- Email: [email protected]
Conditional loss of progranulin in neurons is not sufficient to cause neuronal ceroid lipofuscinosis-like neuropathology in mice
Accepted Manuscript Conditional loss of progranulin in neurons is not sufficient to cause neuronal ceroid lipofuscinosis-like neuropathology in mice
Terri L. Petkau, Jake Blanco, Blair R. Leavitt PII: DOI: Reference:
S0969-9961(17)30139-0 doi: 10.1016/j.nbd.2017.06.012 YNBDI 3982
To appear in:
Neurobiology of Disease
Received date: Revised date: Accepted date:
14 March 2017 1 June 2017 20 June 2017
Please cite this article as: Terri L. Petkau, Jake Blanco, Blair R. Leavitt , Conditional loss of progranulin in neurons is not sufficient to cause neuronal ceroid lipofuscinosis-like neuropathology in mice, Neurobiology of Disease (2017), doi: 10.1016/j.nbd.2017.06.012
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT
Conditional loss of progranulin in neurons is not sufficient to cause
T
neuronal ceroid lipofuscinosis-like neuropathology in mice.
aCentre
CR
IP
Terri L. Petkaua, Jake Blancoa, and Blair R. Leavitta,b,c,§
for Molecular Medicine & Therapeutics, Department of Medical Genetics,
of Neurology, Department of Medicine, University of British Columbia Hospital,
M
bDivision
AN
Avenue, Vancouver, BC, Canada V5Z 4H4
US
University of British Columbia, and Children’s and Women’s Hospital, 980 West 28 th
Research Centre, University of British Columbia, Vancouver, BC V6T 1Z3,
PT
cBrain
ED
S 192 - 2211 Wesbrook Mall, Vancouver, BC, Canada V6T 2B5
CE
Canada §Corresponding
Author: Blair R. Leavitt, Centre for Molecular Medicine &
AC
Therapeutics, Department of Medical Genetics, University of British Columbia, and Children’s and Women’s Hospital, 980 West 28th Avenue, Vancouver, BC, Canada V5Z 4H4. Tel: 604-875-3801, Fax: 604-875-3819, email: [email protected]
ACCEPTED MANUSCRIPT
Abstract Progranulin deficiency due to heterozygous null mutations in the GRN gene is a common cause of familial frontotemporal lobar degeneration (FTLD), while homozygous
T
loss-of-function GRN mutations cause neuronal ceroid lipofuscinosis (NCL). Aged
IP
progranulin-knockout mice display highly exaggerated lipofuscinosis, microgliosis, and
CR
astrogliosis, as well as mild cell loss in specific brain regions. Progranulin is a secreted glycoprotein expressed in both neurons and microglia, but not astrocytes, in the brain.
US
We generated conditional progranulin-knockout mice that lack progranulin in nestin-
AN
expressing cells (Nes-cKO mice), which include most neurons as well as astrocytes. We confirmed near complete knockout of progranulin in neurons in Nes-cKO mice, while
M
microglial progranulin levels remained similar to that of wild-type animals. Overall brain
ED
progranulin levels were reduced by about 50% in Nes-cKO, and no Grn was detected in primary Nes-cKO neurons. Nes-cKO mice aged to 12 months did not display any
PT
increase in lipofuscin deposition, microgliosis, or astrogliosis in the four brain regions
CE
examined, though increases were observed for most of these measures in Grn-null animals. We conclude that neuron-specific loss of progranulin is not sufficient to cause
AC
similar neuropathological changes to those seen in constitutive Grn-null animals. Our results suggest that increased lipofuscinosis and gliosis in Grn-null animals are not caused by intrinsic progranulin deficiency in neurons, and that microglia-derived progranulin may be sufficient to maintain neuronal health and homeostasis in the brain.
Keywords: frontotemporal lobar degeneration, neuronal ceroid lipofuscinosis, progranulin, conditional knockout mice, neuropathology, nestin
2
ACCEPTED MANUSCRIPT
Introduction Progranulin is a pleiotropic, secreted growth factor with widespread expression throughout the periphery and in the brain (Petkau & Leavitt 2014). Within the central
T
nervous system, progranulin is a putative lysosome-associated protein (Tanaka et al
IP
2014, Tanaka et al 2017) expressed in most neuronal populations and in microglia
CR
(Baker et al 2006, Chen-Plotkin et al 2010, Petkau et al 2010), and reduced expression
US
of progranulin is associated with multiple neurodegenerative diseases (Petkau & Leavitt 2014). Most notably, heterozygous loss-of-function mutations in the progranulin gene
AN
(GRN) are a common cause of familial frontotemporal lobar degeneration (FTLD)
M
(Baker et al 2006, Cruts et al 2006), while homozygous loss of GRN expression causes neuronal ceroid lipofuscinosis (NCL) (Smith et al 2012).
ED
Progranulin expression, both in humans and in mice, is present in a punctate,
PT
perinuclear pattern in neurons throughout the brain, with highest expression occurring in the thalamus and CA3 layer of the hippocampus, and without obvious changes in
CE
expression in response to injury (Petkau et al 2010). In microglia, expression of
AC
progranulin is notably stronger than in neurons, and is further increased when microglia are activated after stress or injury (Chen-Plotkin et al 2010, Naphade et al 2010, Petkau et al 2010).
Mice homozygous for a targeted deletion of the mouse progranulin gene (Grn) develop a robust neuropathological phenotype with age (Ahmed et al 2010, Ghoshal et al 2012, Petkau et al 2012, Wils et al 2012, Yin et al 2010b), though heterozygous Grntargeted mice do not develop this phenotype (Ahmed et al 2010, Petkau et al 2012).
3
ACCEPTED MANUSCRIPT
Aged Grn-null mice display exaggerated deposition of the aging pigment lipofuscin and accumulation of a NCL-like storage material, as well as robust astrogliosis and microgliosis in multiple brain regions (Ahmed et al 2010, Ghoshal et al 2012, Petkau et
T
al 2016, Wils et al 2012, Yin et al 2010b). In addition, mild neuronal cell loss (Ghoshal et
IP
al 2012) and alterations in neuronal morphology have also been reported (Petkau et al
CR
2016, Petkau et al 2012). Grn-null immune cells, such as macrophages and microglia, show a hyper-inflammatory response to various stimuli (Martens et al 2012, Suh et al
US
2012, Yin et al 2010a), as well as altered lysosomal biogenesis (Tanaka et al 2013a,
AN
Tanaka et al 2013b, Tanaka et al 2017). Grn deficiency in neurons has been associated with reduced neuronal survival (De Muynck et al 2013, Guo et al 2010, Kleinberger et al
M
2010, Ryan et al 2009, Van Damme et al 2008), reduced neurite outgrowth (Van
ED
Damme et al 2008), and altered lysosome function (Tanaka et al 2017). In this study, we sought to determine the relative contribution of progranulin
PT
derived from neurons to the robust neuropathological phenotype present in constitutive
CE
Grn-null mice. We created neuron-specific conditional Grn knockout mice (Nes-cKO) and evaluated whether they developed lipofuscinosis, microgliosis, and astrogliosis in
AC
the brain with age. Surprisingly, we find that near complete loss of Grn expression in most neurons is not sufficient to reproduce any of the neuropathological phenotypes observed in constitutive Grn-null mice, and conclude that microglia-derived Grn is sufficient to maintain brain homeostasis in terms of the outcomes measured here. Methods Mice
4
ACCEPTED MANUSCRIPT
The generation of ‘floxed’ progranulin-targeted (Floxed) mice was previously described (Petkau et al 2013). Conditional nestin-knockout (Nes-cKO) mice were generated by crossing homozygous Grnflox/flox mice on the C57Bl/6 background to
T
transgenic mice expressing Cre recombinase under the control of the rat nestin
IP
promoter and enhancer (generous gift from Dr. E. Simpson; originally from The Jackson
CR
Laboratory, strain name: B6.Cg-Tg(Nes-cre)1Kln/J). Final experimental cohorts were generated by crossing Grnflox/flox, Cre+ animals to Grnflox/flox, Cre- littermates. Genotyping
US
was performed on tail tip DNA at wean and confirmed on a second DNA sample at sacrifice using the following primer sequences: Grn common forward primer: 5’-
AN
CGGAACACAGTGTCCAGATG-3’; Grn intron 2 reverse primer: 5’-
M
ATCAACCAAAGGGTCTGTGC-3’; Grn exon 5 reverse primer: 5’-
ED
GTGGCAGAGTCAGGACATTCAAACT-3’; Cre forward primer: 5’GCGGTCTGGCAGTAAAAACTATC-3’; Cre reverse primer: 5’-
PT
GTGAAACAGCATTGCTGTCACTT-3’. To generate Het-Nes-cKO mice, we crossed
CE
Grnflox/flox, Cre+ mice to constitutive Grn-/- mice and genotyped the offspring for the presence of the Cre transgene.
AC
Mice were housed on ventilated racks in specific pathogen-free barrier facility with a 12hr light/dark cycle. Mice were group-housed with their littermates to a maximum of four mice per cage. All animal procedures were done with the approval of the Canadian Council for Animal Care and the University of British Columbia’s Animal Care Committee.
5
ACCEPTED MANUSCRIPT
Protein Extraction and quantification of Grn by ELISA Whole brain lysate was prepared by homogenizing previously snap-frozen brains in a rotor-stator homogenizer for 30 sec. in 1mL of complete lysis buffer (50mM Tris-
T
HCl, 1% Triton-X, 150mM NaCl, Halt phosphatase inhibitor cocktail (Thermo Fisher
IP
Scientific), Halt protease inhibitor cocktail (Thermo Fisher Scientific)). Primary cortical
CR
neurons were generated from embryonic day 17.5 pups as previously described. Neurons cultured for 7 days in 6-well plates were washed twice with ice-cold PBS
US
before being scraped into 60L of complete RIPA buffer (50mM Tris-HCl, 1% Triton-X,
AN
150mM NaCl, 1% deoxycholic acid, 1mM EDTA, Halt phosphatase inhibitor cocktail (Thermo Fisher Scientific), Halt protease inhibitor cocktail (Thermo Fisher Scientific))
M
with DNase added. The lysate was collected into pre-chilled microcentrifuge tubes and
ED
stored at -80oC until use. Total protein was assayed using Bradford reagent (BioRad). Primary microglia cultures were generated as previously described (Connolly et al
PT
2016). Conditioned media was collected after 24hrs and stored at -20oC until used.
CE
The quantity of Grn in whole brain lysate, cell lysate, or conditioned media was determined by an enzyme-linked immunosorbent assay (ELISA) using a commercially
AC
available kit (Mouse progranulin ELISA; Adipogen, Korea). Microglia supernatant samples were diluted 1:5; for neuronal cell lysates, 20 g of total protein was loaded per well, and for whole brain lysate, 100g of protein was used. All samples were run in duplicate. The ELISA was conducted according to the manufacturer’s instructions. Data represent the average of 2-8 samples per condition, and all conditions that were compared directly were run on the same plate.
6
ACCEPTED MANUSCRIPT
RNA isolation and qPCR For analysis of Grn mRNA, cortical tissue from 4 Floxed, 5 Nes-cKO, and 2
T
GrnKO mice at 4-6 months of age was collected and immediately frozen at −80°C.
IP
Samples were homogenized with a bead homogenizer in lysis buffer followed by total
CR
RNA extraction (PureLink RNA mini kit; Invitrogen) performed according to the manufacturer’s instructions. Reverse transcription of all samples was carried out using
US
the Superscript VILO kit (Invitrogen) according to the manufacturer's instructions, using
AN
1g of total RNA as input for cDNA synthesis. Following this, cDNA was diluted 1:10 in ddH2O for a total input of 5ng into the quantitative PCR reaction, done using FastSybr
M
(Applied Biosystems) and conducted on a Step-One ABI System (Applied Biosystems).
