Abstract

   It has been previously shown that (never in mitosis gene A)‐related
   kinase 2 (NEK2) is upregulated in multiple myeloma (MM) and contributes
   to drug resistance. However, the mechanisms behind this upregulation
   remain poorly understood. In this study, it is found that amplification
   of NEK2 and hypermethylation of distal CpG islands in its promoter
   correlate strongly with increased NEK2 expression. Patients with NEK2
   amplification have a poor rate of survival and often exhibit TP53
   deletion, which is an independent prognostic factor in MM. This
   combination of TP53 knockout and NEK2 overexpression induces asymmetric
   mitosis, proliferation, drug resistance, and tumorigenic behaviors in
   MM in vitro and in vivo. In contrast, delivery of wild type p53 and
   suppression of NEK2 in TP53^−/− MM cell lines inhibit tumor formation
   and enhance the effect of Bortezomib against MM. It is also discovered
   that inactivating p53 elevates NEK2 expression genetically by inducing
   NEK2 amplification, transcriptionally by increased activity of cell
   cycle‐related genes like E2F8 and epigenetically by upregulating DNA
   methyltransferases. Dual defects of TP53 and NEK2 may define patients
   with the poorest outcomes in MM with p53 inactivation, and NEK2 may
   serve as a novel therapeutic target in aggressive MM with p53
   abnormalities.

   Keywords: amplification, multiple myeloma, NEK2, TP53, transcriptional
   regulation
     __________________________________________________________________

   Patients with NEK2 amplification are more prevalent in the TP53‐deleted
   subgroup of multiple myeloma (MM) and have a poor rate of survival.
   Dysfunctional p53 elevates NEK2 expression genetically by inducing NEK2
   amplification, transcriptionally by mediating E2F8 activity, and
   epigenetically by increasing the expression of DNMTs. NEK2 may serve as
   a novel therapeutic target in aggressive MM with p53 abnormalities.

   graphic file with name ADVS-9-2104491-g004.jpg

1. Introduction

   Multiple myeloma(MM) is a cancer of terminally differentiated plasma
   cells and is the second most common hematological malignancy.^[ [64]^1
   ^] The pathogenesis of MM involves several genetic alterations.^[
   [65]^2 ^] These changes include primary cytogenetic abnormalities
   (e.g., translocations involving chromosome 14q and trisomies of
   odd‐numbered chromosomes) and secondary lesions (e.g., gain of
   chromosome 1q and loss of chromosome 17p).^[ [66]^3 ^]

   Cancer‐related genes are broadly grouped into oncogenes and tumor
   suppressor genes. The abnormal activation of oncogenes such as CCND1,
   CCND2, CCND3, and FGFR3 has been reported in MM.^[ [67]^2 , [68]^4 ^]
   DNA gains and losses that result in copy number alterations cause
   oncogene activation and tumor suppressor gene inactivation; these are
   the driving events leading to the development and progression of MM.^[
   [69]^1 , [70]^2 , [71]^5 ^] For example, the amplification of
   chromosome 1q, which harbors a number of potentially relevant oncogenes
   such as CKS1B,^[ [72]^6 ^] ILF2,^[ [73]^7 ^] ANP32E,^[ [74]^8 ^] and
   PDZK1,^[ [75]^9 ^] contributes to MM cell proliferation. Deletion of
   17p, where TP53 is located, also plays an important role in MM
   progression.^[ [76]^10 ^] However, how these oncogenes collaborate with
   tumor suppressor genes to accelerate MM initiation and progression is
   still poorly understood.

   We have previously shown that increased chromosomal instability (CIN)
   signature is linked to drug resistance (DR) in MM.^[ [77]^11 ^] (Never
   in mitosis gene A)‐related kinase 2 (NEK2), a CIN gene located at
   1q32.2, is the most significant. NEK2 is a serine/threonine kinase^[
   [78]^13 ^] that induces tumor cell proliferation, metastasis and drug
   resistance through regulation of several oncogenes or cell
   cycle‐related molecules including AKT, β‐catenin and enhancer of zeste
   homolog 2 (EZH2)^[ [79]^15 ^] in MM and other types of cancer.^[
   [80]^11 , [81]^16 ^] It is associated with poor outcomes and drug
   resistance due to activating efflux drug pumps,^[ [82]^11 ^]
   autophagy,^[ [83]^14 ^] and ALDH1A1 in MM.^[ [84]^12 ^] Although
   several critical substrates and molecules downstream of NEK2 are
   involved in tumorigenesis, the mechanisms by which NEK2 is activated
   and cooperates with other molecules—especially key tumor suppressor
   genes like TP53—in MM cells are largely unknown.

   Since 1q amplification and 17p deletion are both markers of the poorest
   prognosis in MM patients, we hypothesize that activated NEK2 and
   inactive p53 may have synergistic effects in MM progression. In this
   study, we determine the correlation of genetic aberrations and the
   functional relationship between TP53 and NEK2 in MM in vitro and in
   vivo.

2. Results

2.1. NEK2 Amplification and Promoter Hypermethylation Correlate with its
Upregulation and are Associated with Poor Prognosis in MM

   Genetic lesions (including translocation, amplification and mutations)
   and epigenetic changes (e.g., DNA methylation, microRNA regulation, and
   transcriptional regulation) cause aberrant gene expression.^[ [85]^17
   ^] To investigate the mechanisms underlying aberrant expression of NEK2
   in MM, we first analyzed DNA copy number variations (CNVs) and
   expression of NEK2 by exome‐sequencing and RNA‐sequencing of 575
   primary MM samples from the Multiple Myeloma Research Foundation (MMRF)
   CoMMpass database. We found that 24.2% (139/575) of the patients showed
   NEK2 amplification (Amp) (Figure [86] 1A), which correlated strongly
   with elevated NEK2 expression (Figure [87]1B). Moreover, a Kaplan‐Meier
   survival analysis of the CNVs database containing 573 MM patients
   revealed that patients with NEK2 amplification had a significantly
   shorter overall survival (OS) (NEK2^Amp group, median, 40 months) than
   the patients with a normal copy number (NEK2^N group, median, 60
   months, Figure [88]1C). Subsequent fluorescence in‐situ hybridization
   (FISH) experiments using a probe for NEK2 and chromosome 1 control
   (CEP1) further confirmed these findings (Figure [89]1D). Compared with
   healthy donors (HD), NEK2 amplification (defined as ≥20% of CD138^+
   cells with three or more FISH signals) was detected in 23.5% (5/17),
   75% (3/4), and 87.5% (7/8) of newly diagnosed MM patients (AD),
   relapsed diagnosed MM patients (RD) and MM cell lines, respectively
   (Figure [90]1D and Table [91]S1, Supporting Information), suggesting a
   link between NEK2 amplification and poor prognosis in MM.

Figure 1.

   Figure 1
   [92]Open in a new tab

   NEK2 amplification in MM patients is associated with poor outcomes. A)
   Assessment of NEK2 CNV distribution in MM patients (n = 575). B) The
   correlation between CNV and NEK2 mRNA expression, and data presented as
   mean ± SD, n = 573, p‐values are calculated using unpaired t test,
   ****p < 0.0001. C) Kaplan‐Meier analysis of overall survival in MM
   patients with or without NEK2 CNV amplification (n = 574). D) Left:
   Representative images of FISH analysis using probes specific for CEP1
   (Green) and NEK2 (Red) in eight MM cell lines, healthy donors (HD, n =
   4), and newly‐diagnosed (AD, n = 17) and relapsed (RD, n = 4) MM
   patients. Upper: normal diploid. Middle: three copies. Lower: four
   copies (NEK2 amplification). Right: scatter plot showing statistical
   results of the proportion of cells with different NEK2 copy numbers in
   MM patients and MM cell lines. Data presented as mean ± SD, p‐values
   are calculated using unpaired t test, *p < 0.05, **p < 0.01,
   ***p < 0.001. The MSP results of NEK2 proximal and distal CpG islands
   in E) seven MM cell lines and F) three MM patients. M, MSP products
   using methylation‐specific primers; U, MSP products using
   methylation‐unspecific primers.

   Next, to determine the status of NEK2 mutations in MM cell lines, we
   analyzed all NEK2 exons and its promoter region from −1018 to +1 bp in
   eight MM cell lines. One point mutation (A→G, N354S) of seven exons was
   found only in the U266 cell line (Figure [93]S1A,B, Table [94]S2,
   Supporting Information), suggesting that the mutation may not be
   responsible for upregulation of NEK2 mRNA. Furthermore, two CpG islands
   in the regions from −700 to −500 bp (distal) and −200 to +1 bp
   (proximal) were observed in the NEK2 promoter region (Figure [95]S1B,
   Supporting Information). Methylation‐specific PCR (MSP) revealed
   unmethylation of the proximal CpG island and partial methylation of the
   distal CpG island in seven MM cell lines and three MM patients
   (Figure [96]1E,F). The sequencing of MSP products further showed that
   all the cytosine residues were converted to thymine except for those in
   methylated CpG dinucleotides, indicating the presence of methylated
   cytosines in these CpG dinucleotides (Figure [97]S1C, Supporting
   Information). These results suggest that amplification and distal
   methylation of NEK2 might be associated with its aberrant
   overexpression in MM.

2.2. NEK2 Amplification Correlates with Deletion/Mutation of TP53 in MM

   It was previously shown that loss of TP53 is an independent prognostic
   factor in MM^[ [98]^10 ^c] and that p53 can bind to the promoter of
   NEK2.^[ [99]^18 ^] These findings led us to assess the potential
   association between TP53 and NEK2 by using the CNVs database,
   containing 548 MM patients (excluding 25 MM patients with TP53
   amplification and two patients with NEK2 deletion), from the MMRF
   CoMMpass database. First, we divided the samples into quartiles based
   on their expression levels of TP53 (high = TP53^H , low = TP53^L ) and
   NEK2 (high = NEK2^H , low = NEK2^L ). TP53 deletion (TP53^Del ) was
   observed in 9.3% (51/548) of MM samples and contributed to aberrantly
   high expression of NEK2 (Figure [100] 2A). Among the TP53 ^Del MM
   patients, the frequency of NEK2 amplification (NEK2^Amp ) was ~25.5%
   (13/51) and was also linked to high expression of NEK2 (Figure
   [101]S2A, Supporting Information). In addition, the proportion of
   patients with NEK2 amplification increased in patients with TP53
   deletion (Figure [102]S2B, Supporting Information). The analysis of the
   CNVs database further showed that patients with concomitant TP53
   deletion and NEK2 amplification (TP53^Del&NEK2^Amp ) had a
   significantly shorter OS (median, 32.4 months) than those in the
   TP53^Del (49 months), NEK2^Amp (44.8 months) and no TP53^Del and
   NEK2^Amp (TP53^N &NEK2^N ) groups (not reached) (Figure [103]2B).
   Similarly, patients with concomitant TP53^Del and NEK2^H had a
   significantly shorter OS (median, 25.4 months) than others (TP53^N &
   NEK2^L and TP53^Del & NEK2^L , not reached; TP53^N & NEK2^H , 55.7
   months; Figure [104]S2C, Supporting Information). We also found that
   mutation of TP53 (TP53 ^Mut) was observed in 5.4% (40/743) of MM
   samples and was associated with aberrantly high expression of NEK2
   (Figure [105]2C) using the MMRF CoMMpass database of 743 cases. The
   results further showed that patients with both TP53 ^Mut and NEK2^H had
   a significantly shorter OS (median, 25 months) than those without the
   two defects simultaneously (TP53^N & NEK2^L and TP53^Mut & NEK2^L , not
   reached; TP53^N & NEK2^H , 54.8 months; Figure [106]2D, Supporting
   Information 1).

