Notch is a classic and conservative signaling pathway. The target genes of Notch signaling mainly regulate the phenotype of CSCs, cell survival and differentiation and cell cycle and apoptosis (Andersson et al., 2011). At the same time, Notch is a regulator of CSCs and the TME, and its positive and negative effects have been confirmed by many studies (Weng et al., 2004; Venkatesh et al., 2018). Notch signaling mainly regulates three subgroups in the TAIM: regulatory T cells (Tregs), myeloid-derived suppressor cells (MDSCs) and T cells (Grazioli et al., 2017). In Tregs, Notch can also cascade with NF-κB transcription factors to play a regulatory role. In the TAIM, the classic NF-κB pathway is at the core of Treg functions regulated by Notch1 and Notch3 (Ferrandino et al., 2018). In leukemia, Notch1 can regulate the expression of NF-κB subunits directly to regulate T cells (Osipo et al., 2008) or can bind with the NF-κB subunit to regulate transcription indirectly in different environments and cells (Song L.L. et al., 2008). Notch can also enhance Treg activity through the C-X-C chemokine receptor type 4 (CXCR4) pathway and affect MDSCs through the interleukin (IL) -6-STAT3 axis, thereby promoting the formation of ITAIMs (Weng et al., 2004; Li et al., 2018).

HH signaling plays an important role in embryonic development, but its abnormal activation has been shown to be related to the development of a variety of tumors (Katoh, 2019). In recent years, HH signaling has also been found to be related to ITAIM formation and immune escape. Hanna identified that HH signaling alters a critical kinomic signature that enables macrophages to assume the alternative M2 phenotype (Hanna et al., 2019). Chakrabarti’s study showed that HH signaling can induce the expression of programmed death-ligand 1 (PD-L1) in gastric tumor cells and can protect cancer cells from CTL-induced apoptosis (Chakrabarti et al., 2018). It has also been reported that in patients with advanced basal cell carcinoma, the application of HH inhibitors can improve the therapeutic effect of immune checkpoint inhibitors (Nikanjam et al., 2019). In glioma, HH signaling can recruit tumor-associated macrophages (TAMs) and can cause the immunosuppression MDSCs by releasing C-C motif chemokine ligand (CCL) 2/3 (Mi et al., 2020), thereby inducing PD-L1 expression, inducing stromal cells to produce IL10, and promoting the transcription factor forkhead box P3 (FOXP3) expression in Tregs (Grund-Groschke et al., 2019). HH signaling can also be activated by many other intracellular signals, including transforming growth factor β (TGF-β)-, kirsten rat sarcoma viral oncogene (KRAS)-mitogen activated protein kinase (MAPK)/extracellular regulated protein kinases (ERK)-, PI3K- AKT-, insulin-like growth factor—and tumor necrosis factor (TNF) -α induced mammalian target of rapamycin (mTOR)/ribosomal S6 kinase 1 (S6K1) activation (Takebe et al., 2015; Giroux-Leprieur et al., 2018).

With the advancement of sequencing technology and the comprehensive understanding of cancer genomes, abnormal Wnt signaling has been observed in many cancer entities (Ruan et al., 2020; Yang et al., 2020). In the past few years, three types of Wnt cascade signaling have been described. Among these pathways, the typical Wnt signaling pathways related to tumors involves the activation of the T-cell-specific β-catenin–transcription factor–lymphoid enhancer-binding factor pathway, which plays a key role in immune regulation (Zhong and Virshup, 2020). The Wnt/β-catenin pathway has been identified as one of the important signaling pathways related to immune escape. In creating ITAIMs, Wnt signaling acts Wnt signaling on CD103⁺ dendritic cells (DCs) by inducing the expression of the transcription inhibitor ATF3, reducing the production of CCL4, C-X-C motif chemokine ligand (CXCL) -9 and -10 and inhibiting the activation and infiltration of CD8⁺ T cells (Fu et al., 2015). At the same time, in a pan-cancer analysis based on TCGA database information, Wnt/β-catenin pathway activation is usually associated with low T-cell infiltration and immune rejection (Luke et al., 2019). β-catenin signaling activation reduces the recruitment of DCs in hepatocellular carcinoma, thereby impairing T-cell activity (Ruiz de Galarreta et al., 2019). β-catenin signaling can also inhibit innate immunity by regulating interferon regulatory factor (IRF) 3, thereby downregulating the production of type I interferon (IFN) in colorectal cancer (Ding et al., 2018). Wnt signaling also affects TAMs through Snail to produce IL-1β and increase β-catenin activity (Spranger and Gajewski, 2015) to maintain Wnt signaling activation (Boulter et al., 2015).

