Could an FDA-approved drug that trains the immune system on tumors benefit the brain? According to a report in the January 18 Nature Medicine, an antibody that boosts the cancer-fighting ability of T cells also prods them to issue a rallying cry that calls peripheral macrophages to the central nervous system. Scientists led by Michal Schwartz, Weizmann Institute of Science, Rehovot, Israel, report that these immune cells then clear Aβ plaques and improve memory in mouse models of Alzheimer’s disease.

“These data add to the growing body of evidence suggesting that peripheral adaptive immunity plays a role in the pathophysiology of Alzheimer’s disease,” said Guillaume Dorothee, INSERM, Paris, who was not involved in the work. “In line with other recent studies, this suggests that immunomodulatory strategies in the periphery may have therapeutic potential in AD.” The ideal immune target and strategy still remain to be determined, he said.

Previously, Schwartz and colleagues reported essentially the same outcomes in mouse models when they used genetic methods to temporarily deplete regulatory T cells in the periphery—effectively easing the brakes on the immune system (Sep 2015 news). This response depended on a burst of interferon-γ (IFN-γ) from circulating effector T cells. The cytokine stimulated the choroid plexus to recruit monocyte-derived macrophages from the periphery to the brain. There, the myeloid cells surrounded and cleared Aβ plaques. In the current study, the group wanted to test a more clinically relevant strategy for eliciting an IFN-γ response from effector T cells.

The FDA-approved melanoma drug pembrolizumab from Merck elicits a similar IFN-γ response. Known as KEYTRUDA, this antibody neutralizes the programmed T cell death 1 (PD-1) receptor on effector T cells. PD-1 normally keeps T cell activity in check and suppresses tumor-fighting activity. It is known as an immune checkpoint. Without PD-1, these cells release IFN-γ and can once again kill tumors (Mamalis et al., 2014). In this paper, the research group tested a similar anti-PD-1 antibody, specifically for use in animals.

To investigate, first author Kuti Baruch from Schwartz's group, collaborating with Ido Amit's immunogenomics group, tested the effects of the anti-PD-1 antibody in 10-month-old 5XFAD mice, which had accumulated significant cerebral Aβ plaques. The researchers injected the mice intraperitoneally twice, three days apart, with either the PD-1 antibody or an IgG control. Some animals got a second round of treatment a month later.

The anti-PD-1 antibody appeared to elicit a robust IFN-γ response. A week after the first injection, more CD4+ T cells from the treated mice were producing IFN-γ and RNA sequencing revealed an IFN-γ-associated expression profile at the choroid plexus. The researchers isolated the choroid plexus from mouse brain and analyzed it, but did not compare it to other tissues. More myeloid cells infiltrated the brains of treated mice, while astrogliosis and Aβ plaque load in the hippocampi and cerebral cortices fell by half. Mice that got a second round of injections wound up with even less Aβ.

These pathology benefits seemed to translate to behavior, the scientists reported. A month after the first round of injections, treated mice remembered the location of a hidden platform in a radial arm water maze better than untreated controls. Mice given a second round of treatment performed almost as well as wild types. Cognitive deficits returned in mice that got only one set of injections, suggesting that repeat dosing is needed to maintain benefits.

The results appeared to extend to other mouse models of Alzheimer’s. In eight- or 11-month-old APP/PS1 mice, the anti-PD-1 antibody reduced Aβ plaque area and number by at least half. No behavioral assays were reported for these mice.

“This is the first time immunotherapy based on immune checkpoint blockade is suggested in the context of a neurodegenerative disease,” Schwartz told Alzforum, adding, “Since it is based on an existing FDA-approved therapy for cancer, it can potentially be immediately tested in patients suffering from Alzheimer’s disease.” She pointed out that the therapy is not directed against any specific disease pathology, but rather helps the immune system “cleanse” the brain of toxic materials, including Aβ. She plans to explore whether blocking other immune checkpoints treats AD mouse models. A Merck representative said the company has no current plans to test pembrolizumab in AD.

