Showing papers in "Reviews in The Neurosciences in 2003"

The human amygdala: an evolved system for relevance detection.

Abstract: Evidence from pioneering animal research has suggested that the amygdala is involved in the processing of aversive stimuli, particularly fear-related information. Fear is central in the evolution of the mammalian brain: it is automatically and rapidly elicited by potentially dangerous and deadly events. The view that the amygdala shares the main characteristics of modular systems, e.g. domain specificity, automaticity, and cognitive impenetrability, has become popular in neuroscience. Because of its computational properties, it has been proposed to implement a rapid-response 'fear module'. In this article, we review recent patient and neuroimaging data of the human brain and argue that the fundamental criteria for the amygdala to be a modular system are not met. We propose a different computational view and suggest the notion of a specific involvement of the human amygdala in the appraisal of relevant events that include, but are not restricted to, fear-related stimuli. Considering the amygdala as a 'relevance detector' would integrate the 'fear module' hypothesis with the concept of an evolved neural system devoted to the processing of a broader category of biologically relevant stimuli. In primates, socially relevant events appear to have become, through evolution, the dominant elements of the amygdala's domain of specificity.

Dopamine, acetylcholine and nitric oxide systems interact to induce corticostriatal synaptic plasticity

Abstract: Two distinct forms of synaptic plasticity have been described at corticostriatal synapses: long-term depression (LTD) and long-term potentiation (LTP). Both these enduring changes in the efficacy of excitatory neurotransmission in the striatum have a major impact on the physiological activity of the basal ganglia and are triggered by the stimulation of complex and independent cascades of intracellular second messenger systems. Along with the massive glutamatergic inputs originating from the cortex, striatal neurons receive a myriad of other synaptic contacts arising from different sources. In particular, while the nigrostriatal pathway provides this brain area with dopamine (DA), intrinsic circuits are the main source of acetylcholine (ACh) and nitric oxide (NO). The three neurotransmitter systems interact with each other to determine whether corticostriatal LTP or LTD is triggered in response to repetitive synaptic stimulation. Two distinct subtypes of striatal interneurons produce ACh and NO in the striatum. These interneurons are activated by the cortex during the induction phase of striatal plasticity, and stimulate, in turn, the intracellular changes in projection neurons required for LTD or LTP. Interneurons, therefore, exert a feedforward control of the excitability of striatal projection neurons by ensuring the coordinate expression of two alternative forms of synaptic plasticity at the same type of excitatory synapse. The integrative action exerted by striatal projection neurons on the converging information arising from the cortex, nigral DA neurons, and from ACh- and NO-producing interneurons dictates the final output of the striatum to the other structures of the basal ganglia.

Regulation of spine morphology and synaptic function by LIMK and the actin cytoskeleton.

Abstract: Filamentous actin (F-actin) is highly enriched in the dendritic spine, a specialized postsynaptic structure on which the great majority of the excitatory synapses are formed in the mammalian central nervous system (CNS). The protein kinases of the Lim-kinase (LIMK) family are potent regulators of actin dynamics in many cell types and they are abundantly expressed in the CNS, including the hippocampus. Using a combination of genetic manipulations and electrophysiological recordings in mice, we have demonstrated that LIMK-1 signaling is important in vivo in the regulation of the actin cytoskeleton, spine morphology, and synaptic function, including hippocampal long-term potentiation (LTP), a prominent form of long lasting synaptic plasticity thought to be critical to memory formation. Our results provide strong genetic evidence that LIMK and its substrate ADF/cofilin are involved in spine morphology and synaptic properties and are consistent with the notion that the Rho family small GTPases and the actin cytoskeleton are critical to spine structure and synaptic regulation.

Time, space and hippocampal functions.