The activation of STAT3 signaling in CSCs is key to tumor progression because it promotes immunosuppressive factors and angiogenesis (Sasidharan Nair et al., 2018; Wang Y. et al., 2018). Many oncogenic signaling pathways, such as RAS/BRAF/MEK, can participate in the regulatory network of a TAIM. They cooperate with STAT3 and NF-κB signaling to regulate the expression of cytokines [IL-6, IL-1, IL-10, TNF, and vascular endothelial growth factors (VEGFs)] (Sasidharan Nair et al., 2018; Ji et al., 2019). Activated STAT3 signaling can promote the accumulation of immunosuppressive CD163⁺ TAMs and has an inhibitory effect on T cells. Depletion of CD163⁺ TAMs can enhance the infiltration of CD8⁺ T cells into melanoma and promote CD8⁺ T-cell-mediated tumor regression in mice (Etzerodt et al., 2019). T helper cell (Th) 2 and 17, Tregs and MDSCs produce inflammatory cytokines such as IL-1, IL-17, IL-10, TGF-β, and VEGFs through STAT3 signaling. These cytokines can promote the M2 polarization of macrophages and the proliferation of MDSCs, accelerating the formation of ITAIMs and the proliferation of tumor cells (Owen et al., 2019). The activation of STAT3 has been shown to inhibit the expression of STAT1, IRF7 and IRF 9 genes, type I IFN signaling and antitumor immune responses (Lu et al., 2018). Loannis confirmed that STAT3 signaling promotes PD-L1 expression and suppresses the immune response in breast cancer. In addition, STAT3 downregulates the expression of CXCL-10 and reduces the cytotoxicity induced by natural killer cells (NK) (Clark et al., 2019).

USP4 not only plays a role in promoting cancer in a variety of tumor tissues but also indicates a poor prognosis when highly expressed (Yun et al., 2015; Cao et al., 2016; Qiu C. et al., 2018). USP4 can promote cancer in various ways, such as inhibiting SMAD4 monoubiquitination and promoting activin and BMP signaling (Zhou et al., 2017), promoting breast cancer metastasis via the p21-activated kinase 5-aspartyl aminopeptidase-USP4 axis (Geng et al., 2020) and promoting the invasion of breast cancer cells via the relaxin/TGF-β1/SMAD2/MMP-9 signaling pathway (Cao et al., 2016). In recent studies, we found that the USP4 are also involved in the regulation of the components in SRSs. USP4 promotes liver cancer metastasis by upregulating TGF-β signaling to induce the epithelial-mesenchymal transition (Qiu C. et al., 2018) and activate the ERK pathway by regulating TGF-β to promote the development of glioblastoma (Zhou et al., 2019). Both TGF-β and ERK signaling can activate HH signaling in CSCs. USP4 also stabilize β-catenin in brain metastasis lung adenocarcinoma to affect metastasis ability (Hwang et al., 2016). In colorectal cancer, USP4 upregulates Wnt/β-catenin signaling through deubiquitination and facilitates the nuclear localization of β-catenin (Yun et al., 2015; Nguyen et al., 2019). Recent studies have found that USP4 can deubiquitinate the transcriptional factor retinoic acid-related orphan receptor γt in an inflammatory environment and play a positive regulatory role in Th17 cell function (Yang et al., 2015). Th17 cells can produce inhibitory inflammatory cytokines through STAT3 signaling. In addition, USP4 stabilizes IRF8 through K48-linked deubiquitination and enhances the immunosuppressive function of Tregs (Lin et al., 2017). In summary, USP4 can strengthen the SRSs and directly act on immune cells to promote the ITAIM.

