Caffeine stimulates locomotor activity in the mammalian spinal cord via adenosine A1 receptor-dopamine D1 receptor interaction and PKA-dependent mechanisms

Abstract

Caffeine is a potent psychostimulant that can have significant and widely variable effects on the activity of multiple neuronal pathways. The most pronounced caffeine-induced behavioral effect seen in rodents is to increase locomotor activity which has been linked to a dose-dependent inhibition of A1 and A(2A) receptors. The effects of caffeine at the level of the lumbar spinal central pattern generator (CPG) network for hindlimb locomotion are lacking. We assessed the effects of caffeine to the locomotor function of the spinal CPG network via extracellular ventral root recordings using the isolated neonatal mouse spinal cord preparation. Addition of caffeine and of an A1 receptor antagonist significantly decreased the cycle period accelerating the ongoing locomotor rhythm, while decreasing burst duration reversibly in most preparations suggesting the role of A1 receptors as the primary target of caffeine. Caffeine and an A1 receptor antagonist failed to stimulate ongoing locomotor activity in the absence of dopamine or in the presence of a D1 receptor antagonist supporting A1/D1 receptor-dependent mechanism of action. The use of caffeine or an A1 receptor blocker failed to stimulate an ongoing locomotor rhythm in the presence of a blocker of the cAMP-dependent protein kinase (PKA) supporting the need of this intracellular pathway for the modulatory effects of caffeine to occur. These results support a stimulant effect of caffeine on the lumbar spinal network controlling hindlimb locomotion through the inhibition of A1 receptors and subsequent activation of D1 receptors via a PKA-dependent intracellular mechanism.

Keywords: Adenosine; Caffeine; Dopamine; Locomotion; Mouse; Spinal cord.

Figure 1. Experimental setup

A: Suction recording electrodes were placed to monitor motor activity from ventral root nerve before, during and after perfussion of drugs. B: Extracellular recordings from rL2, and lL2 ventral roots after application of 6μM NMDA and 9μM 5-HT and 18μM dopamine (DA), showing locomotor-like activity characterized by left–right alternation in a P2 spinal cord. Raw (upper two traces) and rectified (lower two traces) recordings are shown. C: Representative segment of a rectified and smoothed trace from an actual control ventral root recording showing the locomotor-related parameters and how there were measured.

Figure 2. Effects of different concentrations of caffeine on locomotor behavior

A: Circular plots showing the effects of increasing concentrations of caffeine (1, 10, 50 and 100μM) to the phasing between rL2 and lL2 or rL2 and rL5 ventral roots (Note: cycles of motor burst activity located at 0.5 are considered in alternation while 0 is considered synchronous activity). Cycles of activity located outside the inner circle have a correlation coefficient (r value) which is significant with the type of locomotor-like activity to which it is related (alternation/synchronization). Concentrations of caffeine up to 50μM did not disrupt ongoing drug-induced locomotor behavior in most preparations (7/8). B panels: Time-course plots showing the effects of the application increasing concentrations of caffeine (1, 10 and 50μM) on locomotor cycle period. Each point represents 1 minute worth of recording. Comparing concentrations that did not disrupt ongoing locomotor behavior, application of 50μM consistently and reversibly decreased cycle period in most preparations (7/8).

Figure 3. Effects of caffeine on locomotor-related output parameters

A panels: raw (top), rectified and integrated (bottom) traces showing the effects of caffeine on pharmacologically induced (9μM 5-HT, 6μM NMDA, 18μM dopamine) locomotor activity recorded from the lumbar ventral roots of an isolated neonatal mouse spinal cord preparation. B panels: time-course plots showing a decrease in locomotor burst amplitude (B₁) which is not statistically significant and a decrease in motor burst duration (B₂) and cycle period (B₃) which was statistically significant (50μM, 20 minutes; n = 7). Each point represents 1 minute worth of recording. C: pooled data, averaged of 5 minutes worth of recordings in each condition, showing a significant decrease in burst duration and cycle period after the application of caffeine (n = 7). *Significantly different from control.

