Microglia Prune Synapses in a Subtype of Frontotemporal Dementia

Like overzealous gardeners, microglia have been found to cut back synapses in mouse models of Alzheimer’s and Huntington’s disease. Now, researchers led by Eric Huang at the University of California, San Francisco, report that a similar process may occur in mice that lack progranulin. Deficiency of this gene causes as many as 20 percent of familial FTD cases. In progranulin knockout mice, microglial activity ramped up. These immune cells overproduced complement proteins, which tagged synapses for destruction, while also dialing up lysosomal digestion. As a result, these microglia devoured more synapses than did microglia in control brains. “This study provides additional evidence that aberrant activation of microglia can indeed drive neurodegeneration,” Huang told Alzforum.

The findings may be applicable to human disease, the authors note, because autopsy data from people with this form of FTD revealed excessive complement-labeled synapses in the frontal cortex. In addition, increased cerebrospinal fluid levels of the complement proteins C1q and C3 correlated with cognitive decline in these patients, Huang and colleagues found. The finding hints that these proteins could have potential as biomarkers of disease progression, Huang noted.

Jonathan Kipnis at the University of Virginia, Charlottesville, agreed the data add to an emerging picture that puts microglia at the center of synaptic loss in multiple disorders. “I think it is clear now that microglia and their phagocytic activity could be real [therapeutic] targets in neurodevelopmental and also in late-stage neurodegenerative diseases,” he wrote to Alzforum.

No Complement, No Problem.

Year-old mice lacking progranulin (middle) lose synapses (green) in the thalamus, but if they also lack complement protein C1q (right), they preserve as many synapses as controls (left). [Courtesy of Cell, Lui et al.]

In the past decade, researchers including Beth Stevens at Boston Children’s Hospital have developed the idea that microglia and astrocytes are responsible for the synaptic pruning that occurs as the mammalian brain develops (see Dec 2000 news; Nov 2007 conference news; Mar 2015 conference news). Recently, multiple groups extended these findings to disease states, reporting that this pruning pathway reactivates in some neurodegenerative disorders and even in conditions such as obesity (see Aug 2013 conference news; Dec 2014 news; Nov 2015 conference news). 

Huang and colleagues wondered if the same thing might happen in progranulin-deficient animals. After all, microglia produce most of the progranulin in the brain (see Zhang et al., 2014). Previous studies by the authors and others had seen neuroinflammation and microglial activation in the absence of this protein, suggesting progranulin normally suppresses these cells (see Yin et al., 2010; Mar 2011 news; Oct 2012 news). However, the mechanisms were unclear.

To pin down what progranulin might be doing, first author Hansen Lui, now at the University of California, Berkeley, compared gene expression in the cerebral cortex, hippocampus, and cerebellum of progranulin knockout mice (GRN-/-) to that of wild-type mice at various ages. Progranulin knockouts develop neuroinflammation and lose inhibitory synapses as they age. Their neuronal circuits become overly excitable, and they groom themselves excessively. Compulsive behaviors mark FTD as well. GRN-/- mice die about 200 days sooner than wild-types. Lui and colleagues found that with age, GRN-/- microglia overexpressed genes involved in innate immunity and lysosomal digestion. The data implied that these two processes played a key role in pathogenesis.

To confirm this, the authors examined mouse brains and found that at 16 months of age, GRN knockouts had four times the normal amount of microglia. These cells contained enlarged lysosomes, suggesting digestion of cellular waste might be perturbed. In keeping with this, GRN-/- microglia from older mice ingested and processed a fluorescent dye more quickly in culture than wild-type microglia did. Likewise, in co-cultures with neurons, the GRN-/- microglia gobbled more synapses, accumulating about 60 percent more synaptic markers than did normal microglia.

In addition to the lysosomal changes, microglia from frontal cortex, hippocampus, and other regions of older GRN-/- mice produced more complement than did microglia from wild-types. The change was most pronounced in the thalamus, where GRN knockout mice pumped out up to eight times more C1q and 100-fold more C3 at 18 months of age. All these regions are sites of pathology in FTD as well. The authors focused on the thalamus in subsequent experiments. They found that thalamic synapses sported increasing amounts of these complement proteins as the mice aged. Moreover, GRN knockout microglia were overeager in lysosomal processing of C3, a process that releases active fragments of this complement factor, including iC3b. GRN-/- microglia made four times as much iC3b as wild-types. This is likely due to faster lysosomal digestion in the knockouts, Huang said. The lysosomal defect might boost the innate immune system as well as stimulating digestion of synapses, he suggested. 

