Tubulin tyrosine ligase like 12, a TTLL family member with SET- and TTL-like domains and roles in histone and tubulin modifications and mitosis

Abstract

hTTLL12 is a member of the tubulin tyrosine ligase (TTL) family that is highly conserved in phylogeny. It has both SET-like and TTL-like domains, suggesting that it could have histone methylation and tubulin tyrosine ligase activities. Altered expression of hTTLL12 in human cells leads to specific changes in H4K20 trimethylation, and tubulin detyrosination, hTTLL12 does not catalyse histone methylation or tubulin tyrosination in vitro, as might be expected from the lack of critical amino acids in its SET-like and TTLL-like domains. hTTLL12 misexpression increases mitotic duration and chromosome numbers. These results suggest that hTTLL12 has non-catalytic functions related to tubulin and histone modification, which could be linked to its effects on mitosis and chromosome number stability.

Figure 1. hTTLL12, domain organisation and similarity with other proteins.

A. Schematic representation. Red indicates the SET-like domain (amino acids 91–249), blue the TTL-like domain (amino acids 353–642), yellow or grey the predicted ATP and Mg²⁺ binding amino acid motifs that are respectively conserved (WICK^416–419, SKYI^450–453, DIRY^470–473, EVN^605–607), or non-conserved (SLDT^426–429 and RAMYAVD^578–584) in hTTLL12. B, C. Phylogenetic trees of the TTL domains of the human TTL family based on Alignment S1 (B) and the SET-like domain of hTTLL12 and human SET domains based on Alignment S2 (C). Bootstrap values are provided for significant nodes when they are >80%. Multiple sequence alignments are available as data.

Figure 2. hTTLL12 changes H4K20me3 levels and co-immunoprecipitates with HP1γ, H3 and H4K20me3.

A. hTTLL12 (hTTLL12_A-E) or Control (Control_A, _B, _D, _E) clones were harvested in Laemmli buffer and subjected to SDS-PAGE and WB. Representative WBs (upper part; hTTLL12_A and Control_E) are shown for hTTLL12 (30 µg protein/lane) or H4K20me3 and TBP (20 µg protein/lane). Similar WBs were probed with different antibodies, quantified by densitometry and normalized to TBP levels (lower part), to determine methylated H4K20me3, H4K20me1, H4K20me2, H3K9me2, H3K9me3, H3K4me2, H3K4me3 and H3K27me2 levels. Data represent average protein levels ± SEM in hTTLL12 lysates relative to average levels in Control lysates (n = 2 except for H4K20me3 (n = 5) and H3K9me3 (n = 3)). B. HEp-2 cells were transfected with 10 nM control siRNAs (siLuciferase, siCtrl, siGFP or siSilencer) or 10 nM hTTLL12-specific siRNA (sihTTLL12_2-5). 72 hours after transfection, cells were lysed in Laemmli buffer and analysed by SDS-PAGE and WB. Chemiluminescent signals were detected using a Versadoc image station (hTTLL12) or film (H4K20me3). Representative WBs (upper part) for hTTLL12 (40 µg of total protein/lane), H4K20me3 (5 µg total protein/lane) and respective TBP signals in siLuciferase, siSilencer and sihTTLL12_2 lysates. H4K20me3 levels were quantified using Quantity One software and normalized to TBP (lower left part). Data represent average H4K20me3 levels ± SEM in sihTTLL12 lysates relative to average levels in control siRNA lysates (siControl; n = 3). * Statistically significant difference to the levels in siControl (P<0.05, Student's t-test). To evaluate H3K9me3 levels (lower right part), HEp-2 cells were transfected with non-targeting control siRNA (siGFP or siCtrl) or hTTLL12-specific siRNA (20 nM sihTTLL12_3 or 10 nM sihTTLL12_4). 48 hours after transfection, cells were harvested in Laemmli buffer and subjected to SDS-PAGE and WB. H3K9me3 levels were quantified by densitometry and normalized to TBP. Data represent average protein levels ± SEM in sihTTLL12 lysates relative to average levels in non-targeting siRNA lysates (siControl; n = 2). C. WB showing co-immunoprecipitation of HP1γ, Histone H3 and H4K20me3 with M2-Flagged hTTLL12 from nuclear extract of stable clone hTTLL12_A. Similar results were obtained with two other clones (hTTLL12_B and hTTLL12_C, data not shown). Lane 1: 1% of the input; lane 2: IP of M2-Flagged hTTLL12 in the presence of 500 µg M2-peptide, lane 3: IP M2-Flagged hTTLL12 in the absence of M2-peptide. * Statistically significant (P<0.05, Student's t-test) difference to the levels in Control or siControl.

