Review
Mechanisms of Selenoprotein P translation regulation by long noncoding RNA
2025 Volume 5 Issue 1 Pages rev19-rev23
Details
2025 Volume 5 Issue 1 Pages rev19-rev23
Selenium (Se) is one of the essential trace elements in the body. Se is present in proteins in the form of selenocysteine (Sec), in which the sulfur of cysteine (Cys) is replaced by Se. These proteins are referred to as selenoproteins. There are 25 selenoproteins in the human genome, and they play important roles in various physiological functions, including as an antioxidant and in the synthesis of thyroid hormones. Sec is inserted into selenoproteins using the Sec insertion sequence (SECIS), which is located in the 3′ untranslated region. We have identified an antisense long noncoding RNA, CCDC152, which binds mRNA of selenoprotein P (SELENOP), one of the plasma selenoproteins. CCDC152 inhibits the binding of SECIS binding protein 2 (SBP2), which is a key protein for selenoprotein translation, to SECIS by direct interaction with SELENOP mRNA. Inhibiting the formation of the SBP2 and SECIS complex by CCDC152 reduces the binding of ribosomes to SELENOP mRNA and suppresses the translation step of SELENOP. As a result, CCDC152 causes a decrease in SELENOP protein levels independent of SELENOP mRNA levels. No impact was observed on the protein and mRNA expression levels of other selenoproteins. This review describes the mechanism of SELENOP protein suppression by CCDC152.
Selenium transport protein selenoprotein P (SELENOP) contains the essential trace element selenium (Se) in the form of selenocysteine (Sec), which is an amino acid in which the sulfur of cysteine is replaced by Se. It has been reported that SELENOP is increased in the blood of patients with type 2 diabetes mellitus (T2DM). Excess SELENOP induces insulin resistance and impairs insulin secretion, resulting in aggravation of T2DM [1,2]. Therefore, SELENOP is expected to become a new therapeutic target of T2DM.
Selenoproteins, including SELENOP, are synthesized by a unique translation mechanism. In the 3′ untranslated region (UTR) of the mRNA encoding selenoprotein, there is a Sec insertion sequence (SECIS) forming a stable loop structure. Sec-tRNAs, specific elongation factor eEFSec, and SECIS binding protein 2 (SBP2) bind to SECIS and form a complex. This complex allows for the insertion of Sec into the UGA codon, which is usually recognized as a termination codon [3]. In the process of analyzing the SECIS sequence, we identified a gene of unknown function, coiled-coil domain-containing protein 152 (CCDC152), which has a complementary sequence to the SECIS-containing 3′ UTR of SELENOP mRNA. The overexpression of CCDC152 in human hepatocellular carcinoma-derived HepG2 cells expressing SELENOP causes a decrease in protein levels without altering SELENOP mRNA levels. Since CCDC152 is mainly found in the nucleus, the CCDC152 protein was not detectable when overexpressed in HEK293 cells, which do not express CCDC152. In addition, CCDC152 RNA binds to SELENOP mRNA; hence, it is possible that CCDC152 functions as an RNA. In HepG2 cells overexpressing CCDC152, the binding of SELENOP mRNA to ribosomes and SBP2 was decreased [4]. Based on these functions, we named CCDC152 a long noncoding RNA inhibitor of selenoprotein P translation (L-IST) (Figure).
In this review, we describe the translation mechanism of selenoprotein in the presence of SECIS and explain the action of L-IST that we found. We also describe L-IST as a new therapeutic target for reducing the increase of T2DM.
Se deficiency is known to cause severe cardiomyopathy and increase the incidence of cancer [5,6]. There are 25 selenoproteins in the human genome, and they play important roles in various physiological functions. Selenoproteins include antioxidant proteins, such as glutathione peroxidase (GPx) and thioredoxin reductase (TrxR), and proteins related to thyroid hormone synthesis, such as iodothyronine deiodinase. Thus, selenoproteins play an important role in the protection of cells from oxidative stress and in the process of energetic metabolism. The insertion of Sec into proteins occurs during the translation phase, and Sec has also been called the “21st amino acid that can be translated” [7]. The complex of SBP2 and Sec-tRNASec binds to a hairpin structure called SECIS, which is essential for Sec translation, and Sec is inserted into the UGA codon [8]. The strength of the binding between SECIS and SBP2 depends on the 3D structure of SECIS [9]. Therefore, it has been suggested that the efficiency of selenoprotein translation is dependent on the 3D structure and not the SECIS sequence [9].
