TRIM22 Activates MAPK Signaling by Regulating the Transcription of SPHK2 and Accelerating the Degradation of Raf-1 in Glioblastoma

Tripartite motif (TRIM) 22 and mitogen-activated protein kinase (MAPK) signaling pathways play a critical role in tumor growth and therapeutic resistance of glioblastoma (GBM) respectively. However, the molecular mechanism between TRIM22 and MAPK signaling remains to be claried. We constructed TRIM22 knockout cell lines for molecular biology experiments, detected potential DNA fragments binding to TRIM22 by ChIP-Seq technology, and veried the sequencing results by ChIP-qPCR and CUT&Tag technology. In addition, we constructed different TRIM22 mutants to detect the binding of proteins in MAPK signaling pathway. Finally, the therapeutic effect was veried in NOD/SCID mice. The difference between the two groups of data conforming to the normal distribution was tested by Student t-test. by Western blot in U251MG and A172. c Western blot analysis of Raf-1 protein in different modied cells treated with cycloheximide (CHX; 25 μg/mL) for 0, 8, 16, and 24 h. d ubiquitination of Raf-1 in an in vitro assay. e In vivo ubiquitination assay of Raf-1 in modied U251MG, A172 and TJ905. Co-IP was used to detect the exogenous binding of TRIM22 and Raf-1 in 293T cells transfected with Flag-TRIM22 and HA-Raf-1. using


Results
Here, we found for the rst time that TRIM22 acts as a transcription factor in the nucleus and binds to exon 2 of the transcript (NM_001204160) of SPHK2 gene to regulate its expression by ChIP-Seq technology, thus indirectly affecting the downstream MAPK signaling pathway. Knockout of TRIM22 using Cas9-sgRNAs resulted in decreased mRNA level of SPHK2 in GBM cells, while overexpression of TRIM22 enhanced it. The ERK1/2 driven luciferase reporter construct identi ed TRIM22 as a potential activator of MAPK signaling. Knockout and overexpression of TRIM22 regulate the inhibition and activation of MAPK signaling through its RING-nger domain. Co-immunoprecipitation demonstrated that TRIM22 bound to the negative regulator Raf-1 of MAPK signaling and accelerated its degradation by inducing K48-linked ubiquitination. The combination of the two is related to the CC domain and SPRY domain of TRIM22 and the C1D domain of Raf-1. TRIM22 also forms a complex with the downstream regulator ERK1/2 of MAPK and promotes K63-linked ubiquitination, resulting in the phosphorylation of ERK1/2. In addition, in vitro and in situ xenotransplantation models, SPHK2 inhibitor (K145), ERK1/2 inhibitor (Selumetinib) and non-phosphorylated mutant Raf-1 S338A inhibited the growth promoting properties of TRIM22 in GBM cell line.

Conclusions
In conclusion, our study shows that TRIM22 regulates SPHK2 transcription as a transcription factor, indirectly affects MAPK signaling, and activates MAPK signaling through post-translational modi cation of two critical regulators of MAPK signaling in GBM cells.

