Splice variation of the mu-opioid receptor and its effect on the action of opioids

Summary points

1. An individual’s response to opioids is influenced by a complex combination of genetic, molecular and phenotypic factors.

2. Intra- and inter-individual variations in response to mu opioids have led to the suggestion that mu-opioid receptor subtypes exist.

3. Scientists have now proven that mu-opioid receptor subtypes exist and that they occur through a mechanism promoting protein diversity, called alternative splicing.

4. The ability of mu opioids to differentially activate splice variants may explain some of the clinical differences observed between mu opioids.

5. This article examines how differential activation of splice variants by mu opioids occurs through alternative mu-opioid receptor binding, through differential receptor activation, and as a result of the distinct distribution of variants located regionally and at the cellular level.

Introduction

There is a growing appreciation of the complexity affecting an individual’s response to opioids, as well as a greater understanding of the differential response that individuals may have to different opioids. The complexity of an individual’s response to opioids is influenced not only by genetic and molecular factors but also by phenotypic determinants such as psychosocial factors, gender and pain sensitivity. Although great efforts are being made to understand these factors, we have yet to identify predictable data that can contribute to a clinically meaningful treatment algorithm for personalising the prescribing of opioids. Such an algorithm would help to guide the prescribing of opioids so that patients receive the opioid analgesics that are most suited to them.

Researching the factors that influence an individual’s response to opioids involves investigating a wide range of parameters that include:

1. Genetic factors – deoxyribonucleic acid (DNA) sequence, genetic variation, gene–gene interaction and splice variation.

2. Pharmacokinetic and pharmacodynamic factors – absorption, distribution, metabolism and excretion of opioids, drug activation of the receptor, influence of receptor dimerisation and incomplete cross tolerance.

3. Characterisation of pain phenotype, including type and chronicity of pain, pain sensitivity and brain connectivity.

4. Clinical factors – disease status, concomitant medication, biochemical and psychosocial factors, diet and so on.

5. Epigenic factors – the dynamic post-translational modifications that affect activation and activity of gene products.

A number of reviews have looked at the strength of evidence for many of these factors.^1–5 In this review, we focus on the impact of splice variation of the mu-opioid receptor (MOR) on the clinical effect of opioids.

Mu opioids activate MORs

At the heart of an individual’s clinical response to opioids is the binding of an opioid to, and activation of, an opioid receptor. The majority of opioids in clinical use are termed ‘mu opioids’ due to their selectivity for MOR in receptor binding assays. Morphine is the prototypical mu opioid but other well-known mu opioids include codeine, diamorphine, morphine-6-glucuronide, oxycodone, fentanyl and methadone. Their interaction with the MOR is responsible for both analgesic effect and adverse events, such as respiratory depression, hallucinations, itch and constipation.

Intra- and inter-individual variation in response to mu opioids

Clinicians have long been aware that a patient may respond differently to two mu opioids; a patient might achieve adequate pain relief and be spared side effects with one mu opioid, but they may continue to have pain and experience intolerable side effects with another. Furthermore, clinicians recognise that between individuals there may be a differential response to a given mu opioid, whereby one patient may benefit while another does not. These intra- and inter-individual variations in response to mu opioids have led to the suggestion that MOR subtypes exist. Without subtypes, it is difficult to explain how supposedly similar drugs acting on the same receptor could give rise to such variable responses.

MOR subtypes

Mu, delta and kappa opioid receptors were discovered in the 1970s.^6–8 They belong to the super-family of seven transmembrane G-protein coupled receptors; they were among the first neurotransmitter receptors to be identified and their discovery led to the detection of endogenous opioid peptides.^9–12 Opioid receptor subtypes were first proposed in the 1980s when binding assays suggested the presence of multiple classes of binding site for opiates and enkephalins,¹³ and subsequent pharmacological studies supported this possibility.^13–18 However, when the mu, delta and kappa opioid receptors were cloned in the early 1990s, they were found to originate from distinct opioid receptor genes, OPRM1,^19–22 OPRD1^23,24 and OPRK1, respectively.^25–27 This discovery left scientists wondering how a single MOR gene, OPRM1, could generate a protein product that enabled such variable clinical response. Twenty years on, scientists have proven not only that MOR subtypes exist but also that they occur through a mechanism promoting protein diversity, called alternative splicing.^28–30

Alternative splicing

‘Alternative splicing’ is the term given to the process in which a single gene gives rise to multiple protein products. This is made possible during the transcription of DNA into messenger ribonucleic acid (mRNA). DNA is comprised of genes, and genes are composed of exons, separated by introns; exons are the actual DNA sequences included in the translation of mRNA into the protein product, whereas introns are regions of DNA that are usually spliced out and removed during the process. ‘Splicing’ is the process of combining exons together, and ‘alternative splicing’ is the inclusion of different combinations of exons within mRNA. Alternative splicing gives rise to different protein products contributing to protein diversity (Figure 1).