Abstract: The hippocampus is one of the most researched structures of the brain. Studies of lesions in humans, primates and rodents have suggested to some that the primary role of the hippocampus is to act as a temporary memory buffer which is required for the consolidation of long-term memory. The famous case study of patient H.M., in particular, seemed to suggest that the hippocampus was of crucial importance for memory formation. However, recordings of single neurons in freely moving rodents did not support this notion. In such recordings, neurons were found that were active predominately when the animal passed through a particular area in space. Consequently, these neurons were termed 'place cells' and a theory was developed that suggested that the hippocampus acts as a 'cognitive map' that is required for spatial orientation. It was then found that H.M. had significant damage to his temporal lobes that included the amygdala, rhinal cortices, and other areas. Further case studies and selective hippocampal lesions in primates resulted in much milder amnestic symptoms, and lesions of defined cortical areas in the temporal lobes showed that a number of functions previously attributed to the hippocampus were in fact linked to these areas. Further analysis of neuronal activity in the hippocampus showed that not only is spatial information represented there, but also additional information, such as speed of movement, direction of movement, match or non-match detection, olfactorial identification, and others. In addition, it was found that selective lesions of the hippocampus in rodents impaired spatial navigation and memory formation only mildly. Only simultaneous lesions of several cortical areas in conjunction with the hippocamus could reproduce the impairments and symptoms that were previously thought to be observed after hippocampal lesions alone. In conclusion it is proposed that information processing and memory formation is shared by several brain areas that act as a functional system. This review presents evidence from many different studies that the hippocampus is part of this system and plays a supportive role in associating complex multimodal information and laying down new memory traces. In addition, the concept of allocating specific functions (such as the development of a cognitive map) exclusively to the hippocampus is rejected.

Training in cortical control of neuroprosthetic devices improves signal extraction from small neuronal ensembles.

Abstract: We have recently developed a closed-loop environment in which we can test the ability of primates to control the motion of a virtual device using ensembles of simultaneously recorded neurons /29/. Here we use a maximum likelihood method to assess the information about task performance contained in the neuronal ensemble. We trained two animals to control the motion of a computer cursor in three dimensions. Initially the animals controlled cursor motion using arm movements, but eventually they learned to drive the cursor directly from cortical activity. Using a population vector (PV) based upon the relation between cortical activity and arm motion, the animals were able to control the cursor directly from the brain in a closed-loop environment, but with difficulty. We added a supervised learning method that modified the parameters of the PV according to task performance (adaptive PV), and found that animals were able to exert much finer control over the cursor motion from brain signals. Here we describe a maximum likelihood method (ML) to assess the information about target contained in neuronal ensemble activity. Using this method, we compared the information about target contained in the ensemble during arm control, during brain control early in the adaptive PV, and during brain control after the adaptive PV had settled and the animal could drive the cursor reliably and with fine gradations. During the arm-control task, the ML was able to determine the target of the movement in as few as 10% of the trials, and as many as 75% of the trials, with an average of 65%. This average dropped when the animals used a population vector to control motion of the cursor. On average we could determine the target in around 35% of the trials. This low percentage was also reflected in poor control of the cursor, so that the animal was unable to reach the target in a large percentage of trials. Supervised adjustment of the population vector parameters produced new weighting coefficients and directional tuning parameters for many neurons. This produced a much better performance of the brain-controlled cursor motion. It was also reflected in the maximum likelihood measure of cell activity, producing the correct target based only on neuronal activity in over 80% of the trials on average. The changes in maximum likelihood estimates of target location based on ensemble firing show that an animal's ability to regulate the motion of a cortically controlled device is not crucially dependent on the experimenter's ability to estimate intention from neuronal activity.

The role of neurotransmitters in neurite outgrowth and synapse formation

Abstract: Besides a well-established role in neuronal communication in the adult central nervous system, neurotransmitters have diverse tasks in the embryonic brain, ranging from early developmental functions in morphogenesis /13/, to later functions in target selection and synapse formation /87/. For example, growth cones of developing neurons are known to release transmitters /26,36,88,110,115/ and respond to transmitters released from other neurons /35,44,59, 61,70/. Moreover, depletion of transmitters during embryonic development results in developmental deficits of the brain /21,48,84,109/, suggesting that transmitters have crucial roles as morphogens and/or neurotrophic factors. Although recently the idea of neurotransmitters being important for neural development has been challenged /99/, there is a vast amount of literature that seems to support the hypothesis that neurotransmitter release in the developing central nervous system is crucial for proper brain development. In this review we focus on the roles that neurotransmitters play in neurite outgrowth, target selection and synapse formation, with particular emphasis on the effects of the transmitters serotonin and dopamine.

Cytoarchitecture of the avian optic tectum: neuronal substrate for cellular computation.

Abstract: To analyse cellular computation in the vertebrate brain, a thorough knowledge of the underlying anatomy, physiology and connectivity of the neuronal substrate is essential. This review compiles data on one of the best known structures of the vertebrate brain, the optic tectum of birds. The functions of this structure are multifold, but can be attributed largely to orientation and the basic analysis of sensory data in a spatial context. In the tectum, a wealth of data on physiology and anatomy has been gathered over more than a century and provides an excellent background for computational studies. The analysis of the optic tectum is facilitated by several principles of organisation, including the retinotopic input and the highly laminated layout with separated input and output layers. Moreover, the molecular mechanisms guiding the development and connectivity have been analysed in detail. As the avian tectum and the mammalian superior colliculus are partly homologous, the cellular mechanisms unraveled in the tectum can also be transferred to the colliculus and thus contribute to the understanding of the vertebrate visual system in general.

