Common Neural Substrates for the Addictive Properties of Nicotine and Cocaine

See allHide authors and affiliations

Science  03 Jan 1997:
Vol. 275, Issue 5296, pp. 83-86
DOI: 10.1126/science.275.5296.83


Regional brain activation was assessed by mapping of Fos-related protein expression in rats trained to self-administration of intravenous nicotine and cocaine. Both drugs produced specific overlapping patterns of activation in the shell and the core of the nucleus accumbens, medial prefrontal cortex, and medial caudate areas, but not in the amygdala. Thus, the reinforcing properties of cocaine and nicotine map on selected structures of the terminal fields of the mesocorticolimbic dopamine system, supporting the idea that common substrates for these addictive drugs exist.

Nicotine is critical in the maintenance of tobacco smoking (1). Recent observations indicate that nicotine, like cocaine, activates the mesocorticolimbic dopamine (DA) system (2). This suggests similarities between the neuroactive properties of cocaine and nicotine but does not show whether the reinforcing properties of these two drugs involve similar neural substrates.

Experiments with animals that voluntarily press a lever to receive cocaine infusions strongly indicate that the mesocorticolimbic DA system is also a key neuroanatomical substrate for drug-seeking behavior itself (3). Because nicotine is intravenously self-administered in rats (4, 5), a study was designed to investigate whether the same set of neurons, a target of the mesocorticolimbic DA system, is activated by self-administration of nicotine and of cocaine. Overlaps in brain activation maps between cocaine and nicotine self-administration might identify a common substrate for cocaine and nicotine addiction.

Neuronal activation of the rat brain can be measured by mapping the expression of the immediate-early gene c-fos (6). Acute injection of cocaine and nicotine is known to produce transient increases of the expression of c-fos protein (Fos) and other Fos-related antigens (FRAs) in the nucleus accumbens and caudate region (7, 8). Newly synthetized Fos and FRAs heterodimerize with members of the Jun family to form the activating protein-1 (AP-1) complexes, which are important transcriptional regulators in neurons (69). Some FRAs, such as the 35-kD component, do not behave as immediate-early genes but their products, once induced, may last for several days (9).

Here, a computer-based detailed regional mapping of FRAs-like immunoreactive (FRAs-LI) profiles, as well as measurements of AP-1 binding, were performed in the brain of control rats and in rats killed after the last session of nicotine or cocaine self-administration (10). FRAs-LI was specifically used because the effects of cocaine on Fos-LI are known to be attenuated by repeated administration (7, 9).

Rats were first trained to press a lever for food, then, after surgical implantation of a catheter in the jugular vein, for intravenous nicotine or cocaine administration (11). The control animals were divided into two main groups: (i) untrained rats, which were exposed to the operant box daily and did not receive any training before or after surgery; and (ii) saline rats, which learned to press the lever for food and then received intravenous saline instead of nicotine or cocaine. An additional group of saline rats, acute nicotine rats, passively received the same amount of intravenous nicotine as rats that were actively self-administering nicotine, but only during the last session. In these animals, nicotine would act on brain structures that produce effects on Fos and FRAs expression that are unlikely to be related to the reinforcing properties of nicotine; indeed, aversive-like effects after acute nicotine or cocaine administration are known (12). Lever-press behavior was stably maintained for 2 weeks in rats that received nicotine or cocaine, whereas the performance of saline rats was significantly reduced (P < 0.01) and that of untrained rats was almost absent (Fig. 1).

Fig. 1.

(A) Average responses (active lever presses per hour) for each experimental group (n = 4 to 13 rats) during the last session in operant boxes after a 2-week training period (11). Untr., untrained control rats; Sal., Nic., and Coc. rats trained to lever press for food and then for intravenous saline, nicotine (0.03 mg/kg per infusion), or cocaine (0.25 mg/kg per infusion), respectively. Asterisk indicates P < 0.01 versus untrained control rats. (B) Self-administration records for one representative animal per group. Each vertical mark indicates the delivery of intravenous infusion after completion of two (cocaine) or three (nicotine and saline) lever responses.

Levels of FRAs-LI in the infralimbic cortex, shell and core of the nucleus accumbens, and medial caudate, but not the amygdala, were significantly increased in rats self-administering nicotine and cocaine when compared with levels in saline rats (P < 0.05; Fig. 2E). Nicotine, but not cocaine, strongly activated the superior colliculus and anterior cingulate cortex, whereas cocaine alone was effective in the lateral caudate (P < 0.05). Acute nicotine infused to saline rats during the last session produced significant effects in the superior colliculus (P < 0.05) and to a lesser extent in the anterior cingulate cortex, reaching the values observed in rats self-administering nicotine (Fig. 2E). Because acute passive administration of nicotine produces the same effects on FRAs-LI as seen here to be produced by self-administration, activation of the superior colliculus may not be related to nicotine-reinforcing properties. Clear-cut differences (P < 0.01) between all groups of rats trained to press the lever and those that were untrained were observed in all structures studied, except for the lateral caudate and superior colliculus. Thus, training for lever pressing produced a certain degree of activation in structures that were strongly activated by drug self-administration. No significant correlation was found between the behavioral performance of each treatment group and the FRAs-LI expression in all the brain structures under study.

Fig. 2.

