www.nature.

com/mp Molecular Psychiatry

EXPERT REVIEW OPEN

Cerebral dopamine neurotrophic factor protects and repairs

dopamine neurons by novel mechanism
1 1✉
Päivi Lindholm and Mart Saarma

© The Author(s) 2021

Midbrain dopamine neurons deteriorate in Parkinson’s disease (PD) that is a progressive neurodegenerative movement disorder.
No cure is available that would stop the dopaminergic decline or restore function of injured neurons in PD. Neurotrophic factors
(NTFs), e.g., glial cell line-derived neurotrophic factor (GDNF) are small, secreted proteins that promote neuron survival during
mammalian development and regulate adult neuronal plasticity, and they are studied as potential therapeutic agents for the
treatment of neurodegenerative diseases. However, results from clinical trials of GDNF and related NTF neurturin (NRTN) in PD have
been modest so far. In this review, we focus on cerebral dopamine neurotrophic factor (CDNF), an unconventional neurotrophic
protein. CDNF delivered to the brain parenchyma protects and restores dopamine neurons in animal models of PD. In a recent
Phase I-II clinical trial CDNF was found safe and well tolerated. CDNF deletion in mice led to age-dependent functional changes in
the brain dopaminergic system and loss of enteric neurons resulting in slower gastrointestinal motility. These defects in Cdnf−/−
1234567890();,:

mice intriguingly resemble deficiencies observed in early stage PD. Different from classical NTFs, CDNF can function both as an
extracellular trophic factor and as an intracellular, endoplasmic reticulum (ER) luminal protein that protects neurons and other cell
types against ER stress. Similarly to the homologous mesencephalic astrocyte-derived neurotrophic factor (MANF), CDNF is able to
regulate ER stress-induced unfolded protein response (UPR) signaling and promote protein homeostasis in the ER. Since ER stress is
thought to be one of the pathophysiological mechanisms contributing to the dopaminergic degeneration in PD, CDNF, and its
small-molecule derivatives that are under development may provide useful tools for experimental medicine and future therapies
for the treatment of PD and other neurodegenerative protein-misfolding diseases.

Molecular Psychiatry (2022) 27:1310–1321; https://doi.org/10.1038/s41380-021-01394-6

INTRODUCTION [8, 9]. Lewy body pathology can be widespread in the central
Increased life expectancy and a growing aging population are nervous system (CNS) as well as in the peripheral nervous
leading to an increase in the incidence of age-related diseases, system (PNS) including the enteric nervous system (ENS) [10]. The
including Parkinson’s disease (PD) which affects 1% of population non-motor symptoms of PD can be related to the dysfunction
over 60 years of age [1], and with more than 6 million people of DA and other neurotransmitter systems, such as the
diagnosed with PD globally [2]. PD is a progressing neurodegen- noradrenergic and cholinergic systems [4]. However, the neuro-
erative movement disorder, in which midbrain dopamine (DA) pathological mechanisms behind the non-motor symptoms are
neurons in the substantia nigra (SN) degenerate and die. Major largely unknown.
motor symptoms of PD are slowness of movement, resting tremor, Treatments are available that can improve motor symptoms of
rigidity, and postural instability that appear when there is about PD in most patients, but no disease-modifying therapy exists.
30% loss of DA neurons in the SN and 50–60% reduction in striatal Future therapies should include interventions that slow down or
DA axon terminals [3]. Patients with PD also suffer from non-motor prevent the degeneration and death of DA neurons, regenerate
symptoms, including constipation, hyposmia, depression, lack of the remaining DA neurons and increase their functional activity.
motivation, sleep disorders, and cognitive decline that signifi- They should also alleviate non-motor symptoms of PD. Neuro-
cantly decrease quality of life [4, 5]. trophic factors (NTFs) hold great promise as drugs that could
Although a few toxins and genetic mutations are known to promote neuroprotection of DA neurons, and even have the
cause PD, the etiology is unknown in majority of cases. While capacity to regenerate them. NTFs are small, secreted proteins
precise mechanisms of DA neuron death are unclear, increasing that promote neuronal survival, regulate development, function
body of evidence suggests that protein aggregation, mitochon- and maintenance of neurons, and advance neuronal recovery
drial dysfunction, inflammation, and reduced growth factor levels from injury [5, 11, 12]. Glial cell line-derived neurotrophic factor
are involved in the molecular pathogenesis of PD [6, 7]. (GDNF) family ligands (GFLs) GDNF (Figs. 1C, D and 3A) and NRTN
Aggregation of misfolded α-synuclein (αSyn), a major component have been shown to be efficient in protecting DA neurons in
of intraneuronal Lewy bodies, may possibly cause endoplasmic rodent and non-human primate (NHP) models of PD, but have
reticulum (ER) stress in DA neurons leading to neuronal death only shown modest effects in Phase II clinical trials in PD patients

Institute of Biotechnology, Helsinki Institute of Life Science, University of Helsinki, FI-00014 Helsinki, Finland. ✉email: mart.saarma@helsinki.fi
1

Received: 14 June 2021 Revised: 9 November 2021 Accepted: 15 November 2021

Published online: 14 December 2021
 P. Lindholm and M. Saarma
1311
[19]. Another important aspect is limited diffusion of GDNF and
NRTN in brain parenchyma that can decrease target engagement
[5, 14]. A major limitation of NTF therapy is the requirement for
their intracranial delivery using invasive brain stereotactic surgery,
as NTF proteins do not cross the blood–brain barrier (BBB). In
order to find out the real value of NTF therapy, several factors
should be taken into consideration. Firstly, treatment should be
started as soon as possible following the clinical diagnosis of PD.
However, currently this is regulated by ethical considerations,
which do not allow invasive surgery for the treatment of early
stage PD patients. Secondly, gene technology and protein design
can be used to improve the therapeutic and pharmacokinetic
properties of NTFs. Thirdly, it is possible to search for new trophic
factors and neurotrophic small molecules with better therapeutic
properties.
We have discovered a protein with NTF properties, named
cerebral dopamine neurotrophic factor (CDNF) [20], that together
with the related mesencephalic astrocyte-derived neurotrophic
factor (MANF, also known as ARMET) [21], form a novel
evolutionarily conserved family of unconventional NTFs [22–27].
CDNF and MANF have neurotrophic properties but they otherwise
dramatically differ from other known NTFs (Table 1). They have a
unique structure, mode of action and they can promote cellular
protein homeostasis by regulating ER stress, regulate inflamma-
tion and support neuron survival in animal models of PD [22–27].
Surprisingly, variants of CDNF can cross through the BBB thus
opening a new possibility for a systemic administration of this
neurotrophic drug [28]. In this review, we discuss the structure,
cellular effects, biology, and therapeutic potential of CDNF.
We also briefly introduce characteristic features of MANF in order
to give an overview of CDNF/MANF protein family.

CDNF and MANF are structurally unique proteins regulating

ER homeostasis
The three-dimensional structures of mature CDNF and MANF
proteins consist of a unique combination of two domains, an
amino-terminal (N-terminal) saposin-like domain and a carboxy-
terminal (C-terminal) SAF-A/B, Acinus, and PIAS (SAP) domain [29–
Fig. 1 Structural features of human CDNF, MANF, and GDNF 31] (Fig. 1A). The domains are connected by a flexible linker region
proteins. A CDNF (PDB ID: 4BIT [30]) is a monomeric protein. It has suggesting that they can perform separate functions [29–31].
an amino-terminal saposin-like domain that may mediate interac- Since saposin-like proteins usually interact with lipids or mem-
tion with lipids, and a carboxy-terminal SAP (SAF-A/B, Acinus, and branes, it is probable that the N-terminal domain mediates the
PIAS) domain. The CXXC motif (CRAC and CKGC in CDNF and MANF, CDNF/MANF interaction with lipids [31]. Indeed, MANF was shown
respectively) forming a cysteine bridge is located in the C-terminal to directly bind sulfoglycolipid 3-O-sulfogalactosylceramide (sulfa-
domain. Cysteine bridges stabilizing the 3D structure are shown in
tide) possibly via its N-terminal domain [32]. The C-terminal SAP-
yellow. An ER retention signal (KTEL) is in the C-terminus of CDNF. B
Primary structure of CDNF and MANF. CDNF/MANF proteins have an domain is important for the neuroprotective activity of MANF,
N-terminal signal peptide directing them to the ER (Pre). Conserved since it can independently promote the survival of neurons in vitro
cysteine residues in mature CDNF (green) and MANF (blue) are [29]. In their primary structure, CDNF/MANF proteins have eight
indicated as yellow bars, and disulfide bridges as black connecting cysteine residues with conserved spacing, which are important for
lines. Human mature CDNF and MANF consist of 161 and 158 amino the protein fold (Fig. 1B). Three intramolecular disulfide bonds
acid residues, respectively, and the amino acid identity between stabilize the saposin fold of the N-terminal domain and a fourth
them is 59%. C Two GDNF (PDB ID: 1AGQ [126]) monomers disulfide bond can be formed in a CXXC motif in the SAP-
(molecular mass 15 kDa; indicated in blue and red) are connected by domain [31]. When the CXXC motif was mutated, neuroprotec-
an intermolecular disulfide bridge (in yellow) to form a homodimer.
tive activity of MANF was lost indicating that this motif is crucial
D GDNF primary structure contains a signal sequence (Pre) directing
it to the secretory pathway, a pro-sequence that is enzymatically for the biological activity of MANF [33]. At the very C-terminal
cleaved releasing mature GDNF (red) with seven conserved end, CDNF and MANF have an ER retrieval sequence resembling
cysteines (in yellow). Number of amino acid residues is indicated. the canonical lysine-aspartic acid-glutamic acid-leucine (KDEL)
sequence preventing protein secretion from the ER [34, 35]
(Figs. 1A, B and 3C). In support for the role of KDEL-receptors
[5, 13, 14]. GDNF has not been shown to be neuroprotective in the (KDEL-Rs) in regulating CDNF and MANF secretion, deletion of
rodent αSyn model of PD, where αSyn was overexpressed by viral the C-terminal KDEL-like sequence increases their release from
vectors [15], but it in vitro and in vivo protects DA neurons from cells [33–37]. Human CDNF has potential sites for N-linked and
accumulation of misfolded αSyn [16]. Why have GDNF and NRTN O-linked glycosylation but glycosylation is not required for its
given modest therapeutic effects in clinical trials so far? One of the secretion [20, 38].
reasons is that patient populations with advanced PD were treated In cells, CDNF and MANF reside mainly in the lumen of the ER
in the Phase II clinical trials [17, 18]. Five years after clinical [39, 40] where, especially MANF and likely CDNF have an
diagnosis, PD patients have almost no striatal dopaminergic fibers important role in regulating of ER protein homeostasis and
left and have pronounced loss of DA neuron cell bodies in the SN promoting cell survival under ER stress [23, 26]. ER stress is a

