CME Slides Forum - B. Diamond

A joint Congress by ERA-EDTA and ISN

B CELLS AND THE KIDNEY

Betty Diamond, Manhasset, USA

Chair: Leishi Li, Nanjing, China

Francesco Paolo Schena, Bari, Italy

diamond

Dr Betty Diamond
The Feinstein Institute for Medical Research
Center of Autoimmune and Musculoskeletal Diseases
Manhasset, NY, USA

So what I hope to do today is to talk, give a review of much of what you know about B cells but give a new prospective on it that I think will have bearing on SLE pathogenesis and on targeting B cells therapeutically in SLE. So you know that B cells have many functions in the immune system and I want to suggest to you that while they’re importance as antigen presenting cells has been highlighted in many autoimmune diseases, in lupus perhaps their most important function is in antibody production. Of course, they make the protective antimicrobial antibodies that we all rely on to protect us against infection but they also make protective autoantibodies. So I think the data are now very clear that in fact there are IgM antinuclear antibodies that are important in clearing apoptotic debris in a non-inflammatory manner. In fact, lupus disease begins with the switch from IgM to IgG.

Then they’re important in immune complexes in epitope spreading so antibodies play an important role in lupus.

You’ve heard that lupus affects many target organs and you obviously know that it affects the kidneys but as far as we’re aware, antibodies initiate all tissue injury in lupus. So after there starts being an inflammatory response in the kidney, there can be chemokines that draw in monocytes and T cells and whatever but tissue injury begins with antibodies in lupus.

So, antibodies begin in the serum, there’s epitope spreading, there’s switching from IgM to IgG, there are antibodies in tissues and in kidneys, we know that these are anti-DNA and animal models, they’re anti-SM. The antibodies cross-react with antigens that are always present in the kidneys like laminin, α-actinin is once the immune complexes are formed there is activation of Fc-receptors which seems to be a critical part and then there is an inflammatory cascade that importantly includes tumour necrosis factor and ends in tissue damage.

So if we want to think about B cells and what goes wrong in the B cell repertoire selection and B cell activation in lupus, we need to begin thinking about how a B cell repertoire is formed. You know that it’s formed by gene rearrangements that occur in B cell development that make a VDJ combination for a heavy chain and a VJ combination for a light chain and the heavy chain and the light chain randomly associate, so in fact one gets a repertoire that is full of auto-reactive cells.

As these cells are ready to leave the bone marrow and move to the spleen and from the spleen to other peripheral lymphoid organs, there is a moment, a period of selection where when the B cell encounters antigen which is most often autoantigen it can be tolerised.

There really are 3 mechanisms for tolerance. There’s deletion which is the elimination of the cell through an apoptotic pathway, there’s energy and there’s receptor editing which seems to be the first pathway chosen to tolerise a B cell.

Receptor editing is really just re-expression of the RAG complex, so that one forms another light chain that may give the antibody a different specificity, so that it’s no longer auto-reactive, no longer encounters antigen as it moves to the periphery and no longer is deleted. The choice of tolerance mechanism depends on the strength of B cell receptor signalling that obviously depends on the affinity of the B cell receptor for antigen and the local concentration of antigen but it also depends on a signalling cascade within the cell.

What we’re learning is that in fact, many of the susceptibility genes in lupus and you’ve just heard about --- but many of the others are actually genes involved in this signalling cascade.

So what we know is that an immense part of the repertoire is auto-reactive and as the B cell matures the auto-reactivity gets eliminated. There is a defect in lupus and the defect accounts not only for the increased auto-reactivity but probably also for the mild immunodeficiency that is present in lupus and in many other autoimmune diseases.

So what we think and what our data would tell us is that in lupus patients the antibodies arise in a T independent fashion in germinal centres. A T helper cell and antigen come together with the B cell and a cognate interaction and there is class switching and semantic mutation. The semantic mutation can also be responsible for generating auto-reactivity and there has to be some regulation of the B cells that acquire auto-reactivity in a germinal centre response.

We know that BAFF is a B cell survival factor that is very important from the immature B cell stage on in B cell survival. We also know that there are multiple tolerance checkpoints during B cell maturation and activation that prevent auto-reactive cells from moving to the next stage.

This is at the immature to transitional, transitional to naïve entering into follicles, entering germinal centres, becoming plasma cells, entering into a memory pool.

There is a checkpoint at each of those moments but we also know that in fact, one can abrogate those checkpoints, that there are B cell intrinsic and B cell extrinsic factors that can make those checkpoints malfunction. These include too much cytokine, too much costimulation, estrogen, activating DNA, BAFF and some that may work indirectly like low TNF levels.

So the model that I want to present is different from the model that you’ve always heard for lupus which is that lupus B cells are hyperactivatable.

What I want to suggest is that in fact they have a low signalling strength when they’re immature and all of the lupus susceptibility alleles suggest this and that this low signalling strength allows more auto-reactive cells into the immunocompetent repertoire. This is just showing that in fact, a low signal strength means that in lupus patients auto-reactive B cells are less likely to be tolerised and BLK, PTPN22 and Csk, the lupus susceptibility alleles all decrease the strength of B cell receptor signalling.

Then what happens which may not be genetic is that usually the FcRIIB on B cells downregulates B cell receptor signalling and in fact, leads to the death of plasma cells for reasons that aren’t understood but again, I’ll suggest are not genetic. In many, many lupus patients FcRIIB is downregulated, so it doesn’t function to keep the B cell in check once it’s an activatable B cell and it doesn’t function to eliminate the B cell once it’s a plasma cell.

