HALLMARKS OF LATE-ONSET ALZHEIMER’S DISEASE IN A …
1
HALLMARKS OF LATE-ONSET ALZHEIMER’S DISEASE IN A HUMANIZED MOUSE MODEL For further information, please see: • MODEL-AD: www.modelad.org • AMP-AD knowledge portal: www.ampadportal.org • JAX AD models: https://www.jax.org/alzheimers • AlzForum research models: http://www.alzforum.org/ research-models Kevin P. Kotredes , 1 Christoph Preuss, 1 Ravi Pandey, 1 Paul Territo, 2 Adrian Oblak, 2 Rima Kaddurah-Daouk, 6 Matthias Arnold, 7 Nicholas Seyfried, 8 Duc Duong, 8 Harriet Williams, 1 Bruce Lamb, 2 Stacey Rizzo, 3 Gregory Carter, 1 Mike Sasner, 1 and Gareth Howell 1 of the MODEL-AD Center 1,2,3,4,5 ABSTRACT Alzheimer’s disease (AD) is a debilitating neurodegenerative disorder affecting nearly 50 million patients worldwide, with no effective treatment currently available. Of this AD population, those suffering from late-onset Alzheimer’s Disease (LOAD) represent the largest subset of dementia patients. Unfortunately, current animal models do not fully recapitulate LOAD, thus are not ideal for therapy development. If progress is to be made in the pursuit to cure AD, researchers and clinicians require better resources to understand this heterogenous pathology. Thus, the Model Organism Development and Evaluation for Late-onset AD consortium (MODEL-AD) is charged with creating, defining, and distributing novel mouse models of LOAD carrying human-relevant risk factors for broad use. Characterization of these variants in aging mice will reveal more appropriate disease pathways useful in the treatment of LOAD. RESULTS As expected, all cohorts displayed increased frailty index scores, a measure of aging and vulnerability, by 24 months however in the absence of additional environmental or genetic risk factors, B6.APOEe4.Trem2*R47H mice did not display penetrant behavioral phenotypes, but did exhibit an increase in mortality. B6.APOEe4.Trem2*R47H and B6J.APOEe4 mice showed decreased levels of lipoproteins at multiple time points compared to control mice - an effect of human APOE alleles reported previously. PET data showed changes in glucose uptake over time in B6.APOEe4.Trem2*R47H mice, but MRI showed no gross abnormalities of the brain. RNA sequencing revealed age- and sex-related changes in B6.APOEe4.Trem2*R47H brain samples compared to controls, as well as a novel splicing event in Trem2 as a result of the R47H mutation. Analysis of 6-month-old animals revealed differences in serum metabolite levels and in brain proteomes compared to controls. Neuropathology analysis in brain tissue is in progress, as are cytokine panels in brain homogenates and plasma. CONCLUSIONS The MODEL-AD consortium has established a new mouse strain to study the effects of two strong risk factors for LOAD. Characterization of this model is ongoing and will serve as a backbone strain for the further addition of LOAD risk alleles to more closely align phenotypes in the mouse to outcomes observed in human AD • APOEe4 allele mice showed expected aberrant cholesterol levels. • Age-dependent increases in frailty across all strains, however no significant differences in working memory. • Phenotypic changes in both brain perfusion and metabolism by 8 months • Decreased survival of B6J.hAT mice at 24 months • In vivo imaging, transcriptomic, histological, and biochemical analysis continues on all time points Astrocytes/Microglia Neurons Vascular Integrity Aggregate Clearance DAPI GFAP IBA1 S100B DAPI NeuN Ctip2 DAPI CD31/Lectin Fibrin IBA1 DAPI X34 Lamp IBA1 1The Jackson Laboratory, Bar Harbor, Maine; 2Stark Neuroscience Research Institute, Indianapolis, Indiana; 3University of Pittsburgh, Pittsburgh, Pennsylvania; 