Education: B cells are not a single subpoulation

Lehmann-Horn K, Kinzel S, Weber MS. Deciphering the Role of B Cells in Multiple Sclerosis-Towards Specific Targeting of Pathogenic Function.Int J Mol Sci. 2017 Sep 23;18(10). pii: E2048. doi: 10.3390/ijms18102048.

B cells, plasma cells and antibodies may play a key role in the pathogenesis of multiple sclerosis (MS). This notion is supported by various immunological changes observed in MS patients, such as activation and pro-inflammatory differentiation of peripheral blood B cells, the persistence of clonally expanded plasma cells producing immunoglobulins in the cerebrospinal fluid, as well as the composition of inflammatory central nervous system lesions frequently containing co-localizing antibody depositions and activated complement. In recent years, the perception of a respective pathophysiological B cell involvement was vividly promoted by the empirical success of anti-CD20-mediated B cell depletion in clinical trials; based on these findings, the first monoclonal anti-CD20 antibody-ocrelizumab-is currently in the process of being approved for treatment of MS. In this review, we summarize the current knowledge on the role of B cells, plasma cells and antibodies in MS and elucidate how approved and future treatments, first and foremost anti-CD20 antibodies, therapeutically modify these B cell components. We will furthermore describe regulatory functions of B cells in MS and discuss how the evolving knowledge of these therapeutically desirable B cell properties can be harnessed to improve future safety and efficacy of B cell-directed therapy in MS.
I am doing a presentation and am trying to see where the road blocks are to the understanding of MS. 

Reading through this review on B cells, which you can read if you are interested, it gives me the feeling that I am banging my head against a brick wall. 
I think there is a failure to appreciate that you cannot simply view B cells as a single population. Because if you do you cannot hope to understand what is going on.

For years we have been viewing T cells as a mixed group of cells, T helper cells (CD4), cytotoxic (CD8) and regulator cells and more, but the biggest road block to understanding MS is viewing B cells as a single cell population, which many studies do.

As a card carrying T cell immunologists, I must admit I do not understand the complexities of the B cell response, but I am trying to learn. So maybe you can learn too.

There are many, many subtypes of B cells and here are a few

B cells are born in the bone marrow from stem cells and then go through a series changes in their coats and expressing some markers. The graph below is simplified as is much more complex 

It is clear that pre-B cells form immature B cells in the Bone Marrow, these change into mature B cells, which shuttle between the bone marrow and lymphoid tissue, such as the spleen and lymph nodes. 

In some studies they are called follicular cells. Following stimulation, they proliferate in structures called germinal centres.

Germinal centre cells differentiate into either memory B cells or plasmablasts that can turn into antibody forming..plasma cells,….

When B cells are depleted they repopulate in a stereotyped way. 

The immature cells enter the blood to fill the space and then there is maturation This masks the very slow repopulation of memory cells, from the lymphoid organs, which is where we have said the action is. All drugs that work in MS deplete memory B cells, except natalizumab which get trapped in the blood so they can’t get into the brain.

Memory B Cells are Major Targets for Effective Immunotherapy in Relapsing Multiple Sclerosis.

Clearly not many people buy it…until they do. They may be clutching at straws

I will explain how this happens,  but I hope you can see that alemtuzumab wipes out memory B cells, but if you look at the CD19 population it is relatively meaningless in terms of understanding what is going on. So MS B cells are not a single population

About the author



    • Yes anti CD20 is in trial for NMO, but not sure where the idea that it doesn't work in MS comes from. The trial was positive MEDI-551

    • Thanks there you would e too expensive to develop for MS…however you now have chemical CD19…its called cladribine

  • Gibberish….Indeed.
    Maybe T cell immunologists are reading:-).

    However, I see the the blog administrators are tinkering again.

    Originally you could tick something like "bad", "OK" and "good", but it was clear that the trolls were simply ticking "bad" every day.

    So to foil this behaviour we got rid of "bad", and "OK" to only give positive choices were simply "good", "very good" or "excellent".

