of 29
Wareham
et al. Molecular Neurodegeneration (2022) 17:23
https://doi.org/10.1186/s13024-022-00524-0
REVIEW
Solving neurodegeneration: common
mechanisms and strategies for new treatments
Lauren
K. Wareham
1
, Shane A. Liddelow
2
, Sally Temple
3
, Larry I. Benowitz
4
, Adriana Di Polo
5
, Cheryl Wellington
6
,
Jeffrey L. Goldberg
7
, Zhigang He
8
, Xin Duan
9
, Guojun Bu
10
, Albert A. Davis
11
, Karthik Shekhar
12
, Anna La
Torre
13
,
David C. Chan
14
, M.
Valeria Canto‑Soler
16
, John G. Flanagan
15
, Preeti Subramanian
18
, Sharyn Rossi
18
,
Thomas Brunner
17
, Diane E. Bovenkamp
18
and David J. Calkins
1*
Abstract
Across neurodegenerative diseases, common mechanisms may reveal novel therapeutic targets based on neuronal
protection, repair, or regeneration, independent of etiology or site of disease pathology. To address these mechanisms
and discuss emerging treatments, in April, 2021, Glaucoma Research Foundation, BrightFocus Foundation, and the
Melza M. and Frank Theodore Barr Foundation collaborated to bring together key opinion leaders and experts in the
field of neurodegenerative disease for a virtual meeting titled “Solving Neurodegeneration”. This “think
‑tank” style
meeting focused on uncovering common mechanistic roots of neurodegenerative disease and promising targets for
new treatments, catalyzed by the goal of finding new treatments for glaucoma, the world’s leading cause of irreversi‑
ble blindness and the common interest of the three hosting foundations. Glaucoma, which causes vision loss through
degeneration of the optic nerve, likely shares early cellular and molecular events with other neurodegenerative dis‑
eases of the central nervous system. Here we discuss major areas of mechanistic overlap between neurodegenerative
diseases of the central nervous system: neuroinflammation, bioenergetics and metabolism, genetic contributions, and
neurovascular interactions. We summarize important discussion points with emphasis on the research areas that are
most innovative and promising in the treatment of neurodegeneration yet require further development. The research
that is highlighted provides unique opportunities for collaboration that will lead to efforts in preventing neurodegen‑
eration and ultimately vision loss.
Keywords:
Neurodegeneration, Alzheimer’s Disease, Glaucoma, Parkinson’s Disease, Huntington’s Disease, Genetics,
Metabolic stress, Neuro
‑regeneration, Neuro
‑replacement, Neurovascular coupling, Biomarker, Cell‑replacement,
Detection, Glia, Imaging, Model Systems, Organoids
© The Author(s) 2022.
Open Access
This article is licensed under a Creative Commons Attribution 4.0 International License, which
permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the
original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or
other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line
to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory
regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this
licence, visit
http:// creat iveco mmons. org/ licen ses/ by/4. 0/
. The Creative Commons Public Domain Dedication waiver (
http:// creat iveco
mmons. org/ publi cdoma in/ zero/1. 0/
) applies to the data made available in this article, unless otherwise stated in a credit line to the data.
Background
A wide spectrum of neurodegenerative disorders affects
the central nervous system (CNS), causing a break
-
down in connectivity and communication between
neurons integral to sensory, motor, and cognitive pro
-
cesses including vision, hearing, movement, speech and
language, memory, and others. This breakdown in neu
-
ronal connection is characterized by the progressive
degradation of synapses and axons that lead to eventual
neuronal death. Cases of neurodegeneration and demen
-
tia worldwide are predicted to rise dramatically with the
aging population, posing a significant threat to global
healthcare systems [
1
,
2
]. Although neurodegenerative
diseases are highly complex and can be etiologically dis
-
tinct, uncovering commonalities in disease mechanisms
and pathologies may yield a deeper understanding of
the triggering events in neurodegeneration and generate
Open Access
*Correspondence: david.j.calkins@vumc.org
1
Department of Ophthalmology and
Visual Sciences, Vanderbilt Eye
Institute, Vanderbilt University Medical Center, Nashville, TN, USA
Full list of author information is available at the end of the article
Page 2 of 29
Wareham
et al. Molecular Neurodegeneration (2022) 17:23
opportunities for novel pan-neurodegenerative therapeu
-
tic avenues.
Main text
Etiological features of neurodegenerative disorders
Alzheimer’s disease and related dementias
The symptoms associated with neurodegenerative dis
-
ease are largely dependent on the CNS tissue affected,
which varies across diseases such as Alzheimer’s Disease
(AD), Huntington’s Disease (HD), Parkinson’s Disease
(PD), and Amyotrophic lateral sclerosis (ALS). Although
each neurodegenerative disease is distinct in terms of eti
-
ology, severity, and rate of progression, shared molecu
-
lar changes and mechanisms can be identified offering
potential avenues for research across multiple diseases.
Alzheimer’s Disease represents the most common form
of dementia, predominantly afflicting the aged popula
-
tion [
3
]. Over time, patients develop gradual but pro
-
gressive memory loss and cognitive decline associated
with the degeneration of neurons [
4
]. In AD, severity of
symptoms is correlated with pathophysiological events
caused by protein aggregations in the cerebral cortex
[
5
8
]. These have been shown histologically as the depo
-
sition of β-amyloid (Aβ) aggregated fibrils and plaques,
and neurofibrillary tangles containing hyperphosphoryl
-
ated Tau protein [
5
]. Amyloid precursor protein (APP)
can be cleaved to form varying lengths (from 38 to 43
amino-acids) of Aβ peptides [
9
]. Aβ monomers can bind
to one another to eventually form oligomers and insolu
-
ble plaques. The deposition and accumulation of Aβ oli
-
gomers is generally accepted as central to pathogenesis of
AD and the most toxic to neurons; however, other patho
-
logical events such as tau aggregation, as well as neuro
-
inflammation also play a major role and contribute to
synaptic loss and neurodegeneration [
3
].
