PDLIM2: Signaling pathways and functions in cancer suppression and host immunity

Abstract

PDZ and LIM domains-containing proteins play pivotal functions in cell cytoskeleton organization, cell polarization and differentiation. As a key member of the family, PDLIM2 regulates stability and activity of transcription factors such as NF-κB, STATs and β-catenin, and thus exert it functions in inflammation, immunity, and cancer. PDLIM2 functions as a tumor suppressor in multiple tissues and it is often genetically mutated or epigenetically silenced in human cancers derived from lung, breast, ovarian and other histologies. However, in certain types of cancers, PDLIM2 may promote cancer cell proliferation and metastases. Therefore, PDLIM2 is added to a long list of genes that can function as tumor suppressor or oncogenic protein. During tumorigenesis induced by oncogenic viruses, PDLIM2 is a key target. Through promotion of NF-κB/RelA and STAT3 degradation, PDLIM2 enhances expression of proteins involved in antigen presentation and promotes T-cell activation while repressing multidrug resistance genes, thereby rendering mutated cells susceptible to immune surveillance and cytotoxicity mediated by immune cells and chemotherapeutic drugs. Intriguingly, PDLIM2 in alveolar macrophages (AMs) plays key roles in monitoring lung tumorigenesis, as its selective genetic deletion leads to constitutive activation of STAT3, driving monocyte differentiation to AMs with pro-tumorigenic polarization and activation. PDLIM2 has also been explored as a therapeutic target for cancer therapy. At the end of this review, we provide perspectives on this important molecule and discuss the future directions of both basic and translational studies.

1. Introduction

PDZ domains are fundamental building blocks in the organization of protein complexes [1,2]. They are composed of ~80–90 amino acids and form a classical two α-helical/six β-strand structure. In the mouse, 928 PDZ domains have been recognized in genome-encoded 328 proteins, which exist in single or multiple copies or in combination with other interaction modules [3]. By analyzing 72 PDZ domains corresponding to 2998 ligands, Tonikian et al. suggested the 16 classes of PDZ domains based on C-terminal motifs [4]. In addition to partner proteins, PDZ domains also bind phosphatidylinositides and cholesterol. Through their ligand interactions, PDZ domain proteins are critical for cellular trafficking and the surface retention of various ion channels. PDZ proteins can also contribute to cytoskeletal dynamics by mediating interactions critical for maintaining cell–cell junctions, cell polarity, and cell migration [1-5].

In contrast, the LIM domain is considered to be a tandem zinc-finger structure that functions as a modular protein-binding interface. LIM domains are present in many cellular and viral proteins that possess diverse normal cellular functions as regulators of gene expression, cytoarchitecture, cell adhesion, cell motility and signal transduction, and they are involved in pathological functions such as oncogenesis, leading to human disease [6,7].

The PDZ and LIM domains (PDLIM) are interacting protein structural modules shared by various proteins [8,9]. At this time, the family of proteins containing both PDZ and LIM domains consists of seven members, PDLIM1/Elfin, PDLIM2, PDLIM3/ALP, PDLIM4/Ril, PDLIM5/ENH, PDLIM6/ZASP/Cypher, and PDLIM7/Enigma [8,10,11]. Due to the presence of the PDZ domain, these proteins possess the ability to interact with certain other structural domains on other proteins, to exert different functions. These functions are often related to cell polarity, intercellular junctions, and control of proliferation, cellular migration, cell differentiation and recognition of immune cells [8-12].

The PDZ and LIM domain-containing protein 2, PDLIM2, also known as Mystique or SLIM, is classified as a member of the actinin-associated LIM family of proteins [13,14]. In human, the gene encoding for PDLIM2 is located on chromosome 8p21, and the protein is an IGF-IR-regulated adapter protein located at the actin cytoskeleton and necessary for the migratory capacity of epithelial cells [15]. Later a number of studies reached the conclusion that this protein is a cytoskeletal and nuclear effector that regulates the activity of several transcription factors (e.g., NF-κB, STATs), and its deregulation has been associated with oncogenesis.

In this review, we will provide an overview of the gene and its encoded protein PDLIM2, its tissue distribution, cellar localization, and general functions. Then we will focus on regulation of its expression, signaling pathways and functions of PDLIM2 in cancer initiation and progression, and in host immunity. Then, its potential use as a target for therapeutic applications will be discussed.

2. The gene, it encoded protein as E3 ligase and cellular localizations

PDLIM2 was identified in a cDNA library of transcripts expressed in the tissues of the rat eye irido-corneal angle by Torrado and his colleagues in 2004 [13], and by other teams and named as Mystique [14,15], or SLIM [16,17]. The human PDLIM2 is located on chromosome 8p21.2 [15], while the murine gene is located on chromosome 14.