USP7 has been found to be highly expressed in a variety of tumors and is associated with poor prognosis (Zhang et al., 2016; Wang X. et al., 2018; Xia et al., 2019b; Zhang M.J. et al., 2019). The regulatory role of USP7 in TAIMs has attracted increasing attention, and it can also affect the SRSs. In T-cell leukemia, USP7 can regulate Notch1 at the transcriptional and post-translational levels and lead to increased expression of Notch1 target genes (Shan et al., 2018; Jin et al., 2019). After inhibiting USP7, Notch1 is ubiquitinated and degraded (Jin et al., 2019). In colorectal cancer, USP7 activates Wnt signaling through β-catenin deubiquitination to promote the growth of ectopic transplantation tumors (Novellasdemunt et al., 2017). It can also prevent the ubiquitination of NF-κB and its regulator NEK2 (Colleran et al., 2013; Franqui-Machin et al., 2018), activating NF-κB signaling. In addition, monoubiquitinated PTEN can be deubiquitinated by USP7, leading to the elimination of PTEN in the nucleus and activating the PI3K/AKT/mTOR signaling pathway. PTEN/PI3K/AKT, as one of the SRS pathways of CSCs, has been proven to maintain stemness and promote tumor growth in a variety of cancers (Li et al., 2017; Hu et al., 2019; Luongo et al., 2019). Classic PTEN/PI3K/AKT signaling plays a core role in maintaining the characteristics and stability of Tregs (Brandmaier et al., 2017). USP7 can also reduce and stabilize the ubiquitination of FOXP3 directly, enhancing the immunosuppressive ability of Tregs (van Loosdregt et al., 2013). Tregs suppress the immune response by expressing cytotoxic T-lymphocyte-associated protein 4 (CTLA-4), IL-10, TGFβ, etc. This immunosuppressive property promotes the immune escape of tumor cells (Li C. et al., 2020). In addition, immunosuppressive PD-1 signaling prevents CK2-mediated degradation of PTEN by reducing the expression of casein kinase 2 (CK2) (Patsoukis et al., 2013). In summary, USP7 can directly upregulate the SRSs of multiple cancers and enhance the effect of immunosuppressive cells.

USP14 has been extensively studied because of its overexpression in lung cancer (Han et al., 2019), breast cancer (Xia et al., 2019a), ovarian cancer (Luo et al., 2019), and oesophageal squamous cell carcinoma (Sha et al., 2019). It may promote tumorigenesis in approximately 61% of cancers (Liu B. et al., 2019). USP14 inhibitors have also shown treatment efficacy in research settings (Xia et al., 2019a). We focus on the mechanism of USP14 involved in the regulation of SRS and ITAIM. In terms of WNT/β-catenin signaling, USP14 is overexpressed in lung adenocarcinoma and promotes cancer cell proliferation by upregulating β-catenin (Wu et al., 2013). Another study found that USP14 deubiquitinates the k63-linked polyubiquitin and dissociates from DVL. The DVL is subsequently phosphorylated and then activates and upregulates Wnt signaling (Jung et al., 2013). USP14 also promotes cell adhesion-mediated drug resistance by enhancing Wnt signaling in multiple myeloma cells (Xu et al., 2017). It has been reported that the inflammation regulation mediated by USP14 through lipopolysaccharides is related to the activation of NF-κB (Liu et al., 2017). Furthermore, USP14 also accelerates the degradation of IκBα mediated by the proteasome to promote the activation of NF-κB (Li M. et al., 2019). In addition, Xu found that USP14 cause resistance to cisplatin through the activation of AKT signaling in gastric cancer (Fu et al., 2018). In summary, the research on USP14 has made great progress in delineating its roles in the proliferation, migration and autophagy of different cancer cells. Although USP14 inhibitors have also been shown to be useful in cancer treatment, broad space for research on USP14 with respect to ITAIM regulation remains open because of its activation and upregulation of SRS.