ED
Quantification of mRNA levels was accomplished using the standard curve method, with amplification of target mRNA and control genes in separate wells. Each sample was run
PT
in duplicate. The relative amount of mRNA in each well was calculated as the ratio
CE
between Grn mRNA (forward: 5’-CTGTAGTGCAGATGGGAAATCCTGCT-3’; reverse: 5’-GTGGCAGAGTCAGGACATTCAAACT-3’) and a normalization factor created using
AC
two control genes, Usf1 (forward: 5’-CCTGTGGCGTGGCAGTCT-3’; reverse: 5’TGCACGCCCACACTGTTT-3’) and Paklip1 (forward: 5’CCCCAAGTGGAGGGAAGTACA-3’; reverse: 5’-TGCCCAGCCGATAGACATC-3’) based on GeNorm (Vandesompele et al 2002). Values are presented as % Floxed control.
Immunohistochemistry
7
ACCEPTED MANUSCRIPT
Serial 25m floating sections were placed in net-well inserts and washed for 10min in phosphate-buffered saline (PBS). Endogenous peroxidase activity was quenched with 1% H2O2 for 45min. After a 15min wash in PBS with 0.1% Triton X (PBS-
T
T), sections were blocked in 5% normal serum and 5% bovine serum albumin diluted in
IP
PBS-T, followed by overnight incubation shaking at room temperature in primary
CR
antibody diluted in 5% normal serum and PBS-T. After two 15min washes in PBS-T, secondary antibody diluted in 1% normal serum and PBS-T was applied for 2hrs
US
shaking at room temperature. Sections were washed for 30min in PBS before an
AN
amplification step was performed using an avidin–biotin–horseradish peroxidase complex kit (Vector Laboratories). Colorimetric detection was achieved with the
M
peroxidase substrate kit Vector DAB (Vector Laboratories) according to the
ED
manufacturer’s instructions. Sections were mounted by hand on onto glass slides (Fisherbrand Superfrost Plus) and dried overnight before being dehydrated through a
PT
series of alcohols and xylene, and cover-slipped with DEPEX (Electron Microscopy
CE
Sciences). Antibodies used were as follows: the neuronal marker NeuN (Chemicon, Millipore, 1:2000, mouse monoclonal), the microglia marker Iba1 (Wako; 1:2000, rabbit
AC
polyclonal), the astrocyte marker GFAP (Sigma; 1:2000, mouse monoclonal), and appropriate biotinylated secondary antibodies (Vector, 1:2000). Immunofluorescent staining was performed by incubating mounted, 25m serial sections in PBS-T for 30 minutes, followed by 1hr in blocking solution (5% normal donkey serum in PBS-T) and subsequent overnight incubation at 4oC with primary antibodies in 2% normal serum. Primary antibodies included NeuN (mouse monoclonal,
8
ACCEPTED MANUSCRIPT
Millipore, 1:1000), Iba1 (rabbit polyclonal, Wako, 1:1000), and Grn (sheep polyclonal, R&D Systems, 1:250). Section were washed twice for 5min each prior to incubation with appropriate secondary antibodies for 1hr at room temperature in 2% normal donkey
T
serum (Donkey anti-mouse Alexa 488, donkey anti-rabbit Alexa 488, donkey anti-sheep
IP
Alexa 594, Molecular Probes, 1:1000). After washing again twice in PBS for 5 min,
CR
sections were incubated with DAPI (4',6-diamidino-2-phenylindole, dihydrochloride) in PBS at 1:10 000 for 5 min, washed twice in PBS again for 5 min each, and then
US
dehydrated and coverslipped with fluorescent DEPEX mounting medium.
AN
Sections to be assessed for autofluorescence were mounted onto glass slides, washed in PBS-T for 30 min, and then stained with DAPI in PBS at 1:10 000 for 5 min.
ED
Image acquisition and analysis
M
Slides were washed in twice in PBS for 5 min prior to coverslipping.
PT
Images were acquired as previously described (Petkau et al 2016). Integrated optical density measurements of signal intensity were acquired as previously described
CE
(Petkau et al 2010) to quantify autofluorescence, a surrogate for lipofuscin and/or NCL-
AC
like material deposition, in 4 images per mouse taken from 2-7 mice per genotype. For quantification of colorimetric stains (Iba1 and GFAP), a threshold was set that pseudocolored stained areas within a defined region of interest in each image. The average percent thresholded area for 4 images per mouse taken from 2-7 mice per genotype is reported. Statistical analysis
9
ACCEPTED MANUSCRIPT
All statistical comparisons were performed as a one-way analysis of variance (ANOVA) with Bonferroni post-hoc analysis to compare individual means to control and correct for multiple comparisons (Prism 6, Graphpad Software Inc.). A p-value less than
IP
T
0.05 was considered significant.
CR
Results Grn is ‘knocked down’ in Nes-cKO mice
US
First, we verified that Floxed mice expressed Grn in the brain at levels comparable to WT mice; ie. that the inserted loxP sites in the targeted allele do not
AN
interfere with normal gene expression. Grn levels quantified by ELISA on whole brain
M
homogenate showed similar levels between Floxed mice and WT mice (Fig.1A; p > 0.05
ED
by Bonferroni post-hoc analysis). Subsequent analyses used Floxed mice as controls, or data from Floxed and WT mice was combined. As expected, no Grn expression was
PT
detected in GrnKO mice (Fig.1A). Whole brain Grn levels were reduced by approximately 50% in Nes-cKO mice compared to Floxed mice (Fig.1A). Analysis of Grn
CE
mRNA expression in the cortex of Floxed and Nes-cKO mice showed a reduction of Grn
AC
in Nes-cKO mice to about 20% of Floxed mice (Fig.1B). Once again, no Grn expression was detected from GrnKO mice. Next, to examine Grn expression in specific cell types, we generated primary cultures of cortical neurons and whole brain microglia and measured Grn in the cell lysate (Fig.1C) or conditioned media (Fig.1D) respectively. As expected, Nes-cKO cortical neuron cultures expressed no Grn, while cultures derived
10
ACCEPTED MANUSCRIPT
from Floxed mice expressed Grn at WT levels (Fig.1D). Grn secreted into the media from microglia, was similar between Nes-cKO and Floxed cultures (Fig.1D). We next evaluated Grn expression by immunohistochemistry in brain sections
T
from Nes-cKO and Floxed mice. Similar to previous reports evaluating Grn staining in
IP
WT mice (Petkau et al 2010), robust Grn staining was present in NeuN-labeled neurons
CR
throughout the brain, including in the CA3 layer of the hippocampus of Floxed mice (Fig.2A-C). Brightly staining puncta that do not co-localize with NeuN correspond to
US
microglia (Fig1C; arrow). Brightly staining, Grn-positive microglia were easily identified
AN
in brain sections from Nes-cKO mice (Fig.2F, arrows), but neuronal Grn staining was drastically reduced, albeit not completely absent (Fig.2E). Given that no Grn was
M
detected in primary neuronal cultures from Nes-cKO mice, and that Grn is a secreted
ED
protein that can be taken up by neurons in culture (Hu et al 2010), it is likely that very low level of Grn staining present in neurons in Nes-cKO mice is internalized, microglia-
PT
derived Grn. Microglia in the cortex, labeled with Iba1 (Fig.2G), were co-labeled with
CE
Grn (Fig.2H), as expected, and looked similar in Nes-cKO mice (Fig.2J-L). Overall, our data show that Nes-cKO mice express little to no Grn in neurons, while microglial Grn
AC
expression is unaffected.
Neuropathology in Nes-cKO mice does not replicate that of GrnKO mice We previously completed a detailed analysis of behavior and neuropathology in aged GrnKO mice on a mixed genetic background (Petkau et al 2012) and on two different inbred background strains (Petkau et al 2016). We chose three robust
11
ACCEPTED MANUSCRIPT
neuropathological endpoints that clearly distinguish GrnKO animals on a C57Bl/6 background from their WT littermates and assessed them in Nes-cKO and matching Floxed and GrnKO littermates at 12 months of age. We analyzed lipofuscinosis,
T
astrogliosis, and microgliosis in four brain regions: the thalamus, CA3 layer of the
IP
hippocampus, striatum, and cortex.
CR
Lipofuscin is an autofluorescent pigment that accumulates with age, most predominantly in post-mitotic cells such as neurons (Sulzer et al 2008). Lipofuscin
US
accumulation, measured by quantifying autofluorescence, is detectable in the brains of
AN
Floxed mice (Fig.3), including in the thalamus (Fig.3A), CA3 layer of the hippocampus (Fig.3E), the cortex (Fig.3I), and the striatum (Fig.3M). Lipofuscin accumulation
M
occurred at levels similar to Floxed mice in all four brain regions in Nes-cKO mice
ED
(Fig.3B, F, J, N), but at significantly higher levels in GrnKO mice in the thalamus (Fig.3C, quantified in D) and CA3 region (Fig.3G, quantified in H). Lipofuscin in the
PT
cortex and striatum was not significantly higher in GrnKO animals compared to Floxed
CE
or Nes-cKO mice, though overall levels of lipofuscin in these regions was considerably lower than in the thalamus and CA3 layer of the hippocampus. Thus, despite clear
AC
evidence of increased lipofuscin deposition in two brain regions in GrnKO mice, no significant change in lipofuscin was observed when Grn was specifically knocked down in neurons.
Astrogliosis was assessed by quantifying the stained area of GFAP immunoreactivity in the same four brain regions. GFAP immunoreactivity in the thalamus was limited to occasional cells only in a given field of view in both Floxed
12
ACCEPTED MANUSCRIPT
(Fig.4A) and Nes-cKO (Fig.4B) mice, but large patches of reactive astrocytes occurred in GrnKO mice (Fig.4C) and overall percent stained area was significantly higher in GrnKO mice (Fig.4D). In the hippocampus, GFAP immunoreactivity was overall much
T
higher than in other brain regions. The stained area was significantly higher in GrnKO
IP
mice than in Floxed or Nes-cKO mice (Fig.4, E-H). GFAP immunoreactivity was also
CR
significantly higher in Nes-cKO mice than in Floxed mice (Fig.4H). The cortex and striatum showed a similar pattern to the thalamus, where Floxed (Fig.4I, M) and Nes-
US
cKO (Fig.4J, N) showed only occasional or sporadic GFAP-immunoreacitve cells, while
AN
large patches of reactive astrocytes could be found in GrnKO mice (Fig.4K, O). Quantification of GFAP immunostaining showed significantly higher percent area
M
stained in GrnKO mice compared to Floxed and Nes-cKO mice in both the cortex
ED
(Fig.4L) and striatum (Fig.4P). Similar to lipofuscinosis, while GrnKO mice show a clear neuropathological phenotype of increased astrocytosis in multiple brain regions, no
PT
changes were apparent in Nes-cKO mice.