As an early study have indicated that PDLIM2 was induced by the growth factor Insulin-like growth factor 1 (IGF-I), and by cell adhesion [15]. The production of this protein in turn, promotes cell attachment and migration and suppresses anchorage-independent growth in fibroblast cells. The authors concluded that Mystique/PDLIM2 is an IGF-IR–regulated adapter protein located at the actin cytoskeleton that is necessary for the migratory capacity of epithelial cells [15].

Cytokine signaling are controlled by several mechanisms, one of which is the ubiquitin-proteasome pathway, which modulates the turnover of cytokine receptors and activated Janus kinases (JAKs). The specificity of the ubiquitin pathway is achieved through various E3 ligase complexes that recognize and interact with distinct target proteins. E3 ligases confer specificity to ubiquitination of target proteins by recognizing target substrates and mediating transfer of ubiquitin from an E2 ubiquitin-conjugating enzyme to substrate [18]. Interestingly, LIM, RING, as well as PHD, are all zinc-binding domains that ligate two zinc ions. Unlike the classical zinc fingers, those domains do not bind to DNA, but rather function as mediator with other proteins [19]. While LIM domain proteins have diverse functions as regulators, RING domain proteins are usually associated with ubiquitination, and the domain is a feature of a class of E3 ubiquitin protein ligases. As for the interaction of these two domains, it is interesting to note that RING finger LIM-domain-interacting protein (RLIM) is an E3 ubiquitin ligase that leads to the ubiquitination and degradation of several LIM domain proteins, such as such as LIM-homeodomain transcription factors and LIM-only proteins [20].

As one key early finding, PDLIM2 has been identified as a nuclear ubiquitin E3 ligase that negatively regulates STAT signaling by Tanaka, Soriano and Grusby [16]. However, this enzyme activity of PDLIM2 has not been verified by another team independently. This is one of the essential tasks to be completed by investigators in the field. Assuming that it can be confirmed, due to its E3 ligase activity and key downstream targets including NF-κB and STAT proteins, it is not too surprising to find that this molecule would play important roles in host immunity and autoimmune diseases [16,21]. However, PDLIM2 E3 ligase activity is not consistent with findings in some other studies. For example, the LIM domain is not required for viral TAX repression, bur for bringing the protein to the nucleus [22,23]. In addition, another study found that another protein (MKRN2) that regulate p65 subunit of NK-κB is a novel ubiquitin E3 ligase [24]. Why would cells need two E3 ligases in the same signaling pathway?

Another protein working with PDLIM2 in the innate immune response is an enzyme named NAD(P)H:quinone oxidoreductase 1 (NQO1). This enzyme normally protects normal cells against oxidative stress and toxic quinones. Kimura and collaborators found a novel role of NQO1 in suppressing Toll-like receptor (TLR)-mediated innate immune responses [25]. NQO1-deficient macrophages selectively produced excessive amounts of IL-6, IL-12, and GM-CSF on lipopolysaccharides (LPS) stimulation, and the deletion of NQO1 in macrophages exacerbated LPS-induced septic shock. NQO1 interacted with the nuclear IκB protein IκB-ζ, promoted IκB-ζ degradation in a ubiquitin-dependent manner. Unlike IκBα in repressing NF-κB, IκB-ζ is required for the induction of certain NF-κB target genes. PDLIM2 participates in NQO1-dependent IκB-ζ degradation [25].

PDLIM2 can be localized in both cytoplasm and in the nucleus. The localization of the protein affects its functions. In monocytes/macrophages, sequestration of PDLIM2 in the cytoplasm is associated with adhesion and increased nuclear activity of NF-κB [26]. O'Connor and colleagues have also shown that PDLIM2 is highly expressed in a subset of triple negative breast cancer (TNBC). In those cancer cells, PDLIM2 was restricted to the cytoplasm/membrane of the cells and excluded from the nucleus [27]. In breast cancer cell lines, PDLIM2 retention in the cytoplasm was controlled by cell adhesion, and translocation to the nucleus was stimulated by insulin-like growth factor-1 or TGFβ. In another study, authors found that PDLIM2 plays some essential functions in cell polarization in 3D cultures by feedback regulation of the β1-integrin–RhoA signaling axis [28].

3. PDLIM2 function mostly as tumor suppressor in cancers

The functions of PDLIM2 have been investigated in a number of types of cancer. Our series of studies with genetic and epigenetic approaches convincingly demonstrated that PDLIM2 function as a tumor suppressor in breast cancer, colorectal cancer (CRC) and lung cancer [29-31]. We first showed that PDLIM2 expression is down regulated at transcriptional level via DNA hypermethylation at the gene promoter region in colon cancer [29]. Importantly, the PDLIM2 expression was sufficient to suppress in vitro anchorage-independent growth and in vivo tumor formation of these malignant cells. For this reason, we assigned a novel role of tumor suppressor to PDLIM2 for the first time [29,30]. In the case of breast cancer, PDLIM2 is repressed in both estrogen receptor-positive and estrogen receptor-negative breast cancer cells, suggesting one important mechanism for the constitutive activation of NF-κB [30]. Finally, we showed that its global or lung epithelial-specific deletion of PDLIM2 in mice causes increased lung cancer development [31].

