184
Brain Research, 131 (1977) 184 190 .( Elsevier/North-Holland Biomedical Press
Memory impairment correlates closely with cycloheximide dose and degree of inhibition of protein synthesis
ELTON E. QUINTON and NEAL R. KRAMARCY Department of Psychology, Neuropsychopharmacology Program, University of Louisville, Louisvilh,, Ky. 40208 (U.S.A.)
(Accepted April 20th, 1977)
Over the past decade, many studies have reported that the administration of cycloheximide (CYC), a potent inhibitor of protein synthesis, shortly before or immediately after training on a learning task severely impairs development of long-term memory for that task 2. The correlation between CYC's effect on cerebral protein synthesis (CPS) and on memory has been widely accepted as a causal relationship, and thus the CYC data would seem to support the hypothesis that the development of long-term memory is dependent upon efficient CPS during the critical posttraining 'consolidation' period I . However, the assumed causal relationship between CYC's effect on protein synthesis and its effect on memory has been challenged by recent reports that CYC impairs catecholamine turnover a,~ and that CYC-induced amnesia can be prevented 1~ or reversed 4,9 by adrenergic agonists. It has been suggested that the memory impairment produced by CYC may be due to an impairment of catecholamine function and not specifically due to an impairment of CPS 9. This challenge necessitates a close evaluation of the strength of the relationship between the dose of CYC, the degree of inhibition of CPS at the time of training, and the degree of memory impairment. If inhibition of CPS is primarily responsible for the memory impairment, then the degree of inhibition of CPS and the degree of memory impairment should vary together as a function of the CYC dose. The existing literature on CYC provides very little information concerning the strength of the relationship between CYC dose, degree of memory impairment, and degree of inhibition of CPS. In the past, the primary concern has been to establish that inhibition of CPS at a critical time led to memory impairment. Therefore most researchers have used high doses of CYC to ensure maximal inhibition of CPS, and then manipulated various behavioral parameters. The behavioral studies which have administered more than one dose of CYC are difficult to interpret collectively because the dose ranges have been limited, different species and strains of animals have been used, different behavioral tasks have been used, and the necessary biochemical data have not been reported. Two previous experiments have attempted to study both memory impairment
185 and CPS using different doses of CYC, but these studies were concerned with other issues and therefore their experimental designs were not adequate to contribute data relative to the present issue. More specifically, these studies used only 2 (ref. 12) or 3 (ref. 1) doses of CYC for both memory and CPS determinations and did not relate the degree of memory impairment to the degree of inhibition of CPS during training or to CYC dose. In one report I no behavioral data were presented and the behavioral results were only casually discussed. Furthermore, the procedure for determining CPS in both studies was not suitable for estimating the degree of inhibition of CPS at the time of training. Thus, in one study ~, animals were given very brief training 30 min alter CYC, but CPS was determined 1 h or more after CYC, following a 30-min pulselabel. In the other study 12, CPS was determined 30 min after CYC following a 20-min pulse-label, but training occurred between 10 and 15 min after CYC. The experiments reported in this paper were initiated to specifically study the relationship between the dose of CYC, the degree of inhibition of CPS at the time of training, and the degree of memory impairment. This report is also concerned with determining whether the recovery of CPS following CYC closely parallels the recovery of memory processes. Our conclusions from these experiments are specific to our experimental paradigm. However, we chose an animal strain and behavioral task for these experiments that have been widely used in memory studies and have been shown to be sensitive to CYC. This permits a comparison of our results with a large body of existing literature. The animals used in this study were C57BL/6J male mice (Jackson Laboratories) 8-10 weeks old. The passive avoidance apparatus was very similar to the one described by Jarvik and Kopp~L It consisted of a two-chambered box, with the smaller chamber made from transparent plastic and the larger chamber made from black opaque plastic. A 40 W light bulb was positioned 45 cm above the transparent chamber. The walls and floor of the larger chamber were covered with two pairs of metal plates. The metal plates were connected to the constant current shock apparatus. Training was accomplished by placing the animal into the smaller chamber facing away from the opening leading into the larger chamber and allowing it to step spontaneously into the larger chamber. When the animal made contact with the rear set of plates in the larger chamber, it received a foot shock. (0.3 mA) which continued until the animal escaped back into the smaller chamber. It was then