Abstract

   The transmembrane and coiled‐coil domains 3 (TMCO3) are highly
   expressed in many tumors. However, the underlying mechanisms governing
   the way in which TMCO3 affects the progression of hepatocellular
   carcinoma (HCC) remain unclear. This study screens out the molecule
   TMCO3 with high N6‐methyladenosine (m^6A) modification level in tumor
   samples compared to the adjacent non‐cancerous tissues of three pairs
   of HCC patients through Methylated RNA Immunoprecipitation Sequencing
   (MeRIP‐seq) and RNA sequencing (RNA‐seq). Subsequently, the oncogenic
   effect of TMCO3 in HCC is verified through in vivo and in vitro
   experiments. AlkB Homolog 5 (ALKBH5), an m^6A demethylase of TMCO3 is
   then screened out. The following experiments demonstrate that TMCO3 can
   activate AKT directly through the Phosphatidylinositol‐3–Kinase (PI3K)
   pathway, thus promoting the progression of HCC. Meanwhile, the
   phosphorylation site on TMCO3: the 85^th amino acid‐serine, and
   mutation of this site can directly impair the activity and membrane
   translocation of AKT is found. Finally, the carcinogenic effect of
   TMCO3 is further elucidated in HCC through the orthotopic treatment
   model and the hydrodynamic tail vein injection treatment model. The
   findings can provide a potential target for targeted AKT treatment in
   patients with HCC and verify a possible prognostic marker in HCC.

   Keywords: AKT, HCC, m^6A, MK2206, TMCO3
     __________________________________________________________________

   TMCO3 as a downstream target of ALKBH5 is found, which regulates its
   expression in an m^6A‐dependent manner, and then a new phosphorylation
   site on TMCO3 is identified and determines the specific mechanism of
   TMCO3 with AKT. Moreover, the combined treatment of targeting the TMCO3
   and inhibition of AKT in mice models with HCC achieves great results.

   graphic file with name ADVS-12-2504187-g005.jpg

1. Introduction

   HCC as the most common type of liver cancer is the fourth leading cause
   of cancer‐related death worldwide.^[ [40]^1 ^] The high mortality rate
   of HCC is mainly due to late‐stage HCC and a high rate of metastasis or
   recurrence. Despite the advanced surgical techniques currently
   available, there are still some patients with late‐stage HCC or
   metastasis that are difficult to cure clinically, necessitating the
   importance of investigating the molecular mechanisms of HCC
   progression.^[ [41]^2 ^]

   TMCO3 is a member of monovalent cation: proton antiporter 2 (CPA2)
   family of transporter proteins.^[ [42]^3 ^] Members of this family
   typically couple the export of monovalent cations, such as potassium or
   sodium, to the import of protons across cellular membranes.^[ [43]^4 ^]
   Members of this family have been poorly studied in cancer area, and
   tumor research on TMCO3 have been negligible. Hence, it is particularly
   important and urgent to explore the mechanism of TMCO3 in HCC.

   RNA modifications have always occupied a significant position in the
   field of epigenetics, and the diversity and extensiveness of its
   modification types, especially research of the mechanism in tumors, are
   particularly important.^[ [44]^5 ^] m^6A plays a crucial role in this
   process. As the most abundant type of mRNA modifications, m^6A
   modifications regulated by “writers” (METTL3/METTL14/WTAP), “erasers”
   (FTO/ALKBH5), and “readers” (YTHDF1‐3/YTHDC1‐2) that respectively
   catalyze m^6A methylation and demethylation, and recognize
   m^6A‐modified RNA.^[ [45]^6 ^] To determine the regulator of TMCO3, our
   experiments confirmed that ALKBH5 is capable of demethylating the
   modification at 348^th site on TMCO3, and Insulin‐like Growth Factor 2
   mRNA‐Binding Protein 2 (IGF2BP2） can recognize and stabilize TMCO3
   mRNA.

   Activation of the PI3K/AKT pathway is dependent on the stimulation from
   cytokines such as Insulin‐like Growth Factor 1 (IGF‐1), Epidermal
   Growth Factor (EGF), and others.^[ [46]^7 ^] Activation of AKT itself
   relies on the catalysis by molecules including the PI3K subunits,
   3‐Phosphoinositide‐Dependent Protein Kinase‐1 (PDK1), and mammalian
   Target of Rapamycin Complex 2 (mTORC2).^[ [47]^8 ^] A prerequisite for
   AKT activation is its translocation from the cytoplasm to the membrane,
   where it aggregates. Following this aggregation, AKT binds to
   PtdIns(3,4,5)P[3] (PIP[3]), starting the phosphorylation of AKT through
   the activation of the T308 site by PDK1, then mTORC2 completes AKT
   activation by stimulating the S473 site.^[ [48]^9 ^] Our research
   discovered that under the stimulation of IGF‐1, TMCO3 could promote the
   membrane accumulation of AKT and participate in its activation.
   Subsequently, we found that activation is due to the stimulation of the
   85th serine residue on TMCO3, thus adding a new potential target for
   upstream regulatory molecules of AKT.

   In summary, our research marked the first discovery that TMCO3 was
   highly expressed and exhibited a high m^6A methylation level in HCC.
   The oncogenic effects of TMCO3 were ascertained through both in vivo
   and in vitro experiments, and ALKBH5 was also verified as a m^6A
   demethylase to downregulate the TMCO3 expression. Then, we found that
   TMCO3 could facilitate the membrane translocation and phosphorylation
   levels of AKT. Finally, our hypothesis was confirmed by targeting TMCO3
   knockdown and inhibiting the AKT activity in animal experiments. This
   study demonstrates a new molecule that regulates AKT via the PI3K
   pathway, suggesting that TMCO3 can serve as a novel biomarker for HCC
   and potentially provide a therapeutic target for HCC treatment.