A number of other studies from different groups, even though mostly through association, confirmed this concept and extended to other types of cancers. Jiang and coauthors first confirmed that PDLIM2 was downregulated in HCC tissues and cells and found that lower PDLIM2 expression was associated with worse prognosis in hepatocellular carcinoma (HCC) patients. Studies in vitro showed that PDLIM2 prevents the malignant phenotype of HCC cells by negatively regulating β-catenin [32]. There are some hints as tumor suppressor for other types of cancer. Zhao and colleagues demonstrated that PDLIM2 is decreased in both ovarian high-grade serous carcinoma and in various human ovarian cancer cell lines. Further analysis revealed that PDLIM2 is epigenetically repressed in ovarian cancer development and inhibition of PDLIM2 promoted ovarian cancer growth both in vivo and in vitro via NOS2-derived nitric oxide signaling, leading to recruitment of M2 type macrophages [33]. One study suggests that PDLIM2 acts as tumor suppressor in gastric cancer [34]. In summary, these studies together suggest that PDLIM2 is a tumor suppressor in other cancer types in addition to breast, lung and ovarian cancers.

It is important to point out that contradictory evidence exists as PDLIM2 may act to promote a tumor growth and metastasis in some cancers [35]. It is clear that PDLIM2 is highly expressed in some invasive cancer cells [36]. High levels of expression of PDLIM2 and PDLIM7, especially PDLIM7, could serve as adverse prognostic factors for acute myeloid leukemia [11]. Global proteome and phospho-proteome analysis showed that PDLIM2 was a commonly upregulated protein in both tumors and that knockdown of PDLIM2 leads to significant reductions in cellular proliferation in merlin-deficient meningioma and schwannoma [37]. As stated that PDLIM2 is highly expressed in certain invasive cancers. In this case, other distinct complex proteins and signaling pathways may dictate the functions of PDLIM2 [38]. In addition, PDLIM2 suppression efficiently reduces tumor growth and possibly metastasis of prostate cancer [39], and kidney cancer [40]. At this time, it is unclear how PDLIM2 could function as tumor suppressor in normal cells while promote tumor cell proliferation and metastasis in certain cancer cells. Yet it is clear that PDLIM2 may act on certain targets and signaling pathway to inhibit transformation of normal cells, while on different targets and signaling pathways of cancer cells to promote their growth. There is a long list of genes that can function as tumor suppressor in normal tissue and promote tumor growth in cancer cells [41,42]. The exact mechanisms for PDLIM2 to play dual roles need further studies in the near future.

4. PDLIM2 as a key target in virus-induced tumorigenesis

Seven human viruses cause about 10–15% of total human cancers. This diverse group of viruses reveals unexpected connections between innate immunity, immune sensors and tumor suppressor signaling network that control both viral infection and cancer development [43]. Infection by these oncogenic viruses often lead to an innate immune response that must be rapid and broadly targeted. However, it must be tightly regulated to avoid the detrimental effects of unregulated expression of inflammatory cytokines (such as interferons) to the host. Multiple mechanisms act in such tight regulation. The alternate way is that viruses trigger host negative regulatory mechanisms to suppress the host innate immune response. We will discuss relevant events related to infection of human papillomavius (HPV), Hepatitis C virus (HCV), Human T-cell leukemia virus type I (HTLV-I), and Kaposi sarcoma herpesvirus (KSHV) and dynamic changes of PDLIM2 in the host cells.

HTLV-I Tax oncoprotein deregulates cellular signaling for oncogenesis and induces leukemic transformation [44,45]. Our study showed that Tax was negatively regulated by PDLIM2, which promoted Tax K48-linked polyubiquitination. Consistently, PDLIM2 suppressed Tax-mediated signaling activation, cell transformation, and oncogenesis both in vitro and in animals [46]. We have also shown that HTLV-1-mediated repression of PDLIM2 involves transcriptional inhibition through promoter DNA methylation but independent of the viral oncoprotein Tax [22]. We further studied molecular determinants of PDLIM2 in suppressing HTLV-I Tax-mediated tumorigenesis. PDLIM2 binds to Tax directly and mediated by a putative α-helix motif of PDLIM2 at amino acids 236–254. Although the C-terminal LIM domain of PDLIM2 was not required for Tax binding, it was important for PDLIM2 to interact with the nuclear matrix. Thus, the LIM domain was essential for PDLIM2-mediated Tax repression [23].

