Ith exposure to effects of cocaine. Taken together with data that describe association between human CSMD1 variants and addiction-related phenotypes, these combined results support likely contributions of moderate individual differences in levels of CSMD1 expression to individual differences in levels of responses to rewarding and addictive drugs. Conditioned place preference tests combinations of drug reward and the memory processes that are triggered by prior rewarding drug-associated experiences. It is thus conceivable that the reduced CPP in homozygous knockouts could receive contributions from the observed reductions in performance on mnemonic testing. Others’ results in which csmd1 knockouts on mixed genetic backgrounds fail to display substantially altered memory in tests of novel object recognition or differences in sucrose preference fail to provide evidence for large confounding influences in these mice [21,27]. Alterations in cocaine-induced place preference in the heterozygous csmd1 knockout mice that express little difference in mnemonic testing suggest that we cannot ascribe the CPP findings in these mice to mnemonic influences of more moderatelyaltered levels of csmd1 expression, however.PLOS ONE | DOI:10.1371/ICG-001 cost journal.pone.0120908 July 14,11 /CSMD1 Variants and AddictionThe modest changes in Morris water maze performance in the csmd1 homozygous knockout mice with virtual elimination of CSMD1 expression are consistent with other data from mice and humans. These results accord with the prominent CSMD1 expression in memoryassociated brain regions, especially the hippocampus. They are consistent with a report of more frequent CSMD1 copy number variants in individuals with mild cognitive impairment or dementia, when compared to controls [25]. The overall data, and contrasting data for missense or other common variation, support the idea that variation in levels of CSMD1 expression attributable to 5′ ICG-001 solubility haplotypes in this gene provides a significant molecular source of common functional human individual difference at the CSMD1 locus. This positive set of translational data, some of which achieve only nominal statistical significance, contrasts with others’ inabilities to detect significant differences in prepulse inhibition in csmd1 knockout animals with mixed genetic backgrounds [21,27]. Taken together, these data support the possibility that human CSMD1 variation might play disproportionate roles in schizophrenia vulnerabilities due to influences on the cognitive difficulties that are commonly experienced by individuals with this diagnosis [34], though recent analyses also point to roles for CSMD1 variation in core schizophrenic psychotic features and in disease severity, providing a caution to accepting this simple explanation [16]. The intron 5 SNP rs10503223 and the intron 3 trinucleotide repeat annotated as rs71534387 are each associated with modest-to-moderate individual differences in levels of expression of CSMD1 mRNA that provide results at the margins of Bonferroni-corrected statistical significance. Variants in 5′ aspects of many genes are associated with their levels of expression [35]. The relatively large set of brains available here for these “level of expression” association studies, the relatively 5′ location of the genomic markers that display nominallysignificant association and the lack of common missense variants in the 5′ exons of this gene provide support for common level of expression variation in CSMD1. T.Ith exposure to effects of cocaine. Taken together with data that describe association between human CSMD1 variants and addiction-related phenotypes, these combined results support likely contributions of moderate individual differences in levels of CSMD1 expression to individual differences in levels of responses to rewarding and addictive drugs. Conditioned place preference tests combinations of drug reward and the memory processes that are triggered by prior rewarding drug-associated experiences. It is thus conceivable that the reduced CPP in homozygous knockouts could receive contributions from the observed reductions in performance on mnemonic testing. Others’ results in which csmd1 knockouts on mixed genetic backgrounds fail to display substantially altered memory in tests of novel object recognition or differences in sucrose preference fail to provide evidence for large confounding influences in these mice [21,27]. Alterations in cocaine-induced place preference in the heterozygous csmd1 knockout mice that express little difference in mnemonic testing suggest that we cannot ascribe the CPP findings in these mice to mnemonic influences of more moderatelyaltered levels of csmd1 expression, however.PLOS ONE | DOI:10.1371/journal.pone.0120908 July 14,11 /CSMD1 Variants and AddictionThe modest changes in Morris water maze performance in the csmd1 homozygous knockout mice with virtual elimination of CSMD1 expression are consistent with other data from mice and humans. These results accord with the prominent CSMD1 expression in memoryassociated brain regions, especially the hippocampus. They are consistent with a report of more frequent CSMD1 copy number variants in individuals with mild cognitive impairment or dementia, when compared to controls [25]. The overall data, and contrasting data for missense or other common variation, support the idea that variation in levels of CSMD1 expression attributable to 5′ haplotypes in this gene provides a significant molecular source of common functional human individual difference at the CSMD1 locus. This positive set of translational data, some of which achieve only nominal statistical significance, contrasts with others’ inabilities to detect significant differences in prepulse inhibition in csmd1 knockout animals with mixed genetic backgrounds [21,27]. Taken together, these data support the possibility that human CSMD1 variation might play disproportionate roles in schizophrenia vulnerabilities due to influences on the cognitive difficulties that are commonly experienced by individuals with this diagnosis [34], though recent analyses also point to roles for CSMD1 variation in core schizophrenic psychotic features and in disease severity, providing a caution to accepting this simple explanation [16]. The intron 5 SNP rs10503223 and the intron 3 trinucleotide repeat annotated as rs71534387 are each associated with modest-to-moderate individual differences in levels of expression of CSMD1 mRNA that provide results at the margins of Bonferroni-corrected statistical significance. Variants in 5′ aspects of many genes are associated with their levels of expression [35]. The relatively large set of brains available here for these “level of expression” association studies, the relatively 5′ location of the genomic markers that display nominallysignificant association and the lack of common missense variants in the 5′ exons of this gene provide support for common level of expression variation in CSMD1. T.