Page 3/23
Background Mitogen activated protein kinase (MAPK) cascade regulates a variety of cell biological processes, such as proliferation, differentiation, apoptosis and key signaling pathways of stress response under normal and pathological conditions [1][2][3]. Studies have shown that MAPK is highly active in human glioblastoma (GBM) and plays a central role in many other active carcinogenic pathways [4][5][6]. For example, MAPK and PI3K signaling pathways cooperate to cause the pathogenesis of GBM [7]. However, the mutation or ampli cation of MAPK signaling subunits is rare in GBM which indicates that the abnormal activation of MAPK signaling may be due to the deregulation of pathway or oncogene.
Ubiquitination is one of the most common and important types of post-translational modi cation [8,9].
The most well studied polyubiquitin chain types include lysine 48 (K48) and lysine 63 (K63) linked. K48linked polyubiquitin chains mainly target degraded proteins, while K63-linked polyubiquitin chains regulate kinase activity, signal transduction and endocytosis [10,11]. Post translational modi cations, especially ubiquitin modi cations, have become key participants in MAPK activation. For instance, LZTR1 promotes the polyubiquitination and degradation of Ras through the ubiquitin proteasome pathway, resulting in the inhibition of Ras/MAPK signaling [12].
The tripartite motif (TRIM) proteins, which regulate MAPK activity through ubiquitination, are becoming key regulators in the development of a variety of cancers, including GBM [9]. TRIM-FLMN protein TRIM45 directly interacts with RACK1 and negatively regulates PKC mediated MAPK signaling pathway [13]. GPER reduces the Bim protein through MAPK/Erk-TRIM2 signal axis and promotes the resistance of ER+ breast cancer cells to tamoxifen [14]. TRIM22 is rarely studied in most human carcinogenesis, and until now the relationship between TRIM22 and MAPK signaling pathway has not been reported. It is reported that TRIM22 regulates protein degradation and activity as an E3 ubiquitin ligase in the cytoplasm [15], but more and more studies have found that TRIM22 is mainly located in the nucleus [16]. Using ChIP-Seq, CUT&Tag-qPCR, CRISP/Cas9 technologies, our study found that TRIM22 acts as a transcription factor in the nucleus to regulate the expression of sphingosine kinase 2 (SPHK2), and then indirectly activate the MAPK pathway downstream of SPHK2. In addition, our study found that TRIM22, as a classical E3 ubiquitin ligase, promotes K48-linked ubiquitination of Raf-1 and K63-linked ubiquitination of ERK1/2, thus directly affecting the MAPK pathway. At present, SPHK2 and MAPK inhibitors have been widely used in clinical trials and achieved good curative effects [17,18]. Our discovery that TRIM22 directly and indirectly regulates SPHK2/MAPK signaling pathway provides a new idea for GBM treatment.

Lentivirus and plasmid transfection
TRIM22 gene was knocked out by CRISPR/cas9 technology. The Cas9 and single-guide RNA (sgRNAs) lentiviruses were designed and constructed by Genechem Co., Ltd (Shanghai, China). The sequences of sgRNAs used are listed in Additional le 1: Table 1. Cell lines were screened with puromycin.

Subcellular fractionation
Nuclear and cytoplasmic fractions from U251MG, A172, TJ905, H4, P1 and P2 were isolated using Nuclear and Cytoplasmic Extraction Kit purchased from Beyotime (Shanghai, China), according to the manufacturer's instructions. Subcellular distribution of proteins was determined using western blot analysis. GAPDH and Histone H3 served as loading controls for cytosolic and nuclear fractions, respectively.

Cleavage Under Targets and Tagmentation (CUT&Tag)
CUT&Tag was carried out according to the instructions of the following kits: Hyperactive. In-Situ ChIP Chromatin immunoprecipitation sequencing and qPCR (ChIP-Seq and ChIP-qPCR) ChIP-seq and ChIP-qPCR was completed with the assistance of Kangchen Biotechnology Co., Ltd (Shanghai, China). The detailed steps of the experiment are described in the Additional le 1: materials and Methods.
Pink box represents SPHK2 gene and gray box is the 2000 BP promoter region upstream of SPHK2. Green box represents a transcript of SPHK2 NM_001243876, orange box represents another transcript NM_001204160, blue box represents the position of peak in ChIP-Seq results. The above gene sequence and transcript sequence information refer to GRCh37/hg19 genome version information, and the sequence is obtained from UCSC database (UCSC Genome Browser http://genome.ucsc.edu/index.html).
NM_001243876 has 6 exons and 5 introns in total, and peak is located in its intron 2 region. NM_001204160 has 6 exons and 5 introns in total. Peak is located in most of the sequences of exon 2 and part of the sequences of intron 2. According to the Peak site, NM_001204160 was selected as the study subject. Detailed site information is shown in Additional le 1: Table 3.

Dual-Luciferase reporter assay
The ERK1/2 re y-luciferase and renilla reporter plasmids were transfected into each modi ed cell lines. The cells were cultured for 24 hours after plasmid transfection and the uorescence intensity of each treatment group was detected by Dual-Luciferase Reporter Assay Kit (Promega).