4Sage Bionetworks, Seattle, Washington; 5University of California, Irvine, California; 6 Duke University, Durham, North Carolina; 7 Helmholtz University, Berlin, Germany; 8 Emory University, Atlanta, Georgia, USA. Figure 7. Immunohistochemistry pipeline. Brain hemispheres sectioned in series at 25µm was stained with various antibody combinations, as outlined in the table above. (Hip.:hippocampus; Ctx.:cortex) METHODS Humanized APOEe4 and the Trem2*R47H mutation, two prominent genetic risk factors for LOAD, were inserted into the genome of C57BL/6J mice to create the B6.APOEe4.Trem2*R47H model. Cohorts of B6.APOEe4.Trem2*R47H (with controls) were established at multiple sites and aged to 4, 8, 12, and 24 months prior to behavioral testing. At the JAX campus, in vivo behavior and wellness assays measured activity (open field and running wheels), frailty, coordination (rotarod), and cognitive ability (spontaneous alternation). Repeated congruently at Indiana University (IU) facilities were in vivo MRI and PET studies using [18F]-FDG and [64Cu]- PTSM. Post-mortem analyses at both sites included blood chemistry, brain immunohistochemistry, transcriptomics, metabolomics, and proteomics. Analyses of subsequent RNA sequencing data in combination with behavioral and molecular phenotypes has yielded additional insight into the transcriptional differences influencing animal health. Neuronal loss Plaques Tangles Gliosis Cognitive Impairments Synaptic Loss Vascular compromise Transcriptomic analysis In vivo imaging 4mo 8mo 12mo 24mo JAX IU IU Figure 1. Blood biochemistry profiling. Non-fasted blood was taken at harvest from mice receiving normal diet. Serum was separated and profiled for analytes (low- density lipoprotein shown). (Age-dependent differences within genotype determined by ANOVA: *p<0.05; **p<0.01; ***p<0.001.) Figure 2. Assessment of frailty and working memory. Health assessment, measured by frailty assay (left), and hippocampal working memory, as measured by percent spontaneous alternations in a continuous alternation y-maze task (right). (Age-dependent differences within genotype determined by ANOVA: *p<0.05; **p<0.01; ***p<0.001.) 45162 0 5 10 Frailty Index Score 4 month 24 month C57BL/6J APOE4 Trem2*R47H APOE4. Trem2*R47H Female *** *** ** **** C57BL/6J APOE4 Trem2*R47H APOE4. Trem2*R47H 0 5 10 15 LDL (mg/dL) Female 4 month 24 month C57BL/6J APOE4 Trem2*R47H APOE4. Trem2*R47H C57BL/6J APOE4 Trem2*R47H APOE4. Trem2*R47H 0 5 10 Frailty Index Score Male *** **** **** **** 4 month 24 month C57BL/6J APOE4 Trem2*R47H APOE4. Trem2*R47H C57BL/6J APOE4 Trem2*R47H APOE4. Trem2*R47H 0 5 10 15 LDL (mg/dL) Male ** 4 month 24 month C57BL/6J APOE4 Trem2*R47H APOE4. Trem2*R47H C57BL/6J APOE4 Trem2*R47H APOE4. Trem2*R47H 0 20 40 60 80 100 Percent Alternation (%) Female ** 4 month 24 month C57BL/6J APOE4 Trem2*R47H APOE4. Trem2*R47H C57BL/6J APOE4 Trem2*R47H APOE4. Trem2*R47H 0 20 40 60 80 100 Percent Alternation (%) Male 4 month 24 month C57BL/6J APOE4 Trem2*R47H APOE4. Trem2*R47H C57BL/6J APOE4 Trem2*R47H APOE4. Trem2*R47H Figure 3. Male mice expressing human both APOE4 and Trem2*R47H alleles are more likely to succumb by 24 months of age than C57BL/6J counterparts. Survival of unaffected animals comprising the 24 month cross-sectional cohort. Statistical differences in morbidity determined by log-rank test. 0 6 12 18 24 0 20 40 60 80 100 Months Percent survival Female C57BL/6J (-) Trem2*R47H (0.2791) APOE4 (0.2543) APOE4.Trem2*R47H (0.0670) 0 6 12 18 24 0 20 40 60 80 100 Months Percent survival Male C57BL/6J (-) Trem2*R47H (0.19) APOE4 (-) APOE4.Trem2*R47H (0.0185) 0.0 0.5 1.0 1.5 2.0 18F-FDG PET SUVR (region/cerebellum) Male 12 month APOE4 APOE4/Trem2 Trem2 C57BL/6J ** * Caudate Putamen Cerebellum Cingulate Cortex Corpus Callosum Hippocampus Medial Orbital Cortex Parietal Association Cortex Prelimbic Cortex Primary Motor Cortex Thalamus Visual Cortex 0.0 0.5 1.0 1.5 2.0 18F-FDG PET SUVR (region/cerebellum) Female 12 month APOE4 APOE4/Trem2 Trem2 C57BL/6J * * ** ** * ** ** * * * * * Caudate Putamen Cerebellum Cingulate Cortex Corpus Callosum Hippocampus Medial Orbital Cortex Parietal Association Cortex Prelimbic Cortex Primary Motor Cortex Thalamus Visual Cortex Figure 4. In vivo neuroimaging of aged mice reveals variations in regional glucose metabolism due to changes in expression of APOE4 and Trem2*R47H risk alleles at 12 months. Positron emission tomography (PET; red scale) of radioactive 18F-FDG marker was used to measure tissue glucose uptake, guided by magnetic resonance imaging (MRI) (black and white) mapping to brain regions of interest, indicated by bregma coordinates (far left). Intensity of PET signal in brains regions, normalized to cerebellum, are quantified. Post-mortem autoradiography of coronal brain tissue is represented (rainbow; far right). Genotype-dependent differences determined by ANOVA: *p<0.05; **p<0.01; ***p<0.001. Plaques/Tau Additional Thios APP AT8 LFB/CV H&E Prussian/Iron blue Figure 5. Transcription profiles of LOAD mice carrying APOE4 and Trem2*R47H risk alleles. PCA analysis of transcriptomic data from all subjects at 4, 8, 12, and 24 months. 4 months 8 months 12 months 24 months Figure 6. Metabolomic and proteomic analysis of blood serum and brain tissue. PCA analysis of analyte levels found in the blood serum and brain tissue of C57BL6/J, 5xFAD, and B6.APOEe4.Trem2*R47H (hAT) female and male mice at 6 months of age. BRAIN METABOLOMES SERUM METABOLOMES B6 F B6 M 5xFAD F 5xFAD M hAT F hAT M BRAIN PROTEOMES Acknowledgements MODEL-AD was established with funding from The National Institute on Aging (U54 AG054345- 01, U54 AG054349-01). Aging studies are also supported by the Nathan Shock Center of Excellence in the Basic Biology of Aging (NIH P30 AG0380770).
Transcript of HALLMARKS OF LATE-ONSET ALZHEIMER’S DISEASE IN A …
HALLMARKS OF LATE-ONSET ALZHEIMER’S DISEASE IN A HUMANIZED MOUSE MODEL
For further information, please see:• MODEL-AD: www.modelad.org• AMP-AD knowledge portal: www.ampadportal.org• JAX AD models: https://www.jax.org/alzheimers• AlzForum research models: http://www.alzforum.org/research-models
Kevin P. Kotredes,1 Christoph Preuss,1 Ravi Pandey,1 Paul Territo,2 Adrian Oblak,2 Rima Kaddurah-Daouk,6 Matthias Arnold,7 Nicholas Seyfried,8 Duc Duong,8 Harriet Williams,1 Bruce Lamb,2 Stacey Rizzo,3 Gregory Carter,1 Mike Sasner,1 and Gareth Howell1 of the MODEL-AD Center1,2,3,4,5
ABSTRACTAlzheimer’s disease (AD) is a debilitating neurodegenerative disorderaffecting nearly 50 million patients worldwide, with no effective treatmentcurrently available. Of this AD population, those suffering from late-onsetAlzheimer’s Disease (LOAD) represent the largest subset of dementiapatients. Unfortunately, current animal models do not fully recapitulate LOAD,thus are not ideal for therapy development. If progress is to be made in thepursuit to cure AD, researchers and clinicians require better resources tounderstand this heterogenous pathology. Thus, the Model OrganismDevelopment and Evaluation for Late-onset AD consortium (MODEL-AD) ischarged with creating, defining, and distributing novel mouse models of LOADcarrying human-relevant risk factors for broad use. Characterization of thesevariants in aging mice will reveal more appropriate disease pathways useful inthe treatment of LOAD.