While AD accounts for 60–80% of dementia cases, vas
-
cular cognitive impairment and dementia (VCID) are the
second leading cause of dementia [
10
]. Recent mount
-
ing evidence supports an underlying vascular element
in the pathophysiology of AD [
11
]; abnormal microvas
-
culature in AD patients is present post-mortem in the
brains of patients [
11
13
]. In fact, the role of cerebrovas
-
cular alterations in dementia-associated neurodegenera
-
tive diseases has been highlighted as a primary cause of
cognitive impairments and as a factor that contributes
directly to dementia associated with neurodegeneration
[
14
,
15
].
PD, the second most common form of neurodegen
-
erative disorder [
16
], is also characterized by progres
-
sive loss of neurons. Neurodegeneration in PD leads to
the impairment of basal ganglia in the brain, present
-
ing in the clinic as difficulty with motor-movement,
cognitive impairment, autonomic failure and other
neuropsychiatric symptoms [
17
]. Similar to AD, PD
symptoms also correlate with aggregates of misfolded
protein, in this instance α-synuclein, leading to the sub
-
sequent formation of Lewy bodies [
18
]. PD falls under an
umbrella of synucleinopathies which also include multi
-
ple system atrophy and dementia with Lewy bodies [
19
].
Among neurodegenerative disorders, ALS is the most
rapid to progress to fatality; where PD and AD symptoms
can begin in a prodromal period that can last many years,
ALS can begin and span to death in under 2–3 years [
20
].
ALS manifests as widespread motor neuron abnormali
-
ties involving the brain, spinal column and peripheral
neuromuscular system; speech impairment, difficulty
swallowing followed by progressive paralysis of the arms
and legs are common [
20
]. Progress in therapeutics for
ALS patients is slow due to the complexity and hetero
-
geneity of disease mechanisms. Some 15% of ALS cases
are familial can be directly attributed to disease-caus
-
ing alleles of genes such as
SOD1
,
TARDBP
,
FUS
, and
OPTN
[
20
]. Pathological mechanisms in ALS include
metabolic impairment (gross mitochondrial morphologi
-
cal and functional changes), glutamate-induced excito
-
toxicity, and neuroinflammation [
20
]. Again, in line with
other neurodegenerative diseases, ALS pathophysiology
also includes protein aggregation, this time of the TAR
DNA-binding protein 43 (TDP43) which can occur in
sporadic and familial forms of ALS [
21
].
The etiologies of AD, PD, ALS, and other related
dementias are highly complex. In addition to the patho
-
physiological changes seen post-mortem, such as dep
-
osition of insoluble protein aggregations, there are
overlapping and common mechanisms of neurodegen
-
eration that include neuroinflammatory, metabolic, neu
-
rovascular, and genetic factors.
Neurodegeneration of the visual system
Glaucoma is the leading cause of irreversible blindness
worldwide [
22
]. The disease encompasses a group of
optic neuropathies that lead to the progressive degen
-
eration of retinal ganglion cells (RGCs), the output neu
-
rons of the retina, along with their axons which form the
optic nerve - the sole neuronal projection to the brain’s
higher vision centers. Like many other neurodegenera
-
tive diseases, glaucoma is associated with increasing age;
as our population ages, it is estimated that approximately
112 million people will be affected worldwide by 2040
[
23
]. Besides age, elevation in intraocular pressure (IOP)
is amongst other prominent risk factors for the disease
which include race, severe myopia, central corneal thick
-
ness, and genetic predisposition to congenital glaucoma.
Forms of glaucoma are classified clinically according to
a key anatomic feature of the anterior segment, the irido
-
corneal angle, which is defined by the angle formed where
Page 3 of 29
Wareham
et al. Molecular Neurodegeneration (2022) 17:23
the iris and cornea meet. In the most prevalent form of
the disease, primary open-angle glaucoma (POAG), the
angle is open but there is a progressive resistance within
the aqueous humor outflow pathways that gradually leads
to an increase in IOP. However, not all glaucoma patients
suffer from elevations in IOP; normotensive glaucoma
patients never experience increases in IOP [
24
,
25
] and
conversely some patients with extremes in IOP at risk for
glaucoma do not exhibit neurodegeneration [
26
]. IOP
remains the only treatable risk factor, and although inter
-
ventions in the clinic such as IOP-lowering drops or IOP-
lowering surgery are available, many patients progress
with neurodegeneration of the visual projection despite
treatment [
27
]. As advances are made in research we are
beginning to understand that glaucoma is characterized
by the
sensitivity
of the optic projection to IOP, rather
than IOP itself [
27
]. How this sensitivity begins or evolves
throughout disease progression, or which IOP-independ
-
ent mechanisms are at play remain to be determined but
may hold the key to early detection and prevention in the
disease.
