Content uploaded by Torsten Bossing
Author content
All content in this area was uploaded by Torsten Bossing on Dec 21, 2017
Content may be subject to copyright.
Distinct phenotypes of three-repeat and four-repeat human tau in a
transgenic model of tauopathy
Megan A. Sealey
a
, Ergina Vourkou
c
, Catherine M. Cowan
a
, Torsten Bossing
b
, Shmma Quraishe
a
,
Sofia Grammenoudi
c
, Efthimios M.C. Skoulakis
c
, Amritpal Mudher
a,
⁎
a
Centre for Biological Sciences, University of Southampton, Southampton SO17 1BJ, UK
b
School of Biomedical and Healthcare Sciences, Plymouth University, PL6 8BU, UK
c
Division of Neuroscience, Biomedical Sciences Research Centre “Alexander Fleming”, Vari 16672, Greece
abstractarticle info
Article history:
Received 1 December 2016
Revised 12 April 2017
Accepted 10 May 2017
Available online 11 May 2017
Tau exists as six closely related protein isoforms in the adult human brain. These are generated from alternative
splicingof a single mRNA transcriptand they differ in the absenceor presence of two N-terminal and three or four
microtubule binding domains. Typically all six isoforms have been considered functionally similar. However,
their differential involvement in particular tauopathies raises the possibility that there may be isoform-specific
differences in physiological functionand pathological role. To explore this,we have compared the phenotypesin-
duced by the 0N3R and 0N4R isoforms in Drosophila. Expression of the 3R isoform causes more profound axonal
transport defects and locomotor impairments, culminating in a shorter lifespan than the 4R isoform. In contrast,
the 4R isoform leads to greater neurodegeneration and impairments in learning and memory. Furthermore, the
phosphorylation patterns of the two isoforms are distinct, as is their ability to induce oxidative stress. These dif-
ferences are not consequent to different expression levels and are suggestive of bona fide physiological differ-
ences in isoform biology and pathological potential. They may therefore explain isoform-specific mechanisms
of tau-toxicity and the differential susceptibility of brain regions to different tauopathies.
Crown Copyright © 2017 Published by Elsevier Inc. This is an open access article under the CC BY-NC-ND license
(http://creativecommons.org/licenses/by-nc-nd/4.0/).
Keywords:
3R tau
4R tau
Alzheimer's disease
Drosophila
Isoforms
Tauopathy
1. Introduction
Transcripts from the single microtubule associated protein tau
(MAPT)-encoding gene on human chromosome 17q21.1 are spliced
into six isoforms in the adult brain (Andreadis, 2005). These tau iso-
forms, ranging in size from 352 to 441 amino acids, arise because of al-
ternative splicing of exons 2, 3 and 10 leading to the absence or
presence of 1 or 2 N-terminal domains and 3 or 4C-terminal microtu-
bule bindingrepeats (Goedert et al., 1989). They are commonly referred
to as 0N3R, 1N3R, 2N3R, 0N4R, 1N4R and 2N4R tau. Furthermore, tau
isoforms undergo a variety of post-translational modifications including
Ser/Thr andTyr phosphorylation, acetylation and SUMOylation.Some of
these modifications occur physiologically and are regulated during de-
velopment and aging; others occur in pathological conditions and are
implicated in tau-mediated toxicity (reviewed in (Medina et al., 2016;
Huefner et al., 2013)).
It has been argued that regulation of alternative splicing during de-
velopment is a mechanism for radically altering the function of tau pro-
tein. This may be reflected in expression of 3R isoforms early in human
brain development during axon path finding when a more dynamic cy-
toskeleton is required and then transitioning to expression of 4R iso-
forms post neurite elaboration, when a more stable network has been
established (Andreadis, 2005). Accordingly, a main distinction of tau
isoforms involves differentiation of the microtubule-binding repeats.
This likely underlies differences in isoform physiology and pathological
potential as they ostensibly interact with distinct or partially overlap-
ping membrane-associated, cytosolic and cytoskeletal proteins. In fact,
differences in microtubule binding properties were well-known
(Goode et al., 2000), but several studies have now demonstrated addi-
tional isoform-specific differences including: the propensity of tau to
aggregate (Adams et al., 2010), differential templated seeding capabili-
ties (Dinkel et al., 2011), intra-neuronal re-localisation during tangle
formation (Hara et al., 2013; Liu and Gotz, 2013), interactions with dis-
tinct cellular binding partners (Bhaskar et al., 2005; Liu et al., 2016),
phosphorylation potential and the impact of these differences on their
biochemical properties (Combs et al., 2011).
The ratio of 3R to 4R isoforms in the adult human brain is approxi-
mately 1. The equimolar isoform ratio is disrupted in some familial
tauopathies due to splicing mutations, which lead to elevation of the
4R tau isoforms (Andreadis, 2005). Even in Alzheimer's Disease (AD)
there is evidence of impaired 3R/4R ratio in tangle bearing neurons
(Niblock and Gallo, 2012; Park et al., 2016). The fact that disrupting
Neurobiology of Disease 105 (2017) 74–83
⁎Corresponding author.
E-mail address: A.Mudher@soton.ac.uk (A. Mudher).
Available online on ScienceDirect (www.sciencedirect.com).
http://dx.doi.org/10.1016/j.nbd.2017.05.003
0969-9961/Crown Copyright © 2017 Published by Elsevier Inc. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
Contents lists available at ScienceDirect
Neurobiology of Disease
journal homepage: www.elsevier.com/locate/ynbdi
the isoform ratio is associated with disease, demonstrates the impor-
tance of maintaining the 3R:4R isoform balance in healthy neurons. Ad-
ditionally, not all isoforms are present in tau aggregates that
characterise particular tauopathies, including sporadic forms. In AD for
example, all tau isoforms form filaments, whereas in others they are
comprised predominately of either 3R (e.g. Pick's disease), or 4R iso-
forms (e.g. Progressive Supranuclear Palsy, Corticobasal Degeneration,
Argyrophilic Grain Disease) (Rabano et al., 2013; Spillantini and
Goedert, 2013).
When divergent phenotypes are reported in animal models of
tauopathy, the particular isoform expressed is not typically considered
as a reason for the discrepancy. Here we highlight this by systematically
assessing isoform-specific phenotypes in Drosophila and demonstrate
that distinct tau isoforms can have significantly different effects in iden-
tical assays. This may shed light on the role of isoform-specific differ-
ences in the divergent pathogenic profiles of tauopathies where one of
these isoforms predominates.
2. Materials and methods
2.1. Flies
Female Drosophila melanogaster expressing either the motor neu-
ron-specific driver D42-GAL4, pan-neuronal driver Elav
C155
-GAL4 and
Elav
C155
-GAL4/TubGAL80
ts
(Bloomington Stock Centre), sensory neu-
ron driver panR7-GAL4 or retinal photoreceptor driver GMR-GAL4
were crossed with male flies transgenic for UAS-human 0N3R tau
(UAS-hTau
0N3R
), or UAS-human 0N4R tau (two 4R transgenic lines
were used and they are referred to as hTau
0N4R
and hTau
0N4R⁎
;they
are distinct transgene insertions presenting similar expression levels –
see Suppl. Fig. 1), or with wild-type Oregon-R male flies (WT). All trans-
genic lines and drivers were obtained from the Bloomington Stock cen-
tre (USA), except the UAS-htau
0N4R
and UAS-htau
0N4R⁎
lines, which
were originally generated by Prof. Mel Feany (Brigham and Women's
Hospital, Boston, USA).