“The two phenotypes together create a perfect storm,” Huang told Alzforum.

Do these changes in complement and lysosomes affect synapse loss directly in vivo? To find out, the authors crossed C1q knockouts with GRN-/- mice. In contrast to the latter, which had lost a third of their synapses by 19 months of age, the double knockouts maintained normal synaptic density in the thalamus even past that age (see image above). The lack of complement also improved neuronal function and the overall health of the animals. Brain slices from double knockouts fired normally. They lived about 100 days longer than the GRN-/- mice and groomed no more often than wild-type animals.

Preliminary data tie these findings to FTD pathology. The authors examined autopsy samples from the frontal cortices of 19 patients with GRN mutations and found up to fourfold higher microglial density compared to controls, and profuse C1q at synapses. In cerebrospinal fluid, absolute levels of complement proteins did not differ between FTD patients and controls, however, the amount of C1q and C3 in patient CSF climbed as cognition faltered. If that relationship holds in larger samples, complement proteins could become biomarkers of disease progression or therapeutic benefit, Huang suggested.

The findings distinguish FTD from AD. Complement has been linked to synapse loss in Alzheimer’s disease (see Apr 2016 news), but the authors found a different pattern of neuroinflammation and biomarkers in the two disorders, suggesting distinct mechanisms might be at work. In AD brains, microglial density was similar to that in controls, and most of the immune cells clustered around plaques. In AD CSF, C1q levels were lower than in controls, and kept going down. Huang noted that complement proteins deposit in amyloid plaques, perhaps explaining their scarcity in CSF just as Aβ42 goes down in AD CSF. In addition, Huang suggested that complement factors in AD and progranulin-deficient FTD might come from different sources. In AD brains, microglia do not appear to overproduce complement, rather, it may come from astrocytes or the bloodstream, Huang speculated. In future work, he plans to look for elevated complement in patient CSF and brain tissue samples from other subtypes of FTD.

Many other questions remain. In GRN-/- mice, the authors found C1q equally distributed among excitatory and inhibitory synapses but, curiously, microglia only pruned the latter. Other signals present at the synapse might protect excitatory synapses, Huang suggested. Huang will examine conditional GRN-/- mice that lack the protein only in microglia or only in neurons, and he intends to test a C1q-blocking antibody from Annexon Biosciences, South San Francisco, to see if the treatment can preserve synapses and improve behavior. Other researchers noted that the links between lysosomal defects and the upregulation of complement proteins are unclear and deserve further investigation. Yoshinori Tanaka at the Tokyo Metropolitan Institute of Medical Science pointed out that C1q production also rises in some lysosomal storage diseases (see Ohmi et al., 2003). 

Huang also wants to follow up on the lysosomal phenotype and dissect how progranulin regulates lysosomal digestion. He suspects he will find links between progranulin-deficient forms of FTD and autoimmune disease. Enhanced processing of C3 by lysosomes in T cells has been found to worsen inflammation in people with autoimmune arthritis (see Liszewski et al., 2013). Moreover, FTD patients have an increased incidence of autoimmune disorders. Alzheimer’s disease also was recently reported to share genetic risk factors with some autoimmune conditions (see Apr 2016 news). 

Tim Sargeant at the South Australian Health & Medical Research Institute, Adelaide, pointed out that complete loss of progranulin causes the lysosomal storage disease neuronal ceroid lipofuscinosis. Loss of several other proteins produce similar lysosomal defects and neurodegeneration, he added. “It is therefore possible the effects observed in this paper are from a lysosomal defect that can be caused by deficits in a wide range of lysosomal proteins. These observations … show targeting of lysosomal function in late onset neurodegenerative disease could offer a feasible therapeutic strategy,” he wrote to Alzforum (see full comment below).—Madolyn Bowman Rogers