The optic nerve head (ONH), where over 1.5 million
unmyelinated RGC axons converge to exit the globe
and form the optic nerve proper in humans, is a critical
juncture for pathogenic neurodegenerative processes
that occur in glaucoma. The vulnerability of axons at
this site is by virtue of the unique structure and physi
-
ology of the ONH [
28
31
]. There, a complex interplay
is seen between neuronal, glial, vascular, and biome
-
chanical components that can change with age to influ
-
ence sensitivity of the optic projection to any given IOP
[
28
,
29
,
32
,
33
]. All tissues in the human body show nat
-
ural variations in stiffness, and changes in this stiffness
occur naturally with aging, but can also be exacerbated
as a result of inflammatory events (i.e., increased depo
-
sition of collagen and extracellular matrix components
by cells, or proliferation of glia, namely astrocytes). In
addition, remodeled tissue and increased stiffening act
as environmental cues to further drive inflammation
[
34
]. There appears to be an interplay between inflam
-
mation and cellular biomechanics that may be relevant
in glaucoma and tissues of the ONH [
34
]. Changes in
the retina and ONH associated with mechanosensi
-
tivity [
35
,
36
], as well as alterations in ocular stiffness
with age [
37
], have been independently investigated in
glaucoma pathogenesis, along with extracellular matrix
deposition due to inflammation. Making the connec
-
tion between tissue biomechanics and inflammation
as a key molecular driver of pathogenesis may uncover
novel areas of therapeutic intervention in glaucoma. It
is also becoming apparent, in a range of neurodegen
-
erative diseases, that the immune and glial responses
are not dependent on any one genetic mutation or
predisposition for disease – making understanding of
these mechanisms important for all patients.
The variety in etiology of glaucoma combined with the
ineffectiveness of IOP-lowering drugs for many patients
suggests multiple mechanisms of neurodegeneration.
By considering glaucoma a neurodegenerative disease,
research into the triggers (i.e., early molecular events)
and drivers of neurodegeneration can identify novel
areas of therapeutic intervention to preserve and restore
vision. In addition, the optic projection is an accessible
extension of the CNS that allows investigators to directly
visualize CNS neurons and define mechanisms that may
be leveraged for understanding other neurodegenerative
diseases.
Mechanisms of progression
It is no coincidence that as humans age, so too does the
incidence of neurodegenerative disease as homeostatic
cellular mechanisms begin to malfunction, and new cel
-
lular functions associated with diseases arise. Neuro
-
degeneration involves complex interactions between
adjacent cells and their axonal projections; neurons have
both proximal and distal regions that have distinct cel
-
lular environments and in turn distinct mechanisms of
disease pathology. Furthermore, the CNS does not always
act in isolation; the peripheral nervous system (PNS) and
peripheral immune system are increasingly implicated as
active players in the degeneration of the CNS. Identify
-
ing molecular commonalities will enhance understand
-
ing of neurodegenerative events, which could then be
harnessed in the design of broad-stroke therapeutics for
neurodegenerative mechanisms across multiple diseases.
To reach this goal of broadly applicable therapeutics for
neurodegenerative disease some knowledge gaps remain:
(i) common molecular events in the early stages of dis
-
ease progression, i.e., triggering events that tip the scale
in an amplification cascade that leads to neurodegen
-
eration, (ii) events in progression that catalyze already
existing neurodegenerative events, (iii) which cell types
are involved, (iv) common pathological endpoints, i.e.,
how can we back-track from these events to prevent or
replace diseased tissue, and finally (v) discerning which
events are pro-degenerative vs. reparative or even pro-
regenerative. As a collective, we have identified several
common mechanistic areas of focus that may provide
potential pan-neurodegenerative therapeutic strate
-
gies. These include: environmental factors, neuroinflam
-
mation, metabolic stress, neurovascular coupling, and
genetic contributions to disease (Fig.
1
).
Environmental contributions to neurodegeneration
Environmental factors can have a profound impact
on cellular and epigenetic contributions to disease
Page 4 of 29
Wareham
et al. Molecular Neurodegeneration (2022) 17:23
progression. For example, these factors include age, diet,
exercise, and exposure to neurotoxic substances that can
act to trigger and/or exacerbate underlying neurode
-
generative events. As such, environmental factors play
a role in many of the shared degenerative mechanisms
discussed below. Across many diseases, age is a primary
risk factor and tissues that are comprised of postmitotic
cells, such as neurons in the brain and retina, are par
-
ticularly sensitive to the effects of aging [
38
]. Hallmarks
of aging cells include genomic instability, epigenetic
alterations, dysregulated signaling pathways, and mito
-
chondrial dysfunction. Changes that occur with age can
impact homeostatic functions in cells, rendering them
sensitive to neurodegeneration. Other external factors,
such as diet and exercise, are proving to be crucial fac
-
tors in maintaining CNS health [
39
,
40
]. Micronutrients,
such as vitamins and trace elements are integral to many
key biological processes, such as mitochondrial ATP
production and immune responses, which in turn affect
CNS physiology [
39
]. Recognizing the role that external
factors play in degeneration and the impact on cellular
mechanisms as outlined blow (i.e., signaling pathways
such as neuroinflammation, metabolism, mitochondrial
dysfunction), will help to provide novel therapeutic strat
-
egies for neurodegenerative diseases.