On the role of somatostatin in seizure control: clues from the hippocampus

Abstract: The role of the hippocampal somatostatin (somatotropin release-inhibiting factor, SRIF) system in the control of partial complex seizures is discussed in this review The SRIF system plays a role in the inhibitory modulation of hippocampal circuitries under normal conditions: 1) SRIF neurons in the dentate gyrus are part of a negative feedback circuit modulating the firing rate of granule cells; 2) SRIF released in CA3 interacts both with presynaptic receptors located on associational/commissural terminals and with postsynaptic receptors located on pyramidal cell dendrites, reducing excitability of pyramidal neurons; 3) in CA1, SRIF exerts a feedback inhibition and reduces the excitatory drive on pyramidal neurons Significant changes in the hippocampal SRIF system have been documented in experimental models of temporal lobe epilepsy (TLE), in particular in the kindling and in the kainate models SRIF biosynthesis and release are increased in the kindled hippocampus, especially in the dentate gyrus This hyper-function may be instrumental to control the latent hyperexcitability of the kindled brain, preventing excessive discharge of the principal neurons and the occurrence of spontaneous seizures In contrast, the hippocampal SRIF system undergoes damage in the dentate gyrus following kainate-induced status epilepticus Although surviving SRIF neurons appear to hyperfunction, the loss of hilar SRIF interneurons may compromise inhibitory mechanisms in the dentate gyrus, facilitating the occurrence of spontaneous seizures In keeping with these data, pharmacological activation of SRIF1 (sst2) receptors, ie of the prominent receptor subtype on granule cells, exerts antiseizure effects Taken together, the data presented suggest that the hippocampal SRIF system plays a role in the control of partial complex seizures and, therefore, that it may be proposed as a therapeutic target for TLE

Spatial patterns of spontaneous local field activity in the monkey visual cortex.

Abstract: Spontaneous activity among visually responsive neurons is often considered to consist of random neural events, or perhaps to reflect an irrelevant by-product of brain homeostasis. However, recent studies have emphasized that such ongoing activity is strongly synchronized over large cortical distances, and can have a marked impact on the responsiveness of neurons to visual stimuli, suggesting that such activity may indeed be highly relevant to the brain's interpretation of its sensory input. In the current study, we examined the spatiotemporal nature of local field potential (LFP) fluctuations in the visual cortex of two macaque monkeys that were awake, but in a state of relaxation with minimal visual stimulation. Using an array of 16 electrodes spaced by several millimeters, we simultaneously monitored the LFP at many sites over a large region of the visual cortex. In agreement with the literature, we found that the coherence in the raw LFP signal fell off quickly with both frequency and distance. However, when we examined slower fluctuations in the LFP power, we found that power signals, including those derived from the high y-range frequencies, had high coherence that fell off only very slowly with cortical distance. Finally, we performed an additional experiment, with several electrodes placed on either side of a sulcus, to demonstrate that the decline in local field synchrony with cortical distance was so reliable that the interruption in the cortical sheet corresponding to the opening of the sulcus could be easily identified by monitoring just a few minutes of spontaneous LFP activity. These experiments reveal that a significant portion of spontaneous LFP fluctuations in the visual cortex is contributed by global mechanisms, imposing synchrony that is, first and foremost, a function of cortical separation between any two points.

Genetic programming by the proteolytic fragments of the amyloid precursor protein: somewhere between confusion and clarity.

Abstract: Mice engineered to overexpress disease-causing mutant amyloid precursor proteins (APP) display plaque deposition, but lack the hyperphosphorylated tau and massive neuronal loss characteristic of Alzheimer's disease (AD). Global gene expression profiles of brain regions from AD patients show upregulation of proapoptotic and inflammatory genes and down-regulation of neurotrophic, MAPK, phosphatase, and synaptic genes, while a profile of mice overexpressing a mutant APP shows the opposite trends in apoptotic and neurotrophic genes. The proteolytic fragments of the amyloid precursor protein have distinct biological actions. Both the gamma-secretase cleaved COOH-terminal fragment (CTFgamma) and the alpha-secretase cleaved NH2-terminal of APP (sAPPalpha) can regulate gene expression. While Abeta and CTFgamma can lead to toxicity and cell death, sAPPalpha promotes neurite outgrowth, enhances memory, and protects against a variety of insults, including Abeta toxicity. In AD, Abeta levels increase while sAPPalpha levels decrease. These subtleties in the levels of APP cleavage products are not reproduced in mice overexpressing mutant APP. In fact, the gene expression changes driven by sAPPalpha, such as increases in transthyretin and insulin-like growth factor 2, may protect these mice from high levels of Abeta.