Molecular Psychiatry (2022) 27:1310 – 1321

 P. Lindholm and M. Saarma
1312
Table 1. General properties of CDNF, GDNF, and NRTN proteins.
CDNF GDNF NRTN References
Protein family CDNF/MANF TGF-β TGF-β [20, 123, 124]
Structure Saposin-like domain and Cystine knot Cystine knot [29–31, 123, 124]
SAP-domain
Polypeptide Pre-CDNF Prepro-GDNF Prepro-NRTN [20, 123, 125]
Number of amino acids in mature 161 134 102 [20, 123, 124]
protein
Active conformation Monomer Homodimer, disulfide- Homodimer, disulfide- [30, 126, 127]
linked linked
Molecular mass 18 kDa 32 kDa 25 kDa [20, 123, 124]
Calculated pI 7.7 9.44 9.01 [31, 128, 129]
Heparin binding Weak Strong Very strong [130, 131]
Diffusion in brain tissue Good Limited Very limited [84, 100, 132]
Solubility Good Good Poor [20, 133]
Stability Good Good Good [101, 134]
Inhibits cell death Yes Yes Yes [20, 135, 136]
Regulates UPR Yes ? ? [70, 121]

condition where protein-folding capacity of the ER is over- interaction with BiP was not required for its neuroprotective
whelmed resulting in accumulation of unfolded proteins in the activity [63]. Further studies demonstrated that MANF directly
lumen. It can be due to various physiological and pathological binds to the luminal domain of IRE1α [65]. MANF binding
conditions, including increased demand of protein secretion, decreased ER stress-induced oligomerization and phosphorylation
synthesis of mutant proteins, hypoxia, nutrient deprivation, or of IRE1α, leading to attenuation of UPR [65]. Under homeostatic
depletion of ER calcium. To overcome ER stress, an adaptive signal conditions, BiP binds to the luminal domain of IRE1α, PERK, and
transduction pathway termed the unfolded protein response ATF6 keeping them inactive, whereas in ER stress BiP is dissociated
(UPR) is activated to restore ER protein homeostasis by increasing triggering the activation of UPR sensors [41]. MANF was shown to
expression of chaperones to improve protein folding capacity, to compete with BiP for the interaction with IRE1α suggesting that
attenuate translation to reduce protein folding load, and to MANF is able to bind and regulate IRE1α activity only when BiP is
enhance ER-associated protein degradation (ERAD) to remove dissociated, as is the case in ER stress [65]. Thus, IRE1α could act as
misfolded proteins [41]. Three ER transmembrane proteins MANF receptor in the ER and MANF, by moderating IRE1α activity
inositol-requiring enzyme 1α (IRE1α; also known as ERN1), protein could promote cell survival during ER stress [65] (Fig. 2B). The
kinase R-like ER kinase (PERK; also known as EIF2AK3) and biological function of MANF in regulating ER protein homeostasis
activating transcription factor 6 (ATF6) function as sensors for was further supported by protein–protein interaction studies
disturbances in ER protein homeostasis in mammalian cells, and suggesting that MANF is a member of a large multiprotein
their activation induces UPR signaling [41–44] (Fig. 2A). If recovery complex of ER chaperones [63]. A recent report demonstrated that
of ER homeostasis fails, UPR can become chronic leading to MANF can function as a chaperone in the ER, although it does not
apoptosis [45]. UPR has been associated with pathophysiology of show structural or sequence homologies to known chaperone
several neurodegenerative protein-misfolding diseases, including families [66].
PD [46–48]. In ER stress-related disease models in vivo, expression of
Based on the structural homology between CDNF and MANF, endogenous CDNF was reported to increase after cerebral or
we can hypothesize that their molecular mechanism of cytopro- myocardial ischemia [67, 68]. In vitro, ER stress-inducing tunica-
tective action has some similar features. Both CDNF and MANF are mycin treatment increased CDNF expression in cardiomyocytes
widely expressed in mammalian tissues although with differential [69] but not in an osteosarcoma-derived cell line [58]. Thus,
levels [20, 49, 50] suggesting tissue-specific functions. The MANF responsiveness of CDNF to ER stress may depend on cell type.
promoter contains ER stress response elements recognized by However, intracellular CDNF was cytoprotective against ER stress
UPR-induced transcription factors [39, 51, 52] and its expression is and able to regulate UPR. Overexpression of CDNF alleviated ER
increased in ER stress-related conditions [39, 53–55]. Biological stress-induced astrocyte damage, and attenuated the expression
importance of endogenous MANF for the maintenance of ER of ER stress-induced apoptotic proteins in neurons [70, 71]. What is
protein homeostasis was demonstrated in conventional and more, CDNF overexpression may induce a mild adaptive
pancreas-specific MANF knockout mice, where chronic UPR conditioning UPR that prepares cells to encounter ER stress and
activation contributes to the loss of pancreatic insulin producing protects cells in this way [70]. Whether CDNF can regulate UPR via
beta cell mass and development of diabetes mellitus-like binding to UPR sensors, similarly to the interaction of MANF and
condition [56, 57]. In cultured cells, silencing of MANF led to IRE1α, is unknown (Fig. 2B).
activation of UPR and increased susceptibility to ER stress-induced Although CDNF and MANF are largely retained in cells, their
cell death [58]. UPR activation was also detected in Caenorhabditis secretion is increased in ER stress when ER calcium is depleted
elegans [59, 60] and Drosophila melanogaster [61] due to the loss [36, 62, 67]. Secreted CDNF and MANF may function as autocrine
of functional MANF. MANF interacts with an ER chaperone BiP or paracrine trophic factors, promoting cell survival. In accordance
[62, 63], and was shown to prolong BiP interaction with its clients with their potential trophic activities, endogenous CDNF and
thus promoting protein-folding homeostasis in the ER [64]. We MANF can be detected in human serum [72, 73]. Circulating
recently observed that intracellular MANF is able to promote the concentrations of CDNF were not altered in PD patients while
survival of cultured neurons by a mechanism relying on the MANF concentrations were significantly increased and positively
activity of either IRE1α or PERK pathways [63]. However, MANF correlated with the Beck Depression Inventory scoring, which is

Molecular Psychiatry (2022) 27:1310 – 1321

 P. Lindholm and M. Saarma
1313

used to measure the severity of depression. This suggests that seemingly different cytoprotective activities of CDNF and MANF
further studies would be useful to test whether blood MANF levels engage the same or different intracellular signaling pathways and
can be used as a clinical marker of PD [73]. It has been proposed molecular mechanisms is under investigation.
that serum MANF functions as a systemic regulator of inflamma- In contrast to classical NTFs, publications demonstrating
tion and metabolic homeostasis, thus protecting against age- survival-promoting effects of extracellular CDNF and MANF on
related deterioration [74]. naive neurons are limited. Exogenous CDNF was able to support
the development and survival of enteric DA neurons originating
Extracellular trophic activities and plasma membrane from enteric neural crest-derived cells in vitro [75], whereas it did
receptors not support the survival of cultured postnatal midbrain DA
Evidently CDNF and MANF can protect neurons as extracellular neurons [76]. CDNF promoted neither the survival of superior
trophic factors, as demonstrated for example in animal models of cervical ganglion (SCG) neurons, motoneurons, nor dorsal root
PD (as discussed later in detail), and as potential intracellular ganglion neurons in contrast to nerve growth factor (NGF) [20].
regulators of protein homeostasis in the ER. Whether these two MANF protein added to the cell culture was unable to promote

Molecular Psychiatry (2022) 27:1310 – 1321

 P. Lindholm and M. Saarma
1314
Fig. 2 General cellular scheme of unfolded protein response (UPR), and a potential mechanism how MANF and CDNF are regulating UPR
in the ER. A UPR is activated by ATF6, PERK, and IRE1α sensors located in the ER membrane of mammalian cells. In nonstressed conditions, ER
chaperone BiP associates to the luminal domain of IRE1α, PERK, and ATF6 keeping them inactive. When unfolded proteins accumulate in the
ER lumen causing ER stress, BiP is dissociated from the sensors, favoring activation of UPR. Unfolded proteins may also directly bind and
activate IRE1α and PERK [137–139]. Upon activation, IRE1α forms homodimers and oligomers leading activation of its cytosolic kinase domain,
trans-autophoshorylation and stimulation of its ribonuclease (RNase) activity. The active RNase domain of IRE1α removes an intron from XBP1
mRNA leading to the expression of transcription factor XBP1s, which induces transcription of genes related to ER quality control, ER-
associated degradation (ERAD), and lipid synthesis. The RNase of IRE1α may also degrade ER-targeted mRNAs and miRNAs through regulated
IRE1-dependent decay (RIDD), thus decreasing protein folding demand. IRE1α can—via adapter TRAF2—regulate c-Jun N-terminal kinase
(JNK) activation and apoptosis pathways, and NF-κB activation and pro-inflammatory signaling. Activated PERK phosphorylates α-subunit of
eukaryotic initiation factor 2 (eIF2), leading to transient arrest of translation initiation and decreased general protein synthesis. PERK also
phosphorylates transcription factor nuclear factor, erythroid 2-related factor 2 (NRF2) that regulates antioxidant response genes [140, 141].
Translation of ATF4 transcription factor is favored in conditions of limited eIF2α. ATF4 induces transcription of genes involved in protein
folding, redox control, amino acid metabolism and autophagy. Under prolonged ER stress, ATF4 induces pro-apoptotic transcription factor
CCAAT/enhancer-binding protein homologous protein (CHOP). Upon activation, ATF6 translocates to the Golgi where it is cleaved by
endopeptidases, releasing ATF6(N) fragment that functions as a transcription factor. ATF6(N) induces expression of XBP1 mRNA and
components of ERAD. XBP1s and ATF6(N) can induce MANF expression. For in-depth discussion of UPR please see excellent reviews [142–144].
B MANF directly interacts with the ER luminal domain of UPR sensor IRE1α. MANF binding decreases ER stress-induced oligomerization and
phosphorylation of IRE1α, leading to attenuation of UPR. BiP prevents MANF interaction with IRE1α, while MANF at physiological
concentrations does not affect BiP–IRE1α interaction, which suggests that MANF binds and regulates the sensor activity after dissociation of
BiP [65]. Similarly to MANF, CDNF may interact with a UPR sensor to regulate UPR.