Immuno uorescence (IF), immunohistochemistry (IHC), Co-immunoprecipitation (Co-IP) and Western blotting (WB)
IF, IHC and WB were performed as previously described [19,20] and all antibodies used are described in Additional le 1: Table 4. To assess in vivo ubiquitination, modi ed cells were treated with 20 μM MG132 (Apexbio; Houston, TX, USA) for 12 h before lysis, followed by western blot analysis and Co-IP." Yeast one hybrid Different modi ed potential DNA-binding proteins, the Prey, are expressed as fusions to the GAL4 activation domain in pGADT7-Rec2. The different modi ed target DNA sequence, or Bait Sequence, is cloned into pHIS2.1 reporter vector as tandem repeats three times. The two plasmids were transferred into Yeast strain Y187 and cultured in selected media. 3-amino-1,2,4-triazole (3-AT) is used to inhibit low levels of His3p expressed in the absence of an activating prey protein.

Animal studies
Male four-week-old NOD/SCID mice were purchased from Shenzhen Huafukang Bioscience Co., Inc., (Shenzhen, China) and randomly divided into 10 mice in each group. Different modi ed GBM cells expressing luciferase were implanted into the brain of mice by stereotactic injection (2 mm posterior and 1 mm lateral of the bregma). The tumor cells (3X10 6 /5ul PBS) were injected into the mouse brain (depth: 2.8mm) using a micro-syringe within 2 minutes. After the operation, the skull was sealed with bone wax, and the incision was sutured and 0.5ml glucose was injected into the abdominal cavity. Fluorescence bioluminescence detection equipment (IVIS Lumina III) was used to detect tumor formation at 7, 15 and day according to the weight of mice until the end of the experiment.

Statistical analysis
All experiments were performed at least three independent biological replicates. The data are expressed as mean ± standard deviation. Student t-test was used for statistical analysis between the two groups, and one-way ANOVA was used between multiple groups. The F-test is used to check whether the data conform to the normal distribution. SPSS for Mac software (26.0) was used for statistical analysis. P < 0.05 was considered statistically signi cant.

Results
TRIM22 acts as a transcription factor to regulate the transcription of SPHK2 in the nucleus We collected glioma samples of different grades from Xijing Hospital for detection (NBT; n = 10, WHO II; n=92, WHO III; n=65, WHO IV; n=70). We found that TRIM22 was highly expressed in GBM (Fig. 1a) and mainly located in the nucleus (Fig. 1b). In addition, cell lines U251MG, A172, TJ905, H4 with high expression of TRIM22 and primary GBM cell lines P1 and P2 were selected. Western blot (Fig. 1c) and immuno uorescence (Fig. 1d) showed that the expression of TRIM22 was mainly located in the nucleus.
To study the role of TRIM22 in the nucleus, we used ChIP-Seq and enrichment analysis to nd that TRIM22 binds to the DNA fragment of VEGF pathway related genes (MAPKAPK3, NFATC2, PP3CC, PRKCA, SPHK1, SPHK2), which may play the role of transcription factors (Fig. 1e, Additional le 1: Fig.  S1a). In addition, ChIP-qPCR (Fig. 1f) in 293T and Yeast one hybrid (Fig. 1g) in vitro showed that TRIM22 could bind to the gene fragment of SPHK2. These results show that TRIM22 acts as a transcription factor to regulate the transcription of SPHK2 in the nucleus.