Comments

  1. This recent paper by Lui and colleagues provides valuable insight into how progranulin deficiency causes neurodegenerative disease. This study shows that progranulin deficiency causes activation of microglia and that the complement system is at least partly responsible for neurodegeneration in the complete gene knockout mouse model. How directly or uniquely progranulin is involved in this process, however, is still unclear. It is known that complete loss of progranulin results in a lysosomal storage disease (LSD) called neuronal ceroid lipofuscinosis, and various LSDs caused by deficiency of multiple other proteins, such as cathepsin D (Partanen et al., 2008), NPC1 (Yamada et al., 2001), glucocerebrosidase (Farfel-Becker et al., 2011), and beta-hexosaminidase (Sargeant et al., 2011) also result in relatively selective neurodegeneration, neuronal cell death, and neuroinflammation within the VPM/VPL of the murine thalamus (Sargeant 2016). It is therefore possible the effects observed in this paper result from a lysosomal defect that can be caused by deficits in a wide range of lysosomal proteins. These observations, along with other lines of evidence (such as lysosomal storage in FTD caused by CHMP2B mutation [Sep 2015 news; Clayton et al., 2015]) show targeting of lysosomal function in late-onset neurodegenerative disease could offer a feasible therapeutic strategy.

    References:

    Partanen S, Haapanen A, Kielar C, Pontikis C, Alexander N, Inkinen T, Saftig P, Gillingwater TH, Cooper JD, Tyynelä J. Synaptic changes in the thalamocortical system of cathepsin D-deficient mice: a model of human congenital neuronal ceroid-lipofuscinosis. J Neuropathol Exp Neurol. 2008 Jan;67(1):16-29. PubMed.

    Yamada A, Saji M, Ukita Y, Shinoda Y, Taniguchi M, Higaki K, Ninomiya H, Ohno K. Progressive neuronal loss in the ventral posterior lateral and medial nuclei of thalamus in Niemann-Pick disease type C mouse brain. Brain Dev. 2001 Aug;23(5):288-97. PubMed.

    Farfel-Becker T, Vitner EB, Pressey SN, Eilam R, Cooper JD, Futerman AH. Spatial and temporal correlation between neuron loss and neuroinflammation in a mouse model of neuronopathic Gaucher disease. Hum Mol Genet. 2011 Apr 1;20(7):1375-86. Epub 2011 Jan 20 PubMed.

    Sargeant TJ, Wang S, Bradley J, Smith NJ, Raha AA, McNair R, Ziegler RJ, Cheng SH, Cox TM, Cachón-González MB. Adeno-associated virus-mediated expression of β-hexosaminidase prevents neuronal loss in the Sandhoff mouse brain. Hum Mol Genet. 2011 Nov 15;20(22):4371-80. Epub 2011 Aug 18 PubMed.

    Sargeant TJ. Commentary: Possible involvement of lysosomal dysfunction in pathological changes of the brain in aged progranulin-deficient mice. Front Aging Neurosci. 2016;8:11. Epub 2016 Feb 3 PubMed.

    Clayton EL, Mizielinska S, Edgar JR, Nielsen TT, Marshall S, Norona FE, Robbins M, Damirji H, Holm IE, Johannsen P, Nielsen JE, Asante EA, Collinge J, FReJA consortium, Isaacs AM. Frontotemporal dementia caused by CHMP2B mutation is characterised by neuronal lysosomal storage pathology. Acta Neuropathol. 2015 Oct;130(4):511-23. Epub 2015 Sep 10 PubMed.

    View all comments by Timothy Sargeant

  2. In this paper, Dr. Lui and co-workers demonstrated that progranulin (PGRN) deficiency facilitated synaptic pruning resulting from increased C1qa production in cultured microglia and in mice.

    PGRN haploinsufficiency, resulting from a heterozygous mutation in the PGRN gene (GRN), causes frontotemporal lobar degeneration (FTLD) characterized by cytoplasmic inclusions containing TAR DNA binding protein 43 (Baker et al., 2006; Cruts et al., 2006). Patients with a homozygous mutation in GRN exhibit neuronal ceroid lipofuscinosis (NCL), a lysosomal storage disease (Smith et al., 2012).

    Previous studies indicated a role for PGRN in lysosomal function (Tanaka et al., 2013; Zhou et al., 2015). Aged, PGRN-deficient mice present excessive neuroinflammation in the ventral thalamus that is likely due to lysosomal defects in the brain because NCL model mice are particular vulnerable in this region of the brain (Tanaka et al., 2014; Kielar et al., 2007; Partanen et al., 2008; von Schantz et al., 2009). 

