The family of zinc finger proteins is composed of regulatory proteins that participate in many molecular and cellular pathways and is considered to be one of the most profuse regulatory protein families in eukaryotic cells, with more than 200 members. Different members of the zinc finger protein family contain multiple cysteine (Cys) and/or histidine (His) residues that require one or more zinc ions to stabilize their structures. The majority of the zinc finger proteins play important roles in DNA binding, RNA binding, RNA packaging, and protein–protein interactions. The family is divided into three sub-groups according to structure and the number of zinc fingers (Laity et al., 2001).

The zinc finger and BTB domain containing 16 (Zbtb16) gene was first described in humans in 1993, where its encoded protein was found to be fused in-frame with the retinoic acid receptor alpha (RARα) in a patient diagnosed with a rare form of acute promyelocytic leukemia (APL) with a reciprocal chromosomal translocation t(11;17). The protein was named the promyelocytic leukemia zinc finger (PLZF), because it was the first zinc finger family member exhibiting a pronounced effect on the pathology of promyelocytic leukemia (Chen et al., 1993).

In humans, the gene is localized to chromosome 11q23 among a cluster of genes, all related to the zinc finger family. The gene has three splice variants (transcripts), all encoding functional proteins, and composed of seven exons with shared homology between exons 3 and 6. The first transcript (ENST00000335953) is 2,477 bp in length and encodes a 673 aa protein. The second transcript (ENST00000392996) is 2,249 bp in length and encodes a 673 aa protein. The third transcript (ENST00000310883) is 1,728 bp in length and encodes a 550 aa protein. In mouse, the gene is located in chromosome 9, and is composed of seven exons. However, unlike its human analog, it only has one transcript (ENSMUST00000093852) that is 5,100 bp in length and encodes a 673 aa protein, which is similar in length to the human analog transcripts 2 and 3.

The N-terminus of the protein (as seen in Figure 1) contains a Drosophila-like bric à brac, tramtrack, and broad-complex domain called BTB (Zollman et al., 1994). This domain is also known as the poxvirus and zinc finger (POZ) domain and is highly conserved in mammals and plays a major role in DNA looping as well as protein dimerization and interaction to form multi-protein complexes (Bardwell and Treisman, 1994; Yoshida et al., 1999). The middle region of the protein has an RD2 domain that is less characterized and understood than the BTB/POZ domain, although mutations in this domain modulate the transcriptional activity of PLZF (Kang et al., 2003). The C-terminus of the protein contains nine Krüppel-like C₂H₂ zinc fingers that facilitate sequence-specific DNA binding to its target genes allowing PLZF to act as a transcriptional repressor (Li et al., 1997). The promoter region of human PLZF starts about 6 kb upstream of the ATG starting codon. The first TATA box in the human genomic DNA sequence for the PLZF gene appears at position 6,108 in the 5′ UTR. A very GC rich region is also found about 2 kb upstream of the ATG codon.

The involvement of PLZF in blocking the differentiation of promyelocytes and causing leukemia led the way for researchers to investigate its role in many biological activities regulating cellular proliferation and differentiation. Earlier reports confined the expression of PLZF to stem cells and early progenitor cells (Shaknovich et al., 1998), but it is now known that PLZF is expressed in many of the CNS cells, hematopoietic cells in the bone marrow, glandular cells in the gallbladder, islets of Langerhans in the pancreas, respiratory epithelial cells, myocytes in the heart and skeletal muscles, endometrial stroma in the uterus, glomeruli and renal tubules, glandular cells of many parts of the gastrointestinal tract, prostate, and endocrine glands (Uhlén et al., 2005).

A model of PLZF (ZBTB16) gene knockout was developed in 2000 by researchers at Memorial Sloan-Kettering Cancer Center (Cornell University). They replaced the second exon of full-length PLZF with the neomycin resistance gene. PLZF^−/− mice displayed major musculoskeletal-limb defects with homeotic transformations of vertebral segments, deformed cartilage and skeleton patterning, and alterations in digit formation that is caused by an inhibition of apoptosis in hindlimb early progenitor cells in the autopod region possibly due to Hox gene regulation (Barna et al., 2000). Moreover, PLZF was shown to be specifically essential for spatial colinear expression of HoxD genes that are expressed in the hindbrain. This was achieved by the active recruitment of histone deacetylases (HDACs), possibly through the PLZF-mediated interaction with polycomb family members such as Bmi-1, to the promoters of the Hox gene clusters affecting only the development of the posterior regions of the limb bud (Barna et al., 2002).