TRIM22 binds to the transcriptional region of SPHK2
To study the detailed binding sites of TRIM22 and SPHK2, according to the information of multiple transcripts of SPHK2, the transcription start site (TSS) and the position of peak in the sequencing results, four pairs of different primers were designed to explore potential binding sites (Additional le 1: S Table 3, Additional le 1: Fig. S1b). In 293T cells, ChIP-qPCR showed that TRIM22 bound most closely to the site interval of peak (Additional le 1: Fig. S1c). In addition, different TRIM22 truncates (Flag-TRIM22-ΔRING, Flag-TRIM22-ΔB-Box, Flag-TRIM22-ΔCC and Flag-TRIM22-ΔSPRY) were transfected into U251MG, A172, TJ905, H4 and two primary cells P1 and P2 for CUT&Tag. We were surprised to nd that both TRIM22 ( Fig. 2a) and its truncate (Additional le 1: Fig. S2a-b, Additional le 1: Fig. S3a-b) bind to SPHK2-4. In Yeast one hybrid, three repeated SPHK2-4 were used as bait sequence, and we also obtained the results consistent with those in vivo (Fig. 2b). To further explore how TRIM22 regulates the expression of SPHK2, Using CRISPR/Cas9 technology, we constructed TRIM22-knockout (KO) cell line (Additional le 1: Fig.  S1d). qPCR results showed that knockout of TRIM22 reduced the mRNA expression of SPHK2, while overexpression of TRIM22 could promote its expression in four cell lines and two primary cells (Fig. 2c).
These results suggest that TRIM22 protein positively regulates the expression of SPHK2.
TRIM22 positively regulates SPHK2/MAPK signaling pathway through K48-linked ubiquitination of Raf-1 Our results show that TRIM22 binds transcripts of SPHK2 and positively regulates SPHK2 expression. The enrichment results of ChIP-Seq also showed that TRIM22 may affect the VEGF pathway, and although both MAKP and PI3K/AKT pathways are in the VEGF pathway, only the MAPK pathway is downstream of SPHK2. (Additional le 1: Fig. S1a). To explore the effect of TRIM22 on SPHK2/MAPK pathway, using an ERK1/2 dependent transcriptional reporter containing three copies of an ERK1/2 response element located upstream of luciferase. Dual luciferase reporter assay showed that knockout of TRIM22 inhibited the activity of MAPK signal (Fig. 3a). Knockout of TRIM22 inhibited the expression of SPHK2/MAPK pathway proteins, including SPHK2, Ras, P-Raf-1 Ser338 , P-MEK1/2 Ser217/221 and P-ERK1/2 Thr202/Tyr204 . Although the expression of phosphorylated protein was down regulated after knockout of TRIM22, there was no signi cant change in the expression of MEK1/2 and ERK1/2 (Fig. 3b, Additional le 1: Fig. S4a). Interestingly, the expression of Raf-1 was up-regulated but the mRNA level did not change (Fig. 3c). In addition, knockout of TRIM22 attenuated the degradation of Raf-1 (Fig. 3d, Additional le 1: Fig. S4b) and reduces K48-linked ubiquitination of Raf-1 (Fig. 3e, Additional le 1: Fig.  S4c). These results suggest that TRIM22 regulates SPHK2/MAPK pathway and TRIM22 deletion attenuates K48-linked ubiquitination of Raf-1.