Though Schwartz’ data imply that boosting systemic immunity could help clear Aβ when a full load of amyloid pathology is present in the brain, Dorothee found a different result at an earlier stage of pathogenesis in mice, when plaques first appeared (Apr 2015 conference news). In that soon-to-be-published study, microgliosis rose and performance on behavioral outcomes improved when regulatory T cells were stimulated, not inhibited. Disease worsened in their absence. “To me this suggests a complex and dynamic process that involves multiple immune effectors, with changes depending on the state of neuroinflammation and the stage of disease progression,” Dorothee said.

Gabriela Constantin, University of Verona, Italy, agreed that PD-1 inhibition should be studied at earlier stages of AD in models (see full comment below). Constantin and Dorothee noted that removal of β-amyloid by any means has yet to show a clinical benefit in human trials. They proposed testing PD-1 therapy in tau models, which Schwartz is currently doing. In addition, Constantin said that PD-1 blockade exaggerates inflammatory disease in animal models (Salama et al., 2003). Researchers should investigate how such outcomes will affect AD, she cautioned. Side effects of pembrolizumab include inflammatory reactions in the lung, liver, and other organs.—Gwyneth Dickey Zakaib


  1. This work by Baruch et al. sheds further light on the role of immune mechanisms in the pathogenesis of Alzheimer's disease. The authors found that inhibition of the programmed death-1 (PD-1) pathway using an anti-PD-1 antibody induces an interferon (IFN)-γ–dependent systemic immune response, followed by increased expression of ICAM-1 and CCL2 on choroid plexuses and recruitment of monocyte-derived macrophages to the brain. These data are interesting and in line with previous studies from the same group showing that circulating monocytes reduces AD pathology and that boosting immune responses is beneficial in brain repair.

    PD-1 inhibitors are currently in use for the treatment of several types of cancer and the study by Baruch et al. opens an attractive perspective for the use of these drugs in patients with AD.

    A remarkable finding of this study is that the treatment with an anti-PD-1 antibody leads to drastic reduction of Aβ accumulation in 5XFAD mice during later stages of disease (10 months) that are normally characterized by severe cerebral Aβ plaque pathology. Also, after treatment with the anti-PD-1 antibody, the performance levels were com­parable to those of wild-type mice in the radial arm water maze task. The relevance of these findings needs to be further confirmed in other animal models of AD showing not only Aβ accumulation but also tau pathology. Other cognitive tests may reinforce these results by showing a more consistent effect on the cognitive deficit after PD-1 inhibition. Aβ clearance in human subjects during late disease stages has no significant therapeutic effect and therefore it would be of interest to clarify the effect of PD-1 inhibition on earlier disease phases in AD models.

    It has been recently shown that PD-1 blockade attenuates the function of myeloid-derived suppressor cells (Yu et al., 2015). Therefore, future studies are needed to understand which is the type of myeloid cells that enter the CNS after PD-1 inhibition and that are apparently responsible for the therapeutic effect of the anti-PD-1 antibody.

    PD-1 is a member of the CD28 superfamily that delivers negative signals on interaction with its two ligands, PD-L1 and PD-L2. Blocking PD-1 signaling enhances proliferation of T and B cells, increases the production of numerous cytokines and heightens autoimmune responses.  Impairment of the PD-1 pathway leads to accelerated and more severe disease in other models of inflammatory diseases, including brain chronic inflammatory diseases (Salama et al., 2003; Kroner et al., 2009; Bu et al., 2011). Thus, the effect of PD-1 inhibition on AD pathogenesis needs to be further investigated also for its potential activating effect on inflammation mechanisms with a negative role in AD.


    . PD-1 blockade attenuates immunosuppressive myeloid cells due to inhibition of CD47/SIRPα axis in HPV negative head and neck squamous cell carcinoma. Oncotarget. 2015 Dec 8;6(39):42067-80. PubMed.

    . Critical role of the programmed death-1 (PD-1) pathway in regulation of experimental autoimmune encephalomyelitis. J Exp Med. 2003 Jul 7;198(1):71-8. PubMed.

    . Accelerated course of experimental autoimmune encephalomyelitis in PD-1-deficient central nervous system myelin mutants. Am J Pathol. 2009 Jun;174(6):2290-9. Epub 2009 May 14 PubMed.

