Gaussian Process-based prediction of memory performance and biomarker status in ageing and Alzheimer's disease-A systematic model evaluation.

Neuroimaging markers based on Magnetic Resonance Imaging (MRI) combined with various other measures (such as genetic covariates, biomarkers, vascular risk factors, neuropsychological tests etc.) might provide useful predictions of clinical outcomes during the progression towards Alzheimer's disease (AD). The use of multiple features in predictive frameworks for clinical outcomes has become increasingly prevalent in AD research. However, many studies do not focus on systematically and accurately evaluating combinations of multiple input features. Hence, the aim of the present work is to explore and assess optimal combinations of various features for MR-based prediction of (1) cognitive status and (2) biomarker positivity with a multi-kernel learning Gaussian process framework. The explored features and parameters included (A) combinations of brain tissues, modulation, smoothing, and image resolution; (B) incorporating demographics & clinical covariates; (C) the impact of the size of the training data set; (D) the influence of dimensionality reduction and the choice of kernel types. The approach was tested in a large German cohort including 959 subjects from the multicentric longitudinal study of cognitive impairment and dementia (DELCODE). Our evaluation suggests the best prediction of memory performance was obtained for a combination of neuroimaging markers, demographics, genetic information (ApoE4) and CSF biomarkers explaining 57% of outcome variance in out-of-sample predictions. The highest performance for Aß42/40 status classification was achieved for a combination of demographics, ApoE4, and a memory score while usage of structural MRI further improved the classification of individual patient's pTau status.

Epistatic interaction of genetic depression risk variants in the human subgenual cingulate cortex during memory encoding.

Recent genome-wide association studies have pointed to single-nucleotide polymorphisms (SNPs) in genes encoding the neuronal calcium channel CaV1.2 (CACNA1C; rs1006737) and the presynaptic active zone protein Piccolo (PCLO; rs2522833) as risk factors for affective disorders, particularly major depression. Previous neuroimaging studies of depression-related endophenotypes have highlighted the role of the subgenual cingulate cortex (CG25) in negative mood and depressive psychopathology. Here, we aimed to assess how recently associated PCLO and CACNA1C depression risk alleles jointly affect memory-related CG25 activity as an intermediate phenotype in clinically healthy humans. To investigate the combined effects of rs1006737 and rs2522833 on the CG25 response, we conducted three functional magnetic resonance imaging studies of episodic memory formation in three independent cohorts (N=79, 300, 113). An epistatic interaction of PCLO and CACNA1C risk alleles in CG25 during memory encoding was observed in all groups, with carriers of no risk allele and of both risk alleles showing higher CG25 activation during encoding when compared with carriers of only one risk allele. Moreover, PCLO risk allele carriers showed lower memory performance and reduced encoding-related hippocampal activation. In summary, our results point to region-specific epistatic effects of PCLO and CACNA1C risk variants in CG25, potentially related to episodic memory. Our data further suggest that genetic risk factors on the SNP level do not necessarily have additive effects but may show complex interactions. Such epistatic interactions might contribute to the 'missing heritability' of complex phenotypes.

Prefrontal dopamine and the dynamic control of human long-term memory.

Dopaminergic projections to the prefrontal cortex support higher-order cognitive functions, and are critically involved in many psychiatric disorders that involve memory deficits, including schizophrenia. The role of prefrontal dopamine in long-term memory, however, is still unclear. We used an imaging genetics approach to examine the hypothesis that dopamine availability in the prefrontal cortex selectively affects the ability to suppress interfering memories. Human participants were scanned via functional magnetic resonance imaging while practicing retrieval of previously studied target information in the face of interference from previously studied non-target information. This retrieval practice (RP) rendered the non-target information less retrievable on a later final test-a phenomenon known as retrieval-induced forgetting (RIF). In total, 54 participants were genotyped for the catechol-O-methyltransferase (COMT) Val(108/158)Met polymorphism. The COMT Val(108/158)Met genotype showed a selective and linear gene-dose effect on RIF, with the Met allele, which leads to higher prefrontal dopamine availability, being associated with greater RIF. Mirroring the behavioral pattern, the functional magnetic resonance imaging data revealed that Met allele carriers, compared with Val allele carriers, showed a greater response reduction in inhibitory control areas of the right inferior frontal cortex during RP, suggesting that they more efficiently reduced interference. These data support the hypothesis that the cortical dopaminergic system is centrally involved in the dynamic control of human long-term memory, supporting efficient remembering via the adaptive suppression of interfering memories.

The BDNF-Val66Met polymorphism: implications for susceptibility to multiple sclerosis and severity of disease.

Neurodegeneration following inflammatory injury is considered to be a pathological correlate of irreversible disability in patients with multiple sclerosis. The availability of neurotrophins could influence the probability or rate of disease progression and the time of onset. The BDNF-Val66Met-polymorphism leads to altered intracellular transport and secretion of BDNF, and is thus a logical candidate for a gene that influences susceptibility and, more specifically, the clinical course of multiple sclerosis. In order to test this hypothesis we genotyped the polymorphism in 951 UK multiple sclerosis trio families, but found no evidence for association before (p=0.63) or after stratification for clinical course (p=0.73).