CE
We evaluated microgliosis using Iba1 immunoreactivity in the thalamus, hippocampus, cortex and striatum. In the thalamus, Iba1-stained microglia were
AC
noticeably more numerous and had larger cell bodies in GrnKO mice (Fig.5C) compared to either Floxed (Fig.5A) or Nes-cKO (Fig.5B) mice, and the percent Iba1-stained area was significantly higher in GrnKO animals (Fig.5D). Qualitatively, a similar change in cell morphology (larger cell bodies and a less ramified appearance) could be appreciated in the GrnKO mice in the CA3 region of the hippocampus (Fig.5G), cortex (Fig.5K), and striatum (Fig.5O) compared to both Floxed (Fig.5A, E, M) and Nes-cKO
13
ACCEPTED MANUSCRIPT
(Fig.5B, F, N) mice, though quantification of the percent Iba1-stained area did not show a significant difference between the genotypes (Fig.5H, L, P). Although microgliosis was not as pronounced in GrnKO mice in this study as in previous studies (Petkau et al
T
2016, Petkau et al 2012), increased Iba1 immunoreactivity was observed in the
IP
thalamus of GrnKO mice but was distinctly absent from Nes-cKO mice.
CR
Further reduction of Grn levels in Het-Nes-cKO mice has no effect on neuropathology Finally, we considered the possibility that the Grn-dependent neuropathological
US
abnormalities of increased lipofuscinosis, astrogliosis, and microgliosis might only occur
AN
when overall Grn levels are reduced below the level seen in Nes-cKO mice, which is approximately 50% of WT mice (Fig.1A). To test this hypothesis, we generated mice
M
carrying one constitutively recombined, Grn-null allele and one Grn ‘floxed’ allele
ED
(Grnflox/-) along with the nestin-driven Cre transgene (Het-Nes-cKO). These mice are constitutively heterozygous Grn-null mice with further neural-specific Grn knockdown in
PT
Cre-expressing mice. When whole brain Grn levels were measured in brain lysate by
CE
Grn ELISA, we found the overall level of Grn in Het-Nes-cKO mice to be approximately 25% of WT or Floxed levels, or 50% less again than Nes-cKO animals (Fig.6A). We
AC
quantified lipofuscinosis (Fig.6B), astrogliosis (Fig.6C), and microgliosis (Fig.6D) in the thalamus of Grn-Het and Het-Nes-cKO animals and compared it to WT, Floxed, NescKO, and GrnKO animals. Even with just 25% of normal Grn levels in the brain, and likely nearly no Grn produced by neurons (based on near-complete loss of Grn expression in primary neurons even in Nes-cKO mice, see Fig.1C), we observed no
14
ACCEPTED MANUSCRIPT
significant difference in any of these measures in Het-Nes-cKO animals compared to Floxed controls. Overall, the data indicate that neural-specific knockdown of Grn in mice does not
T
cause the same neuropathological phenotypes that are observed in constitutive Grn-null
CR
IP
animals.
Discussion
US
This is the first study, to our knowledge, to assess the phenotype of conditional
AN
Grn-null animals where Grn has been selectivity knocked out of neural cells in mice. Despite showing near-complete knockout in isolated neurons and substantially reduced
M
overall brain Grn levels, Nes-cKO mice do not develop any of the neuropathological
ED
phenotypes that are present in aged Grn-null mice. Given that Grn expression in Nes-cKO neurons is essentially absent, the
PT
remaining Grn present in the brain is likely derived from microglia. Although neurons
CE
greatly outnumber microglia in terms of absolute numbers in the brain, Nes-cKO mice Grn expression levels are reduced only by about 50% compared to WT mice,
AC
suggesting that microglia contribute the other 50%. Progranulin is a secreted protein (Suh et al 2012) and at least one putative Grn receptor on neurons has been identified (Hu et al 2010), so it is feasible that microglial-derived Grn could compensate for loss of intrinsic Grn expression in neurons. If this is the case, it suggests an important role for microglia-derived Grn, and future experiments evaluating neuropathology in microgliaspecific Grn knockout mice could greatly improve our understanding of how different
15
ACCEPTED MANUSCRIPT
sources of brain Grn contribute to brain homeostasis. Furthermore, given that there are considerable efforts being invested in evaluating Grn-modulating therapeutics and their clinical potential, it will be of substantial clinical importance to ensure that drug screens
T
for such compounds consider Grn regulation in microglia as well as in neurons.
IP
Lipofuscin accumulation, an early and robust phenotype in constitutive Grn-null
CR
mice (Ahmed et al 2010, Ghoshal et al 2012, Petkau et al 2016, Petkau et al 2012, Wils et al 2012, Yin et al 2010a, Yin et al 2010b), occurs largely in post-mitotic neurons.
US
Lipofuscinosis is often caused by dysfunctional lysosomal components (Jalanko &
AN
Braulke 2009), and recent studies have implicated a role for Grn in lysosomal function (Tanaka et al 2014, Tanaka et al 2017). That Nes-cKO mice do not display any signs of
M
increased lipofuscin deposition implies that cell-intrinsic Grn production is not required
ED
to maintain lysosomal function. Furthermore, it suggests that Grn may contribute to lysosomal function via external signaling pathways. It remains unclear whether intrinsic
PT
Grn production in neurons has additional and/or alternative roles in neuronal health, or if
CE
the cellular source of Grn is dispensable and all Grn-dependent functions in neurons are carried out via extracellular Grn production and signaling. Data from FTLD patients with
AC
GRN mutations shows, quite strikingly, that human progranulin expression actually increases in affected brain regions, and that this increase can be attributed to increased expression from infiltrating microglia (Chen-Plotkin et al 2010). This data alone provides strong support for the first alternative, that Grn plays additional and/or alternative roles in neuronal cell biology that have yet to be fully elucidated.
16
ACCEPTED MANUSCRIPT
An alternative explanation for the discrepancy in neuropathology between mice constitutively lacking Grn and mice with neural-specific loss of Grn is that the phenotype in Grn-null animals is developmental. This hypothesis predicts that any Grn expression
T
in the brain, or perhaps even transient expression of Grn in the brain during
IP
development, is sufficient to prevent the development of lipofuscinosis, astrogliosis, and
CR
microgliosis later in life. It remains to be tested, but this hypothesis is perhaps supported by data from heterozygous Grn-null mice (Ahmed et al 2010, Filiano et al 2013) and
US
Het-Nes-cKO mice (Fig.6), which shows quite clearly that the Grn-dependent
AN
neuropathological phenotypes measured here are not dose dependent. In conclusion, neural-specific knockdown of Grn in mice does not recapitulate the
M
neuropathological phenotypes seen in constitutive Grn-null animals. Despite being a
ED
relatively minor fraction of cells in the brain, microglia can contribute a substantial portion of overall brain Grn levels, which is seemingly sufficient to maintain brain
PT
homeostasis, or at least to prevent exaggerated lipofuscinosis, microgliosis, and
CE
astrogliosis in aged mice. This work highlights the importance of evaluating Grnmodulating therapies in microglia as well as neurons, and suggests that Grn may play
AC
alternative or additional roles in neuronal function.
Acknowledgements This work was supported by the Alzheimer Society of Canada (doctoral trainee award to TP), and by the Canadian Institute for Health Research (BRL operating grant #97857).
17
AC
CE
PT
ED
M
AN
US
CR
IP
T
ACCEPTED MANUSCRIPT
18
ACCEPTED MANUSCRIPT
REFERENCES
AC
CE
PT
ED
M
AN
US
CR
IP
T
Ahmed Z, Sheng H, Xu YF, Lin WL, Innes AE, et al. 2010. Accelerated lipofuscinosis and ubiquitination in granulin knockout mice suggest a role for progranulin in successful aging. Am J Pathol 177: 311-24 Baker M, Mackenzie IR, Pickering-Brown SM, Gass J, Rademakers R, et al. 2006. Mutations in progranulin cause tau-negative frontotemporal dementia linked to chromosome 17. Nature 442: 916-9 Chen-Plotkin AS, Xiao J, Geser F, Martinez-Lage M, Grossman M, et al. 2010. Brain progranulin expression in GRN-associated frontotemporal lobar degeneration. Acta Neuropathol 119: 111-22 Connolly C, Magnusson-Lind A, Lu G, Wagner PK, Southwell AL, et al. 2016. Enhanced immune response to MMP3 stimulation in microglia expressing mutant huntingtin. Neuroscience 325: 74-88 Cruts M, Kumar-Singh S, Van Broeckhoven C. 2006. Progranulin mutations in ubiquitinpositive frontotemporal dementia linked to chromosome 17q21. Curr Alzheimer Res 3: 485-91 De Muynck L, Herdewyn S, Beel S, Scheveneels W, Van Den Bosch L, et al. 2013. The neurotrophic properties of progranulin depend on the granulin E domain but do not require sortilin binding. Neurobiol Aging Filiano AJ, Martens LH, Young AH, Warmus BA, Zhou P, et al. 2013. Dissociation of frontotemporal dementia-related deficits and neuroinflammation in progranulin haploinsufficient mice. J Neurosci 33: 5352-61 Ghoshal N, Dearborn JT, Wozniak DF, Cairns NJ. 2012. Core features of frontotemporal dementia recapitulated in progranulin knockout mice. Neurobiol Dis 45: 395-408 Guo A, Tapia L, Bamji SX, Cynader MS, Jia W. 2010. Progranulin deficiency leads to enhanced cell vulnerability and TDP-43 translocation in primary neuronal cultures. Brain Res 1366: 1-8 Hu F, Padukkavidana T, Vaegter CB, Brady OA, Zheng Y, et al. 2010. Sortilin-mediated endocytosis determines levels of the frontotemporal dementia protein, progranulin. Neuron 68: 654-67 Jalanko A, Braulke T. 2009. Neuronal ceroid lipofuscinoses. Biochim Biophys Acta 1793: 697-709 Kleinberger G, Wils H, Ponsaerts P, Joris G, Timmermans JP, et al. 2010. Increased caspase activation and decreased TDP-43 solubility in progranulin knockout cortical cultures. J Neurochem 115: 735-47 Martens LH, Zhang J, Barmada SJ, Zhou P, Kamiya S, et al. 2012. Progranulin deficiency promotes neuroinflammation and neuron loss following toxin-induced injury. J Clin Invest 122: 3955-9 Naphade SB, Kigerl KA, Jakeman LB, Kostyk SK, Popovich PG, Kuret J. 2010. Progranulin expression is upregulated after spinal contusion in mice. Acta Neuropathol 119: 123-33