2. Results

2.1. TMCO3 Presents a High m^6A Modification Level and High Expression Level
in HCC

   To identify potential targets that may be associated with m^6A
   modification in HCC, three pairs of tumor and adjacent non‐cancerous
   tissues from HCC patients were collected (All three patients have a
   history of chronic viral hepatitis B, and two of them are accompanied
   by liver cirrhosis) and then both MeRIP‐seq and RNA‐seq were conducted
   (Figure [49]1A). The overlap of up‐regulated genes in the results from
   both MeRIP‐seq and RNA‐seq was analyzed, revealing that 455 genes
   exhibited statistically significant differences (Figure [50]1B). Among
   these genes, we identified 10 genes that demonstrated higher expression
   in tumor tissues compared to adjacent non‐cancerous tissues, with TMCO3
   ranking at the forefront (Figure [51]1C). It is noteworthy that TMCO3
   has been seldom reported in relation to tumors, and even in the realm
   of other diseases, thus piquing our interest. Subsequently, pan‐cancer
   data from the TCGA database also revealed that in most cancer species,
   TMCO3 mRNA levels were significantly elevated in tumor tissues compared
   to adjacent non‐cancerous tissues (Figure [52]S1A, Supporting
   Information). Next, we analyzed the abundance of m^6A modification
   levels in TMCO3 across tumor and adjacent non‐cancerous tissues from
   three pairs of patients. The results suggested that the m^6A peak level
   of TMCO3 in tumor tissues was significantly higher than that in
   adjacent non‐cancerous tissues, which matched our previous overlap
   analysis (Figure [53]1D). Then, we analyzed large‐scale data from 92
   GSE datasets and observed significant alterations in TMCO3 mRNA levels
   in HCC tissues in 63 of these datasets (Figure [54]1E). Notably, in 57
   out of these datasets, the TMCO3 mRNA level was significantly elevated
   in HCC tissues, and in three representative datasets: TCGA‐LIHC,
   [55]GSE14520, and ICGC‐LIRI‐JP, TMCO3 mRNA levels were all increased in
   tumors (Figure [56]1F). Results from qRT‐PCR (quantitative
   reverse‐transcription PCR) and Western blots assay also showed that
   both TMCO3 mRNA and protein levels were significantly elevated in
   various HCC cell lines compared to normal liver cells (Figures
   [57]S1B,C, Supporting Information). High TMCO3 expression was related
   to poor overall survival in TCGA‐LIHC and ICGC‐LIRI‐JP datasets
   (Figures [58]S1D,E, Supporting Information).

Figure 1.

   Figure 1
   [59]Open in a new tab

   TMCO3 presents a high m^6A modification level and high expression level
   in HCC. A) Schematics of MeRIP‐seq and RNA‐seq were performed on tumor
   and adjacent noncancerous tissues of three patients. B) Wayne diagram
   of gene intersections with statistical differences in MeRIP‐seq and
   RNA‐seq. C) The heat map shows the top 10 differentially expressed
   genes in gene intersections. D) m^6A levels expression peak of TMCO3 in
   tumor and adjacent noncancerous tissues of three pairs of patients. E)
   RNA‐seq data from 92 GSE subsets showed the relative expression and
   p‐value of TMCO3 in tumor and adjacent noncancerous tissues. F) TMCO3
   expression levels in tumor and adjacent noncancerous tissues in the
   three classic GSE subsets (TCGA‐LIHC, [60]GSE14520, ICGC‐LIRI‐JP).

   These data suggested that TMCO3 showed a high m^6A modification level
   and was highly expressed in HCC cell lines and HCC patients.

2.2. TMCO3 Promotes HCC Proliferation and Metastasis In Vivo and In Vitro

   To delve deeper into the function of TMCO3 in HCC, we hydrodynamically
   co‐injected the c‐Met and N90‐β‐catenin constructs together with
   PT3‐Ctrl/TMCO3(Mouse) plasmids in C57BL/6 mice (Figure [61]2A). Mice
   were closely observed and humanely euthanized upon developing
   substantial abdominal tumors and reaching a state of severe ill health.
   Interestingly, we observed that in the 8th week, mice in the
   PT3‐TMCO3(Mouse) group showed mortality, and these mice presented a
   more significant liver tumor burden compared to the control group
   (Figure [62]2B). This was accompanied by increased liver‐to‐body weight
   ratio and death rate in PT3‐TMCO3(Mouse) group (Figure [63]S2A–C,
   Supporting Information). Subsequently, we constructed stable cell lines
   with MHCC97H overexpressing TMCO3 and HLF knockdown TMCO3 respectively,
   and verified their level of mRNA expression (Figure [64]2C; Figure
   [65]S2D, Supporting Information). First, MHCC97H cells overexpressing
   TMCO3 were subcutaneously injected into nude mice. After four weeks,
   the group with TMCO3 overexpression showed larger tumor volume and
   increased tumor weight compared to the control group (Figure [66]2D–F).
   Then we injected HLF cells with TMCO3 knockdown in the same way, the
   results were opposite (Figure [67]S2E–G, Supporting Information).
   Further to investigate the effect of TMCO3 on HCC metastasis, the
   MHCC97H cell line control and overexpressing TMCO3 groups were used to
   construct orthotopic xenograft models and nude mice tail vein models.
   After six weeks, the overexpressing TMCO3 group substantially enhanced
   the capacity of HCC cells to form secondary lesions in the liver
   compared to the control group (Figure [68]2G). In the nude mice tail
   vein models, the mice injected with TMCO3 overexpression cells showed
   more lesions and nodules compared to the control group after eight
   weeks (Figure [69]2H–J). In vitro experiments, colony formation assays
   and CCK‐8 assays indicated that overexpression of TMCO3 caused
   increased cell proliferation and colony numbers in MHCC97H cells, and
   knockdown TMCO3 in HLF cells resulted in the opposite results
   (Figure [70]2K,L; Figure [71]S2H,I, Supporting Information). Then, the
   migration and invasion assays showed the increased ability of migration
   and invasion in MHCC97H overexpressing TMCO3 cells, and similar effects
   were observed in wound healing assays (Figure [72]2M,N). The opposite
   effects were found in HLF knockdown TMCO3 cells (Figure [73]S2J,K,
   Supporting Information).

Figure 2.

   Figure 2
   [74]Open in a new tab

   TMCO3 promotes HCC proliferation and metastasis in vivo and in vitro.
   A) Diagram of HTVi‐induced HCC models in C57BL/6 mice. B) The gross
   images of tumors in β‐catenin+c‐MET+PT3‐Ctrl group and
   β‐catenin+c‐MET+PT3‐TMCO3 (Mouse) group. C) The relative TMCO3 mRNA
   expression level in MHCC97H‐Ctrl group and MHCC97H‐TMCO3 group. D) The
   gross images of subcutaneous tumors in MHCC97H‐Ctrl and MHCC97H‐TMCO3
   groups. E) The tumor weights of subcutaneous tumors in MHCC97H‐Ctrl and
   MHCC97H‐TMCO3 groups. F) The tumor volumes of subcutaneous tumors in
   MHCC97H‐Ctrl and MHCC97H‐TMCO3 groups. G) Representative images of the
   bioluminescence, gross, and HE staining of orthotopic xenograft tumors
   in MHCC97H‐Ctrl and MHCC97H‐TMCO3 groups. H) Representative images of
   the bioluminescence, gross, and HE staining of lung metastasis models
   in MHCC97H‐Ctrl and MHCC97H‐TMCO3 group. I) The statistical chart of
   metastatic nodules of the lung in MHCC97H‐Ctrl and MHCC97H‐TMCO3
   groups. J) The schematic drawing of the incidence of lung metastasis in
   MHCC97H‐Ctrl and MHCC97H‐TMCO3 groups. K) Colony formation assays of
   MHCC97H cells in Ctrl and TMCO3 groups. L) CCK‐8 assays of MHCC97H
   cells in Ctrl and TMCO3 groups. M) Migration and invasion assays of
   MHCC97H cells in Ctrl and TMCO3 groups. N) Wound healing assays of
   MHCC97H cells in Ctrl and TMCO3 groups.