Neuroinflammation
Inflammatory events that influence the CNS (what is
sometimes referred to as “neuroinflammation”) have
multifaceted outcomes, which can be neuro-protective,
neuro-regenerative and neurodegenerative, defined by
location, timing, and duration. Inflammation outside of
the CNS involves the infiltration of circulating mono
-
cytes and other immune cells, whereas inflammation
within the CNS is usually (but not always) independent
of peripheral inflammatory infiltration and involves resi
-
dent glia, such as microglia and astrocytes [
41
]. Neuro
-
inflammation in neurodegenerative disease was always
Fig. 1
Common mechanisms of neurodegeneration. Across neurodegenerative diseases, five main areas of mechanistic overlap exist, these
include: (1) environmental factors such as diet, age, and, exercise; (2) metabolic stress, e.g., mitochondrial dysfunction, increased reactive oxygen
species (ROS); (3) genetic contributions, e.g., genome
‑wide association study‑linked risk alleles (GWAS), sex
‑linked genetic contributions; (4)
neurovascular coupling, e.g., breakdown of the blood‑brain‑barrier and dysfunctional neurovascular coupling and; (5) neuroinflammation, e.g.,
infiltration of peripheral immune cells, and increased glial reactivity. Environmental factors contribute to all mechanistic areas of degeneration
Page 5 of 29
Wareham
et al. Molecular Neurodegeneration (2022) 17:23
assumed to be merely a response of the system to other
pathophysiological events. However, emerging data from
preclinical and clinical studies across a range of neurode
-
generative diseases including AD, PD and Huntington’s
Diseases, ALS, and multiple sclerosis, among others,
have established that immune-mediated events can trig
-
ger and drive pathogenesis [
42
45
].
Increasing age is associated with increased low-grade
chronic inflammation, or inflamm-aging [
46
] due to dys
-
regulation of immune [
47
], glia [
48
,
49
], or metabolic
homeostasis [
50
]. In humans, age leads to elevations in
circulating inflammatory markers such as C-reactive
protein [
51
] and inflammatory cytokines [
52
,
53
]. Dys
-
functional inflammatory responses that occur with aging
alone may act to induce or simply aggravate inflamma
-
tory events already underway in neurodegeneration.
Such dysfunction in immune surveillance (usually con
-
ducted by microglia and astrocytes) that occurs with age
may be the instigator in triggering prolonged inflamma
-
tion. In AD, a hallmark of disease pathology is the pres
-
ence of neuroinflammation in the brain, which appears to
manifest as reactive responses by astrocytes and micro
-
glia [
54
]. Elevation in pro-inflammatory cytokines in the
brains of AD patients leads to an accumulation of Aβ
and Tau plaques which ultimately result in neuronal loss
[
55
57
]. Neuronal injury due to accumulating Aβ exists
in a perpetuating cycle whereby production of inflam
-
matory cytokines causes release of neurotoxic Aβ, which
in turn triggers reactive microglia to release more pro-
inflammatory cytokines [
56
,
58
]. In AD, microglia are
the primary cell type that engulfs and proteolyzes neuro
-
toxic Aβ [
3
]. Since Aβ plaques are difficult to break down,
the efficiency of the microglial clearance dissipates with
time leading to increased amyloid and enhanced release
of pro-inflammatory cytokines [
57
]. As such, microglial
responses are likely neuroprotective in the early stage but
neurotoxic in the late stage of AD [
59
].
In humans, inheritance of the apolipoprotein E ε4
(
APOE4
) allele strongly increases the chance of devel
-
oping AD [
60
]. The reactive response of microglia and
astrocytes in the brain is increased in human patients
and mouse models expressing the
APOE4
allele. APOE4
alters the baseline pro-inflammatory response even in the
absence of disease, suggesting that APOE4 may indeed
cause dysfunctional inflammatory responses that trig
-
ger neurodegeneration [
61
63
]. Furthermore, APOE4
is correlated with dysfunctional microglial clearance of
Aβ [
64
]. Although the majority of people carrying the
APOE4
genetic variant have an enhanced predisposition
for AD, the effect size is lower or absent in populations of
people with African ancestry compared with Europeans
or Chinese [
65
]. For example, some South American non-
industrialized populations appear to benefit from APOE4
in order to survive parasitic infection in early childhood,
with no apparent adverse AD-associated effects in aged
individuals [
66
].This lack of association of the allele with
disease highlights how genetic variation, environmental
factors and epigenetics may affect gene-associations of
disease.
Similarly, in age-related macular degeneration (AMD)
and glaucoma,
APOE4
is protective against the disease
[
67
,
68
]. The reason why this inverse relationship is seen
in retinal disease and a positive correlation with disease is
seen in AD is intriguing. In a mouse model of AMD, mice
with the human
APOE4
variant had
less
reactive micro
-
glia [
69
]. Reactive microglia in the retina are already
proven to be pathological in glaucoma, so perhaps less-
reactive glia in the retina are protective in the case of
APOE4
variants whereas dysfunctional microglia in AD
are detrimental. A deeper understanding of evidence
across disease pathologies like this that will enhance our
understanding of glaucoma as a neurodegenerative dis
-
ease and will allow us to understand how neuroinflam
-
matory events contribute to disease pathology across the
spectrum of human populations.
Not all disease-linked mutations cause direct responses
from cells to increase inflammatory mediators. In ALS
patients, harboring genetic mutations in the superoxidase
dismutase enzyme (SOD1) accounts for about 5% of ALS
cases. These mutations do not alter the basal microglial
or astrocyte transcriptome, but instead drastically lower
the astrocyte threshold to inflammation making them
poised to respond faster and more aggressively [
70
]. Such
studies highlight the importance of investigating prodro
-
mal and secondary inflammatory responses and func
-
tions in cells expressing disease-associated mutations.
In PD, similarly to AD, protein aggregations are a key
pathological element; post-mortem examination has
identified aggregations of α-synuclein in Lewy bodies of
patients with the disease [
71
]. These protein aggregates
that accumulate in the neurons of the substantia nigra
are unable to be cleared, triggering neurodegeneration.
Since the discovery of high numbers of reactive microglia
in postmortem brain tissue of PD patients, it has been
suggested that neuroinflammatory events could be the
initial instigator of pathogenic mechanisms in PD [
72
].