2.2. Larval locomotion analysis
As previously described (Sinadinos et al., 2012), wandering third in-
star larvae were allowed to crawl freely in a 10 cm × 10 cm plate, filled
to a depth of approximately 4 mm with dark blue agarose (1% agarose,
0.1% alcian blue), within a bioassay room kept at 21 °C, 30–40% humid-
ity, and controlled lighting conditions. After a 6-minute acclimatisation
period, larvae were placed at the centre of the plate, and were filmed for
2 min using an Ikegami digital video camera and 5 mm camera lens
(Tracksys, UK). 4 such plates were filmed simultaneously. Ethovision
movement analysis software (Noldus Information Technology) was
then used to measure the following parameters of locomotion: velocity
(mm/s); meander, measured as the angle deviated from the straight
path per cm travelled (degrees/cm); and angular velocity (degrees/s).
Further, the frequency of contractions of the body-wall muscles (con-
tractions/min) was measured semi-manually from these video record-
ings, with the experimenter blinded to condition. GraphPad Prism was
used to calculate standard error of the mean, unpaired 2-tailed students
t-test, and/or 1-way ANOVA on the resulting data, as appropriate.
2.3. Adult climbing assay
This assay was performed on cohorts of 15 adult flies, which had
been allowed to mate for 1–2 days after eclosion and then separated
Fig. 1. Drosophila expressing hTau
0N3R
but not hTau
0N4R
in motor neurons displaya motor phenotype, and impaired fast axonaltransport. Locomotor performance in a free-crawling test is
significantly impaired in hTau
0N3R
-expressing larvae compared to controls in terms of meander (A), and frequency of body-wall contractions (C). In contrast locomotion of hTau
0N4R
-
expressing larvae is not different from driver controls. A similar trend is seen for velocity of movement (B) but the differences between genotypes are not significant. (D) In larval
motor neurons, vesicular aggregates (indicative of axonal transport defect) were found in hTau
0N3R
-expressing larvae but not in controls or hTau
0N4R
-expressing larvae. Very few
aggregates were found in hTau
0N4R
-expressing larvae compared to hTau
0N3R
-expressing larvae. Scale bar = 10 μm. For meander : WT v s hTa u
0N3R
p= 0.0009; hTau
0N3R
vs hTau
0N4R
p= 0.0001; WT vs hTau
0N4R
p= 0.4801. For velocity: WT vs hTa u
0N3R
p= 0.1508 (ns); hTau
0N3R
vs hTau
0N4R
p= 0.0928 (ns); WT vs hTau
0N4R
p= 0.7304 (ns). For body wall
contractions: WT vs hTau
0N3R
p= 0.0013; hTau
0N3R
vs hTau
0N4R
p= 0.0001; WT vs hTau
0N4R
p= 0.803 (ns) (Unpaired two-tailed t-tests; n=9–10 per assay). For axonal transport:
WT vs hTau
0N3R
p= 0.0041; hTau
0N3R
vs hTau
0N4R
p= 0.0306; WT vs hTau
0N4R
p= 0.0116 (Unpaired two-tailed t-tests or one-way Anova Bonferroni's multiple comparisons test
n= 5). hTau
0N3R
= {w / +; D42-GAL4 / +; UAShtau0N3R / +}. hTau
0N4R
= {w / +; D42-GAL4 / +; UAShTau0N4R / + −parental line hTau
0N4R
}. WT = {w / +; D42-GAL4 / +; + /
+} on an OreR background.
75M.A. Sealey et al. / Neurobiology of Disease 105 (2017) 74–83
by sex and housed in their testing cohorts. Each week, 6–9hintothe12-
hour light cycle of the flies, they were anaesthetised very briefly(b5s)
with CO
2
and placed in a measuring cylinder in an assay room withcon-
trolled lighting conditions,temperature (23 °C) and humidity (30–40%).
They were given 15 min to recover from anaesthesia and to acclimatise
to the assay room. The measuring cylinder was tapped 3 times upon a
mouse pad to send the flies to the bottom, a video recording was carried
out and paused 10s later when the analysis was conducted. Flies rested
for 2 min, and the procedure was repeated 2 more times. Flies were then
placed onto fresh food until the following week.
2.4. Survival assay
Three cohorts of 10 male flies of each genotype were separated
0–3 days post-eclosion and then transferred to new food twice a week
and scored for deaths three times a week. Flies were housed in a room
with controlled lighting conditions, temperature (23 °C) and humidity
(30–40%). A Kaplan-Meier survival curve was plotted and a Log-rank
(Mantel-Cox) test was performed on the data using GraphPad Prism
software.
2.5. Adult learning and memory assay
To obtain animals for learning and memory assays UAS-hTau
0N3R
and UAS-hTau
0N4R
males were crossed en masse with Elav
c155
GAL4;
+;TubGAL80
ts
at 18 °C. Upon eclosion they were collected in fresh bot-
tles and tau expression wasinduced by placing the adult flies at 30 °C for
12 days withbottle changes every 3 days. On the 11th day the flies were
separated in groups of 50–70 animals in vials and placed back at 30 °C
overnight. All animals were placed in fresh food vials 1–2 h before con-
ditioning. Conditioning assays were performed under dim red light at
24 °C–25 °C and 65%–75% humidity. All experiments were carried out
in a balanced manner, where all genotypes involved in an experiment
were tested per day. Classical learning refers to Pavlovian olfactory aver-
sive conditioning and was performed using the aversive odors benzal-
dehyde (BNZ) and 3-octanol (OCT) diluted in oil (6% v/v for BNZ and
50% v/vforOCT)asconditionedstimuli(CS+andCS−) with the elec-
tric shock unconditioned stimulus (US). For training, a group of 50–70
flies was first exposed to the CS + odor for 40 s paired with 90 V
shock (consisting of twelve 1.25s pulses with 4.5 s inter-pulse intervals,
therefore 8 US/CS pairings were delivered within 40 s of odor presenta-
tion) and then 30 s of air. Subsequently, flies were exposed to the
CS−for 40 s without shock and then 30 s of air. Each experimental trial
included two reciprocal groups, with the CS+ and CS −odors switched.
Three minutes after conditioning, both groups of flies were tested simul-
taneously for preferential avoidance of the conditioned odorant.
For 24-hour memory experiments, flies were submitted to 12 US/CS
pairings per round and five such rounds of training with a 15-minute
inter-round interval. The flies were stored at 18 °C for 24 h and then
transferred to a T-maze apparatus and allowed to choose between the
two odors for 90 s. A performance index (PI) was calculated as the frac-
tion of flies that avoided the CS+ minus the fraction that avoided the
CS−odors divided by the total number of flies in the experiment. A
final PI is the average of the scores from the two groups of flies trained
with either benzaldehyde or 3-octanol as CS + and ranges from 0 to 100.