    In the present study, C1qa production and synaptic pruning increased in the lysosome-compromised, PGRN-deficient primary microglia. Moreover, these increases occurred in the ventral thalamus of aged and lysosome-defective PGRN-deficient mice, but not in younger ones. Since C1qa production also increases in mouse models of lysosomal storage diseases such as mucopolysaccharidoses I and IIIB (Ohmi et al., 2003), increased C1qa production and synaptic pruning activity in PGRN-deficient microglia might result from lysosomal defects. Studying synaptic pruning and C1qa production in the ventral thalamus of other NCL model mice might be useful to elucidate how PGRN regulates synaptic pruning.

    Importantly, loss of C1qa could partially suppress pathological changes in PGRN-deficient mice. This suggests that therapies targeting the complement pathway in microglia might improve neurodegenerative diseases resulting from PGRN deficiency or lysosomal defects, though the relationship between the latter and increased C1qa production remains to be elucidated. Further investigation into the regulation of the complement pathway in microglia might lead to new strategies to mitigate neurodegeneration.

    References:

    Baker M, Mackenzie IR, Pickering-Brown SM, Gass J, Rademakers R, Lindholm C, Snowden J, Adamson J, Sadovnick AD, Rollinson S, Cannon A, Dwosh E, Neary D, Melquist S, Richardson A, Dickson D, Berger Z, Eriksen J, Robinson T, Zehr C, Dickey CA, Crook R, McGowan E, Mann D, Boeve B, Feldman H, Hutton M. Mutations in progranulin cause tau-negative frontotemporal dementia linked to chromosome 17. Nature. 2006 Aug 24;442(7105):916-9. PubMed.

    Cruts M, Gijselinck I, van der Zee J, Engelborghs S, Wils H, Pirici D, Rademakers R, Vandenberghe R, Dermaut B, Martin JJ, van Duijn C, Peeters K, Sciot R, Santens P, De Pooter T, Mattheijssens M, Van den Broeck M, Cuijt I, Vennekens K, De Deyn PP, Kumar-Singh S, Van Broeckhoven C. Null mutations in progranulin cause ubiquitin-positive frontotemporal dementia linked to chromosome 17q21. Nature. 2006 Aug 24;442(7105):920-4. PubMed.

    Smith KR, Damiano J, Franceschetti S, Carpenter S, Canafoglia L, Morbin M, Rossi G, Pareyson D, Mole SE, Staropoli JF, Sims KB, Lewis J, Lin WL, Dickson DW, Dahl HH, Bahlo M, Berkovic SF. Strikingly different clinicopathological phenotypes determined by progranulin-mutation dosage. Am J Hum Genet. 2012 Jun 8;90(6):1102-7. PubMed.

    Tanaka Y, Matsuwaki T, Yamanouchi K, Nishihara M. Increased lysosomal biogenesis in activated microglia and exacerbated neuronal damage after traumatic brain injury in progranulin-deficient mice. Neuroscience. 2013 Oct 10;250:8-19. PubMed.

    Zhou X, Sun L, Bastos de Oliveira F, Qi X, Brown WJ, Smolka MB, Sun Y, Hu F. Prosaposin facilitates sortilin-independent lysosomal trafficking of progranulin. J Cell Biol. 2015 Sep 14;210(6):991-1002. PubMed.

    Tanaka Y, Chambers JK, Matsuwaki T, Yamanouchi K, Nishihara M. Possible involvement of lysosomal dysfunction in pathological changes of the brain in aged progranulin-deficient mice. Acta Neuropathol Commun. 2014 Jul 15;2:78. PubMed.

    Kielar C, Maddox L, Bible E, Pontikis CC, Macauley SL, Griffey MA, Wong M, Sands MS, Cooper JD. Successive neuron loss in the thalamus and cortex in a mouse model of infantile neuronal ceroid lipofuscinosis. Neurobiol Dis. 2007 Jan;25(1):150-62. Epub 2006 Oct 12 PubMed.

    von Schantz C, Kielar C, Hansen SN, Pontikis CC, Alexander NA, Kopra O, Jalanko A, Cooper JD. Progressive thalamocortical neuron loss in Cln5 deficient mice: Distinct effects in Finnish variant late infantile NCL. Neurobiol Dis. 2009 May;34(2):308-19. PubMed.

