Specificity protein 1

Human Sp1 is encoded by the SP1 gene located on chromosome 12 (Gaynor et al. 1993). Sp1 comprises 785 amino acid residues and is a member of the Sp transcription factor family (Beishline and Azizkhan-Clifford 2015). Sp proteins are similar to Drosophila melanogaster Krüppel protein, and Sp/Krüppel-like factors (KLFs) are well conserved among mammals (Wierstra 2008; McConnell and Yang 2010). Currently, nine human Sp and 17 human KLF proteins have been characterized (Beishline and Azizkhan-Clifford 2015; McConnell and Yang 2010). All KLFs have three well-conserved C₂H₂-type zinc finger domains (ZFDs) which are essential for their binding to double-stranded DNA (Fig. 1). Sp1 also contains multiple transcription activation domains comprising glutamine-rich and/or serine/threonine-rich regions (Fig. 1, designated as regions A and B). These domains interact with various proteins – including TBP, TAF4, and p300 – that function as transcriptional activators (Wierstra 2008; Beishline and Azizkhan-Clifford 2015). Sp1 associates with multiple proteins involved with diverse functions including general transcription factors, specific transcription factors, chromatin remodeling factors, tumor suppressors, DNA repair factors, cell cycle factors, and nuclear factors (Beishline and Azizkhan-Clifford 2015). However, the specific interfaces of these protein–protein interactions have, for the most part, not been identified (Beishline and Azizkhan-Clifford 2015). Additionally, Sp1 can oligomerize to enhance its transcription-stimulating capacity (Wierstra 2008). Domains C and D depicted in Fig. 1 are generally dispensable for the transcriptional activation ability of Sp1; however, they are required for synergistic activation of transcription when consensus binding sites are clustered in gene promoter regions (Wierstra 2008).

The primary function of Sp1 is the positive regulation of basal transcription. This function can be regulated via posttranslational modifications, including phosphorylation, glycosylation, acetylation, and SUMOylation (Wierstra 2008; Beishline and Azizkhan-Clifford 2015). These modifications can positively or negatively regulate the expression of target genes by affecting Sp1 functional activation, DNA binding activity, and localization. Additionally, physical interactions with other proteins and posttranslational modifications can render Sp1 more prone to proteasomal degradation (Wierstra 2008; Beishline and Azizkhan-Clifford 2015).

Sp1 is a ubiquitous transcription factor involved in the regulation of numerous important genes. Therefore, Sp1 is essential for normal cell processes such as cell cycle progression, DNA repair, inflammation, and metabolism (Wierstra 2008; Beishline and Azizkhan-Clifford 2015). Regarding diseases, the roles of Sp1 in cancer biology have been well described elsewhere (Wierstra 2008; Beishline and Azizkhan-Clifford 2015). Furthermore, Sp1 is overexpressed in multiple cancer types and contributes to malignant phenotypes (Beishline and Azizkhan-Clifford 2015). Sp1 is also involved in the overexpression of c-Myc and is associated with chromosomal rearrangement of the MYC locus in Burkitt’s lymphoma (Wierstra 2008). As described below, Sp1 also mediates diverse transcriptional regulations in cancer cells in response to hypoxia – a major factor within tumor cell biology (Koizume and Miyagi 2016). Additionally, Sp1 plays an important role in the progression of Huntington’s disease, as mutant huntingtin protein fails to interact with Sp1 (Wierstra 2008). This mutant huntingtin protein fails to activate cystathionine γ-lyase expression, resulting in insufficient levels of neuronal cysteine and thereby contributing to neurodegeneration (Paul et al. 2014).

Hypoxia inducible factors (HIFs) are major regulators of hypoxia-driven transcriptional induction (Koizume and Miyagi 2016). HIFs bind the constitutively expressed aryl hydrocarbon receptor nuclear translocator (ARNT) to form heterodimeric complexes. HIF–ARNT complexes can bind hypoxia response elements (HREs) within gene promoters to enhance transcription. However, Sp1 is also involved in inducible gene expression in response to hypoxia in both normal and disease-associated cells. Thus, hypoxia-associated effects on Sp1 likely contribute to the biology of conditions including cancer and cardiovascular diseases (Koizume and Miyagi 2016). Mechanisms of hypoxia-driven transcriptional activation via Sp1 are diverse and can be categorized into several groups (Koizume and Miyagi 2016). For example, Sp1 can regulate authentic HRE-dependent gene expression. Alternatively, Sp1 may activate transcription of target genes by multiple mechanisms in an HRE-independent manner. For example, physical interaction with HIFs, self-assembly, posttranslational modification, protein upregulation, and direct activation of HIF1A genes are all possible mechanisms for Sp1-driven transcriptional control in cells exposed to hypoxia (Koizume and Miyagi 2016). The relative roles of these Sp1-dependent mechanisms in cell biology compared with those of authentic HRE-driven expression are obscure. However, these effects may be cancer-type- and context-dependent, with further details requiring investigation.