   Altogether, our results showed TMCO3 could promote the progression and
   metastasis of HCC both in vivo and in vitro.

2.3. ALKBH5 Downregulates TMCO3 in an m^6A‐Dependent Manner

   Considering that we identified the molecule TMCO3 through MeRIP‐seq and
   RNA‐seq of HCC patient tumor and adjacent non‐cancerous tissues,
   existing m^6A regulators were preliminary to explore what type of m^6A
   regulator affects TMCO3. We detected the protein level of TMCO3 in
   MHCC97H cells after siRNA‐mediated knockdown of several common m^6A
   regulators, including METTL3, METTL14, WTAP, FTO, and ALKBH5 (Figure
   [75]S3A, Supporting Information). The western blot results indicated
   that only si‐ALKBH5 resulted in an elevation in the protein level of
   TMCO3, while the others showed no significant influences (Figure
   [76]3C,D). RT‐qPCR results revealed that TMCO3 mRNA level was
   significantly upregulated upon ALKBH5 inhibition in both MHCC97H cells
   and HLF cells, whereas overexpressing of ALKBH5 resulted in
   downregulation of TMCO3 mRNA level (Figure [77]3A,B). In 293 T cells,
   exogenous plasmids of ALKBH5 and TMCO3 were co‐transferred. The western
   blots result also showed that the protein level of TMCO3 decreased with
   the increase of ALKBH5 concentration dose (Figure [78]S3B, Supporting
   Information). Then we constructed psiCHECK2 plasmid containing TMCO3 wt
   region and conducted the dual luciferase reporter gene experiment in
   293 T cells. The results showed that the activity of Renilla luciferase
   in the ALKBH5 overexpression group was significantly downregulated
   (Figure [79]S3C, Supporting Information). Given that ALKBH5 plays a
   pivotal role as an “eraser” in m^6A modifications, an inactive
   ALKBH5‐mut (mutant) was used to test whether ALKBH5‐mediated TMCO3
   regulation was dependent on m^6A modification. When ALKBH5‐mut was
   overexpressed in both MHCC97H and HLF cells, the protein level of TMCO3
   was downregulated only in the ALKBH5‐wt group, not in the ALKBH5‐mut
   group (Figure [80]3E). Subsequent RNA decay assays demonstrated that
   the stability of TMCO3 mRNA significantly decreased upon overexpression
   of ALKBH5 in MHCC97H cells, while it increased following the knockdown
   of ALKBH5 in HLF cells (Figure [81]3F). Furthermore, m^6A‐specific
   antibodies in the TMCO3 mRNA within both MHCC‐97H and HLF cells were
   significantly enriched (Figure [82]3G). Further to explore the specific
   sites of m^6A modification on TMCO3, several potential sites on the CDS
   region of TMCO3 were predicted through the online prediction website
   SRAMP analysis ([83]http://www.cuilab.cn/sramp). Based on the results,
   several potential m^6A sites were identified with high or very high
   confidence (Figure [84]S3C, Supporting Information). To confirm that
   these potential sites are critical for TMCO3 regulation, point
   mutations were introduced into putative m^6A sites: MUT‐210, MUT‐267,
   MUT‐348, and MUT‐465 (Figure [85]3H). Subsequently, the dual luciferase
   reporter assays showed that ALKBH5 did not bind to TMCO3 when the
   MUT‐348 site was mutated in both MHCC97H cells and HLF cells
   (Figure [86]3I). Given that “readers” are directly involved in the m^6A
   modification process, we next explore potential readers that may
   regulate TMCO3. RT‐qPCR results revealed that in MHCC97H cells, upon
   knocking down IGF2BP1, IGF2BP3, YTHDF1, YTHDF2, YTHDF3, eIF3, HNRNPC,
   HNRNPG, YTHDC1, and YTHDC2, the mRNA level of TMCO3 did not change to
   any significant extent (Figures [87]S3D,E, Supporting Information).
   However, after knocking down IGF2BP2, the TMCO3 mRNA level and protein
   level were significantly decreased (Figure [88]S3F,G, Supporting
   Information; Figure [89]3J). To further confirm that it was IGF2BP2
   that regulated TMCO3 rather than other readers, we knocked down several
   readers that play a role in regulating mRNA translation in the MHCC97H
   cell line, such as IGF2BP1, IGF2BP3, YTHDF1, EIF3A, and YTHDC2. The
   western blots result showed that the knockdown of these several readers
   did not change the protein level of TMCO3 (Figure [90]S3H, Supporting
   Information). Moreover, the RIP‐qPCR results confirmed that the TMCO3
   RNA binds to IGF2BP2 in MHCC97H cells and HLF cells (Figure [91]3K).
   The RNA‐pull‐down assay was performed to confirm the physical
   interaction between TMCO3 and IGF2BP2 in MHCC97H cells and HLF cells
   (Figure [92]3L).

Figure 3.

   Figure 3
   [93]Open in a new tab

   ALKBH5 downregulates TMCO3 in an m^6A‐dependent manner. A) The relative
   ALKBH5 and TMCO3 mRNA levels of MHCC97H cells after ALKBH5
   overexpression or knockdown. B) The relative ALKBH5 and TMCO3 mRNA
   expression levels of HLF cells after ALKBH5 overexpression or
   knockdown. C) The protein levels of ALKBH5 and TMCO3 in MHCC97H cells
   after ALKBH5 overexpression or knockdown. D) The protein levels of
   ALKBH5 and TMCO3 in HLF cells after ALKBH5 overexpression or knockdown.
   E) The protein levels of ALKBH5 and TMCO3 of HLF cells in Ctrl,
   ALKBH5‐WT, and ALKBH5‐H204A groups. F) The relative mRNA expression
   levels of TMCO3 after actinomycin D treatment in MHCC97H and HLF cells.
   G) MHCC97H and HLF cells were transfected with Ctrl or ALKBH5 plasmids
   and then subjected to MeRIP‐qPCR assays. H) The Schematic drawing of
   psiCHECK2‐TMCO3 plasmid and the potential m^6A modified site. I) The
   relative luciferase activities of different mutations in TMCO3 CDS
   region after ALKBH5 expression in MHCC97 and HLF cells. J) The relative
   TMCO3 mRNA levels of MHCC97H and HLF cells after IGF2BP2 knockdown. K)
   RIP assays were carried out using anti‐IGF2BP2 antibody, and the RNA
   was extracted from protein G‐agarose with anti‐IGF2BP2 antibody, or
   with control IgG. Then qPCR assays were performed. L) The western
   blotting results of RNA‐pull down in MHCC97 and HLF cells.

   These results revealed that ALKBH5 downregulated TMCO3 in an
   m^6A‐dependent manner, and this effect was mediated by IGF2BP2
   recognition.