    Partanen S, Haapanen A, Kielar C, Pontikis C, Alexander N, Inkinen T, Saftig P, Gillingwater TH, Cooper JD, Tyynelä J. Synaptic changes in the thalamocortical system of cathepsin D-deficient mice: a model of human congenital neuronal ceroid-lipofuscinosis. J Neuropathol Exp Neurol. 2008 Jan;67(1):16-29. PubMed.

    Ohmi K, Greenberg DS, Rajavel KS, Ryazantsev S, Li HH, Neufeld EF. Activated microglia in cortex of mouse models of mucopolysaccharidoses I and IIIB. Proc Natl Acad Sci U S A. 2003 Feb 18;100(4):1902-7. Epub 2003 Feb 7 PubMed.

    View all comments by Yoshinori Tanaka

  3. This study is very novel as it demonstrates a role for progranulin in regulating lysosomal function and complement system activation in microglia. The authors further showed that complement activation leads to circuit-specific pruning of synapses by microglia. The data is consistent with previous findings that progranulin is essential for proper lysosomal function during aging and that progranulin loss leads to microglial activation and proinflammatory phenotypes. Microglial activation and lysosomal dysfunction have also been known to be implicated in FTLD caused by progranulin mutations.

    This study also suggests that complement protein levels in the CSF correlate with the decline on the MMSE in FTLD-GRN mutant carriers, indicating that CSF complement might be used as a biomarker for this disorder. Interestingly, this CSF change is not seen in AD patients, supporting disease specificity of complement activation in FTLD. Deleting the C1qa gene partially rescued microglial phenotypes in GRN-/- mice, indicating that complement inhibition could potentially be beneficial for FTLD patients with PGRN mutations.

    The connection between lysosomal dysfunction and complement activation seems to be unclear at present. The authors did show that GRN-/- microglia are more efficient in processing materials via the endolysosomal pathway and suggest that this might facilitate lysosomal processing and activation of the complement protein C3. However, the upregulation of C1qa and C3 also seem to be at the transcriptional level—how progranulin loss leads to transcriptional upregulation of complement factors is still unclear. Likewise, how progranulin regulates lysosomal function during aging is still a mystery.

    View all comments by Fenghua Hu

  4. This is a very interesting paper using sophisticated techniques to explore the potential role of progranulin in suppression of microglial activation. The association of genetic mutations leading to low levels of progranulin in some neurological diseases make this an important area to investigate. The authors present transcriptome analyses and immunohistochemical data that show brain region-specific increases in some complement components as well as increased microglial lysosomal size and divergent morphology with aging in progranulin deficit mice. Experiments with BSA-conjugated dye (DQ-BSA) were interpreted to indicate that the Grn-/- microglia are more active (i.e., ingest cargo more readily and more efficiently degrade it). A major and important conclusion of the paper is that Grn reduction leads to impaired regulation of the microglial response to injury (or aging) and thus enhanced synaptic pruning (preferentially of interneuron synapses) leading to hyperactivity in injured or aging animals.

    While an experimentally beautiful paper, the concluded link between greater C3 in the lysosomes and synaptic pruning as suggested by the authors in the results and discussion will require more mechanistic analysis, although indeed there is no need for the ascribed link between these two observations for the validity of their major observation. The authors try to connect intracellular cleavage of C3 in the microglia and subsequent secretion of the cleaved product iC3b with enhanced synaptic pruning. It should be clarified that the activation of complement through C1 (a macromolecular complex comprised of C1q, C1r, and Cls), ultimately leads to the cleavage of C3 to expose the C3 thioester enabling covalent attachment to the activating component, thereby “tagging” it for phagocytosis. If the proposed intracellularly cleaved C3 does not bind immediately to a protein (or sugar) moiety, that reactive thioester binds water and, even if then released from the cell, can no longer covalently bind a target (an extracellular neuronal synapse), tagging it for CR3-mediated elimination by the microglia. As a result, if C3 is cleaved within the lysosome (which can occur by enzymes other than the complement generated C3 cleaving enzyme, as the authors refer to) it will be inactive (in terms of tagging a synapse) by the time it reaches the extracellular space. Rather than the internally cleaved C3 as the mediator in synapse elimination, since C3 is known to be (and verified by the authors) upregulated in the injured brain, there should be plenty of extracellular C3 to become cleaved by the complement pathway at the synapse. Thus, an intriguing future investigation is to identify how the “tagged” synapse became altered to induce binding of C1 and thus activation of the system. In addition, it will be interesting to see if C3a, the smaller C3 cleavage fragment, is having an effect in this system.