Figure 4. Effects of DPCPX, an A₁ receptor antagonist, on locomotor-related output parameters

A: time-course plots showing a decrease in locomotor burst amplitude which is not statistically significant and a significant decrease in motor burst duration and cycle period (1μM, 20 minutes; n = 5). Each point represents 1 minute worth of recording. B: pooled data, averaged of 5 minutes worth of recordings in each condition, showing a significant decrease in burst duration and cycle period after the application of DPCPX (n = 5). *Significantly different from control.

Figure 5. Effects of DPCPX, an A₁ receptor antagonist, and caffeine on locomotor-related output parameters

A: time-course plots not showing an effect on locomotor burst amplitude and a significant decrease in motor burst duration and cycle period after the application of DPCPX however caffeine did not exert any additional effects on any of the parameters measured (50μM caffeine; 1μM DPCPX; 20 minutes; n = 5). Each point represents 1 minute worth of recording. B: pooled data, average of 5 minutes worth of recordings in each condition, showing a significant decrease in burst duration and cycle period after the application of DPCPX but no additional effects after the application of caffeine in the presence of DPCPX (n = 5). C: Dose-response analysis of the effects of DPCPX on cycle period before and after the addition of caffeine showing that a concentration of DPCPX of 1μM produced the most significant effect while occluding the effects of caffeine in a reversible manner (n = 3 in each concentration). *Significantly different from control; #Significantly different from DPCPX.

Figure 6. Effects of SCH58261, an A_2A adenosine receptor antagonist, on locomotor-related output parameters

A: time-course plots showing no effects on either locomotor burst amplitude, motor burst duration and cycle period after the application of SCH58261 (1μM, 20 minutes; n = 5). Each point represents 1 minute worth of recording. B: pooled data, averaged of 5 minutes worth of recordings in each condition, showing no effects on either locomotor burst amplitude, motor burst duration and cycle period after the application of SCH58261 (1μM, 20 minutes; n = 5). *Significantly different from control.

Figure 7. Effects of adenosine, a broad spectrum adenosine receptor agonist, on locomotor-related output parameters

A: time-course plots showing no significant effects on locomotor burst amplitude and a significant increase in motor burst duration and cycle period after the application of adenosine (100μM, 20 minutes; n = 4). Each point represents 1 minute worth of recording. B: pooled data, average of 5 minutes worth of recordings in each condition, showing a significant increase in burst duration and cycle period after the application of adenosine (n = 4). *Significantly different from control.

Figure 8. Effects of CPA, a specific A₁ adenosine receptor agonist, on locomotor-related output parameters

A: time-course plots showing no effect in locomotor burst amplitude and a significant increase in motor burst duration and cycle period after the application of CPA (1μM; 20 minutes; n = 4). Each point represents 1 minute worth of recording. B: pooled data, average of 5 minutes worth of recordings in each condition, showing a significant increase in burst duration and cycle period after the application of CPA (n = 4). *Significantly different from control. C: Dose-response analysis of the effects of CPA on cycle period before and after the addition of caffeine showing that a concentration of CPA of 1μM produced the most significant effect while occluding the effects of caffeine in a reversible manner (n = 3 in each concentration). * Significantly different from control; # Significantly different from CPA; + Significantly different from CPA + caffeine.