    . Impairment of the programmed cell death-1 pathway increases atherosclerotic lesion development and inflammation. Arterioscler Thromb Vasc Biol. 2011 May;31(5):1100-7. Epub 2011 Mar 10 PubMed.

  2. Michal Schwartz’s lab recently generated convincing evidence that impaired immune responses contribute to Alzheimer’s disease-like pathogenesis in animal models. Following their recent publication on Treg cells (Baruch et al., 2015), in the current paper the authors demonstrate a key role for the inhibitory PD-1 signaling on disease pathogenesis in APP-PS1 transgenic mice. Interfering with such signaling allowed macrophage infiltration in brains, lowered plaque load, and ameliorated cognitive performance, in an IFN-γ dependent manner.

    The results are intriguing and of high clinical relevance. Supporting the notion that the adaptive immune system plays a regulatory and/or inhibitory role in the development of the disease, we have recently shown that ablation of lymphocytes from the APP-PS1-dE9 mouse model via cross-breeding or chimerism results in lower plaque load and higher microglial phagocytosis (Späni et al., 2015). The exact mechanisms of the therapeutic effect described by Baruch et al. remain to be elucidated.

    We propose the following points for discussion:

    1. PD-1 signaling was shown to promote Treg induction, proliferation and regulatory function (Francisco et al., 2009; Park et al., 2015; Punkosdy et al., 2011). It would be therefore interesting to know whether the effect of PD-1 blockade was mediated via Treg cells. A further detailed analysis of cell expression of PD-1 and immunophenotyping of treated mice would help clarify this point.

    2. In previous work, transient ablation of Treg cells allowed both myeloid and T cell entry into the brain. In the present paper, while myeloid infiltration to the CNS could be clearly demonstrated, there is no analysis of the lymphocyte compartment in brain. PD-1 treatment could have impacted the migration of T (reg) cells to amyloid burdened sites. On the other hand, blocking PD-1 signaling might have changed the phenotype of resident intracerebral T cells which, as we have recently shown, are mostly low IFN-γ producers (Ferretti et al., 2016).

    3. It will be extremely important to assess whether the therapeutic effects are mediated by antigen-specific, i.e. CD4-Aβ-specific, T cells.

    Even though the current results have provided a strong proof of principle, further evidence will be required for clinical translation of these findings. In particular, one should consider the following:

    1. There is very little, and conflicting, evidence on PD-1 alteration in AD. Saresella et al. have demonstrated that PD-1 expression is not altered in PBMCs from AD patients but is specifically increased on Treg cells. The same group went on to show that there is an impairment of PD-1 signaling in CD4-Aβ-specific T cells (Saresella et al., 2010; Saresella et al., 2012). 

    2. The authors have shown that PD-1 blockade results in a systemic IFN-γ immune response, which mediated the beneficial effects of the treatment. Modulating IFN-γ expression appears therefore as an attractive therapeutic target for treating AD. We note, however, that increased IFN-γ systemic levels are linked to inflammatory conditions; indeed, immune-related adverse events (including colitis, rash and pneumonitis) have been described in patients under PD-1 inhibitor treatment (see Luke and Ott, 2015, for a recent review). Such adverse events might be exacerbated in aged patients.

    3. Most manipulations of the adaptive immune system published so far have been performed on the 5xFAD, the APP-PS1, or the APP-PS1-dE9 mice, all carrying mutated presenilin genes. Since presenilin has been shown to regulate T cell development and function, some effects (such as the altered Treg and PD-1 signaling) might be model-specific. Similar results should be replicated in APP-only tg lines; in this regard, the ArcAβ model (Knobloch et al., 2007) has proven in our hands especially useful since it harbors robust CNS infiltration of T cells (Ferretti et al., 2016).

    To conclude, the work adds important insight into the complexity of the immunology of neurodegeneration, well beyond the dichotomy M1-M2, or inflammatory-anti-inflammatory. As a field, we need a better understanding of the complex interplay between innate and adaptive responses, including both effector and regulatory functions. 


    . Breaking immune tolerance by targeting Foxp3(+) regulatory T cells mitigates Alzheimer's disease pathology. Nat Commun. 2015 Aug 18;6:7967. PubMed.