KSHV is the most common cause of malignancies among AIDS patients. However, the mechanisms by which KSHV induces tumorigenesis remained to be uncovered. Recently, we demonstrated that one key underlying mechanism is through transcriptional repression of PDLIM2 gene [47]. It needs to be emphasized that PDLIM2 downregulation is essential for KSHV-induced persistent activation of NF-κB and STAT3, and subsequent tumorigenesis and tumor maintenance. Our mechanistic studies indicated that PDLIM2 transcriptional repression by KSHV involves DNA methylation in the promoter. Some readers may wonder, how does KSHV regulate DNA methylation? Many viruses encode proteins to reprogram host gene through epigenetic mechanisms [48]. For KSHV, this is regulated at several levels. One key factor encoded by the virus, the latency-associated nuclear antigen (ORF73), leads to CpG methylation by interacting with DNMT3a, the cellular de novo DNA methyltransferase, and recruiting it to certain cellular gene promoters that become methylated and repressed [49]. In this context, it is worth stating that hypomethylation-mediated oncogene activation, and conversely hypermethylation-mediated tumor suppressor gene silencing, are known to be one of the major mechanisms of cancer development [50]. In summary, our study not only improved our understanding of KSHV pathogenesis but also provided therapeutic strategies for KSHV-mediated cancers [47].

Other investigators found that HCV suppresses the innate immune response by upregulation of PDLIM2, independent of the host interferon response [51]. As for HPV infection, it was found that PDLIM2 was elevated about 3.25-fold when comparing HPV16 clearer to HPV negative; about 2.4-fold when comparing HPV 16 -persister to HPV negative [52]. How this upregulation of PDLIM2 would be related to later transformation of liver cells and cervical cells remains to be investigated.

Three viral oncoproteins, HTLV-I Tax, adenovirus type 9 E4-ORF1 and high-risk human papillomavirus E6, encode a related carboxyl-terminal PDZ domain-binding motif (PBM) that could mediate interactions with some cellular PDZ proteins. Later studies indicated that many other viruses also encode PBM-containing proteins that bind to cellular PDZ proteins [53]. Cellular PDZ protein targets of viral PBMs commonly control tight junction formation, cell polarity establishment, and apoptosis. In future, it would be interesting to see if PDLIM2 could interact with other viral PBM proteins and modulate its functions in cellular polarity and host cell survival.

5. Signaling pathways and potential functional partners

What signals can regulate PDLIM2 under normal physiological and pathophysiological conditions? The first set upstream signals are epigenetic and genetic regulations. PDLIM2 transcription can be regulated by multiple epigenetic mechanisms. The PDLIM2 gene promoter is methylated in lung cancer and in several other cancer types [29,30,54]. We showed that promoter DNA hypermethylation downregulates the transcription of PDLIM2 in T-cell leukemia and breast cancers [22,29,30]. Zhao and colleagues showed that this epigenetic mechanism is also involved in silencing PDLIM2 expression in ovarian cancer [33]. Wurster et al. have reported that expression of PDLIM2 at both mRNA and protein levels is lost in the majority of classical Hodgkin lymphoma (cHL) cell lines and Hodgkin and Reed–Sternberg (HRS) cells of nearly all cHL primary samples. This loss is associated with PDLIM2 genomic alterations, promoter methylation and altered splicing [55]. MicroRNAs (miR-221 or miR-222) negatively regulate PDLIM2 by binding to the 3′ untranslated region of the mRNA in colorectal cancer cells [56]. In addition, PDLIM2 acts as tumor suppressor and long noncoding RNA OR3A4 also downregulates PDLIM2 expression in gastric cancer [34].

Vitamin D can upregulate the transcription of PDLIM2 gene in breast cancer cells by two mechanisms [57]. First, vitamin D is a pleiotropic hormone that exerts its effects on a wide range of tissues. It acts through vitamin D receptor (VDR) dependent and an element responsive to VDR has been identified in the promoter of the gene. Secondly, vitamin D also induces demethylation of the PDLIM2 gene promoter, leading to enhanced transcription.