SPHK2 and MAPK inhibitors attenuated TRIM22-promoted proliferation of GBM
It has been reported that TRIM22 is highly expressed in GBM and promotes the proliferation of GBM [21]. To explore whether TRIM22 promotes GBM proliferation through SPHK2/MAPK signaling pathway, K145, SPHK2 inhibitor and Selumetinib, a highly potent non-ATP competitive MEK1/2 inhibitor, were used for in vivo and in vitro experiments. Overexpression of TRIM22 activated MAPK signal, while the uorescence intensity decreased signi cantly after treatment with K145 and Selumetinib for 24 hours (Fig. 4a, Additional le 1: Fig. S5a). K145 and Selumetinib also inhibited the growth rate of cells promoted by TRIM22 (Fig. 4b, Additional le 1: Fig. S5b) and the expression rate of Ki67 (Fig. 4c, Additional le 1: Fig.  S5c). In vivo, consistent with previous reports, A172 cell line and primary cell line P1 overexpressing TRIM22 accelerated the proliferation of GBM 15 days after tumor implantation. The tumor growth rate of NOD/SCID mice injected with K145 and Selumetinib daily intraperitoneally was slower (Fig. 4d) Fig. 4e). These results illustrate TRIM22 promotes proliferation of GBM through SPHK2/MAPK signaling pathway.
TRIM22 ubiquitinates Raf-1 TRIM22 contains four structures: RING, B-Box, CC and SPRY. To explore which structure is critical for TRIM22-mediated activation of MAPK signaling, we tested TRIM22-ΔRING, TRIM22-ΔB-Box, TRIM22-ΔCC and TRIM22-ΔSPRY in ERK1/2 luciferase reporter assays compared with TRIM22-FL and EV groups. The results showed that compared with TRIM22-FL group, only TRIM22-ΔRING group signi cantly reduced luciferase activity (Fig. 5a, Additional le 1: Fig. S6a). Furthermore, overexpression of TRIM22 increased the expression of key proteins in SPHK2/MAPK pathway, while the lack of RING domain inhibited this change. The change trend of Raf-1 protein was opposite to that of other proteins (Fig. 5b, Additional le 1: Fig. S6b). In CHX assay, overexpression of TRIM22 accelerated the degradation of Raf-1 and the results were similar in TRIM22-ΔRING group and Flag-EV group (Fig. 5c, Additional le 1: Fig. S6c). Both in vitro (Fig. 5d) and in vivo (Fig. 5e, Additional le 1: Fig. S6d), overexpression of TRIM22 induced endogenous K48-linked ubiquitination of Raf-1, while the RING domain deletion had no effect. These results illustrate that the change of Raf-1 is caused by ubiquitin degradation pathway induced by TRIM22.
To investigate whether TRIM22 binds directly to Raf-1. We performed Co-IP experiments using 293T cells transfected with Flag-TRIM22 and HA-Raf-1 and the results show that Flag and HA bring down each other (Fig. 5f). Furthermore, the physical binding of endogenous TRIM22 and Raf-1 was also con rmed in four cell lines and two primary cells (Fig. 5g, Additional le 1: Fig. S6e). To further study which domain of the two proteins plays a crucial role in binding, a series of deletion mutants of TRIM22 and Raf-1 were constructed and transfected into 293T cells. The CC and SPRY domains of TRIM22 are the necessary conditions for bringing down HA-Raf-1, while the C1d domain of Raf-1 is the necessary condition for bringing down f Flag-TRIM22 (Fig. 5h). Thus, TRIM22 promotes proteasomal-mediated degradation of Raf-1 via its RING domain, a negative regulator of MAPK signaling.

Raf-1 mediates the proliferation of GBM promoted by TRIM22
Phosphorylation at S338 site activates Raf-1 protein [22]. To further study the role of Raf-1 in TRIM22 promoted proliferation of GBM, we constructed a mutant of phosphorylated Raf-1, and the serine at site 338 was mutated to alanine (Raf-1 S338A ). The plasmid was packaged with lentivirus and transfected into cells for stable expression. Luciferase Report showed that Raf-1 S338A inhibited the activation of MAPK pathway caused by overexpression of TRIM22 (Fig. 6a, Additional le 1: Fig. S7a). The growth of GBM cells (Fig. 6b, Additional le 1: Fig. S7b) and positive rate of Ki67 (Fig. 6c, Additional le 1: Fig. S7c) promoted by overexpression of TRIM22 was also inhibited by Raf-1 S338A . Moreover, in vivo, compared with Flag-TRIM22 group, Raf-1 S338A group showed signi cant inhibitory effect at 15 days after intracranial tumor implantation (Fig. 6d) Fig. 6e). These results indicated that Raf-1 is a key effector in TRIM22promoted growth of GBM cell populations in vitro and in vivo.
TRIM22 promotes K63-linked ubiquitination of ERK1/2 TRIM22 knockout decreased the expression of key proteins of SPHK2/MAPK pathway, including SPHK2, Ras and phosphorylated MEK1/2 and ERK1/2 (Fig. 3b, Additional le 1: Fig. S4a), while overexpression of TRIM22 promoted its up regulation (Fig. 5b, Additional le 1: Fig. S6b). The parallel change trend of these proteins and TRIM22 makes us speculate whether TRIM22 may directly regulate the activation of these proteins. To explore whether TRIM22 directly regulates these key proteins, anti-TRIM22 antibody was used for Co-IP experiment. We found that ERK1/2 was associated with TRIM22 in four GBM cell lines and two primary GBM cells (Fig. 7a, Additional le 1: Fig. S8a). The same results were obtained by Co-IP with anti-ERK1/2 antibody (Fig. 7b, Additional le 1: Fig. S8b). Furthermore, we found that K63 ubiquitination levels of ERK1/2 paralleled TRIM22 protein levels. TRIM22 knockout reduced K63-linked ubiquitination levels of ERK1/2 (Fig. 7c, Additional le 1: Fig. S8c) and overexpression TRIM22 promotes it. The speci c mechanism may also be related to the RING domain of TRIM22 (Fig. 7d, Additional le 1: Fig. S8d). These results suggested that TRIM22 might regulate activation of MAPK through ERK1/2 activation.

