Volume 347:362-364

August 1, 2002

Number 5

Knocking Out the DREAM to Study Pain

Although some mechanisms by which acute pain evolves into a chronic syndrome are known, many others, including changes in the brain stem, thalamus, and cerebral cortex, are not understood. Moreover, even though chronic neuropathic and inflammatory pain can be relieved to some extent with opiate compounds, these syndromes are still poorly understood. The treatment options for pain with a central cause are even more bleak. Now, however, the revolution in genomic engineering has opened up new possibilities for studying the mechanisms of acute and chronic pain and designing novel and more selective drugs.

Prodynorphin is the precursor of dynorphin, an opioid neuropeptide. Expression of the prodynorphin gene is controlled by a calcium-regulated transcription factor that binds to a downstream regulatory element (DRE) of the gene and is termed DRE antagonistic modulator (DREAM). When DREAM binds to DRE, it inhibits transcription of the prodynorphin gene (Figure 1A).¹ In view of the role of dynorphin in spinal and supraspinal modulation of afferent nociceptive signals and its likely involvement in chronic-pain syndromes, the DREAM transcriptional regulator should be an important target for assessing nociceptive transmission.

View larger version (74K): [in this window] [in a new window]

Figure 1. Effect of the Presence (Panel A) and Absence (Panel B) of Downstream Regulatory Element Antagonistic Modulator (DREAM) on Nociceptive Processing after Noxious Stimulation of the Skin with a Pin. In Panel A, the prodynorphin gene is blocked by DREAM, and the nociceptive signal is transmitted by substance P (SP)–containing dorsal-root neurons essentially unaltered through the spinal cord. This activated state is characterized by a large depolarizing potential and multiple spike discharges. Such activity subsequently activates systems governing descending regulatory control and thalamic nuclei, triggering the perception of pain. In Panel B, DREAM has been knocked out, and the release of dynorphin from spinal cord interneurons becomes a prominent part of the nociceptive response. This inactivates spinal-projection neurons, leading to a greatly reduced excitatory response. Subsequently, there is almost complete inactivation of descending regulatory and pain-processing systems. In addition to raising issues related to the specific functions of DREAM, this simple scheme emphasizes the importance of neurophysiological research involving this model and hypotheses to guide drug design.

Cheng et al.² deleted DREAM in mice by targeted disruption of the gene and performed a number of groundbreaking studies of the mechanisms by which DREAM regulates nociception. The DREAM-knockout animals had essentially normal motor control, spatial learning, and anxiety in open field tests but significantly reduced responses to acute thermal and mechanical stimulation of the skin and viscera and decreased responses in neuropathic and inflammatory models of chronic pain (Figure 1B). These features make DREAM-knockout mice ideal for assessing transmitter systems that regulate nociception and reduce pain behavior mediated by kappa-opioid receptors.

Opioids interact with receptors at all levels of nociceptive and pain processing. The principal opioid receptors are the mu, kappa, and delta subtypes. The effectiveness of compounds such as morphine, which are selective for mu-opioid receptors, reflects the wide distribution of mu-opioid receptors and the ability of such compounds to block the transmission of the nociceptor to the spinal cord and to control the perception of pain in the cerebral cortex. Dynorphin is released by interneurons in the spinal cord, binds to kappa-opioid receptors on spinal-projection neurons, and has analgesic effects. It appears, however, that dynorphin is not expressed by nociceptors, is elevated in chronic-pain syndromes, and through its actions at N-methyl-D-aspartate (NMDA) receptors,³ may actually enhance nociceptive transmission, under some circumstances, rather than reduce it. For these reasons, the expression of dynorphins must be under the tight control of DREAM.

Critical evidence of a specific role of DREAM in pain processing² was provided by gel-shift assays showing calcium- and dose-dependent binding of prodynorphin DRE to DREAM and a lack of DREAM in the DREAM-knockout mice. The DREAM-knockout mice had elevated levels of messenger RNA (mRNA) for prodynorphin in the spinal cord, normal levels of mRNA for the opioids pro-opiomelanocortin and proenkephalin and the DRE-containing c-fos, and no changes in the levels of kappa-opioid or NMDA receptors in the spinal cord. The decreased responses to pain in DREAM-knockout mice were mediated by kappa-opioid receptors, as was shown with the kappa-selective antagonist norbinaltorphimine dihydrochloride.