Many cancers arise from sites of infection, chronic irritation and inflammation [58]. Under these physiological and pathological conditions, reactive oxygen species (ROS) and reactive nitrogen species (RNS) are produced in cells [59,60]. It has been hypothesized long ago that ROS and RNS are responsible for the association between chronic inflammatory diseases and increased tumor incidence [61]. A few years ago, a team showed that myeloid cell-derived ROS could indeed induce mutagenesis in epithelial cells [62]. ROS induced epigenetic and genetic changes in aging cells and during human carcinogenesis [63]. As aforementioned, mechanisms leading to decreased or no expression of PDLIM2 in lung and other cancer cells include both genetic and epigenetic mechanisms in both human and mouse [30,31,64], that may be induced by inflammation and other environmental stress that produce high levels of ROS and RNS. These epigenetic and genetic changes take place in a kinetics with days or even months. In contrast, another ROS-induced signaling pathway may happen within hours. ROS may inhibit KEAP1, which in turn activate NRF2 signaling pathway [65]. Nrf2 activation can drive macrophage polarization and cancer cell epithelial-mesenchymal transition (EMT) during interaction [66]. Indeed, recent studies suggest that ROS-led NRF2-BACH1 pathway promotes lung cancer metastasis [67,68]. During lung tumorigenesis, PDLIM2 expression in macrophages is downregulated by ROS-activated transcription repressor BTB and CNC homology 1 (BACH1) [69]. Based on these and other studies, a schematic presentation of putative signaling pathways initiated from ROS/RNS to the inhibition of PDLIM2 expression is illustrated in Fig. 1. However, it is important to point out that more studies in epithelial cells before we have a full understanding on the signaling pathways of and roles of ROS in tumorigenesis in epithelial cells.

Fig. 1. A schematic illustration of a model of signaling transduction pathways leading to suppression of PDLIM2 expression in chronic inflammation and tumorigenesis.

The chronic inflammation (such as cigarette smoking) leads to enhanced expression of ROS/RNS, which in turn induce rapid NRF2-BACH1 signaling pathways in monocytes and macrophages, as well as in epithelial and cancer cells, that inhibit the expression of PDLIM2. Mean-while, ROS/RNS can induce epigenetic and genetic changes in epithelial cells and/or tumor cells, and results in inhibition of PDLIM2 expression in these cells, albeit with slower kinetics.

Much better evidence showed three key signaling molecules downstream of PDLIM2: RelA/NF-κB (in the nucleus), STATs, and β-catenin. NF-κB could be the epicenter of this pathway. Understanding the regulatory mechanisms for the NF-κB is key to control inflammation [70,71]. IκBα maintains NF-κB in an inactive form in the cytoplasm of unstimulated cells, whereas nuclear NF-κB in activated cells is degraded by PDLIM2 [21]. Tanaka and the team have made big efforts to search for other proteins that can regulate the p65 subunit of NF-κB. They have made three findings in vitro. First, it is interesting to note that MKRN2 is a novel ubiquitin E3 ligase for the p65 subunit of NF-κB in the nucleus and negatively regulates inflammatory responses [24]. Secondly, they found that PDLIM1, another member of the PDLIM family, inhibits NF-κB-mediated inflammatory signaling by sequestering the p65 subunit of NF-κB in the cytoplasm [72]. Third, they found that PDLIM7 synergizes with PDLIM2 and p62/Sqstm1 to inhibit inflammatory signaling via promotion of degradation of the p65 subunit of NF-κB [73]. It would be interesting to investigate if these four molecules would co-exist in the same cancer cells and how they would work together as a team in regulating NF-κB activity in anti-inflammatory signaling.

The second downstream target is STATs. There are six members in the STAT family, important molecules in immunity and other fundamental biological processes [74]. So exactly which members are the targets of PDLIM2? So far, STAT1 [16], STAT2 [51], STAT3 [47,75] and STAT4 [16,64], have been demonstrated to be the downstream targets for PDLIM2. No connections of STAT5 and STAT6 as PDLIM2 targets have been reported yet.

β-catenin may be a third downstream target. O'Connor, Cox and collaborators showed that PDLIM2 is expressed in about 60% of triple-negative breast cancer (TNBC) [27]. PDLIM2 is restricted to the cytoplasm in TNBC tumors, and its shuttling from cytoplasm to nucleus is stimulated by adhesion and growth factors. Even more importantly, PDLIM2 expression is sufficient to activate β-catenin. Using shRNA technology, the authors showed that PDLIM2 suppression inhibits 3D and colony formation in vitro and growth of TNBC cells in vivo. In contract, as discussed earlier, another study suggests that PDLIM2 may negatively regulate β-catenin and function as tumor suppressor in hepatocellular carcinoma (HCC) [32]. These authors showed that PDLIM2 is downregulated in HCC tissues and cells relative to normal counter-parts. The authors concluded that β-catenin is a downstream effector of PDLIM2-mediated inhibition of cell invasion and metastasis of HCC. At this time, we are still puzzled on how PDLIM2 activates β-catenin in TNBC while it suppresses β-catenin in HCC.

Oh et al. have identified the loss of PDLIM2 in metastatic colorectal cancers (CRC) by exome and transcriptome sequencing [76]. However, cancer cells learn tricks to overcome this roadblock. As it turned out, miR-221 and miR-222 act in a positive feedback loop to increase expression levels of RelA and STAT3 in CRC cells [56]. Human colorectal cancer tissues had higher levels of miR-221 and miR-222 than nontumor colon tissues; increases correlated with increased levels of RelA and STAT3 mRNAs. Levels of PDLIM2 mRNA were lower in CRC than nontumor tissues.