19
ACCEPTED MANUSCRIPT
AC
CE
PT
ED
M
AN
US
CR
IP
T
Petkau TL, Hill A, Leavitt BR. 2016. Core neuropathological abnormalities in progranulin-deficient mice are penetrant on multiple genetic backgrounds. Neuroscience 315: 175-95 Petkau TL, Leavitt BR. 2014. Progranulin in neurodegenerative disease. Trends Neurosci 37: 388-98 Petkau TL, Neal SJ, Milnerwood A, Mew A, Hill AM, et al. 2012. Synaptic dysfunction in progranulin-deficient mice. Neurobiol Dis 45: 711-22 Petkau TL, Neal SJ, Orban PC, MacDonald JL, Hill AM, et al. 2010. Progranulin expression in the developing and adult murine brain. J Comp Neurol 518: 393147 Petkau TL, Zhu S, Lu G, Fernando S, Cynader M, Leavitt BR. 2013. Sensitivity to neurotoxic stress is not increased in progranulin-deficient mice. Neurobiol Aging Ryan CL, Baranowski DC, Chitramuthu BP, Malik S, Li Z, et al. 2009. Progranulin is expressed within motor neurons and promotes neuronal cell survival. BMC Neurosci 10: 130 Smith KR, Damiano J, Franceschetti S, Carpenter S, Canafoglia L, et al. 2012. Strikingly different clinicopathological phenotypes determined by progranulinmutation dosage. Am J Hum Genet 90: 1102-7 Suh HS, Choi N, Tarassishin L, Lee SC. 2012. Regulation of progranulin expression in human microglia and proteolysis of progranulin by matrix metalloproteinase-12 (MMP-12). PLoS One 7: e35115 Sulzer D, Mosharov E, Talloczy Z, Zucca FA, Simon JD, Zecca L. 2008. Neuronal pigmented autophagic vacuoles: lipofuscin, neuromelanin, and ceroid as macroautophagic responses during aging and disease. J Neurochem 106: 24-36 Tanaka Y, Chambers JK, Matsuwaki T, Yamanouchi K, Nishihara M. 2014. Possible involvement of lysosomal dysfunction in pathological changes of the brain in aged progranulin-deficient mice. Acta Neuropathol Commun 2: 78 Tanaka Y, Matsuwaki T, Yamanouchi K, Nishihara M. 2013a. Exacerbated inflammatory responses related to activated microglia after traumatic brain injury in progranulin-deficient mice. Neuroscience 231: 49-60 Tanaka Y, Matsuwaki T, Yamanouchi K, Nishihara M. 2013b. Increased Lysosomal Biogenesis in Activated Microglia and Exacerbated Neuronal Damage after Traumatic Brain Injury in Progranulin-Deficient Mice. Neuroscience Tanaka Y, Suzuki G, Matsuwaki T, Hosokawa M, Serrano G, et al. 2017. Progranulin regulates lysosomal function and biogenesis through acidification of lysosomes. Hum Mol Genet Van Damme P, Van Hoecke A, Lambrechts D, Vanacker P, Bogaert E, et al. 2008. Progranulin functions as a neurotrophic factor to regulate neurite outgrowth and enhance neuronal survival. J Cell Biol 181: 37-41 Vandesompele J, De Preter K, Pattyn F, Poppe B, Van Roy N, et al. 2002. Accurate normalization of real-time quantitative RT-PCR data by geometric averaging of multiple internal control genes. Genome Biol 3: RESEARCH0034
20
ACCEPTED MANUSCRIPT
AC
CE
PT
ED
M
AN
US
CR
IP
T
Wils H, Kleinberger G, Pereson S, Janssens J, Capell A, et al. 2012. Cellular ageing, increased mortality and FTLD-TDP-associated neuropathology in progranulin knockout mice. J Pathol 228: 67-76 Yin F, Banerjee R, Thomas B, Zhou P, Qian L, et al. 2010a. Exaggerated inflammation, impaired host defense, and neuropathology in progranulin-deficient mice. J Exp Med 207: 117-28 Yin F, Dumont M, Banerjee R, Ma Y, Li H, et al. 2010b. Behavioral deficits and progressive neuropathology in progranulin-deficient mice: a mouse model of frontotemporal dementia. FASEB J 24: 4639-47
21
ACCEPTED MANUSCRIPT
FIGURE LEGENDS
CR
IP
T
Figure 1. Progranulin is knocked down in Nes-cKO mice. (A) Grn measured from whole brain lysate by ELISA in 3-4 month old mice. Data are presented as mean SEM. N = 7 WT, 5 Floxed, 2 Nes-cKO, and 4 GrnKO. (B) Grn transcript measured by qPCR from RNA extracted from cortex and normalized to 3 control genes (see Methods). Data are percent control and presented as mean SEM. N = 4 Floxed, 5 Nes-cKO, and 2 GrnKO. (C) Grn measured by ELISA from primary neuron lysate. N = 11 WT and Floxed combined, 4 Nes-cKO, and 5 GrnKO wells. Data are presented as mean SEM. (D) Grn from conditioned media from primary microglia cultures measured by ELISA and normalized to total protein from lysed cells per well. Data are presented as mean SEM. N = 6 Floxed, 6 Nes-cKO, and 2 GrnKO.
ED
M
AN
US
Figure 2. Progranulin immunoreactivity is highly reduced in neurons but not microglia in Nes-cKO mice. (A-C) Double immunofluorescence shows NeuN-positive neurons (A) co-labeled with Grn (B) in the CA3 region of the hippocampus in Floxed mice. (D-F) In Nes-cKO mice, NeuN immunoreactivity of CA3 pyramidal neurons is similar to Floxed mice (D), but Grn immunoreactivity (E) is highly reduced, though not absent. Distinct, bright Grn-immunoreactive puncta outside of the pyramidal neuron cell layer (arrows) correspond to microglia. (G-I) Double immunofluorescence shows Iba1positive microglia (G) co-labeled with Grn (H) in the cortex in Floxed mice. (J-L) Double immunofluorescence in the cortex of Nes-cKO mice for Iba1 (J) and Grn (K) shows colabeling of microglia, similar to Floxed mice. Scale bar = 50m.
AC
CE
PT
Figure 3. Lipofuscin deposition is not increased in the brains of Nes-cKO mice. (A-C) Representative images of lipofuscin accumulation, measured as autofluorescence, in the thalamus of 12-month-old Floxed (A), Nes-cKO (B), and GrnKO (C) mice. (D) Quantification of relative lipofuscin fluorescence in the thalamus from Floxed, Nes-cKO, and GrnKO mice. (E-G) Representative images of lipofuscin accumulation in the CA3 region of the hippocampus of 12-month-old Floxed (E), NescKO (F), and GrnKO (G) mice. (H) Quantification of relative lipofuscin fluorescence in the CA3 region of the hippocampus from Floxed, Nes-cKO, and GrnKO mice. (I-K) Representative images of lipofuscin accumulation in the cingulate cortex of 12-monthold Floxed (I), Nes-cKO (J), and GrnKO (K) mice. (L) Quantification of relative lipofuscin fluorescence in the cingulate cortex from Floxed, Nes-cKO, and GrnKO mice. (M-O) Representative images of lipofuscin accumulation in the striatum of 12-month-old Floxed (M), Nes-cKO (N), and GrnKO (O) mice. (P) Quantification of relative lipofuscin fluorescence in the striatum from Floxed, Nes-cKO, and GrnKO mice. Data represent the mean SEM measured in arbitrary units (a.u.). N = 4 Floxed, 6 Nes-cKO, and 3 GrnKO mice. For each region, 4 images per region per animal were quantified. *p<0.05; **p<0.001; ***p<0.0001 by Bonferroni post-hoc analysis. Scale bar = 50m.
22
ACCEPTED MANUSCRIPT
US
CR
IP
T
Figure 4. Astrocytosis is not increased in the brains of Nes-cKO mice. (A-C) Representative images of astrocytosis, measured as percent stained area for GFAP immunoreactivity, in the thalamus of 12-month-old Floxed (A), Nes-cKO (B), and GrnKO (C) mice. (D) Quantification of GFAP immunoreactivity in the thalamus from Floxed, Nes-cKO, and GrnKO mice. (E-G) Representative images of GFAP immunoreactivity in the CA3 region of the hippocampus of 12-month-old Floxed (E), Nes-cKO (F), and GrnKO (G) mice. (H) Quantification of GFAP immunoreactivity in the CA3 region of the hippocampus from Floxed, Nes-cKO, and GrnKO mice. (I-K) Representative images of GFAP immunoreactivity in the cingulate cortex of 12-month-old Floxed (I), Nes-cKO (J), and GrnKO (K) mice. (L) Quantification of GFAP immunoreactivity in the cingulate cortex from Floxed, Nes-cKO, and GrnKO mice. (M-O) Representative images of GFAP immunoreactivity in the striatum of 12-month-old Floxed (M), Nes-cKO (N), and GrnKO (O) mice. (P) Quantification of GFAP immunoreactivity in the striatum from Floxed, NescKO, and GrnKO mice. Data represent the mean SEM. N = 4 Floxed, 6 Nes-ckO, and 3 GrnKO mice. For each region, 4 images per region per animal were quantified. *p<0.05; **p<0.001; ***p<0.0001 by Bonferroni post-hoc analysis. Scale bar = 50m.
AC
CE
PT
ED
M
AN
Figure 5. Microgliosis is not increased in the brains of Nes-cKO mice. (A-C) Representative images of microgliosis, measured as % Iba1 immunoreactive area, in the thalamus of 12-month-old Floxed (A), Nes-cKO (B), and GrnKO mice (C). (D) Quantification of Iba1 immunoreactivity in the thalamus from Floxed, Nes-cKO, and GrnKO mice. (E-G) Representative images of Iba1 immunoreactivity in the CA3 region of the hippocampus of 12-month-old Floxed (E), Nes-cKO (F), and GrnKO (G) mice. (H) Quantification of Iba1 immunoreactivity in the CA3 region of the hippocampus from Floxed, Nes-cKO, and GrnKO mice. (I-K) Representative images of Iba1 immunoreactivity in the cingulate cortex of 12-month-old Floxed (I), Nes-cKO (J), and GrnKO (K) mice. (L) Quantification of Iba1 immunoreactivity in the cingulate cortex from Floxed, Nes-cKO, and GrnKO mice. (M-O) Representative images of Iba1 immunoreactivity in the striatum of 12-month-old Floxed(M), Nes-cKO (N), and GrnKO (O) mice. (P) Quantification of Iba1 immunoreactivity in the striatum from Floxed, NescKO, and GrnKO mice. Data represent the mean SEM. N = 4 Floxed, 6 Nes-ckO, and 3 GrnKO mice. For each region, 4 images per region per animal were quantified. *p<0.05; **p<0.001; ***p<0.0001 by Bonferroni post-hoc analysis. Scale bar = 50m. Figure 6. Further reduction of overall Grn levels in Het-Nes-cKO mice is still not sufficient to cause neuropathological changes. (A) Grn measured from whole brain lysate by ELISA in 3-4 month old mice. Data are presented as mean SEM. N = 7 WT and 5 Floxed (combined), 5 Het and 7 Flox-het (combined), 2 Nes-cKO, 7 Het-NescKO, and 4 GrnKO mice. (B) Quantification of lipofuscin accumulation in the thalamus, measured as autofluorescence. (C) Quantification of GFAP immunoreactivity in the thalamus. (D) Quantification of Iba1 immunoreactivity in the thalamus. For (B-D), N = 4 Floxed, 7 Het, 6 Nes-cKO, 7 Het-Nes-cKO, and 4 GrnKO mice. For each region, 4
23
ACCEPTED MANUSCRIPT
AC
CE
PT
ED
M
AN
US
CR
IP
T
images per region per animal were quantified. *p<0.05; **p<0.001; ***p<0.0001 by Bonferroni post-hoc analysis. Scale bar = 50m.
