Raging autoimmune T cells were fingered as the culprits in the toxicity of the failed first-generation Alzheimer disease (AD) vaccine. Even so, one research group suggests that some alleged immune attackers could turn out to have a softer, more nurturing side. If a paper published online in Nature Neuroscience on January 15 receives independent confirmation, then T cells might yet turn out to give critical support to the process of neurogenesis in the hippocampus of adult mice. In the new work, Michal Schwartz and colleagues from the Weizmann Institute of Science in Rehovot, Israel, show that the absence of T cells in two mouse immunodeficiency models is associated with lower levels of neurogenesis as measured by bromodeoxyuridine incorporation in vivo. Transfusions of T cell-rich spleen cells from normal mice reversed the deficit. Transgenic mice that have a surfeit of central nervous system-directed immune cells (in the form of myelin-reactive T cells) displayed more active neurogenesis, and better performance in a spatial learning and memory task compared to wild-type controls. Increased neurogenesis was accompanied by microglial activation and expression of brain-derived neurotrophic factor in these mice. At this point, the paper is not directly relevant to AD, as current data on AD and neurogenesis remain preliminary.

The new work appears to support Schwartz’s proposal, based on her previous work, that well-behaved, self-directed T cells confer a “protective autoimmunity” that can promote the health and repair of the nervous system. In this scheme, beneficial activation of microglia by T cell cytokines leads to the production of neurotrophic factors. But this idea is controversial, especially as it relates to the proclivities of myelin-reactive T cells. Schwartz and colleagues report that these cells ameliorate the effects of spinal cord injury (Hauben et al., 2000), where others find just the opposite (see ARF related news story). Few laboratories have as yet found corroborative evidence for Schwartz’s hypothesis, and the present paper cites a preponderance of her own papers as prior evidence. Even so, the new work provides additional evidence for some involvement of T cells in beneficial interactions with the nervous system. Much of that evidence is indirect at this stage, but the paper should serve to fuel interest in the role of the peripheral immune system, and particularly CNS-directed T cells, in neurodegeneration and repair.

Work by Fred Gage and colleagues previously showed that the birth of new neurons in the dentate gyrus of rat hippocampus could be stimulated in an enriched environment (Kempermann et al., 1997). In the current paper, dual first authors Yaniv Ziv and Noga Ron used this paradigm to search for a role for T cells in the process. Rats that were kept in plain accommodations, or treated to more stimulating cages, were dosed with BrdU to label dividing cells, and then killed after one week. Staining hippocampi with an anti-BrdU antibody showed a significant increase in cells labeled with both BrdU and the neuronal marker NeuN. The number of microglia was also increased, and a subset displayed some evidence of activation by T cell cytokines, with enhanced expression of MHC class II and insulin-like growth factor. The additional observation of the occasional T cell in the brain parenchyma suggested the hypothesis that T cells could influence the brain environment, perhaps rendering it more conducive to neurogenesis.

To further probe the connection between T cells and neuron birth, the researchers employed a novel strategy of studying hippocampal neurogenesis in two immunodeficient strains of mice. The severe combined immunodeficiency (SCID) mouse and the nude mouse both lack T cells. Both exhibited lower levels of neurogenesis compared to wild-type strain-matched mice, as measured by the detection of cells positive for both BrdU and NeuN or doublecortin in the hippocampus. In either strain, transfer of T cell-containing spleen preparations enhanced neurogenesis. T cells seemed to be required for the response to environment, since enrichment failed to increase neurogenesis in SCID mice, in contrast to its positive effect on immunocompetent mice.

If T cells in the CNS promote neurogenesis, then mice with many CNS-directed T cells might be particularly well endowed with new neurons. The authors tested this hypothesis by comparing two strains of mice transgenic for T cell receptor genes. One strain carries a transgene for a T cell receptor against myelin basic protein (TMBP), and 98 percent of the animal’s T cells are myelin-reactive. The other strain (TOVA) carries an ovalbumin-specific T cell receptor transgene. The TMBP mouse had significantly more BrdU/DCX double-labeled cells than its strain-matched control, while the TOVA mice had far fewer. The enhancement in TMBP mice seemed to be mediated by microglia, since the inhibitor minocycline reduced the level of neurogenesis. The authors showed no direct measurement or localization of T cells in the brain, so exactly how T cells might mediate this effect is unclear.