    . Reduced β-amyloid pathology in an APP transgenic mouse model of Alzheimer's disease lacking functional B and T cells. Acta Neuropathol Commun. 2015 Nov 11;3:71. PubMed.

    . PD-L1 regulates the development, maintenance, and function of induced regulatory T cells. J Exp Med. 2009 Dec 21;206(13):3015-29. Epub 2009 Dec 14 PubMed.

    . PD-1 upregulated on regulatory T cells during chronic virus infection enhances the suppression of CD8+ T cell immune response via the interaction with PD-L1 expressed on CD8+ T cells. J Immunol. 2015 Jun 15;194(12):5801-11. Epub 2015 May 1 PubMed.

    . Regulatory T-cell expansion during chronic viral infection is dependent on endogenous retroviral superantigens. Proc Natl Acad Sci U S A. 2011 Mar 1;108(9):3677-82. Epub 2011 Feb 14 PubMed.

    . PD1 negative and PD1 positive CD4+ T regulatory cells in mild cognitive impairment and Alzheimer's disease. J Alzheimers Dis. 2010;21(3):927-38. PubMed.

    . A potential role for the PD1/PD-L1 pathway in the neuroinflammation of Alzheimer's disease. Neurobiol Aging. 2012 Mar;33(3):624.e11-22. Epub 2011 Apr 22 PubMed.

    . PD-1 pathway inhibitors: the next generation of immunotherapy for advanced melanoma. Oncotarget. 2015 Feb 28;6(6):3479-92. PubMed.

    . Intracellular Abeta and cognitive deficits precede beta-amyloid deposition in transgenic arcAbeta mice. Neurobiol Aging. 2007 Sep;28(9):1297-306. Epub 2006 Jul 31 PubMed.

    . T-cell brain infiltration and immature antigen-presenting cells in transgenic models of Alzheimer's disease-like cerebral amyloidosis. Brain Behav Immun. 2016 May;54:211-25. Epub 2016 Feb 9 PubMed.

  3. This is an excellent study by Dr. Schwartz's group. It conclusively demonstrates that systemic immunoneutralization of PD-1 in T lymphocytes would trigger IFN-γ dependent increase in the CD45highCD11b+ (monocyte-derived macrophage) population in the brain, which leads to clearance of cerebral amyloid in two separate transgenic mouse models of AD.

    The authors also demonstrate improved memory following repeated, but not with single, treatment of anti-PD1 antibody. This study well aligns with recent reports on PD1 inhibition and induction of anti-tumor immunity (Lesokhin et al., 2015; Peng W et al., 2012). It clearly appears that PD1 inhibition induces IFN-γ response, which is responsible for recruitment of monocytes from the periphery into the CNS. However, it would be interesting to see if PD1 blockage would prime CD45lowCD11b+ CNS resident microglial population to become phagocytic and clear amyloid. This question is prompted because, in Figure 1C, PD1 neutralization leads to notable reduction in CD45lowCD11b+ population (from 82.3 percent to 74.2 percent after day 7 or 71.2 percent after day 14, which is about an 8-10 percent decrease in microglial population).

    Does that mean the resident microglial population is decreasing, or are they starting to overexpress CD45 because of PD1 inhibition and IFN-γ upregulation? If PD1 suppression also has an effect on CNS resident microglial population, then it would be interesting to test if they, too, become phagocytic cells in clearing amyloid.

    The second point is that previous studies from our group (Bhaskar et al., 2010; Maphis et al., 2015) and Sue Griffin’s group (Li et al., 2003; Sheng et al., 2000) have suggested that overactive microglia may induce collateral damage, increasing tau phosphorylation and reduced synaptophysin levels by activation of the neuronal IL-1R-p38 MAPK pathway. It would be interesting to see how long CD45highCD11b+ cells are maintained after PD1 blockage and whether or not their levels come down once the amyloid is cleared from the brain. Finally, some of the genes (supplementary fig 2 a), specifically, microglia/monocyte specific gene Cx3cr1, show increased expression in CD45high population after day 14 of PD1 inhibition. Would this suggest that the CD45high peripherally-derived monocytes acquire microglial phenotype (increased CX3CR1) and therefore become more responsive to neuronal derived CX3CL1?

