Elsevier

Neurobiology of Aging

Volume 59, November 2017, Pages 197-209
Neurobiology of Aging

Regular article
Antibody recognizing 4-sulfated chondroitin sulfate proteoglycans restores memory in tauopathy-induced neurodegeneration

https://doi.org/10.1016/j.neurobiolaging.2017.08.002Get rights and content

Abstract

Chondroitin sulfate proteoglycans (CSPGs) are the main active component of perineuronal nets (PNNs). Digestion of the glycosaminoglycan chains of CSPGs with chondroitinase ABC or transgenic attenuation of PNNs leads to prolongation of object recognition memory and activation of various forms of plasticity in the adult central nervous system. The inhibitory properties of the CSPGs depend on the pattern of sulfation of their glycosaminoglycans, with chondroitin 4-sulfate (C4S) being the most inhibitory form. In this study, we tested a number of candidates for functional blocking of C4S, leading to selection of an antibody, Cat316, which specifically recognizes C4S and blocks its inhibitory effects on axon growth. It also partly blocks binding of semaphorin 3A to PNNs and attenuates PNN formation. We asked whether injection of Cat316 into the perirhinal cortex would have the same effects on memory as chondroitinase ABC treatment. We found that masking C4S with the Cat316 antibody extended long-term object recognition memory in normal wild-type mice to 24 hours, similarly to chondroitinase or transgenic PNN attenuation. We then tested Cat316 for restoration of memory in a neurodegeneration model. Mice expressing tau with the P301S mutation showed profound loss of object recognition memory at 4 months of age. Injection of Cat316 into the perirhinal cortex normalized object recognition at 3 hours in P301S mice. These data indicate that Cat316 binding to C4S in the extracellular matrix can restore plasticity and memory in the same way as chondroitinase ABC digestion. Our results suggest that antibodies to C4S could be a useful therapeutic to restore memory function in neurodegenerative disorders.

Introduction

Memory loss is a largely unmet medical challenge. Halting brain degeneration would be the most effective form of treatment, but it has yet to be achieved. An alternative is to enable the brain to function despite the degenerative damage. Plasticity enables the brain to bypass damage, but it is limited in adults by inhibitory proteoglycans in the brain extracellular matrix (ECM) situated in perineuronal nets (PNNs), which surround the key regulatory neurons.

PNNs are dense pericellular ECM structures with a structure similar to cartilage that develop at the time of closure of critical periods for plasticity. PNNs are a highly organized complex, mainly composed of hyaluronan, chondroitin sulfate proteoglycans (CSPGs), link proteins, and tenascin-R. The form of PNN most studied is those surrounding parvalbumin-positive (PV) interneurons recognized by the lectin Wisteria floribunda (WFA), although other forms of PNN exist around pyramidal and other neurons (Matthews et al., 2002). In the cortex, PNNs surround PV interneurons which are known to influence plasticity and excitability, and PNNs are involved in the termination of plasticity at the end of critical periods (Fawcett, 2015, Pizzorusso et al., 2002, Sugiyama et al., 2009). PNNs are also involved in memory. Depletion of chondroitin sulfate (CS) with the CS-digesting enzyme, chondroitinase ABC (ChABC) prolonged object recognition (OR) memory in normal mice (Romberg et al., 2013), restored lost memory in a neurodegeneration model in tauopathy mice and Abeta transgenic mice (Vegh et al., 2014, Yang et al., 2015), and affected fear memory extinction (Gogolla et al., 2009), demonstrating that CSPGs are the key molecules in regulating this form of memory and that the CS glycosaminoglycan (GAG) chains digested by ChABC are their effectors. Animals deficient in the PNN component Crtl1 link protein have very attenuated PNNs and show the same changes in memory and plasticity as animals treated with ChABC despite having unaltered CSPGs in the central nervous system (CNS), demonstrating that it is CSPGs in PNNs that are controllers of plasticity and memory (Carulli et al., 2010, Romberg et al., 2013). However, in order to reactivate plasticity, ChABC has to be injected directly into the brain parenchyma where it digests a region of up to 1 mm across, and the effect lasts for around 3 weeks. ChABC is therefore not a practicable treatment for memory restoration in neurodegenerative diseases. This study investigates antibody treatment as an alternative.