2.4. TMCO3 Binding with AKT Directly and Increases AKT Phosphorylation in a
PI3K Dependent Manner

   TMCO3, a member of the monovalent cation: proton antiporter 2 (CPA2)
   transporter proteins family, which has been less studied in tumors. The
   oncogenic role of TMCO3 in HCC has been previously reported, but the
   specific mechanisms are still unclear.^[ [94]^10 ^] The further to
   explore the mechanism through which TMCO3 promotes the progression of
   HCC, immunoprecipitation (IP) followed by MS was performed to identify
   potential TMCO3 interaction proteins, AKT appeared at the top of the
   ranking (Figure [95]4A,B; Table [96]S4, Supporting Information). The
   interaction between TMCO3 and AKT was verified in 293 T cells, and then
   verified their co‐expression in the cytoplasm and membrane through
   immunofluorescence (IF) assays (Figure [97]4C,D). To ascertain whether
   TMCO3 could interact directly with AKT, we further constructed and
   purified GST‐AKT from E.coli and purified Flag‐TMCO3 from 293 T cells,
   then GST pull‐down assays were performed (Figure [98]4E). The results
   showed that TMCO3 could bind to AKT directly. Given the recruitment of
   AKT to the cell membrane and binding to PIP[3] is the initial step of
   AKT activation. To assess test whether TMCO3 could promote the binding
   of AKT and PIP[3], the PIP[3] pull‐down assays showed that the binding
   of AKT and PIP[3] was gradually increased in a TMCO3 dose‐dependent
   manner (Figure [99]4F). Moreover, to observe whether TMCO3 could bind
   to PIP[3], we conducted a lipid strip assay. The results indicated
   that, in contrast to the positive binding of GST‐AKT to both PIP[2] and
   PIP[3], FLAG‐TMCO3 did not exhibit binding to PIP[3]. (Figure
   [100]S4‐1A,B, Supporting Information).

Figure 4.

   Figure 4
   [101]Open in a new tab

   TMCO3 binding with AKT directly and increases AKT phosphorylation in a
   PI3K dependent manner. A) The silver staining diagram of IP. B)
   Secondary peptide distribution of mass spectrometry results. C) The
   interaction between TMCO3 and AKT was confirmed in 293 T cell line. D)
   Immunofluorescence showed that TMCO3 and AKT were colocalized in the
   cytoplasm. E) GST‐pull‐down assay showed the direct binding between
   TMCO3 and AKT. F) PIP[3] pull‐down assay showed the interaction between
   AKT and PIP[3] after transfecting TMCO3. G) MHCC97H cells with
   overexpression of TMCO3 or control were pre‐treated with Pictilisib or
   DMSO control for 2 h before IGF‐1 (100 nm) stimulation. Then western
   blots were performed. H) HLF cells with overexpression of TMCO3 or
   control were pre‐treated with Pictilisib or DMSO control for 2 h before
   IGF‐1 (100 nm) stimulation. Then western blots were performed.

   Next, we tested whether TMCO3 has an effect on downstream signaling in
   PI3K‐AKT pathway. RNA sequencing was performed in TMCO3‐overexpression
   and control cells. Kyoto Encyclopedia of Genes and Genomes (KEGG)
   pathway enrichment analysis and Gene Set Enrichment Analysis (GSEA)
   indicated that genes regulated by TMCO3‐overexpression exhibited the
   greatest enrichment in the PI3K‐AKT pathway (Figure [102]S4‐1C,D,
   Supporting Information). Further to investigate whether TMCO3 affects
   the expression and activation levels of the AKT protein, we assessed
   the protein expression levels of AKT, p‐AKT‐T308, and p‐AKT‐S473 in
   MHCC97H and HLF cell lines with stably overexpressed and knocked‐down
   TMCO3. Western blot results indicated that the expression levels of AKT
   protein remained unaltered with either overexpression or knockdown of
   TMCO3. However, the p‐AKT‐T308 and p‐AKT‐S473 levels were increased
   with TMCO3 overexpression and decreased upon TMCO3 knockdown (Figure
   [103]S4‐1E,F, Supporting Information). Meanwhile, we carried out CO‐IP
   experiments on TMCO3 with some classic PI3K subunits and other classic
   molecules that can affect AKT activation (classic regulatory subunits
   and other proteins: P110α, P110α, p85α, PDK1, PHLPP2, and SIN1).
   Unfortunately, the results showed that these proteins seem didn't bind
   to TMCO3 (Figure [104]S4‐1G, Supporting Information). Next, we also
   explored whether TMCO3's promotion of AKT activation depends on its
   function as an antiporter. We added Dimethylamiloride (which is an
   antiporter inhibitor) to the cell lines overexpressing TMCO3 in MHCC97H
   and HLF respectively. The western blot results showed that compared
   with the group only overexpressing TMCO3, the AKT activation level in
   the group with Dimethylamiloride added didn't seem to change (Figure
   [105]S4–2A, Supporting Information). Therefore, we speculate that TMCO3
   may not affect the AKT activation level through its antiporter
   function. Subsequently, we investigated whether the regulation of AKT
   phosphorylation by TMCO3 was dependent on the demethylation
   modification of ALKBH5 through western blots assay.^[ [106]^11 ^] The
   results demonstrated that upon transfection with ALKBH5‐wt, the protein
   level of TMCO3 was downregulated, and the phosphorylation levels at the
   T308 and S473 sites of AKT were significantly reduced compared to the
   control group. When the ALKBH5‐mut was transfected, the protein level
   of TMCO3 showed a notable recovery compared to the wild‐type ALKBH5
   transfection group, and the phosphorylation at the T308 and S473 sites
   of AKT was also partially restored. These findings suggest that the
   regulation of AKT phosphorylation levels by TMCO3 was contingent upon
   the demethylase activity of ALKBH5 (Figure [107]S4‐2B, Supporting
   Information).

   IP assays then showed that the interaction between TMCO3 and AKT was
   enhanced with increasing IGF stimulation time and the dosage of TMCO3,
   respectively (Figure [108]S4‐2C,D, Supporting Information). Since IGF
   is a major cytokine that stimulates the activation of the PI3K‐AKT
   pathway, we speculated whether TMCO3 is dependent on the activation of
   PI3K pathway to increase the phosphorylation level of AKT. To verify
   this hypothesis, we treated MHCC97H and HLF cells with PI3K inhibitor
   Pictilisib, or DMSO control. In the Pictilisib treatment group,
   overexpression of TMCO3 did not affect p‐AKT‐T308 and p‐AKT‐S473 levels
   (Figure [109]4G,H). This result confirmed that AKT activation by TMCO3
   is contingent upon the activation of the PI3K pathway. Subsequently, we
   performed IP assays in MHCC97H and HLF cells. The results demonstrated
   that the binding between TMCO3 and AKT was significantly enhanced under
   IGF stimulation, but this effect was obviously weakened after the
   treatment of Pictilisib (Figure [110]S4‐2E,F, Supporting Information).

