Impact of mRNA and protein isoforms in lipoprotein metabolism and how to modulate them

Abstract

Purpose of review

More than 95% of human genes undergo alternative pre-mRNA processing based on cell type, developmental stages, and environmental stimuli, among other factors. Not all alternatively spliced mRNAs are translated to proteins, and some of the noncoding mRNA isoforms play vital roles in cellular homeostasis. This review summarizes protein coding and noncoding RNA isoforms reported for key genes involved in lipoprotein metabolism, and emerging technologies that can be exploited to specifically induce a desired isoform.

Recent findings

As sequencing technologies become more accessible, more variations in gene transcripts are being detected. Publicly available databases collate these as they arise, but not all of them are captured. Additionally, the function, if any, of many of these alternatively spliced transcripts is currently unknown. Novel strategies to investigate specific transcripts are also continuously evolving.

Summary

Most human genes are alternatively spliced, generating various mRNAs and protein isoforms. Any cis or trans factors that alter the balance of these isoforms can have deleterious effects. The fundamental knowledge on the role of each isoform in maintaining cellular health is currently lacking. Emerging technologies which allow modulation of natural mRNA splicing can be used to further our understanding of natural isoform expression and function.

INTRODUCTION

Alternative splicing, a well known mechanism of gene regulation, facilitates the generation of multiple mRNA isoforms from a single gene. While the concept itself is not new, research findings have revealed how alternative splicing is regulated by various stimuli, including environmental factors [1] and drug-induced signals [2]. It has also been observed that simple physiological activities, such as exercise [3], or a modest dietary change [4] can trigger the expression of specific isoforms. For example, the levels of low-density lipoprotein receptor mRNA isoforms lacking exons 4 or 12 are lower in individuals who have undergone exercise training compared to those without training [3]. Interestingly, these two isoforms are significantly higher in adults with high cholesterol intake [3].

Various mRNA isoforms reported for each gene are captured in public databases such as Ensembl [5^▪] and the National Centre for Biotechnology Information (NCBI) and assign them as protein-coding or noncoding based on the open reading frames. Likewise, there are many protein isoforms documented in public databases such as The Human Protein Atlas (proteinatlas.org). However, it is important to note that, to date, not all protein-coding transcripts have assigned functional roles. Understanding the function of individual isoforms and circumstances that lead to their induction can be important when exploring therapeutic strategies. 

Box 1.

no caption available

In this review, we discuss the alternative mRNA isoforms reported for the four major proteins involved in lipoprotein metabolism: low-density lipoprotein receptor (LDLR), HMG-CoA reductase (HMGCR), proprotein convertase subtilisin/kexin type 9 (PCSK9), and apolipoprotein B (APOB). Despite each gene generating several mRNA isoforms in cells naturally, there is often only one predominant isoform expressed at high levels and the choice of this/the major isoform is mostly spatially and temporally regulated. We describe approaches that can induce or switch a particular mRNA/protein isoform by manipulating pre-mRNA splicing (Fig. 1) and outline the advantages and limitations of these approaches. These isoform-switching approaches differ fundamentally from gene knockdown and we discuss how these insights can be leveraged for clinical applications.

FIGURE 1.

Overview of various methodologies to switch mRNA and protein isoforms. The illustration depicts both natural and artificial mechanisms that enhance the expression of various isoforms or specific isoforms. Created in BioRender. Cale, J. (2026) https://BioRender.com/h2f7vr5.

MAJOR PROTEINS IN LIPOPROTEIN METABOLISM AND THEIR NATURAL ISOFORMS

Low-density lipoprotein receptor

The low-density lipoprotein receptor (LDLR) is mainly expressed in the liver and is important for clearing LDL from the blood. The LDLR is encoded by the LDLR gene, the MANE (Matched Annotation from NCBI and EMBL-EBI) transcript (ENST00000558518.6) has 18 exons, encoding 860 amino acids (aa). Twenty-eight LDLR transcripts, both protein-coding [transcript contains an open reading frame (ORF)] and noncoding, have been reported in the Ensembl database (Table 1). In addition to the exons, both 5΄ and 3΄ untranslated regions of the LDLR mRNA are heavily alternatively spliced. Of these 28 isoforms, it is uncertain how many are translated and the resulting LDLR protein isoforms function, if at all. An additional seven naturally occurring LDLR transcripts, not in public databases, were reported by Tveten et al. [6], however, the complete transcripts were not sequenced and resulting proteins were not investigated. Of all the reported LDLR transcripts, those missing exons 4 or 12 are of particular interest as they are associated with hypercholesterolemia, presumably indicating that the encoded protein isoforms are non-functional [3].

Table 1.

Transcript isoforms reported for LDLR, HMGCR, PCSK9, and APOB genes

Gene name (gene ID)	Number of exons in MANE select (transcript ID, Ensembl)	Number of transcripts with ORF	Number of noncoding transcripts	Number of transcripts not in database
LDLR (ENSG00000130164)	18 (ENST00000558518.6)	21	7	7 [6]
HMGCR ( ENSG00000113161)	20 (ENST00000287936.9)	30	8	None
PCSK9 (ENSG00000169174)	12 (ENST00000302118.5)	6	6	4 [14]
APOB (ENSG00000084674)	29 (ENST00000233242.5)	2	2	None

APOB, apolipoprotein B; ORF, open reading frame; HMGCR, 3-hydroxy-3-methylglutaryl-CoA reductase; LDLR, low-density lipoprotein receptor; PCSK9, proprotein convertase subtilisin/kexin type 9.