In another study, authors explored the use of exosome-mediated miR-222 transferring for insight into NF-κB-mediated breast cancer metastasis [77]. These authors found that exosomal miR-222 is highly expressed in breast cancer patients with lymphatic metastasis. The miRNA miR-222 promotes the aggressivity of breast cancer cells by directly downregulating PDLIM2 leading to activation of NF-κB signal pathway [77]. Cox and others showed that PDLIM2 can be a marker of adhesion and β-catenin activity in triple-negative breast cancer [27]. This study showed that PDLIM2 expression defines a subset of triple-negative breast cancer that may benefit from targeting the β-catenin and adhesion signaling pathways.

Studies have suggested complex functions of PDLIM2 in breast cancer. We reported that PDLIM2 is repressed in both estrogen receptor-positive and estrogen receptor-negative breast cancer cells, suggesting one important mechanism for the constitutive activation of NF-κB [30]. In another study, authors explored the use of exosome-mediated miR-222 transferring for insight into NF-κB-mediated breast cancer metastasis [77]. These authors found that exosomal miR-222 is highly expressed in breast cancer patients with lymphatic metastasis. The miR-222 promoting the aggressivity of breast cancer cells by directly downregulating PDLIM2 leading to activation of NF-κB signal pathway [77].

6. The role of PDLIM2 in host immune system

Studies have implicated important roles for PDLIM2 in the host immune system. These include the host immunity against invading viruses, innate and adaptive antitumor immunity against precancerous cells and tumor cells. Data at both the mRNA expression and protein level profiling indicate that it is expressed at enhanced levels in lymphoid tissues [78], suggesting that this protein may play some important functions in the immune system and host immunity.

6.1. PDLIM2 on innate and adaptive immunity

Investigators have explored the functions of PDLIM2 in the host immunity. Genetic studies in mice revealed that PDLIM2 is not required for the development of immune cells and immune tissues/organs [16]. However, studies with viruses have shown that upregulation of PDLIM2 suppresses host innate immune responses [51]. In a different kind of study involving innate immune response, PDLIM2 protects articular chondrocytes from lipopolysaccharide-induced apoptosis, degeneration and inflammatory injury through down-regulation of NF-κB signaling. The overexpression of PDLIM2 decreased LPS-induced production of interleukin (IL)-1β, IL-6 and TNF-α [79].

Mechanistically, PDLIM2 inhibits innate immunity by inhibiting Toll-like receptor (TLR) signaling through the degradation of the p65 subunit of NF-κB. It promotes K48-linked polyubiquitination on p65 peptide through the LIM domain of PDLIM2 [21]. PDLIM2 inhibits T helper 17 (Th17) cell development and granulomatous inflammation through degradation of STAT3 [75]. Results from study using PDLIM2 knockout mice showed that PDLIM2 deficiency in CD4⁺ T cells enhances Th1 and Th17 cell differentiation but has no obvious effects on Th2 and Treg cell differentiation. Interestingly, PDLIM2-deficient mice displayed increased susceptibility to experimental autoimmune encephalitis (EAE). PDLIM2 expression in CD4⁺ T cells is critical for EAE suppression as shown by adoptive CD4⁺ T-cell transfer [64].

Our recent study demonstrated that PDLIM2 in cancer cells enhances expression of genes related to antigen presentation and promotes T-cell activation, and represses genes conferring multidrug resistance and promotion of cancer-progression, thereby rendering mutated cells and tumor cells vulnerable to immune surveillance and improved therapies with immune cells and chemotherapeutic drugs [31].

Even though some progresses have been made regarding to functions of PDLIM2 in immunity, these pieces of evidence are fragmented, and more systemic investigations are needed. It would be highly important to study roles of PDLIM2 in T cell-mediated immunity with PDLIM2-knockout specifically in T cell lineages.

6.2. The functions of PDLIM2 in alveolar macrophages and lung tumorigenesis

We have demonstrated that, PDLIM2 functions as a tumor suppressor particularly important in lung tumorigenesis, through promoting the ubiquitination and proteasomal degradation of nuclear STAT3 and RelA [31,47]. However, how tumor cells interact with immune cells, particularly innate immune cells, and what roles they play in this process was not well studied yet. We recently showed that PDLIM2 can function as a checkpoint of alveolar macrophages (AMs) important for lung tumor suppression [69]. Using mice with PDLIM2 knockout-restricted in myeloid cells (PDLIM2^mKO), we showed that AM-intrinsic PDLIM2 expression is critical for phagocytosis function of those macrophages and for restricting the pro-tumorigenic activation and CTL suppression activity of AMs during lung tumorigenesis, spontaneously or induced by carcinogen urethane (Figs. 1 and 2). Mechanistically, myeloid-PDLIM2 exerts a lung tumor-suppressive role mainly through targeting STAT3. Finally, as we have discussed in the previous Section, PDLIM2 expression in AMs is downregulated by a transcription repressor BACH1 which is activated by ROS. Therefore, chronic inflammation-induced ROS/RNS and thus reduced PDLIM2 in myeloid cells may be one key mechanism that promotes lung tumorigenesis.