    A second consideration in the mechanistic analysis (which is critical for designing a successful therapeutic intervention) is that much of the neuronal circuit alternations and behavioral aberrations could be due not solely to the synapse loss, but also to the generation of C5a and its proinflammatory activities, which would amplify the proinflammatory activation state of local microglia. This could be investigated by assessing synapse loss and behavioral alterations in C5aR-/- animals. The generation of C5a and its consequences certainly should be considered by the field while moving forward in any additional studies of complement activation in the brain, as selective intervention at various points in this pathway will likely be critical for a successful therapeutic.

    View all comments by Andrea Tenner

Make a Comment

To make a comment you must login or register.

References

News Citations

  1. Immune Proteins Play Role in Brain Development and Remodeling
  2. San Diego: MHC Class I and Complement—Holding Down Second Jobs in the Synapse
  3. Microglia Rely on Mixed Messages to Select Synapses for Destruction
  4. Curbing Innate Immunity Boosts Synapses, Cognition
  5. Neurons Cave When Astrocytes Heap on the Complements
  6. Microglia Control Synapse Number in Multiple Disease States
  7. Progranulin—Curbs Phagocyte Appetites, Protects Neurons?
  8. Microglial Progranulin Douses Neural Inflammation
  9. Paper Alert: Microglia Mediate Synaptic Loss in Early Alzheimer’s Disease
  10. New Genetic Method Connects Immune Genes to Alzheimer’s

Paper Citations

  1. Zhang Y, Chen K, Sloan SA, Bennett ML, Scholze AR, O'Keeffe S, Phatnani HP, Guarnieri P, Caneda C, Ruderisch N, Deng S, Liddelow SA, Zhang C, Daneman R, Maniatis T, Barres BA, Wu JQ. An RNA-sequencing transcriptome and splicing database of glia, neurons, and vascular cells of the cerebral cortex. J Neurosci. 2014 Sep 3;34(36):11929-47. PubMed.
  2. Yin F, Dumont M, Banerjee R, Ma Y, Li H, Lin MT, Beal MF, Nathan C, Thomas B, Ding A. Behavioral deficits and progressive neuropathology in progranulin-deficient mice: a mouse model of frontotemporal dementia. FASEB J. 2010 Dec;24(12):4639-47. PubMed.
  3. Ohmi K, Greenberg DS, Rajavel KS, Ryazantsev S, Li HH, Neufeld EF. Activated microglia in cortex of mouse models of mucopolysaccharidoses I and IIIB. Proc Natl Acad Sci U S A. 2003 Feb 18;100(4):1902-7. Epub 2003 Feb 7 PubMed.
  4. Liszewski MK, Kolev M, Le Friec G, Leung M, Bertram PG, Fara AF, Subias M, Pickering MC, Drouet C, Meri S, Arstila TP, Pekkarinen PT, Ma M, Cope A, Reinheckel T, Rodriguez de Cordoba S, Afzali B, Atkinson JP, Kemper C. Intracellular complement activation sustains T cell homeostasis and mediates effector differentiation. Immunity. 2013 Dec 12;39(6):1143-57. Epub 2013 Dec 5 PubMed.

Further Reading

Primary Papers

  1. Lui H, Zhang J, Makinson SR, Cahill MK, Kelley KW, Huang HY, Shang Y, Oldham MC, Martens LH, Gao F, Coppola G, Sloan SA, Hsieh CL, Kim CC, Bigio EH, Weintraub S, Mesulam MM, Rademakers R, Mackenzie IR, Seeley WW, Karydas A, Miller BL, Borroni B, Ghidoni R, Farese RV Jr, Paz JT, Barres BA, Huang EJ. Progranulin Deficiency Promotes Circuit-Specific Synaptic Pruning by Microglia via Complement Activation. Cell. 2016 May 5;165(4):921-35. Epub 2016 Apr 21 PubMed.

Annotate

To make an annotation you must Login or Register.