   In summary, these results indicated that TMCO3 directly binds to AKT
   and activates AKT in a PI3K‐dependent manner.

2.5. The IGF‐1‐Stimulated Phosphorylation on the Serine 85 Site of TMCO3
Facilitates the Membrane Localization and Activation of AKT

   As a multi‐transmembrane protein, TMCO3 is predominantly localized in
   the plasma membrane and cytoplasm. A prerequisite for AKT activation is
   its transfer from cytoplasm to the inner surface of plasma membrane
   first, followed by its binding to PIP[3], which in turn activates the
   T308 site by PDK1.^[ [111]^12 ^] We thus hypothesized that TMCO3 might
   increase the membrane location of AKT. Through cytoplasm‐membrane
   protein fractionation assays in MHCC97H and HLF cells, overexpression
   of TMCO3 was found to significantly enhance the membrane localization
   of AKT during stimulation of IGF‐1, while TMCO3 knockdown had the
   opposite effect (Figure [112]5A,B; Figure [113]S5A,B, Supporting
   Information). Consistently, our IF experiments confirmed this
   hypothesis, demonstrating that overexpression of TMCO3 could indeed
   promote the membrane localization of AKT (Figure [114]5C,D). Based on
   previous studies, little is known about the function of TMCO3 as a
   membrane protein.^[ [115]^13 ^] To further explore the specific region
   of TMCO3 that interacts with AKT, we constructed truncated plasmids of
   TMCO3, namely TMCO3(1‐124), TMCO3(1‐204), TMCO3(204‐677), and
   TMCO3(124‐677), based on the functional domain distribution of TMCO3.
   These constructs were co‐transfected with AKT in 293T cells. IP results
   revealed that AKT binds only to TMCO3(1‐124), TMCO3(1‐204), and the
   full‐length TMCO3, indicating that the TMCO3(1‐124) region may be
   responsible for the interaction between TMCO3 and AKT (Figure
   [116]S5C,D, Supporting Information). Therefore, we further hypothesized
   whether TMCO3 itself possessed phosphorylation sites that could be
   activated, thus playing a kinase receptor‐like role in triggering AKT
   phosphorylation. To examine this hypothesis, we used the Cell Signaling
   Technology (CST) MS database ([117]https://www.phosphosite.org) to
   predict potential phosphorylation residues in the TMCO3 CDS domain
   (Figure [118]S5E, Supporting Information). Further to determine which
   residue was the requisite target, we generated the phosphor‐deficient
   mutants for TMCO3 (S to A) and observed that TMCO3‐S85A, not S134A or
   S638A, could markedly decrease p‐AKT‐T308 and p‐AKT‐S473 levels in
   MHCC97H cells (Figure [119]S5G, Supporting Information). Subsequently,
   the co‐IP data revealed that compared with the empty vector control,
   transfection of TMCO3‐S85A reduced serine/threonine phosphorylation
   levels of AKT, as detected by immunoblotting with
   phospho‐Thr/Ser–specific antibodies (Figure [120]S5H, Supporting
   Information). Then, we selected homologous sequences from various
   species sources in the Uniprot database. After aligning them using
   Clustal Omega, we intercepted the sequence interval of 1–124, the amino
   acid sequence of the 1–124 segment of TMCO3 is highly conserved.
   Moreover, the 85th serine site is also highly conserved across
   different species, which also indirectly indicates that this region is
   most likely the main region for activating AKT (Figure [121]5E; Figure
   [122]S5E, Supporting Information). To further confirm that the 85th
   serine is the specific site of phosphorylation on TMCO3, we generated a
   site‐specific antibody that recognizes the phosphorylated S85 site on
   TMCO3. The western blots results demonstrated that, following the
   addition of IGF‐1 stimulation, overexpression of TMCO3 led to a
   significant increase in p‐TMCO3‐S85 level, and it resulted in a notable
   enhancement in p‐AKT‐T308 and p‐AKT‐S473 levels (Figure [123]5F). In
   addition, we conducted kinase activity assays to observe differences in
   kinase activity among various groups by comparing the residual amounts
   of ATP after enzymatic reactions in each group. The results indicated
   that, compared to the control group, the TMCO3‐wt group exhibited a
   significant decrease in the residual ATP levels. The TMCO3(1‐124) group
   showed a trend consistent with that of the TMCO3‐wt group. However,
   when the 85th serine site was mutated, the residual ATP levels showed
   no significant difference compared to the control group
   (Figure [124]5G). Next, the TMCO3 knockout MHCC97H cells were
   constructed, and western blot results revealed that p‐TMCO3‐S85 level
   was virtually undetectable after the knockout of TMCO3. The p‐TMCO3‐S85
   level in the TMCO3 knockout cell lines modestly increased following the
   TMCO3‐S85A restoration, while it significantly rose after TMCO3‐S85D
   was restored. The p‐AKT‐S473 level mirrored the trend seen with
   p‐TMCO3‐S85 (Figure [125]S5I, Supporting Information). Hence, these
   findings implied that IGF‐1 mediates the activation of phosphorylation
   on the 85th serine site of TMCO3, which in turn, facilitates the
   membrane localization and activation of AKT.

Figure 5.

   Figure 5
   [126]Open in a new tab

   The IGF‐1‐stimulated phosphorylation on the serine 85 site of TMCO3
   facilitates the membrane localization and activation of AKT. A) The
   Cytoplasm and membrane protein fractionation assay was performed in
   MHCC97H cells after overexpression of TMCO3 or control. B) The
   Cytoplasm and membrane protein fractionation assay was performed in HLF
   cells after overexpression of TMCO3 or control. C) Immunofluorescence
   images showed the location of AKT in MHCC97H cell after overexpression
   of TMCO3. D) Immunofluorescence images showed the location of AKT in
   HLF cell after overexpression of TMCO3. E) Schematic diagram of TMCO3
   85th amino acid in different species. F) The western blot results
   showed the protein levels of p‐TMCO3‐S85 and other targets in MHCC97H
   and HLF cells after IGF‐1 stimulation. G) The chemiluminescent kinase
   activity assay reflected the differences in the residual ATP levels
   among the various groups.

