The TP53 gene is highly conserved among eukaryotic organisms. In humans, it is located on chromosome 17 (17p13.1) (Lane et al., 2010) and comprises approximately 20,000 base pairs, consisting of 11 exons and 10 introns. The p53 protein is recognized as a crucial tumor suppressor, playing a vital role in regulating key cell cycle processes and responding to cellular stress (Lane and Crawford, 1979; Linzer and Levine, 1979; Kastan et al., 1991). One of the effects of mutations in the TP53 gene is the abnormal expression of its protein, which leads to disruption of the cell cycle and, consequently, to the development of neoplasia. Mutations in the TP53 gene have been found in over 50% of human cancers (Lane and Levine, 2010). It acts primarily as a transcription factor in the form of a tetramer (Khosravi et al., 1999; Ashcroft et al., 1999), stimulating the expression of genes responsible for DNA repair or apoptosis (Kastan et al., 1991).

The 5′ cap-dependent translation initiation process is the most efficient mechanism of protein synthesis for most mRNA transcripts present in the cell. However, the efficiency of this process may be reduced or abolished under conditions of cellular stress, when most of the cell’s vital functions are limited and the action of translation initiation factors, such as eIF4F or eIF2α, is blocked. In such a situation, it is possible to bypass initiation involving the 5′ cap and initiate translation using specialized structures called internal ribosome entry sites (IRES), which recruit a small ribosome subunit to the start codon (Pelletier and Sonenberg, 1988; Spriggs et al., 2008). Thanks to the IRES structure, it is possible to maintain the continuous synthesis of some critical proteins, regardless of unfavorable physiological conditions within the cell. These proteins regulate the cell cycle, apoptosis, cell response to stress conditions, or the process of carcinogenesis. It has been estimated that approximately 10%–15% of cellular transcripts have the ability to translate independently of the cap (Lacerda et al., 2017; Marques et al., 2022). Among them are proteins that are crucial for sustaining life, such as p53 (Ray et al., 2006; Candeias et al., 2006; Harris et al., 2018). This strategy is also employed by some viruses, which, after blocking the cap-dependent mechanism — e.g., by deactivating certain translation factors and utilizing highly efficient IRES structures — redirect the cellular protein biosynthesis machinery to meet their own needs (Pelletier and Sonenberg, 1988; Dutkiewicz et al., 2016). There are also other ways of cap-independent translation initiation, e.g., leaky scanning or re-initiation (Merrick WC, 2004).

In the 1980s, it was observed that shortened mRNA variants can be produced from the TP53 gene, which encodes different proteins (reviewed in Marcel et al., 2011a). This led to the hypothesis that there are several isoforms of the p53 protein. Subsequent years of research on the TP53 gene expression have shown that, as a result of the use of different transcriptional promoters, alternative splicing, and different translation initiation sites, it is possible to synthesize many isoforms of the p53 protein (Khoury and Bourdon, 2010) (Figure 1; Table 1).

To date, at least 13 isoforms of the p53 protein have been identified, differing in their domain content and, consequently, in their location and functions within the cell, but also in their copy numbers and relative proportions within the total cellular p53 pool (Marcel et al., 2011a; Guo et al., 2024; Senturk et al., 2014). These proportions depend on changing conditions and states of the cell, such as stress, cell cycle phase, or age. It has also been observed that the synthesis of individual isoforms in cells depends on tissue type and can be regulated under various conditions of cellular stress (Powell et al., 2008; Bourougaa et al., 2010; Marcel et al., 2011a). Furthermore, studies indicate that p53 protein isoforms may play an important role in regulating p53 activity within the cell (Bourdon et al., 2005).

N-terminal isoforms can be divided into “long” and “short” ones. “Long” are p53FL variants and Δ40p53, starting at AUG1 or 2. “Short” refers to those that start with the AUG 3 or 4: Δ133p53 and Δ160p53. C-terminal isoforms are indicated with suffixes: α, β, γ, or ψ, depending on the region of translation termination (Figure 1B) and are characterized by different subcellular localization and limited oligomerization capacity (Bourdon et al., 2005; Marcel et al., 2011a). In most cells, p53α, Δ133p53, and p53β are mainly found in the cell nucleus, but small amounts can also be detected in the cytoplasm. The p53γ isoform can be located in both the nucleus and the cytoplasm, while the Δ133p53γ isoform occurs only in the cytoplasm.

Δ40p53 lacks a transactivation domain, so its ability to activate gene transcription is severely limited. Since there is no amino terminus to recruit chromatin remodeling factors, Δ40p53 cannot regulate gene expression as effectively as p53FL (Courtois et al., 2002; Powell et al., 2008; Olivares-Illana and Fahraeus, 2010).

Another consequence of the absence of this domain is the loss of protein-protein interactions with HDM2 (human double minute 2 protein), which, in turn, means that this isoform is not ubiquitinated and degraded by the proteasome, unlike full-length p53. As a result, the half-life of Δ40p53 in the cell is significantly longer, at 9.5 h (Marcel and Hainaut, 2009). Similar to p53FL, the Δ40p53 isoform has a DBD domain and an oligomerization domain, thanks to which it can form part of p53 tetramers and disrupt their action as a transcription factor.