Reactivation of memory: role of medial temporal lobe and prefrontal cortex.

Abstract: Memory retrieval is to bring the remembered information on-line or to reactivate the information. The critical determinant of memory retrieval mechanisms is whether the information has been maintained on-line or off-line, regardless of whether it is long-term memory or short-term, working memory. Similar reactivation processes occur during retrieval from long-term memory and from working memory when online maintenance has been interrupted. The reactivation is achieved by interactions between the posterior association areas, medial temporal lobe and prefrontal cortex. Posterior association areas maintain the representations of remembered information and are reactivated at retrieval. The medial temporal lobe is primarily involved in retrieval from off-line memory and triggers the reactivation by associating a whole set of features and episodes during encoding of the information. The prefrontal cortex is involved in retrieval from both on-line and off-line memory. It controls reactivation by setting up retrieval mode, starting retrieval attempt, and monitoring the contents of reactivated information. The prefrontal cortex also controls the selection of task-relevant information from information maintained on-line.

Input prediction and autonomous movement analysis in recurrent circuits of spiking neurons.

Abstract: Temporal integration of information and prediction of future sensory inputs are assumed to be important computational tasks of generic cortical microcircuits. It has remained open how cortical microcircuits could possibly achieve this, especially since they consist--in contrast to most neural network models--of neurons and synapses with heterogeneous dynamic responses. It turns out, however, that the diversity of computational units increases the capability of microcircuit models for temporal integration. Furthermore the prediction of future input may be rather easy for such circuits since it suffices to train the readouts from such microcircuits. In this article we show that very simple readouts from a generic recurrently connected circuit of integrate-and-fire neurons with diverse dynamic synapses can be trained in an unsupervised manner to predict movements of different objects, that move within an unlimited number of combinations of speed, angle, and offset over a simulated sensory field. The autonomously trained microcircuit model is also able to compute the direction of motion, which is a computationally difficult problem ('aperture problem') since it requires disambiguation of local sensory readings through the context of other sensory readings at the current and preceding moments. Furthermore the same circuit can be trained simultaneously in a supervised manner to also report the shape and velocity of the moving object. Finally it is shown that the trained neural circuit supports novelty detection and the generation of 'imagined movements'. Altogether the results of this article suggest that it is not necessary to construct specific and biologically unrealistic neural circuit models for specific sensory processing tasks, since 'found' generic cortical microcircuit models in combination with very simple perceptron-like readouts can easily be trained to solve such computational tasks.

Non-pharmacological interventions in cognitively impaired and demented patients--a comparison with cholinesterase inhibitors.

Abstract: The present paper reviews studies examining the effects of non-pharmacological stimulation, i.e. bright light, physical activity and tactile stimulation (touch), on cognition, affective behaviour, and the sleep-wake rhythm of impaired and demented elderly, both in a qualitative (narrative) and quantitative (meta-analytic) manner. An extensive search through eight bibliographic data bases (PubMed, Web of Science, ERIC, PsychINFO, Psyndex, Cinahl, Biological Abstracts and Rehabdata) was performed up to August 2002. The primary criterion for inclusion in this review was that studies provided sufficient data to calculate effect-sizes. In the qualitative analysis, all three types of stimulation appeared to improve cognitive functioning. Disturbances in behaviour react positively to bright light and tactile stimulation. Bright light was also beneficial to sleep. Tactile stimulation had, moreover, a beneficial influence on the patient-caretaker relationship. A comparison was made with several representative papers published since 1991 on the effects of acetylcholinesterase inhibitors on cognition and behaviour with representative papers on non-pharmacological stimulation interventions. Data indicated that improvements in cognition and affective behaviour by non-pharmacological interventions (d' = 0.32) and by cholinesterase inhibitors (d' = 0.31) were of similar effect-size. Possible mechanisms underlying the non-pharmacological stimulation effects are discussed and suggestions offered for future research.

Neurobehavioral characteristics of mice with modified intermediate filament genes.