Figure 3. hTTLL12 alters detyrosinated tubulin levels and co-immunoprecipitates with α-tubulin.

A. hTTLL12 (hTTLL12_A-E) or Control (Control_A, _B, _D, _E) clones were harvested in Laemmli buffer and analysed by SDS-PAGE and WB. Representative WBs (left part) for detyrosinated tubulin (deY-tub) and TBP (40 µg total protein/lane; Control_D and hTTLL12_B shown). deY-tub levels from similar WBs were quantified by densitometry and normalised to TBP (right part). Data represent average detyrosinated tubulin levels ± SEM in hTTLL12 lysates relative to average levels in control lysates (n = 2). * Statistically significant difference to the controls (P<0.05, Student's t-test). B. HEp-2 cells were transfected with 10 nM control siRNAs (siLuciferase, siCtrl, siGFP or siSilencer), 10 nM hTTLL12-specific siRNA (sihTTLL12_1-6), or 10 nM positive control siRNA (siTTL_1-2). 72 hours after transfection, cells were lysed in Laemmli buffer and analysed by SDS-PAGE and WB. Chemiluminescent signals were detected using a Versadoc image station. Representative WBs (left part) for detyrosinated tubulin (40 µg of total protein/lane) and TBP signals (note that lanes 1–3 of the TBP blot are identical to the ones shown in the upper left part of Figure 2B) in siLuciferase, siSilencer, sihTTLL12_2 and siTTL_2 lysates. Detyrosinated tubulin levels were quantified using Quantity One software and normalized to TBP (right part). Data represent average detyrosinated tubulin levels ± SEM in sihTTLL12 and siTTL lysates relative to average levels in control siRNA lysates (siControl; n = 5). * Statistically significant difference to the levels in siControl (P<0.05, Student's t-test). C. Representative WB showing the co-immunoprecipitation of α-tubulin with Flag-hTTLL12 immunoprecipitated from the stable clone hTTLL12_A (lane 3). Similar results were obtained with 4 other clones (hTTLL12_B-E, data not shown). Parental HEp-2 cells (lane1) and a stable HEp2-clone overexpressing Flag-NudCD2 (lane 2) were used as negative controls. IgG HC: IgG heavy chain.

Figure 4. Cell cycle phase dependence of the levels of detyrosinated tubulin (deY-tub) and H4K20me3 in clones overexpressing hTTLL12 compared to control clones.

A. FACS scans of cells released from a double thymidine block at the indicated times. The times shown are those at which the cells were predominantly in S, G2/M and G1, except for the 18 hour time point. One representative synchronisation is shown. B. WBs, from one experiment, of whole cell lysates from cells synchronised in the G1, S and G2/M phases, respectively. The graphs are the averages from 3 independent experiments. * p<0.05. C. Proportion of cells in G1 at different times after release from double thymidine block. The average from 3 different experiments (up to 18 h) is shown. c2, c4 = DNA complements as measured by propidium iodide.

Figure 5. hTTLL12 up and down regulation prolong mitotic duration of HEp-2 cells.