24
AC
Fig. 1
CE
PT
ED
M
AN
US
CR
IP
T
ACCEPTED MANUSCRIPT
25
AC
CE
PT
ED
M
AN
US
CR
IP
T
ACCEPTED MANUSCRIPT
Fig. 2
26
AC
Fig. 3
CE
PT
ED
M
AN
US
CR
IP
T
ACCEPTED MANUSCRIPT
27
AC
Fig. 4
CE
PT
ED
M
AN
US
CR
IP
T
ACCEPTED MANUSCRIPT
28
AC
Fig. 5
CE
PT
ED
M
AN
US
CR
IP
T
ACCEPTED MANUSCRIPT
29
AC
Fig. 6
CE
PT
ED
M
AN
US
CR
IP
T
ACCEPTED MANUSCRIPT
30
ACCEPTED MANUSCRIPT
Conditional loss of progranulin in neurons is not sufficient to cause neuronal ceroid lipofuscinosis-like neuropathology in mice
Highlights:
T
IP
CR
US AN M ED PT CE
Conditional nestin-specific progranulin knockout (Nes-cKO) mice were generated Nes-cKO mice express no neuronal progranulin Neuropathology in Nes-cKO mice does not mimic the NCL phenotype of Grn-null mice Microglia-derived progranulin is sufficient to prevent a NCL phenotype
AC
31
Terri L. Petkau, Jake Blanco, Blair R. Leavitt PII: DOI: Reference:
S0969-9961(17)30139-0 doi: 10.1016/j.nbd.2017.06.012 YNBDI 3982
To appear in:
Neurobiology of Disease
Received date: Revised date: Accepted date:
14 March 2017 1 June 2017 20 June 2017
Please cite this article as: Terri L. Petkau, Jake Blanco, Blair R. Leavitt , Conditional loss of progranulin in neurons is not sufficient to cause neuronal ceroid lipofuscinosis-like neuropathology in mice, Neurobiology of Disease (2017), doi: 10.1016/j.nbd.2017.06.012
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT
Conditional loss of progranulin in neurons is not sufficient to cause
T
neuronal ceroid lipofuscinosis-like neuropathology in mice.
aCentre
CR
IP
Terri L. Petkaua, Jake Blancoa, and Blair R. Leavitta,b,c,§
for Molecular Medicine & Therapeutics, Department of Medical Genetics,
of Neurology, Department of Medicine, University of British Columbia Hospital,
M
bDivision
AN
Avenue, Vancouver, BC, Canada V5Z 4H4
US
University of British Columbia, and Children’s and Women’s Hospital, 980 West 28 th
Research Centre, University of British Columbia, Vancouver, BC V6T 1Z3,
PT
cBrain
ED
S 192 - 2211 Wesbrook Mall, Vancouver, BC, Canada V6T 2B5
CE
Canada §Corresponding
Author: Blair R. Leavitt, Centre for Molecular Medicine &
AC
Therapeutics, Department of Medical Genetics, University of British Columbia, and Children’s and Women’s Hospital, 980 West 28th Avenue, Vancouver, BC, Canada V5Z 4H4. Tel: 604-875-3801, Fax: 604-875-3819, email: [email protected]
ACCEPTED MANUSCRIPT
Abstract Progranulin deficiency due to heterozygous null mutations in the GRN gene is a common cause of familial frontotemporal lobar degeneration (FTLD), while homozygous
T
loss-of-function GRN mutations cause neuronal ceroid lipofuscinosis (NCL). Aged
IP
progranulin-knockout mice display highly exaggerated lipofuscinosis, microgliosis, and
CR
astrogliosis, as well as mild cell loss in specific brain regions. Progranulin is a secreted glycoprotein expressed in both neurons and microglia, but not astrocytes, in the brain.
US
We generated conditional progranulin-knockout mice that lack progranulin in nestin-
AN
expressing cells (Nes-cKO mice), which include most neurons as well as astrocytes. We confirmed near complete knockout of progranulin in neurons in Nes-cKO mice, while
M
microglial progranulin levels remained similar to that of wild-type animals. Overall brain
ED
progranulin levels were reduced by about 50% in Nes-cKO, and no Grn was detected in primary Nes-cKO neurons. Nes-cKO mice aged to 12 months did not display any
PT
increase in lipofuscin deposition, microgliosis, or astrogliosis in the four brain regions
CE
examined, though increases were observed for most of these measures in Grn-null animals. We conclude that neuron-specific loss of progranulin is not sufficient to cause
AC
similar neuropathological changes to those seen in constitutive Grn-null animals. Our results suggest that increased lipofuscinosis and gliosis in Grn-null animals are not caused by intrinsic progranulin deficiency in neurons, and that microglia-derived progranulin may be sufficient to maintain neuronal health and homeostasis in the brain.
Keywords: frontotemporal lobar degeneration, neuronal ceroid lipofuscinosis, progranulin, conditional knockout mice, neuropathology, nestin
2
ACCEPTED MANUSCRIPT
Introduction Progranulin is a pleiotropic, secreted growth factor with widespread expression throughout the periphery and in the brain (Petkau & Leavitt 2014). Within the central
T
nervous system, progranulin is a putative lysosome-associated protein (Tanaka et al
IP
2014, Tanaka et al 2017) expressed in most neuronal populations and in microglia
CR
(Baker et al 2006, Chen-Plotkin et al 2010, Petkau et al 2010), and reduced expression
US
of progranulin is associated with multiple neurodegenerative diseases (Petkau & Leavitt 2014). Most notably, heterozygous loss-of-function mutations in the progranulin gene
AN
(GRN) are a common cause of familial frontotemporal lobar degeneration (FTLD)
M
(Baker et al 2006, Cruts et al 2006), while homozygous loss of GRN expression causes neuronal ceroid lipofuscinosis (NCL) (Smith et al 2012).
ED
Progranulin expression, both in humans and in mice, is present in a punctate,
PT
perinuclear pattern in neurons throughout the brain, with highest expression occurring in the thalamus and CA3 layer of the hippocampus, and without obvious changes in
CE
expression in response to injury (Petkau et al 2010). In microglia, expression of
AC
progranulin is notably stronger than in neurons, and is further increased when microglia are activated after stress or injury (Chen-Plotkin et al 2010, Naphade et al 2010, Petkau et al 2010).
Mice homozygous for a targeted deletion of the mouse progranulin gene (Grn) develop a robust neuropathological phenotype with age (Ahmed et al 2010, Ghoshal et al 2012, Petkau et al 2012, Wils et al 2012, Yin et al 2010b), though heterozygous Grntargeted mice do not develop this phenotype (Ahmed et al 2010, Petkau et al 2012).
3
ACCEPTED MANUSCRIPT
Aged Grn-null mice display exaggerated deposition of the aging pigment lipofuscin and accumulation of a NCL-like storage material, as well as robust astrogliosis and microgliosis in multiple brain regions (Ahmed et al 2010, Ghoshal et al 2012, Petkau et
T
al 2016, Wils et al 2012, Yin et al 2010b). In addition, mild neuronal cell loss (Ghoshal et
IP
al 2012) and alterations in neuronal morphology have also been reported (Petkau et al
CR
2016, Petkau et al 2012). Grn-null immune cells, such as macrophages and microglia, show a hyper-inflammatory response to various stimuli (Martens et al 2012, Suh et al
US
2012, Yin et al 2010a), as well as altered lysosomal biogenesis (Tanaka et al 2013a,
AN
Tanaka et al 2013b, Tanaka et al 2017). Grn deficiency in neurons has been associated with reduced neuronal survival (De Muynck et al 2013, Guo et al 2010, Kleinberger et al
M
2010, Ryan et al 2009, Van Damme et al 2008), reduced neurite outgrowth (Van
ED
Damme et al 2008), and altered lysosome function (Tanaka et al 2017). In this study, we sought to determine the relative contribution of progranulin
PT
derived from neurons to the robust neuropathological phenotype present in constitutive
CE
Grn-null mice. We created neuron-specific conditional Grn knockout mice (Nes-cKO) and evaluated whether they developed lipofuscinosis, microgliosis, and astrogliosis in
AC
the brain with age. Surprisingly, we find that near complete loss of Grn expression in most neurons is not sufficient to reproduce any of the neuropathological phenotypes observed in constitutive Grn-null mice, and conclude that microglia-derived Grn is sufficient to maintain brain homeostasis in terms of the outcomes measured here. Methods Mice
4
ACCEPTED MANUSCRIPT
The generation of ‘floxed’ progranulin-targeted (Floxed) mice was previously described (Petkau et al 2013). Conditional nestin-knockout (Nes-cKO) mice were generated by crossing homozygous Grnflox/flox mice on the C57Bl/6 background to
T
transgenic mice expressing Cre recombinase under the control of the rat nestin
IP
promoter and enhancer (generous gift from Dr. E. Simpson; originally from The Jackson
CR
Laboratory, strain name: B6.Cg-Tg(Nes-cre)1Kln/J). Final experimental cohorts were generated by crossing Grnflox/flox, Cre+ animals to Grnflox/flox, Cre- littermates. Genotyping
US
was performed on tail tip DNA at wean and confirmed on a second DNA sample at sacrifice using the following primer sequences: Grn common forward primer: 5’-
AN
CGGAACACAGTGTCCAGATG-3’; Grn intron 2 reverse primer: 5’-
M
ATCAACCAAAGGGTCTGTGC-3’; Grn exon 5 reverse primer: 5’-
ED
GTGGCAGAGTCAGGACATTCAAACT-3’; Cre forward primer: 5’GCGGTCTGGCAGTAAAAACTATC-3’; Cre reverse primer: 5’-
PT
GTGAAACAGCATTGCTGTCACTT-3’. To generate Het-Nes-cKO mice, we crossed
CE
Grnflox/flox, Cre+ mice to constitutive Grn-/- mice and genotyped the offspring for the presence of the Cre transgene.
AC
Mice were housed on ventilated racks in specific pathogen-free barrier facility with a 12hr light/dark cycle. Mice were group-housed with their littermates to a maximum of four mice per cage. All animal procedures were done with the approval of the Canadian Council for Animal Care and the University of British Columbia’s Animal Care Committee.
5
ACCEPTED MANUSCRIPT
Protein Extraction and quantification of Grn by ELISA Whole brain lysate was prepared by homogenizing previously snap-frozen brains in a rotor-stator homogenizer for 30 sec. in 1mL of complete lysis buffer (50mM Tris-
T
HCl, 1% Triton-X, 150mM NaCl, Halt phosphatase inhibitor cocktail (Thermo Fisher
IP
Scientific), Halt protease inhibitor cocktail (Thermo Fisher Scientific)). Primary cortical
CR
neurons were generated from embryonic day 17.5 pups as previously described. Neurons cultured for 7 days in 6-well plates were washed twice with ice-cold PBS
US
before being scraped into 60L of complete RIPA buffer (50mM Tris-HCl, 1% Triton-X,
AN
150mM NaCl, 1% deoxycholic acid, 1mM EDTA, Halt phosphatase inhibitor cocktail (Thermo Fisher Scientific), Halt protease inhibitor cocktail (Thermo Fisher Scientific))
M
with DNase added. The lysate was collected into pre-chilled microcentrifuge tubes and
ED
stored at -80oC until use. Total protein was assayed using Bradford reagent (BioRad). Primary microglia cultures were generated as previously described (Connolly et al
PT
2016). Conditioned media was collected after 24hrs and stored at -20oC until used.
CE
The quantity of Grn in whole brain lysate, cell lysate, or conditioned media was determined by an enzyme-linked immunosorbent assay (ELISA) using a commercially
AC
available kit (Mouse progranulin ELISA; Adipogen, Korea). Microglia supernatant samples were diluted 1:5; for neuronal cell lysates, 20 g of total protein was loaded per well, and for whole brain lysate, 100g of protein was used. All samples were run in duplicate. The ELISA was conducted according to the manufacturer’s instructions. Data represent the average of 2-8 samples per condition, and all conditions that were compared directly were run on the same plate.
6
ACCEPTED MANUSCRIPT
RNA isolation and qPCR For analysis of Grn mRNA, cortical tissue from 4 Floxed, 5 Nes-cKO, and 2
T
GrnKO mice at 4-6 months of age was collected and immediately frozen at −80°C.
IP
Samples were homogenized with a bead homogenizer in lysis buffer followed by total
CR
RNA extraction (PureLink RNA mini kit; Invitrogen) performed according to the manufacturer’s instructions. Reverse transcription of all samples was carried out using
US
the Superscript VILO kit (Invitrogen) according to the manufacturer's instructions, using
AN
1g of total RNA as input for cDNA synthesis. Following this, cDNA was diluted 1:10 in ddH2O for a total input of 5ng into the quantitative PCR reaction, done using FastSybr
M
(Applied Biosystems) and conducted on a Step-One ABI System (Applied Biosystems).
