Oxygen uptake by human erythrocytes was examined both experimentally and theoretically in terms of the influence of unstirred solvent layers that are adjacent to the cell surface. A three dimensional cylindrical model was compared to less complex spherical and cylindrical and coordinate schemes. Although simpler and faster, the plane sheet and spherical algorithms are inadequate representations when unstirred solvent layers are considered. The cylindrical disk model most closely represents the physical geometry of red cells and is required for a quantitative analysis. In our stopped-flow, rapid mixing experiments the thickness of the unstirred layer expands with time as the residual turbulence decays. This phenomenon was quantitated using a formulation that is based on previously developed hydrodynamic theories. An initial 10('-4) cm unstirred layer is postulated to occur during mixing and to expand rapidly with time by a (t)('0.5) function when flow stops. Oxygen release time courses for human erythrocytes were also examined and analyzed by this three dimensional model. At high concentrations of external sodium dithionite, the effects of unstirred layers were less important but still necessary for a quantitative understanding of the reaction. In this case, the kinetics were influenced more by chemical reaction parameters, both for the extracellular O(,2)-dithionite reaction and for the intracellular O(,2)-hemoglobin reaction, and rigorous theories of these processes were required for fits to the observed data. A cooperative equilibrium scheme was developed which includes corrections for the Bohr effect in the physiological range of extracellular pH (pH 7.0-7.8). The relative importance of extracellular diffusion processes and chemical reaction parameters was examined in terms of red blood cell morphology and physiology. Effects on cell size were expressed primarily in oxygenation kinetics. Increasing internal heme concentration caused a proportional decrease in both uptake and release rates. Factors that alter the O(,2)-hemoglobin equilibrium influenced the rates of O(,2) release more than the rates of O(,2) uptake. For all of these experiments, the three dimensional cylinder scheme described quantitatively both the rates and overall shapes of the time courses.

Modern measurements of membrane, ionic channel molecules consist of whole cell current relaxations obtained using various voltage pulse protocols, current noise power spectra from small numbers of channels, and single channel intervals in the form of waiting, closed, and open time histograms. A state-variable description was developed for predicting these measurements from Markovian state models, and it was used to analyze the parameter identifiability properties of the models and measurements. Computer programs were developed for estimating model parameters; these included maximum likelihood estimation for the single channel data. In the case of the Ca channel, the parameters from a variety of models were found to be faster when obtained from whole cell rather than single channel measurements. The latter measurements were low pass filtered at 1 kHz in order to reliably detect the (TURN) 1/2 picoamp openings; the errors that resulted were elucidated by running Monte Carlo simulated channels through the threshold detection algorithm. Intervals were distorted by the absence of brief open and closed times as well as falsely prolonged open and closed times. An analytical technique was developed for computing the resulting distorted histograms, and after correction there was reasonable agreement between the measurements. Other errors found using single channel simulations included the appearance of a bimodal amplitude distribution from filtered channels having a single amplitude and the effects of noise and finite record length recording. Of six models considered for describing activation, a 4-state model best fit the whole cell relaxations resulting from simple, short pulse protocols, and it also predicted results from 2-pulse protocols and temperature experiments. Some inactivation models were eliminated because they were unable to simulate recover from inactivation and single channel failure traces. A family of models was suggested that describes the present experimental results, and possible tests for future model discrimination were evaluated using simulation. Two relatively insensitive tests were the 2-pulse test for "coupled" inactivation and the single channel Hi-Lo sort method for determining Ca-accumulation-dependent processes. It is concluded that the combination of parameter estimation and simulation is a useful tool for interpreting results and helping to plan experiments and analyses.