Like dysfunctional neuroinflammatory mechanisms in
AD, the same “missing-link” question can be posed for
PD: are neuroinflammatory events responsible for mis
-
folding of proteins, i.e., triggers of the disease, or are they
secondary to protein aggregations? Interestingly, there
have been studies correlating the use of non-steroidal
anti-inflammatory drugs (NSAIDs) with the prevention
or delay of PD [
73
]. Similarly, the glucagon-like 1 pep
-
tide receptor agonist, NYL01, originally developed to
combat inflammation in diabetes, has proved beneficial
Page 6 of 29
Wareham
et al. Molecular Neurodegeneration (2022) 17:23
in limiting microglia cytokine release and astrocyte
reactivity in mouse models of PD [
74
], as well as in the
bead-occlusion model of glaucoma [
75
]. These find
-
ings highlight neuroinflammation and systemic immune
responses as active contributors to progression of disease
and the importance of understanding crosstalk between
the CNS, PNS and vascular system in disease. Below we
discuss the role of additional factors such as mitochon
-
drial pathology, in diseases such as PD.
While these results suggest that NSAID reduce sys
-
temic inflammation associated with PD progression, they
do not resolve why, in general, anti-inflammatory therapy
for neurodegenerative diseases often ends up fruitless.
Indeed, anti-inflammatory or antioxidant therapies for
neurodegenerative diseases in clinical trials have often
been disappointing. An important factor in the role of
neuroinflammation in neurodegeneration is timing. It is
possible that neuroinflammatory responses have a time
and a place for beneficial effects, yet drastic detrimental
effects when activated and persisting at the wrong time
in disease.
Until recently glaucoma was not considered an inflam
-
matory disease largely due to the supposed immune
privilege state of the retina arising from the blood-retinal-
barrier (BRB). However, there is accumulating evidence
to the contrary in studies from both animal models of
the disease and in human patients [
76
,
77
]. Most of our
understanding of how the immune system responds in
glaucoma has been derived from animal models where
onset of elevated IOP leads to early and almost immediate
increase in microglial activation and reactivity [
78
83
].
In post-mortem tissue from human patients, reactive
microglia in the ONH are evident [
84
,
85
]. Inflamma
-
tion in glaucoma appears to be paradoxical; there is a
basal level of intrinsic immune surveillance and reactiv
-
ity that is required to maintain homeostasis, which can
even stimulate regeneration (see below) and yet, too much
stimulation of inflammatory pathways is associated with
degenerative events. In the retina and optic nerve, resi
-
dent glia (microglia, astrocytes, and Müller glia) act as the
immune surveillance and maintain homeostasis by clear
-
ing cellular debris, releasing neuroprotective factors, and
maintaining homeostasis [
86
,
87
]. A sudden insult, such
as an increase in IOP can tip the balance and trigger resi
-
dent glia to adopt a reactive pro-inflammatory, degenera
-
tive state. In addition to resident immune surveillance,
there is clinical evidence of transient optic disc microhe
-
morrhages in patients independent of IOP, indicating a
clear breach of the blood-retinal-barrier (BRB) and infil
-
tration of circulating immune cells that are associated
with disease progression [
88
94
].
The infiltration of circulating immune cells through
BRB rupture may also lend some explanation to an
autoimmune component of the disease seen in animal
models and human patients [
95
]. Serum auto-immu
-
noglobulins against heat-shock proteins (HSPs) have
been found in the retina of animals and humans with
glaucoma, and inoculation of rodents with HSP60 and
HSP27 induces optic neuropathy [
96
,
97
]. A link between
IOP elevation, intact commensal microflora, and T-cell
activation may in part explain HSP-derived autoim
-
mune reactivity. Gut microbiome-sensitized
CD4
+
T-cell infiltration into the retina promotes the progres
-
sive degeneration of the retina and optic nerve after
microbead-induced IOP elevation [
98
]. After IOP insult,
T cells specifically reactive to HSPs infiltrate the retina;
germ-free mice did not show any evidence of neurode
-
generation after IOP elevation [
98
]. These results provide
evidence that T cells reactive to host microflora mediate
prolonged degeneration of the optic nerve after injury.
How circulating immune cells affect resident glial
responses and to what extent factors released by these
cells encourage neurodegeneration remain uncertain. It
is possible that infiltrating cells could promote regenera
-
tion of cell processes lost by acute retinal inflammation.
In the PNS, the innate immune response to injury plays
an essential role in enabling sensory and motor neu
-
rons to regenerate axons back to their peripheral targets
[
99
]. Interestingly, a spike in IOP can also cause an ini
-
tial influx of macrophages and neutrophils that express
molecules (e.g., oncomodulin and SDF1) that can ini
-
tially stimulate growth of the axon [
100
103
], leading to
the questions of what determines cellular release of pro-
regenerative molecules vs. pro-degenerative molecules
under stress conditions and whether there are cells that
can be coaxed towards pro-regenerative states through
release of specific inflammatory factors. In glaucoma,
involvement of the inflammatory response in disease
progression is indisputable, but more research into the
pleiotropic role of immune cells is warranted.
Increasing knowledge of the role of astrocytes and
microglia in disease has led to the identification of a pro-
reactive sub-state of astrocytes (triggered by reactive
microglia) that play a key role in driving retinal degenera
-
tion by release of toxic lipids [
104
,
105
]. Astrocytes have
been identified as important early responders to unilat
-
eral IOP elevation and optic nerve injury by redistribut
-
ing metabolic resources to the site of injury to promote
optic nerve health [
106
]. Understanding how reactive
astrocyte sub-states can drive disease states, or play pro
-
tective roles, is fundamental to advancing our under
-
standing of inflammation in disease.