Fig. 2. A model showing how PDLIM-2 expression in tumor-associated macrophages impacts tumorigenesis process in the lung.

PDLIM2 expression in alveolar macrophages (AM) induce the expression of both CD80 and PD-L1 that promote phagocytosis of these macrophages and get rid of mutated cells. AM-intrinsic PDLIM2 expression is critical for phagocytosis function of those macrophages and for restricting the pro-tumorigenic activation and CTL suppression activity of AMs during lung tumorigenesis, spontaneously or induced by carcinogens. However, when cells are exposed to chronic inflammation such as smoking, AMs produces ROS and trigger signaling pathways that would downregulate the expression of PDLIM2 in AM, polarize these cells to tumor-promoting phenotype. Overall, this leads to enhanced tumorigenesis in the lung. This figure is modified from Li L. et al., JCI Insight, 2021 [69].

Most recently, we showed that AMs inherently express PD-L1 for optimal protective immunity and tolerance [80]. The constitutively expressed PD-L1 renders AM superior phagocytic ability, and the capacity to repress CTLs by cis- and trans-interacting with CD80 and PD-1, respectively. These findings uncover a unique characteristic of these macrophages and an innate immune function of PD-L1 and CD80, thus open up a new window for further understanding of lung physiology, pathology and PD-L1-based immunotherapy for cancer.

7. PDLIM2 as a therapeutic target in combination regimens of cancer therapeutics

The potential use of PDLIM2 as a target in cancer therapeutics just began to be explored. In those cancers where PDLIM2 functions as a tumor suppressor gene and often downregulated by epigenetic mechanisms as shown in lung and breast cancers [29,30], and in HTLV-1 virus-induced leukemia [23,49], we have shown a causative role of PDLIM2 epigenetic repression in lung cancer, and this downregulation led to resistance to chemo- and immuno-therapies [31]. In addition, vitamin D could upregulate PDLIM2 transcription through demethylation of its methylated promoter in cancer [57]. It is feasible to apply inhibitors for DNA methylation and restore the expression and functionality of PDLIM2 in those cancers. Some of these small molecule inhibitors have been approved by FDA and other authorities to treat patients with certain types of cancer [81,82]. Our study has shown a great potential of triple combination of epigenetic drugs, chemotherapeutic drugs and immune checkpoint blockade for lung cancer (Fig. 3) [31]. In this context, it is worth pointing out that epigenetic drugs are anti-cancer agents with immunomodulating potential, highly useful in combination immunotherapy [83].

Fig. 3. Mechanistic rationale for triple combination therapy strategy for lung cancer using anti-PD-1, chemotherapeutic drug and epigenetic agent.

The upregulation of PDLIM2 by an epigenetic drug in cancer would inhibit the levels and activities of NF-κB/RelA and STAT and sensitize the cancer cells to chemotherapy and T cells-mediated immunotherapy that is enhanced by anti-PD-L1 checkpoint blockade. As for example of epigenetic drugs, they can be inhibitors of DNA methylation, inhibitors of histone deacetylases (HDACs), or even vitamin D that possesses the function of DNA demethylation. This figure is modified from Sun F. et al., Nat Comm., 2019 [31].

It is possible to deliver the PDLIM2 gene by nanoparticles to the tumor tissue in order to restore its tumor suppressor function in cancer cells and/or AMs. Our ongoing study shows promising therapeutic results in multiple tumor models (Sun F et al., manuscript in preparation). Alternatively, it could be delivered using a viral vector (e.g., oncolytic virus) in order to achieve synergistical action for optimized immunotherapy of cancer [84,85].

8. Concluding remarks and future perspective

Evidence clearly shows that PDLIM2 functions in host innate immunity. PDLIM2 may inhibit host innate immunity during viral infection [51]. It inhibits Th17 cell development [75], and expression of PDLIM2 in CD4⁺T cells inhibits Th1 and Th17 cell development [64]. Yet there is little direct evidence on how PDLIM2 would affect antitumor adaptive immunity, an area worth to be further explored in the near future. Indeed, many critical questions remain unanswered at this time. One major aim of this review is to discuss some challenging questions for us to ask and answer in the near future.