    These are some of the important questions this study opens up to our AD community to explore and understand. While it is obvious that PD1 inhibition and upregulation of IFN-γ is well-established, it would have been nice to see systemic immune-neutralization with anti-PD1 antibody indeed reduce the levels of PD1 on T cells.


    . On being less tolerant: enhanced cancer immunosurveillance enabled by targeting checkpoints and agonists of T cell activation. Sci Transl Med. 2015 Mar 25;7(280):280sr1. PubMed.

    . PD-1 blockade enhances T-cell migration to tumors by elevating IFN-γ inducible chemokines. Cancer Res. 2012 Oct 15;72(20):5209-18. Epub 2012 Aug 20 PubMed.

    . Regulation of tau pathology by the microglial fractalkine receptor. Neuron. 2010 Oct 6;68(1):19-31. PubMed.

    . Reactive microglia drive tau pathology and contribute to the spreading of pathological tau in the brain. Brain. 2015 Jun;138(Pt 6):1738-55. Epub 2015 Mar 31 PubMed.

    . Interleukin-1 mediates pathological effects of microglia on tau phosphorylation and on synaptophysin synthesis in cortical neurons through a p38-MAPK pathway. J Neurosci. 2003 Mar 1;23(5):1605-11. PubMed.

    . Interleukin-1 promotes expression and phosphorylation of neurofilament and tau proteins in vivo. Exp Neurol. 2000 Jun;163(2):388-91. PubMed.

  4. This work may generate a breakthrough in the field. However, I have one concern, i.e., the use of APP/PS1 mice as a model of AD. APP and PS1 contain 1 and 9 transmembrane domains, respectively. Overexpression of these proteins is likely to induce a non-specific ER stress, resulting in unfolded protein response (UPR), transcriptional suppression via XBP-1 rearrangement, IP3 receptor activation, i.e., intraneuronal calcium elevation, etc.

    Therefore, many of the molecular processes occurring in these brains could be artifacts. This is why we generated single APP knock-in mice (Saito et al., 2014). The mice are available and are now being used by more than 140 laboratories all over the world.

    Our collaborators will soon clarify whether the work by Schwartz's lab is valid and relevant.


    . Single App knock-in mouse models of Alzheimer's disease. Nat Neurosci. 2014 May;17(5):661-3. Epub 2014 Apr 13 PubMed.

Make a Comment

To make a comment you must login or register.


News Citations

  1. Does Peripheral Immune Activity Tame Alzheimer’s Disease?
  2. Could Adaptive Immunity Set the Brakes on Amyloid?

Research Models Citations

  1. 5xFAD (B6SJL)
  2. APPswe/PSEN1dE9 (line 85)

Paper Citations

  1. . Targeting the PD-1 pathway: a promising future for the treatment of melanoma. Arch Dermatol Res. 2014 Aug;306(6):511-9. Epub 2014 Mar 11 PubMed.
  2. . Critical role of the programmed death-1 (PD-1) pathway in regulation of experimental autoimmune encephalomyelitis. J Exp Med. 2003 Jul 7;198(1):71-8. PubMed.

Further Reading


  1. . Regulation of Neuroinflammation through Programed Death-1/Programed Death Ligand Signaling in Neurological Disorders. Front Cell Neurosci. 2014;8:271. Epub 2014 Sep 3 PubMed.
  2. . IFN-γ Production by amyloid β-specific Th1 cells promotes microglial activation and increases plaque burden in a mouse model of Alzheimer's disease. J Immunol. 2013 Mar 1;190(5):2241-51. PubMed.
  3. . Respiratory infection promotes T cell infiltration and amyloid-β deposition in APP/PS1 mice. Neurobiol Aging. 2014 Jan;35(1):109-21. PubMed.
  4. . PD1 negative and PD1 positive CD4+ T regulatory cells in mild cognitive impairment and Alzheimer's disease. J Alzheimers Dis. 2010;21(3):927-38. PubMed.

Primary Papers

  1. . PD-1 immune checkpoint blockade reduces pathology and improves memory in mouse models of Alzheimer's disease. Nat Med. 2016 Feb;22(2):135-7. Epub 2016 Jan 18 PubMed.