The authors have declared that no competing interests exist.
Conceived and designed the experiments: JB HX GUH. Performed the experiments: JB HX. Analyzed the data: JB HX ADA. Contributed reagents/materials/analysis tools: ADA. Wrote the paper: JB HX GUH.
Progressive Supranuclear Palsy (PSP) is a neurodegenerative disorder characterised by intracellular aggregation of the microtubule-associated protein tau. The tau protein exists in 6 predominant isoforms. Depending on alternative splicing of exon 10, three of these isoforms have four microtubule-binding repeat domains (4R), whilst the others only have three (3R). In PSP there is an excess of the 4R tau isoforms, which are thought to contribute significantly to the pathological process. The cause of this 4R increase is so far unknown. Several lines of evidence link mitochondrial complex I inhibition to the pathogenesis of PSP. We demonstrate here for the first time that annonacin and MPP+, two prototypical mitochondrial complex I inhibitors, increase the 4R isoforms of tau in human neurons. We show that the splicing factor SRSF2 is necessary to increase 4R tau with complex I inhibition. We also found SRSF2, as well as another tau splicing factor, TRA2B, to be increased in brains of PSP patients. Thereby, we provide new evidence that mitochondrial complex I inhibition may contribute as an upstream event to the pathogenesis of PSP and suggest that splicing factors may represent an attractive therapeutic target to intervene in the disease process.
Tauopathies are a heterogeneous group of neurodegenerative diseases with the common feature of intracellular aggregation of the microtubule associated protein tau. They include, but are not limited to, Alzheimer's Disease, Progressive Supranuclear Palsy (PSP), Argyrophilic Grain Disease (AGD), Corticobasal Degeneration (CBD), Pick's Disease and some other forms of frontotemporal dementias. Different tauopathies vary significantly in their clinical and pathological phenotype
In the human central nervous system there are six predominant splicing variants of the
Across different tauopathies the isoform constitution varies. A common classification of tauopathies, therefore, is between the 3R isoform and the 4R isoform tauopathies
Alternative splicing of exon 10 is regulated by a combination of
There are several lines of evidence suggesting a role for dysfunction of the mitochondrial respiratory chain, particularly of mitochondrial complex I, in the pathogenesis of PSP. A study using transmitochondrial cytoplasmic hybrid (cybrid) cell lines expressing mitochondrial genes from persons with PSP found complex I activity to be reduced
Nunc Nunclon Delta 6-well (for protein and mRNA) or 48-well (for cell assays) plates (Thermo Fisher Scientific, Waltham, MA, USA) were coated with 100 µg/ml poly-L-lysine (Sigma-Aldrich, St. Louis, MO, USA) and 5 µg/ml fibronectin (Sigma-Aldrich). LUHMES (Lund Human Mesencephalic) cells, derived from female human embryonic ventral mesencephalic cells by conditional immortalization
Human fresh frozen brain sections of the
RNA from human tissue samples was extracted by grinding the tissue in liquid nitrogen to a powder and then dissolving it in the RA1 buffer supplied as part of the NucleoSpin RNA (Macherey Nagel, Düren, Germany) RNA extraction kit +1% (v/v) 2-Mercaptoethanol (Sigma-Aldrich). RNA from cells was extracted by scraping the cells from the culture plate with RA1 buffer +1% (v/v) 2-Mercaptoethanol. The remaining extraction procedure was according to the manufacturer's instructions for the NucleoSpin RNA kit. RNA concentrations were determined using the NanoDrop 2000c Spectrophotometer (Thermo Fisher Scientific). The RNA was then transcribed into cDNA with the iScript cDNA Synthesis Kit (BioRad, Berkeley, CA, USA) using the manufacturer's instructions. Real-Time PCR was performed on the Applied Biosystems StepOnePlus (Life Technologies) system using TaqMan Universal Master Mix II and TaqMan primers against total