Figure 2.

   Figure 2
   [107]Open in a new tab

   NEK2 expression is elevated in MM patients with TP53 lesions. A) NEK2
   mRNA levels in MM patients with (n = 51) or without (n = 497) TP53
   deletion, and data presented as mean ± SD, p‐values are calculated
   using unpaired t test, ****p < 0.0001. B) Kaplan‐Meier analyses of
   overall survival in MM patients with TP53^N & NEK2^N (normal), TP53^N &
   NEK2^Amp (Amplification), TP53^D (deletion) &NEK2^N , and TP53^D &
   NEK2^Amp (n = 548). C) NEK2 mRNA levels in MM patients with (n = 40) or
   without (n = 703) TP53 mutation, and data presented as mean ± SD,
   p‐values are calculated using unpaired t test, ****p < 0.0001. D)
   Kaplan‐Meier analyses of overall survival in MM patients with wild type
   TP53 (TP53 ^N) & low expression NEK2 (NEK2^L ), TP53^N & high
   expression NEK2 (NEK2^H ), mutant TP53 (TP53^mut ) & NEK2^L and
   TP53^mut & NEK2^H (n = 743). E) Correlation between NEK2 and TP53 mRNA
   expression in the TT2 and TT3 trial form GEP data ([108]GSE2658, n =
   559, p‐values are calculated Pearson correlation coefficient ). F)
   Kaplan‐Meier analyses of overall survival in TT2 and TT3 MM patients (n
   = 559) with high NEK2 (NEK2^H ) & low TP53 expression (TP53^L ), NEK2^H
   & high TP53 expression (TP53^H ), low NEK2 (NEK2^L ) & TP53^L and
   NEK2^L and TP53^H . G) Scatter plot showing the percentage of cells
   with amplified NEK2 copy number in MM patients with or without TP53
   deletion. Data presented as mean ± SD, p‐values are calculated using
   unpaired t test, **p < 0.01, ***p < 0.001. H) Scatter plot showing the
   percentage of cells with amplified NEK2 copy number in TP53 ^WT or
   TP53^−/− MM cells. Data presented as mean ± SD, p‐values are calculated
   using unpaired t test, *p < 0.05, **p < 0.01. I,J) Representative
   immunofluorescence images and statistical analysis for p53 (Green) and
   NEK2 (Red) protein expression in MM newly diagnosed patients (AD, n =
   51) and relapsed patients (RD, n = 16, p‐values are calculated using
   one‐way ANOVA with Bonferroni correction).

   To further examine the relationship between NEK2 and TP53 in other
   cancer types, we analyzed the copy number and mRNA levels of NEK2 in 24
   different cancer types with TP53 abnormalities based on The Cancer
   Genome Atlas (TCGA) database. The cancer types included adrenocortical
   carcinoma, bladder urothelial cancer, breast invasive cancer,
   esophageal carcinoma, glioblastoma multiforme, kidney renal papillary
   cell carcinoma, liver hepatocellular carcinoma (LIHC), lung
   adenocarcinoma (LUAD), pancreatic adenocarcinoma, sarcoma, skin
   cutaneous melanoma, and stomach adenocarcinoma. We found that NEK2
   expression was significantly upregulated in TP53^Del or TP53^Mut cancer
   samples compared to samples with normal TP53 genetic status (Figure
   [109]S2D, Supporting Information). The copy number of NEK2 was also
   amplified and NEK2 mRNA levels were increased in TP53 ^Del samples from
   these cancer types, except for LIHC and LUAD (Figure [110]S2E,
   Supporting Information). These results implied that TP53 deletion or
   mutation (TP53 ^Del/Mut) correlates with NEK2 amplification and
   elevates NEK2 expression in MM and other cancer types.

   To assess the effect of NEK2 and TP53 expression and their roles in
   prognosis, we applied the Gene Expression Programming (GEP) database of
   559 MM samples ([111]GSE2658^[ [112]^5 ^]). As shown in Figure [113]2E,
   TP53 expression (low in MM) correlated inversely with NEK2 expression
   in MM patients. Additionally, the MM patients with TP53^L and
   concomitant NEK2^H expression had a significantly inferior OS (median,
   40.8 months), while patients with either NEK2 ^L or TP53^H had better
   outcomes (TP53^H & NEK2^L and TP53^L & NEK2^L , not reached; TP53^H &
   NEK2^H , 49 months; Figure [114]2F).

   To further confirm whether NEK2 was amplified in TP53^Del patients, we
   performed FISH using probes targeting the TP53 and NEK2 gene loci in MM
   clinical samples. The results showed ≥20% of CD138^+ cells with three
   or more signals marked as NEK2 loci in 100% (4/4) of TP53^Del MM
   samples (P31‐34, Table [115]S1, Supporting Information), but not in
   seven TP53 ^WT samples (P25‐30, Table [116]S1, Supporting Information)
   (Figure [117]2G). We also found that the number of cells containing
   three copies of NEK2 was significantly greater in TP53^Del MM cell
   lines (including ARP1 and KMS11) than in TP53 ^WT MM cell lines
   (including MM.1S and H929) (Figure [118]2H).

   Finally, to verify whether the loss of p53 is linked to elevated NEK2
   expression, we performed immunofluorescence (IF) to examine p53 and
   NEK2 protein levels in 51 newly diagnosed and 16 relapsed MM patients.
   As shown in Figure [119]2I,J, 9.8% (5/51) of newly diagnosed patients
   and 31.3% (5/16) of relapsed patients exhibited low p53 protein
   expression and high NEK2 protein expression, while in 66.7% (34/51) of
   new patients and 43.8% (7/16) of relapsed patients, the reverse was
   true. Experiments with Fisher's exact probability test implied that p53
   protein expression correlated inversely with NEK2 protein expression in
   MM. These results suggested that patients of the TP53^Del/Mut MM
   subgroup are prone to NEK2 amplification and upregulation, leading to
   poor prognosis.

2.3. TP53 Deletion/Mutation Upregulates NEK2 in Myeloma Cells

   To establish a causal relationship between TP53 deletion/mutation and
   NEK2 expression in MM, we induced p53 expression in H929 and ARP1 cell
   lines using different concentrations of nutlin‐3a. A dose‐dependent
   decrease of NEK2 mRNA and protein levels and a dose‐dependent increase
   of TP53 and MDM2 mRNA and protein levels were observed in H929 cells
   (Figure [120] 3A,B) but not in TP53^−/− ARP1 cells (Figure [121]3C,D).

Figure 3.

   Figure 3
   [122]Open in a new tab

   NEK2 expression is reduced and correlates inversely with loss of wild
   type TP53in MM cells. A) Relative mRNA levels of NEK2 and TP53 in the
   H929 TP53^WT MM cell line were detected with qPCR after treatment with
   nutlin‐3a for 48 h at concentrations of 0, 2, and 4 µM. Data presented
   as mean ± SD, p‐values are calculated using unpaired t test, n = 3,
   **p < 0.01. B) Relative protein levels of p53, NEK2, and GAPDH in
   TP53^WT H929 cells after treatment with 0, 2, and 4 µM of nutlin‐3a for
   48 h, as determined with immunoblotting. C) Relative mRNA levels of
   NEK2 and TP53 in the ARP1 TP53^−/− MM cell line were detected with qPCR
   after treatment with nutlin‐3a for 48 h at concentrations of 0, 2, and
   4 µM. D) Relative protein levels of p53, NEK2, and GAPDH in TP53^−/−
   ARP1 cells after treatment with 0, 2, and 4 µM of nutlin‐3a for 48 h,
   as determined with immunoblotting. E) The mRNA levels of TP53 and NEK2
   in H929 cells with or without TP53 deletion edited by CRISPR/Cas9, as
   determined with qPCR. Data presented as mean ± SD, p‐values are
   calculated using unpaired t test, n = 3, ***p < 0.001. F) p53 and NEK2
   protein levels in H929 cells with or without CRISPR‐Cas9‐mediated TP53
   deletion, as determined with immunoblotting. G) TP53 and NEK2 mRNA
   levels in MM.1s cells with or without CRISPR‐Cas9‐mediated TP53
   deletion, as determined with qPCR. Data presented as mean ± SD,
   p‐values are calculated using unpaired t test, n = 3, *p < 0.05,
   ***p < 0.001. H) The protein levels of p53 and NEK2 in MM.1s cells with
   or without CRISPR‐Cas9‐mediated TP53 deletion, as determined with
   immunoblotting. I) Relative mRNA levels of TP53 and NEK2 in ARP1 cells
   were detected with qPCR 72 h after infection with virus containing WT
   or mutant TP53 (E285K)‐expressing vectors. Data presented as mean ± SD,
   p‐values are calculated using unpaired t test, n = 3, *p < 0.05. J)
   Relative protein levels of p53, NEK2, and GAPDH in ARP1 cells were
   detected with immunoblotting 72 h after infection with virus containing
   WT or mutant TP53 (E285K)‐expressing vectors. K) The mRNA levels of
   TP53 and NEK2 in H929 cells with or without TP53 mutation, as
   determined with qPCR. L) The protein levels of p53 and NEK2 in H929
   cells with or without TP53 mutation, as determined with immunoblotting.
   M) The mRNA levels of TP53 and NEK2 in TP53‐deleted H929 (H929‐TP53^KO)
   cells with or without TP53 mutation, as determined with qPCR. N) The
   protein levels of p53 and NEK2 in TP53‐deleted H929 (H929‐TP53^KO)
   cells with or without TP53 mutation, as determined with immunoblotting.

   We also assessed the effect of knocking out (KO) TP53 in HEK293 cells
   using the CRISPR‐Cas9 gene editing technology. Cells infected with
   virus containing CRISPR‐Cas9 or CRISPR‐Cas9‐TP53 vector were
   successfully established and were dubbed HEK293‐Ctrl and
   HEK293‐TP53^KO, respectively (Figure [123]S3A–C, Supporting
   Information). TP53 deletion in HEK293 cells (HEK293‐TP53^KO) caused
   NEK2 mRNA and protein levels to increase dramatically when compared
   with the control cells (Figure [124]S3B,C, Supporting Information).
   Subsequently, p53 was successfully knocked out in H929 cells and
   knocked down in MM.1s cells established using the CRISPR‐Cas9 system.
   These p53‐deficient cell lines were named H929‐TP53^KO and
   MM.1s‐TP53^KO, respectively. We observed increased NEK2 mRNA and
   protein levels in both H929‐TP53^KO cells (Figure [125]3E,F) and
   MM.1s‐TP53^KO cells (Figure [126]3G,H).