We also have the problem that in lupus the antibodies that are made form immune complexes that contain DNA and RNA and these immune complexes are proinflammatory and therefore, further alter selection.

So what I would suggest to you is that a normal B cell has a high signal early on so that a strong B cell receptor signal and that that allows for the deletion of the auto-reactive cells and then a lower signal later which controls the activation of the B cell. In lupus it’s the exact opposite that early on there’s a low signal, auto-reactive cells survive and as it matures later on there’s a much stronger signal, so they’re hyperactivatable and the plasma cells live longer.

So B cells I’m trying to suggest are clearly important in lupus because antibodies are important, why has B cell deletion, ablation failed? Why has BAFF blockade failed?

I’m going to give you just a few data slides and then go back to what I think the principles are but in fact I think that B cell ablation failed because we didn’t think well about it and we didn’t use all the information that we knew when we went into the clinic with this therapy. So if you use cyclophosphamide or you use an anti-B cell antibody to ablate B cells, they’re ablated and then over the next month they return to normal.

But in fact, if you look at the serum of mice that have this ablation, you see that this is actually looking in the spleen at the DNA reactive B cells that there’s a higher number of DNA reactive B cells. If one tries to examine this by going to a transgenic mouse that has the heavy chain of an anti-DNA antibody, where that heavy chain is present in about 5-10% of the B cells and can associate with light chains that confer no DNA binding or light chains that confer low affinity or high affinity DNA binding and we know the light chains that do this, one can deplete the B cells there and look at the reconstituting repertoire.

What you see is that the reconstituting repertoire has many more high affinity anti-DNA B cells.

If you think about it, we know that every time there is B cell depletion there are high levels of BAFF and you see high levels of BAFF in these animals

and if you treat with an anti-BAFF agent what you see is that the DNA reactive antibodies are no longer able to mature to immunocompetence.

We also asked whether antigen was involved in this and you see that if you give DNAs, so that autoantigen is no longer available for positive selection of B cells, you also restore a normal repertoire.

If you look at serum titres, you can see that BAFF-R-Ig and that DNAs decreased anti-DNA antibodies to baseline levels and inactivated DNAs didn’t do that. What that means is that under conditions of low BAFF antigen functions towards negative selection but under conditions of high BAFF antigen functions for positive selection.

We in fact, had all of this data available without having done the experiment when we went to do B cell depletion studies in humans and I would suggest to you that they reconstituted with the more autoimmune repertoire in lupus patients.

So what are the targets in lupus? Well, there are many targets and many still are focused on B cells and I believe that the B cell is a very important target in lupus because we know that antibodies are the first autoantibodies, are the first abnormality you see in lupus patients years before you get clinical disease. We know that antibodies are the initiator of all tissue injury in lupus.

So it’s very important but in fact, we have to think hard about how to target B cells in lupus. We have to understand which B cells are producing the pathogenic antibodies, indeed whether the B cells in the kidney are of importance or not and I would suggest not. Are flares caused by naive or memory B cells? Because that tells you which population you need to target and the more focused your therapy, the less you’re going to have to deal with high levels of BAFF. To really understand which B cells are targeted by ablation therapies I think that in lupus we need to target plasma cells because I think that these activating immune complexes set up a proinflammatory milieu that is immunogenic where it should be tolerogenic and the immune complexes themselves and the BAFF that they create and the IFN that they also induce cause IgM to IgG switching independent of T cell help and I think that it’s the switch from M to G that makes many antibodies pathogenic move from being in fact protective to pathogenic.

I guess what I re ally think is that we need to think about how to target B cell receptor signalling, so that we can make small changes in the strength of that B cell receptor signal because if we could do that not only could we eliminate the auto-reactivity that takes years to move from benign auto-reactivity ANA positivity to tissue damage but we could also probably restore lupus patients to full immunocompetence. So that’s all I want to say that I think that we really need to think about a different model for B cells in lupus but continue to think of them as a very important focus for a therapy that could in fact, be a non-immunosuppressive therapy in lupus patients. Thank you.

Chairman: Thank you for that provocative stimulating presentation. Questions? Could you please state your name and institution?

Question: - -- from South Africa. Thank you for that wonderful presentation. You mentioned a part of your hypothesis using Fc fragment or Fc-receptor can you modulate Fc receptor to see whether they will change the response to B cell?

Dr. Diamond: So I think the question is can you modulate the Fc receptor on B cells is that what the question is? So one of the observations that has recently been made about the way IV gammaglobulin works is that it in fact, increases FcRIIB on B cells and makes it harder to activate them and makes it easier to kill them once they’re plasma cells. Because once the cell is a plasma cell there are only two things that I’m aware of that really stop its survival. One is enormous depletion of BAFF. So you really have to get BAFF levels all the way down and the other is engagement of the Fc-receptor on the B cell will kill the plasma cell. So regulating the longevity of plasma cells is very important and the B cell depleting therapies that we’ve used so far deplete B cells but don’t get rid of these immune complexes that are pro-inflammatory because the plasma cells aren’t depleted and continue to make these antibodies that activate not just the dendritic cells that are important in determining the survival of B cells but actually dendritic cells in skin, kidneys and other inflamed tissues.