2.6. Tau solubility assay to enrich for oligomeric tau species
This assay enriches for insoluble oligomeric tau species as described
(Cowan et al., 2015). A total of 10 fly heads were homogenized in 40 μl
of TBS/sucrose buffer (50 mM Tris-HCl pH 7.4, 175 mM NaCl, 1 M su-
crose, 5 mM EDTA and protease inhibitor cocktail). The samples were
then spun for 2 min at 1000 g and the pellet discarded. The supernatant
was spun at 186,000 g for 2 h at 4 °C. The resulting supernatant was
“S1”–the aqueous soluble fraction. The pellet was re-suspended at room
temperature in 5% SDS/TBS buffer (50 mM Tris-HCl pH 7.4, 175 mM
NaCl, 5% SDS) and spun at 186,000 g for 2 h at25 °C. The resulting super-
natant was “S2”–the SDS-soluble, aqueous-insoluble fraction. The pel-
let was resuspended at room temperature in 5% SDS/TBS buffer (50 mM
Tris-HCl pH 7.4, 175 mM NaCl, 5% SDS and protease inhibitor cocktail)
and spun at 186,000 g for 2 h at 25 °C as a wash spin; following which
the supernatant was discarded. This pellet was then re-suspended in
8Murea,8%SDSbuffer(50mMTris-HClpH7.4,175mMNaCl,8%
SDS, 8 M urea and protease inhibitor cocktail) and agitated for
12–18 h at room temperature (“S3”). All samples were diluted in
2 × Laemmli buffer and boiled for 5 min. “S1”and “S2”were loaded
equally (equivalent volumes) whereas double the amount of “S3”was
loaded compared to “S1”and “S2”. The S3 fraction was then quantified
as a proportion of the sum total of all three fractions.
2.7. Protein oxidation assay (OxyBlot)
For each condition, 5 heads of 1 day-old flies were homogenized in
30 μl OxyBlot buffer (150 mM NaCl, 50 mM MES, 1% triton-X 100, 1%
SDS, 2% β-mercaptoethanol, protease inhibitor cocktail). Homogenates
were centrifuged for 5 min at 5000g, and the pellets discarded. 10 μlho-
mogenate was used for a carbonyl derivatisation reaction with the
OxyBlot kit (Millipore), according to the manufacturer's instructions.
Briefly, we added 10 μl 12% SDS, 20 μl DNPH (or negative control
Fig. 2. Differential Tau isoform toxicity in adult Drosophila. (A) Survival curves for Elav-
GAL4 driven hTa u
0N3R
,hTau
0N4R
and WT male flies (n= 30). HTau
0N3R
flies have
significantly shorter lifespan com pared with both ht au
0N4R
and WT flies (Log-rank,
Mantel-Cox test p= 0. 0001). B) Comparison of the climbing ability with age over
6 weeks for hTau
0N3R
, hTau
0N4R
and WT flies (n= 30). (2-way ANOVA; p= 0.0002).
Error bars are plotted as ±S.E.M. hTau
0N3R
= {w / +; Elav-GAL4 / +; UAShtau0N3R /
+}. hTau
0N4R
= {w / +; Elav-GAL4 / +; UAShTau
0N4R
/+−parental line hTau
0N4R
}.
WT = {w / +; Elav-GAL4 / +; + / +} on an OreR background.
76 M.A. Sealey et al. / Neurobiology of Disease 105 (2017) 74–83
provided) and incubated 15 min at room temperature; then added 14 μl
of neutralizing solution. 10 μlofthisfinal labelled product was applied
to nitrocellulose membrane (Amersham) using a slot blot apparatus
(BioRad). Membranes were probed with anti-DNP antibody (Millipore,
1:150), and signal was detected using fluorescently conjugated anti-
rabbit secondary antibody (LICOR) and a LiCor scanner with Odyssey
software. Resulting band densities were measured using Image J
software.
2.8. Western blotting
Western blotting wasperformed to assess total tau levels, phosphor-
ylation and solubility state of tau. For Western blot analysis of larval
samples, 10 3rd instar larvae or heads of 1-3d adult flies were pooled
and homogenized in 200 μl1×Laemmlibuffer,boiledat95°Cfor
5 min and centrifuged for 5 min at 14000 RPM, at RT. Proteins were sep-
arated by SDS-PAGE according to standard methods, and transferred to
PVDF membrane by semi-dry transfer Anti-Syntaxin (Developmental
Hybridoma Bank) at 1:3000 was used as loading control. Primary
antibodies were used as follows: anti-human tau (Dako, 1:15,000
or T46, 1:3000). The phosphorylation-specific anti-tau antibodies
Ser396/Ser404 (PHF-1) (a gift from Peter Davies, USA, 1:500),
Ser396 (Source Biosciences, 1:2000), Ser202/Thr205 (AT8) (Thermo
Scientific, 1:1000), Thr212/Ser214 (AT100) (Pierce Endogen, 1:1000),
dephosphorylated at Ser199/Ser202/Thr205 (Tau-1) (Millipore, 1:2000),
pS262 (Invitrogen, 1:1000). MC1 (a gift from Peter Davies, USA), was
used at 1:200. Secondary antibodies were at 1:5000 and the signal de-
tected by chemiluminescence (ECL plus).
2.9. Axonal transport studies
Wandering third instar (L3) larvae (day 5) were anaesthetised by
placing larvae in a chamber containing cotton wool soaked in
diethylether vapour for 15 min. Larvae were immobilised on glass slides
in 1% agarose ventral face up and mounted under coverslips. Peripheral
nerves were analysed between the 2nd and 4th denticle bands. For total
area acquisition, vGFP accumulates were imaged at × 63 on an
Axioplan2 Epifluorescence Microscope (Zeiss) and thresholded in
Metamorph software (Molecular Devices, CA, USA). n=5foreach
genotype.
2.10. Immuno-histochemistry
Anaesthetized Flies were decapitated and the brains dissected in
PBS. Brains were fixed for 20 min with 4% formaldehyde in PBS with
0.4% Triton-X100, 10 mM EGTA and 50 mM MgCl
2
added. Brains were
washed five times and incubated for 1 h with 10% Newborn Calfserum
in PBS-T (PBS with 0.4% Triton-X100). Primary antibodies (anti-RFP,
mouse, 1:100, abcam; anti-Tau, 1:2000, DAKO; anti-chaoptin, 1:50,
DSHB, Iowa) were incubated at 4 °C. Washing consisted of five repeti-
tions of 3 rinses and 20 min incubation with PBS-T. Secondary antibod-
ies coupled to Alexa 488 or 568 (1:500 in PBS-T) were incubated
overnight at 4 °C. After the final wash brains were embedded in
Vectashield/70% Glycerol (3:1). For every genotype five brains were re-
corded using a Zeiss 710 confocal microscope using. Controls were im-
aged first and experimental brains were imaged with the same
settings. Images were assembled using Photoshop.