the survival of naive DA or SCG neurons, in contrast to GDNF and material. In hippocampal samples of PD patients, CDNF levels
NGF [29, 63]. Compared to naive neurons, the survival-promoting were increased while GDNF levels were decreased suggesting that
effects of CDNF and MANF have been more prominent on injured these factors could represent potential targets for modification to
or stressed neurons [20, 29, 63]. For example, exogenous CDNF help attenuate cognitive decline in PD [81].
protected DA neurons against toxicity of αSyn oligomers [30]. Biological functions of CDNF in the nervous system has been
CDNF also protected hippocampal cells against synaptotoxicity of studied using mouse and zebrafish knockout models [75, 82, 83].
amyloid-β peptide oligomers likely through regulation of ER stress These studies indicate that CDNF expression is important for the
[77]. In addition to neurons, MANF has various effects on non- development and maintenance of various neuronal types and
neuronal cells. Exogenous MANF stimulated the proliferation of circuits rather than specifically for DA neurons. Although CDNF
mouse and human pancreatic beta cells [56, 57, 78] that, protects midbrain DA neurons in rodent models of PD [20, 84–89],
compared many other cell types, have high physiological ER no gross anatomical changes were observed in the midbrain
stress due to synthesis and secretion of insulin [79]. MANF also DA system of conventional Cdnf−/− mice [83]. Numbers of DA
protected cultured embryonic DA neurons against ER stress and neurons in the SNpc, density of tyrosine hydroxylase (TH)- or
decreased induction of UPR genes via a mechanism dependent on dopamine transporter (DAT)-positive fibers in the striatum, or
either IRE1 or PERK pathways [63] suggesting that exogenous striatal DA or DA metabolite levels did not differ between Cdnf−/−
MANF, similarly to intracellular MANF, can promote neuron and Cdnf+/+ mice [83]. However, Cdnf deletion did lead to
survival through regulating UPR. How could exogenous MANF changes of dopaminergic neurotransmission, as amphetamine
regulate UPR signaling in the ER? Bai and colleagues provided one administration induced an increased hyperlocomotor response,
possible answer to this by proposing that extracellular MANF possibly resulting from altered function of DAT in the dopami-
bound to sulfatide can be endocytosed to cells where it mediates nergic axon terminals in striatum of Cdnf−/− mice [83]. Expression
cytoprotection by promoting ER homeostasis [32] (Fig. 3B). The of UPR genes was not altered in the SN or striatum of Cdnf−/−
molecular mechanism of the potential endocytosis of MANF- mice, suggesting that CDNF expression is not essential for the
sulfatide and subsequent molecular events remain to be resolved. maintenance of ER protein homeostasis in the midbrain DA
Protein receptors proposed to interact with CDNF and MANF on system [83]. Further characterization of Cdnf−/− mice demon-
the PM are KDEL-R and neuroplastin (NPTN) [36, 67, 80] (Fig. 3B). strated the importance of Cdnf expression for the development
KDEL-R is mainly localized in the Golgi but it was also detected in and maintenance of neurons in the ENS. Cdnf−/− mice suffered
the PM where it could bind CDNF and MANF through C-terminal from an age-dependent loss of enteric neurons due to increased
KDEL-like sequences [36, 67] (Fig. 3B). Protective effects of neurodegeneration and autophagy observed selectively in the
exogenous CDNF against myocardial ischemia/reperfusion injury submucosal plexus of the intestinal wall, leading to slowed
was dependent on the presence of the C-terminal lysine- gastrointestinal motility [83]. Cdnf expression was found to be
threonine-glutamic acid-leucine (KTEL) sequence and PI3K-Akt necessary for the normal development and survival of enteric DA
signaling pathway [67]. However, the C-terminal arginine- neurons since Cdnf deletion resulted in loss of DA neuronal
threonine-aspartic acid-leucine (RTDL) sequence of MANF was markers in the submucosal plexus [75]. The observed ENS defect
dispensable for its neuroprotective activity in a model of cerebral in Cdnf−/− mice was not only for DA neurons as the numbers of
ischemia [33], suggesting alternative mechanisms for exogenous NOS-, GABA-, and CGRP-expressing neurons were also decreased
MANF activity. Recently, NPTN was identified as a novel PM [75]. The data suggest that the observed functional changes in
receptor for MANF [80] (Fig. 3B). Direct binding of MANF to NPTN the brain dopaminergic system and loss of ENS neurons in
decreased ER stress-mediated inflammation and cell death [80]. Cdnf−/− mice resemble deficiencies observed in early stage PD
However, it is unclear whether NTPN is the major PM receptor for [83]. In a human population study, mutations in CDNF gene were
MANF. Different from MANF, CDNF does not bind NPTN or not identified in patients with early-stage PD [90]. However, a
sulfatide [32, 80], suggesting that cell surface receptors for CDNF trend towards susceptibility to PD was observed in subjects
remain to be discovered. carrying an allele of an intronic CDNF single nucleotide
polymorphism (SNP) [90].
CDNF and MANF knockout neuronal phenotypes Zebrafish cdnf mutants generated using CRISPR/Cas9-genome
There are only few studies reporting endogenous levels of CDNF editing were viable, fertile, and had no gross morphologic
in patients with PD, obviously due to limited availability of tissue phenotype [82]. Importantly, loss of cdnf caused impairments

Molecular Psychiatry (2022) 27:1310 – 1321

 P. Lindholm and M. Saarma
1315

Fig. 3 Plasma membrane receptors of GFLs and MANF/CDNF. A GDNF Family Ligands (GFLs): GDNF; NRTN; artemin; and persephin, function
as homodimers to activate transmembrane receptor tyrosine kinase RET. Binding of GFLs to RET is mediated by GDNF family receptor-α
(GFRα1–4) co-receptors, which selectively interact with the GFLs. Ligand binding leads to homodimerization and autophosphorylation of RET,
resulting activation of multiple intracellular signaling cascades. GDNF binding to RET is mediated by GFRα1, and leads to activation of Akt,
MAPK and c-Src pathways, promoting neuronal survival and regeneration. B Neuroplastin (NPTN) is a novel receptor for MANF (1). Activation
of NPTN induces NF-κB transcription factor and expression of pro-inflammatory cytokines. MANF binding to NPTN decreases pro-
inflammatory response and protects cells against ER stress-induced inflammation and cell death [80]. MANF binds sulfatide (2). Sulfatide is
present in the plasma membrane (PM) of neurons and other cell types, suggesting that MANF can interact with sulfatide on the PM. MANF
bound to sulfatide can be endocytosed to promote ER homeostasis [32]. The molecular mechanism of the potential endocytosis of MANF-
sulfatide and subsequent molecular events are unclear. CDNF and MANF may bind to the KDEL-receptor (KDEL-R) on the PM via a C-terminal
KDEL-like sequence (3) [36, 67]. Exogenous CDNF promoted calcium homeostasis and mitochondrial maintenance in cardiomyocytes under ER
stress conditions by a mechanism dependent on its KDEL-like sequence suggesting that KDEL-R is binding CDNF [67]. The protective effect of
CDNF was mediated by PI3K/Akt signaling [67]. C MANF, CDNF, and KDEL-R are induced by ER stress. In unstressed cells, CDNF and MANF are
retained in the ER by KDEL-R, whereas in ER stress resulting from the depletion of ER calcium, they are released from cells. In ER stress, KDEL-R
possibly localizes to the PM where it may bind extracellular CDNF and MANF.

in dopaminergic, histaminergic, and GABAergic neurotransmitter expression of tyrosine hydroxylase 2 which functions in DA
systems in selective brain areas, indicating that CDNF is synthesis [82]. Alterations in the neurotransmitter networks were
important in shaping the structure of neurotransmitter circuits associated with abnormal behavior, including impaired social
in these fish CNS [82]. In the brain, cdnf deletion led to increased cohesion and anxiety-related risk taking in adult cdnf mutants

Molecular Psychiatry (2022) 27:1310 – 1321

 P. Lindholm and M. Saarma
1316

Fig. 4 CDNF promotes survival of dopamine neurons. A Midbrain dopamine (DA) neurons project from substantia nigra (SN) to the
putamen forming nigrostriatal pathway. B In Parkinson’s disease (PD), midbrain DA neurons degenerate leading to motor and non-motor
symptoms. Cell bodies of DA neurons are located in the substantia nigra pars compacta (SNpc). C CDNF prevents neurodegeneration and
induces functional recovery of injured DA neurons in animal models of PD. Therapeutic CDNF could reduce ER stress and
neuroinflammation that are thought to be involved in the neuropathogenesis of PD. Differently from GDNF, CDNF does not show
survival-promoting effects on naive DA neurons.