The biosynthesis of p53 protein and its isoforms is strictly controlled by the presence of stable secondary structure elements located within the 5′ non-coding region of p53 mRNA (Vilborg et al., 2010; Błaszczyk and Ciesiołka, 2011; Gorska et al., 2013). The length of the 5′UTR and the secondary structure of this region significantly impact the migration of the 43S PIC complex, which binds to mRNA to identify the AUG initiation codon (Leppek et al., 2018). 5′ UTR regions that are highly structured and thermodynamically stable can cause delays in scanning by the ribosomal complex and lead to a decrease in translation efficiency. Additionally, the initiation of p53 isoform translation may also be controlled via IRES elements present in the 5′ terminal region of p53 mRNA (Ray et al., 2006; Candeias et al., 2006). Furthermore, the high structural complexity of the 5′ UTR region also influences its interaction with RNA-binding proteins (RBPs), which can either enhance or suppress translation initiation (Takagi et al., 2005; Leppek et al., 2018; Swiatkowska et al., 2016). More below.

In the 5′ terminal regions of p53 mRNA (variants transcribed from P0 and P1), there are two characteristic hairpin-like structural motifs: G56–C169, where the codon AUG1 for the p53 protein is located, and U180–A218, which interacts with the HDM2. Additionally, this region contains an IRES element, which is involved in regulating p53 translation (Figure 1C).

The role of individual elements of the 5′ mRNA structure (Figures 1C,D) in the translation efficiency of the main isoform p53α and Δ40p53α has been described in detail in several studies (Gorska et al., 2013; Swiatkowska et al., 2016; Szpotkowska et al., 2019), and for the translation of shorter isoforms – Δ133p53 and Δ160p53 – in the work of Żydowicz-Machtel et al. (2021). The structural context of the start codon, the length of the UTR, and the presence of key structures that bind the primary regulators of translation of a given mRNA are all significant factors affecting translation efficiency. In the case of p53FL variants, the structure responsible for interactions with HDM2 is the U180–A218 hairpin.

Recent work suggests that the action of the U180-220A structure, known also as BOX-1, may be a potential riboswitch sensitive to the cascade of signals triggered by DNA damage (Chen et al., 2025). In its inactive, blocked state, the apical loop interacts with the sequence above and is hidden from interaction with the HDM2 protein. In its active state, it is released from this interaction to favor interaction with the protein and enhance translation efficiency. In this work, the mechanism blocking protein binding is a loop-loop interaction between three nt residues of the apical loop of the aforementioned BOX-1 hairpin, with the nt residues of the apical loop of another hairpin 1–40, which in the structure taking into account the 5′UTR sequence corresponds to the side loop of the hairpin-like structure: G56–C169 (Figure 1C – the interaction is marked with blue stripes).

One explanation for the claim that p53 protein isoforms may be important in regulating p53 activity in the cell (Bourdon et al., 2005) is the tetramer theory. According to this theory, the p53 protein acts as a transcription factor in the form of a tetramer, which can include different types of p53 isoforms, as long as they contain an oligomerization domain and are located in the nucleus. The composition of the tetramer, therefore, depends on the availability of individual isoforms at a given time and place. The way in which this factor acts depends on which isoforms form the tetramer, sometimes contrary to what we would expect from the main protein isoform. For example, suppose the tetramer is formed by isoforms that lack a transactivation domain. In that case, it will not act as a transcription factor, and if the tetramer is formed by isoforms lacking the HDM2 protein-binding domain, it will not undergo ubiquitination. Its duration and action in the cell will be prolonged. In this case, depending on the composition of the p53 isoform pool in the cell, p53 tetramers will act differently.

Most often, when N-terminal isoforms lacking the transactivation domain form a p53 tetramer, they activate other genes less effectively or not at all, which may result in cell cycle arrest, inhibition of apoptosis, and enhanced proliferation, regardless of whether DNA mutations or viral infection have occurred. It is easy to imagine the future fate of the cell and the organism when p53 function is impaired: unchecked cancer development, viral infection, or accelerated cellular aging. Sometimes, however, the action of isoforms can be beneficial. For example, when the alarm signal read by p53 was false (i.e., it came from a situation that is not an actual threat). In case of C-terminal isoforms that lack the oligomerization domain, they will not participate in tetramer formation, but their action will occur at a different level and in a different way.

The mechanisms responsible for generating over a dozen isoforms from a single gene comprise the use of different transcriptional promoters (DNA level), alternative splicing (pre-mRNA level), and different translation initiation sites (mRNA level). In vitro studies have demonstrated that translation initiation at different sites on full-length p53FL mRNA is possible due to specific RNA structures, and that these structures are also responsible for the rate and efficiency of target protein isoform formation.

Unfortunately, some studies of the structure and role of RNA transcripts and their fragments (not only in the case of p53) do not take into account the presence of sequences upstream the start codon, which is a grave mistake, as the presence of an adjacent RNA sequence, additionally rich in G and C residues, may be of great importance in shaping the structure of the fragment under study. The same applies to studies of 3′UTRs. If we do not account for adjacent coding sequences, we may obtain false results. This should also be kept in mind when studying reporter transcripts, i.e., those where the coding region contains a reporter gene sequence that can alter the structural context of the studied UTR fragment.