We compared the SECIS sequences of mRNAs encoding 25 human selenoproteins in SelenoDB 2.0 (http://selenodb.crg.eu) [10] with the sequences of all transcripts using BLAST analysis. We found a novel gene, CCDC152, containing an antisense sequence to the SECIS sequence of the SELENOP. CCDC152 contains a part of the coding sequence of SELENOP and an antisense sequence against the 3′ UTR region containing the SECIS sequence, and both sequences were completely complementary to each other. The results of the homology analysis of CCDC152 show that it is present from fish to mammals.
Analysis of CCDC152 expression levels in various cultured cells revealed that CCDC152 was expressed in human neuroblastoma SH-SY5Y cells, glioma U87MG cells, and human T lymphocytoma Jurkat cells, but only slightly in hepatocarcinoma-derived HepG2 cells, which express and secrete SELENOP. Comparison of expression levels in various tissues in mice revealed that SELENOP was expressed mainly in the liver, small intestine, and kidney and that CCDC152 was most highly expressed in the testes and also expressed in the liver, kidney, and white fat tissue. In high-fat, high-sucrose dietinduced diabetic model mice, there was an increase in SELENOP expression in the liver, while CCDC152 in the liver tended to be decreased.
CCDC152 has a complementary sequence to SELENOP mRNA; therefore, we overexpressed CCDC152 in HepG2 cells expressing SELENOP and examined its effect on SELENOP protein expression. We observed a reduction in SELENOP protein levels and no change in mRNA levels. The overexpression of CCDC152 and SELENOP mRNA in HEK293 cells, which express neither CCDC152 nor SELENOP mRNA, also caused a decrease in SELENOP protein levels without a change in mRNA levels, as observed in HepG2 cells. Hence, the mRNA level–independent decrease in SELENOP protein levels by CCDC152 was not specific to a particular cell type. Next, we examined the effect of CCDC152 on other selenoproteins. There were no changes in the protein and mRNA levels of selenoproteins such as GPx4 and TrxR1 other than SELENOP, suggesting that CCDC152 specifically affects the amount of SELENOP protein and mRNA.
There are more than 200 members of the CCDC protein family. It has been reported that CCDC80 and CCDC134 inhibit the activity of ERK [11, 12] which protein suppresses SELENOP transcription by promoting FoxO3a translocation out of the nucleus. Therefore, CCDC80 and CCDC134 may enhance the transcription of SELENOP. On the other hand, the suppression of SELENOP protein expression by CCDC152 overexpression is not involved in transcriptional regulation. Therefore, it is likely that SELENOP protein levels are regulated by a different mechanism than the previously known CCDC protein family. Because of the presence of an open reading frame (ORF) between bases 71 and 835 in the 5′ region of CCDC152, we investigated whether the CCDC152 protein or RNA could reduce the protein content of SELENOP. To validate protein translation from the ORF, we inserted an HA-tag at the 3′ end of the ORF. However, western blotting with an antibody against the HA-tag did not detect the CCDC152 encoding protein. Next, we analyzed the effect of the CCDC152 deletion mutant on the SELENOP protein. The CCDC152 (∆500) mutant, in which up to 500 bases were deleted from the 5′ end, including the putative initiation ATG codon, retained the ability to suppress SELENOP protein expression. However, the CCDC152 (∆600) mutant, in which up to 600 bases were deleted from the 5′ end, did not have the same suppressive effect. Therefore, the SELENOP-reducing effect of CCDC152 is possibly through its action not as a protein but as an RNA.
Next, we focused on the translation step, with particular emphasis on the step involved in the binding of mRNA to ribosomes. Our results showed that CCDC152 exerts an inhibitory effect on protein synthesis in a manner independent of mRNA levels. Polysome analysis was performed to evaluate the binding of SELENOP mRNA to ribosomes. We found that the number of ribosome-bound SELENOP mRNAs was markedly diminished in CCDC152-overexpressing HepG2 cells. In contrast, GPx4 mRNA, another selenoprotein mRNA, did not show a significant change in its binding to ribosomes when CCDC152 was overexpressed.