Another role for PLZF also emerged after assessing Gli3 and PLZF double knockout mouse embryos, where their cooperation was shown to be required for proximal cartilage condensations in the hindlimb by orchestrating the necessary dissemination of chondrocyte progenitor cells in the proximal limb bud, independent of proximal-distal patterning (Barna et al., 2005). PLZF expression is also important in osteoblastic differentiation, since it is an upstream regulator of core-binding factor 1 (CBFA1/Runx2). CBFA1 affects many important factors during the pluripotent differentiation of human mesenchymal stem cells (hMSCs) into the osteogenic lineage such as collagen 1A1 (COL1A1), alkaline phosphatase (ALP), and osteocalcin (OCN). Loss of PLZF expression during the early stages of pluripotent differentiation of hMSCs leads to the disruption of spinal ossification (Ikeda et al., 2005).

PLZF^−/− male mice show partial sterility, which is caused by markedly impaired spermatogenesis. Spermatozoal maturation in the seminiferous tubules was found to be distinctly decreased as a result of increased apoptosis in spermatogonial cells (Kelly and Daniel, 2006). This confirmed previous studies suggesting a dynamic role for PLZF in maintaining the undifferentiated state of spermatogonial cells, where the loss of this expression disrupts the balance between stem cell self-renewal and differentiation (Buaas et al., 2004). It was later confirmed that PLZF binds to the c-Kit promoter in a spatial and temporal-dependent fashion, to inhibit the maintenance and proliferation of postnatal germ cells (Filipponi et al., 2007).

PLZF^−/− mice were also found to be more prone to viral infections. Exposing neonatal wild-type and PLZF^−/− mice to the Semliki Forest virus (SFV) was lethal as measured by survival curves. However, pretreating the mice with interferon (IFN) for 6 h prior to infection protected the wild-type neonatal mice from infection, with survival times exceeding 3 weeks. On the other hand, mice lacking PLZF expression died after 6 days of infection, with viral loads in their vital organs about a thousand times higher than wild-type mice after only 48 h post-infection. This occurred despite the fact that both genotypes expressed comparable levels of IFNs after infection, showing that activation of certain anti-viral genes was impaired in the absence of PLZF. Similarly, introducing the encephalomyocarditis virus (EMCV) via intraperitoneal injections to wild-type and PLZF^−/− mice yielded equivalent results when comparing survival curves between the two genotypes (Xu et al., 2009). This demonstrates the requirement for PLZF in the IFN-mediated anti-viral innate immune response in vivo. The impairment of the IFN response was due to the failure to induce a subset of IFN-stimulated genes (ISGs), accounting for the fatal phenotype in PLZF^−/− mice. Moreover, IFN-induced activation of natural killer (NK) cells was also impaired in PLZF^−/− mice, which was attributed to the marked decrease in CXCL10 expression in NK-cell-rich organs in these animals (Xu et al., 2009).

One of the ways the cell uses to control the expression profile of different proteins is to regulate steps in the processing of the primary RNA transcript to mature mRNA. Control of RNA processing before protein synthesis is a very effective method for gene regulation and silencing, and allows for rapid and effective silencing of gene expression. Eukaryotic cells employ many processes to achieve this type of regulation, such as m7G 5′ capping, 3′ tail polyadenylation, and intron splicing. MicroRNAs are also key regulators of post-transcriptional events via binding complementary sequences in the 5′ or 3′ UTR of their target mRNA transcripts to block translation and/or cause Ago2-dependent degradation (Hansen et al., 2011).

Promyelocytic leukemia zinc finger has been shown to regulate the expression of miR-146a in vitro in a PLZF-positive HEL cell line. PLZF binds directly to the promoter of miR-146a and inhibits its transcription, leading to an up-regulation of CXCR4 protein. CXCR4 is a chemokine that is crucially involved in the mobilization of normal hematopoietic as well as leukemic cells. It is also required for the proliferation, differentiation, and maturation of CD34⁺ megakaryocytic-progenitor cells, as demonstrated by an increase in the levels of the megakaryocytic markers CD9, CD41, and CD61(Labbaye et al., 2008).

Promyelocytic leukemia zinc finger has also been shown to target the promoter region of miR-221/222 to inhibit their transcription in different PLZF-transduced melanoma cell lines. PLZF-positive melanocytes show a less malignant phenotype. Conversely, increased levels of miR-221/222, in PLZF-negative cells, correlate with increased tumorigenicity of these cells, as reflected by a decrease in G1 phase and a corresponding increase in S and G2-M phases, as well as an enhanced proliferation rate, invasiveness, motility, and anchorage-independent growth. The inhibition of these miRNAs causes an up-regulation of c-Kit, which is required for melanogenesis (Felicetti et al., 2008).