Fig. 3. The expression of 4 repeat human Tau disrupts R7 sensory neurons more severely than 3 repeat human Tau. R7 sensory neurons in the Drosophila visual system express the
membrane marker myristolated-red fluorescent protein (myr-RFP) together with hTau
0N3R
(upper row), hTau
0N4R
(middle row) or on its own (RFP, bottom row). Adult brains were
dissected 5 days (first column), 20 days (second and third column) or 40 days (fourth and fifth column) after eclosion. Brains were stained with antibodies against RFP or human tau
(indicated at the top of each column). Images show axons in the medulla. Scale Bar = 10 μm. (A–C) In 5 day-old brains, RFP expression in R7 axons is not affected by the expression of
either tau isoform. (D–F) In 20 day-old brains expression of hTau
0N3R
(D) and hTau
0N4R
(E) results in a weaker membrane RFP signal than in controls (F) but axons are still intact. (G,
H) Expression of hTau
0N3R
(G) is stronger than h Tau
0N4R
(H) and hTau
0N3R
shows a tendency to form aggregates at synapses. (I–K) In 40 day-old brains expressing either isoform (I, J)
membrane RFP is severely reduced (compar e I, J with K). In particular, hTau
0N4R
expression results in loss of RFP expression in broad areas in the medulla. (L, M) hTau
0N3R
(L)
aggregates along the axons and at synapses and is still expressed stronger than hTau
0N4R
(M). hTau
0N3R
= {w / +; panR7-GAL4 / +; UAShtau
0N3R
/ +}. hTau
0N4R
= {w / +; panR7-
GAL4; UAShTau
0N4R
/+−parental line hTau
0N4R
}. WT = {w / +; panR7-GAL4; + / +} on an OreR background.
77M.A. Sealey et al. / Neurobiology of Disease 105 (2017) 74–83
Fig. 5. No differences between 3R and 4R transgenics in the amount of insoluble tau oligomers formed with age. (A) Representative Western blots of soluble (S1), SDS-soluble (S2) and
SDS-insoluble (S3) fractions generated from adult heads following Elav-GAL4 driven hTau
0N3R
and hTau
0N4R
expression in newly eclosed young (0 weeks) and old (6 weeks) flies. Some
insoluble tau oligomeric speciesare detected in bothyoung and old flies.B) However quantification of S3 fractionrelative to sum totalof all fractions showsthere is no significantdifference
in the amount of insoluble tau between hTau
0N3R
and hTau
0N4R
expressing flies or with age in either line (n= 4). Unpaired, two-tailed t-tests used to test for significance. Error bars are
plotted± S.E.M. hTau
0N3R
= {w / +; Elav-GAL4/ +; UAShTau
0N3R
/ +}. hTau
0N4R
= {w / +; Elav-GAL4/ +; UAShTau
0N4R
/+−parentalline hTau
0N4R
}. WT = {w / +; Elav-GAL4/ +; + /
+} on an OreR background.
Fig. 4. The expression of 4 repeat human Tauimpairs learning and memory but expression of 3R tau does not. Learning andassociative memory was probed in transgenic lines in which
adult specific expression of hTau
0N3R
,hTau
0N4R
was driven by Elav-GAL4 / TubGAL80
ts
. The transgenes were induced progeny of these crosses raised at 18 °C by transferring to 30 °Cfor
12 days prior to testing. Expression of hTau
0N4R
caused severe impairment in learning (pb0.001, Dunnett's test, nN12 per genotype) (A), and memory (pb0.001, Dunnett's test,
nN16)(B), but expressionof hTau
0N3R
did not affecteither learning (p= 0.4585,Dunnett's test, nN12)(C),or LTM (p= 0.142, Dunnett's test,nN14 (D). hTau
0N3R
= {w / +; Elav-GAL4 /
TubGAL80
ts
/+;UAShtau
0N3R
/ +}. hTau
0N4R
= {w / +; GAL4 / TubGAL80
ts
/+;UAShTau
0N4R⁎
/+−parental line hTau
0N4R⁎
}.
78 M.A. Sealey et al. / Neurobiology of Disease 105 (2017) 74–83
3. Results
3.1. Human 3-repeat tau and human 4-repeat tau expression present differ-
ent phenotypes
To test whether the 0N3R and 0N4R isoforms yield identical, similar
or distinct effects on larval mobility (Sinadinos et al., 2012), the UAS-
hTau
0N3R
and UAS hTau
0N4R
transgenes were expressed in motor neu-
rons under D42-GAL4. Using a semi-automated method to track larval
locomotion (Sinadinos et al., 2012), we confirmed our previous obser-
vations that expression of hTau
0N3R
in larval motor neurons manifests
in locomotor defects. Significant impairments were evident in two lar-
val locomotor parameters: meander and contractions (Fig. 1A–C),
which arise from impaired axonal transport (Fig. 1D). In contrast, ex-
pression of hTau
0N4R
did not result in significant locomotor deficits, or
cause axonal transport impairments as profound as those induced by
hTau
0N3R
(Fig. 1A–D).
Isoform-specific phenotypes were also revealed in adult animals
upon transgene expression with Elav-GAL4. Premature lethality was ap-
parent in hTau
0N3R
-expressing adult flies earlier than hTau
0N4R
-express-
ing animals (Fig. 2A). In addition, isoform-specific differential effects
were revealed on a negative geotaxis locomotor assay (Mudher et al.,
2004) in adult flies expressing pan-neuronally the two tau isoforms.
The climbing ability of flies expressing hTau
0N3R
starts to decline at
1 week and diminishes rapidly as animals progress to week 5 and 6,
when the majority of the flies are virtually immobile (Fig. 2B). By com-
parison, the climbing ability of flies expressing hTau
0N4R
begins to dete-
riorate significantly one week later (week 2) and even at week 5, many
of the flies remain mobile (Fig. 2B). Similar results were observed with
an independent UAS-0N4R transgene insertion (Suppl. Fig. 2),
supporting the idea that this difference between the tau isoforms is of
biological significance and not because of differential expression levels
due to transgene insertion. Therefore, expression of 0N3R tau appears
to precipitate more severe effects than 0N4R expression in the same
neurons.
Collectively the results indicate differential effects of the two tau iso-
forms on survival and larval and adult locomotion. Therefore, we won-
dered whether such differential isoform-specific effects may be
revealed in additional neuronal subpopulations.
We selected a subset of eye sensory neurons to assay the effects of
these isoforms since the fly retina has been used extensively to study
tau-dependent neurodegeneration. Each tau isoform was co-expressed
with membrane-tagged RFP (myristolated-RFP) in R7 sensory receptor
neurons (pan-R7-GAL4). Degenerationwas not apparent at 5 days post-
eclosion evidenced by anti-RFP staining of axons following expression
of either isoform (Fig. 3A–C). By day 20 however, degeneration was ap-
parent in the sensory neurons expressing either of the isoforms, but
with those expressing hTau
0N4R
(Fig. 3E, H) presenting more extensive
aberrations than those expressing hTau
0N3R
(Fig. 3D, G). By 40 days
post-eclosion, the sensory neurons expressing htau
0N4R
had largely
degenerated with only few axons remaining (Fig. 3J, M). In contrast,
more axons remained in animals expressing hTau
0N3R
at this time
time-point (Fig. 3I, L). This data indicates that when expressed in the
adult visual sensory neurons, hTau
0N4R
exhibits a stronger neurodegen-
erative phenotype than hTau
0N3R
. Interestingly, in 40 day-old hTau
0N4R
brains myr-RFP expression is nearly absent(Fig. 3J), but accumulation of
tau along the axons persists (Fig. 3M), indicating loss of membrane in-
tegrity leaving ‘ghost’axonal scaffolds behind. We confirmed the loss
of membrane integrity by expressing the tau isoforms in all photorecep-
tors (under GMR-GAL4) and using antibodies against the membrane
glycoprotein Chaoptin (Hirai-Fujita et al., 2008). In 40 day-old hTau
0N4R
expressing optic lobes, Chaoptin is completely absent, whereas al-
though severely reduced in hTau
0N3R
expressing neurons, it is still de-
tectable (Suppl. Fig. 3). Therefore, the results from both myr-RFP and
Chaoptin membrane markers confirm that although tau expression dis-
rupts membrane integrity, the severity is isoform-specificwith
hTau
0N4R
precipitating a stronger neurodegenerative phenotype than
hTau
0N3R
.