[82]. Mutant fish were also more susceptible to drug-induced number of dopaminergic fibers in the striatum, or the striatal
seizures. Interestingly, the observed behavioral phenotypes of concentrations of DA or its metabolites in adult mice [95].
cdnf mutant fish are reminiscent of human neuropsychiatric Although chronic activation of UPR was detected in the brain
conditions, such as schizophrenia [82], in accordance with the tissue of Manf−/− mice, it did not result in neurodegeneration [95].
suggested association between a CDNF SNP and schizophrenia In contrast to observations in Manf knockout mice, Drosophila
susceptibility in humans [91]. Manf, encoded by a single homolog of human MANF/CDNF,
Homozygous loss-of-function mutations of the human MANF appears to be essential for the maintenance of DA neurites and
gene were reported as a cause of childhood diabetes, and were DA levels in the fly [96]. In DmManf mutant larvae, the volume of
mechanistically connected to ER stress and impaired beta cell DA neurites was diminished whereas somas were maintained,
function [92]. A homozygous MANF mutation was also associated suggesting that DA neurites degenerate before cell bodies [96],
with mild intellectual disability, microcephaly, and deafness [93], thus resembling degeneration of DA neurons in PD. UPR-related
suggesting that MANF has a role in brain development and genes were upregulated in DmManf mutant embryos indicating
normal auditory function. In accordance, Manf inactivation in mice ongoing UPR [61]. Larval lethality of DmManf zygotic mutants was
resulted in a hearing loss [94]. However, characterization of rescued with ubiquitously expressed human MANF or CDNF,
conventional and CNS-specific Manf knockout mice indicated that indicating that DmManf and human MANF and CDNF are
endogenous MANF is not required for the maintenance of functionally conserved [97]. Also in zebrafish, studies of manf
midbrain DA neurons [95]. CNS-specific deletion of Manf in mice knockdown using antisense splice-blocking morpholino oligonu-
did not affect the number of TH-positive DA neurons in the SNpc, cleotides suggested that MANF is involved in the regulation of DA

Molecular Psychiatry (2022) 27:1310 – 1321

 P. Lindholm and M. Saarma
1317
neuron development and maintenance [98]. In the manf-1 mutant neurons in animal models of PD is limited when the neurotoxin-
C. elegans worms neuronal development was normal; however, induced lesion is severe [5]. Importantly, Wang et al. [105]
there was loss of manf-1 activated ER stress and UPR [59, 60], observed, using a rat 6-OHDA model of PD, that AAV8-CDNF
resembling observations in Manf−/− mice and supporting the role administration significantly improved motor function and
of MANF as a regulator of ER homeostasis. increased TH levels in rats with mild 6-OHDA-induced lesions,
but it had limited therapeutic effects in rats with severe lesions
CDNF effects in animal models of Parkinson’s disease [105]. Lentiviral vector-mediated overexpression of CDNF or MANF
In patients with PD, DA neurons located in the SN and projecting alone in the SN showed differential protection of dopaminergic
to the striatum degenerate and die [7] (Fig. 4A, B). In animal function in the 6-OHDA model of PD [87]. While overexpression of
models of PD, degeneration of DA neurons can be induced using CDNF in the SN both reduced amphetamine-induced rotational
neurotoxins 6-hydroxydopamine (6-OHDA) and 1-methyl-4-phe- behavior and loss of striatal TH-positive innervation, overexpres-
nyl-1,2,3,6-tetrahydropyridine (MPTP) [99]. In the first in vivo study, sion of MANF in the SN only protected TH-positive cells in the
a single injection of CDNF before the delivery of 6-OHDA into the nigra [87]. However, combined nigral overexpression of CDNF and
striatum significantly reduced amphetamine-induced ipsilateral MANF led to a robust reduction in amphetamine-induced
turning behavior and almost completely protected nigral DA rotations and protection of both DA cells and their fibers,
neurons in a rat model of PD [20]. When administered 4 weeks indicating that CDNF and MANF can have synergistic neuropro-
after 6-OHDA, CDNF restored the dopaminergic function and tective effects [87]. Unfortunately, the levels overexpressed
prevented the degeneration of DA neurons at least as efficiently as CDNF and MANF in the brain tissue were not reported [87], thus
GDNF [20]. In the following study, the neuroprotective effects of hampering comparisons of their neuroprotective effects. When
2-week striatal infusions of CDNF, MANF, and GDNF were GDNF was overexpressed in the SN it was unable to direct
compared in a rat 6-OHDA model [84]. CDNF rescued 6-OHDA- regeneration of TH-positive axons [106]. Since CDNF is not
lesioned nigral DA neurons and TH-positive fibers in the striatum, anterogradely transported from SN to striatum [102], its effects
whereas MANF and GDNF had no significant effect in these may resemble those of GDNF i.e., have full neuroregenerative
measures [84]. The volume of distribution for injected MANF in the potential only when delivered to the striatum. These data
striatum was larger than that of CDNF, and both MANF and CDNF indicate that CDNF and MANF have differential modes of action
diffused significantly better than GDNF [84, 100]. Intrastriatally and encourages using a combination of different growth factors
injected CDNF similarly to GDNF was retrogradely transported for the treatment of PD. Indeed, an additive neurorestorative
to the SN [84, 101], whereas CDNF injected to SN was not effect of CDNF and GDNF was demonstrated in the 6-OHDA
anterogradely transported to the striatum [102]. model of PD in rats [89]. Experiments on cell lines and DA
Airavaara et al. [85] demonstrated that striatal administration of neurons have clearly shown that CDNF and GDNF have
CDNF was neuroprotective and neurorestorative for the TH- completely different modes of action. These additive effects
positive cells in the nigrostriatal DA system in a mouse MPTP observed in a rat PD model also indicated different mechanisms
model of PD. Jiaming and Niu [103] evaluated the therapeutic of action for CDNF and GDNF [89]. Both CDNF and GDNF were
effects of CDNF-expressing bone marrow-derived mesenchymal able to activate the survival-promoting PI3 kinase-Akt signaling
stem cell (MSC) injections. Using intrastriatal, intraventricular, and pathway, but only CDNF decreased the levels of ER stress
intravenous routes of CDNF-MSC administration, they showed markers ATF6 and BiP, in addition to the level of phosphorylated
neurotrophic effects of CDNF-MSC grafts in a rat 6-OHDA model of eukaryotic initiation factor 2 α subunit (eIF2α) downstream of
PD by intrastriatal and intra-lateral ventricular transplantation the UPR sensor PERK [89]. In 6-OHDA-treated PC12 cells, a
routes. Since CDNF is mostly an intracellular protein, it was of cellular model of PD, CDNF treatment increased cell viability
great interest to test CDNF effects using gene therapy approaches. through upregulating ratio of anti-apoptotic Bcl-2/pro-apoptotic
Bäck and colleagues [86] studied the neuroprotective effect of Bax proteins and downregulating caspase-3 activity, thus
adeno-associated virus (AAV) serotype 2 vector expressing CDNF resembling the function of NTFs [107].
in a rat 6-OHDA model of PD. Elevated levels of CDNF expression Several in vitro studies have indicated that CDNF may provide a
in the striatum resulted in a marked decrease in amphetamine- novel therapy for neuroinflammation related to the microglia.
induced ipsilateral rotations [86]. However, compared to studies In microglial cells, CDNF attenuated the production of pro-
using CDNF protein delivery [20, 84], gene therapy of CDNF inflammatory cytokines prostaglandin E2 and interleukin-1β (IL-
provided only partial protection of DA neurons and their fibers 1β) as well as remarkably suppressed the phosphorylation of c-Jun
[86]. One reason for this can be the retention of CDNF inside the N-terminal kinase (JNK) [108]. Nadella et al. [109] found that in the
cells with very limited diffusion of CDNF in the striatum [86]. Ren 6-OHDA-lesioned rats, CDNF overexpressed from a plasmid vector
et al. [88] examined the neuroprotective and functional restorative reduced nitrosative stress, glial markers, and IL-6 levels in the SN,
effects of CDNF overexpression in the striatum via gene therapy but not TNFα and IL-1β levels, suggesting that CDNF may be a
with an AAV2-CDNF vector in 6-OHDA-lesioned rats. In addition to potential novel agent for the treatment of neuroinflammation
the significant restoration of TH-immunoreactive nigral neurons seen in the PD.
and striatal fibers, positron emission tomography (PET) imaging of We still have very limited information about the effects of CDNF
DA transporters revealed functional recovery of the nigrostriatal on nigral DA neurons in NHPs. CDNF therapeutic effects were first
DA system [88]. Compared to the study by Bäck et al. [86] the studied in a unilateral 6-OHDA lesion model of PD in marmoset
prominent neuroprotection by CDNF in the study by Ren et al. [88] monkeys and compared with the effects of GDNF [110]. This study
may be ascribed to the optimal expression level and greater also monitored the severity of 6-OHDA lesions and treatment
spreading of CDNF in the striatum. Hao et al. [104] demonstrated effects in vivo using 123I-FP-CIT (a DAT ligand) SPECT [110]. This
robust long-term overexpression of MANF in rat striatum using analysis showed a significant increase of DAT binding activity in
AAV9 vector-mediated gene delivery. In a 6-OHDA model of PD, lesioned monkeys treated with CDNF, whereas no statistical
intrastriatal delivery AAV9-MANF provided significant protection difference was observed in the GDNF-treated group [110]. In a
for nigral DA neurons and promoted regeneration of striatal DA more recent study, CDNF restored SN DA neuron integrity when
fibers and increase in striatal DA levels [104]. Striatal MANF effects of CDNF and GDNF were compared in a rhesus monkey
overexpression by AAV9 vector led to increased MANF levels also MPTP model of PD [111]. The animal data together demonstrate
in the SN, suggesting that MANF was retrogradely transported that CDNF not only protects but also restores the function of DA
from the striatum to SN, thus providing local protection for nigral neurons by regulating ER stress, neuroinflammation, and counter-
neurons [104]. The ability of GDNF and related NTFs to rescue DA acting cell death (Fig. 4C).