removed to its home cage to await testing. Test procedures were the same as training except that no shock was administered. Step-through latencies (STL) for training and test trials were recorded automatically and were defined as the time required for the animal to touch the rear plates in the larger chamber after being placed in the smaller chamber. On the test trial, if an animal did not enter the larger chamber within 300 sec, it was removed from the apparatus and assigned an STL of 300 sec. CYC (ICN) was dissolved in 0.9 o~ saline to a concentration of 10 mg/ml. The mice were injected subcutaneously in the anterior dorsal region. The doses of CYC administered were 150, 75, 30, 15, 7, or 3.5 mg/kg. All the above doses were administered to different groups of mice trained 0.5, 0.75, or I h after injection. Additional groups of animals were given 150 mg/kg 1.5 or 3 h before training, or 75 mg/kg 1.5 h
186 before training. Two control groups were given saline (0.3 ml) 0.5 or 1 h before training, and another control group received no injection before training: All groups were tested in the behavioral apparatus 72 h after training. Cerebral protein synthesis was monitored by determining the amount of radioactive precursor incorporated into protein during a 15-min pulse-label. Mice were injected with saline or one of the CYC doses (above) and sacrificed by decapitation 0.5, 0.75, 1, 1.5, 3, 4.5, or 6 h later. Three additional groups were given 150 mg/kg, 75 mg/kg, or saline and sacrificed 9 h later. Fifteen minutes before sacrifice all mice were injected intrapluraUy with 5 #Ci of L-[l-14C]valine (36 mCi/mmole, Schwartz/Mann). After decapitation each cerebrum (whole brain minus olfactory lobes, cerebellum, and lower brain stem) was quickly removed and homogenized in 3 ml 0.1 N NaOH (4 °C). A 1-ml aliquot of this homogenate was then subjected to trichloroacetic acid (TCA) precipitation and washing, as described previously10. The final pellet was solubolized in 2 ml 0.1 N NaOH and aliquots of this solution (TCA precipitate; protein) and the TCA supernatant ('free' valine) for each brain were added individually to scintillation vials and counted in a Beckman LSI00 scintillation counter to 1.0 ~ standard error. The CPMs were corrected for quench, averaged, and converted to DPM. Ten #1 aliquots of the protein-NaOH solution from each brain were used to determine the protein content of each brain s. Protein synthesis during the 15 min incorporation interval was thus expressed as DPM/mg protein/15 min. An analysis of the training trial STLs indicated that the pretraining injection of CYC had no significant effect on performance (F(1,.~75)~0.75, P ~:~ 0~25). The test results are summarized in Table I. When CYC was injected 30 min before training, all drug groups except the 3.5 mg/kg groups had shorter test STLs than controls, indicating impaired memory. A Kruskal-Wallis one-way analysis was applied to these data and was significant (/-/5 = 36.4, P < 0.001 ). This suggests that the degree of memory impairment was dose-dependent. This suggestion is further supported by the results of individual comparisons (Mann-Whitney U-test). The STLs of the 150 mg/kg group was significantly shorter than the STLs of all other groups except the 75 mg/kg group; the 75 mg/kg group had shorter STLs than the 15,7 and 3.5 mg/kg groups; the 30 mg/kg group had shorter STLs than the 7 and 3.5 mg/kg groups; and the 15 mg/kg group h a d shorter STLs than the 3.5 mg/kg group (for all comparisons, P • 0.05, one -tail). When CYC was administered 45 rain before training only the groups given 150 and 75 mg/kg had STLs significantly less than controls (Mann-Whitney U-test, P ~ 0.05). The STL of the 150 mg/kg groups was also significantly less than each of the other drug groups (for all comparisons P < 0.05), but the performance o f the 75 mg/kg group was not significantly different from the groups given lower doses. Administration of the drug 60 min before training only impaired the test performance of the group given 150 mg/kg. The test STL of this group was also significantly less than the STLs of all other drug groups (P < 0.01, all comparisons). Ninetyminute pretraining injections again affected the test performance o f only the 150 mg/kg group, while the 180-min pretraining injection had no significant effect on test performance. Test performance improved for the groups given 150 mg/kg as the injection-
187 TABLE 1 Median test trial latency (sec) as a fanction o f C YC dose and the injection-training interval
The numbers in parentheses are the number of mice in each group. The median STL of the pooled control groups was 300 (42). C YC dose
150 mg/kg 75 30 15 7 3.5
Injection training interval (rain) 30
45
60
90
12.5" (18) 18 * (18) 23.5* (18) 105 * (18) 200 ** (18) 300 (18)
40.5*(24) 152 *(24) 300 (24) 278 (22) 300 (18) 251 (24)
52.5*(24) 134' (24) 300 (30) 300 (24) 300 (18) 300 (18) 248 (18) 300 (24)
180
300 (18)
• P
Brain Research, 131 (1977) 184 190 .( Elsevier/North-Holland Biomedical Press
Memory impairment correlates closely with cycloheximide dose and degree of inhibition of protein synthesis
ELTON E. QUINTON and NEAL R. KRAMARCY Department of Psychology, Neuropsychopharmacology Program, University of Louisville, Louisvilh,, Ky. 40208 (U.S.A.)