To further explore whether such isoform-dependent differences
persisted in other adult assays, we undertook conditional pan neuronal
expression of both isoforms in the adult CNS because hTau
0N4R
expres-
sion has been reported to yield learning deficits in this assay
(Papanikolopoulou and Skoulakis, 2015). Hence we investigated
whether adult specific expression of 0N3R might also precipitate such
deficits in learning and 24 h memory (long term memory-LTM). Sur-
prisingly, while 0N4R expression impaired associative learning (Fig.
4A) in agreement with prior results (Papanikolopoulou and Skoulakis,
2015; Kosmidis et al., 2010) and LTM was similarly significantly im-
paired (Fig. 4B), expression of hTau
0N3R
did not affect either of these
processes (Fig. 4C, D).
Fig. 6. Qualitative assessment of phosphorylation state in hTau
0N3R
and hTau
0N4R
larvae.
Tau expression wa s driven using th e D42-GAL4 moto r neuron specific driver. Tau
phosphorylation was assessed in wandering third instar larvae. Representative blots are
shown for at least tw o independent experiments. Both isoforms of tau were
phosphorylated to similar extents at the pS262 site (A), AT180 site (D) and pS396 site
(E). However hTau
0N4R⁎
was less phosphorylated at the AT8/Tau-1 site (B and C) and
AT100 site (F). Expression of both isoforms of tau is comparable (G). hTau
0N3R
={w/
+; D42-GAL4 / +; UAShTau
0N3R
/+}.hTau
0N4R
= {w / +; D42-GAL4 / +;
UAShTau
0N4R⁎
/+−parental line hTau
0N4R⁎
}. WT = {w / +; D42-GAL4 / +; + / +} on
an OreR background.
79M.A. Sealey et al. / Neurobiology of Disease 105 (2017) 74–83
Collectively then, our data demonstrate that distinct larval and adult
neuronal populations are differentially sensitive to the neuro-toxic ef-
fects of 0N3R and 0N4R tau isoforms, precipitating phenotypic strength
differences, or lack of discernable phenotypes. This in turn strongly sug-
gests that the effects of tau expression in Drosophila are not merely a
consequence of non-specific toxicity or dysfunction due to overexpres-
sion of an exogenous protein. Rather it is likely that human tau isoforms
interact differentially with the same or different intra-neuronal clients
as suggested by the specificity and range of phenotypes described
above.
3.2. Isoforms engage different mechanisms of toxicity
Differences in phenotypic strength can be precipitated by expression
level differences. However this is unlikely to be key for the differences
we report since the expression of the two tau isoforms is comparable
in the transgenic lines we have employed (Suppl. Fig. 1). To investigate
whether the two tau isoforms act differently at the cellular/molecular
level, we assessed their accumulation, solubility, phosphorylation status
and oxidative stress potential because these are other properties impli-
cated in mechanisms of tau toxicity (Alavi Naini and Soussi-Yanicostas,
2015).
We have previously reported that increased tau levels lead to aggre-
gation (Cowan et al., 2015) and accordingly we have found elevated ac-
cumulation of tau with increasing age in both 0N3R and 0N4R adults
(data not shown). We therefore explored whether the age-dependent
accumulation of tau led to its aggregation. Using a commonly used bio-
chemical insolubility assay that enriches for insoluble oligomeric tau
species (Cowan et al., 2015), we found little evidence of significant
levels of insoluble tau oligomers in the brains of adult flies expressing
0N3R or 0N4R pan-neuronally, even in older flies (Fig. 5A). Though in-
soluble oligomeric tau species were not abundant, substantial isoform-
specific aggregate profiles were not revealed upon fractionation of
brain extracts expressing 0N3R or 0N4R (Fig. 5B). We did not investi-
gate whether larger insoluble tau aggregates, such as tau filaments are
found in these transgenics, because they have been described in some
(Wu et al., 2013), but not all Drosophila tauopathy models (Wittmann
et al., 2001).
Of the six isoforms, 0N3R is most highly phosphorylated (Smith et
al., 1995), but whether the other tau isoforms undergo differential
post-translational modifications has been unclear. Therefore we sought
to determine whether the two isoforms were phosphorylated differen-
tially in larval motor neurons implicated inthe isoform-specificlocomo-
tor behaviours and in adult brains to potentially mirror the isoform-
specific learning, memory and longevity differences. Hence, we
expressed 0N3R or 0N4R specifically in larval motor neurons or adult
brains andthen probed the occupation status of a set of phosphorylation
sites implicated in tauopathies.
Though phosphorylation at many sites was similar in larval motor
neurons, there were some interesting isoform-specific differences.
Phosphorylation at Ser 262 (Fig. 6A), Thr 231 (AT180 - Fig. 6D), and
Ser 396 (Fig. 6E) did not appear significantly differentbetween isoforms
in larvae. However, the signal with the AT100 antibody appeared ele-
vated in hTau
0N3R
-expressing larvae (Fig. 6F). Phosphorylation at the
AT8 site was also elevated in hTau
0N3R
-expressing larvae, while in
hTau
0N4R
-expressing animals AT8 phosphorylation was suppressed as
revealed by the Tau-1 antibody which is reactive to non-phosphorylated
epitopes at the AT8 site (Fig. 6 B and C).
Fig. 7. Quantitative assessment of the phosphorylation state of hTau
0N3R
and hTau
0N4R
in adult brains. Tau expression was driven using the ELA V-GAL4 pan-neuronal driver. Tau
phosphorylation was assessed in 1–3 day adult brain s to compare 0N3R tau with two indep endent p-element insertion lines express ing 0N4R tau (referred to as hTa u
0N4R
and
hTau
0N4R⁎
). Representative blots are shown for at least three independent experiments (A) and their qua ntification is shown in (B). Both isoforms of tau were phosphorylated to
similar extents at most sites except for the Tau-1 and PHF-1 sites which showed greater immunoreactivity in 0N3R tau brains. MC1 immunoreactivity, indicative of misfolded tau, was
greater in 0N3R tau. Expression of total tau levels was comparable between all three lines. hTau
0N3R
= {w / +; ELAV-GAL4 / +; UAShtau0N3R / +}. hTau
0N4R
= {w / +; ELAV-GAL4 /
+; UAShTau
0N4R
/+−parental line hTau
0N4R
} and hTau
0N4R⁎
= {w / +; ELAV-GAL4 / +; UAShTau
0N4R⁎
/+−parental line hTau
0N4R⁎
}. WT = {w / +; ELAV-GAL4 / +; + / +} on an
OreR background.