   We then obtained cell clones from H929‐TP53^KO cells via sequential
   dilution and found that the NEK2 mRNA and protein levels were
   upregulated in clones with TP53 deletion (C1, C2, and C4) and more
   strongly in those with TP53 mutation (C3, C5) (Figure [127]S3D,E,
   Supporting Information). The sequence forms of the DNA and protein from
   these clones were confirmed by sequencing (Figure [128]S3F and
   Additional File [129]S6, Supporting Information). Similar results were
   also seen in HCT116‐TP53^−/− cells (Figure [130]S3G,H, Supporting
   Information) and the Trp53^tm1Tyj TP53‐knockout mouse genetic model
   (Figure [131]S3I,J, Supporting Information).

   In contrast, ectopic expression of WT TP53 in ARP1 cells caused NEK2
   mRNA and protein levels to markedly decrease compared with empty vector
   (ARP1‐EV), while ectopic expression of mutant TP53 with a hotspot
   mutation at E285K exerted no inhibitory effect (Figure [132]3I,J).
   Moreover, ectopic expression of mutant TP53 with hotspot mutations
   observed in MM and other tumors (Y126H, R175H, R282G, and E285K) in
   H929‐TP53^WT and H929‐TP53^KO cells caused slight elevation of NEK2
   mRNA levels, while protein levels were increased dramatically
   (Figure [133]3K–N). These results indicate that TP53 deletion or
   mutation induces NEK2 upregulation.

2.4. NEK2 Activation and p53 Suppression Promote Mitotic Aberrations, Cell
Proliferation, and Tumorigenesis in MM

   To test the effect of dual defects of NEK2 and TP53 on MM, we
   overexpressed NEK2 in H929 cells (H929‐NEK2 OE) and H929‐TP53^KO cells
   (H929‐TP53^KO/NEK2 OE), then performed functional assays in H929‐Ctrl,
   H929‐NEK2 OE, H929‐TP53^KO, and H929‐TP53^KO/NEK2 OE cells after
   determination of TP53 and NEK2 mRNA expression and protein levels
   (Figure [134]S4A,B, Supporting Information).

   We first examined the role of NEK2 in cell growth by using the
   clonogenic soft agar assay and BrdU intake assay, finding that
   H929‐TP53^KO/NEK2 OE cells showed a significant elevation in colony
   formation (117 ± 7) compared with other groups including H929‐NEK2 OE
   (68 ± 9), H929‐TP53^KO (87 ± 5), and H929‐Ctrl (28 ± 2, Figure [135]
   4A,B). The percentage of BrdU‐positive cells in the H929‐TP53^KO/NEK2
   OE group (43.5 ± 0.4%) was also much higher than that for other groups
   including H929‐NEK2 OE (28.4 ± 2.0%), H929‐TP53^KO (34.7%), and
   H929‐Ctrl cells (14.3 ± 0.5%, Figure [136]4C,D). In addition, after
   NEK2 was depleted in H929‐TP53^KO/shNEK2 cells via
   tetracycline‐inducible shRNA‐mediated knockdown, the number of colonies
   (57 ± 1, Figure [137]4A,B) and the percentage of BrdU‐positive cells
   (16.8 ± 1.7%, Figure [138]4C,D) were drastically reduced compared to
   those of H929‐TP53^KO cells.

Figure 4.

   Figure 4
   [139]Open in a new tab

   Dual defects in NEK2 and p53 enhance mitotic abnormalities, cell
   proliferation, apoptosis, and tumorigenesis in MM. A) Representative
   images of clonogenic analysis in H929‐Ctrl, H929‐TP53^KO, H929‐NEK2 OE,
   H929‐TP53^KO/NEK2 OE, and H929‐TP53^KO/shNEK2 cells (images are shown
   in 4x magnification). B) Statistical analysis of clone numbers formed
   in soft agarose of the five cell groups shown in panel (A). Data
   presented as mean ± SD, p‐values are calculated using unpaired one‐way
   ANOVA with Dunnett post‐hoc test, n = 3, *p < 0.05, **p < 0.01,
   ***p < 0.001. C) Representative flow cytometry dot plots for detection
   of BrdU‐positive cells among H929‐Ctrl, H929‐TP53^KO, H929‐NEK2 OE,
   H929‐TP53^KO/NEK2 OE, and H929‐TP53^KO/shNEK2 cells. D) Statistical
   analysis of the number of BrdU‐positive cells among the five cell
   groups shown in panel (C). Data presented as mean ± SD, p‐values are
   calculated using one‐way ANOVA with Dunnett post‐hoc test, n = 3,
   *p < 0.05, **p < 0.01, ***p < 0.001. E) Representative histograms for
   detection of apoptotic cells in H929‐Ctrl, H929‐TP53^KO, H929‐NEK2 OE,
   H929‐TP53^KO/NEK2 OE, and H929‐TP53^KO/shNEK2 cells treated with 4 nM
   BTZ for 48 h. F) Statistical analysis of the percentage of apoptotic
   cells among the five cell groups shown in panel (E) after treatment
   with 4 nM BTZ for 48 h. Data presented as mean ± SD, p‐values are
   calculated using one‐way ANOVA with Dunnett post‐hoc test, n = 3,
   *p < 0.05, **p < 0.01, ***p < 0.001. G) Representative images of tumor
   xenografts from B‐NDG mice with subcutaneous injections of H929‐Ctrl,
   H929‐TP53^KO, H929‐NEK2 OE, H929‐TP53^KO/NEK2 OE, or
   H929‐TP53^KO/shNEK2 cells into the right abdomen (6 mice measured for
   each group). H) Statistical analysis of tumor volumes of xenografts
   from B‐NDG mice as shown in panel (G) (6 mice measured for each group).
   Data presented as mean ± SD, p‐values are calculated using two‐way
   ANOVA with Dunnett post‐hoc test, n = 6, *p < 0.05, **p < 0.01,
   ***p < 0.001. I) Representative images for IHC detection of Ki‐67
   protein in the tumor nodules derived from B‐NDG mice injected
   subcutaneously with H929‐Ctrl, H929‐TP53^KO, H929‐NEK2 OE,
   H929‐TP53^KO/NEK2 OE, or H929‐TP53^KO/shNEK2 cells.

   Subsequently, we used an apoptosis assay to test the sensitivity of
   cells with combined NEK2/TP53 defects to Bortezomib (BTZ, a proteasome
   inhibitor and a first‐line treatment in MM). We found that the
   percentage of apoptotic cells decreased significantly in the
   H929‐TP53^KO group (75.7 ± 3.3%), and more prominently in the
   H929‐TP53^KO/NEK2 OE group (67.2 ± 4.8%), after BTZ treatment (4 or
   2 nM) when compared with the H929‐Ctrl group (96 ± 1.8%)
   (Figure [140]4E,F, Figure [141]S4C, Supporting Information). In
   addition, after NEK2 was depleted in H929‐TP53^KO/shNEK2 cells, the
   percentage of apoptotic cells (91 ± 0.1%) was markedly increased
   compared with that in H929‐TP53^KO cells (Figure [142]4E,F, Figure
   [143]S4C, Supporting Information).

   Finally, we evaluated the effects of combined NEK2 and TP53 defects in
   vivo using MM xenograft models. We found that NEK2 overexpression and
   p53 knockdown separately could increase tumor growth, and this effect
   was significantly enhanced when the defects were combined in the
   H929‐TP53^KO/NEK2 OE group; meanwhile, tumor growth was decreased in
   the H929‐TP53^KO/shNEK2 group (Figure [144]4G,H, Figure [145]S4D,
   Supporting Information). Immunohistochemistry (IHC) confirmed that the
   tumor cells from xenograft nodules were CD138^+ plasma cells with NEK2
   overexpression and p53 depletion (Figure [146]S4E, Supporting
   Information). Also, the number of proliferative (Ki‐67^+) cells
   increased significantly in the H929‐TP53^KO/NEK2 OE group while
   decreasing in the H929‐TP53^KO/shNEK2 group (Figure [147]4I, Figure
   [148]S4F, Supporting Information). Thus, dual defects in NEK2 and TP53
   promote proliferation and tumorigenesis in MM.

2.5. TP53 Deletion Upregulates NEK2 via the Regulation of Chromosomal
Instability‐Related Genes

   Because of the complexity of genetic abnormalities in cancer, we used
   HEK293 cells, with their relatively normal phenotype and genetic
   background, to examine whether p53 deletion or NEK2 overexpression can
   induce CIN. A comparative genomic hybridization (CGH) array was
   performed to scan the entire genome of normal HEK293 cells
   (HEK293‐Ctrl), cells with NEK2 overexpression (HEK293‐NEK2 OE) and
   HEK293‐TP53^KO cells. Notably, compared with HEK293‐Ctrl, several
   high‐risk genetic abnormalities, including deletion of chromosomes 5,
   7q, 8, 9, and 17p and some gains of chromosomes 1, 2q, 3p, 4p, 6, 13q,
   and 14q, were observed in HEK293‐TP53^KO cells (Figure [149] 5A, Table
   [150]S3, Supporting Information). Fewer chromosomal changes were
   observed in HEK293‐NEK2 OE cells compared with HEK293‐Ctrl cells
   (Figure [151]5C, Table [152]S4, Supporting Information). Interestingly,
   chromosome 1q21‐44, where NEK2 is located, was dramatically amplified
   in HEK293‐TP53^KO cells (Figure [153]5B).

Figure 5.

   Figure 5
   [154]Open in a new tab

   TP53 deletion upregulates NEK2 expression by inducing amplification and
   chromosomal instability in MM cells. A) Comparative analysis of
   HEK293‐TP53^KO versus HEK293‐Ctrl cells using CGH array. Red column
   represents gains of chromosomes. Blue column represents deletions.
   Chromosomal aberrations are across the whole genome. B) Schematic
   depiction of NEK2 location in chromosome 1q21.1‐q44. C) Comparative
   analysis of HEK293‐NEK2 OE versus HEK293‐Ctrl cells using CGH array.
   Red column represents gain. Blue column represents deletion. D) Left,
   representative images of FISH analysis using probes targeting TP53
   (17P, Red) and NEK2 (1q32.2, Green) in HEK293‐Ctrl and HEK293‐TP53^KO
   cells. Right, scatter plot showing the percentage of cells from both
   groups with amplified NEK2 copy numbers. Data presented as mean ± SD,
   p‐values are calculated using unpaired t test, n = 3, *p < 0.05,
   **p < 0.01. E) Scatter plot showing the percentage of H929‐Ctrl and
   H929‐TP53^KO cells with amplified NEK2 copy numbers as determined by
   FISH assay. Data presented as mean ± SD, p‐values are calculated using
   unpaired t test, n = 3, *p < 0.05, **p < 0.01. F,G) GSEA of enrichment,
   including chromosome 1q32 and chromosomal instability, from
   differentially expressed genes between HEK293‐Ctrl and HEK293‐TP53^KO
   cells. H) Representative images of spindles (Red: Tubulin), Plasmid‐GFP
   (Green), and nuclei (Blue) in H929‐Ctrl, H929‐TP53^KO, H929‐NEK2 OE,
   H929‐TP53^KO/NEK2 OE, and H929‐TP53^KO/shNEK2 cells. I) Quantification
   showing the percentage of cells with abnormal monopolarity among each
   of the five groups shown in panel (H). All results are shown as means ±
   SD of three independent experiments, and at least 60 spindles per
   experiment were randomly chosen and counted in each experiment.Data
   presented as mean ± SD, p‐values are calculated usingone‐way ANOVA with
   Dunnett post‐hoc test, *p < 0.05.