USP22 was thought to have a cancer-promoting effect, and its gene was confirmed as one of 11 genes called “death-from-cancer” (Glinsky et al., 2005). However, in recent studies, the cancer-promoting or anticancer effects of USP22 were found to depend mainly on its environment (Jeusset and McManus, 2017). In most tumors, high expression of USP22 predicts a poor prognosis (Hu et al., 2015; Lim et al., 2020), and USP22 also been shown to promote tumor proliferation, invasion and drug resistance through a variety of substrates, including PD-L1 (Huang et al., 2019; Huang X. et al., 2020; Wang et al., 2020), RAS activator son of sevenless 1 (SOS1) (Lim et al., 2020), BMI1 (Ma et al., 2017; Qiu G.Z. et al., 2018; Yuan et al., 2019; Qiu et al., 2020), MYC (Liu H. et al., 2019; Zhang K. et al., 2019), and COX-2 (Xiao et al., 2015). These substrates include inhibitory immune checkpoints, transcriptional activators and oncogenes. In non-small cell lung cancer, USP22 can directly deubiquitinate PD-L1 and regulate the level of PD-L1 through the USP22-CSN5-PD-L1 axis, resulting in decreased T-cell proliferation and activation (Wang et al., 2020). Furthermore, studies have shown that knocking out USP22 inhibits the growth of liver cancer in an immune-dependent manner. Tumor immunogenicity and lymphocyte infiltration were enhanced, and the efficacy of PD-L1 immunotherapy and CDDP were also improved (Huang et al., 2019). SOS1, another substrate of USP22, was found to activate RAS signaling and upregulate the PI3K/AKT pathway (Lim et al., 2020). The PI3K/AKT pathway is a carrier of SRSs and promotes Treg activation (Brandmaier et al., 2017). The stem cell factors BMI1 has been proven to be a downstream target of USP22 in glioma, gastric and colorectal cancer. It is upregulated by USP22 deubiquitination (Ma et al., 2017; Qiu G.Z. et al., 2018; Yuan et al., 2019; Qiu et al., 2020). The latest research found that inhibiting BMI1 can lead to the recruitment and activation of CD8⁺ T cells by stimulating IRF3. Targeting BMI1 may eliminate CSCs by silencing PD-L1, and BMI1 inhibition can also attenuate tumor metastasis and recurrence by activating endogenous immunity (Jia et al., 2020). Therefore, USP22 may create ITAIMs through the BMI1 signaling pathway. USP22 can exert its cancer-promoting effect by upregulating the nicotinamide phosphoribosyltransferase (NAMPT)/SIRT1, AKT and ERK signaling pathways involved with c-Myc activation (Liu H. et al., 2019; Zhang K. et al., 2019). In a report, Myc upregulated the expression of CD47 and PD-L1 on the surface of CSCs. CD47 inhibited phagocytosis by binding to sirpα on macrophages (Park and Surh, 2017). In a study using CRISPR to screen the ubiquitination modulators of FOXP3 in Tregs, it was found that USP22 and Atxn713, members of the deubiquitination module of the SAGA stain complex, were positive regulators of FOXP3. In other words, USP22 may inhibit the immune response to tumor cells by directly acting on FOXP3 (Cortez et al., 2020). Furthermore, Han also proved that USP22 can promote tumor development through deubiquitination and immunosuppression in lung adenocarcinoma (Han et al., 2020). USP22 was found to activate MED1 transcriptionally to regulate NK cell development through histone H2A deubiquitination (Zhang Y. et al., 2020).

In summary, the regulation of USP22 in CSCs and TAIMs is abundant and has the potential to suppress immunity in many ways. In some preclinical experiments, USP22 used as a regulator achieved encouraging effects (Li J. et al., 2020).