3-Hydroxy-3-methylglutaryl coenzyme A reductase

The 3-hydroxy-3-methylglutaryl-CoA reductase (HMGCR), found in the endoplasmic reticulum, is the rate limiting enzyme essential for catalyzing HMG-CoA to mevalonate, an important step in cholesterol biosynthesis [7]. The HMGCR is highly sensitive to cholesterol concentration, such that high cholesterol levels reduce HMGCR expression, while low cholesterol levels promote its expression. The enzyme is encoded by the HMGCR gene, where the MANE transcript (ENST00000287936.9) has 20 exons, encoding 888 aa. Currently, 38 protein-coding and noncoding transcripts have been reported on Ensembl (Table 1). The translation start codon is located in exon 2 of the MANE transcript, with most transcripts having one or two exons of various sizes spliced in before exon 2 or in the 3’ untranslated region. This results in many transcripts coding the same full-length protein, although these transcript isoforms are expressed in a tissue-specific manner [8]. Exon 13 is excluded from four of these documented transcripts, which is unsurprising as its splicing is reported to be highly influenced by a single-nucleotide polymorphism (SNP) rs3846662 located in intron 13 [4,9]. HMGCR protein lacking exon 13 has reduced activity in minigene assays [9], and the expression of HMGCR transcripts lacking exon 13 has been associated with reduced responses to Simvastatin in the cohort being studies [10,11]. Apart from those reported in Ensembl, no additional HMGCR transcripts have been reported to date.

Proprotein convertase subtilisin/kexin type 9

Proprotein convertase subtilisin/kexin type 9 (PCSK9) is secreted by the liver and acts as a negative regulator of LDLR expression [12]. Upon binding to LDLR, PCSK9 promotes its degradation, preventing LDLR recycling and elevating LDL in the blood. Levels of LDL in the blood are directly correlated with increased PCSK9 expression, and PCSK9 inhibitors are currently approved to treat familial hypercholesterolemia and hyperlipidemia [12]. The PCSK9 gene encodes the PCSK9 protein and the MANE transcript (ENST00000302118.5) consists of 12 exons, encoding 692 aa. In the Ensembl database, there are 12 transcripts reported for the PCSK9 gene (Table 1). The PCSK9 transcript missing exon 8, reported by Schmidt et. al. [13], was found to be induced by cholesterol supplementation [4]. Additional alternatively spliced transcripts were also detected by Cale et al.[14] in untreated human hepatocellular carcinoma cell line, Huh-7. The Δ8 protein isoform was detected only when the Δ8 PSCK9 transcript became the major isoform within these cells and was shown to have a reduced inhibitory effect on LDLR function [14,15].

Apolipoprotein B

Unlike the proteins mentioned above, APOB is part of the lipoprotein itself and is essential for the assembly and secretion of triglyceride-rich lipoproteins [16]. The two major APOB protein isoforms, APOB-100 and APOB-48, are exclusively expressed in the liver and intestine, respectively, and their roles are well documented [16]. APOB-100 is critical for transporting cholesterol and triglycerides from the liver to peripheral tissues, whereas APOB-48 is responsible for absorbing lipids from the diet and transporting them to the lymphatic system and bloodstream. APOB protein is encoded by the APOB gene, and four transcripts are reported for this gene (Table 1). APOB-100 is produced from a 29 exon MANE select transcript (ENST00000233242.5) encoding 4563 aa and in the presence of RNA editing enzyme APOBEC-1 (apolipoprotein B mRNA editing catalytic subunit 1), the codon for amino acid (Gln) residue 2153, CAA, in exon 26 is edited to UAA (stop codon), generating APOB-48, missing the C-terminal domain necessary for LDLR binding [17,18]. We have not found transcripts not already included in public databases.

Isoform switching

Although the human body is highly sensitive to environmental changes and can adjust the expression of beneficial isoforms accordingly, some genetic variations may impair these adaptive responses. For individuals whose genetic predisposition impair their response to environmental cues, artificially inducing potentially beneficial isoforms could help avert health issues that would otherwise arise. Recent advances in technology have enabled us to specifically enhance processing and synthesis of the desired mRNA isoforms, and consequently, protein isoforms, transiently or permanently, to provide therapeutic benefits.