2.6. Mutation on the Serine 85 of TMCO3 Inhibits the Membrane Location and
Kinase Activity of AKT, Ultimately Impeding HCC Progression

   Previous data verified that phosphorylation on the S85 site of TMCO3
   could promote the membrane localization and activation of AKT. Then, we
   investigated whether mutating S85 site could alleviate this effect.
   Through cytoplasm‐membrane protein fractionation assays and IF assays,
   our results highlighted that compared to the TMCO3‐WT group, the
   distribution of AKT within the cell membrane of the TMCO3‐S85A group
   was significantly reduced, with a corresponding significant decrease in
   the p‐AKT‐S473 level (Figure [127]6A–C; Figure [128]S6A, Supporting
   Information). Next, whether mutations on S85 site could weaken the
   binding of TMCO3 and AKT was ascertained. Co‐IP results showed that
   compared with TMCO3‐WT, the binding of TMCO3‐S85A to AKT was
   significantly weakened, while the binding of TMCO3‐S85D to AKT was
   restored (Figure [129]6D). Then, to ascertain whether TMCO3 could
   directly phosphorylate both the T308 and S473 sites of AKT
   simultaneously, we conducted in vitro kinase assays.^[ [130]^14 ^] The
   results showed that when the S85 site of TMCO3 was mutated, the
   phosphorylation levels at T308 and S473 sites of the wild‐type AKT
   purified protein were down‐regulated to varying degrees. When the T308
   site of AKT was mutated, the activating effect of TMCO3 on
   GST‐AKTT^308A almost completely disappeared. And when the S473 site of
   AKT was mutated, the activating effect of TMCO3 on GST‐AKT^S473A also
   nearly vanished. Combining with the results of the Thi‐ester
   phosphorylation antibody, we concluded that TMCO3 could directly
   phosphorylated the T308 and S473 sites of AKT, and this activating
   effect may be more significant for the phosphorylation of the S473 site
   (Figure [131]6E). The PIP[3] pull‐down assays also showed that
   following the mutation on site 85 of TMCO3, the binding of PIP[3] to
   AKT was notably weakened compared to TMCO3‐WT group, which was
   consistent with the previous results (Figure [132]6F). The better to
   evaluate the effect of mutation on S85 site, TMCO3‐WT, TMCO3‐S85A,
   TMCO3‐S85D, TMCO3‐S134A, and TMCO3‐S638A in MHCC97H‐KO‐TMCO3 cell lines
   were restored.^[ [133]^15 ^] CCK‐8 assays and colony formation assays
   demonstrated that compared to KO+WT group, the cell proliferation
   capacity of KO+S85A group was significantly reduced (Figure [134]S6B,C,
   Supporting Information). The other groups showed similar effects to
   those of KO+WT group. Subsequently, the subcutaneous tumor injection
   model demonstrated findings consistent with in vitro experiments. Tumor
   weight and volume were significantly increased in the KO+WT group as
   compared to the KO group. However, this effect almost vanished after
   TMCO3‐S85A restoration, and the other groups exhibited trends similar
   to the KO+WT group (Figure [135]S6D–F, Supporting Information).
   Collectively, these results support mutation of S85 site on TMCO3
   impairs AKT membrane location and kinase activity, finally negatively
   regulating oncogenic functions of HCC.

Figure 6.

   Figure 6
   [136]Open in a new tab

   Mutation on the Serine 85 of TMCO3 inhibits AKT membrane location and
   kinase activity, ultimately impeding HCC progression. A) The Cytoplasm
   and membrane protein fractionation assay was performed in MHCC97H cells
   after overexpression of TMCO3‐WT, TMCO3‐MUT (S85A), or control. B) The
   Cytoplasm and membrane protein fractionation assay was performed in HLF
   cells after overexpression of TMCO3‐WT, TMCO3‐MUT (S85A), or control.
   C) Immunofluorescence images showed the location of AKT in MHCC97H cell
   after overexpression of TMCO3‐WT, TMCO3‐MUT (S85A), or control. D) IP
   assays showed the interaction between TMCO3 and AKT after transfection
   of TMCO3‐WT, TMCO3‐S85A, TMCO3‐S85D, or control. E) Western blots on
   the products from in vitro kinase assays assessing purified FLAG‐TMCO3/
   TMCO3^S85A phosphorylation of GST‐AKT/AKT^T308A/AKT^S473A. F) PIP[3]
   pull‐down assay showed the interaction between AKT and PIP[3] after
   transfection of TMCO3‐WT and TMCO3‐S85A.

2.7. Combined Treatment with AAV‐shTMCO3 and MK2206 Dramatically Decreases
AKT Phosphorylation and Inhibits Tumorigenesis in Mouse Model

   Next, to investigate whether targeting TMCO3 could augment the
   therapeutic effectiveness of MK2206, an allosteric AKT inhibitor, a
   hydrodynamic model was established for c‐Met/N90‐β‐catenin ‐driving
   hepatocarcinogenesis (Figure [137]7A).^[ [138]^16 ^] Initially, the
   knockdown efficiency of TMCO3 (mouse) in Hep1‐6 cells was demonstrated
   (Figure [139]S7‐1A, Supporting Information).^[ [140]^17 ^] Following
   this, we utilized adeno‐associated virus 8 (AAV8) to package the short
   hairpin RNA (shRNA) targeting TMCO3 and designed a control vector for
   AAV8 as a measure to verify the success of the AAV8 injection (Figure
   [141]S7‐1B, Supporting Information).^[ [142]^18 ^] The results showed
   that compared to the control group, both the AAV‐shTMCO3 group and the
   MK2206 treatment group had significantly lower tumor burdens.^[
   [143]^19 ^] The combined treatment of AAV‐shTMCO3 and MK2206 resulted
   in the most substantial reduction in tumor burden and also exhibited
   the best survival rate (Figure [144]7B–E). Overexpression of c‐Met,
   N90‐β‐catenin, and knockdown of TMCO3 were validated by western blots
   in these four groups (Figure [145]S7‐1C, Supporting Information). The
   results of immunohistochemical (IHC) staining also indicated that the
   expression levels of KI67, TMCO3, and p‐AKT‐S473 were significantly
   downregulated in both the AAV8 and MK2206 groups, and the most
   pronounced downregulation was observed in the combined treatment group
   (Figure [146]7E). Simultaneously, multiplex immunohistochemistry
   (m‐IHC) results also revealed that the decrease in the levels of TMCO3
   and p‐AKT‐S473, as well as the reduction in AKT membrane localization,
   were most significant in the combined treatment group compared to the
   control group (Figure [147]7F; Figure [148]S7‐1D,E, Supporting
   Information). Then, we used the MHCC97H cell line to construct
   orthotopic xenograft models in nude mice and divided these mice into
   four groups as before. The results mirrored our previous findings,
   where the combined treatment group had the lightest tumor burden and
   the fewest intrahepatic metastatic foci (Figure [149]S7‐2A–D,
   Supporting Information). Western blots and IHC staining results
   validated the knockdown efficiency of TMCO3, the expression level of
   473, and the localization of AKT (Figure [150]S7‐2E,F, Supporting
   Information). In conclusion, these data suggested that targeting TMCO3
   combined with MK2206 could significantly decrease AKT phosphorylation
   and inhibit the progression of HCC.

Figure 7.