Discussion
TRIM22 plays a key role as E3 ubiquitin ligase in tumors. Our study found for the rst time that TRIM22 regulates SPHK2 gene transcription as a transcription factor in the nucleus. We found that the main binding gene sites of TRIM22 are located in intron 2 of NM_001243876 transcript and exon 2 of NM_001204160 transcript. In addition, it is found that TRIM22 positively regulates the expression of SPHK2 gene and indirectly affects the activation of MAPK signaling. TRIM22 also directly binds to MAPK pathway core proteins Raf-1 and ERK1/2, regulates degradation of Raf-1 through K48-linked ubiquitination, and regulates activation of ERK1/2 by K63-linked ubiquitination. These results illustrate that TRIM22 directly or indirectly regulates MAPK signaling, thereby promoting GBM proliferation (Fig. 7e).
TRIM22 contains four domains, RING domain, B-Box domain, CC domain and SPRY domain. It has been reported that the RING domain is involved in the function of ubiquitination as a classical domain of the TRIM family [15]. The speci city of the Ub-binding system arises from the direct binding of E3 ligases to their substrates. Our ndings that TRIM22 binds Raf-1 and ERK1/2 and promotes K48-linked and K63linked ubiquitination respectively, are consistent with the functions reported in the literature. In addition, we found that deletion of the RING domain signi cantly inhibited SPHK2/MAPK signaling activation by overexpressing TRIM22, suggesting that the RING domain of TRIM22 may regulate signaling activation in addition to its role as an E3 ubiquitin ligase. Although the RING domain plays a functionally important role and has also been shown to be the main domain promoting proliferation in GBM, the binding of TRIM22 to Raf-1 does not appear to be related to the RING domain. Studies have reported that TRIM22 binds IκBα associated with its B-Box domain, CC domain and SPRY domain [21], while our study found that the binding of TRIM22 and Raf-1 is associated with the CC domain and SPRY domain. On the other hand, we also found that TRIM22 is independent of its four domains when it acts as a transcription factor to regulate SPHK2 transcription. Polymorphisms in the function of this domain of TRIM22 make it necessary to further investigate which region of TRIM22 may be associated with the development of human gliomas.
Raf-1 is a major effector recruited through the GTP-binding protein Ras and activates the MEK-MAP kinase pathway [23]. Activation of Raf-1 involves phosphorylation of multiple activation sites including Ser338, Tyr341, Thr491, Ser494, Ser497, and Ser499 [22]. The ratio of Raf-1 S338 /Raf-1 illustrates the activation of Raf-1, and our study found that the synergistic effect of TRIM22 on both Raf-1 phosphorylation and degradation prompted Raf activation. In addition, the p44/42 MAPK (ERK1/2) signaling pathway is activated in response to a variety of extracellular stimuli. MEK1 and MEK2 activate p44 and p42 by phosphorylating the activation loop residues Thr202/Tyr204 and Thr185/Tyr187, respectively [24,25]. It is believed that ERK1/2 is a key target for cancer diagnosis and treatment [26]. We demonstrate that TRIM22 targets Raf-1 for proteasome-mediated degradation and further enhances MAPK signaling activation by modifying K63-linked polyubiquitin chains on ERK1/2. However, in addition to K48 and K63, K6, K11, K27, K29 and K33-linked polyubiquitin chains also play important roles in the regulation of Raf-1 and ERK1/2, but whether TRIM22 regulates ubiquitination and other phosphorylation sites at these sites of Raf-1 and ERK1/2 also needs further study.
K145 is a selective inhibitor of SPHK2 and shows excellent antitumor activity in vivo and in vitro [27,28]. Liu's group [29] found that K145 can act as a dual pathway inhibitor and inhibit both ART and AKT signaling pathway phosphorylation while inhibiting SPHK2, thereby achieving anti-proliferative and apoptotic effects in leukemia U937 cells. Selumetinib is a highly potent and selective, non-ATPcompetitive MEK1/2 inhibitor [30], and the results from National Cancer Institute showed a good therapeutic effect [31] in a clinical trial of Selumetinib in the treatment of low-grade astroglioma and metastatic low-grade glioma (NCT04166409) conducted on January 3, 2020. Our study found that K145 and Selumetinib could signi cantly inhibit the proliferation of GBM promoted by overexpression of TRIM22 both in vivo and in vitro.

