A complex interplay of sex, genotype, and emotional state can determine how people process pain and respond to an analgesic. That is the conclusion of a new investigation into the underpinnings of pain sensitivity in mice and men. The study, from collaborators Jeffrey Mogil at McGill University in Montreal, Canada; Roger Fillingim at the University of Florida, Gainesville; and Inna Belfer at the University of Pittsburgh, Pennsylvania, reveal that genetic differences in a receptor for the peptide hormone vasopressin influence pain sensitivity, but only in males (human or murine) under stress. Sex, stress levels, and genes also affect the ability of the vasopressin analog desmopressin to reduce inflammatory pain in both mice and people.
The results, published online October 23 in Nature Neuroscience, show that activation of the vasopressin receptor pathway can relieve pain, but that the analgesic effects of desmopressin or other vasopressin-targeted treatments might be easy to miss in a real-world stew of sex differences, treatment anxiety, and genetic variation.
“Characteristics of individuals—how stressed they feel at the time, and their genotype and sex—seem to profoundly influence their response to this drug,” Fillingim told PRF. The finding, he said, highlights the peril of considering a group of individuals in a drug trial as a uniform bunch: Doing that may hide a drug’s effect.
There may be a similar message for genetic association studies, Mogil said. When a gene effect is found in one study but fails to hold up in another, it could be because “these gene effects interact with other things that are different from one study to the next,” he said. “Here is an example of just how complicated it can be.”
In mice, a region of chromosome 10 has been shown to associate with pain sensitivity in models of inflammatory pain (Nair et al., 2011; Wilson et al., 2002). In their new study, Mogil (who is a PRF science advisor) and his colleagues used haplotype mapping to pinpoint the effect to Avpr1a, the gene encoding the vasopressin 1A receptor (V1AR). A pain-resistant mouse strain displayed increased levels of Avpr1a transcripts, especially in the dorsal root ganglion and spinal cord, suggesting that higher receptor expression protects animals from pain. In agreement with that idea, mice lacking the Avpr1a gene showed increased pain behaviors (licking and biting) in response to formalin and capsaicin injection into the hind paw.
To see whether variation in the vasopressin receptor gene associates with pain sensitivity in humans, Belfer and colleagues analyzed two AVPR1A single-nucleotide polymorphisms (SNPs) in a group of healthy volunteers who had previously reported pain ratings in response to capsaicin. In the group as a whole, the researchers saw no association between pain and AVPR1A genotype. However, they found that AVPR1A genotype did correlate with pain ratings for a subset of subjects: For men who reported high stress levels at the time of testing, pain ratings were lower in those with a G allele of the SNP rs10877969.
The team wondered whether treating people with vasopressin might impact capsaicin pain. (Previous studies have indicated that vasopressin relieves acute pain, although results in people have been inconsistent; see Honda and Takano, 2009; Pohl et al., 1996.) When Fillingim and colleagues treated subjects with intranasal desmopressin (a synthetic vasopressin analog), the average capsaicin pain ratings were identical between the treated and untreated groups. However, the drug did help some subjects, namely those who were given desmopressin on their second day of testing; people who received the drug on their first visit experienced no pain relief. Reasoning that the first round of pain testing might be a high-stress situation, the researchers looked for a correlation between stress levels and desmopressin effects. They found that, indeed, subjects who reported high stress when they received desmopressin got little benefit, but subjects who reported low stress had reduced pain. The researchers also saw that stress, vasopressin analgesia, genotype, and sex interacted in these subjects: Analgesia correlated with stress level only in male subjects homozygous for the A allele of rs10877969.
Follow-up studies with male mice recapitulated the interaction of genes, sex, and stress seen in people. The researchers compared mice that had spent time in the testing environment versus those that were new to it, and found that, consistent with their results in people, administering vasopressin relieved capsaicin pain only in habituated (less stressed) mice. Also, as in people, stress and the Avpr1a gene interacted in setting pain sensitivity, because mice lacking Avpr1a had increased capsaicin-induced pain only when the animals were stressed. Looking into mechanism, the investigators found that the receptor knockout mice were deficient in stress-induced analgesia, suggesting a critical role for Avpr1a in endogenous pain control mechanisms.