    I see this issue has been forgotten, obviously it going to be used to say how bad the posts are:-(…Will it change how we write?

  • The issue I have with these posts is that they are irrelevant to me – I'm not a researcher or scientist, just a patient. It's a bit like when I take my laptop in for repair, I don't need to know the ins and outs of all the hardware and software – just get it working properly again. I am interested in trial results based on this sort of work, but lots of colourful blobs marked up with different names don't help my condition at all. Please keep the gibberish option in as some of your posts are gibberish.

    • It is not irrelevant and here is why.

      This post perhaps gets to the core of why MS drugs work and if you accept that the central purpose of each MS treatment is to remove the memory B cells, then when a treatment comes along, you can ask what does it do in terms of memory cell depletion.

      If treatment doesn't kill them very well, then it is likely that the treatment is not going to work very well.

      Therefore it helps with making the right choice.

      You don't believe it… can ask what evidence supports an alternative view.

      Maybe you like the "neuro knows best" approach….

      Maybe they have read the post and are enlighted, but if they haven't and are lazy, they will stick you on a DMT that requires no monitoring and you will be OK, whilst you burn up your neurological research, by which time you options have become more severely limited..

      So you either take the rubbishy drug and hope that you are one of the lucky few where the rubbishy drug works (you could monitor to see if it is working) or you select one where the chances of memory B cell depletion is better and so it has a better chance of it working.

      If the action is at the level of the B cell, why on earth would you want to take an anti-T cell drug….you wouldn't do that…Maybe you wouldn't but the literature is full of people who have been involved in T cell trials that haven't worked and had limited chances of working.

      Another example Ocrelizumab is given every 6 months because the CD19 population is depleted for 6 months and then it returns. It returns because the space in the blood is filled with immatutre and mature B CD19+ cells but there is not a memory B cell in sight maybe for 18 months – 5 years. So you are paying $30,000 every 6 months to kill a cell that is not even there so maybe you would save £240,000 from your purse or for the NHS. Not bothered about that…but if you keep your B cells permanently depleted you will probably suffer the consequence of infections from having no B cells. Maybe if you monitored your B cells you would know it it is worth having another dose

    • It is jibberish be you have not understood the post…maybe we have pitched the level to high….maybe you have not done any home work to get up to speed and we cannot do back to basics over and over again.

      Does this subject need a basic explanation, e.g the car park ad train analogy to explain the difference between relapsing and progressive MS. This takes time to prepare.

    • Surely the comment should not be "jibberish" but "I don't understand" explain in a different way and if enough people click this then it would be worth re-writing

    • The opposite perspective: driving home from hospital this afternoon repeatedly saying to myself 'Thank god for Bart's Blog, and they're telling it the way it is!'
      I'm a patient and don't have a scientific cell ('cuse the pun) in my body. Much of what is posted and the replies are beyond me. What is true is that I benefit from getting a handle on the key underlying pieces of information. Due to your posts MD, I asked about testing for neutralising antibodies today (due second round of Alemtuzumab) to be told they don't test for them. So guess what if/when I require a second lot of Alemtuzumab I'll be requesting the test!
      And now I'll know to check out any future possible DMT for the capacity to destroy B cells.
      I'm going to cause some amusement, scoffing, and incredulity by adding this, but it's my truth: the Blog is a life-line and I wouldn't be in the place I am physically and psychologically without the incredible contribution it has made!
      Hope the whole team has a great weekend. 🙂

    • The second round of alemtuzumab should not be a major problem for the majority of people but could be more of a problem on round three. But they can be tested for by the manufacturer as they have done it but to do that first there needs to an acknowledgement that they exist and that they are an issue. They have the data they just need to be transparent.

    • Hi Fi 5.00pm Well said! re possible future dmt to destroy B cells, you mean memory B cells of course?
      And yes, Me 9.16pm, that sounds like a great idea! Save all that unnecessary drug, loss of immunity and not to mention cost!