RESULTSAs expected, all cohorts displayed increased frailty index scores, a measureof aging and vulnerability, by 24 months however in the absence of additionalenvironmental or genetic risk factors, B6.APOEe4.Trem2*R47H mice did notdisplay penetrant behavioral phenotypes, but did exhibit an increase inmortality. B6.APOEe4.Trem2*R47H and B6J.APOEe4 mice showeddecreased levels of lipoproteins at multiple time points compared to controlmice - an effect of human APOE alleles reported previously. PET datashowed changes in glucose uptake over time in B6.APOEe4.Trem2*R47Hmice, but MRI showed no gross abnormalities of the brain. RNA sequencingrevealed age- and sex-related changes in B6.APOEe4.Trem2*R47H brainsamples compared to controls, as well as a novel splicing event in Trem2 asa result of the R47H mutation. Analysis of 6-month-old animals revealeddifferences in serum metabolite levels and in brain proteomes compared tocontrols. Neuropathology analysis in brain tissue is in progress, as arecytokine panels in brain homogenates and plasma.
CONCLUSIONSThe MODEL-AD consortium has established a new mouse strain to studythe effects of two strong risk factors for LOAD. Characterization of thismodel is ongoing and will serve as a backbone strain for the furtheraddition of LOAD risk alleles to more closely align phenotypes in themouse to outcomes observed in human AD• APOEe4 allele mice showed expected aberrant cholesterol levels.• Age-dependent increases in frailty across all strains, however no
significant differences in working memory.• Phenotypic changes in both brain perfusion and metabolism by 8
months• Decreased survival of B6J.hAT mice at 24 months• In vivo imaging, transcriptomic, histological, and biochemical
analysis continues on all time points
Astrocytes/Microglia Neurons Vascular Integrity Aggregate Clearance
DAPIGFAPIBA1S100B
DAPINeuNCtip2
DAPICD31/LectinFibrinIBA1
DAPIX34LampIBA1
1The Jackson Laboratory, Bar Harbor, Maine; 2Stark Neuroscience Research Institute, Indianapolis, Indiana; 3University of Pittsburgh, Pittsburgh, Pennsylvania; 4Sage Bionetworks, Seattle, Washington; 5University of California, Irvine, California; 6Duke University, Durham, North Carolina; 7Helmholtz University, Berlin, Germany; 8Emory University, Atlanta, Georgia, USA.
Figure 7. Immunohistochemistry pipeline. Brain hemispheres sectioned inseries at 25µm was stained with various antibody combinations, as outlined in thetable above. (Hip.:hippocampus; Ctx.:cortex)
METHODSHumanized APOEe4 and the Trem2*R47H mutation, two prominent geneticrisk factors for LOAD, were inserted into the genome of C57BL/6J mice tocreate the B6.APOEe4.Trem2*R47H model. Cohorts ofB6.APOEe4.Trem2*R47H (with controls) were established at multiple sitesand aged to 4, 8, 12, and 24 months prior to behavioral testing. At the JAXcampus, in vivo behavior and wellness assays measured activity (open fieldand running wheels), frailty, coordination (rotarod), and cognitive ability(spontaneous alternation). Repeated congruently at Indiana University (IU)facilities were in vivo MRI and PET studies using [18F]-FDG and [64Cu]-PTSM. Post-mortem analyses at both sites included blood chemistry, brainimmunohistochemistry, transcriptomics, metabolomics, and proteomics.Analyses of subsequent RNA sequencing data in combination with behavioraland molecular phenotypes has yielded additional insight into thetranscriptional differences influencing animal health.
Neuronal loss
Plaques
Tangles
Gliosis
Cognitive Impairments
Synaptic Loss
Vascular compromise
Transcriptomic analysis
In vivo imaging
4mo 8mo 12mo 24mo
JAX
IU
IU
Figure 1. Blood biochemistry profiling. Non-fasted blood was taken at harvest frommice receiving normal diet. Serum was separated and profiled for analytes (low-density lipoprotein shown). (Age-dependent differences within genotype determinedby ANOVA: *p<0.05; **p<0.01; ***p<0.001.)
Figure 2. Assessment of frailty and working memory. Health assessment,measured by frailty assay (left), and hippocampal working memory, as measured bypercent spontaneous alternations in a continuous alternation y-maze task (right).(Age-dependent differences within genotype determined by ANOVA: *p<0.05;**p<0.01; ***p<0.001.)