   To verify the findings from the CGH array, we subsequently performed
   FISH in HEK293‐Ctrl and HEK293‐TP53^KO cells using probes targeting the
   NEK2 and TP53 DNA regions. The number of cells with two copies of NEK2
   decreased, while the number of cells with three or more copies
   increased significantly in the HEK293‐TP53^KO cells compared with the
   HEK293‐Ctrl cells (Figure [155]5D). A similar pattern was observed for
   H929‐TP53^KO cells (Figure [156]5E). This evidence further suggests
   that p53 deletion is associated with NEK2 amplification in MM.

   Interestingly, differentially expressed genes (Tables [157]S5 and
   [158]S6, Supporting Information) were mainly found in the cytoplasm and
   nucleus. They were mostly involved in spindle formation, chromosomal
   stability and gene transcription, as revealed by R pathway enrichment
   analysis (Table [159]S7 and [160]S8, Supporting Information) when we
   used RNA‐seq to examine the gene expression profiles in HEK293‐Ctrl,
   HEK293‐TP53^KO, and HEK293‐TP53^KO/NEK2 OE cells. In addition, gene set
   enrichment analysis showed 1q32 amplification and major types of gene
   signatures enriched in CIN, cell proliferation, and p53 pathway
   (Figure [161]5F,G, Figure [162]S5A, Supporting Information). The
   expression level of CIN‐related genes was significantly altered in
   HEK293‐TP53^KO cells and more prominently in cells with combined
   defects in NEK2 and TP53 (Figure [163]S5B, Supporting Information). We
   then verified the changes in the expression of CIN‐related genes,
   including TP53, NEK2, BUBR1, HEC1, AURKA, AURKB, and MAD2L1, by using
   qPCR in HEK293 and H929 cells with or without TP53 deletion (Figure
   [164]S5C,D, Supporting Information). Thus, p53 deletion leads to NEK2
   amplification or overexpression, likely through the alteration of
   CIN‐related genes.

   Next, we used a spindle formation assay to demonstrate that the
   percentage of cells with abnormal mitosis (e.g., asymmetric spindle
   division, multipolar division, and nuclear condensation) increased
   significantly in the H929‐NEK2 OE (22.4 ± 1.5%) and H929‐TP53^KO groups
   (21.5 ± 1.5%) when compared to the H929‐Ctrl group (11.9 ± 0.3%), and
   an even greater increase was observed in the H929‐TP53^KO/NEK2 OE group
   (36.1 ± 1.2%, Figure [165]5H,I). We also depleted NEK2 in H929 cells
   with p53 deletion by using shRNA‐mediated NEK2 knockdown (denoted as
   the H929‐TP53^KO/shNEK2 group), which rescued aberrant defects in
   spindle formation (14.3 ± 0.2%, Figure [166]5H,I).

2.6. TP53 Deletion Enhances NEK2 Expression Upregulation of E2F8

   A non‐canonical p53 recognition site (CGCCATGTTGGCCAGGCTGGTCT),
   identical to the site from the huntingtin gene promoter, was previously
   identified at the distal promoter region of NEK2.^[ [167]^18 ^] To test
   the function of this site, we constructed the full‐length NEK2 promoter
   fragment (1098 bp from −1017 to −2 bp relative to the TSS, Figure
   [168]S6A, Supporting Information) and a mutant NEK2 promoter with
   deletion of the p53 binding site and inserted them into the luciferase
   reporter vector PGL3‐enhancer. We then co‐transfected the fusion
   constructs together with the TP53 expression vector into HEK293‐TP53^KO
   and ARP1 cells. Luciferase activity in cells with the full‐length NEK2
   promoter sequence decreased dramatically upon p53 expression, while it
   was mostly unaffected in cells containing the mutant promoter with p53
   binding site deletion (Figure [169] 6A,B, Figure [170]S6B,C, Supporting
   Information). ChIP‐qPCR analysis with an anti‐p53 antibody showed
   enrichment of distal NEK2 promoter sequences compared with the IgG
   isotype control, similar to the positive control p21 gene (also known
   as cyclin dependent kinase inhibitor 1A, CDKN1A) (Figure [171]6C),
   indicating that p53 suppresses NEK2 transcription by directly binding
   to the NEK2 promoter.

Figure 6.

   Figure 6
   [172]Open in a new tab

   p53 deletion promotes aberrantly high NEK2 expression through indirect
   transcriptional regulation. A,B) Luciferase activity driven by NEK2
   promoter with normal (full length) or mutant p53 binding site in
   HEK293‐TP53^KO and ARP1 cells transiently co‐transfected with p53 OE.
   Data presented as mean ± SD, p‐values are calculated using unpaired t
   test, n = 3, **p < 0.01. C) ChIP confirming that the transcription
   factor p53 specifically binds to the NEK2 promoter region in H929
   cells. Data presented as mean ± SD, p‐values are calculated using
   unpaired t test, n = 3, ****p < 0.0001. D) Correlation of expression
   between NEK2 and E2F8 in MM patients based on GEP database
   ([173]GSE2658, n = 559, p‐values are calculated Spearman correlation
   coefficient ). E) The mRNA and protein levels of TP53, CDKN1A (p21),
   and E2F8 in H929 cells with or without p53 deletion. Here and in panels
   (F–J), mRNA and protein data were obtained using qPCR and
   immunoblotting, respectively. Data presented as mean ± SD, p‐values are
   calculated using unpaired t test, n = 3, ***p < 0.001. F) The mRNA and
   protein levels of TP53, CDKN1A (p21), and E2F8 in ARP1 cells with or
   without p53 overexpression. Data presented as mean ± SD, p‐values are
   calculated using unpaired t test, n = 3, *p < 0.05, ***p < 0.001. G,H)
   The mRNA and protein levels of NEK2 and E2F8 in H929‐Ctrl and
   H929‐TP53^KO cells with (H929‐TP53^KO/E2F8 OE) or without E2F8
   overexpression. Data presented as mean ± SD, p‐values are calculated
   using unpaired t test, n = 3, *p < 0.05, **p < 0.01. I,J) The mRNA and
   protein levels of NEK2 and E2F8 in ARP1‐EV and ARP1‐TP53 OE cells with
   (ARP1‐TP53 OE/E2F8 OE) or without E2F8 overexpression. Data presented
   as mean ± SD, p‐values are calculated using unpaired t test, n = 3,
   **p < 0.01. K) ChIP confirmed that the transcription factor E2F8
   specifically binds to the NEK2 promoter region. L) Schematic of p53
   deletion enhancing NEK2 expression through E2F8 upregulation.

   Subsequently, we analyzed the differentially expressed genes (Table
   [174]S9, Supporting Information) obtained from gene expression profiles
   in H929‐Ctrl, H929‐TP53^KO, and H929‐TP53^KO/NEK2 OE cells. CIN‐related
   genes, along with genes in the cyclin and E2F families, were
   differentially expressed in H929‐TP53^KO cells and, to an even larger
   degree, in H929‐TP53^KO/NEK2 OE cells (Figure [175]S6D, Supporting
   Information). These alterations of cyclin and E2F family gene
   expressions in HEK293‐TP53^KO, HEK293‐TP53^KO/NEK2 OE, H929‐TP53^KO,
   and H929‐TP53^KO/NEK2 OE cells were also verified by qPCR (Figure
   [176]S6E, Supporting Information). It was previously reported that NEK2
   can be regulated by E2F1, E2F3, and E2F4 and that these genes are
   targets of p53.^[ [177]^19 ^] By using GEP data ([178]GSE2658) mining,
   we found that the expression of E2F8, but not other E2Fs, positively
   and strongly correlates with NEK2 expression in MM samples (r = 0.79,
   Figure [179]6D, Figure [180]S6F, Supporting Information). TP53 deletion
   in H929‐TP53^KO cells caused the mRNA and protein levels of E2F8 to
   increase significantly when compared with the control cells
   (Figure [181]6E). In contrast, p53 overexpression in ARP1 cells
   (ARP1‐TP53 OE) decreased E2F8 mRNA and protein levels (Figure [182]6F).
   It was previously shown that E2Fs can promote NEK2 transcriptional
   activity by binding to the E2F site in the NEK2 promoter region (−169
   to −179 bp)^[ [183]^19 ^b] (Figure [184]S1B, Supporting Information).
   We found that E2F8 overexpression in H929‐Ctrl and H929‐TP53^KO cells
   significantly upregulated NEK2 levels (Figure [185]6G,H). Similarly,
   E2F8 overexpression in ARP1‐EV and ARP1‐TP53 OE cells also
   significantly upregulated NEK2 expression, while NEK2 levels were
   markedly lower in ARP1‐TP53 OE cells than in ARP1‐EV cells with E2F8
   overexpression (Figure [186]6I,J). Additionally, ChIP‐qPCR analysis
   showed that the E2F8 binding site in the NEK2 promoter was enriched in
   ARP1 cells (Figure [187]6K). Thus, p53 can also inhibit NEK2
   transcription via the regulation of cell cycle pathway‐related genes
   such as E2F8 in MM (Figure [188]6L).

2.7. TP53 Suppresses NEK2 Expression via Downregulation of DNA
Methyltransferases

   It was previously reported that p53 represses NEK2 expression and
   protects its binding region from accumulating DNA methylation.^[
   [189]^18 ^] We also found that DNA methylation was elevated in H929
   cells with TP53 deletion (TP53^KO‐C1 and TP53^KO‐C2; Figure [190]S7A,B,
   Supporting Information). To explore how p53 represses NEK2 expression
   through DNA methylation, we first examined the levels of DNA
   methylation‐related genes (DNMT1, DNMT3a, DNMT3b) and histone
   methylation‐related genes (NSD3, PRMT1, SETD5, and SETD7) and found
   that they were upregulated in H929‐TP53^KO compared with H929‐Ctrl
   cells (Figure [191] 7C,D). Upregulation of the histone
   methylation‐related genes (NSD3, PRMT1, SETD5, and SETD7) caused
   H3K36me3 protein methylation (Figure [192]7E). Subsequently, we showed
   that the expression of DNMT3b was significantly upregulated compared
   with the expression of the other two DNA methyltransferase genes (DNMT1
   and DNMT3a) in H929‐TP53^KO cells (Figure [193]7F,G). Aberrantly high
   expression of DNMT3b and DNMT1 coincided with upregulated NEK2
   expression (Figure [194]S7B, Supporting Information). When DNMT3b and
   DNMT1 were depleted via shRNA, NEK2 mRNA and protein expression levels
   were dramatically downregulated in H929 cells with or without p53
   (Figure [195]7H,I, Figure [196]S7C–F, Supporting Information), while
   lower NEK2 expression levels were found in H929‐Ctrl cells expressing
   wild type p53. Additionally, a co‐immunoprecipitation (co‐IP) assay
   showed that p53 interacted with DNMT1 and DNMT3a, but not with DNMT3b,
   in H929 cell lines (Figure [197]7J). We also confirmed that NEK2
   methylation levels decreased in p53‐overexpressing ARP1 cells (Figure
   [198]S7F, Supporting Information). These results suggested that p53
   protects its binding region in the NEK2 promoter from accumulating DNA
   methylation by repressing DNMT expression (Figure [199]7K).

