The dominant active mutant, daDREAM, blocks the regulated activity of endogenous DREAM/KChIP proteins when transgenic to endogenous mRNA ratio is as low as 1 to 6 . In this study, we show that in L1 transgenic mice the ratio daDREAM/DREAM mRNA is 1.6 to 1 and 1 to 3 in spinal cord and DRG, respectively, indicating that the expression of the dominant active mutant protein is sufficient to block Ca2+- and cAMP-mediated derepression by endogenous DREAM/KChIP proteins. As a result, expression of prodynorphin and BDNF, two potential targets for transrepression by DREAM that are related to pain, are reduced in DRG and spinal cord from L1 transgenic mice. Since the mutation at the LCD domain in daDREAM prevents its interaction with CREB , downregulation of prodynorphin and BDNF in L1 mice is not related to interference with CREB-dependent transcription and is due to direct repression by daDREAM. Importantly, changes in nociception are related to the expression of the transgene in spinal cord and DRG neurons, since sensory thresholds were normal in L26 mice in which daDREAM expression is confined exclusively to telencephalic areas. In the same way, another line of daDREAM mice, L33, without transgene expression in spinal cord and DRG neurons, did not show any change in pain sensitivity to thermal stimulation .
Previous work with DREAM-/- mice  demonstrated the involvement of this transcriptional repressor in pain modulation. Mice lacking DREAM showed a basal state of analgesia and reduced response to inflammatory and neuropathic treatments, that was interpreted as caused by elevated spinal cord levels of dynorphin A peptide observed in these mice . Consistent with this, the reduced expression of prodynorphin in L1 mice could account for the basal state of hyperalgesia found in adult mice. The basal hyperalgesia was paralleled by increased hyperreflexia in the isolated spinal cord without signs of change in the function of sensory afferents, suggesting that the exaggerated sensitivity is related to changes in spinal processing of afferent signals.
Activation of NMDA receptors plays a central role in the transmission of primary sensory information in the spinal cord [reviewed in ]. Recently, it has been shown that DREAM reduces the amplitude of NMDA currents in hippocampal neurons through a Ca2+-sensitive interaction between DREAM and PSD95  or between DREAM and the NR1 subunit of the receptor . On the other hand, it was previously shown that release of dynorphin peptides from presynaptic terminals inhibits glutamatergic trasmission through NMDA receptors . The increased hyperreflexia in isolated spinal cord and the increased basal sensitivity to pain stimulation suggest that a potential reduction in NMDA currents in spinal cord neurons in daDREAM transgenic mice is compensated by the reduced inhibition due to the low expression of prodynorphin.
In addition to the hypothesis of prodynorphin involvement, there is a convergence of data suggesting that transient A-type potassium currents, mediated by Kv4 channels, are responsible for hypersensitivity to acute pain stimuli. Genetic elimination of Kv4.2 reduces A-type currents and increases excitability of dorsal horn neurons, resulting in enhanced sensitivity to noxious stimuli , that resembles the scenario in L1 mice. Moreover, genetic ablation of Kv channels results in decreased expression of DREAM/KChIP proteins , suggesting a genetic auto-regulatory loop between these two gene families. Our results, however, are not consistent with the hypothesis that reduced expression of Kv channels in the spinal cord or malfunction of the associated currents, are responsible for the altered noxious sensitivity in L1 transgenic mice. On the contrary, we found small but significant increase in Kv4.2 and Kv4.3 mRNA levels in the spinal cord and, more important, we found that IA currents in dorsal horn neurons from L1 mice were indistinguishable from those of wild type mice in terms of various properties, including current density and kinetics.
Reduced prodynorphin levels may explain basal hypersensitivity of L1 mice, however, does not account for the reduced behavioral response to inflammation following CFA injection. Importantly, isolated transgenic spinal cords were not capable to show signs of central sensitization, which may be causal for the anomalous response to inflammatory stimuli. Whereas spinal cords from carrageenan-treated wild type mice showed a marked hyperreflexia, L1 mice with or without treatment, showed essentially similar spinal reflexes. Furthermore, persistent low frequency stimulation of primary C-afferents caused a long-term enhancement of DR-VRRs in wild type but not in L1 mice. It is worth noting, that similar protocols of stimulation  have been shown to produce LTP in superficial laminae neurons in a process that resembles central sensitization induced by inflammation.
In order to understand how L1 mice fail to produce a normal process of central sensitization, we considered the reduced expression levels of BDNF in L1 mice. Reduced BDNF levels in vivo confirm previous in vitro studies showing the regulatory effect of DREAM on BDNF promoter activity . Several lines of evidence relate BDNF with central sensitization; i) intrathecal administration of exogenous BDNF produces hyperalgesia in wild type mice, whereas administration of antiserum directed against either BDNF or TrkB receptors prevent inflammation-induced hyperalgesia [29, 36], ii) conditional BDNF knockout mice do not develop hyperalgesia after inflammatory stimuli  and iii) at the functional level, BDNF has been shown to be required for simple forms of spinal reflex plasticity like wind-up  and to enhance the spinal response to sensory inputs [16, 39] through post-synaptic NMDA receptors  or by reversing chloride gradients in central afferent terminals . Here we show that BDNF is required also for longer lasting forms of plasticity in spinal reflexes. BDNF-/- mice did not develop long-term enhancement of ventral root reflexes following low frequency stimulation of C-fibers. Interestingly, BDNF-/+ mice were able to develop a small amplitude enhancement of reflexes, suggesting that the quantitative expression of BDNF in the spinal cord is correlated with the ability to develop long-term forms of spinal plasticity.
DREAM transgenic mice express normal levels of TrkB receptors in the spinal cord and therefore we were able to test the effects of exogenous applications of BDNF to the isolated spinal cord. BDNF produced a similar enhancement of spinal responses to afferent inputs in spinal cords from wild type and L1 mice, suggesting that the spinal cord from L1 mice does not receive enough BDNF from primary afferents after low frequency stimulation of C-fibers or after inflammation. The absence of induction in BDNF expression in dorsal root ganglia from L1 mice following inflammation supports this idea.