Metabolic stress
The energy produced by mitochondria (in the form of
adenosine triphosphate; ATP) is required for synthesis
Page 7 of 29
Wareham
et al. Molecular Neurodegeneration (2022) 17:23
of neurotransmitters, bidirectional axonal transport, res
-
toration of ion gradients, buffering of calcium and the
organization of synaptic vesicles, among other functions
[
107
]. Mitochondria are highly dynamic organelles, and
continuously change their size, shape, number, and cel
-
lular location to meet metabolic demands of neurons. In
addition, mitochondrial fusion, and fission are important
for the inheritance of mitochondrial DNA. There are sev
-
eral important processes that mitochondria can undergo
to meet metabolic demands; however, they can become
dysfunctional in disease [
108
]. Mitochondrial biogenesis
describes the biosynthetic process of increasing mito
-
chondrial number [
107
], while a delicate balance between
fusion and fission allows for the rapid adaptation to meet
metabolic demands [
107
,
109
]. Mitophagy, or mitochon
-
drial degradation and clearance is also imperative to
maintain cellular homeostasis. Finally, mitochondria are
transported along the length of neuronal axons to synap
-
tic terminals and dendrites to provide energy at different
focal locations along the neuron
107
.
Besides the inheritance of genes that can cause mito
-
chondrial disease, increasing age increases spontaneous
mutation of mtDNA [
110
]. Aging can also cause mito
-
chondria to function less efficiently, which results in ele
-
vated production of reactive oxygen species (ROS), that
in turn can trigger further mtDNA mutation, pro-inflam
-
matory signaling, and protein dysfunction. ROS produc
-
tion is an unavoidable byproduct of aerobic respiration
along the electron transport chain, and complexes I and
III account for up to 90% of cellular ROS production
[
111
]. Although ROS are important for cellular signaling,
an imbalance leave mitochondria dysfunctional and less
efficient at producing ATP. In addition, ROS can cause
lipid peroxidation in cell membranes, leading to droplet
accumulation in glia a process that is exacerbated in neu
-
rodegeneration [
112
,
113
].
Mitochondrial dysfunction has been linked to PD,
based on the discovery of the roles of PTEN-induced
putative kinase 1 (PINK1) and parkin (PRKN) in mediat
-
ing mitochondrial mitophagy [
114
]. Mutations in PINK1
(
PINK1
) and PRKN (
PARK2
) genes were among the first
genes to be linked to autosomal recessive PD [
115
,
116
],
and there has been increased focus on mitochondrial
roles of inherited gene mutations in PD [
117
]. For exam
-
ple,
LRRK2
mutations lead to α-synuclein aggregates on
the mitochondrial outer membrane [
118
,
119
]. It should
be noted that PD-associated genes
PINK1
and
LRRK2
are highly enriched in astrocytes over other CNS cells
[
120
,
121
] – again implicating non-neuronal cells and
inflammation in the pathogenesis of this neurodegenera
-
tive disease.
Impaired energy metabolism and defects in expres
-
sion of genes related to mitochondrial bioenergetics
are commonly associated with characteristics of AD
pathology [
122
], including altered mitochondrial bio
-
genesis, mitophagy, fusion/fission and axonal trans
-
port of mitochondria [
122
]. For example, Aβ aggregates
cause increased ROS production that can activate
downstream proteases that act on mitochondrial fis
-
sion/fusion GTPases [
122
]. In the case of mitochon
-
drial transport, Aβ associates with motor machinery
including kinesins [
123
] and dyneins [
124
]. In glau
-
coma, evidence of mitochondrial dysfunction is com
-
monly associated with RGC degeneration. Abnormal
mitochondrial morphology and distribution has been
noted in humans and animal models [
125
,
126
]. In a
model of murine glaucoma, mitochondrial transport
in RGCs (including number of transported mitochon
-
dria, distance transported, and rate of transport) is
affected both in the early and late stages of the disease
[
127
]. Furthermore, aged mice exhibit differences in
mitochondrial transport and are more susceptible to
elevated IOP-driven changes than young mice [
127
].
Elevated IOP also affects mitochondrial bioenergetics
in the visual cortex of the brain in rats; ATP produc
-
tion was reduced, superoxide production was increased
and differential mitochondrial complex activity was
observed [
128
].
More generally in neurodegenerative conditions,
mitochondrial transport might be hijacked to com
-
municate a stress signal after a local lesion or infarct.
Conversely, the movement of mitochondria could be
harnessed therapeutically for viral delivery or to pro
-
mote increased clearance of waste products in dis
-
ease. When mitochondrial dynamics are altered, either
through dysfunction or genetic mutation, the impact
for neurons can be catastrophic. The retina is one of the
most metabolically active tissues and requires precise
regulation of energy supply to meet demands [
129
]. The
unmyelinated portion of the RGC axon in the retina
lacks saltatory conduction and therefore is less efficient
generating action potentials [
107
]. Since RGCs rely
heavily on mitochondria in the unmyelinated segment,
dysfunctional mitochondria lead to optic neuropathies
that result in vision loss. Many of these optic neuropa
-
thies occur through the inheritance of a specific genetic
mutation. For example, mutations in Optineurin
(
OPTN
) affect mitophagy and these have been linked
to incidence of glaucoma [
130
]. Mutations in the
OPA1
gene affect mitochondrial fusion and leads to dominant
optic neuropathy, the most common inherited optic
neuropathy [
131
]. Mitochondrial DNA (mtDNA) can
also harbor mutations that lead to disease, including
Leber’s Hereditary Optic Neuropathy (LHON), which
can occur due to a mutation in any of several mtDNA
genes [
132
,
133
].
