How Does B-Cell Tolerance Contribute to the Protective Effects of Diabetes Following Induced Mixed Chimerism in Autoimmune Diabetes?

  1. F. Susan Wong
  1. Institute of Molecular and Experimental Medicine, Cardiff School of Medicine, Cardiff University, Cardiff, U.K.
  1. Corresponding author: F. Susan Wong, wongfs{at}

In type 1 diabetes, autoimmune-mediated damage and destruction of insulin-producing β-cells lead to a raised blood glucose. Glucose itself is toxic and some recovery of β-cell function is expected in the honeymoon phase after insulin treatment is instituted, but it was previously believed that all β-cells will ultimately be lost. In fact, there are individuals who continue to have some functioning β-cells, even many years after the onset of type 1 diabetes (1,2). This raises the hope that treatment to induce replication or regeneration of β-cells could be instituted in patients with long-standing diabetes, and makes it even more imperative that means of halting autoimmunity are found.

Although nonmyeloablative autologous stem cell transplantation has been carried out, with long-term insulin independence successfully achieved in type 1 diabetes (3,4), many would feel that the risk-benefit ratio of this treatment is not acceptable for general use. Many other strategies, both nonantigen- and antigen-specific therapies, have been trialed in patients with new-onset type 1 diabetes, but none, as yet, has provided a long-term solution to the autoimmune attack on remaining β-cells (5). However, short-term slowing of β-cell loss was seen with T-cell−targeted therapy using nondepleting anti-CD3 (6,7), which was shown to temporarily reduce T cells, but, more importantly, to increase T-cell regulation. In addition, anti-CD20 treatment (rituximab) that targets B cells also temporarily slowed loss of endogenous insulin production (8,9), and currently there is a move to consider therapies that may be combined in order to achieve improved results.

Racine and colleagues have developed a strategy to stop autoimmunity by induction of mixed chimerism in NOD mice. The treatment uses a nonablative therapy comprising nonmyeloablative conditioning with anti-CD3 and anti-CD8 monoclonal antibodies, followed by infusion of major histocompatibility complex-mismatched bone marrow and CD4-depleted spleen cells (10). This therapy is combined with gastrin and epidermal growth factor, and Racine et al. (11) have shown that not only is tolerance induced but also return of β-cells is demonstrated. They have previously shown that T-cell tolerance is induced by this means, but in this issue, Racine et al. (12) suggest that B-cell tolerance is also an essential component of the effectiveness of the treatment. Figure 1 illustrates B-cell processes that contribute to type 1 diabetes, and therefore why this subset of lymphocytes should be considered in immunotherapeutic targeting strategies.

Figure 1

Interactions and functions of B cells that may influence autoimmune diabetes. 1. B cells bind antigen specifically via cell surface immunoglobulin—the specificity of the immunoglobulin directs processing of the protein (15). 2. The B-cell specificity increases the ability of B cells to present protein antigen to CD4 T cells (16). 3. B cells enhance antigen presentation to CD8 T cells (17). 4. B cells may enhance antigen presentation by dendritic cells (18). 5. B cells differentiate to plasma cells, which produce autoantibodies. MHC, major histocompatibility complex; TCR, T-cell receptor.

Racine et al. (12) have taken care to show effects not only on preexisting B cells but also on newly developing B cells. Following induction of tolerance and the infusion of the major histocompatibility complex-mismatched bone marrow, they depleted preexisting autoreactive B cells that included CD19HiCD138Lo preplasma and CD19LoCD138Hi plasma cells, among which are cells that can produce autoantibodies. Anti-insulin antibodies were reduced, although total immunoglobulin G was maintained in the treated mice. Other changes in B-cell subsets were documented, including a rise in immature B cells in the bone marrow, but a decrease in recirculating mature B cells. In addition, NOD mice have a lower proportion of T1 transitional (CD24HiCD21Lo) and higher proportion of T2 transitional (CD24IntCD21HiCD23+) B cells compared with C57BL/6 mice. However, after treatment, an increased ratio of T1/T2 cells in treated mice was dominated by donor T1 B cells, as found in normal C57BL/6 mice. This alteration of ratios of T1 to T2 B cells rebalances the lower T1/T2 transitional B-cell ratio in NOD mice. Increased apoptosis is seen in the host B cells, but how this is effected is not known. The authors speculate that donor B cells rather than donor cytotoxic CD8 T cells are necessary for the cell death, possibly due to the nonautoreactive donor B cells competing with the autoreactive host B cells for survival factors that include B-cell activating factor.

What is interesting about the study by Racine et al. (12) is how the normalization of B-cell subsets toward a profile that appears more like that seen in C57BL/6 mice affects diabetes and how much this contributes to the reduction of autoimmunity. The conditioning strategy used to protect from autoimmune diabetes in this study focuses on depleting T cells with antibodies and this reestablishes deletion of autoreactive T cells in the thymus (13). Although B cells are not directly targeted by the preconditioning, this reestablishment of a “normal” B-cell compartment is clearly also playing a role in the protective effects seen. Is it mainly the reduction in the autoreactive B cells that reduces disease activity and how does the “normalization” of the ratios of B cells contribute to the maintenance of tolerance? How the repopulation of both T- and B-cell subsets occurs after depleting autoreactive T and B cells is likely to be critical to success of treatments for autoimmunity that aim to reduce pathogenic autoreactive cells and “reset” the immune system.

In respect to the anti-CD3 or anti-CD20 treatments used so far in diabetes, which have had no tangible long-term effects, it is likely that this repopulation of the T- or B-cell compartments and “resetting” of the de novo−produced B and T cells have not been optimal. However, the process must be carefully controlled; otherwise, there is the potential for unwanted effects. For example, alemtuzumab (anti-CD52), which depletes both T and B cells, has been successfully used to treat multiple sclerosis. However, in >30% of successfully treated patients, where B cells repopulate earlier after treatment than T cells, thyroid autoimmunity is unmasked. Problems are caused by production of antibodies that stimulate the thyroid-stimulating hormone receptor leading to hyperthyroidism (Graves disease) or hypothyroidism due to antithyroid antibodies (14).

In summary, the study by Racine et al. (12) illustrates an interesting model system where depletion of both autoreactive T and B cells, combined with an active strategy to induce and “reset” both T- and B-cell tolerance after treatment, can induce long-term remission from autoimmune diabetes. The challenge will be to bring treatments that incorporate these features, but that have an acceptable safety profile, to therapy of human type 1 diabetes.

Article Information

Funding. Research in the laboratory of F.S.W. is funded by the Medical Research Council (U.K.), Diabetes UK, and the European Union 7th Framework Programme for Research and Technological Development.

Duality of Interest. No potential conflicts of interest relevant to this article were reported.


  • See accompanying article, p. 2051.

Readers may use this article as long as the work is properly cited, the use is educational and not for profit, and the work is not altered. See for details.


| Table of Contents