RNA processing is a critical step in gene expression. RNA-binding proteins are involved in every aspect of RNA processing including splicing, polyadenylation, nuclear export, translation, localization, and, eventually, RNA turnover. Polyadenylate-binding nuclear protein 1 (PABPN1) is an RNA-binding protein that is predominantly localized to the nucleus where its canonical function is to regulate polyadenosine [poly(A)] tail addition by first stimulating poly(A) polymerase (PAP) activity and then limiting poly(A) tail length (Wahle 1991, Kuhn et al. 2009). PABPN1 is also involved in RNA export from the nucleus, the pioneer round of translation, and targeting some specific RNAs for decay (Banerjee et al. 2013). Although PABPN1 was originally cloned and characterized as an RNA recognition motif (RRM)-containing RNA-binding protein, interest in the protein greatly increased when mutations that modestly expand an N-terminal alanine stretch were identified as causative for a form of late-onset muscle-specific disease termed oculopharyngeal muscular dystrophy (OPMD) (Brais et al. 1998). Many studies since this discovery have focused on understanding how a modest alanine expansion within this ubiquitously expressed protein leads to a muscle disease that impacts a specific subset of muscles. However, the PABPN1 protein is also an essential nuclear polyadenosine RNA-binding protein that plays critical roles in producing properly processed, mature RNAs.
PABPN1 Domain Structure and Function
Poly(A) Tail Addition and Alternative Polyadenylation
The PABPN1 protein plays a variety of roles in RNA processing that are summarized in Fig. 1b. PABPN1 is critical for 3′-end formation including poly(A) site (PAS) selection and subsequent polyadenylation, which are critical for fine-tuning gene expression. Polyadenylation is critical to ensure transcript stability as most RNAs that are not properly polyadenylated are targeted for degradation. PAS site utilization can affect gene expression by influencing the length of the 3′ untranslated region (UTR). A change in 3′-end length could alter the number of binding sites for RNA-binding proteins or microRNAs thus altering regulation of these transcript variants.
Experiments performed in the yeast Schizosaccharomyces pombe revealed that PABPN1 (Pab2 in S. pombe) associates with the C-terminal domain of RNA polymerase II (RNAPII) (Lemieux and Bachand 2009). After RNAPII transcribes the PAS, PAP and cleavage and polyadenylation specificity factors (CPSF) initiate cleavage and polyadenylation. PABPN1 then binds to the nascent poly(A) tail where PABPN1 and CPSF cooperatively stimulate PAP activity (Wahle 1991; Wahle et al. 1993). In addition to stimulating polyadenylation, PABPN1 is thought to act as a molecular ruler to ensure that the proper number of adenosine residues is added to 3′ ends by affecting the interaction between PAP and CPSF (Kuhn et al. 2009). In cultured primary muscle cells, Pabpn1 knockdown results in global poly(A) tail shortening and nuclear retention of polyadenylated RNAs (Apponi et al. 2010). PABPN1, therefore, plays a critical role in global modulation of poly(A) tail length, which in turn affects nuclear export efficiency.
Approximately 50–70% of mammalian transcripts contain multiple poly(A) sites, which are predominantly located within the 3′UTR (Elkon et al. 2013). Choice of PAS determines 3′UTR length and therefore inclusion or exclusion of cis regulatory elements such as microRNA binding sites and motifs recognized by RNA-binding proteins. Binding of microRNAs or RNA-binding proteins regulates gene expression by repressing translation, recruiting or excluding RNA decay machinery, and influencing RNA localization. Studies in cultured cells show that PABPN1 knockdown results in a global shift in PAS usage leading to 3′UTR shortening for a majority of transcripts (Jenal et al. 2012, Li et al. 2015). Similar changes in PAS usage with PABPN1 knockdown were found for a subset of candidate genes in a muscle cell line (de Klerk et al. 2012). The mechanism by which PABPN1 regulates PAS usage has not been determined though the prevailing model suggests that PABPN1 binds to and masks the weaker PAS from cleavage and polyadenylation machinery, thus enhancing distal PAS recognition and utilization (Jenal et al. 2012).
Other Roles in mRNA Processing
In addition to its canonical role in polyadenylation, PABPN1 influences splicing and translation. Although the majority of splicing occurs co-transcriptionally, some 3′ introns are not spliced before polyadenylation is completed. Studies using reporter constructs as well as endogenous targets in the context of PABPN1 knockdown showed that PABPN1 influences splicing of terminal introns (Muniz et al. 2015). Further mutational analysis revealed that the ability of PABPN1 to bind RNA and stimulate PAP is critical for modulating splicing at the 3′ end (Muniz et al. 2015).