Figure 7.

   Figure 7
   [200]Open in a new tab

   p53 suppresses NEK2 expression by regulating the expression of DNMTs.
   A) Detailed BGS analysis confirmed methylation status of the NEK2
   promoter's distal CpG island in H929 cells with or without p53
   deletion. B) Statistical analysis of methylated CpG dinucleotides in
   the NEK2 promoter's distal CpG nucleotides in H929 cells with or
   without p53 deletion. Data presented as mean ± SD, n = 11, p‐values are
   calculated using one‐way ANOVA with Bonferroni correction, **p < 0.01,
   ****p < 0.0001. C) Heatmap of the ratios of the signal intensities of
   differential methylation‐related genes in H929 cells with or without
   p53 deletion. D) The mRNA levels of histone methylation‐related genes
   (NSD3, PRMT1, SETD5, and SETD7) detected by qPCR in H929 cells with or
   without p53 deletion. Data presented as mean ± SD, p‐values are
   calculated using unpaired t test, n = 3, *p < 0.05, **p < 0.01,
   ***p < 0.001. E) H3K36me3 protein levels detected using immunoblotting
   in H929 cells with or without p53 deletion. F,G) The mRNA and protein
   levels of DNMT1, DNMT3a, and DNMT3b in H929 cells with or without p53
   deletion using qPCR and immunoblotting. Data presented as mean ± SD,
   p‐values are calculated using unpaired t test, n = 3, ***p < 0.001.
   H,I) The mRNA and protein levels of NEK2 after shRNA‐mediated DNMT3b
   knockdown in wild type p53 and p53‐deleted H929 cells, detected by qPCR
   and immunoblotting. Data presented as mean ± SD, p‐values are
   calculated using unpaired t test, n = 3, **p < 0.01. J) Endogenous p53
   was pulled down by p53 antibodies in H929 cells with or without p53
   deletion, and the DNMT1, DNMT3a, and DNMT3b proteins were analyzed by
   western blotting. The lysates before IP were used as a positive
   control. K) Dysfunctional p53 elevates NEK2 expression in MM by
   upregulating DNMT expression, thereby increasing methylation of the
   NEK2 promoter.

2.8. TP53 Overexpression Synergistically Interacts with NEK2 Suppression to
Promote Tumor Formation and Reduce Bortezomib Sensitivity

   We next determined the effect of combined p53 overexpression and NEK2
   suppression on chromosome stability, proliferation, and drug resistance
   in TP53^−/− MM cell lines using in vitro and in vivo assays. Spindle
   formation assay experiments showed that the number of cells with
   abnormal mitosis were reduced after p53 ectopic expression in ARP1
   cells (ARP1‐TP53 OE, 17.6 ± 0.7%) compared with control cells (ARP1‐EV,
   19.2 ± 0.9%). Meanwhile, the number of cells with mitotic abnormalities
   was elevated after NEK2 overexpression in ARP1‐TP53 OE cells (ARP1‐TP53
   OE/NEK2 OE, 21.7 ± 0.8%). However, the number of defective cells was
   substantially reduced when NEK2 was depleted by shRNA‐mediated
   knockdown in ARP1‐TP53 OE cells (ARP1‐TP53 OE/shNEK2, 10.3 ± 0.7%)
   (Figure [201] 8A,B). Similarly, BrdU intake assay experiments revealed
   that the number of BrdU‐positive cells decreased significantly in the
   ARP1‐TP53 OE group, but increased in the ARP1‐TP53 OE/NEK2 OE group,
   compared with ARP1‐EV cells. In contrast, the number of defective cells
   decreased dramatically in ARP1‐TP53 OE/shNEK2 cells compared to ARP1‐EV
   or ARP1‐TP53 OE cells (Figure [202]8C,D).

Figure 8.

   Figure 8
   [203]Open in a new tab

   Downregulation of NEK2 in TP53‐deleted MM cells inhibits cell growth
   and decreases drug resistance. A,B) Representative images of spindles
   (Red, Tubulin), GFP (Green), and nuclei (Blue) in ARP1‐Ctrl, ARP1‐TP53
   OE, and ARP1‐TP53 OE/shNEK2 cells. Quantifications show the percentage
   of abnormal monopolarity from three independent experiments. At least
   60 spindles were randomly chosen and counted in each experiment. All
   results are shown as means ± SD of three independent experiments, Data
   presented as mean ± SD, p‐values are calculated using one‐way ANOVA
   with Dunnett post‐hoc test, *p < 0.05. Flow cytometric detection of
   BrdU‐positive cells in ARP1‐Ctrl, ARP1‐TP53 OE, and ARP1‐TP53 OE/shNEK2
   cells. Representative C) dot plots and D) quantification. Data
   presented as mean ± SD, p‐values are calculated using one‐way ANOVA
   with Dunnett post‐hoc test, n = 3, *p < 0.05, **p < 0.01. E)
   Representative histograms for detection of apoptotic cells and F)
   statistical analysis of the percentage of apoptotic cells in ARP1‐Ctrl,
   ARP1‐TP53 OE, ARP1‐TRP53 OE/NEK2 OE, and ARP1‐TP53 OE/shNEK2 cells
   treated with 4 nM BTZ for 48 h. Data presented as mean ± SD, p‐values
   are calculated using one‐way ANOVA with Dunnett post‐hoc test, n = 3,
   **p < 0.01. G) Representative images of tumor xenografts in B‐NDG mice
   after subcutaneous injection of ARP1‐Ctrl, ARP1‐TP53 OE, ARP1‐TP53
   OE/NEK2 OE, or ARP1‐TP53 OE/shNEK2 cells into the right abdomen. Cells
   were treated with BTZ or with PBS as a negative control. Tumor volumes
   of xenografts derived from B‐NDG mice injected with ARP1‐Ctrl,
   ARP1‐TP53 OE, ARP1‐TP53 OE/NEK2 OE, or ARP1‐TP53OE/shNEK2 cells (n =
   4). Cells were treated I) with BTZ or H) with control. Data presented
   as mean ± SD, p‐values are calculated using two‐way ANOVA with Dunnett
   post‐hoc test, n = 6, *p < 0.05, **p < 0.01, ***p < 0.001. J)
   Representative images for IHC detection of Ki‐67 protein in the tumor
   nodules derived from B‐NDG mice injected subcutaneously with ARP1‐Ctrl,
   ARP1‐TP53 OE, ARP1‐TP53 OE/NEK2 OE, or ARP1‐TP53 OE/shNEK2 cells. Cells
   were treated with BTZ (right) or with control (left).

   Subsequently, we tested the sensitivity of cells with combined p53 and
   NEK2 depletion to BTZ by using an apoptosis assay. We found that the
   percentage of apoptotic cells increased significantly in the ARP1‐TP53
   OE/shNEK2 group (30.0 ± 3.04%) after 4 nM BTZ treatment when compared
   with the ARP1‐Ctrl (14.1 ± 1.41%) and ARP1‐TP53 OE (17.3 ± 0.42%)
   groups (Figure [204]8E,F).

   In addition, by using an MM xenograft model with B‐NDG immunodeficient
   mice, we confirmed that the tumor sizes produced from ARP1‐TP53 OE
   cells were much smaller than those from ARP1‐EV or ARP1‐TP53 OE/NEK2 OE
   cells (Figure [205]8G,H). Furthermore, tumors formed by ARP1‐TP53
   OE/shNEK2 cells were the smallest and were much more sensitive to BTZ
   than were ARP1‐TP53 OE cells (Figure [206]8G,I). Subsequent IHC
   analysis also revealed that cells in tumor nodules were CD138^+ (Figure
   [207]S8A, Supporting Information), and the number of Ki‐67^+ cells was
   reduced, suggesting that the recovery of p53 function and further
   suppression of NEK2 are beneficial for the p53‐deleted group of cells
   (Figure [208]8J, Figure [209]S8B, Supporting Information). Thus, stable
   expression of wild type p53 rescues the defects in mitosis,
   proliferation, and tumorigenesis in MM cells and enhances the
   therapeutic effect of BTZ both in vitro and in vivo. This rescue effect
   can be further improved when combined with NEK2 depletion.

3. Discussion

   Collaboration between oncogenes and tumor suppressor genes is an
   important mechanism in the development of MM. Here, we showed that NEK2
   amplification is a major cause for NEK2 upregulation in MM and other
   cancer types, especially in TP53^−/ ^− MM patients. Previously, we had
   reported that NEK2 overexpression induces drug resistance,
   proliferation and CIN in cancer cells.^[ [210]^11 ^] In this study, we
   found that patients with combined defects of NEK2 activation and p53
   inactivation suffer from poor survival and that collaboration between
   NEK2 and TP53 defects augments MM cell growth and drug resistance in
   vitro and in vivo.

   Gene mutations, DNA amplification and promoter methylation status are
   the major causes for aberrantly high expression of candidate tumor
   genes.^[ [211]^17 ^a, [212]^17 ^] Previous studies focused on the
   function, mechanisms and protein stability of NEK2, but how NEK2 is
   activated and upregulated remained poorly understood. In this study,
   using a FISH probe for NEK2 DNA, we examined the copy number of NEK2 in
   MM cells and patients and assessed mutations in the NEK2 promoter and
   exons. We further examined DNA methylation in the proximal and distal
   NEK2 promoter regions in MM cell lines and analyzed the effect of NEK2
   mutation on its expression (data not shown) using genomic data from a
   recent MMRF study (study accession phs000748). Interestingly, we found
   that NEK2 is only amplified in the MM cells of patients who have a poor
   rate of survival, suggesting that NEK2 amplification is the major cause
   for NEK2 activation in MM. To date, NEK2 DNA amplification in tumors is
   still poorly documented due to a lack of appropriate FISH probes for
   detecting the chromosomal copy number of NEK2. In our present study, we
   prepared a DNA probe of NEK2 to detect its copy number with FISH,
   leading us to find that NEK2 amplification occurs in 23.5% of MM (AD)
   patients, similar to the data from MMRF study. Thus, it is highly
   likely that our prepared DNA probe can be used for clinical detection
   of NEK2 amplification in the future.