      Have a great weekend everyone 🙂

  • MDs, you say that MS pathology might be driven by memory B cells. However, when brain damage takes place, like in AIS, all kinds of immune cells, T & B, enter the brain for restoration purposes.

    Does this mean that within an early MS lesion some damaging B cells co-exist and act together with some other protective B cells? Are B cells Janus-cells?

    • Yes there are regulatory B cells and damaging B cells. It is a hole in the idea. Why do B cells accumulate to n the CNS, what makes this special to MS and how is it done.

      I haven't matured the ideas enough…however in my defense when you ask the question about how would a CD4 kill an oligodendrocyte…the answer is equally vague.

    • "how would a CD4 kill an oligodendrocyte"

      You know that i would never ask that question.
      But, don't you think that the contradiction of regulatory and damaging B cells acting side by side, in the light of the indisputable regulatory activity following brain damage, is enough proof that there is no prior damaging activity of B cells whatsoever?

    • Oligos are dead before inflammation starts, remember? Your B-memory cells have killed them, but blood-borne B-cells that inflame the lesion site collect the garbage. Why don't they contribute to the damage instead? They are all supposed to be EBV-infected.

    • If you look in MS tissue there are sick looking oligos surrounded by microglia, they are not necessarily dead and the B cell toxins (e.g antibodies or the high molecular weight toxin reported by Lisak et al.are soluble and so act at some distance from the cells.

    • I guess Lisak et al used B-cells taken from blood directly. Couldn't those toxins be just the manifestation of the regulatory B-cell function? Since oligos are constantly ill anywhere in the CNS, why wouldn't the immune system keep on producing those toxins?

      For your theory to hold, you have to prove that oligos are in perfect health prior to toxin production. However, nothing in the facts available so far can even remotely insinuate that. Activated microglia, B-cell toxins and everything else are always reported with clear damage in sight.

    • What do you suggest we use two photon microscopy on people infected with fluorescent virus to track oligodendrocytes and microglia.

      Activated microglia always has damage in sight, not so sure

  • If the working theory regarding CD 20 B cell depletion is that these cells (specifically the memory B cells) is that the cells harbor EBV, wouldn't a drug targeting CD 19 be even more effective, since EBV is also harbored in plasmablast and plasma cells? These cells are not targeted by the CD 20 drugs, correct?

    Or, would depleting all CD 19 cells be too dangerous, opening the patient up to all kinds of opportunistic infections?

    Interesting that the Intrathecal Rituxamab trials failed, as it would seem that compartmentalized B cells in the CNS would be candidates as primary drivers of the MS disease process. Any thoughts on the failure of these trials (conducted by the NIH, I believe).

    • EBV infects via CD21 which is a late stage mature cell
      marker I think that this then drives the mature cells to differentiate into Germinal centre cells that proliferate and turn into memory cells or plasmablasts. The plasmablasts don't express CD20 and so won't be wiped out by rituximab however it looks like the virus stops the cells turning into plasmablasts and drive the B cell to become a memory cell. Therefore there is no need to evoke the cD19 .However cladribine is a chemical anti cd19.

      The virus lives in the memory B cell doing nothing until it reactivates to create infectious virus.

    • Why didn't rituximab work. It was put into the spinal fluid to get rid of plasma cells in the brain. (A) plasma cells do not express any CD20 and so will never be depleted by the antibody and (b) the flow of the cerebrospinal fluid was down and out into the blood so the rituximab didn't touch the rain b cells and simply wiped out the B cells in the blood. It was known that this approach was not good enough to target non active ms. So it failed in all studies.

      So next they will want to put it in the ventricles and to get a better flow but it still won't kill plasma cells. Why do it I don't know.

      5minutes of thought verse experiments costing thousands..but hey I'm just a heretic who doesn't know any better:-(

    • Apart from cladribine. We might not know if enough gets in the brain but it's all we have at the moment so it's worth a try.

    • Please Dr explain your theory in an article if possible because I have more questions than before. Cars and trains would do too

  • MD, thanks for the replies. I suspected that part of the reason for the Intrathecal Rituxan failures might be due to the drug being washed out of the CNS before it could be effective. Hadn't considered that it doesn't target the cells that needed targeting.