ED
Quantification of mRNA levels was accomplished using the standard curve method, with amplification of target mRNA and control genes in separate wells. Each sample was run
PT
in duplicate. The relative amount of mRNA in each well was calculated as the ratio
CE
between Grn mRNA (forward: 5’-CTGTAGTGCAGATGGGAAATCCTGCT-3’; reverse: 5’-GTGGCAGAGTCAGGACATTCAAACT-3’) and a normalization factor created using
AC
two control genes, Usf1 (forward: 5’-CCTGTGGCGTGGCAGTCT-3’; reverse: 5’TGCACGCCCACACTGTTT-3’) and Paklip1 (forward: 5’CCCCAAGTGGAGGGAAGTACA-3’; reverse: 5’-TGCCCAGCCGATAGACATC-3’) based on GeNorm (Vandesompele et al 2002). Values are presented as % Floxed control.
Immunohistochemistry
7
ACCEPTED MANUSCRIPT
Serial 25m floating sections were placed in net-well inserts and washed for 10min in phosphate-buffered saline (PBS). Endogenous peroxidase activity was quenched with 1% H2O2 for 45min. After a 15min wash in PBS with 0.1% Triton X (PBS-
T
T), sections were blocked in 5% normal serum and 5% bovine serum albumin diluted in
IP
PBS-T, followed by overnight incubation shaking at room temperature in primary
CR
antibody diluted in 5% normal serum and PBS-T. After two 15min washes in PBS-T, secondary antibody diluted in 1% normal serum and PBS-T was applied for 2hrs
US
shaking at room temperature. Sections were washed for 30min in PBS before an
AN
amplification step was performed using an avidin–biotin–horseradish peroxidase complex kit (Vector Laboratories). Colorimetric detection was achieved with the
M
peroxidase substrate kit Vector DAB (Vector Laboratories) according to the
ED
manufacturer’s instructions. Sections were mounted by hand on onto glass slides (Fisherbrand Superfrost Plus) and dried overnight before being dehydrated through a
PT
series of alcohols and xylene, and cover-slipped with DEPEX (Electron Microscopy
CE
Sciences). Antibodies used were as follows: the neuronal marker NeuN (Chemicon, Millipore, 1:2000, mouse monoclonal), the microglia marker Iba1 (Wako; 1:2000, rabbit
AC
polyclonal), the astrocyte marker GFAP (Sigma; 1:2000, mouse monoclonal), and appropriate biotinylated secondary antibodies (Vector, 1:2000). Immunofluorescent staining was performed by incubating mounted, 25m serial sections in PBS-T for 30 minutes, followed by 1hr in blocking solution (5% normal donkey serum in PBS-T) and subsequent overnight incubation at 4oC with primary antibodies in 2% normal serum. Primary antibodies included NeuN (mouse monoclonal,
8
ACCEPTED MANUSCRIPT
Millipore, 1:1000), Iba1 (rabbit polyclonal, Wako, 1:1000), and Grn (sheep polyclonal, R&D Systems, 1:250). Section were washed twice for 5min each prior to incubation with appropriate secondary antibodies for 1hr at room temperature in 2% normal donkey
T
serum (Donkey anti-mouse Alexa 488, donkey anti-rabbit Alexa 488, donkey anti-sheep
IP
Alexa 594, Molecular Probes, 1:1000). After washing again twice in PBS for 5 min,
CR
sections were incubated with DAPI (4',6-diamidino-2-phenylindole, dihydrochloride) in PBS at 1:10 000 for 5 min, washed twice in PBS again for 5 min each, and then
US
dehydrated and coverslipped with fluorescent DEPEX mounting medium.
AN
Sections to be assessed for autofluorescence were mounted onto glass slides, washed in PBS-T for 30 min, and then stained with DAPI in PBS at 1:10 000 for 5 min.
ED
Image acquisition and analysis
M
Slides were washed in twice in PBS for 5 min prior to coverslipping.
PT
Images were acquired as previously described (Petkau et al 2016). Integrated optical density measurements of signal intensity were acquired as previously described
CE
(Petkau et al 2010) to quantify autofluorescence, a surrogate for lipofuscin and/or NCL-
AC
like material deposition, in 4 images per mouse taken from 2-7 mice per genotype. For quantification of colorimetric stains (Iba1 and GFAP), a threshold was set that pseudocolored stained areas within a defined region of interest in each image. The average percent thresholded area for 4 images per mouse taken from 2-7 mice per genotype is reported. Statistical analysis
9
ACCEPTED MANUSCRIPT
All statistical comparisons were performed as a one-way analysis of variance (ANOVA) with Bonferroni post-hoc analysis to compare individual means to control and correct for multiple comparisons (Prism 6, Graphpad Software Inc.). A p-value less than
IP
T
0.05 was considered significant.
CR
Results Grn is ‘knocked down’ in Nes-cKO mice
US
First, we verified that Floxed mice expressed Grn in the brain at levels comparable to WT mice; ie. that the inserted loxP sites in the targeted allele do not
AN
interfere with normal gene expression. Grn levels quantified by ELISA on whole brain
M
homogenate showed similar levels between Floxed mice and WT mice (Fig.1A; p > 0.05
ED
by Bonferroni post-hoc analysis). Subsequent analyses used Floxed mice as controls, or data from Floxed and WT mice was combined. As expected, no Grn expression was
PT
detected in GrnKO mice (Fig.1A). Whole brain Grn levels were reduced by approximately 50% in Nes-cKO mice compared to Floxed mice (Fig.1A). Analysis of Grn
CE
mRNA expression in the cortex of Floxed and Nes-cKO mice showed a reduction of Grn
AC
in Nes-cKO mice to about 20% of Floxed mice (Fig.1B). Once again, no Grn expression was detected from GrnKO mice. Next, to examine Grn expression in specific cell types, we generated primary cultures of cortical neurons and whole brain microglia and measured Grn in the cell lysate (Fig.1C) or conditioned media (Fig.1D) respectively. As expected, Nes-cKO cortical neuron cultures expressed no Grn, while cultures derived
10
ACCEPTED MANUSCRIPT
from Floxed mice expressed Grn at WT levels (Fig.1D). Grn secreted into the media from microglia, was similar between Nes-cKO and Floxed cultures (Fig.1D). We next evaluated Grn expression by immunohistochemistry in brain sections
T
from Nes-cKO and Floxed mice. Similar to previous reports evaluating Grn staining in
IP
WT mice (Petkau et al 2010), robust Grn staining was present in NeuN-labeled neurons
CR
throughout the brain, including in the CA3 layer of the hippocampus of Floxed mice (Fig.2A-C). Brightly staining puncta that do not co-localize with NeuN correspond to
US
microglia (Fig1C; arrow). Brightly staining, Grn-positive microglia were easily identified
AN
in brain sections from Nes-cKO mice (Fig.2F, arrows), but neuronal Grn staining was drastically reduced, albeit not completely absent (Fig.2E). Given that no Grn was
M
detected in primary neuronal cultures from Nes-cKO mice, and that Grn is a secreted
ED
protein that can be taken up by neurons in culture (Hu et al 2010), it is likely that very low level of Grn staining present in neurons in Nes-cKO mice is internalized, microglia-
PT
derived Grn. Microglia in the cortex, labeled with Iba1 (Fig.2G), were co-labeled with
CE
Grn (Fig.2H), as expected, and looked similar in Nes-cKO mice (Fig.2J-L). Overall, our data show that Nes-cKO mice express little to no Grn in neurons, while microglial Grn
AC
expression is unaffected.
Neuropathology in Nes-cKO mice does not replicate that of GrnKO mice We previously completed a detailed analysis of behavior and neuropathology in aged GrnKO mice on a mixed genetic background (Petkau et al 2012) and on two different inbred background strains (Petkau et al 2016). We chose three robust
11
ACCEPTED MANUSCRIPT
neuropathological endpoints that clearly distinguish GrnKO animals on a C57Bl/6 background from their WT littermates and assessed them in Nes-cKO and matching Floxed and GrnKO littermates at 12 months of age. We analyzed lipofuscinosis,
T
astrogliosis, and microgliosis in four brain regions: the thalamus, CA3 layer of the
IP
hippocampus, striatum, and cortex.
CR
Lipofuscin is an autofluorescent pigment that accumulates with age, most predominantly in post-mitotic cells such as neurons (Sulzer et al 2008). Lipofuscin
US
accumulation, measured by quantifying autofluorescence, is detectable in the brains of
AN
Floxed mice (Fig.3), including in the thalamus (Fig.3A), CA3 layer of the hippocampus (Fig.3E), the cortex (Fig.3I), and the striatum (Fig.3M). Lipofuscin accumulation
M
occurred at levels similar to Floxed mice in all four brain regions in Nes-cKO mice
ED
(Fig.3B, F, J, N), but at significantly higher levels in GrnKO mice in the thalamus (Fig.3C, quantified in D) and CA3 region (Fig.3G, quantified in H). Lipofuscin in the
PT
cortex and striatum was not significantly higher in GrnKO animals compared to Floxed
CE
or Nes-cKO mice, though overall levels of lipofuscin in these regions was considerably lower than in the thalamus and CA3 layer of the hippocampus. Thus, despite clear
AC
evidence of increased lipofuscin deposition in two brain regions in GrnKO mice, no significant change in lipofuscin was observed when Grn was specifically knocked down in neurons.
Astrogliosis was assessed by quantifying the stained area of GFAP immunoreactivity in the same four brain regions. GFAP immunoreactivity in the thalamus was limited to occasional cells only in a given field of view in both Floxed
12
ACCEPTED MANUSCRIPT
(Fig.4A) and Nes-cKO (Fig.4B) mice, but large patches of reactive astrocytes occurred in GrnKO mice (Fig.4C) and overall percent stained area was significantly higher in GrnKO mice (Fig.4D). In the hippocampus, GFAP immunoreactivity was overall much
T
higher than in other brain regions. The stained area was significantly higher in GrnKO
IP
mice than in Floxed or Nes-cKO mice (Fig.4, E-H). GFAP immunoreactivity was also
CR
significantly higher in Nes-cKO mice than in Floxed mice (Fig.4H). The cortex and striatum showed a similar pattern to the thalamus, where Floxed (Fig.4I, M) and Nes-
US
cKO (Fig.4J, N) showed only occasional or sporadic GFAP-immunoreacitve cells, while
AN
large patches of reactive astrocytes could be found in GrnKO mice (Fig.4K, O). Quantification of GFAP immunostaining showed significantly higher percent area
M
stained in GrnKO mice compared to Floxed and Nes-cKO mice in both the cortex
ED
(Fig.4L) and striatum (Fig.4P). Similar to lipofuscinosis, while GrnKO mice show a clear neuropathological phenotype of increased astrocytosis in multiple brain regions, no
PT
changes were apparent in Nes-cKO mice.
CE
We evaluated microgliosis using Iba1 immunoreactivity in the thalamus, hippocampus, cortex and striatum. In the thalamus, Iba1-stained microglia were
AC
noticeably more numerous and had larger cell bodies in GrnKO mice (Fig.5C) compared to either Floxed (Fig.5A) or Nes-cKO (Fig.5B) mice, and the percent Iba1-stained area was significantly higher in GrnKO animals (Fig.5D). Qualitatively, a similar change in cell morphology (larger cell bodies and a less ramified appearance) could be appreciated in the GrnKO mice in the CA3 region of the hippocampus (Fig.5G), cortex (Fig.5K), and striatum (Fig.5O) compared to both Floxed (Fig.5A, E, M) and Nes-cKO
13
ACCEPTED MANUSCRIPT
(Fig.5B, F, N) mice, though quantification of the percent Iba1-stained area did not show a significant difference between the genotypes (Fig.5H, L, P). Although microgliosis was not as pronounced in GrnKO mice in this study as in previous studies (Petkau et al
T
2016, Petkau et al 2012), increased Iba1 immunoreactivity was observed in the
IP
thalamus of GrnKO mice but was distinctly absent from Nes-cKO mice.
