Dynamic phosphorylation of Histone Deacetylase 1 by Aurora kinases during mitosis regulates zebrafish embryos development

Histone deacetylases (HDACs) catalyze the removal of acetyl molecules from histone and non-histone substrates playing important roles in chromatin remodeling and control of gene expression. Class I HDAC1 is a critical regulator of cell cycle progression, cellular proliferation and differentiation during development; it is also regulated by many post-translational modifications (PTMs). Herein we characterize a new mitosis-specific phosphorylation of HDAC1 driven by Aurora kinases A and B. We show that this phosphorylation affects HDAC1 enzymatic activity and it is critical for the maintenance of a proper proliferative and developmental plan in a complex organism. Notably, we find that Aurora-dependent phosphorylation of HDAC1 regulates histone acetylation by modulating the expression of genes directly involved in the developing zebrafish central nervous system. Our data represent a step towards the comprehension of HDAC1 regulation by its PTM code, with important implications in unravelling its roles both in physiology and pathology.


Supplementary Figure 1. Aurora kinases A and B phosphorylate HDAC1 in mitosis on serine 406 in vitro
(a) HeLa cells were synchronized at the G1/S boundary or in mitosis as described in Supplementary Methods. Samples were collected every hour and analyzed by western blot with the indicated antibodies. RbAP48 is used as loading control. Cell synchronization was evaluated by FACS. Asterisks indicate the slow-migrating, modified forms of HDAC1.
(b) λ-phosphatase assay on asynchronous and mitotic HeLa cells. The asterisk indicates the hyper-phosphorylated forms of HDAC1. BT-15 recognizes HDAC1 mitotic phosphorylation (Segre' et al., 2016). Cdc25c is used as positive control for the assay, Cyclin B1 as mitotic marker, Vinculin as loading control.
(c) Alignment of human class I HDAC1 (GeneBank CAG_46518), HDAC2 (NCBI NP_001518) and HDAC3 (NCBI NP_003874) protein sequences. Putative target sites identifed by the Phosida bioinformatic tool for Aurora kinases (pink) and Plk1 (green) are indicated. Numbers of target residues are referred to the HDAC1 sequence.
(d) HeLa cells were synchronized in G1/S by thymidine block and released in fresh medium for 4 hours, synchronised in G2/M phase by nocodazole with or without the Aurora kinase inhibitors Hesperadin and ZM-447439 or the Plk1 inhibitor BI-2536; in the last 1 hour of treatment, we added proteasome inhibitor MG132. Mitotic cells were collected after 5.5 hours by mitotic shaking and analyzed by western blot with the indicated antibodies. Cdc25c is a known substrate of Plk1 in mitosis and here is used as positive control for BI2536. H3S10ph is a known substrate of Aurora kinases and here is used as positive control of ZM-447439 and Hesperadin effects. Cyclin B1 is used as marker of mitosis and Vinculin as loading control.
(e) HeLa cells were subjected to two cycles of RNA interference with the indicated siRNA, synchronized at G1/S boundary and released with nocodazole as schematized in the cartoon.
Cells were collected at 9.5 hours by mitotic shaking and analyzed by western blot with the indicated antibodies. BT-15 recognizes HDAC1 mitotic phosphorylation (Segre' et al., 2016).
Cdc25c phosphorylation is used as mitotic marker and Vinculin as loading control.
Densitometric analysis was performed using Image J and the average of two independent experiments was indicated ± SEM.
Samples were loaded on a 14% PAA gel and incubated for autoradiography. H3 is used as a positive control for Aurora kinase activity. Coomassie-stained gels are reported as loading controls. Asterisks mark the bands corresponding to recombinant proteins. A ladder of molecular weight is reported, where KDa are kilo Daltons.
(g) Recombinant Aurora A/TPX2 and Aurora B/INCENP were incubated with 1µg of HDAC1 wild type and the indicated serine-to-alanine mutants as substrates with 5µCi ATP( 32 P) for 1 hour at 30°C. Samples were loaded on a 14% gel and incubated for autoradiography.
Coomassie-stained gels are shown as loading controls.