Protein was extracted from cells using the M-PER Mammalian Protein Extraction Reagent (Thermo Fisher Scientific). The protein solution was frozen at −80°C immediately after retrieval and for a minimum of two hours. The solution was then thawed on ice, vortexed, centrifuged at 5000 g for 15 minutes at 4°C and the supernatant retrieved. Total protein concentrations were determined using the BCA kit (Thermo Fisher Scientific) by heating the samples at 60°C for 30 minutes and measuring the absorption on the NanoDrop 2000c Spectrophotometer (Thermo Fisher Scientific). 20 µg of total protein were then adjusted to equal concentrations between samples by dilution with M-PER and subsequently heated at 95°C for 5 minutes with 1× Roti-Load loading buffer (Carl Roth, Karlsruhe, Germany). SDS-PAGE was performed using precast Gels (anyKD, Bio-Rad) in a tris-glycine running buffer (14.4% glycine, 3% Tris, 1% SDS w/v, Carl Roth). The protein was blotted onto PVDF membrane (Bio-Rad) at 70 V for 65 minutes. The membrane was blocked with 1× Roti-Block solution (Carl Roth) for 1h and then incubated at 4°C overnight under gentle shaking with the primary antibody (
Antigen | Clone | Species | Concentration (v/v) | Company |
Human tau | HT7 | Mouse | 1:1000 | Pierce Antibodies, Thermo |
3-repeat tau | 8E6/C11 | Mouse | 1:500 | Millipore |
4-repeat tau | 1E1/A6 | Mouse | 1:300 | Millipore |
Actin (I-19) | Polyclonal | Goat | 1:2500 | Santa Cruz Biotechnologies |
LUHMES cells were seeded out and differentiated as described above and allowed to adhere to the plate floor for 4 h. siRNA (Sigma-Aldrich) targeted against SRSF2 (final concentration 200 nM) and Lipofectamine RNAiMAX (Life Technologies) (final concentration 1.2 µl/ml) were dissolved in separate aliquots of medium (OptiMEM, Life Technologies). The diluted siRNA was then added to the diluted Lipofectamine RNAiMAX. The combined solution was then allowed to incubate for 20 minutes before being added to the cells.
ATP assays were conducted using the ATP test kit by Lonza according to the manufacturer's instructions. Luminescence was read with the FLUOstar Omega platereader (BMG Labtech). The data was analysed using the MARS Data Analysis Software (BMG Labtech).
Thiazolyl Blue Tetrazolium Blue (MTT) (Sigma Aldrich) was dissolved in sterile PBS to a concentration of 5 mg/ml. This stock solution was added to the cells in culture medium to achieve a final concentration of 0.5 mg/ml. The 48-well culture plate was then incubated at 37°C for 1 h, the medium removed completely and frozen at −80°C for 1 h. The plate was then thawed, 300 µl DMSO (AppliChem, Darmstadt, Germany) was added per well and the plate was shaken to ensure complete dissolution of the violet crystals. 100 µl from each well were transferred to a new 96-well plate and the absorbance was read with the platereader at a wavelength of 590 nm (reference wave length 630 nm). The data was analysed using the MARS Data Analysis Software (BMG Labtech).
Prism 6 (GraphPad Software, La Jolla, CA, USA) was used for statistical calculations and for the creation of line and bar graphs. Results were compared by 2-way ANOVA with Sidak post-hoc test, unless stated otherwise. Data are shown as mean ± SEM. P<0.05 was considered significant.