Molecular Psychiatry (2022) 27:1310 – 1321

 P. Lindholm and M. Saarma
1318
First results of clinical trials encouraging, much more work is needed before BBB-penetrating
Since the mode of action of CDNF differs from that of GDNF, CDNF-derived molecules can be taken to clinical trials. One
NRTN, and other growth factors tested in clinical trials for PD, and important reason for the limited success of clinical development
CDNF was more efficient than GDNF in protecting the function of of NTFs so far is their poor pharmacokinetic characteristics, which
DA neurons in animal models of PD [84, 110] it was important to include inability to cross tissue barriers, poor diffusion in tissues,
test CDNF in clinic. The first clinical Phase I-II, randomized, double- ability to activate several receptors in different tissues and cell
blind study conducted by Herantis Pharma Plc. investigated types, and high costs of the drug [5]. The development of small
the safety and tolerability of intermittent bilateral intraputamenal molecules selectively targeting CDNF receptors with optimized
monthly infusions of CDNF (ClinicalTrials.gov Identifier: pharmacokinetic properties can open a new avenue for the
NCT03295786) [23, 112]. A two-part study in 17 patients with development of disease-modifying treatments of neurodegenera-
advanced PD was carried out in three university hospitals in tive diseases in the future.
Finland and Sweden. During the initial 6-month period, all patients CDNF was also shown to have beneficial effects in animal
received either placebo or CDNF at one of two dose levels. This models of Alzheimer’s disease [120], amyotrophic lateral sclerosis
was followed by a 6-month period, in which all patients received [121], and Huntington’s disease [122]. However, further studies are
CDNF at one of the two dose levels, including the previous required to confirm these exciting results. Taken together, unique
placebo group patients. Treatment was administered via a dose properties of CDNF encourage its testing in different neurological
delivery system using intraputamenal catheters that were diseases, especially in those where neuronal protein homeostasis
implanted into the putamen at the beginning of the study. has been perturbed.
Human recombinant CDNF, used in the study, was produced in a
mammalian cell line and its biological activity was rigorously
tested in neuronal survival assays. Intraputamenal CDNF infusions REFERENCES
were safe and well tolerated, and thus the primary endpoint of the 1. Tysnes OB, Storstein A. Epidemiology of Parkinson’s disease. J Neural Transm
study was met. Exploratory endpoints included UPDRS scores and (Vienna). 2017;124:901–5.
DAT PET, which was performed with a high-resolution research 2. Dorsey ER, Bloem BR. The Parkinson pandemic-a call to action. JAMA Neurol.
2018;75:9–10.
tomography system using DAT radioligand [18F]FE-PE2I [113].
3. Cheng HC, Ulane CM, Burke RE. Clinical progression in Parkinson disease and the
A minimal clinical important difference in Unified Parkinson’s neurobiology of axons. Ann Neurol. 2010;67:715–25.
Disease Rating Scale (UPDRS III) (off) was observed in the CDNF 4. Schapira AHV, Chaudhuri KR, Jenner P. Non-motor features of Parkinson disease.
dose-groups suggesting a potential slowing of disease progres- Nat Rev Neurosci. 2017;18:509.
sion. Furthermore, increased DAT availability in the putamen was 5. Sidorova YA, Saarma M. Can growth factors cure Parkinson’s disease? Trends
observed with PET in some patients that received CDNF Pharm Sci. 2020;41:909–22.
suggesting a potential improvement in dopaminergic function. 6. Allen SJ, Watson JJ, Shoemark DK, Barua NU, Patel NK. GDNF, NGF and BDNF as
Although the study of patient population with advanced PD was therapeutic options for neurodegeneration. Pharm Ther. 2013;138:155–75.
not designed to show efficacy of CDNF, the documented 7. Poewe W, Seppi K, Tanner CM, Halliday GM, Brundin P, Volkmann J, et al.
Parkinson disease. Nat Rev Dis Prim. 2017;3:17013.
improvements in some patients were very encouraging [112].
8. Mercado G, Valdes P, Hetz C. An ERcentric view of Parkinson’s disease. Trends
Mol Med. 2013;19:165–75.
Concluding remarks 9. Colla E. Linking the endoplasmic reticulum to Parkinson’s disease and alpha-
CDNF is an atypical neurotrophic protein that is cytoprotective synucleinopathy. Front Neurosci. 2019;13:560.
both in the ER and as an extracellular factor. In addition to 10. Wakabayashi K, Tanji K, Odagiri S, Miki Y, Mori F, Takahashi H. The Lewy body in
neuroprotective and neuroregenerative activities that, similarly to Parkinson’s disease and related neurodegenerative disorders. Mol Neurobiol.
other NTFs, partially occur via the activation of PI3 kinase-Akt 2013;47:495–508.
pathways [89, 114], CDNF also counteracts cell death by regulating 11. Airaksinen MS, Saarma M. The GDNF family: signalling, biological functions and
UPR pathways in the ER [70, 89]. CDNF protects against toxicity of therapeutic value. Nat Rev Neurosci. 2002;3:383–94.
12. Huang EJ, Reichardt LF. Neurotrophins: roles in neuronal development and
αSyn oligomers in vitro [30], and was recently shown to directly
function. Annu Rev Neurosci. 2001;24:677–736.
interact with αSyn, reduce propagation of αSyn aggregation and 13. Kirkeby A, Barker RA. Parkinson disease and growth factors - is GDNF good
alleviate behavioral deficits induced by αSyn fibrils in mice [115]. enough? Nat Rev Neurol. 2019;15:312–4.
CDNF also reduces the synthesis and release of pro-inflammatory 14. Manfredsson FP, Polinski NK, Subramanian T, Boulis N, Wakeman DR,
cytokines decreasing neuroinflammation [108, 109, 114, 116]. One Mandel RJ. The future of GDNF in Parkinson’s disease. Front Aging Neurosci.
particularly interesting property of CDNF, which differentiates it 2020;12:593572.
from classical NTFs, is that its effects on naive and healthy neurons 15. Decressac M, Ulusoy A, Mattsson B, Georgievska B, Romero-Ramos M, Kirik D,
are low or even absent [20, 29, 63]. This may be very important et al. GDNF fails to exert neuroprotection in a rat alpha-synuclein model of
from the clinical point of view, because it suggests a good safety Parkinson’s disease. Brain 2011;134:2302–11.
16. Chmielarz P, Er S, Konovalova J, Bandres L, Hlushchuk I, Albert K, et al. GDNF/RET
profile for CDNF. Although CDNF has now been successfully tested
signaling pathway activation eliminates Lewy body pathology in midbrain
in rodent and NHP models of PD [20, 76, 84–86, 88, 103, 109– dopamine neurons. Mov Disord. 2020;35:2279–89.
111, 117–119], as well as in Phase I-II clinical trial in patients with 17. Lang AE, Gill S, Patel NK, Lozano A, Nutt JG, Penn R, et al. Randomized controlled
PD [23, 112], several challenges remain. CDNF can regulate UPR trial of intraputamenal glial cell line-derived neurotrophic factor infusion in
pathways, but its receptors and signaling pathways remain poorly Parkinson disease. Ann Neurol. 2006;59:459–66.
described. We also know very little about how CDNF acts in the ER, 18. Olanow CW, Bartus RT, Volpicelli-Daley LA, Kordower JH. Trophic factors for
how it is secreted and whether it has both intracellular and plasma Parkinson’s disease: to live or let die. Mov Disord. 2015;30:1715–24.
membrane receptors. Despite the promising results in animal 19. Kordower JH, Bjorklund A. Trophic factor gene therapy for Parkinson’s disease.
models of PD, NTF- and CDNF-based treatments share a Mov Disord. 2013;28:96–109.
20. Lindholm P, Voutilainen MH, Lauren J, Peranen J, Leppanen VM, Andressoo JO,
fundamental drawback; they require a direct delivery of the
et al. Novel neurotrophic factor CDNF protects and rescues midbrain dopamine
therapeutic protein to the brain through invasive surgery, since neurons in vivo. Nature 2007;448:73–7.
NTFs and CDNF cannot pass through the BBB. We have recently 21. Petrova P, Raibekas A, Pevsner J, Vigo N, Anafi M, Moore MK, et al. MANF: a new
discovered a novel CDNF variant that acts similarly to CDNF, but mesencephalic, astrocyte-derived neurotrophic factor with selectivity for
can efficiently pass through the BBB [28]. Furthermore, our dopaminergic neurons. J Mol Neurosci. 2003;20:173–88.
preliminary data show that this CDNF variant has beneficial 22. Albert K, Airavaara M. Neuroprotective and reparative effects of endoplasmic
effects in both rodent 6-OHDA and MPTP toxin models of PD reticulum luminal proteins - mesencephalic astrocyte-derived neurotrophic
when administered subcutaneously [28]. Although these data are factor and cerebral dopamine neurotrophic factor. Croat Med J. 2019;60:99–108.