CSPGs and PNNs are implicated in neurodegenerative conditions in various ways. CSPGs are abundant in amyloid and neurofibrillary tangles in Alzheimer's disease (AD; reviewed in DeWitt et al., 1993, Kwok et al., 2011) and differentially sulfated CS such as chondroitin 4-sulfate (C4S), chondroitin 6-sulfate (C6S), and unsulfated chondroitin were detected (DeWitt et al., 1993). A reduction of PNNs in brains affected by AD was shown by staining with lectin WFA (Baig et al., 2005, Kobayashi et al., 1989), although many markers were not affected in AD patients and animal models of AD (Morawski et al., 2010b, Morawski et al., 2012b). PV interneurons are largely unaffected by neurofibrillary tangles (Baig et al., 2005, Bruckner et al., 1999, Hartig et al., 2001, Morawski et al., 2010a, Morawski et al., 2012b). This suggests that PNNs may be protective, and there is evidence for regulation of ion homeostasis (Suttkus et al., 2014) and neuroprotection against oxidative stress (Cabungcal et al., 2013, Suttkus et al., 2016). Treatment with chondroitinase does not affect hyperphosphorylated or filamentous tau (Yang et al., 2015), but an effect on Abeta quantity has been reported (Howell et al., 2015).

The functional effect and binding properties of CSPGs are heavily dependent on the patterns of sulfation of their GAG chains (Lin et al., 2011, Malavaki et al., 2008, Miyata and Kitagawa, 2015). The sugars in the GAG chain can be sulfated in the 4, 6, 2 + 6, and 4 + 6 positions on the CSs (C4S, C6S, C2,6S, and C4,6S, also known as CS-A, CS-C, CS-D, and CS-E), respectively, and different patterns of sulfation give motifs that can provide specific binding sites (Mikami and Kitagawa, 2013, Xu and Esko, 2014, Yabuno et al., 2015). C4S is inhibitory to neurite growth while C6S is more permissive to neurite growth and to plasticity (Lin et al., 2011, Miyata et al., 2012, Wang et al., 2008). Thus, mice with a knockout of chondroitin 6-sulfotransferase 1 showed worse regeneration and less plasticity than wild-type (WT) animals (Lin et al., 2011) while chondroitin 6-sulfotransferase 1 overexpressing mice demonstrated enhanced plasticity into adulthood (Miyata et al., 2012). During embryonic development C6S predominates, but at the end of the critical period in the PNN fraction only 4.9% of CS-GAG is 6-sulfated while 80.6 % is the inhibitory 4-sulfated form (Carulli et al., 2010, Deepa et al., 2006). We have recently demonstrated that this trend continues into aging when memory deficits start to appear. C6S was reduced almost to zero in aged rats (12 and 18 months old) when compared to young animals (Foscarin et al., 2017). Moreover, the sulfation pattern of PNN CS-GAGs (which make up 2% of total brain CS-GAG) is different from that of the general brain matrix, giving the PNN specific binding properties (Deepa et al., 2006). Thus, semaphorin 3A (Sema3A) and OTX2 bind specifically to PNNs but not to the general matrix (Beurdeley et al., 2012, Dick et al., 2013, Miyata et al., 2012, Vo et al., 2013), where they are positioned to affect synapse dynamics and PV cell maturation. Both these molecules bind most strongly to 4,6 disulfated CS-GAG. The major CSPG core protein of PNNs is aggrecan, which is necessary for PNN formation (Carulli et al., 2016, Dick et al., 2013, Giamanco et al., 2010, Morawski et al., 2012a) and contains specific glycan modifications in its linkage region (Yabuno et al., 2015). The CS-GAGs carried by aggrecan are probably the key component of PNNs that regulate memory and plasticity, because animals lacking aggrecan in the CNS show the same memory prolongation phenotype as animals with attenuated PNNs due to Crtl1 knockout or ChABC treatment (unpublished observations) (Romberg et al., 2013). Our hypothesis therefore was that masking C4S on aggrecan would make PNNs less inhibitory and would therefore have a similar effect as ChABC in restoration of memory to brains damaged by neurodegeneration or in prolonging memory in normal animals.