Finally, to assess behavioral correlates of neurogenesis, the researchers tested TMBP and TOVA mice in the Morris water maze, a spatial learning and memory test that requires hippocampal input. In all aspects of the water maze, the TMBP mice performed better than their matched controls, while the TOVA mice performed worse. BDNF levels correlated with the water maze results: The TMBP mice had higher levels of BDNF derived from neurons, while the TOVA or SCID had lower levels.

Given that the peripheral immune system can inhibit neurogenesis in the brain via inflammation, the idea that immune cells could also promote neurogenesis in some instances may not be such a surprise. The effects of immunodeficiency and T cell replenishment on neurogenesis in this study are consistent with just such a role. The data do not unequivocally show that T cells are responsible for the neurogenesis. More work will be required to test this intriguing hypothesis, and to ultimately determine how the peripheral immune system can best be manipulated to treat neurodegenerative disease.—Pat McCaffrey


  1. The paper by Ziv et al. brings together two often disparate fields of study: immunology and neuroscience. The group of Michal Schwartz is one of a relatively few in the world who draws on tools of both trades to study how the immune and nervous systems intersect to influence brain function.

    The authors propose the interesting hypothesis that the hippocampal (and olfactory) neurogenesis required for optimal functioning of the adult brain is dependent on cues from peripheral immune cells. It had been shown previously that inflammatory activation of the peripheral immune system can diminish neurogenesis in the brain. This work suggests that the converse, that is, that neurogenesis depends in some way on immune support, may also hold true.

    The authors' use of SCID and nude mice for these studies is quite innovative, and the experiments carefully control for differences in genetic background that are known to influence adult neurogenesis. The decrement in BrdU+ cells, and specifically BrdU/DCX and BrdU/NeuN cells, in the immune-deficient mice is consistent with their hypothesis.

    However, it is the reconstitution experiments examining neurogenesis after replenishing the immune system with normal or T cell-depleted splenocytes that to me forms the crux of this study. Restoration of normal neurogenesis by the introduction of donor splenocytes is the definitive proof that the neuronal precursor cell population is intact and simply requires external activation from the added immune cells. The description of these experiments might have benefited from more detailed display of this data on which to evaluate the results. For instance, in Figure 2 showing the first of the reconstitution experiments, the number of BrdU/DCX+ cells per dentate gyrus after addition of normal splenocytes (panel d) appears comparable to the number of BrdU/DCX+ cells found in unreconstituted SCID mice at the same age (panel a). The intended comparison is to SCID mice reconstituted with T cell-depleted splenocytes, but data for the level of neurogenesis in untreated mice would have been a good control to include in the same panel.

    The reconstitution experiments are especially important in evaluating the data shown in Figure 4, which presents data of the study of neurogenesis in nude mice. The very disrupted DCX staining in the nude mice of Figure 4c suggests that the gene defect in these mice may affect neurogenesis in ways independent of T cell function. After all, the mice are also nude, and have hair follicle deficits that may have nothing to do with alterations in the immune system. For these experiments, the authors do plot the data one would have liked to see for the SCID experiments; specifically, they show untreated nude vs. nude + splenocytes, demonstrating that there is significant recovery of newly dividing cells in the hippocampus. Here it is worth noting that PCNA staining does not equal neurogenesis (PCNA, like BrdU, does not distinguish between cell types), and the experiment could be stronger if it provided the same comparison (untreated nude vs. nude + splenocytes) for BrdU/DCX+ double labeled cells to show a specific effect on neuronal production.

    One goal to tackle for follow-up work is to convincingly connect the function of peripheral T cells to the effector microglia in the brain, and to explain how microglia then act on progenitor cells to increase neurogenesis. The present work shows T cells in the ventricles, where they might be able to directly influence the turnover and/or differentiation of precursor cells in the subventricular zone (the source of neurogenesis for olfactory bulb interneurons). It would be fascinating to know how they signal to microglia in the parenchyma of the brain to activate precursor cells deep in the dentate gyrus. Identifying the signaling factors used to communicate between these distant areas will solve the spatial paradox that exists based on the data available so far.

    This paper should serve to start the neuroscience community thinking more seriously about the interaction of body and mind. There may be a lot more to it than most researchers or clinicians now realize.