(Accepted April 20th, 1977)
Over the past decade, many studies have reported that the administration of cycloheximide (CYC), a potent inhibitor of protein synthesis, shortly before or immediately after training on a learning task severely impairs development of long-term memory for that task 2. The correlation between CYC's effect on cerebral protein synthesis (CPS) and on memory has been widely accepted as a causal relationship, and thus the CYC data would seem to support the hypothesis that the development of long-term memory is dependent upon efficient CPS during the critical posttraining 'consolidation' period I . However, the assumed causal relationship between CYC's effect on protein synthesis and its effect on memory has been challenged by recent reports that CYC impairs catecholamine turnover a,~ and that CYC-induced amnesia can be prevented 1~ or reversed 4,9 by adrenergic agonists. It has been suggested that the memory impairment produced by CYC may be due to an impairment of catecholamine function and not specifically due to an impairment of CPS 9. This challenge necessitates a close evaluation of the strength of the relationship between the dose of CYC, the degree of inhibition of CPS at the time of training, and the degree of memory impairment. If inhibition of CPS is primarily responsible for the memory impairment, then the degree of inhibition of CPS and the degree of memory impairment should vary together as a function of the CYC dose. The existing literature on CYC provides very little information concerning the strength of the relationship between CYC dose, degree of memory impairment, and degree of inhibition of CPS. In the past, the primary concern has been to establish that inhibition of CPS at a critical time led to memory impairment. Therefore most researchers have used high doses of CYC to ensure maximal inhibition of CPS, and then manipulated various behavioral parameters. The behavioral studies which have administered more than one dose of CYC are difficult to interpret collectively because the dose ranges have been limited, different species and strains of animals have been used, different behavioral tasks have been used, and the necessary biochemical data have not been reported. Two previous experiments have attempted to study both memory impairment
185 and CPS using different doses of CYC, but these studies were concerned with other issues and therefore their experimental designs were not adequate to contribute data relative to the present issue. More specifically, these studies used only 2 (ref. 12) or 3 (ref. 1) doses of CYC for both memory and CPS determinations and did not relate the degree of memory impairment to the degree of inhibition of CPS during training or to CYC dose. In one report I no behavioral data were presented and the behavioral results were only casually discussed. Furthermore, the procedure for determining CPS in both studies was not suitable for estimating the degree of inhibition of CPS at the time of training. Thus, in one study ~, animals were given very brief training 30 min alter CYC, but CPS was determined 1 h or more after CYC, following a 30-min pulselabel. In the other study 12, CPS was determined 30 min after CYC following a 20-min pulse-label, but training occurred between 10 and 15 min after CYC. The experiments reported in this paper were initiated to specifically study the relationship between the dose of CYC, the degree of inhibition of CPS at the time of training, and the degree of memory impairment. This report is also concerned with determining whether the recovery of CPS following CYC closely parallels the recovery of memory processes. Our conclusions from these experiments are specific to our experimental paradigm. However, we chose an animal strain and behavioral task for these experiments that have been widely used in memory studies and have been shown to be sensitive to CYC. This permits a comparison of our results with a large body of existing literature. The animals used in this study were C57BL/6J male mice (Jackson Laboratories) 8-10 weeks old. The passive avoidance apparatus was very similar to the one described by Jarvik and Kopp~L It consisted of a two-chambered box, with the smaller chamber made from transparent plastic and the larger chamber made from black opaque plastic. A 40 W light bulb was positioned 45 cm above the transparent chamber. The walls and floor of the larger chamber were covered with two pairs of metal plates. The metal plates were connected to the constant current shock apparatus. Training was accomplished by placing the animal into the smaller chamber facing away from the opening leading into the larger chamber and allowing it to step spontaneously into the larger chamber. When the animal made contact with the rear set of plates in the larger chamber, it received a foot shock. (0.3 mA) which continued until the animal escaped back into the smaller chamber. It was then removed