in vivo model of UUO.

Slide 37

So she’s now moved on to look at some human material and this is a representative biopsy from some tissue from patients who have CAN compared to normal control tissue which is obtained at multi-nephrectomy for carcinomas. What she’s shown is a significant number of CD68 positive macrophages that infiltrate the biopsies that exhibit CAN and in control you see very few. If you do again double labelling for CD68 and also iNOS, you actually find there are macrophages in human kidney, which actually co-localise both and again at low power you can see there are number of yellow CD68 positive, iNOS positive macrophages in a human kidney specimen.

Slide 38

Interestingly, if you quantify the number of peritubular capillaries in this kind of study, we know that patients with CAN have significantly less micro vessels, peritubular capillaries both in the medullar and also in the cortex and again there are some relationships between macrophage infiltration, loss of the peritubular capillaries etc.

Slide 39

So I think in summary I think we now know that macrophages can certainly kill cells and that’s something they can do very well. They’re well equipped to do that and in our hands they can kill tubular cells and endothelial cells. Even though they can make a variety of mediators such as TNF and Fas L, NO appears to be important because if you take away NO, you have a major effect on disease outcome. This NO is involved in tubular cell death and rarefaction of the UUO model and I think an important point is the fact that macrophages are important not just in disease that we tend to think of as immunological, so T cell driven glomerular nephritis, for example. We know the UUO model, it’s the same in a scarred mouse so it doesn’t appear that the T cells are playing a key role there. So it may well be a common effector either of immunological inflammation or non-immunological inflammation, which you may often find in scarring kidneys. I think this is food for thought because it doesn’t mean that it’s a target for future therapy. This is a slightly casually marked, it’s the kind of thing you put at the end of your grant to hope you get some funding from somewhere but I’m going to bring it together in a little scheme which is very simplistic but essentially macrophages are very complicated and knocking them out by ablating them is a bit of a blunder bust technique to try and understand what they’re doing.

Slide 40

Now I’ve touched upon their role in terms of being a sentinel cell and making chemokines and cytokines. I’ve touched upon their role whereby they actually make these mediators that can kill host cells but they do a lot of other things as well. So, for example, they actually can take up cells and that has a major impact upon their phenotype. It re-programs them to what we think of as a reparative phenotype whereby they make TGF-beta, which is also immunosuppressive as well as profibrotic. They can make TNF and also IL-1 ra. I think the future challenge is to really understand more about the macrophages in human disease not just in murine disease but to take the co-culture work, to take looking at their phenotype in human disease to another level. Then begin to understand how can we actually modulate macrophages? It maybe that we might be able to knock down macrophage infiltration in the key disease but we now know that macrophages are probably important in diabetes and we certainly can’t take our diabetic patients and knock down their macrophages in the long-term. So what we have to do is be cleverer and understand the nature of this beast. I’m going to just mention there’s actually now data, there’s one very good paper in JASN I showed in the macrophages are very important for remodelling a scarred tissue. So they can play an important role in actually getting rid of matrix. That’s in the liver, there’s no real data in the kidney as of yet but I think one of the very exciting prospects is that we may be able to take a macrophage, understand the different phenotypes in disease and then try and modulate macrophages to be beneficial and reparative rather than actually toxic and detrimental. But I think we’re not there yet and there’s probably another decade of work despite Neil’s optimism.

Slide 41

Just to mention those people who did the work. Tina Kipari, Jean Francois Cailhier, Anya Adair and Spike Clay did really all the work that I’ve mentioned. Jeremy Duffield who is now in Boston did the NTN study and obviously thanks to all of those who provided the money. Thanks very much.

Slide 42

Chairman: Thank you very much Doctor Hughes. I think, would you like maybe to comment on what’s the difference between a different type of phenotypic macrophage and a dendritic cell?

Dr Hughes: That’s an area where things begin to merge and we often call them sort of dendrophages because there’s definitely a spectrum. What I will say, in our studies actually we can’t find CD11c positive cells in the peritoneum, despite having antibodies that work on myeloid differentiated DCs and it maybe a question of number but I think you’re quite right. I often regard them very much as a spectrum and they’re probably all important.

Question: My name’s John just a question, it’s not come up today but for the panel and the speaker. There’s been a lot of work done on chemokine antagonists to quell inflammation but there’s nothing in the clinic as such. What does the panel think of that? Is it going to work when you have a redundant system and you have a specific chemokine receptor?

Dr Hughes: I think my view is the fact that it’s difficult because there may well be different chemokines that bring in different kinds of macrophages. For example, we actually don’t know at all if when you can take away the injury from an organ, for example, the liver or the kidney, we don’t know whether the reparative macrophages are derived from the population that are there already which can be re-programmed or whether you have recruitment of new cells and we now know in mice that actually there are some differing subsets of monocytes in the blood that require different chemokines to bring them in. So there could a difficulty there but I think it is difficult and I know there’s work from Munich, your lab Detlef whereby if you use, I think it’s a CCR1 blocker, you actually make things worse even though you have  less macrophages, we actually have higher levels of -- expression and you get more tissue injury. So I think it’s not just a numbers game, it’s actually the nature of the cell which is there in the tissue game and I think you’ve got to get both right. Hopefully you should have not the bad guys, more the good guys and if you get that wrong, it maybe detrimental.

Chairman: So Detlef do you have thoughts?

Prof Schlöndorff: I fully agree with that and I think there’s even a further level of complexity and that’s a different vascular bed in the kidney. For example, CCR1 antagonists have absolutely no influence on glomerular localisation of inflammatory cells. They only work on interstitial inflammatory cells and so far, there’s really no good antagonist that will prevent glomerular macrophage infiltration without changing the phenotype of that glomerular macrophage. What we’ve done is was Met-RANTES what I think you referred to that actually worsened the disease because it changed the phenotype of the glomerular macrophage. A lot of work to be done like Jeremy said.

Chairman: It’s very appropriate and European Jeremy that you end with reform rather than abolition. Thank you very much.