   The 17p chromosomal region harbors the gene locus of TP53, an important
   tumor suppressor gene.^[ [213]^21 ^] Deletion of this region is a
   recurrent cytogenetic abnormality present in 10–34% of MM cases along
   with disease progression and is considered an independent factor
   responsible for less favorable clinical outcome in MM patients.^[
   [214]^2 ^a, [215]^4 ^b] In keeping with this notion, the lesions
   associated with short OS in multivariate analysis are +1q and del17p13
   in MM,^[ [216]^21 , [217]^22 ^] suggesting that the combined
   cytogenetic abnormality contributes to the progression of MM. Despite
   these advances, the molecular mechanisms underlying p53's action are
   poorly understood. MDM4, a homolog of MDM2, is located in the +1q
   region and inactivates p53 by binding to and inhibiting its
   transactivation.^[ [218]^22 ^a, [219]^23 ^] Whether this occurs in MM
   is unknown. Our current studies demonstrated that NEK2 amplification
   correlates strongly with TP53 deletion in MM and has a significant
   clinical impact. Previously, it had been reported that NEK2 and p53
   modulate each other. On the one hand, methylation of the distal NEK2
   promoter containing the p53‐binding site affects p53 binding to the
   promoter, resulting in the inability of p53 to attenuate NEK2
   expression.^[ [220]^18 ^] On the other hand, NEK2 can attenuate the
   function of wild type p53 by inhibiting its phosphorylation.^[ [221]^23
   ^]

   In this study, we found that, under normal physiological conditions,
   p53 is an active protein that binds to the transactivation domain in
   the NEK2 promoter. This inhibits NEK2 transcription through the cell
   cycle pathway mediated by p53‐p21‐DREAM‐E2F and protects its distal CpG
   island from increased methylation. The E2F transcriptional factors
   mediate various biological functions involved in cell cycle
   progression.^[ [222]^24 ^] Previous studies suggested that cell cycle
   arrest is achieved through indirect transcriptional repression by
   p53‐p21‐DREAM‐E2F/CHR.^[ [223]^25 ^] In this pathway, the target genes
   with E2F or CHR promoter sites are transcriptionally regulated by the
   DREAM transcriptional inhibitory complex.^[ [224]^25 ^] One E2F binding
   site in the NEK2 promoter region (−169 to −179 bp) was previously
   reported.^[ [225]^19 ^b] Our studies demonstrated, for the first time,
   that E2F8 serves as a transcription factor to induce NEK2 expression
   through the p53/E2F8 pathway in MM. It is well‐known that chromatin
   opening and gene transcriptional activation are regulated by the status
   of histone modifications. H3K36 methylation, which is regulated by
   histone methylation modification‐related genes such as NSD3, PRMT1,
   SETD5, and SETD7, can promote gene transcriptional activation.^[
   [226]^26 ^] Our results showed that mRNA levels of these genes and
   H3K36me3 protein levels increase in TP53‐knockout MM cells
   (H929‐TP53^KO). In mammals, DNA methylation is mainly catalyzed by
   DNMTs including the DNMT1, DNMT2 and DNMT3 families. DNMT1 is a
   maintenance methylase, while DNMT2 can bind to specific heterotopic
   sites on DNA, but its specific role is not clear. DNMT3a and DNMT3b are
   re‐methylases that re‐methylate demethylated CpG sites to participate
   in de novo methylation of DNA. In this study, we found that the
   methylation of CpG sites within the NEK2 promoter distal CpG island,
   which are the p53 binding sites reported in previous studies,^[
   [227]^18 ^] was elevated in p53‐deleted MM cells. The mRNA and protein
   levels of DNMT3b and DNMT1 protein were also upregulated significantly.
   In addition, co‐IP experiments showed that p53 interacts with DNMT1.
   The expression of DNMT1 and DNMT3b was highly positively correlated to
   increased NEK2 expression, and the mRNA and protein levels of NEK2 were
   decreased when DNMT1 or DNMT3b was knocked down with shRNA, especially
   in DNMT3b depleted MM cells. In this study, p53 inhibited the
   expression of the DNMT3b and DNMT1 proteins, reducing methylation of
   the NEK2 promoter CpG island. Thus, wild type p53 binds to NEK2 DNA and
   inhibits its transcription. These results suggest that p53 suppresses
   NEK2 expression by regulating DNMT expression, thereby protecting its
   binding region from accumulating DNA methylation. Together, when TP53
   has gain‐of‐function mutation or loss, the dysfunctional p53 elevates
   NEK2 expression genetically by inducing NEK2 amplification,
   transcriptionally by mediating E2F8 activity and epigenetically by
   repressing DNMT expression (Figure [228] 9 ).

Figure 9.

   Figure 9
   [229]Open in a new tab

   Working model depicting the regulation of NEK2 expression by TP53.

   We found that p53 regulates NEK2 at both the genetic and the
   transcriptional levels in MM based on the following observations.
   First, at the clinical level, deletion of the TP53‐containing
   chromosome occurs in 9.3% (51/548) of MM samples and correlates with
   aberrantly high expression of NEK2. Indeed, we found NEK2 amplification
   in 25.5% (13/51) of MM samples, and it was closely linked to high‐level
   expression of NEK2. Furthermore, experiments with TCGA data mining
   showed that TP53 deletion correlates strongly with NEK2 amplification
   and aberrantly high expression in several cancer types. Thus, TP53
   correlates inversely with NEK2 amplification and overexpression in
   patients with MM and other cancer types. Second, at the genetic level,
   we used CGH and FISH analyses to observe NEK2 amplification in TP53^−/
   ^− cell lines. Third, at the transcriptional level, experiments with
   the luciferase reporter assay and ChIP‐qPCR confirmed that p53 binds
   directly to the NEK2 promoter in MM cells. Wild type p53 inhibits NEK2
   transcription, while mutant p53 significantly increases its expression.
   Finally, at the functional level, combined TP53 deletion and NEK2
   overexpression in WT TP53 MM cell lines induce asymmetric mitosis and
   promote proliferation and tumorigenic activity in vitro and in vivo.
   Conversely, concomitant p53 overexpression and NEK2 inhibition in
   TP53^−/− cells can partially rescue these defects, including drug
   resistance, in MM cells in vitro and in vivo.

   A preclinical study using myeloma demonstrated that adenovirus‐mediated
   delivery of WT TP53 can potently induce apoptosis.^[ [230]^27 ^] Our
   current study also suggests that delivery of WT TP53 and inhibition of
   NEK2 in a TP53 ^−/− MM cell line can suppress tumor formation and
   enhance BTZ's therapeutic effect. Our findings suggest that targeting
   the function of the NEK2 and p53 pathways may have therapeutic values
   by reversing the adverse outcome of MM patients without p53.

   In conclusion, we found that NEK2 amplification leads to NEK2
   upregulation, while combined defects in TP53 and NEK2 can be used as a
   novel marker for poor prognosis in a cohort of MM patients. We also
   validated NEK2 as a novel therapeutic target in the TP53^−/− subset of
   MM and revealed a novel mechanism by which TP53 regulates NEK2 at both
   the genetic and transcriptional levels. Future experimentation should
   examine the clonal evolution of the subgroup containing TP53 and NEK2
   dual defects and their functions in MM initiation, promotion and drug
   resistance. Selective inhibitors should be developed to improve the
   clinical outcomes for these patients.

4. Experimental Section

Human Samples

   Human bone marrow samples were obtained from healthy donors (n = 8) and
   newly diagnosed (n = 82) and relapsed (n = 25) MM patients in the Third
   Xiangya Hospital, Central South University (Changsha, China) and
   Institute of Hematology and Blood Disease Hospital, Chinese Academy of
   Medical Sciences and Peking Union Medical College (Tianjin, China).
   Written informed consent was obtained from all participants. The
   sampling procedure was approved by the Cancer Research Institute of the
   Central South University Medical Ethics Committee. Primary MM cells and
   normal plasma cells were isolated from the mononuclear cells of BM
   samples using CD138 MicroBeads (Miltenyi Biotec, Auburn, CA).

Cell Culture and Reagents

   The human MM cell lines H929, MM.1S, MM.1R, RPMI‐8226, and U266 were
   purchased from the American Type Culture Collection (ATCC, Manassas,
   VA). HEK293, ARP1, and OCI‐my5 cell lines were obtained from the Cancer
   Research Institute of Central South University. HCT116‐TP53^WT and
   HCT116‐TP53^−/− cells were provided by Dr. Tiebang Kang (Sun Yat‐sen
   University Cancer Center, Guangzhou, China) and maintained in McCoy's
   5A medium (#16 600 082, Gibco‐BRL, Grand Island, NY, USA) supplemented
   with 10% fetal bovine serum (FBS) (#04‐001‐1, Biological Industries
   (BI), Kibbutz Beit‐Haemek, Israel) and 1x penicillin and streptomycin
   (P/S) (#15140‐122, Gibco‐BRL). All MM cell lines were maintained in
   RPMI‐1640 medium (#C11875500BT, Gibco‐BRL, Suzhou, China) supplemented
   with 10% FBS and 1x P/S. HEK293 cells were identified by STR analysis
   (Supporting information 2) and maintained in DMEM medium (#01‐051‐1ACS,
   BI) supplemented with 10% FBS and 1x P/S.

   Anti‐NEK2 (D‐8, #sc‐5560), anti‐p21 (F‐5, #sc‐6246), and anti‐E2F2
   (KH95, #sc‐251) and mouse or rabbit secondary antibodies were from
   Santa Cruz Biotechnology (CA, USA). Anti‐p53 (#10442‐1‐AP), anti‐E2F8
   (#13425‐1‐AP), and anti‐GAPDH (#10494‐1‐AP) antibodies were purchased
   from Proteintech (Wuhan, Hubei, China), while nutlin‐3a and doxycycline
   were purchased from Selleckchem (#S8059, Houston, TX, USA). Doxorubicin
   was from Harbin Pharmaceutical Group Bioengineering Co. (Harbin,
   China).

Vectors and Transfections

   NEK2, E2F8, and wild type and mutant TP53 cDNAs containing open reading
   frames (ORF) were subcloned into the pCDH vector tagged with GFP or RFP
   (#CD511B‐1, Addgene, Watertown, MA, USA). NEK2, E2F8, and wild type
   TP53 ORF sequences (Table [231]S10, Supporting Information) were
   amplified by RT‐PCR using RNA purified from HEK293 cells as a template
   with TP53‐ORF primers and confirmed by sequencing. Mutant TP53
   expression vector was obtained from wild type TP53 expression vector by
   site‐directed mutagenesis (Supporting Information 3). The shRNAs (shRNA
   sequence for NEK2 and E2F8 is shown in Table [232]S10, Supporting
   Information) were inserted into EcoRI and Xhol sites of the pLKO vector
   tagged with Tet and puromycin (Addgene). Lentiviruses were packaged in
   HEK293T cells using pMD2G and psPAX2 helper vectors and polybrene (3 µg
   mL^−1) ‐mediated transduction (#H9268‐5G, Sigma‐Aldrich, Missouri,
   USA). Transient transfection was performed using Lipofectamine 3000
   reagent (#L3000015, Invitrogen, California, USA) according to the
   manufacturer's instructions. NEK2 and TP53 overexpressing cells were
   sorted with flow cytometry by GFP or RFP 5 days post‐infection. Cells
   expressing NEK2 shRNAs were selected with 1–2 µg mL^−1 puromycin
   (#A1113803, Invitrogen) for 24 h.