CR
Further reduction of Grn levels in Het-Nes-cKO mice has no effect on neuropathology Finally, we considered the possibility that the Grn-dependent neuropathological
US
abnormalities of increased lipofuscinosis, astrogliosis, and microgliosis might only occur
AN
when overall Grn levels are reduced below the level seen in Nes-cKO mice, which is approximately 50% of WT mice (Fig.1A). To test this hypothesis, we generated mice
M
carrying one constitutively recombined, Grn-null allele and one Grn ‘floxed’ allele
ED
(Grnflox/-) along with the nestin-driven Cre transgene (Het-Nes-cKO). These mice are constitutively heterozygous Grn-null mice with further neural-specific Grn knockdown in
PT
Cre-expressing mice. When whole brain Grn levels were measured in brain lysate by
CE
Grn ELISA, we found the overall level of Grn in Het-Nes-cKO mice to be approximately 25% of WT or Floxed levels, or 50% less again than Nes-cKO animals (Fig.6A). We
AC
quantified lipofuscinosis (Fig.6B), astrogliosis (Fig.6C), and microgliosis (Fig.6D) in the thalamus of Grn-Het and Het-Nes-cKO animals and compared it to WT, Floxed, NescKO, and GrnKO animals. Even with just 25% of normal Grn levels in the brain, and likely nearly no Grn produced by neurons (based on near-complete loss of Grn expression in primary neurons even in Nes-cKO mice, see Fig.1C), we observed no
14
ACCEPTED MANUSCRIPT
significant difference in any of these measures in Het-Nes-cKO animals compared to Floxed controls. Overall, the data indicate that neural-specific knockdown of Grn in mice does not
T
cause the same neuropathological phenotypes that are observed in constitutive Grn-null
CR
IP
animals.
Discussion
US
This is the first study, to our knowledge, to assess the phenotype of conditional
AN
Grn-null animals where Grn has been selectivity knocked out of neural cells in mice. Despite showing near-complete knockout in isolated neurons and substantially reduced
M
overall brain Grn levels, Nes-cKO mice do not develop any of the neuropathological
ED
phenotypes that are present in aged Grn-null mice. Given that Grn expression in Nes-cKO neurons is essentially absent, the
PT
remaining Grn present in the brain is likely derived from microglia. Although neurons
CE
greatly outnumber microglia in terms of absolute numbers in the brain, Nes-cKO mice Grn expression levels are reduced only by about 50% compared to WT mice,
AC
suggesting that microglia contribute the other 50%. Progranulin is a secreted protein (Suh et al 2012) and at least one putative Grn receptor on neurons has been identified (Hu et al 2010), so it is feasible that microglial-derived Grn could compensate for loss of intrinsic Grn expression in neurons. If this is the case, it suggests an important role for microglia-derived Grn, and future experiments evaluating neuropathology in microgliaspecific Grn knockout mice could greatly improve our understanding of how different
15
ACCEPTED MANUSCRIPT
sources of brain Grn contribute to brain homeostasis. Furthermore, given that there are considerable efforts being invested in evaluating Grn-modulating therapeutics and their clinical potential, it will be of substantial clinical importance to ensure that drug screens
T
for such compounds consider Grn regulation in microglia as well as in neurons.
IP
Lipofuscin accumulation, an early and robust phenotype in constitutive Grn-null
CR
mice (Ahmed et al 2010, Ghoshal et al 2012, Petkau et al 2016, Petkau et al 2012, Wils et al 2012, Yin et al 2010a, Yin et al 2010b), occurs largely in post-mitotic neurons.
US
Lipofuscinosis is often caused by dysfunctional lysosomal components (Jalanko &
AN
Braulke 2009), and recent studies have implicated a role for Grn in lysosomal function (Tanaka et al 2014, Tanaka et al 2017). That Nes-cKO mice do not display any signs of
M
increased lipofuscin deposition implies that cell-intrinsic Grn production is not required
ED
to maintain lysosomal function. Furthermore, it suggests that Grn may contribute to lysosomal function via external signaling pathways. It remains unclear whether intrinsic
PT
Grn production in neurons has additional and/or alternative roles in neuronal health, or if
CE
the cellular source of Grn is dispensable and all Grn-dependent functions in neurons are carried out via extracellular Grn production and signaling. Data from FTLD patients with
AC
GRN mutations shows, quite strikingly, that human progranulin expression actually increases in affected brain regions, and that this increase can be attributed to increased expression from infiltrating microglia (Chen-Plotkin et al 2010). This data alone provides strong support for the first alternative, that Grn plays additional and/or alternative roles in neuronal cell biology that have yet to be fully elucidated.
16
ACCEPTED MANUSCRIPT
An alternative explanation for the discrepancy in neuropathology between mice constitutively lacking Grn and mice with neural-specific loss of Grn is that the phenotype in Grn-null animals is developmental. This hypothesis predicts that any Grn expression
T
in the brain, or perhaps even transient expression of Grn in the brain during
IP
development, is sufficient to prevent the development of lipofuscinosis, astrogliosis, and
CR
microgliosis later in life. It remains to be tested, but this hypothesis is perhaps supported by data from heterozygous Grn-null mice (Ahmed et al 2010, Filiano et al 2013) and
US
Het-Nes-cKO mice (Fig.6), which shows quite clearly that the Grn-dependent
AN
neuropathological phenotypes measured here are not dose dependent. In conclusion, neural-specific knockdown of Grn in mice does not recapitulate the
M
neuropathological phenotypes seen in constitutive Grn-null animals. Despite being a
ED
relatively minor fraction of cells in the brain, microglia can contribute a substantial portion of overall brain Grn levels, which is seemingly sufficient to maintain brain
PT
homeostasis, or at least to prevent exaggerated lipofuscinosis, microgliosis, and
CE
astrogliosis in aged mice. This work highlights the importance of evaluating Grnmodulating therapies in microglia as well as neurons, and suggests that Grn may play
AC
alternative or additional roles in neuronal function.
Acknowledgements This work was supported by the Alzheimer Society of Canada (doctoral trainee award to TP), and by the Canadian Institute for Health Research (BRL operating grant #97857).
17
AC
CE
PT
ED
M
AN
US
CR
IP
T
ACCEPTED MANUSCRIPT
18
ACCEPTED MANUSCRIPT
REFERENCES
AC
CE
PT
ED
M
AN
US
CR
IP
T
Ahmed Z, Sheng H, Xu YF, Lin WL, Innes AE, et al. 2010. Accelerated lipofuscinosis and ubiquitination in granulin knockout mice suggest a role for progranulin in successful aging. Am J Pathol 177: 311-24 Baker M, Mackenzie IR, Pickering-Brown SM, Gass J, Rademakers R, et al. 2006. Mutations in progranulin cause tau-negative frontotemporal dementia linked to chromosome 17. Nature 442: 916-9 Chen-Plotkin AS, Xiao J, Geser F, Martinez-Lage M, Grossman M, et al. 2010. Brain progranulin expression in GRN-associated frontotemporal lobar degeneration. Acta Neuropathol 119: 111-22 Connolly C, Magnusson-Lind A, Lu G, Wagner PK, Southwell AL, et al. 2016. Enhanced immune response to MMP3 stimulation in microglia expressing mutant huntingtin. Neuroscience 325: 74-88 Cruts M, Kumar-Singh S, Van Broeckhoven C. 2006. Progranulin mutations in ubiquitinpositive frontotemporal dementia linked to chromosome 17q21. Curr Alzheimer Res 3: 485-91 De Muynck L, Herdewyn S, Beel S, Scheveneels W, Van Den Bosch L, et al. 2013. The neurotrophic properties of progranulin depend on the granulin E domain but do not require sortilin binding. Neurobiol Aging Filiano AJ, Martens LH, Young AH, Warmus BA, Zhou P, et al. 2013. Dissociation of frontotemporal dementia-related deficits and neuroinflammation in progranulin haploinsufficient mice. J Neurosci 33: 5352-61 Ghoshal N, Dearborn JT, Wozniak DF, Cairns NJ. 2012. Core features of frontotemporal dementia recapitulated in progranulin knockout mice. Neurobiol Dis 45: 395-408 Guo A, Tapia L, Bamji SX, Cynader MS, Jia W. 2010. Progranulin deficiency leads to enhanced cell vulnerability and TDP-43 translocation in primary neuronal cultures. Brain Res 1366: 1-8 Hu F, Padukkavidana T, Vaegter CB, Brady OA, Zheng Y, et al. 2010. Sortilin-mediated endocytosis determines levels of the frontotemporal dementia protein, progranulin. Neuron 68: 654-67 Jalanko A, Braulke T. 2009. Neuronal ceroid lipofuscinoses. Biochim Biophys Acta 1793: 697-709 Kleinberger G, Wils H, Ponsaerts P, Joris G, Timmermans JP, et al. 2010. Increased caspase activation and decreased TDP-43 solubility in progranulin knockout cortical cultures. J Neurochem 115: 735-47 Martens LH, Zhang J, Barmada SJ, Zhou P, Kamiya S, et al. 2012. Progranulin deficiency promotes neuroinflammation and neuron loss following toxin-induced injury. J Clin Invest 122: 3955-9 Naphade SB, Kigerl KA, Jakeman LB, Kostyk SK, Popovich PG, Kuret J. 2010. Progranulin expression is upregulated after spinal contusion in mice. Acta Neuropathol 119: 123-33
19
ACCEPTED MANUSCRIPT
AC
CE
PT
ED
M
AN
US
CR
IP
T