These studies also assessed the NMDA receptors in DREAM-knockout mice. The NMDA receptor mediates the entry of calcium into spinal neurons through the activation of glutamatergic nociceptors. In the DREAM-knockout mice, however, a selective NMDA antagonist (MK-801) had no effect on the tail-flick test, which measures the interval between noxious heating of the tail and withdrawal of the tail from the stimulus. Although MK-801 increases the threshold for tail withdrawal in normal animals, the drug did not change the response in knockout mice, which suggests that the actions of dynorphin in the knockout mice are not mediated by an NMDA mechanism, as they are in normal animals. Since DREAM may also be regulated independently of nuclear calcium levels,⁴ the actions of dynorphin in this model may not be directly linked to central sensitization.

Many new questions arise from these pivotal findings. First, what is the specificity of DREAM in nociceptive and pain processing? All pain-induced behavior appears to be interrupted at a nodal point in the spinal cord of these knockout animals. Although this could prove to be an important feature of DREAM, it means that descending inhibitory functions in the brain stem and other functions of central-pain processing will not be available for study because they have been permanently blocked. Regulatory specificity is being uncovered in animals with other knockout targets. For example, deletion of the dopamine- hydroxylase gene widely blocks expression of noradrenaline, yet it inhibits hyperalgesia for thermal but not mechanical stimulation.⁵ Second, what is the involvement of the supraspinal midbrain and medial thalamus in altered pain processing in the DREAM-deficient animals? Although the hippocampus was analyzed for prodynorphin levels, there was no assessment of pain mechanisms in the forebrain in regions established as nociceptive. Third, what agents, processes, and genes regulate the expression of DREAM, and how is it regulated during acute pain in normal animals? Fourth, how can this gene serve as a target for therapy? Compounds that regulate the expression of neuronal genes are not yet available. Indeed, we must evaluate other systems involved in establishing chronic pain using microarray analysis in DREAM-knockout animals during acute and chronic pain.

There are some difficulties inherent in the use of the DREAM-knockout model to study pain processing. Since all nociceptive processing is blocked at a nodal point in the spinal cord, this model may not tell us much about the unique causes of neuropathic and inflammatory pain. DREAM-knockout mice express prodynorphin in the ventral horns, but this has not been observed in normal adult mice. Moreover, chronic pain usually begins in adulthood, yet the effect in knockout mice begins very early in life. Further investigation of these neurons could provide insight into dormant systems that may be useful for targeting drugs. It should also be noted that chronic pain reduces nociceptive responses of the anterior insular and cingulate cortices. Such cortical alterations must be identified in animal models because these are the sites of pain-induced stress and anxiety. Although the direct therapeutic value of the knockout model is limited, targeted delivery of DNA complexes to the central nervous system is possible.⁶

The DREAM-knockout model may be useful for designing drugs that can directly regulate genes, and the side effects of such an approach may be reduced by the use of site-specific expression vectors. There is no doubt that merging drug design with genetic engineering has great potential. The dream is that these engineering techniques will elucidate the mechanisms of the least understood pain syndromes and provide selective targets for effective relief of pain, even for the most intransigent pain syndromes, such as central pain.

Brent A. Vogt, Ph.D.
State University of New York Upstate Medical University
Syracuse, NY 13210

Supported by a grant (NS38485) from the National Institutes of Health.

References

1. Carrion AM, Link WA, Ledo F, Mellstrom B, Naranjo JR. DREAM is a CA²⁺-regulated transcriptional repressor. Nature 1999;398:80-84.[ISI][Medline]

2. Cheng H-YM, Pitcher GM, Laviolette SR, et al. DREAM is a critical transcriptional repressor for pain modulation. Cell 2002;108:31-43.[ISI][Medline]

3. Laughlin TM, Larson AA, Wilcox GL. Mechanisms of induction of persistent nociception by dynorphin. J Pharmacol Exp Ther 2001;299:6-11.[Abstract/Full Text]

4. Ledo F, Carrion AM, Link WA, Mellstrom B, Naranjo JR. DREAM-CREM interaction via leucine-charged domains depresses downstream regulatory element-dependent transcription. Mol Cell Biol 2000;20:9120-9126.[Abstract/Full Text]

5. Jasmin L, Tien D, Weinshenker D. et al. The NK1 receptor mediates both the hyperalgesia and the resistance to morphine in mice lacking noradrenaline. Proc Natl Acad Sci U S A 2002;99:1029-1034.[Abstract/Full Text]

6. Meuli-Simmem C, Liu Y, Yeo TT, et al. Gene expression along the cerebral-spinal axis after regional gene delivery. Hum Gene Ther 1999;10:2689-2700.[ISI][Medline]

Edward E. Rylander, M.D.

Diplomat American Board of Family Practice.

Diplomat American Board of Palliative Medicine.