   TP53 homozygous gene knockout cell lines were established with the
   CRISPR‐Cas9 system. CRISPR‐Cas9‐Lenti‐V2‐TP53, including four sgRNA
   guide sequences and CRISPR‐Cas9‐Lenti‐V2 (as a control), was provided
   by Dr. Tiebang Kang (Sun Yat‐sen University Cancer Center). 3 × 10^5
   HEK293 cells were transfected with four different sgRNA guide
   sequences, and cells were cultured for 3 days. After western blot
   analysis showed the p53 protein level in transfected cells was shown to
   significantly decrease, the target sgRNA was selected to subsequently
   establish cell lines. Transfected HEK293 cells were selected with
   puromycin (1 µg mL^−1) for 24 h and then maintained continuously with
   0.5 µg mL^−1 puromycin. Small single colonies emerging from single
   cells were picked and expanded to derive isogenic cell lines with
   defined mutations. To obtain H929 cell lines with TP53‐knockout by
   using CRISPR, TP53 sgRNA was inserted into the pL‐CRISPR.EFS.GFP vector
   provided by Jiaxi Zhou (State Key Laboratory of Experimental
   Hematology). Next, H929 cells were infected with lentiviruses
   containing pL‐CRISPR.EFS.GFP ‐TP53 plasmids and sorted by GFP 5 days
   after infection. Single colonies were obtained by sequential dilution
   after sorting (Supporting Information 3). Deletion of the TP53 gene in
   selected HEK293 and H929 cell clones was confirmed by qPCR and western
   blotting. The sgRNA guide sequences and genotyping primers are listed
   in Table [233]S10, Supporting Information.

Fluorescence In‐Situ Hybridization

   FISH was performed on interphase nuclei using established methods.^[
   [234]^11 ^] As previously described, all MM cell samples were purified
   using Miltenyi technology (anti‐CD138‐coated magnetic beads) before
   FISH. To detect NEK2 amplification, a CEP1 probe targeting chromosome 1
   (#CHR01‐10‐GR, Empire Genomics, Buffalo, NY, USA) and a bacterial
   artificial chromosome (BAC) at 1q32.2 (#RP11‐1114G13, Invitrogen) were
   purchased. The BAC probe was labeled using the Nick Translation Kit
   (#32‐801300, Abbott, Chicago, USA) with green‐dUTP (#02N32‐050, Abbott)
   or orange PF555‐dUTP (#PK‐PF555‐8‐100, PromoKine, Germany). The status
   of the TP53 gene was analyzed using a DNA probe targeting TP53
   (#TP53‐20‐OR, Empire Genomics). Interphase FISH was performed according
   to the procedure described previously. Briefly, the probes were
   hybridized to CD138^+ cells. The slides were stored at 4 °C until FISH
   analyses were performed. Interphase FISH signals were evaluated in at
   least 200 interphase nuclei in each sample. If at least three copies
   were seen in at least 20% of CD138^+ cells, it was considered evidence
   of gain/amplification. To investigate effects of the magnitude of
   AmpNEK2 on clinical outcomes and each category, AmpNEK2 was divided
   into two categories: 1) Three copies of the NEK2 probe (the percentage
   of clonal plasma cells with at least three copies was <20%) and 2) more
   than three copies of NEK2 probe (the percentage of clonal plasma cells
   with more than three copies was ≥ 20%).

Soft Agar Clonogenicity Assay

   1000 cells per well were seeded in 12‐well plates for double‐layer agar
   cultures for 3 weeks. Cells were resuspended in 0.3% agar (#16 520 100,
   Invitrogen) in RPMI‐1640 medium supplemented with 15% FBS. Cells were
   incubated (37 °C, 5% CO[2]) and fed with the same medium every three
   days on the up‐layer for three weeks. The aggregates of cells > = 50
   cells were defined as colonies. Photographs of all plates were scanned
   with ChemiDoc XRS (Bio‐Rad, California, USA), and the colonies were
   counted using ImageJ (NIH, USA).

BrdU Assay

   For the BrdU assay, all procedures followed the standard protocol with
   the APC BrdU Flow Kit (#552 598, BD, New Jersey, USA). Cells were
   labeled with BrdU in culture medium for 1 h. Subsequently, the
   incorporated BrdU was stained with specific anti‐BrdU fluorescent
   antibodies, and 7‐aminoactinomycin D (7‐AAD) was used to label total
   DNA in conjunction with BrdU staining. Finally, cells stained with BrdU
   and 7‐AAD were examined by flow cytometry and analyzed with FlowJo 10.0
   software.

Apoptosis

   Apoptotic cells were labeled by APC‐conjugated Annexin V (BD). Dead
   cells were labeled by 7‐AAD (BD). Cell staining was performed according
   to the manufacturer's protocol. Labeled cells were then measured by
   CytoFLEX (Beckman Instruments, Inc, CA, USA). The percentages of
   apoptotic cells were calculated using FlowJo software.

Immunofluorescence

   Bone marrow aspirates from human myeloma patients were sorted with
   anti‐CD138 magnetic beads and mounted onto cytospin slides for this
   study. Myeloma cells were fixed in 4% formaldehyde, and primary
   antibodies against NEK2 (mouse anti‐human, D‐8, #sc‐55601, Santa Cruz
   Biotechnology) and p53 (rabbit anti‐human, #10442‐1‐AP, Proteintech)
   were added at a final dilution of 1:100 followed by overnight
   incubation at 4 °C. The secondary antibodies goat anti‐rabbit Alexa
   Fluor 488 (#A21202, Invitrogen) and goat anti‐mouse Alexa Fluor 594
   (#A21207, Invitrogen) were added at a final dilution of 1:1000 for 1 h
   at room temperature. The slides were washed and mounted with DAPI.
   Images were captured using a confocal microscope (Nikon, Tokyo, Japan).

Luciferase Activity Assay

   The NEK2 promoter sequence ranging from −1017bp to −2 bp, with or
   without the p53 binding site, was inserted into a pGL3‐enhancer vector
   (Promega, Wisconsin, USA) and subsequently a luciferase reporter gene
   vector. The luciferase reporter gene constructs were named pGL3‐NEK2‐P6
   and pGL3‐NEK2‐P6M, respectively. Subsequently, pGL3‐control (positive
   control, Promega), pGL3‐NEK2‐P6, pGL3‐NEK2‐P6M, and the internal
   control Renilla (pRL‐null, #E2271, Promega) vectors were co‐transfected
   with wild type pcDNA3.1‐TP53‐FLAG and the TP53 expression vector into
   HEK293‐TP53^KO and ARP1 cells using Lipofectamine 3000. Cells were
   harvested after 48 h cultivation, and a luciferase activity was
   detected using Dual‐Luciferase Reporter Assay System (Promega) and a
   GloMax 20/20 Luminometer (Promega) according to the manufacturer's
   protocol. All samples were done in triplicate. The primers for
   luciferase reporter gene constructs are listed in Table [235]S10,
   Supporting Information.

Chromatin Immunoprecipitation

   The binding of the NEK2 subunit to DNA in TP53‐knockout H929‐T53^KO and
   wild type TP53 H929‐Ctrl MM cell lines was quantified with
   ChIP‐quantitative PCR (ChIP‐qPCR). The chromatin immunoprecipitation
   (ChIP) assay was performed with the EZ‐ChIP kit (#17‐371 RF, Millipore,
   Massachusetts, USA), Briefly, chromatin (5 µg) from the two myeloma
   cell lines was used in the ChIP assay using antibodies (3 µg) against
   p53 (DO‐1, #sc‐126 X, Santa Cruz Biotechnology) or E2F8 (#13425‐1‐AP,
   Proteintech). The ChIP DNA fragments were quantified with the EZ‐ChIP
   kit and the enrichment of DNA fragments containing putative p53 binding
   sites in the gene promoter was quantified by qPCR using a LightCycler
   96 (Roche, Basel, Switzerland). Two specific primers for NEK2, p21, and
   GAPDH (Table [236]S10, Supporting Information) were used to amplify
   fragments containing the predicted p53 or E2F8 binding sites in the
   NEK2 promoter region. As a negative control, GAPDH was also amplified
   with the corresponding primers, while p21 was used as a positive
   control. Values obtained from immunoprecipitated samples were
   normalized to that of their corresponding input samples. Data are
   representative of three separate experiments, and error bars indicate
   mean ± SD. The primers used in the ChIP‐qPCR assay are listed in Table
   [237]S10, Supporting Information.

Comparative Genomic Hybridization

   A CGH array of HEK293‐TP53KO and HEK293‐NEK2 OE versus HEK293‐Ctrl
   cells was performed to identify, in combination with bioinformatics,
   amplified or deleted genes among these three cell types. The SurePrint
   G3 Human CGH Microarray Kit, 2 × 400K chip (Agilent, California, USA)
   to detect genome‐wide differences was used. After the completion of
   hybridization, the array slides were taken out and washed, then placed
   into an Agilent Microarray Scanner for scanning. After scanning, the
   data were interpreted with Agilent Feature Extraction software. The CGH
   differential region was then calculated using Agilent CytoGenomics
   software. The detection method is numbered AG‐GC‐WL01‐01‐2012, and the
   data analysis method is numbered AG‐GC‐DL01‐01‐2010. The whole testing
   process was completed by Capital Bio Technology (Shanghai, China).

Assessment of Mitotic Spindle Phenotypes

   Assessment of mitotic spindle phenotypes was performed according to the
   procedure described previously.^[ [238]^28 ^] Briefly, cells were fixed
   in 4% paraformaldehyde (#P1110, Solarbio, Beijing, China),
   permeabilized by 0.1% Triton X‐100 (Sigma‐Aldrich) and incubated with
   fluorescent anti‐α‐tubulin (#ab7291, Abcam, Cambridgeshire, UK) at
   1:1000 dilution for 2 h, with the secondary antibodies goat anti‐mouse
   Alexa Fluor 594 (#A21207, Invitrogen) at 1:1000 dilution for 1 h and
   0.1 µg mL^−1 DAPI (#D9564‐10MG, Sigma‐Aldrich) for 5 min at room
   temperature. Cells were then analyzed by Axio Imager Z2+Metafer
   fluorescence microscopy (Zeiss, Oberkochen, Germany). Cells showing
   asymmetric spindle division, multipolar division, and nuclear
   condensation were counted in at least 60 cells in each sample and
   expressed as mean ± SD%.

