Correction to: “The Affinity of D₂-Like Dopamine Receptor Antagonists Determines the Time to Maximal Effect on Cocaine Self-Administration.”

The senior authors of the above article [Norman AB, Tabet MR, Norman MK, Fey BK, Tsibulsky VL, and Millard RW (2011) J Pharmacol Exp Ther 338:724–728; doi: 10.1124/jpet.111.183244], Andrew B. Norman and Vladimir L. Tsibulsky, have recently identified data entry and data analysis errors.

In Fig. 2 the data points presented did not correspond to the original data files for the representative sessions shown. The data in the corrected Fig. 2 are from the original session files. These sessions did not have the same magnitudes of maximum response, as shown in the published figure, but the time to maximal effect (T_max) for each antagonist used is largely independent of the dose over the ranges of antagonist doses used in these studies. Therefore, in the corrected Fig. 2 given below, the ordinate axis is normalized to a percent of the maximum effect for each session defined by the best fit model shown by the line through each data set.

Fig. 2.

The time courses of antagonist-induced increases in the cocaine satiety threshold are different. Symbols represent the proportional changes in cocaine satiety threshold from the baseline value after the injection of antagonist. Time 0 is the time of antagonist injection. The doses and T_max values of molindone, eticlopride, nemonapride, and spiperone were 750, 20, 10, and 75 nmol/kg, and 6.1, 26.3, 61.1, and 163.5 min, respectively.

The data included in Fig. 3 were reanalyzed. The updated T_max values were plotted against the log D at pH 7 (retrieved from the CAS registry database) instead of the calculated log D at pH 7.4. The linear regression line has a correlation coefficient of 0.0237 and p = 0.693, indicating there was no correlation between the two sets of values as was stated in the original paper.

We were unable to verify which sessions were used to generate the mean ± SEM T_max values nor the values used to generate the mean ± SEM antagonist K_i or K_d values shown in Fig. 4. Consequently, the T_max values for the antagonists used for the correlation in Fig. 4 were reanalyzed from the original data, but using all sessions across a range of antagonist doses that generated a measurable maximum response. The number of rats used is shown in parentheses as is the total number of sessions used to generate the mean ± SEM T_max values for each antagonist and the range of doses used is also shown: spiperone (13, 49), 40 – 150 nmol/kg; nemonapride (5, 9), 10 – 15 nmol/kg; eticlopride (8, 13), 15 – 30 nmol/kg; haloperidol (8, 33), 40 – 300 nmol/kg; thioridazine (4, 8), 5,000 – 9,000 nmol/kg; triflupromazine (6, 16), 50-750 nmol/kg; olanzapine (12, 35), 200 – 2,000; aripiprazole (6, 50), 100 – 800 nmol/kg; molindone (4, 14), 400 – 1,500 nmol/kg.

Fig. 4.

The T_max of antagonist effect on cocaine self-administration correlates with antagonist affinity for D₂-like dopamine receptors. The symbols represent the mean ± SEM for both the K_i or K_d values and T_max values. The mean K_i or K_d values also include values from the BindingDB (http://www.bindingdb.org) database [Liu T, Lin Y, Wen X, Jorissen RN, and Gilson MK (2007) BindingDB: a web-accessible database of experimentally determined protein–ligand binding affinities. Nucleic Acids Res 35(Database issue): D198–D201] with high outlier values excluded. The linear regression line has a correlation coefficient r = 0.86 and is significant, p = 0.0026.

The authors apologize for these errors. The reanalyzed data are consistent with the conclusions of the original published paper. An investigation by the University of Cincinnati has determined that AB Norman, MK Norman, BK Fey, VL Tsibulsky, and RW Millard did not take part in and were unaware of any inappropriate data handling.

The authors acknowledge the efforts of Hanna N. Wetzel in the data reanalysis.