    So, EBV only resides in memory B cells? Interesting. Any work being done on ways to specifically target only these cells?

  • Great work Md

    "However cladribine is a chemical anti cd19"

    Do you mean cladribine kills memory B cell ? Or only cd 19+?


    • The specificity of cells targeted by cladribine is likely broader than anti-CD20 antibodies and should hit the pro-B population. It certainly hits memory B cells as we will be presenting at ECTRIMS along with target profiling

  • EBV infects via CD21 which is a late stage mature cell
    marker I think that this then drives the mature cells to differentiate into Germinal centre cells that proliferate and turn into memory cells or plasmablasts. The plasmablasts don't express CD20 and so won't be wiped out by rituximab however it looks like the virus stops the cells turning into plasmablasts and drive the B cell to become a memory cell

    EBV is transmitted via oral secretions, enters through the epithelium that lines the oropharynx
    (Waldeyer's ring) and infects naïve B cells, which differentiate through a germinal center
    reaction into memory B cells, where EBV persists for the lifetime of the host [28]. Upon returning
    of EBV-infected memory B cells to the oropharynx these can differentiate into plasma cells
    that subsequently initiate viral replication and release virions that are shed into the saliva [29].

    This article has sought to cast EBV in the role of a stealth virus that has adapted to the biology of the
    normal B cell in order to establish life long, persistent infection. Although the details of the model will
    need refining, the data at hand make such a model virtually inescapable. Most striking is that the virus
    has specific gene transcription patterns uniquely designed for each stage of the process:

    "In this paper we demonstrate that the cells which initiate replication of Epstein-Barr virus (EBV) in the tonsils of healthy carriers are plasma cells (CD38hi, CD10−, CD19+, CD20lo, surface immunoglobulin negative, and cytoplasmic immunoglobulin positive)"

    This article has sought to cast EBV in the role of a stealth virus that has adapted to the biology of the
    normal B cell in order to establish life long, persistent infection. Although the details of the model will
    need refining, the data at hand make such a model virtually inescapable. Most striking is that the virus
    has specific gene transcription patterns uniquely designed for each stage of the process:
    1. The virus will enter any resting B cell and cause it to proliferate, guaranteeing efficient access of
    the virus to the B cell pool and amplifying the viral genome copy number. In doing so, it uses a set of
    genes that recapitulates what is necessary to cause a B cell to become a blast on the way to
    differentiating to a plasma or memory cell.
    2. The virus will infect any resting B cell, but in vivo it ends up in a very specific subset, the resting
    memory B cell – precisely the best place to be for long term persistence. When there, it shuts off gene
    expression so the cells cannot be immunosurveilled. Instead the cells are maintained at stable levels
    for long periods of time as though they were normal memory B cells – precisely the state required for
    benign, long term persistence in a healthy host.
    3. The memory cells appear to go through an activated state when they enter mucosal epithelium where
    they express only the minimal viral information necessary for survival, the LMPl/Th surrogate, the
    LMP2a/BCR surrogate and EBNAl(Qp) to be sure the viral genome is maintained.
    4. When the cells become activated in the mucosal epithelium they behave like normal memory B cells
    in a secondary response. Most of the cells produced undergo terminal differentiation to become plasma cells. The virus responds by reactivating and producing infectious virus so that it can be released into
    the saliva.

    Life Scieaces, Vol. 65, No. 14. pp. 1433-1453, 1999

    Is this contradictory to your reasoning?


  • Thanks for the quotes and for the reader viewing these comments naive cells are the same as mature cells

    No I don't think this is contradictory at all.

    To us it indicates the virus has co evolved with humans. It survives to live another day and hijacks the B cells, it makes B cell growth and differentiation factors and immortalises B cells and it drives cells to house the silent virus and activates to create live virus.