Interferons are a group of cytokines produced by the host cells after infections or oncogenic transformations. They perform a fundamental function in instigating the innate immune response and coordinating the profile of the acquired immune response against a variety of immunological stimuli. IFNs function by activating signal transducer and activator of transcription (STATs) factors to induce the Janus kinase (Jak)–STAT signaling cascade. As discussed earlier, Xu et al. demonstrated that PLZF was required for the anti-viral innate immune response in vivo. Using two renal adenocarcinoma cell lines (RCC1 and ACHN), they showed that PLZF up-regulated many immune response genes after IFN treatment and that those genes contained putative PLZF-binding sites in their promoter regions. Many ISGs failed to respond to high levels of IFN in PLZF^−/− mice after viral infection. The expression levels of Oas1g, CXCL10, Rsad2 (Viperin), and Ifit2 were significantly lower, showing the requirement of PLZF in the induction of the anti-viral response (Xu et al., 2009).

Natural killer cells are a group of cytotoxic lymphocytes that represent a powerful tool of the innate immune system. NK cells can spontaneously lyse cells lacking self-antigen or the major histocompatibility complex (MHC) class-I. PLZF expression affects the effector function of NK cells. NK cells from PLZF^−/− mice, although constituting the same percentage of total lymphocytes as compared to the wild-types, were functionally impaired. This was reflected by their failure to spontaneously lyse Yac-1 cells (an NK cell-specific target because of their missing MHC-1). This was also true after poly I:C-mediated IFN activation. Moreover, Granzyme B (GzmB) induction after IFN activation was also affected, with a lower percentage of PLZF^−/−splenic NK cells showing GzmB positive staining. This impairment of IFN activation was confirmed to be the direct result of lost PLZF expression, which was required by a specific subset of ISGs. In spite of the well-known role of PLZF as a transcriptional repressor, a novel activator role was described for the first time with regards to the IFN signaling pathway, where phosphorylated PLZF binds to the promoter regions of those ISGs and actively drives their transcription (Xu et al., 2009).

Natural killer T-lymphocyte (NKT) cells are a heterogeneous population of T cells that both express surface antigen markers designated for NK cells and recognize the Cd1d antigen that presents many glycoprotein and lipid molecules. Invariant NKT (iNKT) cells are a subgroup of NKT cells with distinctive properties. They express an invariant α chain of the T cell receptor (TCR), a specific TCR-β chain that uses the variable region 8.2, and they can respond to immunological stimuli by cytokine secretion within minutes (Matsuda et al., 2008). PLZF expression was found to be restricted to iNKT cells, with higher concentrations at early stages of thymocyte development and very low expression levels as cells differentiate toward maturation. Loss of PLZF expression in PLZF^−/− mice did not affect the number of mature iNKT cells, but did have a dramatic impact on their function. Most iNKT cells expressed low levels of CD44, CD8, NK1.1, DX5, NKG2D, CD244, CD24, and CD122. Conversely, they expressed high levels of CD69 and failed to remain in the thymus after maturation; instead accumulating in the liver, lymph nodes, and spleen. Functionally, PLZF-deficient iNKT cells failed to secrete both IL-4 and IFNγ after immunological challenge, a hallmark of NKT cells (Savage et al., 2008; Kreslavsky et al., 2009).

A similar scenario was observed in a subset of γδNKT cells that share phenotypic and functional properties with NKT cells and express PLZF endogenously. Vγ1 + Vδ6.3/Vδ6.4+ cells are the predominant subtype of γδNKT, which produce a repertoire of cytokines after innate immune stimuli. Loss of PLZF expression in those cells completely shut down their capacity to secrete these important cytokines. This functional deficiency was shown to be the result of impaired intrinsic development of thymocytes that leads to a reduction of activated CD44^HiCD62^Low cells in the bone marrow. Furthermore, TCR-mediated signaling was shown to be required for PLZF induction, which drives development of γδNKT cells (Kreslavsky et al., 2009).

In 1995, a study described the relatively high expression of PLZF inside the nuclear speckles of CD34⁺ progenitor cells purified from human bone marrow. However, the expression levels of PLZF decreased in these hematopoietic cells as they became more differentiated. The study also noted the possibility of PLZF being involved in a regulatory network of transcription factors to control hematopoiesis (Reid et al., 1995).

