Page 49 - I. Chemical biology and drug discovery

A naturally-derived DYRK1A inhibitor rescues
Down syndrome-related phenotypes
Young-wook Ham 1,2,# , Miri Choi 1,2,# , Ae-kyeong Kim , Joo-Youn Lee , Jung-Nyoung Heo , Min-Hyo Ki , Sang-Bae Han , Kyu-Sun Lee , Sungchan Cho 1,7,*
6
2
3
3
4
5
1 Natural Medicine Research Center, KRIBB, Korea. 2 College of Pharmacy, Chungbuk National University, Korea. 3 Bionanotechnology Research Center, KRIBB, Korea. 4 Chemical Data-Driven
Research Center, KRICT, Korea. 5 Bioorganic Science Division, KRICT, Korea. 6 Center Research Institute, Samjin Pharm. CO.,LTD, Korea. 7 Department of Biomolecular Science, KRIBB School of
Bioscience, KUST, Korea.
ABSTRACT
Dual-specificity tyrosine phosphorylation-regulated kinase 1A (DYRK1A) is a significant pathogenic factor in Down syndrome (DS). Thus, inhibition of DYRK1A is
considered as a therapeutic strategy to modify the disease. Here, we identified Aristolactam BIII as a novel naturally-derived DYRK1A inhibitor through two-step
screening approach using structure-based virtual screening of >400,000 KCB chemical library (1 st step) and cell-based NFAT-RE promoter assay (2 nd step).
Aristolactam BIII potently inhibited the kinase activity of DYRK1A in vitro (IC 50 = 9.67 nM) and effectively suppressed DYRK1A-mediated hyperphosphorylation of Tau
and Presenilin 1 in mammalian cells, which was stronger than those of other DYRK1A inhibitors. Aristolactam BIII rescued neurological and phenotypic defects of Down
syndrome-like Drosophila model. Moreover, oral administration of Aristolactam BIII acutely suppressed Tau hyperphosphorylation in the brain of DYRK1A TG mice. In
the open field test, Aristolactam BIII (1 mg/kg) significantly ameliorated the anxiety-like behavior of DYRK1A TG mice. Together, our results obviously demonstrate that
Aristolactam BIII as a novel DYRK1A inhibitor rescues Down syndrome phenotypes in cells and in vivo and also suggest its therapeutic potential for the treatment of
DYRK1A-related diseases including DS and Alzheimer’s disease (AD).
RESULTS
Figure 2. Aristolactam BIII as a potent inhibitor of DYRK1A in vitro. (A)
Aristolactam BIII inhibited DYRK1A potently in in vitro kinase assays
performed by ThermoFisher Scientific using recombinant human DYRK1A,
DYRK1B, DYRK3, DYRK4, CDK1/cyclin B, CLK1, CK2α1 and GSK-3β
proteins. The mean values and standard deviation (s.d.; error bars) were
determined from two independent assays. The GraphPad Prism 5 program
was used to evaluate IC 50 . (B) The molecular docking simulation predicted
that Aristolactam BIII binds to the ATP-binding site of DYRK1A. (C) Binding
mode of Aristolactam BIII to the ATP-binding pocket of DYRK1A was
predicted. DYRK1A structure in blue ribbon and Aristolactam BIII in yellow
sticks.
Figure 5. Aristolactam BIII rescues phenotypic and neurological defects
in a mnb-overexpressing Drosophila model. (A) The wing area from adult
flies expressing mnb throughout the wing led to distal truncation of the L5
wing vein (red circle). Ten μM of Aristolactam BIII was treated on the L5 wing
vein defect of MS1096>2×mnb flies (n = 236, 286 for each). (B) The eyes of
adult flies overexpressing mnb and/or human Tau under the control of the
eye-specific GMR-Gal4 driver, and of control flies bearing only GMR-Gal4.
Overexpression of mnb aggravated the eye abnormality induced by Tau
overexpression. This eye defect was effectively rescued by feeding with 10
μM of Aristolactam BIII. White dashed lines outline the eye contour. (C) The
brain cortex, primary neuronal cell clusters and axon bundle, as visualized
with EGFP using UAS-Synaptobrevin-EGFP driven by Elav-Gal4 at the late
Figure 3. Aristolactam BIII potently inhibits potentiated Tau and PS1 embryonic stage 18. The pattern of neural connectivity and CNS structure
phosphorylation by DYRK1A in mammalian cells. 293T cells were were severely disorganized in mnb-overexpressing embryos (Elav-
transfected with the indicated plasmids expressing Tau (A), PS1 (C) and/or sytGFP+mnb, DMSO), the neurogenic defects of which were remarkably
DYRK1A. Then cells were treated with the indicated doses of Aristolactam rescued by 10 μM of Aristolactam BIII.
BIII for 6h. Total cell lysates were collected and subjected to western blotting
with anti-Tau, anti-phosphorylated-Tau (at T212), anti-PS1 and anti-DYRK1A
antibodies. The hnRNP A1 protein was analyzed as a loading control.
Western blotting was performed twice, and representative data are presented.
(B, D) The phosphorylated and total proteins of Tau and PS1 from western
blotting were quantified, respectively, and the amount of protein from DMSO-
treated samples was considered to be 100%. The mean values and standard
deviations were determined from two independent experiments.
Figure 1. Identification of Aristolactam BIII. (A) Strategy from virtual
screening to cell-based experimental testing to identify DYRK1A inhibitor hit.
(B) A simplified scheme of the effect of DYRK1A inhibitor determined by
NFATc1-mediated transcriptional activation, utilizing a firefly luciferase
reporter driven by the NFAT response element. ① The DYRK1A inhibitor
inhibits overexpressed DYRK1A which phosphorylates the dephosphorylated
NFATc and subsequently phosphorylated NFATc returns into the cytoplasm.