  2. Excellent article, and studies done were well-controlled. This article provides further validation of the communication between the central nervous system and the immune system. In this article, the authors showed that T cells (of the immune system) that reside in the central nervous system (CNS) can influence both neurogenesis and cognitive functioning independently. Under normal conditions, the resident T cells and microglia in the CNS are barely detectable, but when neurogenesis was enhanced by housing rats in an enriched environment, both T cells and microglia were activated. When neurogenesis was examined in mutant mice deficient in T cells, the authors found that neurogenesis was decreased compared to the control mice. It is interesting that even when the mutant mice were housed in the enriched environment, this did not help in increasing neurogenesis, as is usually seen in normal animals. However, when the mutant mice were injected with "splenocytes" containing replenished T cells, increased neurogenesis was seen when compared to mice depleted for T cells. Furthermore, the authors found that the influence of CNS T cells in neurogenesis is partly mediated by microglial cells, because when they gave the mice minocycline (a drug that inhibits microglial activity), a significant decrease in neurogenesis was seen compared to controls.

    What is also interesting in this study is that mere activation of resident T cells in the CNS does not result in increased neurogenesis. The authors showed that the antigen (or protein) that induces T cell activation has to be specific (i.e., proteins involved in brain plasticity) for effective involvement in neurogenesis. This is demonstrated when they examined two different types of transgenic mice (one that expresses T cell receptors that recognize myelin basic protein—protein that is involved in axonal growth—and another that expresses T cell receptors that recognize ovalbumin—general protein); they found increased neurogenesis and enhance learning ability only in the transgenic mice that express T cell receptors that recognize myelin basic protein when compared to their control counterparts. The thoroughness of the experiments presented in this article provides good evidence that the immune system is involved in maintaining CNS integrity. As the authors suggest, the results of their experiment may partially explain the association between age-related decline in neurogenesis and decreased immune functioning related to aging.

    Further corroboration is needed before the evidence can be convincingly accepted.
    With regard to its applicability in Alzheimer disease, the results of this study
    may be limited. It should, however, trigger thoughts on the role of the aging immune system in influencing brain function, as well as on the bidirectional
    communication between the CNS and the immune system.

  3. This paper supports the general theory of bidirectional and tight communication between the nervous and immune systems. Microglia as components of both organ systems are critical players, presumably translating the T cell presence into a phenotype associated with increased cell proliferation and improved memory. Potential mediators include microglial IGF-1, and neuronally derived BDNF. The use of immune-deficient mouse models, both SCID and nude, coupled with the T cell adoptive transfer experiments and the T/MBP transgenic mouse provides in my mind pretty compelling evidence that brain-directed T cells facilitate neurogenesis and also improve hippocampal function as tested by a standard hippocampal-dependent behavioral task.

    This study is also significant because it raises some interesting possibilities for neurodegenerative disorders in addition to the role of the immune system in maintaining "normal" brain function.
    One intriguing question is whether the brain, during enrichment, expresses signaling or trophic molecules that induce the trafficking of brain-specific T cells into distinct regions of the brain. Is there something different about microglia (and other resident cells) within the brains of these immune-deficient mice that reduces the expression of molecules involved in both basal and enrichment-induced neurogenesis?

    It is tempting to speculate on the ramifications of these results for Alzheimer disease, particularly in view of the interest in vaccines and antibody strategies targeting β amyloid. Are there endogenous T cells circulating in those of us without AD that continuously survey the brain, help maintain neuronal health (and neurogenesis), as well as inducing microglia to mop up amyloid fibrils and microdeposits before they reach plaque status? Moreover, what is the relationship between the cellular and humoral arms of the immune system with respect to both neurogenesis, cognition, and Alzheimer disease? Like any good science, this study opens up the field to more questions than answers.

    View all comments by Matthew During

Make a Comment

To make a comment you must login or register.


News Citations

  1. Spinal Tap Dance—Do Myelin-Reactive T cells Protect or Destroy?

Paper Citations

  1. . Passive or active immunization with myelin basic protein promotes recovery from spinal cord contusion. J Neurosci. 2000 Sep 1;20(17):6421-30. PubMed.
  2. . More hippocampal neurons in adult mice living in an enriched environment. Nature. 1997 Apr 3;386(6624):493-5. PubMed.

Further Reading


  1. . Protective autoimmunity and neuroprotection in inflammatory and noninflammatory neurodegenerative diseases. J Neurol Sci. 2005 Jun 15;233(1-2):163-6. PubMed.

Primary Papers

  1. . Immune cells contribute to the maintenance of neurogenesis and spatial learning abilities in adulthood. Nat Neurosci. 2006 Feb;9(2):268-75. PubMed.