Tobacco use remains the leading preventable cause of death in the world. According to the World Health Organization (WHO), tobacco use is responsible for the death of approximately six million people and causes hundreds of billions of dollars of economic loss worldwide each year[1]. In 2010, the Global Adult Tobacco Survey (GATS) China conducted by the Chinese Center for Disease Control and Prevention reported that 28.1% of adults in China were current smokers, and the prevalence of smoking in men was significantly higher than that in women (52.9% vs. 2.4%)[1]. There was an estimation of 301 million current smokers in China, making this country the largest consumer of tobacco in the world[2]. Cigarette smoking increases the risk of numerous health problems and causes a variety of tobacco-related diseases, including primarily cancers, diabetes, cardiovascular diseases, and chronic lung diseases. Up to half of the world’s 1 billion smokers will eventually die from a tobacco-related disease[1].

Cessation reverses most adverse effects of smoking. The majority of smokers are highly motivated to quit, but confronted with high relapse rate after stopping smoking[3]. The same issue is confronting the patients with tobacco-related diseases. Although some smoking patients quit because of diseases, there is still a considerable proportion of them who persist smoking even after being informed that tobacco smoking is an independent risk factor for their diseases. As several surveys revealed, the percentage of current smokers accounting for the patients with tobacco-related diseases was compared to that of general population[4–6]. Despite great progress in pharmacological treatment of nicotine dependence, the efficacy of available treatments is limited, and only 15–30% of smokers permanently abstain from smoking[3].

Tobacco smoking is believed to be a complex and multifactorial behavior with both genetic and environmental determinants. Recent researches strongly suggested that smokers varied in their underlying genetic susceptibility to smoking addiction, and have found significant genetic influences on several aspects of smoking behaviors, including smoking initiation, smoking maintenance and cessation success[7, 8]. Variations in several genes have been suggested to contribute to smoking behaviors, and these researches have been focused on two broad classes of candidate genes: 1) genes that may influence the response to nicotine (e.g. nicotine metabolism and nicotinic receptors) and 2) genes that may predispose to addictive behavior because of their effects on key neurotransmitter pathways (e.g. dopamine and serotonin)[9, 10].

Although multiple neurobiological mechanisms have been implicated in mediating the reward and reinforcing effects of drug abuse including nicotine, for the past decade increasing studies pointed to the endogenous opioid system, especially the mu-opioid receptor (MOR)[11, 12]. Genetic and pharmacological studies indicated that mu-opioid receptors were mainly involved in nicotine reward[10, 13–16]. These findings have led us to consider the potential role of variations in genes encoding proteins regulating the function of the MOR. Several proteins were manifested to interact with the MOR in in-vitro models and had some opioid-related phenotypes demonstrated in mouse knock-out models[17]. However, it remains uncertain whether and to what extent common genetic variants in the genes of MOR interacting proteins affect smoking behaviors on account of limited population data.

Considering higher prevalence of smoking in Chinese men compared with women, we conducted a cross-sectional study in Chinese male smokers to explore the associations between smoking behaviors and genetic variants in the MOR and MOR-interacting protein genes, including OPRM1 gene encoding mu-opioid receptor and two genes encoding mu opioid-receptor-interacting proteins, namely ARRB2 and HINT1.

The present study focused on the potential association of SNPs in genes encoding MOR (OPRM1) and MOR-interacting proteins (ARRB2 and HINT1) with smoking behaviors. The results generated from Chinese men with tobacco-related diseases suggested that HINT1 rs3852209 was significantly associated with smoking status (current smokers vs. ex-smokers). However, it did not provide evidence for the effect of variations in OPRM1, ARRB2 and HINT1 genes on smoking behaviors including age of smoking initiation, daily cigarette consumption, and FTND score. In addition, the logistic regression analyses revealed that smokers were increasingly probable to achieve smoking cessation with age independent of the other covariates, which was consistent with the results described in previous trials and survey studies[20, 21].

Recently, we have demonstrated that there was a considerable percentage of smoking patients with tobacco-related diseases did not successfully quit smoking, and were still current smokers. Most of these patients were willing to quit, especially after having recognized tobacco smoking as an independent risk factor for their diseases. But unfortunately, they were confronted with as high rate of relapse as general population[22]. From the above, it can be seen why an ex-smoker is defined as one who has quit smoking for at least one year. This assertion is supported by the previous evidence that self-reported cessation for at least one year had the highest likelihood of continuous abstinence of smoking[23]. In the present study, 83.3% of current smokers attempted to quit and used to be abstinent for a short time, but finally relapsed. Although the average age of current smokers was lower than that of ex-smokers, both of current smokers and ex-smokers had the same number of smoking years at enrollment. Thus, it was indicated that current smokers initiated smoking earlier than ex-smokers. Overall, the present study revealed the differences of smoking behaviors between current smokers and ex-smokers, and suggested that current smokers might have more difficulties in quitting.