We first characterized expression of tau isoforms in LUHMES cells, a cell culture line of human mesencephalic neurons, derived from female human embryonic ventral mesencephalic cells by conditional immortalization (Tet-off v-myc over-expression)
When treated with annonacin at a concentration of 25 nM for 48 h from days 8 to 10 post differentiation, LUHMES cells remain 60.7±0.4% viable (MTT assay) with an ATP concentration of 64±1% compared to that of untreated cells (
A) LUHMES neurons were treated with different concentrations of annonacin for 48 h from day 8–10 post differentiation (n = 12). The MTT test, a measure for mitochondrial reducing function, and ATP concentration, are expressed as a relative percentage compared to untreated control cells. B) 4R isoform (exon 10) mRNA is upregulated with annonacin treatment. Quantitative PCR results showing the relative quantity of mRNA for different
Under these conditions we observed the mRNA of the 4R isoforms of tau to be upregulated significantly (
We also observed an upregulation of the 4R tau isoforms on the protein level by Western blot (
We next explored the mechanism of how annonacin induces this isoform change. We tested 10 splicing factors known to influence the inclusion or exclusion of exon 10 in the MAPT gene
Splicing factor | Target |
Effect on exon 10 splicing |
SRSF1 (SRp30a, ASF) | PPE | Inclusion |
SRSF2 (SRp30b, SC35) | SC35-like | Inclusion |
SRSF3 (SRp20) | ND | Exclusion |
SRSF4 (SRp75) | ND | Exclusion |
SRSF6 (SRp55) | ND | Exclusion |
SRSF7 (9G8) | ISS | Exclusion |
SRSF9 (SRp30c) | ND | Inclusion |
SRSF11 (SRp54) | PPE | Exclusion |
TRA2B | PPE | Inclusion |
Source: Adapted from
We found SRSF2 to be the only splicing factor significantly upregulated with annonacin treatment that is known to promote the inclusion of exon 10 (
A) Quantitative PCR results for 9 different splicing factors known to have an effect on exon 10 alternative splicing (
We also tested splicing factor expression levels in human brain tissue of the
Case Number | Diagnosis | Cause of Death | Age at death | Braak Stage | Sex | Postmortem delay (hours: minutes) |
P1 | PSP | “Natural death” | 73 | 2C | Male | 4:20 |
P2 | PSP | Acute heart failure | 70 | 3 | Male | 6:50 |
P3 | PSP | Aspiration pneumonia | 73 | 2 | Male | 6:15 |
P4 | PSP | Urinary tract infection | 70 | 1A | Male | 5:20 |
C1 | Non-demented control | Pancreas carcinoma | 70 | 0 | Male | 7:30 |
C2 | Non-demented control | Prostate cancer | 69 | 0 | Male | 5:55 |
C3 | Non-demented control | Lung emboli (clinical suspicion) | 73 | 0 | Male | 24:45 |
C4 | Non-demented control | Sepsis | 71 | 1 | Male | 7:40 |
C5 | Non-demented control | Myocardial infarction | 67 | 1B | Male | 18:35 |
We tested whether 4R isoform upregulation upon annonacin treatment is a non-specific consequence of neuronal injury, specific to mitochondrial complex I inhibition or even more specific to annonacin. We therefore repeated the experiment with 1-methyl-4-phenylpyridinium (MPP+), another complex I inhibitor, 6-hydroxydopamine (6-OHDA), a neurotoxin known to be neurotoxic primarily through oxidative stress
A) ATP concentration and MTT cell viability in LUHMES cells as measured by the MTT assay for different concentrations of 6-OHDA. Treatment was for 48 h from day 8–10 post differentiation. n = 12. B) ATP concentration and MTT cell viability in LUHMES cells as measured by the MTT assay for different concentrations of MPP+. Treatment was for 48 h from day 8–10 post differentiation. n = 12. C) Quantitative PCR results of MAPT splicing variants for LUHMES cells treated with 10 µM MPP+ for 48 h from day 8–10 post differentiation. 3 biological repeats with 3 technical repeats each. **: p<0.01 vs. untreated cells (dotted line), (2-way ANOVA with Sidak's post-hoc test). D) Quantitative PCR results of MAPT splicing variants for LUHMES cells treated with 20 µM 6-OHDA for 48 h from day 8–10 post differentiation. 3 biological repeats with 3 technical repeats each. *: p<0.05 vs. untreated cells (dotted line), (2-way ANOVA with Sidak's post-hoc test). E) Quantitative PCR results of MAPT splicing variants for LUHMES cells starved of nutrients and glucose for 24 h from day 8–9 post differentiation. 3 biological repeats with 3 technical repeats each. *: p<0.05 vs. untreated cells (dotted line), (2-way ANOVA with Sidak's post-hoc test). F) Quantitative PCR results of SRSF2 for LUHMES cells treated with 10 µM MPP+ or 20 µM 6-OHDA for 48 h from day 8–10 post differentiation or starved for 24 h from day 8–9 post differentiation. 3 biological repeats with 3 technical repeats each. *: p<0.05, **: p<0.01 vs. untreated cells (dotted line), (2-way ANOVA with Sidak's post-hoc test).