    We think the advantage to the human is that this provides for long lasting immunity, the price is a few cancers, glandular fever and autoimmunity, but as the autoimmunity often occurs at an age post reproductive age remembering most people would be teenagers in our history, it would not be selected against. A testable hypothesis but by no means

  • Thanks everyone for all this ebv info. I'm doing my best to keep up and getting there slowly. Bought a good book too:

    (not sure if that counts as advertising and will be deleted? But I see author John Oxford is a fellow Barts prof so maybe that's allowed)

    Sounds like ECTRIMS quite the place to be this year, look forward to learning more about cladribine thanks MD (rare trip out of the mousehole, you'll have to be nice to all those important neuros and smooth pharma reps) and ATA188 too.

    Re ATA188 it's a tad ironic that after all these years of trying to deplete T cells someone is (hopefully) getting good results doing the polar opposite with CD8. Oh, talking of which, isn't there this big T v B debate too?

    Re MD 12.35 ebv co-evolved with humans and lifelong immunity. Do people produce memory B cells at all without ebv? If so, what is the difference between a normal mem B and ebv mem B?

    Thanks again MD, Luis and Nissan.

    • what is the difference between a normal mem B and ebv mem B?
      I don't know by I suspect longevity.

      There are people without EBV, they will have B cell memory but will it be the same…we have the hypothesis and are trying to test this, we have bloods coming.

    • "I don't know by I suspect longevity."

      For example, one study found
      IgG1+ memory B cells specific for the 1918 pandemic
      strain of influenza virus circulating in the blood of survivors
      90 years after primary exposure to the virus50


  • Do people produce memory B cells at all without ebv?

    Sure ,if you catch an infection (other than ebv):
    Memory B cells are a B cell sub-type that are formed within germinal centers following primary infection and are important in generating an accelerated and more robust antibody-mediated immune response in the case of re-infection (also known as a secondary immune response)

    " what is the difference between a normal mem B and ebv mem B?"

    "EBV is transmitted via oral secretions, enters through the epithelium that lines the oropharynx
    (Waldeyer's ring) and infects naïve B cells"

    Mem b are, not infected with ebv

    • "Mem b are, not infected with ebv"

      "EBV episome replication during B cell division is now known to be inefficient, resulting in some descendant B cells becoming EBV-free after a few dozen divisions. EBV-free memory B cells in the CNS may thus have descended from a memory B cell which matured while containing EBV episomes, enabling its B cell receptor to recognize “forbidden” MS-causing antigens in the CNS, even if EBV is absent from this site."

      •Epstein–Barr virus (EBV) is involved in the pathogenesis of MS.
      •EBV DNA generally cannot be detected in central nervous system.
      •EBV may enable the recognition of “forbidden” antigens by memory B cells (Pender's hypothesis).
      •Inefficient EBV episome replication during B cell division results in some descendant B cells becoming EBV-free.
      •Memory B cells in the CNS are mostly EBV-free and can recognize “forbidden” MS-causing antigens in the CNS.

    • The question is why the CNS, the same argument for EBV can be made for other conditions such as arthritis, where does the tissue selectivity occur?

    • It's a bit of a leap to extrapolate from the freshwater hydra to the human but there's obviously a lot of work going on investigating the microbiome and disease. Whether anything useful results is another matter, I try to keep an open mind but to me it's the latest bandwagon to be jumped on.

  • Being a beginner and trying to make a sense from it all… I understand that you believe that there is only B cell involvement in the autoimmuninity and there is no role for T cells and their interaction. And that B cell depleting therapies such as ocrelizumab and cladribine are successful in the way that they approach the resolution of the disease but fail because the CNS is very complex. Is that so? Why B cell therapies are not 100% successful to other autoimmmune diseases? Is it because it is hard to remove all memory B cells, do we need a wider B cell depletion agent or does it mean that there are other factors in the immune system that stay untouched by only Bcell targeting therapies? Is it ever possible to achieve a wider depletion that is not harmful to the overal health?

