- Analysis of phosphopeptides in IκBα by AurkC using MALDI-TOF MS
In a previous study, the interaction between IκBα and AurkC was identified using the CUPID system.
The result was demonstrated through typical experiments, including co-immunoprecipitation and use
of a mammalian two-hybrid system. The interaction between IκBα and AurkC, which are serine and
threonine kinases, leads to IκBα phosphorylation. Interestingly, although IκBα is known to be
phosphorylated at S32 and S36, AurkC phosphorylates IκBα only at S32, but the phospho-band was
thick and shifted to identify a novel phosphorylation site 17. Thus, we screened the IκBα
phosphorylation site to identify the carcinogenic mechanism of IκBα-AURKC binding and designed a
new IκBα phosphorylation site with AurkC using MALDI-TOF MS and nanoLC-quadrupole orbitrap
MS/MS (Supporting Figure 1). Phospho-IκBα was enriched to analyze the phosphopeptides using MS.
Trifluoroacetic acid and the alpha-cyano-4-hydroxycinnamic acid matrix commonly used for peptide
analysis were utilized instead of dissolving the 2,5-dihydroxybenzoic acid matrix in a phosphoric acid
(PA) solution. To identify the specific phosphorylation site in IκBα using AurkC, IκBα
phosphorylated by IKKβ and unphosphorylated IκBα were used as negative controls. We detected 11
of the 22 tryptic peptides in phosphorylated IκBα using MALDI-TOF MS and also found 3
phosphorylated peptides. The tryptic peptides containing missed cleavages and modified peptides,
including oxidized (+16 Da) and carbamidomethylated (+57 Da) IκBα, are shown with their
theoretical molecular weights. The phosphopeptide increased about 80 Da because HPO3- binds to
serine and threonine. The increase of 80 Da in the T6 (amino acids 30–38) peptide including S32 was
observed at m/z 1069.40 (P1). We expected that the original T6 peptide (989.44 Da) would be
phosphorylated at S32. However, the original T6 peptide peak was not found because it underwent
various modifications and missed cleavages (Supplementary Figure 2 and Table 1).
The spectra of IκBα phosphorylated by AurkC were compared with those of unphosphorylated IκBα
and IκBα phosphorylated by IKKβ. We found a novel peak only in IκBα phosphorylated by AurkC
(Figure 1). m/z 1731.81 (P2) and m/z 2340.15 (P3) were expected to increase about 80 Da each from
the T9-10 peptide (amino acids 54–67) and T20 peptide (amino acids 246–264) (Supplementary
Figure 2), indicating that IκBα has a novel AurkC phosphorylation site (Supplementary Table 2).
In this study, the phosphopeptide analysis was performed in positive ion detection mode without
enrichment.
- Profiling of phosphopeptides in IκBα by AurkC using nanoLC-quadrupole orbitrap MS/MS
LC-quadrupole orbitrap MS/MS (LC MS/MS) analysis (Q-Exactive Plus) was performed to confirm
the MALDI-TOF results. The peptide mixture was generated by an in-gel digestion process and then
dissolved in 2% PA. LC MS/MS was performed to separate the tryptic peptides and identify the
separated peptides. The Mascot data search program was operated under tryptic mis-cleavage 2,
methionine oxidation, and serine/threonine phosphorylation conditions.
The LC MS/MS data were matched to 83% of the sequence of IκBα, except that the largest molecular
weight peptide (265–314; 5848.57 Da) and short LTL peptide (315–317; 346.23 Da) from the C-
terminal region were produced by trypsin. Moreover, 63 phosphopeptides obtained from the Mascot
search results had the expected cut-off value (10-5). Phosphorylated S32-, S63-, and S262-containing
peptides of various lengths were confirmed. In conclusion, we observed S32 (Supplementary Figure 3),
S63, and S262 phosphorylation sites in IκBα used by AurkC (Figure 2). Phospho-S63 and phospho-
S262 have not been reported in IκBα until now.
- AURK phosphorylates the new phosphorylation site of IκBα.
In vitro kinase assays were used to confirm phosphorylation of the newly discovered S63 and S262
sites. No anti-S63 or -S262 antibody exists, so an antibody was produced. First, S32 of representative
IκBα was phosphorylated to confirm that IκBα was phosphorylated by AURK (Figure 3A). AurkC
was reacted with inactivated IκBα. As a result, we confirmed that AurkC phosphorylated IκBα at S32.
Based on these results, IκBα S63 and S262, which were newly detected by AURK, were
phosphorylated (Figure 3B). All three AURK families phosphorylated IκBα at S63 and S262, as well
as at S32. The highest affinity was for AurkA and the lowest was for AurkC. Therefore, AURK
confirmed that the IκBα phosphorylation site was phosphorylated.
- Ser-63 and 262 mutations of IκBα inhibit cell proliferation in breast cancer cells
Novel S63 and S262 mutants were constructed by single point mutations; serine was substituted with
alanine to confirm the intracellular mechanism (Figure 4A). p65 transcriptional activity was confirmed
to examine the effect of S63 and S262 phosphorylation on NF-κB activity in MDA-MB 231 cells
(Figure 4B). p65 transcriptional activity in the S63A and S262A mutants was decreased compared
with that in the control, and p65 activity in the S63A mutant was further reduced compared with that
in the S262A mutant. Inhibition of NF-κB activity was closely related to survival of cancer cells,
confirming the viability and proliferation of the breast cancer cells. MDA-MB 231 and MCF-7 cells
showed decreased tumor cell viability in the S63A and S262A mutants (Figure 4C). Both cell lines
decreased significantly, and cell viability was decreased slightly in the S63A mutant compared to the
S262A mutant. Tumor cell proliferation was also inhibited by both mutations (S63A and S262A;
Figure 4D). However, synergistic effects in the S63A and S262A double mutants were not confirmed
in any of the three experiments. Nevertheless, the phosphorylation of IκBα at S63 and S262 not only
replaced IκBα S32, but also regulated downstream NF-κB activity (Supplementary Figure 4A).
- S63A and S262A of IκBα induced necroptosis in MDA-MB 231 cells
The degree of apoptosis in the mutants was confirmed, as the single point mutation in IκBα affected
the viability and proliferation of the breast cancer cell lines. Apoptosis was measured in each mutant
using the MDA-MB 231 cell line (Figure 5A). Cell death increased 3–4 fold in the S63A or S262A
mutant compared to the control. However, apoptosis increased in the S63A and S262A groups
compared to the control group, but no significant difference was observed between the mutant groups
and the double mutant group. Interestingly, necrosis changed more than apoptosis in the cell
population. In other words, cells that were dead by apoptosis also underwent propidium iodide staining,
but the cell population was confirmed to proceed to Q3-Q1 as well as Q3-Q4-Q2. Therefore,
programmed necrosis or necroptosis in the population increased for Q1. Thus, the expression and
phosphorylation of RIP3 and MLKL, which are related to necroptosis, were confirmed (Figure 5B).
The phosphorylation of RIP3 and MLKL increased 2–2.5 times compared to the control. The mutant
group also showed increased expression of cleaved caspase-3 and -8 compared to the control group.
Based on these results, S63 and S262 of IκBα also have important effects, including apoptosis and
necroptosis, upon IκBα phosphorylation.