A. Example frames from live cell movies of a HEp-2 cell stained with vital Hoechst. The time between nuclear envelope breakdown (NEB, first arrowhead, 0 min) and anaphase onset (second arrowhead) are indicated. Upper panel: fluorescent images (Hoechst), middle panel: phase contrast images, lower panel: overlay. B–C. Cumulative frequency (plot, right axis) and frequency distribution (histogram, left axis) of mitotic duration for hTTLL12 clones (B) and siRNA transfected HEp-2 (C). B. hTTLL12 (hTTLL12_A-E) or Control (Control_A-E) clones were seeded in 6-well plates. 48 hours post seeding, cells were treated with vital Hoechst and analysed (Experimental Procedures). Parental HEp-2 cells (light green), and the averages of the hTTLL12 (red) and Control (dark green) clones are plotted. C. HEp-2 cells were transfected separately with 12.5 nM sihTTLL12_1-6 or controls (siLuciferase, siCtrl, siScramble). 72 hours post transfection, cells were stained with vital Hoechst and analysed. HEp-2 cells; light green; averages of sihTTLL12s; red; averages of siControls; dark green. More detailed information and statistical analysis can be found in Figure S1.

Figure 6. hTTLL12 lacks HMTase activity.

The assays included histones purified from calf thymus (A, lanes 2, 3,5, 6, 8–11; B lanes 1, 3, 5–9, 11) as substrate and proteins to be tested for HMT activity that were purified from bacterial, viral and mammalian cell expression systems. The upper panels in each part (A & B) are Coomassie Blue stained SDS-polyacrylamide gels of the expressed proteins and the lower panels are the corresponding fluorograms centred on the core histones (Histones). The proteins tested for HMTase activity were expressed and purified from appropriate pGEX vector transformed E. coli [GST-NSD1 (1700–1987) (G-NSD1), GST-TTLL12 (50–250) (G-TTLL12)], recombinant vaccinia virus Ankara strain (MVA) infected mammalian cells [vTTLL12, vTTLL12 and vTTLL12 + E (a fraction that contains EEF1A1)] and HEp-2 cell clones transformed with pSG5-puro-Flag (CON-D) or pSG5-puro-Flag TTLL12 (L12-C & L-12-D). The HEp-2 proteins were purified from cytoplasmic (c-CON-D, c-L12-c, c-L12-D), nuclear (n-CON-D, n-L12-C, n-L12-D) and 1M KCl (1M-L12-C) fractions. 10 µl reactions were loaded on 15% SDS-PAGE gels. In A, the approximate amounts of protein used per reaction were: GST-NSD1 (1700–1987) (1 µg, lanes 1, 2; 0.5 µg, lane 3), GST-TTLL12 (50–250) (0.5 µg, lanes 4, 5; 0.25 µg, lane 6), TTLL12 (3 µg, lanes 7, 8; 1.5 µg lane 9), TTLL12 + EEF1A1 (2 µg, lanes 10, 12) and TTL (4 µg, lane 11). In B, they were GST-NSD1 (1700–1987) (2 µg, lanes 1, 12), TTLL12 (0.25 µg, lanes 4–6; 0.005 µg lanes 8–10, 0.001 µg lane 11). TTLL12 was not detected in the equivalent fractions purified from the control HEp-2 clone (CON-D). In B, lanes 2–11, the band migrating slightly faster than GST-NSD1 (1700–1987)) is IgH from the affinity column that is eluted under the harsh denaturing conditions used for sample preparation for SDS-PAGE.

Figure 7. hTTLL12 lacks tubulin tyrosine ligase activity, as shown with tubulin tyrosine ligase (TTL) assays using cell extracts (A, B) or purified proteins (C).

(A) TTL assays on extracts of transfected HEp-2 cells. Control reactions lacked cell extract (1), ATP (2) or purified tubulin (3). Test reactions contained extracts from cells transfected with: no DNA (4, mock), pc-AS 153 (5, antisense hTTLL12), pcDNA3-hTTLL12 (6), pSG5-puro-hFlag TTLL12 (7, Flag-hTTLL12), pc-AS TTL (8, antisense hTTL), and pcDNA3-TTL (9). (B) Expression of TTLL12 and TTL in transfected cell extracts. WBs were used to detect hTTLL12, TTL and TBP (loading control) in the extracts used for the activity tests. (C). TTL activity of purified proteins. The in-vitro reactions contained the indicated quantities (µg) of TTL, hTTLL12 and hTTLL12 complexed to EEF1A1 that had been purified from mammalian cells infected with corresponding recombinant Ankara strain (MVA) vaccinia viruses. The counts were normalised to 100 for 10 µg TTL. Error bars indicate standard deviations.