Supplementary Figure 2. Overexpression of hHDAC1 in zebrafish embryos
(a) Western blot analysis of protein extracts from control embryos or Hdac1 morphant embryos. On the left samples were collected at 24, 48 or 72 hpf and blotted with zfHdac1; asterisk indicates the zfHDAC1 band. On the right samples were collected at 72 hpf and blotted with Flag. Anti-Histone3 (H3) antibody was used as a loading control.
(b) The BT-15 epitope recognizing the KRISI Aurora consensus motives of HDAC1 is reported. The non-conserved residues between human and zebrafish HDAC1 are shown. The amino acids are indicated according to the universal one letter code.

Supplementary Figure 3. Aurora kinases phosphorylate hHDAC1 in zebrafish embryos
(a) Immunofluorescence microscopy of 24 hpf zebrafish embryos. Embryos were injected at one-cell stage with scramble MO or hdac1 MO alone and in combination with hHDAC1 wt, S406A or S406E mutants, collected at 24 hpf, fixed in 4% paraformaldehyde and then immunostained with the specific antibody. Scale bar corresponds to 50 µm.
(b) Immunofluorescence microscopy of zebrafish embryos injected at one-cell stage with hdac1 MO and hHDAC1 wt, collected at gastrulae stage (80% epiboly) and stained with the indicated antibody. Scale bar corresponds to 100 µm. Left Panel: Immunohistochemistry staining of 72 hpf uninjected embryos (from Figure 3a) with haematoxylin (blue) and acetylated-histone H4 antibody (clone T25) (brown) (Ronzoni et al., 2005). Right Panel: The diencephalon area was identified with a pink line and used for the analysis. Aperio ImageScope software, used for quantification, identified blue pixel as negative, yellow pixels as weak positive, orange pixels as positive and red pixels as strong positive acetylated-histone H4. The graph expresses the distribution of total pixels as percentage.

Cell synchronization
Cells were synchronized at the G1/S boundary by double thymidine block: cells were treated with 2 mM thymidine overnight, followed by wash and release in fresh medium, for two consecutive nights. To synchronize cells in mitosis, the second day, they were released in medium containing 330 nM nocodazole for 9.5 hours and mitotic, rounded-up cells were harvested by shaking them off culture dishes. The G2-enriched population is the one that remains attached to the plate upon nocodazole treatment and mitotic shaking. G1 pure population was obtained from mitotic cells released for 2 hours from nocodazole block in order to re-enter synchronously in a new G1.

Production and purification of GST tag proteins
GST tag proteins were produced in bacteria and purified with Glutathion® beads (Sigma).
Proteins were eluted using reduced glutathione and the GST tag was removed through proteolytic cleavage with PreScission protease (produced as recombinant protein at our campus facilities).

Quantification of purified recombinant protein was performed by SDS-PAGE and gel staining
with Coomassie Brilliant Blue, by comparison with known amounts of BSA protein.

In vitro Aurora kinase assay
Aurora kinase assays were performed as previously described (Santaguida et al., 2010).
Briefly, 1 µg of purified substrate (HDAC1, HDAC3 or histone H3) was incubated with 50 ng of recombinant-purified Aurora A/TBX2 or Aurora B/INCEP kinase in the presence of 5 µCi of γ-32pATP at 30°C for 1 hour and analyzed by SDS-PAGE followed by autoradiography. For the enzymatic activity, the scintillation counts were normalized versus the amount of immunoprecipitated proteins.

Plasmids and vectors
For expression in human cells the pBJ5-HDAC1-Flag was used. For expression of recombinant proteins in bacteria the pGEX-6P1-GST-HDAC1 and pGEX-6P1-GST-HDAC3 plasmids were used. Plasmids for production of Aurora kinases have been previously described (Santaguida et al., 2010;Sessa et al., 2005). Point mutant HDAC1 proteins were generated by PCR site directed mutagenesis of pBJ5-HDAC1 wild type using primers carrying the desired mutations (see Supplementary Table 1).
For mRNA transcription and injection in zebrafish embryos, the human HDAC1 wild type, S406A and S406E constructs were subcloned into pCS2+ expression vector (Addgene) using BamHI/XhoI restriction enzymes.