Putting all of the data together, Mogil and his collaborators suggest a mechanism by which vasopressin, genotype, sex, and stress interact: Stress activates endogenous vasopressin analgesia through V1AR, depending on sex and AVPR1A genotype. Administering vasopressin can relieve pain through the same mechanism, but only if the pathway has not already been activated by stress.
Fillingim says the results motivate him to assess stress, and perhaps try to ameliorate it, in analgesic trials. In this study, stress appeared to be masking drug effects by activating the same analgesic mechanism. “But there might be other scenarios in which stress, or anxiety, or negative expectations might be activating neurochemistry that interferes with the analgesic effect of a drug,” he noted. (For more on the effect of negative expectations on drug efficacy, see PRF related news story.)
The results also underscore the idea that “pain circuitry in males and females is not the same,” Mogil said. In further support of that idea, he and his colleagues published another study this week, showing evidence that, while spinal pain processing involves toll-like receptor 4 (TLR4) in male mice, in females TLR4 does not play a role (Sorge et al., 2011).
Comments
Marshall Devor, Hebrew University of Jerusalem
This paper is quite
This paper is quite complicated, but solid, all in all interesting, and maybe even a beacon in the emerging area of pain genetics.
In the Introduction, Mogil and coworkers make the point that this is one more recent success of the strategy of using genomewide searches in mice (linkage analysis was used here) and then testing candidate genes found for relevance to human cohorts. Hidden behind this statement is the fact that genomewide association studies can be done directly in human populations, but they are expensive. Funding agencies have supported hundreds of such GWAS for a variety of medical conditions, but for reasons that are not entirely clear to me, pain studies have been excluded. This is despite the huge burden of chronic pain in society, and the enormous financial costs involved in treatment and compensation. There is still not enough awareness that chronic pain is not "just" a symptom, but a major biomedical problem in its own right. I know of only one GWAS on pain published so far.
A second interesting point has to do with the gene itself. Avpr1a is the gene that codes for the 1A vasopressin receptor. Polymorphisms in this same gene have been reported to associate with various biological phenotypes, as acknowledged in the opening sentence of the paper. But most readers will not be aware of what a large and quirky variety of phenotypes is on the list. After a quick look in PubMed, I came up with the following traits associated with Avpr1a, in addition to pain from capsaicin irritation in relaxed male mice: generosity, hypertension, alcohol drinking, maternal behavior, altruism, creative dance performance (in human dancers), fear conditioning, sleep, and circadian rhythms. Do these phenotypes have something deep in common among them? Perhaps, but not necessarily.
Whole-genome scans are the way to go when one is looking for novel connections, unbiased by the currently known physiology. But the fact that a particular genetic polymorphism occurs somewhat more in one cohort than another and is hence "associated" with phenotypic differences between the two does not necessarily mean that the gene "mediates" that trait, if the word "mediates" is used in the sense of "causes" or "depends on." The reason for linkage may be quite distant from the neural networks that "cause" the trait, especially a complex trait like pain. I like to relate the following scenario to students: "Imagine a gene that codes for a muscle protein. Boys born with a particular allele of this gene might be muscular, handsome, athletic, and successful with girls. The resulting self-confidence might increase the likelihood that they engage in dangerous sports, and hence the likelihood of injuries and of pain. For this reason, the muscle protein is a ‘pain gene,’ and will likely come up with significant linkage in a GWAS of young men with and without frequent pain. But does the muscle protein ‘mediate‘ pain?" There is association, for sure, but uncovering the reason for the association may be difficult and not necessarily rewarding. Avpr1a is a pain gene, but only time will tell if it is an important new player in the physiology of pain. The authors’ strategy of returning once again to mice after checking the gene in people, and finding evidence for predicted analgesic effects of gene-related compounds, is a promising sign that it might, in fact, open a new page in pain physiology.