Abstract: Intermediate proteins comprise cytoskeletal elements that preserve the shape and structure of neurons. These proteins have been proposed to be involved in the onset and progression of amyotrophic lateral sclerosis (ALS), mainly characterized by motoneuron atrophy and paresis. In support of this hypothesis are the findings that genetically modified mice for intermediate filaments successfully mimic certain neuropathological aspects of ALS, such as reduced axonal caliber and retarded conduction speed in peripheral nerves, although often without leading to paresis. Nevertheless, even in those models with no overt phenotype, the involvement of intermediate proteins in motor function is underlined by the deficits in tests of balance and equilibrium revealed in mice containing transgenes for neurofilament of heavy molecular weight (NFH), alpha-internexin, peripherin, and vimentin. In addition, spatial learning was impaired in transgenic mice expressing transgenes for NFH and NFM, similar to the memory deficits reported in patients with ALS.

Design for a brain revisited: the neuromorphic design and functionality of the interactive space 'Ada'.

Abstract: While much is now known about the operation and organisation of the brain at the neuronal and microcircuit level, we are still some way from understanding it as a complete system from the lowest to the highest levels of description. One way to gain such an integrative understanding of neural systems is to construct them. We have built the largest neuromorphic system yet known, an interactive space called 'Ada' that is able to interact with many people simultaneously using a wide variety of sensory and behavioural modalities. 'She' received 553,700 visitors over 5 months during the Swiss Expo.02 in 2002. In this paper we present the broad motivations, design and technologies behind Ada, and discuss the construction and analysis of the system.

Spike-driven synaptic plasticity for learning correlated patterns of mean firing rates.

Abstract: Long term synaptic changes induced by neural spike activity are believed to underlie learning and memory. Spike-driven long-term synaptic plasticity has been investigated in simplified situations in which the patterns of mean rates to be encoded were statistically independent. An additional regulatory mechanism is required to extend the learning capability to more complex and natural stimuli. This mechanism can be provided by those effects of the action potentials that are believed to be responsible for spike-timing dependent plasticity. These effects, when combined with the dependence of synaptic plasticity on the post-synaptic depolarization, produce the non-monotonic learning rule needed for storing correlated patterns of mean rates.

Neural Architectures for Robot Intelligence

Abstract: We argue that direct experimental approaches to elucidate the architecture of higher brains may benefit from insights gained from exploring the possibilities and limits of artificial control architectures for robot systems. We present some of our recent work that has been motivated by that view and that is centered around the study of various aspects of hand actions since these are intimately linked with many higher cognitive abilities. As examples, we report on the development of a modular system for the recognition of continuous hand postures based on neural nets, the use of vision and tactile sensing for guiding prehensile movements of a multifingered hand, and the recognition and use of hand gestures for robot teaching. Regarding the issue of learning, we propose to view real-world learning from the perspective of data-mining and to focus more strongly on the imitation of observed actions instead of purely reinforcement-based exploration. As a concrete example of such an effort we report on the status of an ongoing project in our laboratory in which a robot equipped with an attention system with a neurally inspired architecture is taught actions by using hand gestures in conjunction with speech commands. We point out some of the lessons learnt from this system, and discuss how systems of this kind can contribute to the study of issues at the junction between natural and artificial cognitive systems.

Analysis of evoked and emergent patterns of stimulus-related auditory cortical activity.

Abstract: Cortical activity contains both evoked patterns and emergent patterns of stimulus-related activity. Here we compared evoked and emergent patterns in the primary auditory cortex, field AI, of the gerbil by studying the differential effects of diluting spatial information about the patterns on their geometric dissimilarity by randomly removing channels from the recording data. This identified the sets of most relevant channels for the discrimination of stimuli in both types of patterns. In the evoked patterns the sets of most discriminative channels were found to be focally organized at locations corresponding to the thalamically relayed input into the cortical tonotopic map. In the emergent patterns the sets of most discriminative channels were broadly distributed and held no apparent relationship to the tonotopic map. The results indicate the coexistence in the same neuronal tissue of a topographic mapping principle for the evoked activity and a holographic mapping principle for the emergent activity.

Properties of a temporal population code.

Abstract: The temporal patterning of neuronal activity may play a substantial role in the representation of sensory stimuli. One particular hypothesis suggests that visual stimuli are represented by the temporal evolution of the instantaneous firing rate averaged over a whole population of neurons. Using an implementation in a cortical type network with lateral interactions, we could previously show that this scheme can be successfully applied to a pattern recognition task. Here, we use a large set of artificially generated stimuli to investigate the coding properties of the network in detail. The temporal population code generated by the network is intrinsically invariant to stimulus translations. We show that the encoding is invariant to small deformations of the stimuli and robust with respect to static and dynamic variations in synaptic strength of the lateral connections in the network. Furthermore, we present several measures which indicate that the encoding maps the stimuli into a high-dimensional space. These results show that a temporal population code is a promising approach for the encoding of relevant stimulus properties while simultaneously discarding the irrelevant information.