Western blot analysis
For western blot analysis, cells were lysed in denaturing SDS lysis buffer [one volume of

Immunoprecipitations
Cells were lysed in non-denaturing E1A buffer [(50 mM Hepes pH 7.5, 250 mM NaCl, 0.1% NP-40) PMSF, leupeptin, aprotinin, sodium orthovanadate, sodium fluoride and NEM were freshly added immediately before use]. Immunoprecipitation (IP) was performed incubating protein extracts with the antibody of interest at ratio of 3-4 mg Ab/mg of crude extract for 16 hours at 4 °C on rotation. Then protein-A/sepharose-beads (slurry 50%) were added to the samples for 2 hours at 4 °C. Beads were then extensively washed with cold E1A buffer, loaded on SDS-PAGE and analyzed by western blot with the indicated antibodies. Total extracts (input) were loaded as a control.

Zebrafish embryo cell dissociation
Embryos were anesthetized with tricaine and decapitated. After transient storage in PBS on ice, heads were incubated in trypsin-EDTA (Lonza) for 30 min at 37°C. Afterward, heads were transferred to fresh PBS and mechanically disgregated. This treatment led to a single-cell suspension. Cells were harvested by centrifugation for 3 min at 3000 × g at 4°C, resuspended in PBS and used for further experiments.

FACS analysis
For FACS analysis, cell suspension obtained as above mentioned was fixed in ice-cold 70% ethanol and incubated over night on ice. DNA was stained with a solution containing 2.5 µg/ml propidium iodide (PI) (Sigma) and 250 µg/ml RNase A. Samples were acquired using a FACSCalibur flow cytometer (Becton Dickinson), and data were analyzed using CellQuest software (Becton Dickinson).

RNA extraction
Embryos were anesthetized with tricaine and decapitated. After transient storage in PBS on ice, heads were disaggregated in TRIzol® and RNA was extracted following the manufacturing instructions (TRI Reagent® Protocol/ Sigma-Aldrich).

Confocal immunofluorescence microscopy
For immunostaining, gastrulae or 24 hpf embryos were fixed in 4% para-formaldehyde (PFA) overnight and permeabilized with acetone at -20°C for 7 minutes. Embryos were then blocked in PBS containing 0.5% Triton-X, 1% DMSO, 1% BSA and 2% sheep serum (PBST) for 2 hours at 4°C and incubated with primary antibody overnight. The next day, embryos were rinsed in PBST and incubated with secondary antibody. Then, embryos were stained with DAPI before mounting for confocal microscopy.
The following secondary antibodies were used: Alexa-488 conjugated mouse IgG, Alexa-433 conjugated rabbit IgG.

Immunohistochemistry (IHC)
For histological sections, 72 hpf embryos were fixed in 4% PFA overnight at 4°C and embedded in 1.2% low-melting agarose: while the agarose solidify, embryos were properly oriented. Then, the embryos-containing agarose blocks were dehidated and included in paraffin before sectioned with microtome. 5µm transversal sections of the heads of the embryos were dehydrated and incubated with Na-citrate buffer (10mM dihydrated Na-citrate pH6, 0,05% Tween20) at 95 °C for 45 minutes, to unmask the antigen from the paraffin. Later they were blocked for 1 hour with blocking solution (PBS, 1% BSA, 2% FBS) and incubated with primary antibody overnight. The day after, sections were washed in PBS and incubated with secondary antibody. The antibody signal was revealed with DAB and haematoxylin staining was performed, then sections were dehydrated and mounting with Eukitt mounting medium.
The percentage of acetylated histones (acH4) strong positive pixels was calculated using Aperio ImageScope software and is reported for every sample as average of at least three