Vertebrate genomes are methylated predominately on C residues at CpG sites. However, the promoter region of all house-keeping genes and approximately 40% of genes with tissue-specific expression patterns (70% of gene promoters in total) associate with typically 1–2 kb methylation-free genomic regions called CpG islands (CGIs) (Deaton and Bird 2011) (Fig. 3). This is important, as decreased methylation status closely correlates with the active state of chromatin, with a lack of methylation facilitating transcriptional induction (Fig. 3). CGIs are generally CpG-rich; thus, it is expected that CGI-associated genes are predominately regulated by Sp1 via GC-boxes (Fig. 3). Indeed, CGIs are enriched with Sp1 binding in vivo (Deaton and Bird 2011 and references therein). Earlier studies have reported that methylation itself does not interfere with Sp1 binding of its consensus binding sites (Holler et al. 1988; Harrington et al. 1988). Therefore, loss of Sp1 interaction with methylated consensus binding sites is not the cause of low expression levels of CGI-methylated genes. However, cytosine residues at CpNpG sites can be methylated in the mammalian genome (Clark et al. 1997). In contrast to CpG sites, methylation of CpNpGs can impair Sp1 binding to double-stranded DNA. These results suggest that cytosine methylation can affect Sp1 binding to consensus binding motifs; however, this depends on the specific nucleotide sequence present.

Open image in new window

Sp1 binding sites within CGIs are critical for maintaining CGIs in a methylation-free state in the promoter of the mammalian adenine phosphoribosyltransferase gene (Macleod et al. 1994; Brandeis et al. 1994), suggesting additional importance of Sp1 in the expression of CGI-associated genes. Further details of the Sp1-dependent demethylation process including whether this mechanism applies to other CGI-associated genes are unclear. However, active transcription at an early developmental stage and cooperative binding of Sp1 with other transcription factors via some cis-acting elements may correlate with protection of CGIs from de novo methylation (Deaton and Bird 2011). Additionally, Sp1 is likely involved in epigenetic gene regulation via chromatin components such as histone deacetylase and DNA methyltransferase (Beishline and Azizkhan-Clifford 2015; Koizume and Miyagi 2016). Overall, Sp1 exerts critical transcriptional regulation through both epigenetic means and simple interactions with target DNA sequences.

The CGIs of many tumor suppressor genes are heavily methylated in cancer cells, with transcriptional suppression occurring following formation of closed forms of chromatin structure (Baylin and Jones 2011). In this inactive chromatin state, Sp1 is expected to be evicted from the gene promoter region (Fig. 3). However, this does not necessarily mean that Sp1 is responsible for the demethylation of CGIs. Rather, Sp1 eviction could result from formation of condensed chromatin associated with the recruitment of methyl-binding domain proteins (Deaton and Bird 2011). Under these circumstances, many other transcriptional activators would also be unable to access gene regulatory regions.

Accumulating experimental evidence suggests that Sp1 may be involved in aberrant gene expression in cancer cells. This Sp1-dependent gene expression contributes to the aberrant protein production required for the expression of malignant phenotypes. However, in contrast to authentic HRE-dependent mechanisms, how and to what extent these Sp1-dependent mechanisms contribute to hypoxia-driven gene transcription in cancer cells is not fully understood. Inhibition of Sp1 function can be achieved by using mithramycin (Vizcaíno et al. 2015), available as a pharmaceutical compound. Mithramycin intercalates double-stranded DNA with GC-rich sequences in the nucleus, thereby blocking the association of these regions with Sp1. Additionally, several other chemical compounds, including curcumin and doxorubicin, downregulate Sp1 expression (Vizcaíno et al. 2015). Developing anticancer drugs targeting transcription factors such as Sp1 is challenging, as such proteins have traditionally been regarded as undruggable, because pharmacological inhibition of protein–protein and protein–DNA interactions occurring over a large surface area is generally difficult to design (Darnell 2002; Hagenbuchner and Ausserlechner 2016). However, such drugs may find clinical use in the future. Nevertheless, care should be taken to limit unwanted side effects given that Sp1 is critical for many normal cell functions and that present known Sp1 inhibitors are not as specific as would be ideal.

Sp1 is a member of the Sp/Krüppel-like factor family and is a crucial transcription factor regulating normal physiological cell functions. Sp1 is also critical for the maintenance of epigenetic state, as it affects the methylation status of CG sequences within gene promoter regions. However, Sp1 also contributes to the progression of diseases such as cancer via characteristic molecular mechanisms. Sp1 overexpression is associated with increased transcription of many cancer-related genes, and Sp1 mediates multiple and characteristic transcription profiles that are responsible for malignant phenotypes. The therapeutic potential of targeting Sp1 has been considered, with multiple pharmaceutical compounds able to inhibit Sp1 function. However, no compounds are clinically applicable at present. Therefore, a greater understanding of Sp1-dependent cancer biology is required. Additionally, the significance of Sp1 actions for normal cell functions means that future successful anti-Sp1 strategies will require careful design to prevent predicted undesired side effects.