Figure 9. Effects of caffeine and DPCPX in the absence of dopamine on locomotor-related output parameters

A: time-course plots from a locomotor rhythm elicited with 5-HT and NMDA in the absence of dopamine showing no effects in locomotor burst amplitude, burst duration or cycle period after the application of caffeine (50μM; 20 minutes; n = 4). Each point represents 1 minute worth of recording. B: pooled data, average of 5 minutes worth of recordings in each condition, showing no significant effects on any of the parameters measured after the application of caffeine in the absence of dopamine (n = 4). C: time-course plots from a locomotor rhythm elicited with 5-HT and NMDA in the absence of dopamine showing no effects in locomotor burst amplitude, burst duration or cycle period after the application of the A₁ adenosine receptor antagonist DPCPX (1μM; 20 minutes; n = 4). Each point represents 1 minute worth of recording. D: pooled data, average of 5 minutes worth of recording in each condition, showing no significant effects on any of the parameters measured after the application of DPCPX in

Figure 10. Effects of caffeine in the presence of SCH23390, a specific D₁ receptor antagonist and sulpiride, a specific D₂ receptor antagonist, on locomotor-related output parameters

A: time-course plots showing no effect in locomotor burst amplitude, burst duration and cycle period after the application of caffeine (50μM; 20 minutes; n = 4) in the presence of SCH23390. Each point represents 1 minute worth of recording. B: pooled data, average of 5 minutes worth of recordings in each condition, showing no significant effects on burst amplitude, burst duration and cycle period after the application of caffeine in the presence of SCH23390 (n = 4). C: time-course plots showing no effect in locomotor burst amplitude, while producing a statistically significant decrease in burst duration and cycle period after the application of sulpiride with not additional effects produced by the application of caffeine (50μM; 20 minutes; n = 4) in the presence of sulpiride. Each point represents 1 minute worth of recording. Each point represents 1 minute worth of recording. D: pooled data, average of 5 minutes worth of recordings in each condition, showing a significant decrease in burst duration and cycle period with no significant effects on burst amplitude after the application of sulpiride with no additional effects produced by the subsequent application of caffeine in the presence of sulpiride (n = 4).* Significantly different from control.

Figure 11. Effects of caffeine in the presence of 14–22 amide, a specific phosphate kinase A (PKA) inhibitor, and of forskolin, a specific activator of the protein adenylyl cyclase, on locomotor-related output parameters

A: time-course plots showing no effect in locomotor burst amplitude, burst duration and cycle period after the application of caffeine (50μM; 20 minutes; n = 4) in the presence of 14–22 amide (1μM). Each point represents 1 minute worth of recording. B: pooled data, average of 5 minutes worth of recording in each condition, showing no significant effects on burst amplitude, burst duration and cycle period after the application of caffeine in the presence of 14–22 amide (n = 4). C: time-course plots showing no effect in locomotor burst amplitude, while producing a statistically significant decrease in burst duration and cycle period after the application of caffeine (50μM; 20 minutes; n = 4) in the absence of dopmine from the locomotor-activating cocktail of drugs (5-HT and NMDA only) with significant effects on burst duration and cycle period by the application of forskolin (5–10μM; 20 minutes; n = 4) in the presence of caffeine. Each point represents 1 minute worth of recording. Each point represents 1 minute worth of recording. D: pooled data, average of 5 minutes worth of recording in each condition, showing a significant decrease in burst duration and cycle period with no significant effects on burst amplitude after the application of forskolin with no additional effects produced by the previous application of caffeine in the absence of dopamine in the perfusate (n = 4).* Significantly different from control.

Figure 12. Proposed cellular mechanisms mediating the modulation of locomotor activity through the activation or inhibition of the A₁ adenosine receptor in the mammalian spinal cord

A: The binding of adenosine to the spinal A₁ adenosine receptor coupled to a G_i protein inhibits D₁ dopamine receptor activity and of adenylyl cyclase lowering the levels of cAMP and thus decreasing the activity of PKA leading to a depression of locomotor activity. B: The binding of caffeine to the A₁ adenosine receptor leads to inhibition of adenosine signaling which no longer suppress the activity of the D₁ receptor leading to the activation of adenylyl cyclase through its coupling to a G_S protein increasing the levels of intracellular cAMP which augments PKA-dependent activity leading to the stimulatory effects of caffeine on locomotor activity.