Conclusion
In conclusion, our study reveals for the rst time an entirely new regulatory mechanism of MAPK activation, nding that TRIM22, which is highly expressed in GBM, regulates the activation of MAPK signaling by both direct (E3 ubiquitin ligase) and indirect (transcription factor) means. TRIM22 induces SPHK2/MAPK signaling activation in GBM, driving tumor growth and progression. Finally, our study de nes TRIM22 as a candidate therapeutic target. Drug inhibition of its E3 ligase activity or TRIM22/SPHK2/MAPK interaction may provide a promising strategy for treating GBM.

Declarations Ethics Approval and Consent to Participate
All primary glioma tissue samples (WHO II; n=92, WHO III; n=65, WHO IV; n=70) and nonneoplastic brain tissue samples (NBT; n=10) were obtained from the Department of Neurosurgery at Xijing Hospital. Ethical approval was obtained from Xijing Hospital Research Ethics Committee, and written informed consent was obtained from each patient. All experimental procedures were approved by the Institutional Animal Care and Use Committee of Fourth Military Medical University. The study was performed in accordance with the Declaration of Helsinki.

Consent for publication
Not applicable.

Availability of data and materials
The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.

Competing interests
The authors declare that they have no competing interests Author's contributions FXW, DYN and SK designed the study, performed the experiments, and prepared the manuscript, and they contributed equally to this work. WJL, GQD, WL, WXQ and LWH were involved in experiment performance and data collection. FXW had written the original draft. FZ were responsible for the supervision of the entire project and were involved in the study design, data interpretation and manuscript preparation. All authors read and approved the nal manuscript. . c Expression of TRIM22 in cytoplasm and nucleus of U251MG, A172, TJ905, H4 and primary GBM cells P1 and P2. d Localization of TRIM22 in six cells. Scale bar: 50 μm. e After 293T cells were transfected with Flag-TRIM22 overexpression plasmid, ChIP-Seq was carried out to analyze the differential genes. f ChIP-qPCR experiment was used to verify the results of ChIP-Seq. g Yeast one hybrid experiment veri ed the binding of TRIM22 and SPHK2 in vitro. (All data represent mean ± SEM n ≥ 3). Student's t test: *P < 0.05. Figure 2 TRIM22 binds to the transcriptional region of SPHK2 and positively regulates its transcription. a In U251MG, A172, TJ905, H4 and primary GBM cells P1 and P2, different binding sites of TRIM22 and SPHK2 (1-4) were detected by CUT&Tag. b In vitro yeast one hybrid was used to detect the binding of different TRIM22 truncated mutants (Flag-TRIM22-ΔRING, Flag-TRIM22-ΔB-Box, Flag-TRIM22-ΔCC and Flag-TRIM22-ΔSPRY) to SPHK2-4. c Effect of knockout or overexpression of TRIM22 on mRNA of SPHK2.