Molecular Psychiatry (2022) 27:1310 – 1321

 P. Lindholm and M. Saarma
1319
23. Huttunen HJ, Saarma M. CDNF protein therapy in Parkinson’s disease. Cell 50. Lindholm P, Peranen J, Andressoo JO, Kalkkinen N, Kokaia Z, Lindvall O, et al.
Transplant. 2019;28:349–66. MANF is widely expressed in mammalian tissues and differently regulated after
24. Jantti M, Harvey BK. Trophic activities of endoplasmic reticulum proteins CDNF ischemic and epileptic insults in rodent brain. Mol Cell Neurosci. 2008;39:356–71.
and MANF. Cell Tissue Res. 2020;382:83–100. 51. Lee AH, Iwakoshi NN, Glimcher LH. XBP-1 regulates a subset of endoplasmic
25. Kim Y, Park SJ, Chen YM. Mesencephalic astrocyte-derived neurotrophic factor reticulum resident chaperone genes in the unfolded protein response. Mol Cell
(MANF), a new player in endoplasmic reticulum diseases: structure, biology, and Biol. 2003;23:7448–59.
therapeutic roles. Transl Res. 2017;188:1–9. 52. Wang D, Hou C, Cao Y, Cheng Q, Zhang L, Li H, et al. XBP1 activation enhances
26. Lindahl M, Saarma M, Lindholm P. Unconventional neurotrophic factors CDNF MANF expression via binding to endoplasmic reticulum stress response ele-
and MANF: structure, physiological functions and therapeutic potential. Neu- ments within MANF promoter region in hepatitis B. Int J Biochem Cell Biol.
robiol Dis. 2017;97:90–102. 2018;99:140–6.
27. Sousa-Victor P, Jasper H, Neves J. Trophic factors in inflammation and regen- 53. Hartley CL, Edwards S, Mullan L, Bell PA, Fresquet M, Boot-Handford RP, et al.
eration: the role of MANF and CDNF. Front Physiol. 2018;9:1629. Armet/Manf and Creld2 are components of a specialized ER stress response
28. Saarma M, Voutilainen MH, Airavaara M, Yu LY, Lindahl M. C-terminal CDNF and provoked by inappropriate formation of disulphide bonds: implications for
MANF fragments, pharmaceutical compositions comprising same and uses genetic skeletal diseases. Hum Mol Genet. 2013;22:5262–75.
thereof. United States Patent Application No 2019;17/043028; 2018. 54. Tadimalla A, Belmont PJ, Thuerauf DJ, Glassy MS, Martindale JJ, Gude N, et al.
29. Hellman M, Arumae U, Yu LY, Lindholm P, Peranen J, Saarma M, et al. Mesen- Mesencephalic astrocyte-derived neurotrophic factor is an ischemia-inducible
cephalic astrocyte-derived neurotrophic factor (MANF) has a unique mechanism secreted endoplasmic reticulum stress response protein in the heart. Circ Res.
to rescue apoptotic neurons. J Biol Chem. 2011;286:2675–80. 2008;103:1249–58.
30. Latge C, Cabral KM, de Oliveira GA, Raymundo DP, Freitas JA, Johanson L, et al. 55. Yu YQ, Liu LC, Wang FC, Liang Y, Cha DQ, Zhang JJ, et al. Induction profile of
The solution structure and dynamics of full-length human cerebral dopamine MANF/ARMET by cerebral ischemia and its implication for neuron protection. J
neurotrophic factor and its neuroprotective role against alpha-synuclein oligo- Cereb Blood Flow Metab. 2010;30:79–91.
mers. J Biol Chem. 2015;290:20527–40. 56. Danilova T, Belevich I, Li H, Palm E, Jokitalo E, Otonkoski T, et al. MANF is
31. Parkash V, Lindholm P, Peranen J, Kalkkinen N, Oksanen E, Saarma M, et al. The required for the postnatal expansion and maintenance of pancreatic beta-cell
structure of the conserved neurotrophic factors MANF and CDNF explains why mass in mice. Diabetes 2019;68:66–80.
they are bifunctional. Protein Eng Des Sel. 2009;22:233–41. 57. Lindahl M, Danilova T, Palm E, Lindholm P, Voikar V, Hakonen E, et al. MANF is
32. Bai M, Vozdek R, Hnizda A, Jiang C, Wang B, Kuchar L, et al. Conserved roles indispensable for the proliferation and survival of pancreatic beta cells. Cell Rep.
of C. elegans and human MANFs in sulfatide binding and cytoprotection. Nat 2014;7:366–75.
Commun. 2018;9:897. 58. Apostolou A, Shen Y, Liang Y, Luo J, Fang S. Armet, a UPR-upregulated protein,
33. Matlik K, Yu LY, Eesmaa A, Hellman M, Lindholm P, Peranen J, et al. Role of two inhibits cell proliferation and ER stress-induced cell death. Exp Cell Res.
sequence motifs of mesencephalic astrocyte-derived neurotrophic factor in its 2008;314:2454–67.
survival-promoting activity. Cell Death Dis. 2015;6:e2032. 59. Hartman JH, Richie CT, Gordon KL, Mello DF, Castillo P, Zhu A, et al. MANF
34. Norisada J, Hirata Y, Amaya F, Kiuchi K, Oh-hashi K. A comparative analysis of the deletion abrogates early larval Caenorhabditis elegans stress response to tuni-
molecular features of MANF and CDNF. PLoS ONE. 2016;11:e0146923. camycin and Pseudomonas aeruginosa. Eur J Cell Biol. 2019;98:151043.
35. Oh-Hashi K, Tanaka K, Koga H, Hirata Y, Kiuchi K. Intracellular trafficking and 60. Richman C, Rashid S, Prashar S, Mishra R, Selvaganapathy PR, Gupta BP. C.
secretion of mouse mesencephalic astrocyte-derived neurotrophic factor. Mol elegans MANF homolog is necessary for the protection of dopaminergic neu-
Cell Biochem. 2012;363:35–41. rons and ER unfolded protein response. Front Neurosci. 2018;12:544.
36. Henderson MJ, Richie CT, Airavaara M, Wang Y, Harvey BK. Mesencephalic 61. Palgi M, Greco D, Lindstrom R, Auvinen P, Heino TI. Gene expression analysis of
astrocyte-derived neurotrophic factor (MANF) secretion and cell surface binding Drosophilaa Manf mutants reveals perturbations in membrane traffic and major
are modulated by KDEL receptors. J Biol Chem. 2013;288:4209–25. metabolic changes. BMC Genomics. 2012;13:134.
37. Galli E, Lindholm P, Kontturi LS, Saarma M, Urtti A, Yliperttula M. Characteriza- 62. Glembotski CC, Thuerauf DJ, Huang C, Vekich JA, Gottlieb RA, Doroudgar S.
tion of CDNF-secreting ARPE-19 cell clones for encapsulated cell therapy. Cell Mesencephalic astrocyte-derived neurotrophic factor protects the heart from
Transplant. 2019;4:413–24. ischemic damage and is selectively secreted upon sarco/endoplasmic reticulum
38. Sun ZP, Gong L, Huang SH, Geng Z, Cheng L, Chen ZY. Intracellular trafficking calcium depletion. J Biol Chem. 2012;287:25893–904.
and secretion of cerebral dopamine neurotrophic factor in neurosecretory cells. 63. Eesmaa A, Yu LY, Goos H, Noges K, Kovaleva V, Hellman M, et al. The cyto-
J Neurochem. 2011;117:121–32. protective protein MANF promotes neuronal survival independently from its
39. Mizobuchi N, Hoseki J, Kubota H, Toyokuni S, Nozaki J, Naitoh M, et al. ARMET is role as a GRP78 cofactor. J Biol Chem. 2021;296:100295.
a soluble ER protein induced by the unfolded protein response via ERSE-II 64. Yan Y, Rato C, Rohland L, Preissler S, Ron D. MANF antagonizes nucleotide
element. Cell Struct Funct. 2007;32:41–50. exchange by the endoplasmic reticulum chaperone BiP. Nat Commun. 2019;10:541.
40. Fernandez A, Guzman S, Cruz Y, Zamorano P. Construction of bicistronic lenti- 65. Kovaleva V, Yu L-Y, Ivanova L, Nam J, Eesmaa A, Kumpula E-P, et al. MANF
viral vectors for tracking the expression of CDNF in transduced cells. Plasmid regulates unfolded protein response and neuronal survival through its ER-
2014;76:15–23. located receptor IRE1α. bioRxiv: 2020:2020.09.22.307744 [Preprint]. 2020.
41. Walter P, Ron D. The unfolded protein response: from stress pathway to Available from: https://www.biorxiv.org/content/10.1101/2020.09.22.307744v1.
homeostatic regulation. Science 2011;334:1081–6. 66. Arrieta A, Blackwood EA, Stauffer WT, Santo Domingo M, Bilal AS, Thuerauf DJ,
42. Harding HP, Zhang Y, Ron D. Protein translation and folding are coupled by an et al. Mesencephalic astrocyte-derived neurotrophic factor is an ER-resident
endoplasmic-reticulum-resident kinase. Nature 1999;397:271–4. chaperone that protects against reductive stress in the heart. J Biol Chem.
43. Bertolotti A, Zhang Y, Hendershot LM, Harding HP, Ron D. Dynamic interaction 2020;295:7566–83.
of BiP and ER stress transducers in the unfolded-protein response. Nat Cell Biol. 67. Maciel L, de Oliveira DF, Mesquita F, Souza H, Oliveira L, Christie MLA, et al. New
2000;2:326–32. cardiomyokine reduces myocardial ischemia/reperfusion injury by PI3K-AKT
44. Haze K, Yoshida H, Yanagi H, Yura T, Mori K. Mammalian transcription pathway via a putative KDEL-receptor binding. J Am Heart Assoc. 2021;10:e019685.
factor ATF6 is synthesized as a transmembrane protein and activated by 68. Zhang GL, Wang LH, Liu XY, Zhang YX, Hu MY, Liu L, et al. Cerebral dopamine
proteolysis in response to endoplasmic reticulum stress. Mol Biol Cell. 1999; neurotrophic factor (CDNF) has neuroprotective effects against cerebral ische-
10:3787–99. mia that may occur through the endoplasmic reticulum stress pathway. Int J
45. Sano R, Reed JC. ER stress-induced cell death mechanisms. Biochim Biophys Mol Sci. 2018;19:1905.
Acta. 2013;1833:3460–70. 69. Liu H, Yu C, Yu H, Zhong L, Wang Y, Liu J, et al. Cerebral dopamine neurotrophic
46. Baek JH, Whitfield D, Howlett D, Francis P, Bereczki E, Ballard C, et al. Unfolded factor protects H9c2 cardiomyocytes from apoptosis. Herz 2018;43:346–51.
protein response is activated in Lewy body dementias. Neuropathol Appl 70. Arancibia D, Zamorano P, Andres ME. CDNF induces the adaptive unfolded
Neurobiol. 2016;42:352–65. protein response and attenuates endoplasmic reticulum stress-induced cell
47. Hetz C, Saxena S. ER stress and the unfolded protein response in neurode- death. Biochim Biophys Acta Mol Cell Res. 2018;1865:1579–89.
generation. Nat Rev Neurol. 2017;13:477–91. 71. Cheng L, Zhao H, Zhang W, Liu B, Liu Y, Guo Y, et al. Overexpression of con-
48. Hoozemans JJ, van Haastert ES, Eikelenboom P, de Vos RA, Rozemuller JM, served dopamine neurotrophic factor (CDNF) in astrocytes alleviates endo-
Scheper W. Activation of the unfolded protein response in Parkinson’s disease. plasmic reticulum stress-induced cell damage and inflammatory cytokine
Biochem Biophys Res Commun. 2007;354:707–11. secretion. Biochem Biophys Res Commun. 2013;435:34–9.
49. Danilova T, Galli E, Pakarinen E, Palm E, Lindholm P, Saarma M, et al. Mesen- 72. Galli E, Härkönen T, Sainio MT, Ustav M, Toots U, Urtti A, et al. Increased cir-
cephalic astrocyte-derived neurotrophic factor (MANF) is highly expressed in culating concentrations of mesencephalic astrocyte-derived neurotrophic factor
mouse tissues with metabolic function. Front Endocrinol. 2019;10:765. in children with type 1 diabetes. Sci Rep. 2016;6:29058.