80 M.A. Sealey et al. / Neurobiology of Disease 105 (2017) 74–83
As with the larval motor neurons, the 0N3R and 0N4R isoformswere
similarly phosphorylated at most epitopes in adult brain, but there were
some interesting differences suggesting that developmental stage influ-
ences tau phosphorylation in an isoform-specific manner. These differ-
ences were genuine because they were evident even when the
independent 0N4R p-element insertion line hTau
0N4R⁎
was used (Fig. 7).
In larval motor neurons, 0N3R tau is more phosphorylated than
0N4R tau at the AT8 site (as evidenced by increased AT8 and decreased
Tau-1 immunoreactivity in 0N3R motor neurons) whereas in adult
brain, 0N4R tau is more phosphorylated than 0N3R tau at these sites
(as indicated by greater Tau-1 immunoreactivity in 0N3R brains)
(Figs. 6 and 7). Similarly the greater phosphorylation of 0N3R tau at
the AT100 site in larval motor neurons is not evident in adult brain
(Figs 6 and 7). Instead, 0N3R tau is more phosphorylated at the Ser
396/404 (detected by PHF-1 with an increased trend seen with an anti-
body that only picks up phosphorylation at Ser 396) than 0N4R tau in
adult brains but not in larval motor neurons (Figs 6 and 7). Significantly,
the 0N3R isoform is much more immunoreactive with the MC1 anti-
body than the 0N4R proteins, suggesting differences in folding or pa-
thology-related structure between the two isoforms.
In summary the data implies that 0N3R and 0N4R isoforms are dif-
ferentially phosphorylated at somebut not all sites in different develop-
mental stages. Whether these site-specific phosphorylation differences
underpin the differential phenotypes precipitated by the two isoforms
is yet to be determined, but is consistent with the data. Moreover, be-
cause the phosphorylation profiles of these tau isoforms on tauopathy
associated sites are not identical, the data support the notion of iso-
form-specific interactions with kinases and phosphatases.
Aside from phosphorylation and aggregation, oxidative stress is an-
other mechanism by which tau mediates toxicity (Dias-Santagata et al.,
2007; Alavi Naini and Soussi-Yanicostas, 2015). We therefore examined
oxidative stress using a commercial assay to detect oxidised proteins in
brain homogenates from flies expressing either tau isoform pan-
neuronally. Though there was clear evidence of oxidative stress in
brains from both transgenic lines, surprisingly, oxidation detected in
hTau
0N4R
-expressing flies was twice as much as of that detected in
hTau
0N3R
-expressing animals (Fig. 8). This again demonstrates a clear
isoform-specific difference, which may underpin some of the isoform-
specific differences in toxicity and neuronal dysfunction described
here by us and by others in various tauopathy models.
4. Discussion
We report here isoform-specific phenotypes in both larval and adult
Drosophila expressing eitherhTau
0N3R
or hTau
0N4R
transgenes. Although
expression level differences may contribute to the phenotypic conse-
quences, isoform-specific differences independent of this were uncov-
ered in this study. These results are in agreement with independent
studies, which have also demonstrated isoform-specific differences in
physiological tau biology, including sub-cellular localisation and func-
tion, and disease-relevant biochemical properties (Liu and Gotz, 2013;
Liu et al., 2016).
Differences in the best-described cellular function of tau, microtu-
bule binding, have long been known, and believed to arise because the
4R isoforms possess an extra microtubule-binding domain enabling
three-fold greater microtubule affinity (Goode et al., 2000). However,
the flanking carboxy-terminal region also differentially regulates micro-
tubule-binding, and curiously appears to influence binding of 3R iso-
forms to a greater extent than 4R isoforms (Goode et al., 2000). In
addition to microtubule-binding, isoform specific differences have
been identified in several other physiological roles attributed to tau.
This includes interacting partners, with 2N4R isoforms exhibiting stron-
ger affinity to proteins implicated in neurodegeneration (Liu et al.,
2016) and interactions with kinases such as Fyn, which binds preferen-
tially to 3R tau (Bhaskar et al., 2005). Differences in sub-cellular
localisation have also been uncovered, wherein 0N isoforms appear
preferentially in the soma and axons, 1N isoforms in the dendrites and
nucleus and 2N isoforms in cell bodies and axons (Liu and Gotz,
2013). Such results challenge the widely-held view that tau behaves
preferentially as an axonal protein engaged mostly in microtubule
stabilisation, and jointly with the data herein, promote the idea that
tau is a protein of multiple functions which are likely sub-served differ-
entially by distinct isoforms.
Further support for this notion is provided byisoform-specific differ-
ences in pathological behaviour including propensity to aggregate
(Adams et al., 2010), morphology of aggregates formed (Adams et al.,
2010) and templated seeding (Dinkel et al., 2011). Templated seeding
of filaments in particular, exhibits striking isoform-specific barriers.
While seeds containing 3R isoforms alone or 3R and 4R isoforms togeth-
er can recruit both 3R and 4R monomers into growing filaments, seeds
comprising just of 4R isoforms can only recruit 4R monomers (Dinkel et
al., 2011). Since hyper-phosphorylationof tau is believed to promote its
aggregation, some studies have explored the impact of pseudo-phos-
phorylation on aggregation in vitro and identified interesting isoform-
specific differences (Combs et al., 2011). Although we have revealed iso-
form-specific, developmental stage-dependent phosphorylation differ-
ences, these do not appear to lead to aggregation differences, at least
within the resolution afforded by our techniques, but clearly appear
consistent with the distinct phenotypic consequences we detailed. Dif-
ferences have also been reported in the sub-cellular localisation of the
differenttau isoforms and how this changes during the evolution of tan-
gle pathology (Hara et al., 2013). Collectively these studies begin to elu-
cidate why different tau isoforms are preferentially affected in different
tauopathies and the cellular/molecular basis for predilection of different
brain regions therein.
The studies discussed above show clear isoform-specificdifferences
in the biochemical and pathological properties of tau; however not
many studies have directly compared and contrasted isoform-specific
Fig. 8. Comparison of the protein oxidation induced by expression of hTau
0N4R
versus
hTau
0N3R
-expressing Drosophila. Elav-GAL4 driven pan -neural expression of either
hTau
0N4R
or hTau
0N3R
induces oxidati ve stress in 1d old adu lt flies as measured by a
commercial Oxyblot assay. However levels of protein oxidation are significantly greater
in hTau
0N4R
versus hTau
0N3R
flies. Graph represents the averag e of 9 experiments.
Results from unpaired t-tests: wild-type vs hTau
0N3R
p= 0.0145; hTau
0N3R
vs hTau
0N4R
p= 0.04; wild-type vs hTau
0N4R
(p= 0.008). hTau
0N3R
= {w / +; Elav-GAL4 / +;
UAShTau
0N3R
/+}.hTau
0N4R
= {w / +; Elav-GAL4 / +; UAShTau
0N4R
/+−parental
line hTau
0N4R
}. WT = {w / +; Elav-GAL4 / +; + / +} on an OreR background.