(All data represent mean ± SEM n ≥ 3). Student's t test: *P < 0.05. Figure 3 TRIM22 delete inhibit SPHK2/MAPK signaling via K48-linked ubiquitination of Raf-1. a Luciferase activity from U251MG, A172, TJ905, H4 and primary GBM cells P1 and P2 transfected with sgRNA-TRIM22, along with a reporter plasmid carrying the ERK1/2 promoter relative to negative control. b The expression of SPHK2/MAPK pathway core protein was detected by Western blot. c The mRNA level of Raf-1 was detected by qPCR. d Western blot analysis of Raf-1 protein in TRIM22-KO cells treated with cycloheximide (CHX; 25 μg/mL) for 0, 8, 16, and 24 h. e IP experiment was used to detect the endogenous K48-linked ubiquitination of Raf-1 after TRIM22 knockout. (All data represent mean ± SEM n ≥ 3). Student's t test: *P < 0.05. SPHK2/MAPK pathway regulates the proliferation of GBM in vivo and in vitro. a Luciferase activity from U251MG and A172 treated with K145 and Selumetinib, along with a reporter plasmid carrying the ERK1/2 promoter relative to negative control. b Growth curves generated using cell counting over 72 h for the cells indicated. c Representative images and quanti cation of Ki-67 immuno uorescence staining from modi ed U251MG and A172 cells. (5µM K145 and 10nM for 24 hours). Scale bar: 50 μm. d Representative images and quanti cation of in vivo imaging. e Kaplan-Meier survival analysis and Logrank test performed with survival data from indicated groups. (All data represent mean ± SEM n ≥ 3).
Student's t test: *P < 0.05. Figure 5 TRIM22 binds to Raf-1 and regulates SPHK2/MAPK signaling through its RING domain. a Luciferase activity from U251MG and A172 transfected with different TRIM22 truncation mutants. b The effects of different TRIM22 truncation mutants on the core protein of SPHK2/MAPK pathway were detected by Western blot in U251MG and A172. c Western blot analysis of Raf-1 protein in different modi ed cells treated with cycloheximide (CHX; 25 μg/mL) for 0, 8, 16, and 24 h. d ubiquitination of Raf-1 in an in vitro assay. e In vivo ubiquitination assay of Raf-1 in modi ed U251MG, A172 and TJ905. f Co-IP was used to detect the exogenous binding of TRIM22 and Raf-1 in 293T cells transfected with Flag-TRIM22 and HA-Raf-1. g exogenous binding of TRIM22 and Raf-1 in U251MG and A172 using anti-TRIM22 and anti-Raf-1 antibodies. h Schematic representation of wild-type TRIM22 and Raf-1 and the indicated deletion mutants. Western blot analysis of Co-IPs in 293T cells transfected with Flag-TRIM22/HA-Raf-1 alone or together with indicated HA-Raf-1/ Flag-TRIM22 constructs. (All data represent mean ± SEM n ≥ 3). Student's t test: *P < 0.05. Raf-1 regulates TRIM22-induced GBM proliferation in vivo and in vitro. a Luciferase activity from U251MG and A172 transfected with Flag-TRIM22 and Raf-1S338A, along with a reporter plasmid carrying the ERK1/2 promoter relative to negative control. b Growth curves generated using cell counting over 72 h for the cells indicated. c Representative images and quanti cation of Ki-67 immuno uorescence staining from modi ed U251MG and A172 cells. Scale bar: 50 μm. d Representative images and quanti cation of Page 22/23 in vivo imaging. e Kaplan-Meier survival analysis and Log-rank test performed with survival data from indicated groups. (All data represent mean ± SEM n ≥ 3). Student's t test: *P < 0.05, ****P < 0.0001.

Figure 7
TRIM22 promotes K63-linked ubiquitination of ERK1/2 through its RING domain. a Anti-TRIM22 antibody was used to detect the binding of TRIM22 to SPHK2/MAPK pathway core protein in U251MG, A172 and TJ905. b Association of TRIM22 with ERK1/2 in U251MG, A172 and TJ905. c-d Western blot analysis for