Petkau TL, Hill A, Leavitt BR. 2016. Core neuropathological abnormalities in progranulin-deficient mice are penetrant on multiple genetic backgrounds. Neuroscience 315: 175-95 Petkau TL, Leavitt BR. 2014. Progranulin in neurodegenerative disease. Trends Neurosci 37: 388-98 Petkau TL, Neal SJ, Milnerwood A, Mew A, Hill AM, et al. 2012. Synaptic dysfunction in progranulin-deficient mice. Neurobiol Dis 45: 711-22 Petkau TL, Neal SJ, Orban PC, MacDonald JL, Hill AM, et al. 2010. Progranulin expression in the developing and adult murine brain. J Comp Neurol 518: 393147 Petkau TL, Zhu S, Lu G, Fernando S, Cynader M, Leavitt BR. 2013. Sensitivity to neurotoxic stress is not increased in progranulin-deficient mice. Neurobiol Aging Ryan CL, Baranowski DC, Chitramuthu BP, Malik S, Li Z, et al. 2009. Progranulin is expressed within motor neurons and promotes neuronal cell survival. BMC Neurosci 10: 130 Smith KR, Damiano J, Franceschetti S, Carpenter S, Canafoglia L, et al. 2012. Strikingly different clinicopathological phenotypes determined by progranulinmutation dosage. Am J Hum Genet 90: 1102-7 Suh HS, Choi N, Tarassishin L, Lee SC. 2012. Regulation of progranulin expression in human microglia and proteolysis of progranulin by matrix metalloproteinase-12 (MMP-12). PLoS One 7: e35115 Sulzer D, Mosharov E, Talloczy Z, Zucca FA, Simon JD, Zecca L. 2008. Neuronal pigmented autophagic vacuoles: lipofuscin, neuromelanin, and ceroid as macroautophagic responses during aging and disease. J Neurochem 106: 24-36 Tanaka Y, Chambers JK, Matsuwaki T, Yamanouchi K, Nishihara M. 2014. Possible involvement of lysosomal dysfunction in pathological changes of the brain in aged progranulin-deficient mice. Acta Neuropathol Commun 2: 78 Tanaka Y, Matsuwaki T, Yamanouchi K, Nishihara M. 2013a. Exacerbated inflammatory responses related to activated microglia after traumatic brain injury in progranulin-deficient mice. Neuroscience 231: 49-60 Tanaka Y, Matsuwaki T, Yamanouchi K, Nishihara M. 2013b. Increased Lysosomal Biogenesis in Activated Microglia and Exacerbated Neuronal Damage after Traumatic Brain Injury in Progranulin-Deficient Mice. Neuroscience Tanaka Y, Suzuki G, Matsuwaki T, Hosokawa M, Serrano G, et al. 2017. Progranulin regulates lysosomal function and biogenesis through acidification of lysosomes. Hum Mol Genet Van Damme P, Van Hoecke A, Lambrechts D, Vanacker P, Bogaert E, et al. 2008. Progranulin functions as a neurotrophic factor to regulate neurite outgrowth and enhance neuronal survival. J Cell Biol 181: 37-41 Vandesompele J, De Preter K, Pattyn F, Poppe B, Van Roy N, et al. 2002. Accurate normalization of real-time quantitative RT-PCR data by geometric averaging of multiple internal control genes. Genome Biol 3: RESEARCH0034
20
ACCEPTED MANUSCRIPT
AC
CE
PT
ED
M
AN
US
CR
IP
T
Wils H, Kleinberger G, Pereson S, Janssens J, Capell A, et al. 2012. Cellular ageing, increased mortality and FTLD-TDP-associated neuropathology in progranulin knockout mice. J Pathol 228: 67-76 Yin F, Banerjee R, Thomas B, Zhou P, Qian L, et al. 2010a. Exaggerated inflammation, impaired host defense, and neuropathology in progranulin-deficient mice. J Exp Med 207: 117-28 Yin F, Dumont M, Banerjee R, Ma Y, Li H, et al. 2010b. Behavioral deficits and progressive neuropathology in progranulin-deficient mice: a mouse model of frontotemporal dementia. FASEB J 24: 4639-47
21
ACCEPTED MANUSCRIPT
FIGURE LEGENDS
CR
IP
T
Figure 1. Progranulin is knocked down in Nes-cKO mice. (A) Grn measured from whole brain lysate by ELISA in 3-4 month old mice. Data are presented as mean SEM. N = 7 WT, 5 Floxed, 2 Nes-cKO, and 4 GrnKO. (B) Grn transcript measured by qPCR from RNA extracted from cortex and normalized to 3 control genes (see Methods). Data are percent control and presented as mean SEM. N = 4 Floxed, 5 Nes-cKO, and 2 GrnKO. (C) Grn measured by ELISA from primary neuron lysate. N = 11 WT and Floxed combined, 4 Nes-cKO, and 5 GrnKO wells. Data are presented as mean SEM. (D) Grn from conditioned media from primary microglia cultures measured by ELISA and normalized to total protein from lysed cells per well. Data are presented as mean SEM. N = 6 Floxed, 6 Nes-cKO, and 2 GrnKO.
ED
M
AN
US
Figure 2. Progranulin immunoreactivity is highly reduced in neurons but not microglia in Nes-cKO mice. (A-C) Double immunofluorescence shows NeuN-positive neurons (A) co-labeled with Grn (B) in the CA3 region of the hippocampus in Floxed mice. (D-F) In Nes-cKO mice, NeuN immunoreactivity of CA3 pyramidal neurons is similar to Floxed mice (D), but Grn immunoreactivity (E) is highly reduced, though not absent. Distinct, bright Grn-immunoreactive puncta outside of the pyramidal neuron cell layer (arrows) correspond to microglia. (G-I) Double immunofluorescence shows Iba1positive microglia (G) co-labeled with Grn (H) in the cortex in Floxed mice. (J-L) Double immunofluorescence in the cortex of Nes-cKO mice for Iba1 (J) and Grn (K) shows colabeling of microglia, similar to Floxed mice. Scale bar = 50m.
AC
CE
PT
Figure 3. Lipofuscin deposition is not increased in the brains of Nes-cKO mice. (A-C) Representative images of lipofuscin accumulation, measured as autofluorescence, in the thalamus of 12-month-old Floxed (A), Nes-cKO (B), and GrnKO (C) mice. (D) Quantification of relative lipofuscin fluorescence in the thalamus from Floxed, Nes-cKO, and GrnKO mice. (E-G) Representative images of lipofuscin accumulation in the CA3 region of the hippocampus of 12-month-old Floxed (E), NescKO (F), and GrnKO (G) mice. (H) Quantification of relative lipofuscin fluorescence in the CA3 region of the hippocampus from Floxed, Nes-cKO, and GrnKO mice. (I-K) Representative images of lipofuscin accumulation in the cingulate cortex of 12-monthold Floxed (I), Nes-cKO (J), and GrnKO (K) mice. (L) Quantification of relative lipofuscin fluorescence in the cingulate cortex from Floxed, Nes-cKO, and GrnKO mice. (M-O) Representative images of lipofuscin accumulation in the striatum of 12-month-old Floxed (M), Nes-cKO (N), and GrnKO (O) mice. (P) Quantification of relative lipofuscin fluorescence in the striatum from Floxed, Nes-cKO, and GrnKO mice. Data represent the mean SEM measured in arbitrary units (a.u.). N = 4 Floxed, 6 Nes-cKO, and 3 GrnKO mice. For each region, 4 images per region per animal were quantified. *p<0.05; **p<0.001; ***p<0.0001 by Bonferroni post-hoc analysis. Scale bar = 50m.
22
ACCEPTED MANUSCRIPT
US
CR
IP
T
Figure 4. Astrocytosis is not increased in the brains of Nes-cKO mice. (A-C) Representative images of astrocytosis, measured as percent stained area for GFAP immunoreactivity, in the thalamus of 12-month-old Floxed (A), Nes-cKO (B), and GrnKO (C) mice. (D) Quantification of GFAP immunoreactivity in the thalamus from Floxed, Nes-cKO, and GrnKO mice. (E-G) Representative images of GFAP immunoreactivity in the CA3 region of the hippocampus of 12-month-old Floxed (E), Nes-cKO (F), and GrnKO (G) mice. (H) Quantification of GFAP immunoreactivity in the CA3 region of the hippocampus from Floxed, Nes-cKO, and GrnKO mice. (I-K) Representative images of GFAP immunoreactivity in the cingulate cortex of 12-month-old Floxed (I), Nes-cKO (J), and GrnKO (K) mice. (L) Quantification of GFAP immunoreactivity in the cingulate cortex from Floxed, Nes-cKO, and GrnKO mice. (M-O) Representative images of GFAP immunoreactivity in the striatum of 12-month-old Floxed (M), Nes-cKO (N), and GrnKO (O) mice. (P) Quantification of GFAP immunoreactivity in the striatum from Floxed, NescKO, and GrnKO mice. Data represent the mean SEM. N = 4 Floxed, 6 Nes-ckO, and 3 GrnKO mice. For each region, 4 images per region per animal were quantified. *p<0.05; **p<0.001; ***p<0.0001 by Bonferroni post-hoc analysis. Scale bar = 50m.
AC
CE
PT
ED
M
AN
Figure 5. Microgliosis is not increased in the brains of Nes-cKO mice. (A-C) Representative images of microgliosis, measured as % Iba1 immunoreactive area, in the thalamus of 12-month-old Floxed (A), Nes-cKO (B), and GrnKO mice (C). (D) Quantification of Iba1 immunoreactivity in the thalamus from Floxed, Nes-cKO, and GrnKO mice. (E-G) Representative images of Iba1 immunoreactivity in the CA3 region of the hippocampus of 12-month-old Floxed (E), Nes-cKO (F), and GrnKO (G) mice. (H) Quantification of Iba1 immunoreactivity in the CA3 region of the hippocampus from Floxed, Nes-cKO, and GrnKO mice. (I-K) Representative images of Iba1 immunoreactivity in the cingulate cortex of 12-month-old Floxed (I), Nes-cKO (J), and GrnKO (K) mice. (L) Quantification of Iba1 immunoreactivity in the cingulate cortex from Floxed, Nes-cKO, and GrnKO mice. (M-O) Representative images of Iba1 immunoreactivity in the striatum of 12-month-old Floxed(M), Nes-cKO (N), and GrnKO (O) mice. (P) Quantification of Iba1 immunoreactivity in the striatum from Floxed, NescKO, and GrnKO mice. Data represent the mean SEM. N = 4 Floxed, 6 Nes-ckO, and 3 GrnKO mice. For each region, 4 images per region per animal were quantified. *p<0.05; **p<0.001; ***p<0.0001 by Bonferroni post-hoc analysis. Scale bar = 50m. Figure 6. Further reduction of overall Grn levels in Het-Nes-cKO mice is still not sufficient to cause neuropathological changes. (A) Grn measured from whole brain lysate by ELISA in 3-4 month old mice. Data are presented as mean SEM. N = 7 WT and 5 Floxed (combined), 5 Het and 7 Flox-het (combined), 2 Nes-cKO, 7 Het-NescKO, and 4 GrnKO mice. (B) Quantification of lipofuscin accumulation in the thalamus, measured as autofluorescence. (C) Quantification of GFAP immunoreactivity in the thalamus. (D) Quantification of Iba1 immunoreactivity in the thalamus. For (B-D), N = 4 Floxed, 7 Het, 6 Nes-cKO, 7 Het-Nes-cKO, and 4 GrnKO mice. For each region, 4
23
ACCEPTED MANUSCRIPT
AC
CE
PT
ED
M
AN
US
CR
IP
T
images per region per animal were quantified. *p<0.05; **p<0.001; ***p<0.0001 by Bonferroni post-hoc analysis. Scale bar = 50m.
24
AC
Fig. 1
CE
PT
ED
M
AN
US
CR
IP
T
ACCEPTED MANUSCRIPT
25
AC
CE
PT
ED
M
AN
US
CR
IP
T
ACCEPTED MANUSCRIPT
Fig. 2
26
AC
Fig. 3
CE
PT
ED
M
AN
US
CR
IP
T
ACCEPTED MANUSCRIPT
27
AC
Fig. 4
CE
PT
ED
M
AN
US
CR
IP
T
ACCEPTED MANUSCRIPT
28
AC
Fig. 5
CE
PT
ED
M
AN
US
CR
IP
T
ACCEPTED MANUSCRIPT
29
AC
Fig. 6
CE
PT
ED
M
AN
US
CR
IP
T
ACCEPTED MANUSCRIPT
30
ACCEPTED MANUSCRIPT
Conditional loss of progranulin in neurons is not sufficient to cause neuronal ceroid lipofuscinosis-like neuropathology in mice
Highlights:
T
IP
CR
US AN M ED PT CE
Conditional nestin-specific progranulin knockout (Nes-cKO) mice were generated Nes-cKO mice express no neuronal progranulin Neuropathology in Nes-cKO mice does not mimic the NCL phenotype of Grn-null mice Microglia-derived progranulin is sufficient to prevent a NCL phenotype
AC
31