Accumulating evidence shows that PDLIM2 functions as a tumor suppressor in a number of histological tissues. The most compelling evidence comes from our genetic and epigenetic studies in lung tumorigenesis and in other cancer models [29-31]. In PDLIM2^mKO mice, it was revealed that AM cell-intrinsic PDLIM2 promotes phagocytosis and restrict pro-tumorigenic activation of these AMs and also suppress the activity of CTLs in lung tumorigenesis either naturally or with urethane induction [69]. However, a better genetic model is to construct conditionally PDLIM2-knockout in epithelial cells in mice and examine the frequency and kinetics of tumor initiation and progression from these PDLIM2-KO epithelial cells. We hope that these types of data would be available from our group and/or from other investigators soon. As for other tumor models, Shi et al. have recently shown that PDLIM2 acts as a tumor suppressor in non-small cell lung cancer via down-regulation of NF-κB signaling in xenograft model [86]. PDLIM2 also suppresses HTLV-1-mediated leukemia development [22,46]. Its role as tumor suppressor in other types of cancer will be strengthened with more direct evidence.

Studies have shown that a protein can function as both tumor suppressor and oncoprotein, depending on the context. Decades ago, adenovirus E1A was originally discovered as an oncoprotein, yet further studies showed that it can function as a tumor suppressor [87]. Recent studies have uncovered many genes (at least 50) that possess both tumor-suppressor and oncogenic functions [41,42]. According to Datta et al., certain tumor suppressor genes may act as “double agents”, playing contrasting roles in vivo, “where either due to haploinsufficiency, epigenetic hypermethylation, or due to involvement with multiple genetic and oncogenic events, they play an enhanced proliferative role and facilitate the pathogenesis of cancer” [42]. In the case of PDLIM2, we know that the promoter of the gene can be hypermethylated, and then it participated in oncogenic pathways such as NF-κB-related tumorigenesis. However, exactly how it can function as oncogenic protein in certain cancer cells needs careful investigation.

The behaviors of PDLIM2 in biological processes could be explained at least partially by those of three well-characterized downstream targets NF-κB, STATs and β-catenin. NF-κB plays Yin-Yang functions in cancer [70,71]. NF-κB family of transcription factors is activated by canonical and non-canonical signaling pathways, which differ in both signaling components and biological functions. Both canonical and non-canonical NF-κB pathways function in immunity and inflammation [88]. It has been known that STAT proteins [89], and β-catenin [90] also play Yin-Yang roles under certain conditions.

In perspectives, the exact mechanisms by which PDLIM2 functions as a tumor suppressor and in host immunity remain to be further elucidated. Clearly, the actions of three downstream targets alone cannot provide all of the answers. There are several other regulatory pathways that act to regulate its functions in different cell types and in different environments. One such possibility is the post-translational modifications (PTMs). PTMs have shown to play key roles in the functions of tumor suppressor proteins such as p53, Rb and PTEN [91]. As a tumor suppressor protein, PDLIM2 could well have similar PTMs. It is highly possible that there are more key downstream signaling molecules to be discovered and characterized. Proteomic approach has been proven to be powerful in leading to these types of discoveries. Using this approach, not only potential post-translational modifications of PDLIM2 could be defined, but also new protein partners could be found, and characterized in combination with other biochemical and molecular approaches.

Alternatively, more cutting-edge technologies of modern biology and medicine, such as CRISPR-Cas9 system [92] and single-cell sequencing-based technologies [93], are widely available now, it is feasible to study gene functions in whole animals and to a degree in human patients too. It is exciting and high time to continue investigating on this protein and its gene, its partners and signaling networks, and we look forward to exploring more of its functions and mechanisms in cancer development and host immunity, and its potential use as a therapeutic target in cancer therapy.

Abbreviations

AMs

alveolar macrophages

BACH1

BTB and CNC homology 1

CTL

cytotoxic T lymphocytes

DNMT

DNA methyltransferase

HPV

human papillomavirus

HTLV-I

Human T-cell leukemia virus type I

HCV

hepatitis C virus

HDAC

histone acetylase inhibitor

KSHV

Kaposi sarcoma herpesvirus

LPS

lipopolysaccharides

NF-κB

Nuclear Factor kappa-light-chain-enhancer of activated B cells

NQO1

NAD(P)H:quinone oxidoreductase 1 (NQO1)

PBM

PDZ domain-binding motif.

PD-1

Programmed cell death protein 1

PD-L1

Programmed death-ligand 1

PDLIM2

PDZ and LIM domain-containing protein 2

PDLI2 KO

PDLIM2 knockout

PDLIM2^mKO

mice mice with PDLIM2 knockout-restricted in myeloid cells

RNS

Reactive nitrogen species

ROS

Reactive oxygen species

STAT

Signal transducer and activator of transcription

TLR

Toll-like receptor

TNBC

Triple negative breast cancer