USP15 has a high degree of homology with the abovementioned carcinogenic USP4, and its gene upregulation has been found in glioblastoma, breast cancer and ovarian cancer (Chou et al., 2017). However, USP15 is the same as USP22 in that its role in cancer depends on the environment (Cheng et al., 2019). USP15 acts as a regulator of the natural immune response in the anti-infection processes of the human body (Torre et al., 2017; Kusakabe et al., 2019; Huang L. et al., 2020). It has been reported that USP15 can participate in regulating multiple stem cell factors, including MDM2 (Zou et al., 2014; Niederkorn et al., 2020), TET2 (Chen L.L. et al., 2020), BMI1 (Zhang L. et al., 2020), NF-κB (Zhou Q. et al., 2020), and TGF-β (Eichhorn et al., 2012). The aforementioned signaling pathways activated by these factors can promote the production of IATIM. MDM2 not only can promote tumor progression by downregulating P53 (Niederkorn et al., 2020) but can also negatively regulate Tregs by acting on the transcription factor NFATc2 (Zou et al., 2014). In breast cancer, USP15 is found to deubiquitinate and stabilize BMI1 protein at lys-81 (Zhang L. et al., 2020). Recent studies have also found that USP15 exerts a deubiquitination effect at the K1299 site of TET2 and negatively regulates the activity of TET2. Therefore, USP15 weakens the TET2-mediated IFN/JAK/STAT pathway and reduces the expression of chemokines and lymphocytic infiltration. Knocking out USP15 improves the response of mice with transplanted melanoma to immunotherapy and prolongs survival (Chen L.L. et al., 2020). In summary, USP15 mainly acts on stem cell factors to affect CSCs and ITAIM, other potential pathways need to be explored and verified.

CYLD plays a major role in cell development and proliferation, and it is also an important inflammation regulator (Hadian et al., 2011). CYLD inhibits the NF-κB pathway by deubiquitinating key factors, including NF-κB essential modulators IKKg and TRAF2 (Kovalenko et al., 2003). In contrast to other oncogenic DUBs of the USP family, CYLD is the DUB with the most typical tumor suppressor effects. Patients with high CYLD expression and multiple myeloma (van Andel et al., 2017), oral squamous cell carcinoma (Shinriki et al., 2018) or breast cancer (Hayashi et al., 2014) have a better prognosis. Several substrates of CYLD have been identified in recent studies, including DVL (Tauriello et al., 2010), TRAF (Brummelkamp et al., 2003), Bcl-3 (Massoumi et al., 2006), IKK (Chu et al., 2006), and p53 (Fernandez-Majada et al., 2016). Some of these substrates act as CSCs factors, which can activate and upregulate the SRSs. For example, the upregulation of CYLD causes DVL deubiquitination, significantly inhibiting the oncogenic Wnt pathway, thereby inhibiting the occurrence of tumors (Caliskan et al., 2017). The downregulation of CYLD promotes the accumulation of β-catenin in the nucleus through ubiquitination of DVL at Lys-63 (Tauriello et al., 2010). In addition, CYLD deubiquitinates TRAF and Bcl-3 proteins directly to downregulate NF-kB activity, inhibiting oncogenic transformation in keratinocytes (Brummelkamp et al., 2003; Massoumi et al., 2006). In salivary gland tumors, CYLD has a negative regulatory effect on NF-kB activity after TNF-α stimulation (Fukuda et al., 2008). Studies have also found that the downregulation of CYLD can promote breast cancer metastasis by activating NF-κB (Hayashi et al., 2014). In the inhibition process of TNF-related apoptosis-inducing ligand (TRAIL)-mediated NF-κB, the direct combination of CYLD and IKK enhances the effect of TRAIL (Chu et al., 2006).

The SRSs and CYLD influence each other. The upstream regulatory mechanism of CYLD transcription and translation has been extensively studied in tumors. A variety of oncogenic signaling pathways can regulate CYLD, and some of them belong to SRSs. Studies have shown that Hes-1, a downstream factor of Notch in T-cell leukemia, can reduce the expression of CYLD (D’Altri et al., 2011). STAT3 signaling can activate miR-21 (Iliopoulos et al., 2010) and miR-181b (Xu et al., 2016) to inhibit CYLD transcription. In chronic leukemia, lymphoid enhancer binding factor 1 (LEBF1) downregulates CYLD at the transcriptional level through the WNT/β-catenin signaling pathway (Liu et al., 2012). HH signaling in basal cell carcinoma downregulates the expression of CYLD through GLI1 under the action of the transcription repressor Snail (Kuphal et al., 2011).

In summary, CYLD, as a tumor-suppressing DUB, has a close relationship with CSCs. It potentially reverses the ITAIM by regulating the SRSs, and even improve the effect of immunotherapy.