Molecular Psychiatry (2022) 27:1310 – 1321

 P. Lindholm and M. Saarma
1320
73. Galli E, Planken A, Kadastik-Eerme L, Saarma M, Taba P, Lindholm P. Increased 96. Palgi M, Lindstrom R, Peranen J, Piepponen TP, Saarma M, Heino TI. Evidence
serum levels of mesencephalic astrocyte-derived neurotrophic factor in subjects that DmMANF is an invertebrate neurotrophic factor supporting dopaminergic
with Parkinson’s disease. Front Neurosci. 2019;13:929. neurons. Proc Natl Acad Sci USA. 2009;106:2429–34.
74. Sousa-Victor P, Neves J, Cedron-Craft W, Ventura PB, Liao CY, Riley RR, et al. 97. Lindstrom R, Lindholm P, Kallijarvi J, Yu LY, Piepponen TP, Arumae U, et al.
MANF regulates metabolic and immune homeostasis in ageing and protects Characterization of the structural and functional determinants of MANF/CDNF in
against liver damage. Nat Metab. 2019;1:276–90. Drosophila in vivo model. PLoS ONE. 2013;8:e73928.
75. Chalazonitis A, Li Z, Pham TD, Chen J, Rao M, Lindholm P, et al. 98. Chen YC, Sundvik M, Rozov S, Priyadarshini M, Panula P. MANF regulates dopa-
Cerebral dopamine neurotrophic factor is essential for enteric neuronal devel- minergic neuron development in larval zebrafish. Dev Biol. 2012;370:237–49.
opment, maintenance, and regulation of gastrointestinal transit. J Comp Neurol. 99. Airavaara M, Parkkinen I, Konovalova J, Albert K, Chmielarz P, Domanskyi A. Back
2020;528:2420–44. and to the future: from neurotoxin-induced to human Parkinson’s disease
76. Jaumotte JD, Saarma M, Zigmond MJ. Protection of dopamine neurons by CDNF models. Curr Protoc Neurosci. 2020;91:e88.
and neurturin variant N4 against MPP+ in dissociated cultures from rat 100. Voutilainen MH, Back S, Porsti E, Toppinen L, Lindgren L, Lindholm P, et al.
mesencephalon. PLoS ONE. 2021;16:e0245663. Mesencephalic astrocyte-derived neurotrophic factor is neurorestorative in rat
77. Zhou W, Chang L, Fang Y, Du Z, Li Y, Song Y, et al. Cerebral dopamine neuro- model of Parkinson’s disease. J Neurosci. 2009;29:9651–9.
trophic factor alleviates Abeta25-35-induced endoplasmic reticulum stress and 101. Matlik K, Vihinen H, Bienemann A, Palgi J, Voutilainen MH, Booms S, et al.
early synaptotoxicity in rat hippocampal cells. Neurosci Lett. 2016;633:40–6. Intrastriatally infused exogenous CDNF is endocytosed and retrogradely trans-
78. Hakonen E, Chandra V, Fogarty CL, Yu NY, Ustinov J, Katayama S, et al. MANF ported to substantia nigra. eNeuro. 2017;4:ENEURO.0128-16.2017.
protects human pancreatic beta cells against stress-induced cell death. Diabe- 102. Albert K, Renko JM, Matlik K, Airavaara M, Voutilainen MH. Cerebral dopamine
tologia 2018;61:2202–14. neurotrophic factor diffuses around the brainstem and does not undergo
79. Eizirik DL, Cardozo AK, Cnop M. The role for endoplasmic reticulum stress in anterograde transport after injection to the substantia nigra. Front Neurosci.
diabetes mellitus. Endocr Rev. 2008;29:42–61. 2019;13:590.
80. Yagi T, Asada R, Kanekura K, Eesmaa A, Lindahl M, Saarma M, et al. Neuroplastin 103. Jiaming M, Niu C. Comparing neuroprotective effects of CDNF-expressing bone
modulates anti-inflammatory effects of MANF. iScience. 2020;23:101810. marrow derived mesenchymal stem cells via differing routes of administration
81. Virachit S, Mathews KJ, Cottam V, Werry E, Galli E, Rappou E, et al. Levels of glial utilizing an in vivo model of Parkinson’s disease. Neurol Sci. 2015;36:281–7.
cell line-derived neurotrophic factor are decreased, but fibroblast growth factor 104. Hao F, Yang C, Chen SS, Wang YY, Zhou W, Hao Q, et al. Long-term protective
2 and cerebral dopamine neurotrophic factor are increased in the hippocampus effects of AAV9-mesencephalic astrocyte-derived neurotrophic factor gene
in parkinson’s disease. Brain Pathol. 2019;29:813–25. transfer in parkinsonian rats. Exp Neurol. 2017;291:120–33.
82. Chen YC, Baronio D, Semenova S, Abdurakhmanova S, Panula P. Cerebral 105. Wang L, Wang Z, Zhu R, Bi J, Feng X, Liu W, et al. Therapeutic efficacy of AAV8-
dopamine neurotrophic factor regulates multiple neuronal subtypes and behavior. mediated intrastriatal delivery of human cerebral dopamine neurotrophic factor
J Neurosci. 2020;40:6146–64. in 6-OHDA-induced parkinsonian rat models with different disease progression.
83. Lindahl M, Chalazonitis A, Palm E, Pakarinen E, Danilova T, Pham TD, et al. PLoS ONE. 2017;12:e0179476.
Cerebral dopamine neurotrophic factor-deficiency leads to degeneration of 106. Kirik D, Rosenblad C, Bjorklund A, Mandel RJ. Long-term rAAV-mediated gene
enteric neurons and altered brain dopamine neuronal function in mice. Neu- transfer of GDNF in the rat Parkinson’s model: intrastriatal but not intranigral
robiol Dis. 2020;134:104696. transduction promotes functional regeneration in the lesioned nigrostriatal
84. Voutilainen MH, Back S, Peranen J, Lindholm P, Raasmaja A, Mannisto PT, et al. system. J Neurosci. 2000;20:4686–700.
Chronic infusion of CDNF prevents 6-OHDA-induced deficits in a rat model of 107. Mei JM, Niu CS. Effects of CDNF on 6-OHDA-induced apoptosis in PC12 cells via
Parkinson’s disease. Exp Neurol. 2011;228:99–108. modulation of Bcl-2/Bax and caspase-3 activation. Neurol Sci. 2014;35:1275–80.
85. Airavaara M, Harvey BK, Voutilainen MH, Shen H, Chou J, Lindholm P, et al. CDNF 108. Zhao H, Cheng L, Liu Y, Zhang W, Maharjan S, Cui Z, et al. Mechanisms of anti-
protects the nigrostriatal dopamine system and promotes recovery after MPTP inflammatory property of conserved dopamine neurotrophic factor: inhibition of JNK
treatment in mice. Cell Transplant. 2012;21:1213–23. signaling in lipopolysaccharide-induced microglia. J Mol Neurosci. 2014;52:186–92.
86. Back S, Peranen J, Galli E, Pulkkila P, Lonka-Nevalaita L, Tamminen T, et al. Gene 109. Nadella R, Voutilainen MH, Saarma M, Gonzalez-Barrios JA, Leon-Chavez BA,
therapy with AAV2-CDNF provides functional benefits in a rat model of Par- Jimenez JM, et al. Transient transfection of human CDNF gene reduces the 6-
kinson’s disease. Brain Behav. 2013;3:75–88. hydroxydopamine-induced neuroinflammation in the rat substantia nigra. J
87. Cordero-Llana O, Houghton BC, Rinaldi F, Taylor H, Yanez-Munoz RJ, Uney JB, Neuroinflammation. 2014;11:209.
et al. Enhanced efficacy of the CDNF/MANF family by combined intranigral 110. Garea-Rodriguez E, Eesmaa A, Lindholm P, Schlumbohm C, Konig J, Meller B, et al.
overexpression in the 6-OHDA rat model of Parkinson’s disease. Mol Ther. Comparative analysis of the effects of neurotrophic factors CDNF and GDNF in a
2015;23:244–54. nonhuman primate model of Parkinson’s disease. PLoS ONE. 2016;11:e0149776.
88. Ren X, Zhang T, Gong X, Hu G, Ding W, Wang X. AAV2-mediated striatum 111. Ganapathy Subramanian K, Dettmer A, Rockcastle N, Zhang Z, Doyle A, Noor-
delivery of human CDNF prevents the deterioration of midbrain dopamine baksh A, et al. Intraputamenal CDNF infusions restore substantia nigra dopa-
neurons in a 6-hydroxydopamine induced parkinsonian rat model. Exp Neurol. mine neuron integrity in the monkey low-dose MPTP model of Parkinson’s
2013;248:148–56. disease. Society for Neuroscience Annual Meeting. Presentation 331:07/L3. 2013.
89. Voutilainen MH, De Lorenzo F, Stepanova P, Back S, Yu LY, Lindholm P, et al. 112. Huttunen HJ, Booms S, Kerstens V, Fazio P, Johansson J, Koskinen J, et al. Phase
Evidence for an additive neurorestorative effect of simultaneously administered I-II first-in-man clinical trial of intraputamenal CDNF in Parkinson’s disease:
CDNF and GDNF in hemiparkinsonian rats: implications for different mechanism topline results of the 12-month treatment period. AD/PD™ Abstract. 2021.
of action. eNeuro. 2017;4:ENEURO.0117-16.2017. 113. Stepanov V, Krasikova R, Raus L, Loog O, Hiltunen J, Halldin C. An efficient one-
90. Choi JM, Hong JH, Chae MJ, Ngyuen PH, Kang HS, Ma HI, et al. Analysis of step radiosynthesis of [18F]FE-PE2I, a PET radioligand for imaging of dopamine
mutations and the association between polymorphisms in the cerebral dopa- transporters. J Label Compd Radiopharm. 2012;55:206–10.
mine neurotrophic factor (CDNF) gene and Parkinson disease. Neurosci Lett. 114. Zhang Y, Xiang Y, Wang X, Zhu L, Li H, Wang S, et al. Cerebral dopamine
2011;493:97–101. neurotrophic factor protects microglia by combining with AKT and by reg-
91. Yang Y, Yu H, Li W, Liu B, Zhang H, Ding S, et al. Association between cerebral ulating FoxO1/mTOR signaling during neuroinflammation. Biomed Pharmac-
dopamine neurotrophic factor (CDNF) 2 polymorphisms and schizophrenia other. 2019;109:2278–84.
susceptibility and symptoms in the Han Chinese population. Behav Brain Funct. 115. Albert K, Raymundo DP, Panhelainen A, Eesmaa A, Shvachiy L, Araujo GR, et al.
2018;14:1. Cerebral dopamine neurotrophic factor reduces alpha-synuclein aggregation
92. Montaser H, Patel KA, Balboa D, Ibrahim H, Lithovius V, Naatanen A, et al. Loss of and propagation and alleviates behavioral alterations in vivo. Mol Ther.
MANF causes childhood-onset syndromic diabetes due to increased endo- 2021;29:2821–40.
plasmic reticulum stress. Diabetes 2021;70:1006–18. 116. Zhao H, Cheng L, Du X, Hou Y, Liu Y, Cui Z, et al. Transplantation of cerebral
93. Yavarna T, Al-Dewik N, Al-Mureikhi M, Ali R, Al-Mesaifri F, Mahmoud L, et al. High dopamine neurotrophic factor transducted BMSCs in contusion spinal cord
diagnostic yield of clinical exome sequencing in Middle Eastern patients with injury of rats: promotion of nerve regeneration by alleviating neuroinflamma-
Mendelian disorders. Hum Genet. 2015;134:967–80. tion. Mol Neurobiol. 2016;53:187–99.
94. Herranen A, Ikaheimo K, Lankinen T, Pakarinen E, Fritzsch B, Saarma M, et al. 117. Huotarinen A, Penttinen AM, Back S, Voutilainen MH, Julku U, Piepponen TP,
Deficiency of the ER-stress-regulator MANF triggers progressive outer hair cell et al. Combination of CDNF and deep brain stimulation decreases neurological
death and hearing loss. Cell Death Dis. 2020;11:100. deficits in late-stage model Parkinson’s disease. Neuroscience 2018;374:250–63.
95. Pakarinen E, Danilova T, Voikar V, Chmielarz P, Piepponen P, Airavaara M, et al. 118. Mei J, Niu C. Effects of engineered conserved dopamine neurotrophic factor-
MANF ablation causes prolonged activation of the UPR without neurodegeneration expressing bone marrow stromal cells on dopaminergic neurons following
in the mouse midbrain dopamine system. eNeuro. 2020;7:ENEURO.0477-19.2019. 6-OHDA administrations. Mol Med Rep. 2015;11:1207–13.