One of the important activators of the PLZF promoter in the early stages of hematopoietic development was found to be Evi-1. Evi-1, which is usually up-regulated in many leukemias, was found to bind, and subsequently activate, the promoter of PLZF in HEL, KG1, and K562 cells in vivo. Although this interaction did not affect the expression level of PLZF, a theory of short cis-acting sequences working in a tissue-specific manner was proposed (Takahashi and Licht, 2002).

In a more recent study, researchers used a xenograft model of immunodeficient mice to study the role of PLZF in early myeloid development. They performed bone marrow transplantation of lineage⁻ (lin⁻) umbilical human cord blood-derived myeloid progenitor cells that had been transduced with PLZF expression/knockdown vectors. They successfully demonstrated that PLZF was able to regulate the differentiation capacity of these myeloid progenitors. PLZF achieved this through the transcriptional modulation of many key transcription factors involved in myeloid differentiation, such as GFI-1, C/EBPα, LEF-1, DUSP6, and ID2. This demonstrates how PLZF-controlled cellular proliferation in early hematopoiesis is essential in establishing a balance between the numbers of progenitor and fully differentiated cells (Doulatov et al., 2009).

The PLZF (ZBTB16) protein was discovered in a patient with APL (which is characterized by abnormal accumulation of immature granulocytes). The cause of the disease was found to be a translocation t(11;17)(q23;q21) between RARα and PLZF. The fusion proteins caused the progression of the leukemia by suppressing many retinoic acid-regulated genes, thereby locking the differentiation of granulocytes into the “promyelocyte” stage. Specifically, the patient had a chromosomal rearrangement that led to a reciprocal translocation where two fusion gene products were produced: PLZF-RARα and RARα-PLZF. PLZF is located in chromosome 11q23 and RARα is located in chromosome 17q21. The first form of the fusion protein “PLZF-RARα” is composed of full-length PLZF until the second zinc finger of the protein, where a breakpoint joins RARα domains B–F. The second fusion protein RARα-PLZF binds the A domain of RARα with the remaining seven zinc fingers (3–9) of PLZF (Chen et al., 1993).

Promyelocytic leukemia zinc finger-RARα works by forming multi-protein complexes with other repressor molecules such as SMRT, Sin3, HDACs, and N-CoR to block myeloid differentiation. Chimeric mice expressing the fusion protein develop a retinoic acid-resistant leukemia that requires the addition of an HDACi to restore retinoic acid sensitivity (He et al., 1998). In vitro, dihydroxyvitamin D₃ (D3) and transforming growth factor β1 (TGFβ-1) drive the differentiation of the human monocytic U937 cell line. Similarly, dimethyl sulfoxide (DMSO) can induce the differentiation of the human promyelocytic leukemia HL-60 cell line. PLZF-RARα expression blocked the D3-TGF-induced monocytic and DMSO-induced granulocytic differentiation of both U937 and HL-60 cells, respectively (Ruthardt et al., 1997). These results suggest that the dominant negative effect of PLZF-RARα may involve the disruption of BTB/POZ-mediated obligate homodimerization, which can cause a loss of wild-type PLZF function. Additionally, U937 cells expressing PLZF-RARα were able to proliferate after prolonged retinoic acid treatment, which may be explained by the ability of PLZF-RARα to induce Myc expression and inhibit dual specificity phosphatase 6 (DUSP6) expression (Rice et al., 2009).

Promyelocytic leukemia zinc finger-RARα was also shown to interact with Bmi-1 and that interaction causes the fusion protein to lose its function. Both PLZF and PLZF-RARα were able to bind Bmi-1 in vitro to form the PRC1 polycomb complex, which can then be recruited to retinoic acid response elements (RAREs), transforming the chromatin in those regions insensitive to remodeling by HDACs. This could explain the non-responsiveness phenotype to all-trans retinoic acid treatment in PLZF-RARα-mediated APL patients (Spector and Lamond, 2011).

In a yeast two-hybrid screening, a new cervical cancer suppressor 3 (CCS3) protein was found to interact with the BTB domain of PLZF, reflecting the high expression of PLZF in normal cells and the absence of expression in cancer cells. CCS3 expression followed a similar pattern in human cervical cells. Restoring CCS3 expression in SiHa cervical cancer cells caused a dramatic change in morphology, a reduction in the numbers of viable cells in culture, and increased apoptosis. CCS3 was shown to achieve this by interacting with PLZF to repress PLZF-target genes, such as CCNA2 (cyclin-A2; Rho et al., 2006).