45162
4mo, 5xFADnull, F4mo, R47H 4mo, 4/4 4mo, A/T24mo, 5xFADnull, F24mo, R47H24mo, 4/424mo, A/T0
5
10
Frai
lty In
dex
Scor
e
4 month 24 month
C57BL/6J
APOE4
Trem2*R
47H
APOE4.
Trem2*R
47H
Female
******
******
C57BL/6J
APOE4
Trem2*R
47H
APOE4.
Trem2*R
47H
4mo, B6, F4mo, R47H, F4mo, 4/4, F4mo, A/T, F24mo, B6, F24mo, R47H, F24mo, 4/4, F24mo, A/T, F0
5
10
15
LDL
(mg/
dL)
Female
4 month 24 month
C57BL/6J
APOE4
Trem2*R
47H
APOE4.
Trem2*R
47H
C57BL/6J
APOE4
Trem2*R
47H
APOE4.
Trem2*R
47H
4mo, 5xFADnull, M4mo, R47H 4mo, 4/4 4mo, A/T24mo, 5xFADnull, M24mo, R47H24mo, 4/424mo, A/T0
5
10
Frai
lty In
dex
Scor
e Male
*** ************
4 month 24 month
C57BL/6J
APOE4
Trem2*R
47H
APOE4.
Trem2*R
47H
C57BL/6J
APOE4
Trem2*R
47H
APOE4.
Trem2*R
47H
4mo, B6, M4mo, R47H, M4mo, 4/4, M4mo, A/T, M24mo, B6, M24mo, R47H, M24mo, 4/4, M24mo, A/T, M0
5
10
15
LDL
(mg/
dL)
Male
**
4 month 24 month
C57BL/6J
APOE4
Trem2*R
47H
APOE4.
Trem2*R
47H
C57BL/6J
APOE4
Trem2*R
47H
APOE4.
Trem2*R
47H
new 4mo, B6, Fnew 4mo, R47H, Fnew 4mo, A, Fnew 4mo, A/T, Fnew 24mo, B6, Fnew 24mo, R47H, Fnew 24mo, A, Fnew 24mo, A/T, F0
20
40
60
80
100
Perc
ent A
ltern
atio
n (%
)
Female
**
4 month 24 month
C57BL/6J
APOE4
Trem2*R
47H
APOE4.
Trem2*R
47H
C57BL/6J
APOE4
Trem2*R
47H
APOE4.
Trem2*R
47H new 4mo, B6, Mnew 4mo, R47H, Mnew 4mo, A, Mnew 4mo, A/T, Mnew 24mo, B6, Mnew 24mo, R47H, Mnew 24mo, A, Mnew 24mo, A/T, M
0
20
40
60
80
100
Perc
ent A
ltern
atio
n (%
)
Male
4 month 24 month
C57BL/6J
APOE4
Trem2*R
47H
APOE4.
Trem2*R
47H
C57BL/6J
APOE4
Trem2*R
47H
APOE4.
Trem2*R
47H
Figure 3. Male mice expressing human both APOE4 and Trem2*R47H alleles aremore likely to succumb by 24 months of age than C57BL/6J counterparts.Survival of unaffected animals comprising the 24 month cross-sectional cohort.Statistical differences in morbidity determined by log-rank test.