② The activated calcineurin by the calcium influx dephosphorylates from
phosphorylated NFATc in the cytoplasm that makes the dephosphorylated
NFATc proteins enter into the nucleus. ③ The increase of NFAT-dependent Figure 6. Oral gavage of Aristolactam BIII suppresses Tau
transcription is measured by a firefly luciferase reporter assay. (C) The high- phosphorylation in the brain of DYRK1A-overexpressing mice. Western
throughput screening to discover DYRK1A inhibitor was performed through blot showing the effect of Aristolactam BIII treatment on hippocampus (A)
cell-based NFAT-RE promoter assay. Human embryonic kidney 293T cells and frontal cortex (C) of DYRK1A-overexpressing mice. The 14- to 15-week-
were co-transfected with NRE-Luc reporter and DYRK1A and treated with IM old DYRK1A TG and littermate (n = 4 for each group) were treated with
(5 μM) and PMA (10 ng/mL) along with 586 candidates at 10 μM for 12 hours Aristolactam BIII (10 mg/kg or 30 mg/kg of body weight) in 50:50 (vol/vol) of
and assayed for firefly luciferase activity using One-Glo reagents. The DMSO:corn oil as vehicle orally, and the hippocampus and frontal cortex
luciferase activity in the DMSO-treated sample was set to 1, and relative were harvested after 30-40 min. The phosphorylation of Tau was analyzed by
luciferase activities were calculated. (D) Structure of Aristolatam BIII. (E) By Figure 4. Inhibitory efficacy of Aristolactam BIII on DYRK1A stronger western blotting with anti phosphorylated-Tau (at residue T231, T212 and
visualizing the translocation of the FLAG-NFATc1 protein, the effect of than other DYRK1A inhibitors and aristolactam analogues in S202/T205) and anti-Tau antibodies. Alpha-tubulin and hnRNP A1 were also
Aristolactam BIII on calcineurin-NFAT signaling regulated by DYRK1A was mammalian cells. (A) Transfected 293T cells were treated with the indicated analyzed as a loading control. (B, D) The phosphorylated and total Tau
examined. The 293T cells were transiently transfected with plasmids doses of Aristolactam BIII, Harmine and CX-4945 for 6 h. Total cell lysates proteins in panel A and C were quantified, respectively, and the average
expressing FLAG-NFATc1 or human DYRK1A for 24 h, and pre-treated with were subjected to western blotting with anti-phosphorylated-Tau (at T212) amount of each protein was calculated. Relative ratios of phosphorylated Tau
Aristloactam BIII (1, 10 μM) for 3 h and then stimulated with IM (5 μM) for 1 h. antibody to compare the inhibitory efficacy among those compounds, and to total Tau (p-Tau/Tau) were presented by setting the Normal-vehicle mice
DYRK1A, Tau and hnRNP A1 were detected using their corresponding as 1.
antibodies. The hnRNP A1 protein was analyzed as a loading control.
Western blotting was performed twice, and representative data are presented.
(B) Comparison of the effect of Aristolactam, Harmine and CX-4945 on
NFATc1-mediated transcriptional activation. Transfected 293T cells were
treated with IM (5 μM) and PMA (10 ng/mL) along with the indicated doses of
Aristolactam BIII, Harmine and CX-4945 for 12 h. Firefly luciferase activities
were measured using One-Glo reagents. Luciferase activity in the sample
with reporter plasmid alone was set to 1, and the relative luciferase activities
were calculated. Means±s.d. were determined from two independent
experiments. (C) Structures of Aristolactam BIII derivatives. (D) Aristolactam Figure 7. Aristolactam BIII improves anxiety-like behavior of DYRK1A-
BIII is the only compound that increases the transcriptional activity by overexpressing mice in the open field test. (A) Open field test of Normal
following NFATc nuclear translocation among eight structural derivatives. and DYRK1A-overexpressing mice orally administrated with vehicle or
Transfected 293T cells were treated with IM (5 μM) and PMA (10 ng/mL) Aristolactam BIII. Ratio of center distance traveled to total distance and ratio
along with the indicated doses of of Aristolactam BIII and eight derivatives for of cumulative duration in center to total cumulative duration are presented.
12 h. Activities of firefly luciferase were measured using One-Glo reagents. The vehicle-treated DYRK1A-overexpressing mice had a deficit relative to
Luciferase activity in the sample with reporter plasmid alone was set to 1, vehicle-treated Normal mice in abnormal general behavior, and the DYRK1A-
and the relative luciferase activities were determined. Means±s.d. were overexpressing mice receiving Aristolactam BIII displayed a significant
calculated from two independent experiments. improvement compared to the vehicle group.
CONCLUSION Contact information
We identified Aristolactam BIII as a novel naturally-derived DYRK1A inhibitor through two-step screening approach using Sungchan Cho, PhD
structure-based virtual screening and cell-based NFAT-RE promoter assay. Aristolactam BIII inhibits selectively DYRK1A in Natural Medicine Research Center, KRIBB, 30 Yeongudanji-ro,
vitro, and has a stronger inhibitory effect on Tau and PS1 phosphorylation and a modulatory effect on DS-related Ochang-eup, Cheongju-si, Chungbuk 28116, Republic of Korea
calcineurin/NFAT signaling than other DYRK1A inhibitors in mammalian cells. Aristolactam BIII rescues neurological and Tel: 043-240-6105, Fax: 043-240-6159
phenotypic defects in a DS-like Drosophila model. Oral administration of Aristolactam BIII rescues aberrant Tau E-mail: sungchan@kribb.re.kr
phosphorylation and behavioral phenotype in DYRK1A BAC transgenic mice.