X-ray absorption fine structures of photolyzed carboxymyoglobin were measured at 4 K and at 40 K. The latter was a 10-s-resolved measurement following photolysis. The goal of the project was to elucidate the dynamic pathways through which ligand binding couples to protein conformation changes. The project involved five major tasks: (1) We constructed an integration spectrometer for x-ray absorption capable of 100 $\mu$s-resolved measurement. (2) Heme protein samples have to be photolyzable on the one hand and possess a high (Fe) fluorescence yield on the other. This is very difficult to achieve with solution samples. The problem was solved by embedding concentrated $(\sim 10$ mM) myoglobin in thin $(\sim 20\ \mu$m) dry polyvinyl alcohol (PVA) films. Carboxymyoglobin embedded in PVA film, has the same properties as in frozen buffer solution as proven by optical absorption, kinetics of CO recombination following photolysis and x-ray absorption spectra. (3) A cryostat was constructed for x-ray absorption measurement, which was also equipped with optical windows for flash photolysis and optical monitoring of the degree of photodissociation in the sample. (4) Synchrotron radiation experiments were performed at Stanford Synchrotron Radiation Laboratory and at National Synchrotron Light Source at Brookhaven. (5) Data analyses showed that the photolyzed carboxymyoglobin ${\rm (MB\sp\ast)}$ is structurally different from either carboxymyoglobin (MbCO) or deoxymyoglobin (Mb): the Fe-C(O) distance is 1.9 A, 2.17 A, $\infty$ A for MbCO, ${\rm Mb\sp\ast}$ and Mb respectively; the corresponding Fe-${\rm N}\sb{\rm imidazole}$ distances are 2.2 A, 2.18 A, 2.1 A and Fe-${\rm N}\sb{\rm porphyrin}$ distances are 1.97 A, 1.98 A, 2.06 A. At 40 K, from 1 s to 10 s following photolysis, 40% of photodissociated CO has recombined with Fe, and the rest of the sample is indistinguishable from ${\rm Mb\sp\ast}.$ The structure of ${\rm Mb\sp\ast}$ described above is in disagreement with a previous report by Powers et al. (Biochemistry 23, 5519 (1984)).

Gramicidin is the best characterized transmembrane ion channel. An opulence of biochemical and physiological data exists and it is the only channel for which a relatively well-defined structure is known. Thus, it has been proven to be an ideal model for actual physiological channels by providing valuable insight into a channel's structural/functional properties. Although much is known about the gramicidin channel, there is still no solid experimental data on the channel molecular dimensions. Therefore, this thesis supplies a very crucial piece of experimental structural data; namely, the direct location of two symmetric cation binding sites within the channel by x-ray diffraction. The binding sites are located at 9.4 $\pm$ 0.4 A about the midpoint of the channel. A previous $\sp{13}$C NMR study (Urry et al., 1982a,b; Urry, 1983) pinpointed cation binding between Trp-11 and Trp-13, and, based on rigid molecular models, this puts the binding sites at 11.0-11.5 A, which is at least 1.2 A larger than the value reported here. This suggests a considerable plasticity of the gramicidin channel consistent with conclusions drawn from recent molecular dynamic simulations (Roux and Karplus, 1988). Also, the smaller value leads to an elegant qualitative picture in which the flexible ends of the channel can conveniently provide the gating mechanism that transmits cations and blocks anions.

A quantitative understanding of the stability and expression of hemoglobins and myoglobins must take into account all of the factors which govern heme affinity and the folding of the globin portions of the molecules. In order to evaluate these factors, a new assay for measuring the rate of hemin loss was developed, and fluorescence techniques were adapted to monitor apoglobin unfolding. Native apomyoglobin has roughly the same size and shape as the holoprotein, including a hydrophobic crevice for heme binding, but its overall structure is less helical and more solvated. Addition of small amounts of denaturant to apomyoglobin produces an intermediate which lacks a well-defined heme pocket but still contains a hydrophobic core made up of three $\alpha$ helices. Further addition of denaturants produces a completely unfolded state. This three state unfolding reaction has traditionally been monitored by changes in circular dichroism but is more easily visualized by changes in tryptophan fluorescence. The heme pocket in native apomyoglobin is stabilized by the presence of hydrophobic amino acids which exclude solvent from the binding site. Introduction of polar amino acids into this region greatly destabilizes native apomyoglobin causing formation of the molten globular intermediate due to solvation of the heme pocket. The binding of hemin to apomyoglobin is very favorable and results in the release of $\sim{-}$77 kJ/mol of Gibbs' free energy under physiological conditions. The association rate constant is $\sim1 \times 10\sp8$ M$\sp{{-}1}$s$\sp{{-}1}$ and little affected by the structure of the apoglobin or reaction conditions. As a result, the equilibrium constant for hemin binding is determined primarily by its dissociation rate constant which is markedly dependent on apoprotein structure and ranges from $$100 h$\sp{{-}1}$ at pH 7, 37$\sp\circ$C. The principle factors determining hemin affinity are: (1) non-specific hydrophobic interactions, which account for $\sim{-}$36 kJ/mol of the free energy change produced by the formation of holomyoglobin from hemin and apoglobin; (2) specific van der Waals and electrostatic interactions between the porphyrin and heme-pocket amino acid residues, which account for another $-$18 kJ/mol; and (3) the covalent bond between His$\sp{93}$(F8) and the iron atom, which accounts for the remaining $\sim{-}$23 kJ/mol.