While the steady-state localization of PABPN1 is nuclear, the protein shuttles between the nucleus and the cytoplasm. Most studies have focused on nuclear roles of PABPN1, but some work has investigated potential roles in translation. PABPN1 has been identified as part of a protein complex containing 5′ cap proteins that is involved in mRNA quality control and early rounds of translation (Ishigaki et al. 2001). Pab2, the PABPN1 ortholog in S. pombe, associates with polysomes (Lemieux and Bachand 2009), suggesting that PABPN1 may play a larger role in translation than previously appreciated. However, the specific mechanism by which PABPN1 participates in regulation of translation remains unknown.
Much of PABPN1 function has been described in the context of mRNAs, but PABPN1 also regulates a variety of noncoding RNAs by recruiting exosome machinery to targets for degradation. PABPN1-mediated hyperadenylation is required for rapid exosome-mediated degradation of intronless, polyadenylated reporter RNAs (Bresson and Conrad 2013). While PABPN1 depletion has only mild effects on steady-state levels of mRNAs, stability of a subset of polyadenylated long noncoding RNAs (lncRNAs) increases in the absence of PABPN1, likely due to loss of exosome-mediated degradation (Beaulieu et al. 2012). Polyadenylation is required for PABPN1-mediated lncRNA regulation, suggesting that polyadenylation or a subset of proteins bound to the poly(A) tail signals for exosome-mediated lncRNA turnover (Beaulieu et al. 2012). Additionally, PABPN1 and PAP hyperadenylate improperly spliced RNAs and target them for degradation by the RNA exosome (Bresson et al. 2015). Other studies of Pab2 in S. pombe have implicated PABPN1 in exosome-mediated poly(A) tail trimming during small nucleolar RNA (snoRNA) processing (Lemay et al. 2010). Therefore, PABPN1, through its participation in polyadenylation, regulates noncoding RNAs and improperly spliced RNAs by targeting these RNAs for decay.
Regulation of PABPN1
Although PABPN1 is critical for multiple steps in mRNA processing, its regulation is not well understood. As demonstrated by studies in cultured cells, PABPN1 is autoregulated via intron retention (Bergeron et al. 2015). When PABPN1 protein levels are high, it binds to an A-rich region in the PABPN1 3′UTR resulting in retention of the 3′-terminal intron. This 3′-terminal intron retention then triggers RNA exosome-mediated PABPN1 RNA degradation.
Tissue-specific PABPN1 regulation in the context of skeletal muscle is a topic of importance as mutations in PABPN1 cause muscle-specific disease. Muscle tissue contains low levels of PABPN1 mRNA and protein relative to levels in non-muscle tissue (Apponi et al. 2013). In fact, the PABPN1 mRNA is unstable in muscle though the mechanisms controlling PABPN1 mRNA stability have not been defined. Furthermore, it has been reported that steady-state levels of PABPN1 mRNA decline with age in human muscle tissue (Anvar et al. 2013). The low levels of PABPN1 in muscle tissue that further decrease with age suggest that muscle-specific regulation of PABPN1 could contribute to OPMD pathology. Thus, further study of tissue-specific and temporal regulation of PABPN1 is needed.
PABPN1 in OPMD
Mutations in the PABPN1 gene cause the late-onset, muscle-specific disease OPMD. Patients with OPMD have weakness in specific muscles including those of the eyelids, pharynx, and proximal limbs that cause eyelid drooping, difficulty swallowing, and loss of mobility (Banerjee et al. 2013). The PABPN1 protein contains a tract of 10 alanine residues in the amino terminus that is expanded to 11–18 alanines in OPMD patients (Fig. 1a) (Brais et al. 1998). The expansion to 11 alanines is inherited in an autosomal recessive manner. But expansions as small as 12 alanines cause autosomal dominant disease (Brais et al. 1998). Thus, a single copy of PABPN1 encoding a mere two additional alanines can cause disease. The alanine expansion does not disrupt any known functional domains within the PABPN1 protein, and it is unknown how this small expansion in the ubiquitously expressed PABPN1 protein causes muscle-specific disease.
One pathological hallmark of OPMD is the presence of insoluble nuclear PABPN1 aggregates in about 5–10% of patient muscle nuclei (Banerjee et al. 2013). Although the role aggregates play in OPMD pathogenesis is unclear, previous studies have shown that muscle from a mouse model of OPMD with a high percentage of aggregate-containing nuclei expressed markers of cell death (Davies et al. 2005). Still other studies have demonstrated that nuclear PABPN1 aggregates can sequester polyadenylated RNAs and other RNA-binding proteins that lead to nuclear retention and RNA splicing defects. For example, sequestration of the splicing factor SC35 in PABPN1 aggregates can cause mis-splicing of muscle-specific mRNAs (Klein et al. 2016); improperly spliced mRNAs have been detected in OPMD patient muscle biopsies as well (Klein et al. 2016). Other RNA-binding proteins such as TDP-43 (Kusters et al. 2009) and hnRNPA1 have also been detected in PABPN1 aggregates (Fan et al. 2003). However, many of these studies utilized expanded PABPN1 overexpression to investigate aggregate formation and contribution to disease progression, which may alter aggregate dynamics and cause phenotypes related to overexpression. In fact, overexpression of both expanded and wild-type PABPN1 results in the formation of insoluble aggregates that sequester poly(A) RNA, PAP, and mis-spliced pre-mRNAs (Fan et al. 2001, Tavanez et al. 2005, Klein et al. 2016). Thus, the exact mechanisms by which aggregates caused by endogenous levels of expanded PABPN1 may contribute OPMD pathology remain unknown.