0 6 12 18 240
20
40
60
80
100
Months
Per
cent
sur
viva
l
Female
C57BL/6J (-)
Trem2*R47H (0.2791)APOE4 (0.2543)
APOE4.Trem2*R47H (0.0670)
0 6 12 18 240
20
40
60
80
100
Months
Per
cent
sur
viva
l
Male
C57BL/6J (-)
Trem2*R47H (0.19)
APOE4 (-)
APOE4.Trem2*R47H (0.0185)
0 .38
-1 .94
-3 .80
C57B L/6 J APOE4 TREM 2*R47H APOE4 .TR EM 2*R47H
Cau
date
Put
amen
(Dor
sal S
tria
tum
)
Cer
ebel
lum
Cin
gula
te C
orte
x
Cor
pus
Cal
losu
m
Hip
poca
mpu
s (C
A-C
A3)
Med
ial O
rbita
l Cor
tex
Parie
tal A
ssoc
iatio
n C
orte
x (L
ater
al-M
edia
l)
Prel
imbi
c C
orte
x
Prim
ary
Mot
or C
orte
x
Thal
amus
Visu
al C
orte
x (P
rimar
y an
d Se
cond
ary)
0.0
0.5
1.0
1.5
2.0
18F-
FDG
PET
SU
VR (r
egio
n/ce
rebe
llum
)
Male 12 month
APOE4 APOE4/Trem2Trem2C57BL/6J
** *
Caud
ate
Puta
men
Cere
bellu
m
Cing
ulat
eCo
rtex
Corp
usCa
llosu
m
Hipp
ocam
pus
Med
ial O
rbita
lCo
rtex
Parie
tal
Asso
ciat
ion
Cort
ex
Prel
imbi
cCo
rtex
Prim
ary
Mot
orCo
rtex
Thal
amus
Visu
al C
orte
x
Cau
date
Put
amen
(Dor
sal S
tria
tum
)
Cer
ebel
lum
Cin
gula
te C
orte
x
Cor
pus
Cal
losu
m
Hip
poca
mpu
s (C
A-C
A3)
Med
ial O
rbita
l Cor
tex
Parie
tal A
ssoc
iatio
n C
orte
x (L
ater
al-M
edia
l)
Prel
imbi
c C
orte
x
Prim
ary
Mot
or C
orte
x
Thal
amus
Visu
al C
orte
x (P
rimar
y an
d Se
cond
ary)
0.0
0.5
1.0
1.5
2.0
18F-
FDG
PET
SU
VR (r
egio
n/ce
rebe
llum
)
Female 12 month
APOE4 APOE4/Trem2Trem2C57BL/6J
**
****
***
***
*
*
*
*
Caud
ate
Puta
men
Cere
bellu
m
Cing
ulat
eCo
rtex
Corp
usCa
llosu
m
Hipp
ocam
pus
Med
ial O
rbita
lCo
rtex
Parie
tal
Asso
ciat
ion
Cort
ex
Prel
imbi
cCo
rtex
Prim
ary
Mot
orCo
rtex
Thal
amus
Visu
al C
orte
x
Figure 4. In vivo neuroimaging of aged mice reveals variations in regionalglucose metabolism due to changes in expression of APOE4 and Trem2*R47Hrisk alleles at 12 months. Positron emission tomography (PET; red scale) ofradioactive 18F-FDG marker was used to measure tissue glucose uptake, guided bymagnetic resonance imaging (MRI) (black and white) mapping to brain regions ofinterest, indicated by bregma coordinates (far left). Intensity of PET signal in brainsregions, normalized to cerebellum, are quantified. Post-mortem autoradiography ofcoronal brain tissue is represented (rainbow; far right). Genotype-dependentdifferences determined by ANOVA: *p<0.05; **p<0.01; ***p<0.001.
Plaques/Tau Additional
ThiosAPPAT8
LFB/CVH&EPrussian/Iron blue
Figure 5. Transcription profiles of LOAD mice carrying APOE4 and Trem2*R47H risk alleles. PCA analysis of transcriptomic data from all subjects at 4, 8, 12, and 24 months.
4 months 8 months 12 months 24 months
Figure 6. Metabolomic and proteomic analysis of blood serum and braintissue. PCA analysis of analyte levels found in the blood serum and brain tissueof C57BL6/J, 5xFAD, and B6.APOEe4.Trem2*R47H (hAT) female and male miceat 6 months of age.
BRAIN METABOLOMESSERUM METABOLOMES
B6 F B6 M 5xFAD F 5xFAD M hAT F hAT M
BRAIN PROTEOMES
AcknowledgementsMODEL-AD was established with funding from The National Institute on Aging (U54 AG054345-01, U54 AG054349-01). Aging studies are also supported by the Nathan Shock Center of Excellence in the Basic Biology of Aging (NIH P30 AG0380770).
![The neuropathological diagnosis of Alzheimer’s disease... · 2019. 8. 2. · can present as early as age 20, with the average age of onset of 46.2years [11]. Early onset Alzheimer’s](https://static.fdocuments.net/doc/165x107/61054b81f9f2b225a23ac0e1/the-neuropathological-diagnosis-of-alzheimeras-disease-2019-8-2-can.jpg)