The androgen-independent prostate cancer cell line DU145 lacks endogenous PLZF expression and that expression could be restored by the ectopic expression of the androgen receptor (AR). Overexpressing PLZF in these cells caused a marked down-regulation of the pre-B-cell leukemia transcription factor (PBX1) and led to decreased proliferation and survival in vitro. However, this was not true in LNCaP cells, which are androgen-dependent prostate cancer cells. This reflects the fact that PLZF-mediated repression is androgen-dependent and indicates how DU145 could escape androgen-based chemotherapy (Park et al., 2003).

A similar study performed on primary malignant melanoma tumors from 41 patients found that PLZF is expressed in the majority of tumors in vivo, but this expression was lost in melanoma cell lines cultured in vitro. The study also showed that PBX1 expression increased as a result of low PLZF levels, which may contribute to the tumorigenesis of melanoma. Additionally, PLZF mRNA expression was used as a prognostic/survival index for patients (Brunner et al., 2008).

The expression of PLZF has been found to be frequently lost in the highly aggressive malignant mesothelioma cells as a result of recurrent focal chromosomal deletions. Restoring the expression of PLZF in these cells caused decreased viability in culture, reduced colony formation in soft agar, increased apoptosis, and an up-regulation of the cleaved Mcl-1 protein, which is a pro-apoptotic marker. Despite earlier reports connecting PLZF expression with Myc and cyclin-A2, no significant difference in the expression levels of these genes was noted in the study (Cheung et al., 2010).

An analysis of the upstream translated region of human PLZF revealed many putative binding sites for the CCAAT displacement protein (CUX1). In vivo interaction between CUX1 and thePLZF promoter was confirmed by electrophoretic mobility shift assay (EMSA) and chip analysis. CUX1 was also able to bind to the promoter region of PLZF in Caco-2/15 and HEK293 cells in vitro. A reporter assay showed that CUX1 blocked the transcriptional repression of PLZF in a dose-dependent manner. When assessing the PLZF gene expression profile in different colon cancer cells, an inverse relation was discovered between PLZF and CUX1 gene expression (Fréchette et al., 2010).

Promyelocytic leukemia zinc finger is a transcription factor implicated in major developmental and biological processes including spermatogenesis, hind limb formation, hematopoiesis, and immune regulation (Figure 2). PLZF achieves its transcriptional regulatory activity by interacting with different partners to form large multi-protein complexes that bind to the regulatory elements in the promoter region of the target genes. Recently, PLZF has become implicated in carcinogenesis as a tumor suppressor gene, because of its ability to regulate the cell cycle and apoptosis in many cell types.

Many biological functions of PLZF in differentiated cells remain to be determined. The number of regulatory factors controlled by PLZF or by its target genes is substantial. Although expressed in many tissue types, its expression level is low compared to the levels in stem cells or progenitor cells. However, this does not mean that it has a limited function in differentiated cells, as can be seen by the large cohort of biological functions listed in Table 2. What is more intriguing is the fact that PLZF expression is lost in many cancer cells (Figure 3). As discussed earlier, the loss of PLZF expression has been correlated to many malignant phenotypes in different cancer cell types, including increased proliferation, enhanced invasiveness and motility, and resistance to apoptosis. All of these features are required for cancer cell survival, therefore PLZF is considered to be a tumor suppressor gene. Thus the requirement to turn off the expression of PLZF for tumorigenesis is evident, but the exact mechanism has yet to be elucidated. Possible mechanisms include epigenetic modifications or up-regulation of other transcription factors in order to silence the unfavorable PLZF expression, microRNAs to block mRNA translation, or signals to reduce the stability of the PLZF protein.

Promyelocytic leukemia zinc finger can also regulate the expression of different microRNA molecules. This ability is important, because it can aid in the understanding of how PLZF achieves the transcriptional repression of many downstream targets. This also adds an extra level of regulation to the already known promoter-dependent transcriptional regulatory activities of PLZF. A single microRNA molecule not only can target a single pre-mature mRNA, but also can regulate a plethora of mRNA molecules, leading to an amplified transcriptional regulatory signal that could affect the basic properties of the cell, including proliferation rate, apoptosis resistance, migration, and invasiveness.

Understanding the biological activities of PLZF and its DNA/protein/miR interaction network would augment our understanding of its complex involvement in many pathways and may delineate the potential of fine tuning its expression profile in cancer cells. This would add to its value as a therapeutic target to modulate malignant cellular characteristics in oncological disease backgrounds.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.