Subsequently, we investigated the binding of SELENOP mRNA to SBP2, which is an essential factor for Sec insertion via SECIS, using an RNA pull-down assay. We found that the overexpression of CCDC152 reduced the binding affinity of SELENOP mRNA to SBP2. Thus, these results showed that CCDC152 inhibits the translational step of SELENOP mRNA, especially the binding of SBP2 to SECIS. Based on these functions, we named CCDC152 a long noncoding RNA inhibitor of selenoprotein P translation [4].
It has been reported that elevated blood SELENOP levels are associated with T2DM [1] and pulmonary hypertension [13]. The development of a method to decrease SELENOP would be a novel therapeutic strategy for those diseases. Therefore, we searched for substances that increase L-IST. We found that epigallocatechin gallate (EGCg), which is known as the main ingredient of green tea with an antidiabetic effect, increased L-IST and decreased SELENOP protein levels without changing SELENOP mRNA levels. Based on these results, we administered EGCg to mice and found that L-IST was increased in the liver of EGCg-treated mice, but SELENOP mRNA levels were unchanged. Mice with increased L-IST by EGCg also had decreased blood levels of SELENOP and blood glucose levels, indicating that L-IST could be a promising therapeutic candidate agent for T2DM treatment [4].
Most of the intergenic regions in the genome have been called “junk regions,” which are presumed to lack any functional significance. However, in recent years, many noncoding RNAs have been found in these junk regions, and their functions have been clarified. The analysis of noncoding RNAs involved in the regulation of protein quantity has been a prominent area of research.
Protein reduction by short RNAs of approximately 20 mer, such as siRNA and miRNA, is mediated by the AGO family protein [14]. In contrast, different mechanisms of translation regulation by long noncoding RNAs have been reported for different types of RNA. Antisense BASE-1 has been demonstrated to increase BASE-1 protein levels by causing mRNA stabilization [15]. Long noncoding RNAs containing SINEB2 sequences promote the translation of target mRNAs by recruiting ribosomes [16]. Competing endogenous RNAs (ceRNAs) are long noncoding RNAs that do not act directly on mRNAs. These ceRNAs trap miRNAs, thereby inhibiting the binding of miRNAs to their target mRNAs and increasing protein levels [17]. The L-IST has a complementary sequence to the SECIS sequence of SELENOP mRNA. L-IST appears to bind directly to SELENOP mRNA, inhibiting the binding of SBP2 to SECIS and of SELENOP mRNA to the ribosome. The mechanism of translational suppression of SELENOP by L-IST differs from the previously reported mechanisms of regulation of protein levels by long noncoding RNAs such as antisense BASE-1. However, Long-noncoding RNAs that bind directly to mRNAs, and inhibit the binding of SECIS to SECIS binding proteins such as SBP2 have not been reported. Therefore, the translational suppression of SELENOP by L-IST is considered to be a novel function of a noncoding RNA.
We are currently analyzing the functional sequence of L-IST and its effect on the translation of other selenoproteins. We have not yet completely determined the precise functional sequences of L-IST. Nevertheless, we have identified the functional sequences that are present in both the 5′ and 3′ regions and not only sequences complementary to SELENOP mRNA [4]. Furthermore, it is becoming obvious that the complementary sequence between L-IST and SELENOP mRNA is not necessary for the suppression of SELENOP mRNA translation by L-IST. We replaced the complementary sequence region to SELENOP mRNA in L-IST with the sequence complementary to the 3′ UTR in other selenoprotein mRNA. The available data are preliminary, but they suggest that RNAs containing antisense sequences of 3′ UTR may also suppress protein levels of other selenoproteins as well as SELENOP. In the future, we will confirm the reproducibility of the study and analyze the effect of antisense long noncoding RNAs on the translation of mRNAs containing SECIS.
We identified a novel noncoding RNA, CCDC152/L-IST, with a sequence complementary to the SECIS sequence of SELENOP mRNA. L-IST reduced SELENOP protein levels by specifically inhibiting the translation step of SELENOP mRNA (Figure) [4]. SELENOP was originally studied as a Se transport protein, but recent studies have shown that excess SELENOP is associated with many diseases, such as T2DM [1,2]. As L-IST can specifically decrease SELENOP protein levels, therapies that increase L-IST may be promising for treating several diseases in patients with high SELENOP levels.
The research on CCDC152/L-IST described in this manuscript was conducted in collaboration with Professor Yoshiro Saito (Tohoku University) and Professor Toshifumi Inada (Tokyo University). We would like to take this opportunity to thank them.
The authors declare no conflict of interest associated with this manuscript.