   Figure 7
   [151]Open in a new tab

   Combined treatment with AAV‐shTMCO3 and MK2206 dramatically decreases
   AKT phosphorylation and inhibits tumorigenesis in mouse models. A) The
   schematic diagram of HTVi‐induced HCC models in C57BL/6 mice and
   treatment groups. B) The liver weight statistics chart in four
   different groups. C) The statistics chart of liver/body weight ratio in
   four groups. D) The survival rates of mice in four groups. E) The gross
   images, HE staining, and IHC staining of four groups. F) The mIHC
   images include relative targets in four groups.

2.8. The Levels of TMCO3 and p‐TMCO3‐S85 are Positively Correlated with HCC
and Predict Poor Prognosis

   The previous data validated the influence of TMCO3 on AKT and its
   inhibitory role in HCC progression through both cell and animal
   experiments. Next, to observe the expression of TMCO3 and its
   associated prognosis in a cohort of patients with HCC from Tongji
   Hospital, we selected tissue microarrays from 123 pairs of HCC patient
   tumors and adjacent non‐cancerous tissues for IHC staining of TMCO3,
   AKT, and P‐AKT‐S473 (Figure [152]8A,B; Table [153]S1, Supporting
   Information). The results showed that the expression levels of TMCO3
   and P‐AKT‐S473 in cancer tissues were significantly higher than those
   in adjacent non‐cancerous tissues (Figure [154]8D). Moreover, in
   samples with a high TMCO3 expression, the proportion of AKT localized
   to the membrane was also notably higher than in samples with a low
   TMCO3 expression (Figure [155]8C). The Pearson correlation test
   indicated a significant positive correlation between the expression
   levels of P‐AKT‐S473 and TMCO3 (Figure [156]8E). Subsequently, western
   blotting was conducted to assess the expression levels of TMCO3 and
   p‐TMCO3‐S85 in 40 pairs of tumor and adjacent non‐cancerous tissues
   from HCC patients. The results revealed that both TMCO3 and p‐TMCO3‐S85
   were significantly increased in cancer tissues compared to adjacent
   noncancerous tissues (Figure [157]8F; Figure [158]S8A,B, Supporting
   Information). Consistent with the analysis of the TCGA database, the
   group with a high TMCO3 expression in the Tongji HCC patient cohort
   exhibited poorer overall survival and disease‐free survival rates
   (Figure [159]8G,H). Cox's multivariate analysis indicated that higher
   TMCO3 expression was an independent risk factor for shorter overall
   survival (Figure [160]8I; Table [161]S2, Supporting Information). In
   summary, these data demonstrated that the levels of TMCO3 and
   p‐TMCO3‐S85 were highly expressed in HCC patients, and also correlated
   with worse OS and DFS in the Tongji HCC patient cohort.

Figure 8.

   Figure 8
   [162]Open in a new tab

   The levels of TMCO3 and p‐TMCO3‐S85 are positively correlated with HCC
   and predict poor prognosis. A) The IHC staining of case 6 of TMCO3,
   AKT, P‐AKT‐S473 targets in tissue microarrays. B) The IHC staining of
   case 107 of TMCO3, AKT, and P‐AKT‐S473 targets in tissue microarrays.
   C) Statistical chart showed the cytoplasmic and membrane localization
   of AKT in groups with high or low TMCO3 expression. D) The statistical
   chart showed the TMCO3 IHC scores in adjacent noncancerous or tumor
   tissues of HCC patients in the Tongji cohort. E) Pearson's correlation
   analysis of the TMCO3 and P‐AKT‐S473 IHC scores in tissue microarrays.
   F) The Statistical chart of TMCO3 and P‐TMCO3‐S85 protein levels in
   adjacent noncancerous or tumor tissues of 40 pairs of HCC patients. G)
   The overall survival with TMCO3 low expression or TMCO3 high expression
   of 123 pairs of HCC patients in the Tongji cohort. H) The disease‐free
   survival with TMCO3 low expression or TMCO3 high expression of 123
   pairs of HCC patients in the Tongji cohort. I) Forest plot of the
   multivariate analysis for overall survival. J) Schematic model of the
   mechanism underlying TMCO3‐drived HCC progression and relative
   treatment.

3. Discussion

   Nowadays, the surgical operations and medicine treatment for HCC
   worldwide have reached a certain bottleneck, with advanced HCC and some
   recurrent and metastatic HCC still associated with high mortality
   rates. Therefore, an urgent understanding of the molecular mechanisms
   underlying HCC progression and therapeutic strategies is needed. This
   study identified TMCO3, a highly expressed protein in HCC, through
   MeRIP‐seq and RNA‐seq of patient tumor tissues. Mechanically, ALKBH5
   downregulates the m^6A modification of TMCO3 mRNA, and IGF2BP2 binds to
   the m^6A site of TMCO3 mRNA, maintaining its stability. TMCO3 promotes
   HCC progression by facilitating the membrane localization of AKT and
   increasing AKT activation. Significantly, targeting TMCO3 combined with
   MK2206, a classic allosteric inhibitor of AKT, effectively inhibits HCC
   progression.^[ [163]^20 ^]

   In recent years, studies on m^6A modification in tumors have been
   emerging, with a growing number of reports focusing on its role in
   HCC.^[ [164]^21 ^] Classic m^6A modifications are primarily conducted
   by “writers” such as METTL3/METTL14/WTAP or “erasers” like ALKBH5/FTO,
   modifying the m^6A sites on targeted gene mRNA; “readers” such as
   IGF2BP1/2/3, YTHDF1/2/3 recognize and bind to the target gene mRNA for
   stabilization or degradation.^[ [165]^22 ^] Our study discovered for
   the first time that ALKBH5 downregulated the m^6A methylation level of
   TMCO3 mRNA, with IGF2BP2 participating in its mRNA stabilization.
   Moreover, the specific site of m^6A modification on TMCO3 mRNA was
   identified, which bears significant importance for our subsequent
   research.