Page 8 of 29
Wareham
et al. Molecular Neurodegeneration (2022) 17:23
Neurovascular coupling
The metabolic demands of the CNS necessitate a tightly
controlled supply of nutrients and metabolites to main
-
tain cellular homeostasis. Neuronal activity (i.e., meta
-
bolic demand) and blood flow (i.e., metabolic supply) are
coupled such that an increase in neuronal activity evokes
increased blood flow to the area [
134
]. This neurovas
-
cular coupling is mediated by multiple cell types that
together comprise the neurovascular unit (NVU) [
135
],
including vascular smooth muscle cells, pericytes and
endothelial cells as well as astrocytes, microglia and oli
-
godendrocytes [
136
139
]. Aside from metabolic support
and waste removal, a major role of the NVU is to main
-
tain the integrity of the blood-brain-barrier (BBB), which
mediates controlled communication between the CNS
and the periphery [
140
,
141
]. The BBB protects the CNS
from the systemic circulation and regulates the transport
of serum factors and neurotoxins, which could perturb
homeostasis [
142
]. The BBB is not passive; the presence
of specialized tight junctions and transporters on luminal
and abluminal membranes along with membrane-bound
enzymes make it a highly selective and metabolic site of
exchange [
143
]. A specialized CNS glymphatic system
involving cerebral spinal fluid, interstitial fluid and lym
-
phatic vessels contributes to the exchange of nutrients
and signalling molecules with clearance products such
as proteins and solutes in the brain parenchyma [
141
].
Recently, an ocular glymphatic system was described as
an eye-to-cerebrospinal fluid (CSF) pathway that sup
-
ports clearance of waste products from the retina and the
vitreous [
144
].
The function of the BBB and glymphatic systems of
the brain and ocular tissues are fundamental to neuronal
health and have implications in the progression of neu
-
rodegenerative diseases. Some 30% of dementia patients
are specified as suffering from VCID, which represents
the second most common cause of dementia after AD
[
145
,
146
]. VCID arises from stroke or other vascular
injuries that cause significant changes to cognitive func
-
tions. VCID shares comorbidity with other common
dementias such as AD. Around 60% of AD patients show
significant signs of VCID [
145
], and VCID may involve
impaired clearing of Aβ, which is also observed in AD
patients [
147
]. Neurodegeneration also involves a com
-
promise or breakdown of the NVU, which can arise from
the disruption of astrocyte connections with blood ves
-
sels [
145
]. Increased reactivity of astrocytes and micro
-
glia leads to changes in morphology that can destabilize
the NVU and compromise the BBB, which initiates of a
pro-inflammatory and pro-degenerative cycle involving
peripheral immune system invasion.
A risk factor for AD, APOE may be protective of the
peripheral vascular system, along with other molecules
such as high-density lipoprotein (HDL). There appears to
be a functional interplay between lipoproteins and how
they modulate the vascular system, and in turn their indi
-
rect effect on neurons in the CNS. APOE peripherally
associates with HDL and has been linked to clearance of
Aβ in vitro [
148
]. While HDL and APOE work together
to help transport beta-amyloid into vessels, the ApoE2
isoform is more effective than other forms of APOE
[
148
]. Thus, HDL could be neuroprotective target in
amyloid-driven disease, as could APOE in the clearance
of α-synuclein in PD.
In glaucoma, although a vascular theory of the disease
has generated some debate over the decades [
149
152
],
the role of cells in the neurovascular unit in the disease
is only recently becoming clear [
32
,
153
]. Glaucoma
involves alterations in the vasculature, both morpho
-
logical (i.e. blood vessel diameter, capillary dropout) and
functional (i.e., NVC dysfunction) [
32
]. Neurovascu
-
lar coupling in the ONH and retina has been elegantly
demonstrated through measurements of hemodynamic
responses to flicker-light stimulation [
154
157
]. In glau
-
coma patients, flicker-light induced retinal vasodilation
is diminished [
158
,
159
]. Interestingly, short-term acute
IOP elevations do not alter flicker-light responses, sug
-
gesting diminished responses in glaucoma are not due to
changes in IOP alone [
155
]. This evidence hints at under
-
lying dysfunction in the NVU, either due to reduced neu
-
ronal activity or altered glial cell function [
160
,
161
].
Recently, an important role for pericytes in coordinat
-
ing NVU responses in the retina has been highlighted as
an integral component of RGC homeostasis and function
[
153
]. Pericytes are highly mobile and interact to finely
tune blood flow through capillaries in the retina through
inter-pericyte tunnelling nanotubes (IP-TNTs), as visual
-
ized though in vivo imaging [
153
]. Pericyte IP-TNTS are
a key component of microcapillary blood flow regulation
and are damaged in ocular hypertension [
162
]. This work
highlights not only a potential role for dysfunctional peri
-
cyte networks in neurodegeneration, but also the accessi
-
bility of the retina as a model for CNS disease. In addition
to neurodegeneration of the retina, a pathogenic role for
APOE4 in pericytes has also been shown in an in vitro
model of cerebral amyloid angiography, reiterating the
important role of pericyte function in neurodegenerative
disease [
163
]. Understanding how pericytes react in reti
-
nal disease could inform mechanisms of neurodegenera
-
tion in AD, PD and traumatic brain injury.