81M.A. Sealey et al. / Neurobiology of Disease 105 (2017) 74–83
phenotypic differences in the same model. Adding to this, we show here
that for some tau-mediated phenotypes, like axonal transport disrup-
tion and adult locomotion, the 3R tau isoform is more detrimental
whereas in other tau-mediated phenotypes, such as learning and mem-
ory and photo-receptor degeneration, it is the 4R tau which is more
toxic. In line with this, we have previously shown that expression of ei-
ther hTau
0N4R
or hTau
2N4R
but not hTau
0N3R
eliminates Drosophila
mushroom bodies (Papanikolopoulou and Skoulakis, 2015;
Grammenoudi et al., 2008; Papanikolopoulou et al., 2010), and that ex-
pression of hTau
0N3R
is associated with dysfunction in the absence of
neuronal death (Mudher et al., 2004; Cowan et al., 2010). Likewise,
many other Drosophila models of tauopathy report degeneration, main-
ly of photoreceptors or other brain regions, but most studies invariably
express 4R isoforms (Dourlen et al., 2016; Chanu and Sarkar, 2016).
Moreover, to our knowledge, no other studies have directly compared
the oxidative potential of 3R and 4R isoforms. We provide evidence
that the 0N4R isoform is more potent at inducing oxidative stress than
0N3R. Oxidative stress has been reported in tauopathies implicating ei-
ther or both 3R and 4R isoforms (reviewed in (Alavi Naini and Soussi-
Yanicostas, 2015)). Whether, as our data suggests, it plays a more pro-
found role in diseases involving 4R remains to be determined. In fact,
of the few studies that have systematically compared 3R and 4R mediat-
ed phenotypes in the same model, work in mice suggests that tipping
the balance towards 4R isoforms was associated with greater pathology
and behavioural defects (Schoch et al., 2016).
The molecular mechanism(s) underpinning the divergent pheno-
types of tau isoforms are unclear at present. Expression level and stabil-
ity differences are likely contributing factors because tau-toxicity is
generally believed to correlate with intraneuronal tau load (Ubhi et al.,
2007). This notion is significantly supplemented and enhanced by our
own results uncovering phenotypic differences between hTau
0N3R
and
hTau
0N4R
, which persisted even in the face of comparable expression
levels. Differences in the epitopes phosphorylated may be another con-
tributing factor since such differences have been associated with differ-
ential toxicity (Brelstaff et al., 2015), or reduced microtubule binding
(Biernat et al., 1993) which underpins tau-mediated neuronal dysfunc-
tion (Cowan et al., 2010; Quraishe et al., 2013).
5. Conclusion
The six tau isoforms are often regarded as the same protein. Indeed
many of their normal biological and pathological characteristics are very
similar. However, there are distinct differences in isoform functional
properties arising from the variable N-terminal domains and 3 or 4 mi-
crotubule-binding domains. This manifests in variations in their post-
translational regulation and in turn their normal cellular functions.
Adding to this, we report here that they are distinctly different in their
pathological potential as well. Such isoform-specific differences need
to be taken into account when interpreting data from experimental
models of tauopathy since they will invariably differ from model to
model. It should also inform tau-centric therapeutic approaches. It re-
mains to be investigated how the tau isoforms contribute to differential
susceptibility of brain region and mechanism of tau-toxicity in different
tauopathies.
Supplementary data to this article can be found online at http://dx.
doi.org/10.1016/j.nbd.2017.05.003.
Acknowledgements
We would liketo thank and acknowledge the Wessex Medical Trust
and Gerald Kerkut Trust. This research has also been co-financed by the
European Union (European Social Fund –ESF) and Greek national funds
through the Operational Program “Education and Lifelong Learning”of
the National Strategic Reference Framework (NSRF) - Research Funding
Program: THALIS –UOA - “Study mechanisms of neurodegeneration in
Alzheimer's disease”.
References
Adams, S.J., DeTure, M.A., McBride, M., Dickson,D.W., Petrucelli, L., 2010. Three repeat iso-
forms of tau inhibit assemblyof four repeat tau filaments.PLoS One 5, e10810. http://
dx.doi.org/10.1371/journal.pone.0010810.
Alavi Naini, S.M., Soussi-Yanicostas, N., 2015. Tau hyperphosphorylation and oxidative
stress, a critical vic ious circle in neurodegenerative tauopath ies? Oxidative Med.
Cell. Longev.:151979 http://dx.doi.org/10.1155/2015/151979.
Andreadis, A., 2005. Tau gene alternative splicing: expression patterns, regulation and
modulation of function in normal brain and neurodegenerative diseases. Biochim.
Biophys. Acta 1739:91–103. http://dx.doi.org/10.1016/j.bbadis.2004.08.010.
Bhaskar,K., Yen, S.H., Lee, G., 2005. Disease-related modificationsin tau affect the interac-
tion betweenfyn and tau. J. Biol.Chem. 280:35119–35125. http://dx.doi.org/10.1074/
jbc.M505895200.
Biernat, J., Gustke, N., Drewes, G., Mandelkow, E.M., Mandelkow, E., 1993. Phosphoryla-
tion of Ser262 strongly reduces binding of tau to microtubules: distinction between
PHF-like immunoreactivity and microtubule binding. Neuron 11, 153–163.
Brelstaff, J., et al., 2015. The fluorescent pentameric oligothiophene pFTAA identifies fila-
mentous tau in live neurons cultured from adult P301S tau mice. Front. Neurosci. 9:
184. http://dx.doi.org/10.3389/fnins.2015.00184.
Chanu, S.I., Sarkar, S., 2016. Targeted downregulation of dMyc suppresses pathogenesis of
human neuronal tauopathies in Drosophila by limiting heterochromatin relaxation
and tau hyperphosphorylation. Mol. Neurobiol. http://dx.doi.org/10.1007/s12035-
016-9858-6.
Combs, B., Voss, K., Gamblin, T.C., 2011. Pseudohyperphosphorylation has differential ef-
fects on polymerization and function of tau isoforms. Biochemistry 50:9446–9456.
http://dx.doi.org/10.1021/bi2010569.
Cowan, C.M., Bossing, T., Page, A., Shepherd, D., Mudher, A., 2010. Soluble hyper-phosphor-
ylated tau causes microtubule breakdown and functionally compromises normal tau in
vivo. Acta Neuropathol. 120:593–604. http://dx.doi.org/10.1007/s00401-010-0716-8.
Cowan, C.M., et al., 2015. Rescue from tau-induced neuronal dysfunction produces insol-
uble tau oligomers. Sci. Report. 5:17191. http://dx.doi.org/10.1038/srep17191.
Dias-Santagata, D., Fulga, T.A., Duttaroy, A., Feany, M.B., 2007. Oxidative stress mediates
tau-induced neurodegeneration in Drosophila. J. Clin. Invest. 117:236–245. http://
dx.doi.org/10.1172/JCI28769.