Figure 1.

Alternative splicing occurs during transcription of DNA into mRNA, leading to an increased diversity of protein products arising from a single gene.

Alternative splicing explains how only a relatively small number of genes in the human genome (approximately 20,000–30,000) give rise to many more protein products; it is thought that approximately 60% of genes produce at least one alternative mRNA.³¹

MOR splice variants

The MOR was cloned in 1993 from a single gene, OPRM1.^19–22 MOR1 is the name given to the primary MOR transcript, that is, the major form of OPRM1. MOR1 comprises four exons that encode architecturally distinct regions of the receptor. Exon 1 codes for the extracellular amino terminus and first transmembrane domain, exons 2 and 3 code for the remaining six transmembrane domains and exon 4 codes for the final 12 amino acids of the intracellular carboxyl terminus.³² (Figure 2)

Figure 2.

The MOR is a G-protein-coupled receptor. It has an extracellular amino terminus and an intracellular carboxyl tail. The receptor spans the cell membrane seven times resulting in three intracellular and three extracellular domains. The extracellular loops and the extracellular amino chain are important for ligand binding. The second and third intracellular loops as well as the carboxyl tail are important in the interaction between the receptors and G-proteins.

Soon after its discovery, the importance of MOR1 in morphine analgesia was confirmed in a series of studies involving MOR1 gene knockout mice models.³³ More was learned later of its structure and function using the novel technique of ‘antisense mapping’, whereby individual exons could be targeted and degraded. These animal studies demonstrated that targeting exon 1 resulted in a loss of morphine activity, despite preserving analgesia from diamorphine and morphine-6-glucuronide (M6G). Moreover, antisense probes targeting exons 2 and 3 had little effect on morphine action, whereas they significantly diminished the activity of diamorphine and M6G.^34,35 These findings have since been confirmed in exon-specific knockout mice models,³⁶ strongly supporting the early assertion that diamorphine and M6G exert their pharmacological actions through receptor mechanisms distinct from those of morphine.³⁷ The possibility that these effects were being mediated via kappa- or delta-opioid receptors was eliminated when diamorphine and M6G-mediated analgesia persisted in a triple knockout mouse model.³⁶

To date, nearly 20 human MOR splice variants have been characterised.^28,38–40 Broadly speaking, these variants fall into two categories: variants associated with exon 1 and those associated with exon 11. Both exons are controlled by distinct promoters, which regulate the expression of their respective variants.

Exon 1–associated splice variants

The majority of exon 1–associated variants involve exon 4 being replaced by alternative exons. These variants give rise to full-length seven transmembrane G-protein coupled receptors. Because of this, the structure of the binding pocket (formed by folding of the seven transmembrane domains) is identical and the variants demonstrate similar affinity and selectivity for mu opioids in receptor binding assays. However, structural differences in the intracellular carboxyl terminus affect signal transduction following receptor activation. Binding assays in human⁴¹ and mice MOR splice variants⁴² reveal major differences in potency and efficacy among these variants. Moreover, there is poor correlation between receptor activation and binding affinity, and the order in which mu opioids exert their maximal effect differs according to the variant.⁴² Thus, the ability of mu opioids to differentially activate splice variants may explain some of the clinical differences observed between the various mu opioids.

Exon 11–associated splice variants

Exon 11 is located upstream of exon 1, and novel exon 11–associated splice variants have been identified in mice⁴³, rats⁴⁴ and humans.⁴⁵ These variants account for approximately one-quarter of the total expression of MOR1 and its variants at the level of mRNA and protein in the whole brain of mice. Evidence from animal models suggest that exon 11–associated variants may be functionally and pharmacologically important, for example, disruption to exon 11 in mice has shown a significant reduction in the analgesic action of diamorphine, M6G and fentanyl, whereas morphine- and methadone-mediated analgesia were unaffected.⁴⁰