Seeking potential treatments for memory restoration, we have tested an inhibitory proteoglycan neutralizing antibody. In this manuscript, we have identified Cat316 as a C4S binding antibody that can modulate the inhibitory properties of the glycan, reduce Sema3A binding, prolong OR memory in normal animals, and restore lost memory in tauopathy animals.

Section snippets

Screening of PNN-blocking antibodies using PNN-HEK cells

PNN-human embryonic kidney (HEK) cells (∼12,000 cells) were plated on poly-L-lysine–coated coverslips for 2 hours and cultured in low-serum medium (0.5 % fetal bovine serum in Dulbecco's modified eagle medium [DMEM]) in the presence of 5% of CO2. Detached cells were then removed with a rinse in fresh warm medium. Potential PNN-blocking antibodies (Cat316 is from Millipore; HAPLN1 is from R&D; 6B4 and 7B7 are from CosmoBio) were diluted into 5 μg/mL in cultured medium before incubation with the

Results

The aim of our experiments was to select and evaluate an antibody that can reactivate plasticity and restore memory through blocking inhibitory proteoglycans in PNNs. We tested several candidates for their ability to block inhibition by PNN proteoglycans, from which we selected Cat316. The ability of Cat316 to bind to proteoglycans and PNNs and to compete with Sema3A binding to PNNs was then evaluated in more detail. We then asked whether Cat316 has the same effects on memory prolongation and

Discussion

PNNs have been implicated in memory through several types of experiment. In OR memory, transgenic attenuation of PNNs or their digestion with ChABC leads to memory prolongation (Romberg et al., 2013). In fear memory, ChABC digestion in the amygdala restores memory erasure (a form of learning) to the infantile pattern (Gogolla et al., 2009). Memory stimulated by environmental enrichment depends on the formation of new inhibitory synapses on PV neurons, and this synapse formation is facilitated

Disclosure statement

JWF is a paid consultant for Acorda Therapeutics Inc, which is involved in the commercial development of chondroitinase. The other authors declare no other competing financial interests.

Acknowledgements

This work was supported by the ERC advanced grant ECM Neuro (294502), by a fellowship to SY from Alzheimer's Research UK (ARUK-RF2016A-1), and by the NIHR Cambridge Biomedical Research Centre. The authors are grateful to Dr Michel Goedert (MRC Laboratory of Molecular Biology, Cambridge, UK) for providing the human P301S tau transgenic mice.

References (52)

  • M. Morawski et al.

    Aggrecan: beyond cartilage and into the brain

    Int. J. Biochem. Cell Biol.

    (2012)
  • M. Morawski et al.

    Neurons associated with aggrecan-based perineuronal nets are protected against tau pathology in subcortical regions in Alzheimer's disease

    Neuroscience

    (2010)
  • M. Morawski et al.

    Perineuronal nets are largely unaffected in Alzheimer model Tg2576 mice

    Neurobiol. Aging

    (2010)
  • T. Vo et al.

    The chemorepulsive axon guidance protein semaphorin3A is a constituent of perineuronal nets in the adult rodent brain

    Mol. Cell Neurosci.

    (2013)
  • S. Yang et al.

    Perineuronal net digestion with chondroitinase restores memory in mice with tau pathology

    Exp. Neurol.

    (2015)
  • B. Allen et al.

    Abundant tau filaments and nonapoptotic neurodegeneration in transgenic mice expressing human P301S tau protein

    J. Neurosci.

    (2002)
  • S. Baig et al.