With MPP+ treatment we observed a significant increase in exon 10 inclusion on the mRNA level by qPCR (
A) 4R isoform protein is upregulated with MPP+ treatment. Western blot for 3R and 4R isoforms of tau protein, as well as total tau (detected with the HT7 antibody). LUHMES cells were either left untreated or treated with 10 µM MPP+. Actin was used as loading control. B) Quantification of
Finally, we explored the role of
In this paper we have been able to add the increase in 4R tau isoforms as an additional feature to the list of characteristics of PSP that annonacin treatment reproduces in cell culture. This makes annonacin treated neurons a good model for PSP and potentially other sporadic 4R tauopathies. It is unique in the fact that it does not rely on any genetic modification of the
However, the effect on alternative splicing is not specific to annonacin. Rather, it seems to be related to mitochondrial complex I inhibition more generally. This is suggested by the fact that we have observed the same increase in 4R tau isoforms with MPP+, another complex I inhibitor. In fact, other features of tauopathy have also been reproduced by other complex I inhibitors
However, it is not yet fully understood to what extent the relative increase in the 4R tau isoform contributes to neurotoxicity or impairment of neural functioning. 4R tau isoform increases are only seen in a selection of tauopathies and are region specific. In Alzheimer's disease, there is no abnormal upregulation of 4R isoforms. In PSP, there is some evidence that the 4R isoform may not be upregulated in the frontal cortex, despite the existence of tau pathology in this region
We have identified SRSF2 as a mediator essential for mitochondrial complex I inhibitor induced exon 10 inclusion. The fact that a knockdown of SRSF2 reverses the annonacin induced increase in 4R tau confirms that SRSF2 plays a necessary role for this isoform shift.
SRSF2 is controlled by several kinases including SRPK, AKT, topoisomerase I and CLK/STY family kinases, as well as lysine acetylation
In our annonacin-treated cell cultures, which might be considered to be an acute model of a sporadic tauopathy, inhibition of SRS2 prevented the 4R isoform shift of tau but not the cell death induced by annonacin. This suggests that, in this model, the 4R tau is not necessary for cell death, since neurons might rather die from reduced energy production
In human PSP patients both the SRSF2 and TRA2B splicing factors are upregulated. This suggests that the 4R upregulation is not exclusively due to complex I inhibition, as in that case we would have expected only SRSF2 to be upregulated. Therefore, exploring upstream events leading to TRA2B upregulation may lead to insights on further reasons for the increase in 4R tau isoforms in some tauopathies. It would also be interesting to compare the splicing factor expression levels in 3R tauopathies versus 4R tauopathies.
If SRSF2 is confirmed to be a key player in mediating the 4R isoform upregulation in PSP and other 4R tauopathies, this would make it a suitable drug target for reducing this isoform shift.
In summary, we can conclude that SRSF2 is a necessary mediator for mitochondrial complex I inhibitor induced tau 4R isoform upregulation. As SRSF2 is also increased in PSP patients this suggests mitochondrial complex I inhibition may play at least a partial role in the pathogenesis of 4R tauopathies such as PSP. However, other mechanisms are also likely to contribute.
Mr Robin Konhäuser and Ms Magda Baba were instrumental in maintaining the cell lines and cell culture. We would like to acknowledge the Netherlands Brain Bank for their generous contribution of the human brain tissue samples.