Molecular Psychiatry (2022) 27:1310 – 1321

 P. Lindholm and M. Saarma
1321
119. Wang L, Wang Z, Xu X, Zhu R, Bi J, Liu W, et al. Recombinant AAV8-mediated 141. Cnop M, Toivonen S, Igoillo-Esteve M, Salpea P. Endoplasmic reticulum stress
intrastriatal gene delivery of CDNF protects rats against methamphetamine and eIF2alpha phosphorylation: the Achilles heel of pancreatic beta cells. Mol
neurotoxicity. Int J Med Sci. 2017;14:340–7. Metab. 2017;6:1024–39.
120. Kemppainen S, Lindholm P, Galli E, Lahtinen HM, Koivisto H, Hamalainen E, et al. 142. Hetz C, Zhang K, Kaufman RJ. Mechanisms, regulation and functions of the
Cerebral dopamine neurotrophic factor improves long-term memory in APP/ unfolded protein response. Nat Rev Mol Cell Biol. 2020;21:421–38.
PS1 transgenic mice modeling Alzheimer’s disease as well as in wild-type mice. 143. Karagoz GE, Acosta-Alvear D, Walter P. The unfolded protein response: detect-
Behav Brain Res. 2015;291:1–11. ing and responding to fluctuations in the protein-folding capacity of the
121. De Lorenzo F, Lüningschrör P, Nam J, Pilotto F, Galli E, Lindholm P, et al. CDNF endoplasmic reticulum. Cold Spring Harb Perspect Biol. 2019;11:a033886.
rescues motor neurons in three animal models of ALS by targeting ER stress. 144. Preissler S, Ron D. Early events in the endoplasmic reticulum unfolded protein
bioRxiv: 2020.05.05.078618 [Preprint]. 2020. Available from: https://www.biorxiv. response. Cold Spring Harb Perspect Biol. 2019;11:a033894.
org/content/10.1101/2020.05.05.078618v2.
122. Stepanova P, Srinivasan V, Lindholm D, Voutilainen MH. Cerebral dopamine
neurotrophic factor (CDNF) protects against quinolinic acid-induced toxicity in ACKNOWLEDGEMENTS
in vitro and in vivo models of Huntington’s disease. Sci Rep. 2020;10:19045. We thank Dr. Mikko Airavaara, Dr. Barry J. Hoffer, Dr. Vera Kovaleva, Dr. Maria Lindahl,
123. Kotzbauer PT, Lampe PA, Heuckeroth RO, Golden JP, Creedon DJ, Johnson EM and Dr. Raimo Tuominen for comments on the manuscript. Images in Figs. 2–4 were
Jr., et al. Neurturin, a relative of glial-cell-line-derived neurotrophic factor. Nature created with BioRender.com. Fig 2A was adapted from “UPR Signaling (ATF6,
1996;384:467–70. PERK, IRE1)”, by BioRender.com (2021). Retrieved from https://app.biorender.com/
124. Lin LF, Doherty DH, Lile JD, Bektesh S, Collins F. GDNF: a glial cell line-derived biorender-templates.
neurotrophic factor for midbrain dopaminergic neurons. Science 1993;260:1130–2.
125. Lonka-Nevalaita L, Lume M, Leppanen S, Jokitalo E, Peranen J, Saarma M. Char-
acterization of the intracellular localization, processing, and secretion of two glial
cell line-derived neurotrophic factor splice isoforms. J Neurosci. 2010;30:11403–13. AUTHOR CONTRIBUTIONS
126. Eigenbrot C, Gerber N. X-ray structure of glial cell-derived neurotrophic factor M.S. prepared an initial draft. M.S. and P.L. together wrote the paper, and P.L.
at 1.9 A resolution and implications for receptor binding. Nat Struct Biol. 1997; prepared the figures. Both authors edited the manuscript and accepted the final
4:435–8. version.
127. Sandmark J, Dahl G, Oster L, Xu B, Johansson P, Akerud T, et al. Structure and
biophysical characterization of the human full-length neurturin-GFRa2 complex:
a role for heparan sulfate in signaling. J Biol Chem. 2018;293:5492–508. FUNDING
128. Lindholm P, Saarma M. Novel CDNF/MANF family of neurotrophic factors. Dev This work was supported by Jane and Aatos Erkko foundation and the Academy of
Neurobiol. 2010;70:360–71. Finland (grant number 1310891). Open Access funding provided by University of
129. Runeberg-Roos P, Piccinini E, Penttinen AM, Matlik K, Heikkinen H, Kuure S, et al. Helsinki including Helsinki University Central Hospital.
Developing therapeutically more efficient neurturin variants for treatment of
Parkinson’s disease. Neurobiol Dis. 2016;96:335–45.
130. Bespalov MM, Sidorova YA, Tumova S, Ahonen-Bishopp A, Magalhaes AC, COMPETING INTERESTS
Kulesskiy E, et al. Heparan sulfate proteoglycan syndecan-3 is a novel receptor P.L. and M.S. are inventors in the CDNF- and MANF-related patents (7,452,969;
for GDNF, neurturin, and artemin. J Cell Biol. 2011;192:153–69. 9,592,270) owned by Herantis Pharma Plc. M.S. is a minority shareholder in Herantis
131. Parkash V, Leppanen VM, Virtanen H, Jurvansuu JM, Bespalov MM, Sidorova YA, et al. Pharma Plc. Intellectual property concerning CDNF fragments belongs to the
The structure of the glial cell line-derived neurotrophic factor-coreceptor complex: University of Helsinki.
insights into RET signaling and heparin binding. J Biol Chem. 2008;283:35164–72.
132. Hamilton JF, Morrison PF, Chen MY, Harvey-White J, Pernaute RS, Phillips H, et al.
Heparin coinfusion during convection-enhanced delivery (CED) increases the
ADDITIONAL INFORMATION
distribution of the glial-derived neurotrophic factor (GDNF) ligand family in rat
striatum and enhances the pharmacological activity of neurturin. Exp Neurol. Correspondence and requests for materials should be addressed to Mart Saarma.
2001;168:155–61.
133. Rosenblad C, Kirik D, Bjorklund A. Neurturin enhances the survival of intrastriatal Reprints and permission information is available at http://www.nature.com/
fetal dopaminergic transplants. Neuroreport 1999;10:1783–7. reprints
134. Piccinini E, Kalkkinen N, Saarma M, Runeberg-Roos P. Glial cell line-derived
neurotrophic factor: characterization of mammalian posttranslational mod- Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims
ifications. Ann Med. 2013;45:66–73. in published maps and institutional affiliations.
135. Horger BA, Nishimura MC, Armanini MP, Wang LC, Poulsen KT, Rosenblad C,
et al. Neurturin exerts potent actions on survival and function of midbrain
dopaminergic neurons. J Neurosci. 1998;18:4929–37.
136. Yu LY, Jokitalo E, Sun YF, Mehlen P, Lindholm D, Saarma M, et al. GDNF-deprived Open Access This article is licensed under a Creative Commons
sympathetic neurons die via a novel nonmitochondrial pathway. J Cell Biol. Attribution 4.0 International License, which permits use, sharing,
2003;163:987–97. adaptation, distribution and reproduction in any medium or format, as long as you give
137. Gardner BM, Walter P. Unfolded proteins are Ire1-activating ligands that directly appropriate credit to the original author(s) and the source, provide a link to the Creative
induce the unfolded protein response. Science 2011;333:1891–4. Commons license, and indicate if changes were made. The images or other third party
138. Karagoz GE, Acosta-Alvear D, Nguyen HT, Lee CP, Chu F, Walter P. An unfolded material in this article are included in the article’s Creative Commons license, unless
protein-induced conformational switch activates mammalian IRE1. eLife. 2017;6: indicated otherwise in a credit line to the material. If material is not included in the
e30700. article’s Creative Commons license and your intended use is not permitted by statutory
139. Wang P, Li J, Tao J, Sha B. The luminal domain of the ER stress sensor protein regulation or exceeds the permitted use, you will need to obtain permission directly
PERK binds misfolded proteins and thereby triggers PERK oligomerization. J Biol from the copyright holder. To view a copy of this license, visit http://creativecommons.
Chem. 2018;293:4110–21. org/licenses/by/4.0/.
140. Cullinan SB, Zhang D, Hannink M, Arvisais E, Kaufman RJ, Diehl JA. Nrf2 is a
direct PERK substrate and effector of PERK-dependent cell survival. Mol Cell Biol.
© The Author(s) 2021
2003;23:7198–209.

Molecular Psychiatry (2022) 27:1310 – 1321