Traditionally, small molecules have been the primary choice for drug development. However, in the early years, they were not widely considered feasible to modulate splicing. This assumption is no longer valid, with the splice modulating small molecule drug, risdiplam [19], which corrects splicing of survival of motor neuron 2 exon 7, receiving approval from the U.S. Food and Drug Administration (FDA) as a treatment for spinal muscular atrophy in August 2020. One advantage of small molecule drugs is often their oral bioavailability; however, after the lead molecules have been identified from a library comprising tens of thousands to millions of molecules, extensive modifications are still required to avoid or minimize toxicity. Several clinical trials using small molecule drugs to modulate splicing were discontinued due to unacceptable risks, such as RG7800 for spinal muscular atrophy [19] and branaplam/LMI070 for Huntington's disease [20]. In addition to splice correction, small molecules that can promote the inclusion of a “poison” exon (e.g., PTC518 for Huntington's disease) [21] or exon skipping (e.g., ICD33 for late-onset Alzheimer's disease) [22] are also under development. It should also be noted that not all pre-mRNA splicing modulation can be achieved using small molecules. As discussed earlier, the proteins involved in lipoprotein metabolism have multiple isoforms, and some of which have been associated with specific health conditions. Hence, it is possible that small molecules could be developed to help prevent the expression of the harmful isoforms or induce beneficial isoforms as appropriate.

The first class of drugs with proven success in modulating mRNA splicing or isoform switching are steric blocking antisense oligomers (ASOs), short (15–30 nucleotides) synthetic nucleic acid analogues. The first report of using ASO to modulate defective β-globin premRNA processing in a cell-free extract was published in 1993 by Kole and Dominski [23]. Despite slow progress in the early years, eleven ASOs have been approved by the US FDA to date as treatments for various conditions [24]. Five of these act through the modulation of pre-mRNA splicing. In 2007, Khoo et al. published the use of exon skipping ASOs to reduce the levels of APOB100 [25]. In addition, we and others have shown that various PCSK9 protein isoforms can be induced by modulating PCSK9 premRNA splicing, allowing these isoforms to be studied [14,15]. The advantage of using ASOs to generate mRNA/protein isoforms is that the behavior of these isoforms can then be assessed at their physiologically relevant/significant concentrations, in contrast to cloning the cDNA into an overexpression plasmid/minigene expression cassette. Furthermore, we can ensure that cells naturally utilize the mRNA isoforms for protein production, rather than relying on artificially induced over-expression of protein through cloning and inducible promoters.

Compared to small molecules, ASOs can be rationally designed to exhibit greater specificity as they exert their actions via Watson–Crick base pairing rules, often with minimal off-target effects with some ASO chemistries. Furthermore, base modifications and backbone chemistries are continually evolving to further increase potency with reduced toxicity and off-target effects [26]. However, the major challenge of all gene and genetic therapies is efficient delivery. This issue has been addressed, at least for liver targeting to some extent by attaching multivalent N-acetylgalactosamine (GalNAc) to the ASO, thereby enhancing delivery and uptake into hepatocytes [27]. Consequently, ASOs are a most suitable class of drugs to modulate genes mainly expressed in the liver.

The aforementioned small molecules and ASO-dependent isoform switching strategies are transient, and the isoform ratios will eventually revert to their original state once the drug is cleared from the cell. However, with emerging technologies, it can be possible to permanently switch expression of desired isoforms on or off using programmable gene-editing technology based on the clustered regularly short palindromic repeats (CRISPR)/CRISPR-associated (Cas) system [28]. This methodology is increasingly being used to assess the effect of mRNA variants on splicing and function in the context of rare disease diagnosis and exploring therapeutics to modify targeted gene expression [29,30^▪]. Should the off-target bystander editing of neighboring bases fall within the acceptable limit, CRISPR could be used to precisely edit the SNPs associated with the induction of undesirable mRNA isoforms in lipoprotein metabolism. Given the concerns around permanent genome editing, an alternative approach is RNA editing, which allows for the transient modification of transcripts without altering the underlying genomic sequence. APOBEC-1-mediated RNA editing is a naturally occurring mechanism for generating two APOB isoforms via the deamination of cytosine to uridine [17,18]. Other emerging technologies, such as adenosine deaminases acting on RNA editing, offer precise conversion of adenosine to inosine [31]. Both technologies can be used to modify undesired SNPs to favor the splicing of particular mRNA isoforms as needed. To our knowledge, neither genome nor mRNA editing technologies are sufficiently mature to be used as routine treatments for individuals prone to diseases of lipoprotein metabolism such as hyperlipidemia. However, there is no doubt they will become an option in the future.

CONCLUSION

Not all naturally occurring mRNA isoforms detected in human cells are captured in public databases likely due to the early data analysis relying on previously annotated variants. The list of both protein coding and noncoding mRNA isoforms is set to grow. Most alternative mRNA isoforms exist at very low levels and many of them may provide regulation as noncoding mRNA. With established technologies, we can now switch isoforms transiently or permanently according to the needs. Understanding the function, if any, of these isoforms may provide the discovery of safer and more effective splice modulating based therapies for human conditions.

Acknowledgements

None.

Financial support and sponsorship

This work is supported by Murdoch University funds to support KM. JC and MA-H salaries are partly funded by Sarepta Therapeutics, Cambridge, MA and Perron Institute for Neurological and Translational Science.

Conflicts of interest

There are no conflicts of interest.

REFERENCES AND RECOMMENDED READING

Papers of particular interest, published within the annual period of review, have been highlighted as:

• ▪ of special interest

• ▪▪ of outstanding interest