    • I am open to T cells being involved and would be unusual if they were not but at present we ask the question are they critical.
      We have made the suggestion at all MS drugs work the same way and the more efficient they are at depleting memory B cells the more efficient they are. They are very effective at controlling relapsing disease

      The potential failings is how much of each agent gets in the brain.

      No treatment is 100% effective and B cell therapies do work in other diseases.

      We dont't need to remove all B cells we need to be more surgical in what we remove

    • "the more efficient they are at depleting memory B cells the more efficient they are."

      Do higher does deplete more B cells?..beta-seron is supposed to be once every 48 hours but seems better 24-36 hours.

    • I don't know it would be interesting to see the differences

      Immunology and Cell Biology (2016) 94, 886–894; doi:10.1038/icb.2016.55; published online 5 July 2016
      Interferon-β therapy specifically reduces pathogenic memory B cells in multiple sclerosis patients by inducing a FAS-mediated apoptosis

      In that study they looked at beteraferon and rebif but did not distinguidh between the two

  • " Is it because it is hard to remove all memory B cells, do we need a wider B cell depletion agent or does it mean that there are other factors in the immune system that stay untouched by only Bcell targeting therapies"

    Long-lived plasma cells (LLPCs) are an unmet therapeutic challenge, and developing strategies
    for their targeting is an emerging goal of autoantibody-mediated diseases such as
    systemic lupus erythematosus (SLE). It was previously shown that plasma cells can be
    depleted by agents such as bortezomib (Bz) or by blocking LFA-1 and VLA-4 integrins.
    However, they regenerate quickly after depletion due to B cell hyperactivity in autoimmune
    conditions. Therefore, we compared different therapies for the elimination of LLPCs combined
    with selective B-cell targeting in order to identify the most effective treatment to eliminate
    LLPCs and prevent their regeneration in lupus-prone NZB/WF1 mice.

    Our data indicate that effective depletion of long-lived and autoreactive plasma cells can
    only be achieved by depletion regimens containing the proteasome inhibitor bortezomib. The
    combination of bortezomib with integrin blockade does not seem to provide an additional synergistic

    • Yes interesting post Luis and well spotted…you may be interested to note we are on the case already and may write more soon.

      You info on PLC G2 was interesting thanks.

  • Sharing is caring

    Thanks a lot… glad to be helping

    Especially because you have no Coi 🙂

    Also they did a SLE trial

    The proteasome inhibitior bortezomib depletes
    plasma cells and ameliorates clinical manifestations
    of refractory systemic lupus erythematosus

    (Its open source)

    They have this b an plasma cells populations followup recovery ,in

    graphs, i think it relates to your work



    • Thanks

      It s not only Hsct…lolllll

      In contrast,
      mice conditionally deleted for Mcl-1 in activation-
      induced cytidine deaminase producing
      cells were able neither to form GC, nor
      to produce memory B cells and GC-derived
      plasma cells. Antigen-specific B cells did notexpand even before the onset of the GC
      reaction, indicating that Mcl-1 dependency
      precedes GC formation

      anti-apoptotic molecule myeloid cell leukemia sequence 1 (Mcl-1

      precedes GC formation. However, immunization
      of mice, in which Mcl-1 was deleted in
      either one or both alleles demonstrated that
      GC B cells are even more sensitive to Mcl-1
      deletion than activated B cells present before
      GC induction.

    • Dont know

      Being involved in systemic inflammation and regulator of

      imumne cells, blocking it would send dangers signal to others

      cells to make maters worse in auto immunne disease context

      (i am speculating really) Cytokines are tricky

      It has been suggested that TNFα antagonists may increase the risk of demyelinating diseases in patients with RA by about 30% [32], however, these data are not supported by others [38]. The overall prevalence of RA and MS is 0.6% and 0.05% respectively [23]. The occurrence of both diseases in the same patient has been reported [51]. This coincidence should be not surprising because both RA and MS share pathogenetic and genetic similarities [51]. In this context patients having one autoimmune disease are at increased risk of developing another. In favor of this, is the existence in our study of two patients with preexisting MS-like lesions.


By MouseDoctor



Recent Posts

Recent Comments