Although the mesolimbic dopamine hypothesis is the most influential theory of nicotine reward and reinforcement, recent developments in the neurobiology of nicotine dependence have identified several other neurotransmitter systems that may contribute to the addictive properties of nicotine as well. In this regard, the brain opioidergic system has been of interest. Nicotine stimulates the release of endogenous peptide β-endorphin which has the high affinity for the mu-opioid receptor[24]. The activated mu-opioid receptors suppress the release of neurotransmitter GABA[25]. Thus, the decrease of GABA level causes the disinhibition of VTA dopamine neurons, which leads to the increase of dopamine release from nucleus accumbens[26]. Presumably, it is the association between the opioidergic system and the mesolimbic dopamine system that appeals to explore the role of mu-opioid receptors in smoking behaviors. The previous studies using inbred and knockout mice strongly indicated that the mu-opioid receptor mediated both the positive and negative reinforcing effects of nicotine. And it was found that nicotine induced reward effect and dependence was substantially attenuated in mu-opioid receptor knock-out mice[10, 13]. All these suggest that mu-opioid receptor, encoded by the gene of OPRM1, is one of the factors contributing to tobacco addiction, and thus the OPRM1 becomes an attractive candidate gene for smoking behaviors in humans. The Asn40Asp (A118G) polymorphism (rs1799971), located in exon I of OPRM1 gene, induces functional change of the mu-opioid receptor and becomes one of the most studied polymorphisms[27]. This mis-sense SNP rs1799971 causes the substitution of an aspartate (Asp) for an asparagine (Asn) at position 40 in the amino-acid sequence. The Asp40 (G) variant, related to the reduced expression of OPRM1, was found to increase the binding affinity of β-endorphin for the mu-opioid receptor by three-fold compared with the wild-type Asn40 (A) OPRM1[13], nevertheless this finding subsequently failed to be replicated[28, 29]. In rodent studies, there was evidence that OPRM1 G allele might play a role in susceptibility to nicotine addiction[11]. In the clinical trial, Lerman and colleagues found that Asn40Asp (A118G) mis-sense SNP in OPRM1 might predict the treatment responses to NRT (nicotine replacement therapy). Specifically, the smokers with one or more copies of G allele (Asp40 variant) were more likely to quit after a course of NRT than individuals with two copies of A allele[16]. Their results might be explained by the fact that smokers with G allele could gain more reward from nicotine provided by NRT and then derived greater benefit from NRT. However, the present study did not provide strong support for the role of the OPRM1 polymorphism (rs1799971) in smoking behaviors. To be specific, there were no statistical associations identified between the OPRM1 Asn40Asp (A118G) polymorphism and smoking status as well as other smoking behaviors (including age at smoking initiation, daily cigarette consumption and nicotine dependence) in Chinese males.

Furthermore, the mu-opioid receptor directly interacts with multiple proteins called MOR interacting proteins (MORIPs), and thus MORIPs gene polymorphisms are considered to influence on the individual susceptibility to drug dependence. Of the interacting proteins, β-arrestin 2 and mPKCI-1 have demonstrable opioid-related phenotypes in mouse knock-out model and other in-vitro models[17]. By binding to phosphorylated MOR, β-arrestin 2 has been found to be an important regulator of signal transduction mediated by opioid receptors through promotion of receptor desensitization and internalization and plays a role in opioid reward[17, 30, 31]. ARRB2 is known as the gene encoding β-arrestin 2. In ARRB2 null mutant animals, MOR did not undergo desensitization and then the increased and prolonged antinociceptive effects were perceived after morphine challenge[32]. In contrast to β-arrestin 2, mPKCI-1 (known as histidine triad nucleotide binding protein 1 in humans) encoded by the gene HINT1 decreases MOR phosphorylation and desensitization. It has been shown that mPKCI-1 knock-out mice could develop tolerance to analgesic effect of morphine much more quickly than wild-type mice[33]. In the present study, the results indicated a significant association between HINT1 variant rs3852209 and smoking status. Compared with CC genotype, the subjects with HINT1 rs3852209 T allele were less possible to achieve smoking cessation. Although it is unknown how HINT1 rs3852209 T allele affects HINT1 gene expression, presumably T allele leads to the deficiency of mPKCI-1 expression, which is incapable to inhibit MOR desensitization. As a result, the individual with T allele develops tolerance to nicotine and is susceptible to nicotine dependence and then less likely to quit smoking. The haplotype analyses did not provide evidence for the relationship between ARRB2 or HINT1 and smoking status, which is consistent with the results from the previous clinical NRT trial[19].

After all, this is the primary study exploring the role of genetic variations of MOR-interacting proteins in smoking behaviors. Several limitations in our study should be considered. First, one major limitation of this study was the limited sample size with the power about 20% (OR: 1.2 - 1.3) calculated by Quanto, which was not sufficient to determine the associations between genetic variations and smoking behaviors and might lead to false negative results which meant missed potential positive associations. Given the minor genetic effects, the results of our study might be caused by the limited sample size. Further studies with sufficient statistical power are required to verify the present results. Second, the mismatch of age at enrollment between current smokers and ex-smokers might influence the results to a certain extent, though the adjustments were adopted. Third, the study focused on the subjects with tobacco-related diseases enrolled from a general hospital. Considering the variety of diseases and differences in severity and course of the diseases, it was difficult to precisely determine the effect of tobacco-related disease on smoking behavior. At last, smoking history was accessed by the participants' self-report without biochemical confirmation for current smoking status, which might possibly cause misclassification.

Despite these limitations, the present study preliminarily investigated the role of genetic variations of the opioid pathway in smoking behaviors of Chinese men. Our results revealed no significant association between common genetic variations in genes encoding MOR and MOR-interacting proteins and smoking behaviors of Chinese men, and presented suggestive evidence for the association between the MOR-interacting protein encoding gene and smoking status (current smokers vs. ex-smokers). Finally, further prospective studies need to be replicated in order to confirm the findings of the present study.