    Loss of perineuronal net N-acetylgalactosamine in Alzheimer's disease

    Acta Neuropathol.

    (2005)
  • C. Bernard et al.

    Otx2-PNN Interaction to regulate cortical plasticity

    Neural Plast.

    (2016)
  • M. Beurdeley et al.

    Otx2 binding to perineuronal nets persistently regulates plasticity in the mature visual cortex

    J. Neurosci.

    (2012)
  • J.H. Cabungcal et al.

    Perineuronal nets protect fast-spiking interneurons against oxidative stress

    Proc. Natl. Acad. Sci. U. S. A.

    (2013)
  • D. Carulli et al.

    Perineuronal nets and CNS plasticity and repair

    Neural Plast.

    (2016)
  • D. Carulli et al.

    Animals lacking link protein have attenuated perineuronal nets and persistent plasticity

    Brain

    (2010)
  • F. de Winter et al.

    The chemorepulsive protein semaphorin 3A and perineuronal net-mediated plasticity

    Neural Plast.

    (2016)
  • F. Donato et al.

    Parvalbumin-expressing basket-cell network plasticity induced by experience regulates adult learning

    Nature

    (2013)
  • S. Foscarin et al.

    Brain Ageing Changes Proteoglycan Sulfation, Rendering Perineuronal Nets More Inhibitory

    (2017)
  • N. Gogolla et al.

    Perineuronal nets protect fear memories from erasure

    Science

    (2009)
  • Cited by (40)

    • Oral treatment of 4-methylumbelliferone reduced perineuronal nets and improved recognition memory in mice

      2022, Brain Research Bulletin
      Citation Excerpt :

      Several proof-of-concept approaches have been developed to manipulate or remove PNNs to increase the neuroplasticity, memory and CNS repair. These include enzymatic degradation of ECM molecules (Romberg et al., 2013; Kwok et al., 2008; Howell and Gottschall, 2012), antibodies blocking PNN inhibitory action (Yang et al., 2017), or genetic modifications such as knockout models with gene deletion of various PNN components (Romberg et al., 2013; Carulli et al., 2010). Genetic attenuation of PNNs through specific knock down of link protein hapln1 in the CNS prolongs memory for familiar objects.

    • The microglial P2Y<inf>6</inf> receptor mediates neuronal loss and memory deficits in neurodegeneration

      2021, Cell Reports
      Citation Excerpt :

      P301S mice lose cortical neurons (Hampton et al., 2010; Yang et al., 2015 & 2017), so we first tested whether genetic ablation of P2ry6 is sufficient to prevent the neuronal loss observed in the P301S mouse model at seven months of age. We found about 15% neuronal loss in the perirhinal cortex of P301S mice compared to WT mice (Figures 3A and 3B), similar to the loss previously reported (Yang et al., 2015, 2017). However, no significant neuronal loss was found in the perirhinal cortex of double-transgenic (P2r6y−/−:P301S+/+) mice compared to WT (P2ry6+/+:P301S−/−) mice (Figures 3A and 3B).

    • Potential involvement of perineuronal nets in brain aging: An anatomical point of view

      2021, Factors Affecting Neurological Aging: Genetics, Neurology, Behavior, and Diet
    • Sex and region-specific effects of high fat diet on PNNs in obesity susceptible rats

      2020, Physiology and Behavior
      Citation Excerpt :

      There is also evidence that diet-induced reductions in PNN number can affect cognition. For example, the antibody, Cat316 attenuates PNN formation when injected into the perirhinal cortex and enhances long-term object recognition memory in mice [58]. Additionally, Reichelt et al., 2019 found that a high-fat high-sugar diet given intermittently during adolescence (PND 28-57) impaired social recognition memory, and the degree of impairment corresponded to degree of reduction in the number of PNNs within the IL-PFC [59].

    • Chondroitin 4-O-sulfation regulates hippocampal perineuronal nets and social memory

      2023, Proceedings of the National Academy of Sciences of the United States of America
    View all citing articles on Scopus
    View full text