(A through D) Open squares on the brain maps represent the anatomical areas chosen for quantitative image analysis of FRAs-LI profiles. The anteroposterior level of each section is defined as distance from the Bregma in stereotaxic coordinates (19): (A) 2.8 mm; (B) 1.70 mm; (C) −6.00 mm; and (D) −2.45 mm. IL, infralimbic cortex; CIG, anterior cingulate cortex, layers II and III; ADs, nucleus accumbens, dorsal shell; AVs, nucleus accumbens, ventral shell; ACo, nucleus accumbens, core; CeA, central nucleus of amygdala; BLA, basolateral nucleus of amygdala; StM, caudate-putamen, medial part; StL, caudate-putamen, laterodorsal part; and SuC, superficial gray of the superior colliculus. (E) Number of FRAs-LI profiles counted within each anatomical area represented in (A) through (D) of rats killed 90 min after the last self-administration session (10). Analysis of variance (ANOVA) indicated a significant group effect in all the brain regions (P < 0.01). Significant differences between groups are indicated as follows: *, P < 0.05 compared to saline; and +, P < 0.05 compared to acute nicotine. (F) FRA immunoblot of nucleus accumbens microdissected 60 min after the last self-administration session. Experimental groups were as defined as in the legend of Fig. 1. Antibody to FRAs was used at 25 μg/ml. (G and H) Representative photomicrographs of FRAs-LI cell nuclei (black-stained dots) in the ventral shell of the nucleus accumbens of one saline rat (G) and one nicotine self-administration rat (H). Scale bar, 20 μm.

Confirmation of these results came from measurements of binding of the AP-1 complex to nuclear DNA in microdissected brain tissue of rats killed 60 min after the last session (13). Both nicotine and cocaine self-administration rats showed significant increases of AP-1 binding in the medial prefrontal cortex and nucleus accumbens when compared with that in saline rats (Fig. 3). These changes were paralleled by increases of the 35-kD FRA bands as detected by immunoblot (Fig. 2F). Differential effects were found in the caudate region, where levels of the AP-1 complex were increased by cocaine but not by nicotine (Fig. 3D) (14).

Fig. 3.

(A) Representative autoradiogram of a bandshift assay of AP-1 binding in prefrontal cortex extracts of rats killed after the end of the last self-administration session. The lower arrow (a) indicates nonspecific binding; SS indicates the supershifted AP-1 complex band. Specificity experiments were performed in the presence of an excess of AP-1 unlabeled oligonucleotide (0.5 pmol; Nic. + CAP-1) or with the use of a labeled mutant oligonucleotide (19). Supershift experiments were performed by addition of 0.2 ml of antibody to FRAs (Nic. + FRA Ab) (9). Gel bands were directly scanned for radioactivity levels with an Ambis scanner. (B through D) Effects of nicotine and cocaine self-administration on AP-1 levels in various brain regions of rats (n = 4 to 7) killed immediately before (pre-session) or 60 min after (post-session) the last self-administration session. ANOVA indicated a group effect (P < 0.05) in the post-session measurements in all three brain structures but indicated a significant pre-session effect only in the nucleus accumbens and caudate-putamen (P < 0.05). Significant differences between groups are indicated as follows: *, P < 0.05 versus saline rats; +, P < 0.05 versus untrained rats; and #, P < 0.05 versus pre-session values.

AP-1 complex binding was also measured in animals killed immediately before the last self-administration session (Fig. 3). These animals were not exposed to the last dose of nicotine or cocaine, and the detection of significant levels of the AP-1 complex indicated the effects of previous repeated exposures to drugs or self-administration procedures. High levels of the AP-1 complex were observed in the nucleus accumbens of cocaine and, to a lesser extent, nicotine self-administration rats (P < 0.05 versus saline rats; Fig. 3C). No difference was measured in the prefrontal cortex region (Fig. 3B). In the caudate region, AP-1 binding was increased in all rats except the untrained ones (P < 0.05; Fig. 3D), which indicates a specific long-term effect of lever-press training.

Overall, the data on AP-1 complex binding suggest that each daily session of nicotine or cocaine self-administration produced a fractional increase of FRAs-LI that accumulated particularly in neurons of the nucleus accumbens but not the prefrontal cortex, forming persistent AP-1 complexes. Thus, the transcriptional regulatory effects of persistent AP-1 complexes may be implicated in the long-term adaptive changes associated with the maintenance of nicotine and cocaine self-administration (3, 7, 9).

The pharmacological effects of cocaine on FRAs-LI expression have been described (7, 9); those of nicotine can be related to the increased intracellular calcium levels produced by activation of nicotine receptors located on the glutamate, acetylcholine, or DA terminals projecting to the target areas (2, 15).

In humans, nicotine and cocaine are compulsively self-administered by cigarette smokers and cocaine users (16). In the rat model, compulsive self-administration of nicotine and cocaine was associated with overlapping activation maps in the shell and core of the nucleus accumbens, medial prefrontal cortex, and medial caudate-putamen, but not the amygdala, corroborating the view that there is a common neuronal substrate for addiction to these two drugs.


  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
  6. 6.
  7. 7.
  8. 8.
  9. 9.
  10. 10.
  11. 11.
  12. 12.
  13. 13.
  14. 14.
  15. 15.
  16. 16.
  17. 17.
  18. 18.
  19. 19.
  20. 20.
View Abstract

Navigate This Article