Electron microscopic studies of the dentate gyrus of the rat have revealed an apparent association between polyribosomes and dendritic spines. The present study was designed to elucidate the nature of this association. Our qualitative observations revealed that polyribosomes appeared primarily in two locations within the dendrite: (1) beneath the base of identified spines just subjacent to the intersection of the spine neck with the main dendritic shaft and (2) beneath mounds in the dendritic membrane which had the appearance of the base of a spine which extended out of the plane of section. To quantitatively define the nature of the apparent association, we attempted to determine (1) the proportion of spines with associated polyribosomes and (2) the proportion of the polyribosomes within dendrites which are associated with spine bases. Evaluation of profiles which were identifiable as spine neck-dendritic shaft intersections in a single section revealed that an average of 12.2% had associated polyribosomes. A serial section analysis revealed a somewhat higher incidence, however. Of a collection of 34 through-sectioned spines, 29% had polyribosomes which were revealed in one or more of the sections comprising the series. To evaluate what proportion of polyribosomes within the dendrite was associated with spines, we evaluated a series of photographs covering approximately 1250 micrometer2 of the dentate molecular layer from five animals, identifying all polyribosomes within dendrites and scoring their location as being (1) under spines, (2) under mounds, or (3) other. An average of 9.6% of the polyribosomes were found under processes identifiable as spine neck-dendritic shaft intersections, while 71.4% of the polyribosomes were found under mounds. Only 19% were not obviously associated with spines or mounds. Spine bases and mounds comprise only 3 of 35%, respectively, of the outline of dendritic profiles, however, indicating that the high incidence of polyribosomes under these elements cannot be accounted for by chance. To attempt to determine whether the mounds represent the base of dendritic spines, 68 mounds in 21 dentritic profiles were selected from the middle of the series of 20 serial sections. Ninetine of these mounds (28%) were continuous with an identified spine, and an additional 31% were continuous with thin processes of the size and appearance of spine necks. Thus, most of the mounds probably do represent the base of spines which extend out of the plane of a single section.

Randomness can be a useful component of computation. Using a computationally minimal, but still biologically based model of the hippocampus, we evaluate the effects of initial state randomization on learning a cognitive problem that requires this brain structure. Greater randomness of initial states leads to more robust performance in simulations of the cognitive task called transverse patterning, a context-dependent discrimination task that we code as a sequence prediction problem. At the conclusion of training, greater initial randomness during training trials also correlates with increased, repetitive firing of select individual neurons, previously named local context neurons. In essence, such repetitively firing neurons recognize subsequences, and previously their presence has been correlated with solving the transverse patterning problem. A more detailed analysis of the simulations across training trials reveals more about initial state randomization. The beneficial effects of initial state randomization derive from enhanced variation, across training trials, of the sequential states of a network. This greater variation is not uniformly present during training; it is largely restricted to the beginning of training and when novel sequences are introduced. Little such variation occurs after extensive or even moderate amounts of training. We explain why variation is high early in training, but not later. This automatic modulation of the initial-state-driven random variation through state space is reminiscent of simulated annealing where modulated randomization encourages a selectively broad search through state space. In contrast to an annealing schedule, the selective occurrence of such a random search here is an emergent property, and the critical randomization occurs during training rather than testing.

Darwinian evolution tends to produce energy-efficient outcomes. On the other hand, energy limits computation, be it neural and probabilistic or digital and logical. Taking a particular energy-efficient viewpoint, we define neural computation and make use of an energy-constrained, computational function. This function can be optimized over a variable that is proportional to the number of synapses per neuron. This function also implies a specific distinction between ATP-consuming processes, especially computation per se vs the communication processes including action potentials and transmitter release. Thus to apply this mathematical function requires an energy audit with a partitioning of energy consumption that differs from earlier work. The audit points out that, rather than the oft-quoted 20 watts of glucose available to the brain [1], the fraction partitioned to cortical computation is only 0.1 watts of ATP. On the other hand at 3.5 watts, long-distance communication costs are 35-fold greater. Other novel quantifications include (i) a finding that the biological vs ideal values of neural computational efficiency differ by a factor of 10^8 and (ii) two predictions of N, the number of synaptic transmissions needed to fire a neuron (2500 vs 2000).