   The PI3K/AKT signaling pathway is participated in the progression of
   various tumors and is also highly activated in HCC.^[ [166]^23 ^] The
   sublocalization of AKT in tumor cells, and its activation of
   phosphorylation, can determine the role of the pathway in the tumor.^[
   [167]^24 ^] In the current mainstream studies, the T308 site activation
   of AKT is primarily induced by PDK1, the S473 site is stimulated by
   mTORC2, and the influence of PTEN on inhibiting T308 site activation.^[
   [168]^25 ^] In addition to these classic ways of activating AKT, there
   are also some non‐classic proteins that activate AKT independently of
   kinase function. For instance, the proto‐oncogene TCL1 can directly
   bind to AKT to promote its oligomerization, thereby altering the
   protein spatial conformation of AKT and facilitating its
   phosphorylation.^[ [169]^26 ^] Moreover, upon stimulation by IL – 1β,
   TRAF6 can undergo oligomerization, followed by K63‐linked
   polyubiquitination. This polyubiquitination does not lead to the
   degradation of TAK1, an interacting protein of TRAF6, via the
   proteasome pathway. Instead, it changes the activity state of TAK1. It
   is precisely due to the autoubiquitination of TRAF6 that the spatial
   conformation of the TRAF6‐TAK1 complex is altered, resulting in the
   activation of TAK1.^[ [170]^27 ^] In our study, we identified TMCO3, an
   oncogenic molecule, from MeRIP‐seq and RNA‐seq and confirmed its
   oncogenic effect. Subsequent mass spectrometry results revealed that
   AKT is a major candidate among binding proteins. Hence, we validated
   the direct bind between TMCO3 and AKT, discovering that TMCO3 binds
   with AKT more effectively when the PI3K pathway is activated.
   Furthermore, considering the cytoplasm and membrane localization of
   TMCO3, TMCO3 was found to facilitate the accumulation of AKT from the
   cytoplasm to the membrane and promote the phosphorylation level of AKT
   in HCC cell lines. Intriguingly, we speculated whether TMCO3 could
   function in a manner akin to that of other kinases to activate AKT and
   verified this hypothesis through phosphorylation mass spectrometry. We
   identified the 85^th serine site on TMCO3 and observed that the
   addition of IGF‐1 stimulation significantly increased the levels of
   p‐TMCO3‐S85. Subsequently, we mutated the 85^th serine site of TMCO3
   from serine to alanine, noticing that TMCO3‐S85A largely abolished the
   function of promoting the membrane localization of AKT and activating
   AKT phosphorylation. This finding not only further elucidates the role
   of TMCO3 in HCC but also largely provides a new regulatory factor for
   the PI3K/AKT pathway, paving the way for future studies.

   Further to highlight the significance of TMCO3 in HCC, we established a
   hydrodynamic‐induced HCC model in C57BL/6 mice and targeted TMCO3 via
   AAV‐shTMCO3, combined with MK2206 to observe the therapeutic effect on
   HCC. The results showed that both the AAV‐shTMCO3 group and the MK2206
   group had certain therapeutic effects, whereas the combined treatment
   group exhibited the most significant effect, with the levels of
   p‐AKT‐T308 and p‐AKT‐S473 most noticeably reduced. Subsequently, we
   constructed an orthotopic xenograft model in nude mice and performed
   similar grouping, yielding consistent results. These findings suggest
   that TMCO3 exerts its corresponding function in vitro, and its
   influence on HCC in vivo has been further validated.

   Next, we performed TMCO3, AKT, and p‐AKT‐S473 staining on tissue
   microarrays of 123 pairs of patients. The results revealed that the
   cell membrane localization of AKT in the group with a high TMCO3
   expression was significantly higher than that in the group with a low
   TMCO3 expression. Moreover, the expression level of TMCO3 was
   positively correlated with p‐AKT‐S473. Subsequently, the protein
   expression of TMCO3 and p‐TMCO3‐S85 in patient tissues was detected.
   Their levels of expression were found to be significantly elevated in
   tumor tissues. Within our patient cohort, high TMCO3 expression also
   indicated poor overall survival and recurrence rates.

   Overall, our study shows that TMCO3, a highly expressed and high
   m^6A‐modification level molecule in HCC, is involved in the regulation
   of AKT through the PI3K/AKT pathway and ultimately promotes the
   progression of HCC (Figure [171]8J). These findings suggest that TMCO3
   serves as a useful biomarker in HCC, and targeting TMCO3 may provide a
   potential therapeutic strategy for HCC, especially in HCC patients with
   a highly activated PI3K/AKT pathway.

4. Experimental Section

Cell Culture and Reagents

   THLE‐3, MHCC97H, Hep3B, HLF, HepG2, Alex, and LM3 cell lines were
   seeded in DMEM supplemented with 10% fetal bovine serum (FBS) (Tsingmu,
   Wuhan, China) at 37 °C in 5% CO2. The siRNAs were designed and provided
   by AuGCT Biotecho and the Lipomaster 3000 Transfection Reagent was
   obtained from Vazyme (Nanjing, China). The optical MEM was purchased
   from ThermoFisher (31985070).

Animals Study

   Male C57BL/6J mice and BALB/c nude mice were obtained from
   GemPharmatech (Nanjing, China) and were maintained in a Specific
   Pathogen Free (SPF) facility. For subcutaneous tumor assay in BALB/c
   nude mice, 1 × 10^6 cells were diluted in 100 µL serum‐free medium and
   then injected in both armpits. For orthotopic xenograft models, 1 ×
   10^6 cells were diluted in 30 µL serum‐free medium and then injected
   into the liver. For tail vein injection models, 1 × 10^6 cells were
   diluted in 100 µL serum‐free medium and then injected in the tail vein.
   For the hydrodynamic tail vein injection (HTVi) induced HCC, 5–6 weeks
   male C57BL/6J mice were used. For each mouse, PT3‐c‐Met
   (20 µg)/PT3‐△N90‐β‐catenin (20 µg)/PT3‐Vector/TMCO3 (20 µg) and SB
   transposase plasmids (1/25 of the total plasmid mass) were dissolved in
   2 mL PBS. The other details of animal studies were provided in the
   Supplementary Information.

Statistical Analysis

   Each value was shown as the average plus or minus the standard
   deviation. To assess statistical significance between two groups,
   Student's t‐tests (for normal distribution) or Wilcoxon signed‐rank
   tests (for matched pairs) were utilized. For multiple groups,
   statistical analysis was performed using either one‐way ANOVA or
   two‐way ANOVA. The Kaplan–Meier method illustrated the survival curves,
   while the log‐rank test evaluated their statistical significance.
   Correlations were evaluated using a Pearson correlation test.
   Appropriate statistical tests were conducted for all Figures, with p
   < 0.05 considered statistically significant: *p < 0.05; **p < 0.01;
   ***p < 0.001; ns: not significant. All statistical values were
   calculated using GraphPad Prism 8.0 software.

   Other materials and methods are presented in the Supplementary
   Information.

Ethics Approval and Consent to Participate

   The ethics committee of Tongji Hospital, affiliated with Tongji Medical
   College at Huazhong University of Science and Technology (Wuhan, China)
   granted approval for the study. All research was conducted in
   accordance with both the Declarations of Helsinki and Istanbul. Written
   consent was given in writing by all subjects (TJ‐IRB20211214). All
   animal experimental procedures were also approved by the Committee on
   the Ethics of Animal Experiments of Tongji Hospital (TJH‐202110039).

Conflict of Interest

   The authors declare no conflict of interest.

Author Contributions

   X.L. and M.H. contributed equally to this work. B.Z., Z.L., and X.C.
   conceived this study; B.Z. and Z.L. provided financial and
   administrative support; X.L. and M.H. conducted most experiments; H.Z.
   and H.L. collected the public data and patients’ data of Tongji cohort;
   X.L. wrote the manuscript. All authors have read and approved the final
   manuscript.

Supporting information

   Supporting Information
   [172]ADVS-12-2504187-s001.docx^ (6.9MB, docx)

Acknowledgements