to its home cage to await testing. Test procedures were the same as training except that no shock was administered. Step-through latencies (STL) for training and test trials were recorded automatically and were defined as the time required for the animal to touch the rear plates in the larger chamber after being placed in the smaller chamber. On the test trial, if an animal did not enter the larger chamber within 300 sec, it was removed from the apparatus and assigned an STL of 300 sec. CYC (ICN) was dissolved in 0.9 o~ saline to a concentration of 10 mg/ml. The mice were injected subcutaneously in the anterior dorsal region. The doses of CYC administered were 150, 75, 30, 15, 7, or 3.5 mg/kg. All the above doses were administered to different groups of mice trained 0.5, 0.75, or I h after injection. Additional groups of animals were given 150 mg/kg 1.5 or 3 h before training, or 75 mg/kg 1.5 h
186 before training. Two control groups were given saline (0.3 ml) 0.5 or 1 h before training, and another control group received no injection before training: All groups were tested in the behavioral apparatus 72 h after training. Cerebral protein synthesis was monitored by determining the amount of radioactive precursor incorporated into protein during a 15-min pulse-label. Mice were injected with saline or one of the CYC doses (above) and sacrificed by decapitation 0.5, 0.75, 1, 1.5, 3, 4.5, or 6 h later. Three additional groups were given 150 mg/kg, 75 mg/kg, or saline and sacrificed 9 h later. Fifteen minutes before sacrifice all mice were injected intrapluraUy with 5 #Ci of L-[l-14C]valine (36 mCi/mmole, Schwartz/Mann). After decapitation each cerebrum (whole brain minus olfactory lobes, cerebellum, and lower brain stem) was quickly removed and homogenized in 3 ml 0.1 N NaOH (4 °C). A 1-ml aliquot of this homogenate was then subjected to trichloroacetic acid (TCA) precipitation and washing, as described previously10. The final pellet was solubolized in 2 ml 0.1 N NaOH and aliquots of this solution (TCA precipitate; protein) and the TCA supernatant ('free' valine) for each brain were added individually to scintillation vials and counted in a Beckman LSI00 scintillation counter to 1.0 ~ standard error. The CPMs were corrected for quench, averaged, and converted to DPM. Ten #1 aliquots of the protein-NaOH solution from each brain were used to determine the protein content of each brain s. Protein synthesis during the 15 min incorporation interval was thus expressed as DPM/mg protein/15 min. An analysis of the training trial STLs indicated that the pretraining injection of CYC had no significant effect on performance (F(1,.~75)~0.75, P ~:~ 0~25). The test results are summarized in Table I. When CYC was injected 30 min before training, all drug groups except the 3.5 mg/kg groups had shorter test STLs than controls, indicating impaired memory. A Kruskal-Wallis one-way analysis was applied to these data and was significant (/-/5 = 36.4, P < 0.001 ). This suggests that the degree of memory impairment was dose-dependent. This suggestion is further supported by the results of individual comparisons (Mann-Whitney U-test). The STLs of the 150 mg/kg group was significantly shorter than the STLs of all other groups except the 75 mg/kg group; the 75 mg/kg group had shorter STLs than the 15,7 and 3.5 mg/kg groups; the 30 mg/kg group had shorter STLs than the 7 and 3.5 mg/kg groups; and the 15 mg/kg group h a d shorter STLs than the 3.5 mg/kg group (for all comparisons, P • 0.05, one -tail). When CYC was administered 45 rain before training only the groups given 150 and 75 mg/kg had STLs significantly less than controls (Mann-Whitney U-test, P ~ 0.05). The STL of the 150 mg/kg groups was also significantly less than each of the other drug groups (for all comparisons P < 0.05), but the performance o f the 75 mg/kg group was not significantly different from the groups given lower doses. Administration of the drug 60 min before training only impaired the test performance of the group given 150 mg/kg. The test STL of this group was also significantly less than the STLs of all other drug groups (P < 0.01, all comparisons). Ninetyminute pretraining injections again affected the test performance o f only the 150 mg/kg group, while the 180-min pretraining injection had no significant effect on test performance. Test performance improved for the groups given 150 mg/kg as the injection-
187 TABLE 1 Median test trial latency (sec) as a fanction o f C YC dose and the injection-training interval
The numbers in parentheses are the number of mice in each group. The median STL of the pooled control groups was 300 (42). C YC dose
150 mg/kg 75 30 15 7 3.5
Injection training interval (rain) 30
45
60
90
12.5" (18) 18 * (18) 23.5* (18) 105 * (18) 200 ** (18) 300 (18)
40.5*(24) 152 *(24) 300 (24) 278 (22) 300 (18) 251 (24)
52.5*(24) 134' (24) 300 (30) 300 (24) 300 (18) 300 (18) 248 (18) 300 (24)
180
300 (18)
• P