MOR variants display regional specificity within the central nervous system

Several of the exon 11–associated variants generate receptor proteins that are structurally identical to MOR1. However, they differ from MOR1, and from each other, in their distribution and expression within specific regions of the central nervous system (CNS).^43,46 This level of regulation suggests differential regional processing even among variants. Indeed, the regulation of MOR splice variants appears to be highly complex. There is evidence for cell-specific splicing, in which different cells express different splice variants,⁴⁷ and for splice variants influencing the position of a receptor pre- and post-synaptically within a cell.⁴⁸

Truncated variants form heterodimers to generate pharmacologically active targets

A proportion of the exon 11–associated variants predict proteins that lack exon 1. These variants encode proteins with six transmembrane domains rather than the full complement of seven. As described above, truncated exon 11–associated variants show regional mRNA distribution⁴³ and they are implicated in diamorphine and M6G-mediated analgesia.⁴⁰ However, they also appear to generate pharmacologically active targets through heterodimerisation. A study in mice demonstrated a novel naltrexone-derived agent binding to, and activating, the heterodimer formed between an exon 11–associated splice variant and a second receptor protein, which may be the orphanin FQ receptor, ORL₁. Not only did this study demonstrate the role of truncated variants in forming novel therapeutic targets, but, by activation of the heterodimer, the novel agent was able to bring about analgesia without mediating any opioid-related side effects. These findings suggest that it may be possible to dissociate the analgesic properties of receptor activation from unwanted side effects.⁴⁹

Splice variants encode single transmembrane protein products

Finally, a third class of variant has been identified composed only of exons 1 and 4. This variant encodes a protein product with a single transmembrane domain. The functional importance of these variants has yet to be understood; however, there is evidence to show that single transmembrane variants modulate MOR1 expression levels in the endoplasmic reticulum by dimerisation with the full-length seven-transmembrane MOR1.⁵⁰

MOR splice variants in the clinical context

A small number of human studies have examined the impact of MOR subtypes in the context of clinical conditions. In a study of CNS cell lines from the subjects infected with human immunodeficiency virus (HIV), researchers demonstrated that there are indeed specific differences in the MOR variant expression profile among CNS cell types (astroglia, microglia and neurons), and that the expression levels of the these variants are differentially regulated by HIV type 1. Importantly, the data suggest that expression of MOR variants may be differentially regulated in the brains of HIV-infected subjects with varying levels of neurocognitive impairment.⁵¹

The association between MOR splice variants and opioid-related side effects has also been examined; Chinese women undergoing gynaecological surgery were genotyped for single nucleotide polymorphisms (SNPs) in the OPRM1 gene. Results have shown an association between a SNP in the splice variant, MOR1X, and occurrence of fentanyl-induced emesis in the postoperative setting.⁵²

Furthermore, in a model of mice with ‘morphine-induced scratching’ (MIS), researchers propose that a MOR variant, MOR1D, is essential for MIS, whereas the major protein transcript of OPRM, MOR1, is necessary for mediating analgesia and that these processes may be independent of each other.⁵³

Conclusion

The pharmacology of mu opioids is highly complex. So too is an individual’s clinical response to opioids. It is clear now that MOR subtypes exist and that they might, in part, explain why there is intra- and inter-individual variation in response to opioids.

A growing body of evidence suggests that MOR splice variants are functionally and pharmacologically active, and that differential activation of variants by mu opioids occurs through alternative binding, through differential receptor activation and as a result of the distinct distribution of variants located regionally and at the cellular level.

We have learned that, in MOR splice variants, exon 1 is necessary for morphine analgesia whereas exons 2 and 3 are required for activity of diamorphine and M6G,^34,35 and that novel exon 11 is important in mediating activity of diamorphine, M6G and fentanyl but not morphine or methadone.⁴⁰ Furthermore, heterodimers of truncated variants have been shown to form novel therapeutic targets, and activation of novel targets has made it possible to dissociate analgesic effect from adverse effect.⁴⁹ These data are exciting because they shed light on potential mechanisms underpinning differential pharmacological properties of the mu-opioid system, and raise the possibility that novel therapeutic agents might be engineered to activate analgesic targets, or be designed on the basis of their relative activation profile rather than simply on their binding affinity.

However, there remains an enormous challenge to turn this data into clinically meaningful algorithms to benefit patients. We will need to understand more about splice variants, realise their pharmacological and functional potential and establish to what extent this mechanism contributes to the overall clinical response of an individual to a given mu-opioid agonist such as morphine.