Dinkel, P.D., Siddiqua, A., Huynh, H., Shah, M., Margittai, M., 2011. Variations in filament
conformation dictate seeding barrierbetween three- and four-repeat tau. Biochemis-
try 50:4330–4336. http://dx.doi.org/10.1021/bi2004685.
Dourlen, P., et al., 2016. Functional screening of Alzheimer risk loci identifies PTK2B as an
in vivo modulator and early marker of tau pathology. Mol. Psychiatry http://dx.doi.
org/10.1038/mp.2016.59.
Goedert, M., Spillantini, M.G., Jakes, R., Rutherford, D., Crowther, R.A., 1989. Multiple iso-
forms of human microtubule-associated protein tau: sequences and localization in
neurofibrillary tangles of Alzheimer's disease. Neuron 3, 519–526.
Goode, B.L., Chau, M.,Denis, P.E., Feinstein, S.C., 2000. Structural and functional differences
between 3-repeat and 4-repeat tau isoforms. Implications for normal tau function
and the onset of neu rodegenetati ve disease. J. Biol. Chem. 275:38182–38189.
http://dx.doi.org/10.1074/jbc.M007489200.
Grammenoudi, S., Anezaki, M., Kosmidis, S., Skoulakis, E.M., 2008. Modelling cell and iso-
form type specificity of tauopathies in Drosophila. SEB Exp. Biol. Ser. 60, 39–56.
Hara, M., Hirokawa, K., Kamei, S., Uchihara, T., 2013. Isoform transition from four-repeat
to three-repeat tau underlies dendrosomatic and regional progression of neurofibril-
lary pathology. Acta Neuropathol. 125:565–579. http://dx.doi.org/10.1007/s00401-
013-1097-6.
Hirai-Fujita, Y., Yamamoto-Hino, M., Kanie, O., Goto, S., 2008. N-Glycosylation of the Dro-
sophila neural protein Chaoptin is essential for its stability, cell surface transport and
adhesive activity. FEBS Lett. 582:2572–2576. http://dx.doi.org/10.1016/j.febslet.2008.
06.028.
Huefner, A., Kuan, W.L., Barker, R.A., Mahajan, S., 2013. Intracellular SERS nanoprobes for
distinction of different neuronal cell types. Nano Lett. 13:2463–2470. http://dx.doi.
org/10.1021/nl400448n.
Kosmidis, S., Grammenoudi, S., Papanikolopoulou, K., Skoulakis, E.M., 2010. Differential
effects of tau on the integrity and function of neurons essential for learning in Dro-
sophila. J. Neurosci. Off. J. Soc. Neurosci . 30:464–477. htt p://dx.doi.org/10.1523/
JNEUROSCI.1490-09.2010.
Liu, C., Gotz, J., 2013. Profiling murine tau with 0N, 1N and 2N isoform-specific antibodies
in brain and peripheral organs reveals distinct subcellular localization, with the 1N
isoform being enriched in the nucleus. PLoS One 8, e84849. http://dx.doi.org/10.
1371/journal .pone.0084849.
Liu, C., Song, X., Nisbet,R., Gotz, J., 2016. Co-immunoprecipitationwith tau isoform-specif-
ic antibodies reveals distinct proteininteractions and highlights a putative role for 2N
tau in disease. J. Biol. Chem. 291:8173–8188. http://dx.doi.org/10.1074/ jbc.M115.
641902.
Medina, M., Hernandez, F., Avila, J., 2016. New features about tau function and dysfunc-
tion. Biomolecules 6. http://dx.doi.org/10.3390/biom6020021.
Mudher, A., et al., 2004. GSK-3beta inhibition reverses axonal transport defects and be-
havioural phenotypes in Drosophila. Mol. Psychiatry 9:522–530. http://dx.doi.org/
10.1038/sj.mp.4001483.
Niblock, M., Gallo, J.M., 2012. Tau alternative splicing in familial and sporadic tauopathies.
Biochem. Soc. Trans. 40:677–680. http://dx.doi.org/10.1042/BST20120091.
Papanikolopoulou, K., Skoulakis, E.M., 2015. Temporally distinct phosphorylations differ-
entiate tau-dependent learning deficits and premature mortality in Drosophila.Hum.
Mol. Genet. 24:2065–2077. http://dx.doi.org/10.1093/hmg/ddu726.
Papanikolopoulou, K., Kosmidis, S., Grammenoudi, S., Skoulakis, E.M., 2010. Phosphoryla-
tion differentiates tau-dependent neuronal toxicity and dysfunction. Biochem. Soc.
Trans. 38:981–987. http://dx.doi.org/10.1042/BST0380981.
82 M.A. Sealey et al. / Neurobiology of Disease 105 (2017) 74–83
Park, S.A., Ahn, S.I., Gallo, J.M., 2016. Tau mis-splicing in the pathogenesis of neurodegen-
erative disorders. BMB Rep.
Quraishe, S., Cowan, C.M., Mudher, A., 2013. NAP (davunetide) rescues neuronal dysfunc-
tion in a Drosophila model of tauopathy. Mol. Psychiatry 18:834–842. http://dx.doi.
org/10.1038/mp.2013.32.
Rabano, A., Cuadros, R., Calero, M., Hernandez, F., Avila, J., 2013. Specificprofile of tau iso-
forms in argyrophylic grain disease. Exp. Neurosci. 7:51–59. http://dx.doi.org/10.
4137/JEN.S12202.
Schoch, K.M., et al., 2016. Increased 4R-tau induces pathological changes in a human-tau
mouse model. Neuron 90:941–947. http://dx.doi.org/10.1016/j.neuron.2016.04.042.
Sinadinos, C., Cowan, C.M., Wyttenbach, A., Mudher, A., 2012. Increased throughput as-
says of locomotor dysfunction i n Drosophila larvae. J. Neurosci. Methods 203:
325–334. http://dx.doi.org/10.1016/j.jneumeth.2011.08.037.
Smith, C.J., Anderton, B.H., Davis, D.R., Gallo, J.M., 1995. Tau isoform expression and phos-
phorylation state during differentiation of cultured neuronal cells. FEBS Lett. 375,
243–248.
Spillantini, M.G., Goedert, M., 2013. Tau pathology and neurodegeneration. Lancet Neurol.
12:609–622. http://dx.doi.org/10.1016/S1474-4422(13)70090-5.
Ubhi, K.K., Shaibah, H.,Newman, T.A., Shepherd, D., Mudher, A., 2007. A comparison of the
neuronal dysfunction caused by Drosophila tau and human tau in a Drosophila model
of tauopathies. Invertebr. Neurosci. 7:165–171. http://dx.doi.org/10.1007/s10158-
007-0052-4.
Wittmann,C.W., et al., 2001. Tauopathyin Drosophila: neurodegeneration without neuro-
fibrillary tangles. Science 293:711–714. http://dx.doi.org/10.1126/science.1062382.
Wu, T.H., et al., 2013. Loss of vesicular dopamine release precedes tauopathy in degener-
ative dopaminer gic neurons in a Drosophila model expressing human tau. Acta
Neuropathol. 125:711–725. http://dx.doi.org/10.1007/s00401-013-1105-x.
83M.A. Sealey et al. / Neurobiology of Disease 105 (2017) 74–83