RNA Sequencing and Data Analysis

   Total RNA was extracted from fresh cells using Trizol (#15 596 018,
   Invitrogen) and quality was assessed via NanoDrop (Thermo Fisher, New
   York, USA). Samples with total RNA greater than 300 ng and RIN (RNA
   integrity number) > 7 were retained for RNA‐seq. MM cell lines
   including HEK293‐Ctrl, HEK293‐TP53^KO, and HEK293‐TP53^KO/NEK2 OE were
   used for RNA‐seq. The total RNA sample was used to construct
   complementary DNA libraries according to the TruSeq RNA Library Prep
   Kit (Illumina, CA, USA) protocol. RNA‐seq was performed on a HiSeq 2000
   (Illumina) and pair‐end sequencing data was generated. Sequence files
   for all samples used in this study have been deposited in the public
   database of National Omics Data Encyclopedia (NODE) under project
   number OEP000456, available at:
   [239]https://www.biosino.org/node/review/detail/OEV000066?code=UZQODSSG
   .

   All sequencing reads were aligned with the reference genome (GRch38)
   using HISAT2,^[ [240]^29 ^] with default options in the StringTie^[
   [241]^30 ^] RNA‐seq workflow.^[ [242]^31 ^] After the removal of
   improperly aligned reads, read count information was extracted from the
   files generated by StringTie with a provided Python script (prepDE.py).
   Finally, EdgeR^[ [243]^32 ^] in R was used to screen differentially
   expressed genes in the samples.

Mouse Xenograft Models

   All animal work was performed in accordance with the guidelines of the
   Institutional Animal Care and local veterinary office and ethics
   committee of the CSU, China (Animal experimental license,
   NO.2019sydw0146) under approved protocol.

   For Figures [244]4 and [245]8, 1 × 10^6 H929‐Ctrl, H929‐NEK2 OE,
   H929‐TP53^KO, H929‐TP53^KO/NEK2 OE, or H929‐TP53^KO/shNEK2 cells (in
   150 µL PBS) were injected subcutaneously into the abdomen of
   immunodeficient female B‐NDG (6–8 weeks old) mice without mature T
   cells, B cells, and NK cells^[ [246]^33 ^] (Biocytogen Co, Beijing,
   China). 1 × 10^6 ARP1‐Ctrl, ARP1‐TP53 OE, ARP1‐TP53/NEK2 OE, or
   ARP1‐TP53 OE/shNEK2 cells (in 150 µL PBS) were injected subcutaneously
   into the abdomen of 6–8 weeks old B‐NDG mice (Biocytogen Co, Beijing,
   China). After 14 days, mice with ≈3 × 3 mm tumor nodules were treated
   with 1 mg kg^−1 BTZ every three days until the end of experiment. Mice
   injected with ARP1‐TP53 OE/shNEK2 cells were fed with water containing
   2 mg mL^−1 doxycycline every 2 days when the size of tumor nodule was
   about 3 × 3mm.

   Tumor burdens were monitored by measuring tumor volumes every 3 days.
   Tumor volumes were calculated according to the equation V = (length ×
   width^2)/2. Expression of NEK2, p53, and Ki‐67 proteins in tumor
   nodules were detected via immunohistochemistry.

Immunohistochemistry Staining

   Tumor xenografts derived from mice in vivo were fixed in formalin for
   48 h. Then, microarray slides were cut at four microns on plus slides.
   IHC was performed using a standard streptavidin‐biotin‐peroxidase
   complex method as previously described. Slides were allowed to air dry
   and placed in a 55–60 °C oven for 30 min. The tissue sections were
   incubated with the primary antibodies anti‐CD138 (#10593‐1‐AP,
   Proteintech), anti‐NEK2 (D‐8, #sc‐55601, Santa Cruz Biotechnology), p53
   (#AMO183, Spectre, Xiamen, China), and Ki‐67 (#AMO383, Spectre) at a
   final dilution of 1:100 overnight at 4 °C. Assessment of the stained
   proteins was carried out by determining both the intensity (0, 1, 2, or
   3) and extent of staining (0, 0%; 1, <10%; 2, 10–50%; 3, >50%) as
   previously described.^[ [247]^34 ^]

Accession Numbers

   Microarray data sets were deposited in the National Center for
   Biotechnology Information's Gene Expression Omnibus (GEO) database
   ([248]http://www.ncbi.nlm.nih.gov/geo) using the accession numbers
   [249]GSE2658^[ [250]^5 ^] .The MMRF CoMMpass database were based on the
   Relating Clinical Outcomes in Multiple Myeloma to Personal Assessment
   of Genetic Profile study (CoMMpassSM, [251]NCT01454297). The MMRF
   genomic data can be found on the GDC Data Portal. To request access to
   protected MMRF data, please apply to dbGaP for access to the MMRF Study
   (study accession phs000748).

Real‐Time Quantitative PCR

   For quantitative analysis of gene expression, total RNA was isolated
   with Trizol. Complementary DNA was synthesized using a first full cDNA
   transcription kit according to the manufacturer's instructions (#K1622,
   Thermo Fisher). Real‐time qPCR for human NEK2, TP53, E2F8, GAPDH, and
   other genes listed below was performed using SYBR Green Super Mixture
   Reagents (#A25742, Invitrogen) on a LightCycler 96 system (Roche). PCR
   was initiated at 95 °C for 2 min to hot‐start the DNA polymerase and
   denature the template, followed by 40 cycles of denaturing at 95 °C for
   15 s and annealing and extension at 60 °C for 1 min. The primers used
   in the qPCR assay are listed in Table [252]S10, Supporting Information.

Immunoblotting

   Cell pellets were lysed with RIPA lysis buffer and 20–40 µg protein
   from each experimental condition were subjected to sodium dodecyl
   sulfate‐polyacrylamide gel electrophoresis, which was then transferred
   onto a polyvinylidene fluoride membrane. The membranes were blocked
   with 5% nonfat dry milk in Tris‐buffered saline solution containing
   0.1% Tween 20 (TBS‐T) for 1 h at room temperature and then incubated
   with the appropriate primary antibodies overnight at 4 °C. Primary
   antibodies were diluted in 5% bovine serum albumin with TBS‐T. The
   membranes were washed with TBS‐T, followed by probing with
   species‐specific secondary HRP‐conjugated antibodies conjugated, which
   were diluted in bovine serum albumin with TBS‐T. Protein bands were
   detected using SuperSignal West Pico Chemiluminescent Substrate
   (Pierce, Rockford, IL, USA)

Methylation‐Specific PCR

   MSP was performed using our previously published protocol.^[ [253]^35
   ^] Briefly, MSP was carried out for 45 cycles using the Ex Taq Hot
   Start DNA polymerase (#RR006A, TaKaRa, Dalian, China) with 10 ng of
   sodium bisulfite‐treated DNA. MSP products (both U and M products) of
   NEK2 from partial samples, which accounted for 20% of all detected
   samples, were purified with Agarose Gel Purified System (#LS1022,
   QIAGEN, Promega) and subsequently sequenced. All sequencing reactions
   were performed by a 377 ABI PRISM DNA Sequencer at the Shanghai
   Invitrogen Company (Shanghai, China). The primers used in the MSP assay
   are listed in Table [254]S10, Supporting Information.

   For Bisulfite Genomic Sequencing, primers were designed with MethPrimer
   2.0 tools online (Table [255]S10, Supporting Information). After PCR
   amplification was performed from deaminated DNA, products were
   gel‐purified and connected to T vector. Ten clones were randomly
   selected for sequencing and analysis by the use of NCBI Nucleotide
   BLAST.

Co‐Immunoprecipitation

   For co‐IP analysis, the H929‐Ctrl and H929‐TP53^KO cell lines were
   used. All procedures followed the standard protocol previously
   reported.^[ [256]^11 , [257]^14 ^] Briefly, cells were lysed in lysis
   buffer for 40 min on ice. The lysates were incubated overnight at 4 °C
   on a rotator with 4 µg of polyclonal anti‐p53 and mouse IgG antibodies
   (Santa Cruz Biotechnology, CA, USA). 50 µL of protein A/G beads
   (Biolinkedin, Shanghai, China) were transferred to the protein‐antibody
   complexes, and immunoprecipitates were collected after 2 h incubation.
   Finally, the immunoprecipitates were resuspended in lithium dodecyl
   sulfate (LDS) sample buffer and heated for 10–12 min at 70 °C for
   analysis by LDS polyacrylamide gel‐electrophoresis, loading equal
   concentrations of protein from the original lysate, and western
   blotting with monoclonal antibodies against p53 (ABclonal, Shanghai,
   China), DNMT1, and DNMT3B (Proteintech).

Statistical Analysis

   Quantitative data are shown as means ± SD. Student's t‐test, ANOVA with
   Dunnett post‐hoc test, Chi‐square and Rank sum tests were used to
   analyze data. To analyze correlation of NEK2 and TP53 expression with
   disease progression, overall survival was measured using the
   Kaplan‐Meier method, and the log‐rank test was used for group
   comparison based on GraphPad Prism 7 software. Significance was set at
   p < 0.05.

Conflict of Interest

   The authors declare no conflict of interest.

Author Contributions

   X.F. and J.G. contributed equally to this work. W.Z., X.F., and J.G.
   conceived of and designed the experiments; X.F., J.G., Y.W., B.M.,
   Y.Z., L.Q., R.L., and J.X. performed the experiments; W.Z., X.F., J.G.,
   Z.Y., and Z.L. analyzed the data; X.L., G.A., G.R., L.Q., and J.Z.
   provided critical materials; W.Z., X.F., and J.G. wrote the manuscript
   and all authors edited the manuscript.

Supporting information

   Supporting Information
   [258]Click here for additional data file.^ (2.7MB, pdf)

   Supplemental Table 1
   [259]Click here for additional data file.^ (19.2KB, docx)

   Supplemental Table 2
   [260]Click here for additional data file.^ (12.7KB, docx)

   Supplemental Table 3
   [261]Click here for additional data file.^ (297.5KB, xls)

   Supplemental Table 4
   [262]Click here for additional data file.^ (73KB, xls)

   Supplemental Table 5
   [263]Click here for additional data file.^ (229.3KB, xlsx)

   Supplemental Table 6
   [264]Click here for additional data file.^ (223.7KB, xlsx)

   Supplemental Table 7
   [265]Click here for additional data file.^ (47.5KB, xls)

   Supplemental Table 8
   [266]Click here for additional data file.^ (52.5KB, xls)

   Supplemental Table 9
   [267]Click here for additional data file.^ (84KB, xlsx)

   Supplemental Table 10
   [268]Click here for additional data file.^ (20.8KB, docx)

   Supplemental Figure 2
   [269]Click here for additional data file.^ (21KB, xls)

   Supplemental materials‐and‐methods 2
   [270]Click here for additional data file.^ (1MB, pdf)

   Supplemental materials‐and‐methods 3
   [271]Click here for additional data file.^ (128.4KB, docx)

   Supplemental materials‐and‐methods 4
   [272]Click here for additional data file.^ (1.2MB, docx)

Acknowledgements