PABPN1 expansion mutations may also interfere with PABPN1 protein function perhaps through contributing to aggregate formation. Several studies support the model by which a loss of PABPN1 function contributes to OPMD pathology. Although PABPN1 is a ubiquitously expressed protein, expansion mutations in PABPN1 cause disease only in a specific subset of skeletal muscle. Experiments performed in human and mouse tissues revealed that PABPN1 mRNA and protein levels are low in muscle tissue relative to non-muscle tissue; PABPN1 levels are lower still in those muscles affected by OPMD (Apponi et al. 2013). PABPN1 knockdown in mouse tibialis anterior muscle leads to shortening of 3′UTRs in mRNAs encoding components of the proteasome complex (Riaz et al. 2016). Many of these targets were detected in sequencing experiments performed in expanded PABPN1 overexpression models, demonstrating that PABPN1 reduction results in similar 3′UTR changes as expression of alanine-expanded PABPN1. Furthermore, overexpression of wild-type PABPN1 in an expanded PABPN1 overexpression mouse model is antiapoptotic and increases muscle strength while aggregate numbers remain unchanged (Davies et al. 2008). This result suggests that expanded PABPN1 is not able to fulfill antiapoptotic functions independent of aggregate numbers. However, expanded PABPN1 cannot rescue splicing defects observed with PABPN1 depletion (Muniz et al. 2015), suggesting that the molecular mechanisms involved in OPMD are complex.
The precise roles for loss or gain of function of PABPN1 in OPMD pathology have yet to be elucidated. It is likely that a combination of PABPN1 nuclear aggregates and loss of PABPN1 function contribute to disease, and the relative contributions may change during disease progression. For example, PABPN1 mRNA levels have been reported to decrease with age in human muscle (Anvar et al. 2013) so a small loss of function may become progressively worse as PABPN1 levels decline. PABPN1 also regulates E3-ligase PAS usage, which in turn regulates PABPN1 protein degradation (Raz et al. 2014). Thus, progressive downregulation of E3-ligase may exacerbate aggregate-related pathology in OPMD. Taken together, the low levels of PABPN1 in muscle and the changes that occur with age may underlie the muscle-specific pathology in OPMD.
Additional Roles for PABPN1
While the majority of PABPN1 studies have focused on general PABPN1 function or PABPN1 within the context of OPMD, recent publications implicate PABPN1 in a range of diseases including cancer, cardiomyopathy, and obesity. A study performed in duodenal cancer cells reveals that excess PABPN1 is exported to the extracellular space in exosomes (Ohshima et al. 2014). PABPN1 expression levels have also been shown to influence PAS usage in dilated cardiomyopathy (Creemers et al. 2016), as well as in lung cancer where lower PABPN1 levels were associated with poor prognosis (Ichinose et al. 2014); in both cases low PABPN1 expression correlated with changes in PAS usage. Recently, single nucleotide polymorphisms in PABPN1 and BCL2L2-PABPN1, which encodes a read-through transcript, were reported to be associated with obesity in Korean women (Lee et al. 2016).
Much of the PABPN1 literature focuses on understanding the canonical roles of PABPN1 in polyadenylation and alternative polyadenylation. Recent studies, however, have found that PABPN1 plays a much broader role in RNA processing and gene regulation than previously appreciated. Historically, many experiments have used in vitro biochemical experiments or immortalized cell lines to study mechanisms of PABPN1-mediated regulation. However, the study of PABPN1 within the context of muscle to understand the molecular mechanisms underlying the muscle specificity of OPMD has also contributed to general knowledge of PABPN1 function. More recent work has shown that PABPN1 function is clinically relevant in other settings such as cancer and heart disease. Therefore, PABPN1 is likely involved in many other cellular processes yet to be discovered.
- Klein P, Oloko M, Roth F, Montel V, Malerba A, Jarmin S, et al. Nuclear poly(A)-binding protein aggregates misplace a pre-mRNA outside of SC35 speckle causing its abnormal splicing. Nucleic Acids Res. 2016; 44:10929–45.Google Scholar
- Kuhn U, Gundel M, Knoth A, Kerwitz Y, Rudel S, Wahle E. Poly(A) tail length is controlled by the nuclear poly(A)-binding protein regulating the interaction between poly(A) polymerase and the cleavage and polyadenylation specificity factor. J Biol Chem. 2009;284(34):22803–14.PubMedPubMedCentralCrossRefGoogle Scholar