Genetic contributors
Characterization of genes responsible for neurodegen
-
erative diseases allows at least partial understanding of
risk through inheritance of disease-associated alleles,
and thus heritability is often used as a population-based
Page 9 of 29
Wareham
et al. Molecular Neurodegeneration (2022) 17:23
measure of risk for developing a particular disease. Her
-
itability is formally defined as the proportion of pheno
-
typic variance due to genetic factors, although it does not
mean that inheritance of a gene will cause disease, and
similarly not all individuals with the disease will carry the
same risk alleles. Progressing from heritability to disease
mechanisms is not a trivial task. One important ques
-
tion to consider is whether the risk allele resides in a gene
directly affecting disease, e.g., is it monogenic in nature
(a “core gene”), or whether it is a mutation in a “periph
-
eral gene” only indirectly affecting the course of disease
through potential regulation of or interaction with core
genes [
164
]. Although genome-wide association studies
(GWAS) have identified novel single nucleotide polymor
-
phisms (SNPs), these have generally not been useful for
generating disease risk predictive models for use in the
clinic [
165
]. One major reason for this is that many neu
-
rodegenerative diseases are polygenic in nature [
166
]. A
better determination of genetic risk of developing dis
-
ease is through the compilation of a polygenic risk score
(PRS). The score considers the small effects of many
genetic variations that contribute to disease risk, better
capturing the polygenicity of a disease. Indeed, capturing
the polygenicity of a disease may lead to the identification
of co-morbidities between diseases and common mecha
-
nisms to combat more generally a broad range of neuro
-
degenerative diseases.
Genome-wide association studies have been critical
for identifying risk factors in AD [
164
] and studies have
highlighted common gene-linked pathways e.g.
APOE4
and the closely associated lipoprotein
CLU
[
167
]. As
noted above,
APOE4
is a shared risk factor for both AD
and Parkinson’s disease dementia (PDD), and there is evi
-
dence for an APOE-genotype effect on multiple aspects
of protein aggregation, inflammation, and neurodegen
-
eration across several distinct diseases including AD
and PDD [
168
171
]. Studies that have combined genetic
risk factors across diseases in mice have provided an
insight into the mechanisms linking APOE genotype to
other neurodegenerative disorders. Transgenic mice that
develop alpha-synuclein pathology (Lewy bodies) have
been genetically crossed to genetic isoforms of the
APOE
gene [
172
,
173
].
APOE2
genotype protects against alpha
synuclein degeneration compared with other
APOE
gen
-
otypes while
APOE4
genetic background had the highest
burden of alpha synuclein pathology [
172
]. These results
raise the questions of whether the effects of the protec
-
tive
APOE
genotype are executed at the gene level or at
the level of protein, which has ramifications for leverag
-
ing genetics to create neuroprotective gene replacements.
Like many genes that putatively harbor disease-asso
-
ciated mutations
APOE
is enriched in astrocytes and
microglia.
The effect of sex differences on neurodegeneration is
intriguing and highly complex. In the CNS, sex differ
-
ences are generated by both long- and short-term epi
-
genetic changes caused by gonadal hormones and their
interaction with transcriptional gene products found
on sex chromosomes [
174
,
175
]. Sex hormones and sex
chromosomes therefore each play a part in the response
of the CNS to diseases and aging [
174
]. Aging and disease
are both associated with changes in levels of hormones,
such as testosterone, estradiol, progesterone, and down
-
stream neuroactive metabolites [
176
]. Primary examples
of changes in levels of hormones are in pregnancy or dur
-
ing menopause with both affecting the process of brain
aging in females [
177
].
Of the studies that have focused on sex differences in
neurodegenerative disease, many have highlighted a clear
role of differences between male and female biology in
disease progression. In these studies
APOE4
increases
the risk of AD to a greater degree in women than in men
[
178
], women are less likely to recover from stroke than
men [
179
], estrogen has proven neuroprotective effects in
females [
180
,
181
], and sex differences exist in the use of
cholinesterase inhibitors for the treatment of AD [
182
].
Interestingly, sex-driven pathophysiological changes in
neurodegenerative disease have also been linked to glial
cell populations [
174
]. Indeed, the sex chromosome
complement determines differences in transcriptional
responses in glia in response to injury or disease [
174
].
Furthermore, downstream metabolites of gonadal hor
-
mones can interact directly with hormone receptors on
many types of glial cells to elicit specific neuroprotective
responses [
174
,
183
]. As well as possible direct effects
of sex hormones on neuronal health, sex hormones can
also affect the vasculature which indirectly affects neu
-
ronal survival. The role of sex hormones in maintain
-
ing the integrity of the BBB has been recently reviewed
[
184
]. Moreover, the vasculature in the can generate sex
hormones locally [
185
]. Sexual dimorphisms are also
abundant in glaucoma; there is increasing evidence that
lifetime exposure to estrogen may alter the pathogenesis
of glaucoma and that estrogen may have a neuroprotec
-
tive effect on progression of POAG [
186
,
187
].
Over the last decade, genetic studies including GWAS
have identified over 260 risk alleles for glaucoma. Studies
of heritability of disease have shown that glaucoma, spe
-
cifically POAG, is one of the most commonly inherited
diseases [
188
]. Family-based linkage analyses have identi
-
fied three monogenic risk genes for the disease:
MYOC,
TBK1
and
OPTN
[
188
]. Monogenic risk factors, how
-
ever, only account for less than 5% of all cases of POAG,
suggesting that risk factors for the disease are polygenic
in nature; high heritability is due to hundreds or maybe
even thousands of gene variants with an additive effect