CETP polymorphisms associate with brain structure, atrophy rate, and Alzheimer's disease risk in an APOE-dependent manner

Two alleles in cholesteryl ester transfer protein (CETP) gene polymorphisms have been disputably linked to enhanced cognition and decreased risk of Alzheimer's disease (AD): the V and A alleles of I405V and C-629A. This study investigates whether these polymorphisms affect brain structure in 188 elderly controls and 318 AD or mild cognitive impairment (MCI) subjects from the Alzheimer's Disease Neuroimaging Initiative cohort. Nominally signficant associations were dependent on APOE ε4 carrier status. In APOE ε4 carriers, the V and A alleles, both of which decrease CETP and increase HDL, associated with greater baseline cortical thickness and less 12-month atrophy in the medial temporal lobe. Conversely, in APOE ε4 non-carriers, the I allele, which increases CETP and decreases HDL, associated with greater baseline thickness, less atrophy and lower risk of dementia. These results suggest CETP may contribute to the genetic variability of brain structure and dementia susceptibility in an APOE-dependent manner.

Assessing the performance of neural encoding models in the presence of noise

An analytical method is introduced for evaluating the performance of neural encoding models. The method addresses a critical question that arises during the course of the development and validation of encoding models: is a given model near optimal in terms of its accuracy in predicting the stimulus-elicited responses of a neural system, or can the predictive accuracy be improved significantly by further model development? The evaluation method is based on a derivation of the minimum mean-square error between actual responses and modeled responses. It is formulated as a comparison between the mean-square error of the candidate model and the theoretical minimum mean-square error attainable through an optimal model for the system. However, no a priori information about the nature of the optimal model is required. The theoretically minimum error is determined solely from the coherence function between pairs of system responses to repeated presentations of the same dynamic stimulus. Thus, the performance of the candidate model is judged against the performance of an optimal model rather than against that of an arbitrarily assumed model. Using this method. we evaluated a linear model for neural encoding by mechanosensory cells in the cricket cercal system. At low stimulus intensities, the best-fit linear model of encoding by single cells was found to be nearly optimal, even though the coherence between stimulus-response pairs (a commonly used measure of system linearity) was low. In this low-stimulus-intensity regime, the mean square error of the linear model was on the order of the power of the cell responses. In contrast, at higher stimulus intensities the linear model was not an accurate representation of neural encoding. even though the stimulus-response coherence was substantially higher than in the low-intensity regime.

Information theoretic analysis of dynamical encoding by four identified primary sensory interneurons in the cricket cercal system

1. The stimulus/response properties of four identified primary sensory interneurons in the cricket cercal sensory system were studied using electrophysiological techniques. These four cells are thought to represent a functionally discrete subunit of the cercal system: they are the only cells that encode information about stimulus direction to higher centers for low intensity stimuli. Previous studies characterized the quantity of information encoded by these cells about the direction of air currents in the horizontal plane. In the experiments reported here, we characterized the quantity and quality of information encoded in the cells' elicited responses about the dynamics of air current waveforms presented at their optimal stimulus directions. The total sample set included 22 cells. 2. This characterization was achieved by determining the cells' frequency sensitivities and encoding accuracy using the methods of stochastic systems analysis and information theory. The specific approach used for the analysis was the "stimulus reconstruction" technique in which a functional expansion was derived to transform the observed spike train responses into the optimal estimate (i.e., "reconstruction") of the actual stimulus. A novel derivation of the crucial equations is presented. The reverse approach is compared with the more traditional forward analysis, in which an expansion is derived that transforms the stimulus to a prediction of the spike train response. Important aspects of the application of these analytical approaches are considered. 3. All four interneurons were found to have identical frequency tuning, as assessed by the accuracy with which different frequency components of stimulus waveforms could be reconstructed with a linear expansion. The interneurons encoded significant information about stimulus frequencies between 5 and 80 Hz, which peak sensitivities at approximately 15 Hz. 4. All four interneurons were found to have identical stimulus/response latencies. The mean latency between a stimulus component and the corresponding elicited spike was 17 ms. All four interneurons also had identical integration times. The integration time, measured by the duration of stimulus, which could affect the probability of spiking, was approximately 50 ms. 5. The accuracy of the encoding can be expressed as a signal-to-noise ratio, where the noise is a scaled difference between the original signal and the best estimate of the signal. Peak signal-to-noise ratios of approximately 1 were obtained for the cells across all stimulus power levels, using only the linear expansion term. Analysis of the data indicated that the consideration of second-order nonlinear transformations of the stimulus would not have increased the calculated encoding accuracy. 6. The encoding accuracy also can be expressed in the information theoretic units of bits/second, which characterizes the information transmission rate of the cell. Bits/second values varied between 10 and 80 for the 22 different cells in our experimental set. The information rate values were highly correlated with the mean spike rates of the interneurons, but were not correlated with the stimulus power levels. However, normalizing the absolute information rates by the mean spike rate in each case yielded a measure of bits/spike that was remarkably invariant across all experiments. The measured bits/spike rate was approximately 1 for all experiments. This result is discussed in the context of recent theoretical studies on optimal encoding. 7. Although the dynamic sensitivities of the four interneurons were identical, their directional sensitivities are known to be orthogonal. Thus the cells are complementary to one another from a functional standpoint: whereas a particular cell will be insensitive to air currents from some directions, one or more of the other three cells will be sensitive to stimuli from those directions...

Information theoretic analysis of dynamical encoding by filiform mechanoreceptors in the cricket cercal system

1. The stimulus/response properties of 20 mechanosensory receptors in the cricket cercal sensory system were studied using electrophysiological techniques. These receptors innervated filiform hairs of various lengths and directional selectivities. Previous studies have characterized the sensitivity of such cells to the direction of air currents and to the amplitude of sinusoidal stimuli. In the experiments reported here, the quantity and quality of information encoded in the receptors' elicited responses about the dynamics of more complex air current waveforms were characterized. 2. Based on a white analysis of receptor response properties, the median frequency of each receptor's frequency tuning curve was found to be strongly correlated with the length of its associated mechanosensory hair. The receptors connected to hairs > 900 microns encoded frequencies below approximately 150 Hz very accurately and the receptors connected to shorter hairs encoded progressively higher bands of frequencies. These results were interpreted within the constraints imposed by the biomechanics of the air current-to-cercus boundary. 3. The encoding accuracy was expressed in the information theoretic units of bits/second, which characterizes the information transmission rate of a receptor. The information rates of the neuronal spike trains ranged from 75 to 220 bits/s. The information transmission rate was not correlated with the length of the mechanosensory hair. The average amount of information transmitted per action potential was negatively correlated with receptor hair length and ranged between 0.6 and 3.1 bits/spike. Decoding of the receptor responses was restricted to linear transformations of the spike trains. 4. The stimulus/response latencies of the different receptors ranged between 5 and 11 ms, and the integration time of the receptors ranged between 8 and 30 ms. The latency of a receptor was only weakly correlated with the length of its associated hair, and a receptor's integration time was correlated with hair length. 5. The stimulus/response phase difference for receptor cells that innervated hairs longer than approximately 800 microns increased with frequency > 50 Hz. The phase responses for receptor cells connected to hairs < 800 microns did not vary for frequencies > 50 Hz.