Learning distinct and complementary feature selectivities from natural colour videos.

Abstract: Many biological and artificial neural networks require the parallel extraction of multiple features, and meet this requirement with distinct populations of neurons that are selective to one property of the stimulus while being non-selective to another property. In this way, several populations can resolve a set of features independently of each other, and thus achieve a parallel mode of processing. This raises the question how an initially homogeneous population of neurons segregates into groups with distinct and complementary response properties. Using a colour image sequence recorded from a camera mounted on the head of a freely behaving cat, we train a network of neurons to achieve optimally stable responses, that is, responses that change minimally over time. This objective leads to the development of colour-selective neurons. Adding a second objective, decorrelating activity within the network, a subpopulation of neurons develops with achromatic response properties. Colour selective neurons tend to be non-oriented while achromatic neurons are orientation-tuned. The proposed objective thus successfully leads to the segregation of neurons into complementary populations that are either selective for colour or orientation.

Wiring the olfactory bulb--activity-dependent models of axonal targeting in the developing olfactory pathway.

Abstract: Two recent experimental studies /20,21/ revealed that odorant-evoked activity-dependent competition is significant in the organisation and maintenance of the olfactory system. In this paper, we investigate the generation of a chemotopic sensory map in the olfactory bulb through three models driven by high-density optical chemosensor arrays which have similar properties to olfactory receptor neurons. By exposing the sensor arrays to various odours, these models were subjected to Hebbian learning to achieve self-organisation, potentially explaining the activity-dependent competition demonstrated by these recent studies. Our final model also predicts a role for periglomerular cells in the formation of the chemotopic sensory map in the olfactory bulb.

Internal simulation of expected sensory experiences before movements get started.

Abstract: The functions of the cortical and subcortical motor areas have been regarded as the execution and control of limb movements We introduce recent neuroimaging evidence in humans that somatic sensation elicited by afferent inputs that signal limb movement also engage the motor areas It is generally understood that somatic sensation can only be experienced once movement has been executed Here we provide evidence that somatic sensation of limb movement can be internally simulated before the movement gets started by recruiting the motor areas It is suggested that the sensory experiences that are expected when movements are actually executed can be internally simulated in the network of the motor areas during motor imagery

Statistical mechanics beyond the Hopfield model: solvable problems in neural network theory.

Abstract: We present four 'case study' examples of solvable problems in the theory of recurrent neural networks, which are relevant to our understanding of information processing in the brain, but which are also interesting from a purely statistical mechanical point of view, even at the level of simple models (which helps in stimulating interdisciplinary work). The examples concern issues in network dynamics, network connectivity, spike timing and synaptic plasticity.

On the choice of a sparse prior

Abstract: An emerging paradigm analyses in what respect the properties of the nervous system reflect properties of natural scenes. It is hypothesized that neurons form sparse representations of natural stimuli: each neuron should respond strongly to some stimuli while being inactive upon presentation of most others. For a given network, sparse representations need fewest spikes, and thus the nervous system can consume the least energy. To obtain optimally sparse responses the receptive fields of simulated neurons are optimized. Algorithmically this is identical to searching for basis functions that allow coding for the stimuli with sparse coefficients. The problem is identical to maximizing the log likelihood of a generative model with prior knowledge of natural images. It is found that the resulting simulated neurons share most properties of simple cells found in primary visual cortex. Thus, forming optimally sparse representations is a very compact approach to describing simple cell properties. Many ways of defining sparse responses exist and it is widely believed that the particular choice of the sparse prior of the generative model does not significantly influence the estimated basis functions. Here we examine this assumption more closely. We include the constraint of unit variance of neuronal activity, used in most studies, into the objective functions. We then analyze learning on a database of natural (cat-cam) visual stimuli. We show that the effective objective functions are largely dominated by the constraint, and are therefore very similar. The resulting receptive fields show some similarities but also qualitative differences. Even for coefficient values for which the objective functions are dissimilar, the distributions of coefficients are similar and do not match the priors of the assumed generative model. In conclusion, the specific choice of the sparse prior is relevant, as is the choice of additional constraints, such as normalization of variance.