All globins consist of eight $\alpha$ helices labelled A through H and connected by loop regions, except mammalian $\alpha$ globins which lack a D helix. Myoglobin is used as a simple model for the study of the hemoglobin subunits to understand the stereochemical mechanisms that regulate function and stability in the globins. The roles of Phe43 (CD1), the D helix (residues 51-55), and Met55 (D5) in modulating the function and stability of sperm whale myoglobin were investigated by site directed mutagenesis. In myoglobin, Phe43$\to$Val, Leu, Ile, and Trp substitutions were constructed; the D helix was deleted and replaced with Ala$\sp{51-55}$ and Ala$\sp{51-54}$Met$\sp{55}$; and Met55(D5)$\to$Arg, Gly, Ala, Leu, Phe, and Trp substitutions were constructed. The role of the D helix (50-54) in $\beta$ subunits of human hemoglobin and the evolutionary significance of its absence from chordate human hemoglobin $\alpha$ subunits were examined by deleting residues 50-55 from $\beta$ subunits, called $\beta$ ($-$D), and inserting those same residues into $\alpha$ subunits, called $\alpha$ (+D). Finally, $\alpha$ His45 (CE3)$\to$Arg and $\beta$ Ser44 (CD3)$\to$Arg, Lys, Arg substitutions were constructed to increase hemin retention in recombinant human hemoglobin subunits. The size, hydrophobicity, and aromaticity of Phe at CD1 are required to regulate ligand binding, inhibit autooxidation, and retain hemin. The D helix appears to stabilize the CD corner and inhibit rapid hemin loss. A large apolar residue at D5 immobilizes the D helix, indirectly slowing hemin loss. Removal of the D helix from $\beta$ subunits also destabilizes bound hemin in both hemoglobin subunits. The absence of a D helix in chordate $\alpha$ subunits appears to be compensated by contacts with the adjacent $\beta$ subunits across the $\alpha$1$\beta$1 interface. This interface appears to stabilize the CE corner to inhibit hemin loss. A large apolar residue is required at positions CD1 and D5 to stabilize the heme binding pocket in apomyoglobin. Arg45 (CD3) stabilizes bound hemin in myoglobin by forming a salt bridge with the heme-6-propionate. His44 (CE3) in $\alpha$ subunits forms a similar salt bridge, whereas Ser45 (CD3) does not in $\beta$ subunits. When Ser44 (CD3) is replaced with His a favorable interaction with the heme-6 propionate appears to occur, stabilizing the $\beta$ globin-heme complex. Since this mutation has little effect on ligand binding and autooxidation, it appears to be a good secondary mutation for stabilizing a cell-free Hb-based blood substitute.

Semiconductor nanocrystals, also known as quantum dots (QDs), are evoking remarkable technological and scientific interest due to their fascinating size-dependent electronic and optical properties. Many biophysical studies to date have used effectively as brighter and more photostable replacement of organic dyes. In this thesis we focus on the most prominent feature of their photophysical properties, a random switching between emitting and non-emitting states, also known as fluorescence intermittency, or "blinking," in order to better understand the mechanism of quantum dot emission and how it reflects interaction with their immediate environment. We first designed and built a total internal reflection florescent microscope (TIRFM) with single molecule detection capabilities and determined optimal conditions for single QD studies. We then used image correlation techniques to show that the change in blinking dynamics could be detected and that it could complicate interpretation of the commonly used analytical techniques that rely on intensity fluctuations as reporters of particle mobility. In particular, we demonstrated that the transport coefficients recovered from fluorescence fluctuation analysis of diffusional mobility using temporal image correlation spectroscopy (TICS) had significant systematic errors due to blinking of the nanoparticles. We then performed a thorough, systematic study of the effects of protons on QDs' photochemical stability by varying the pH of their aqueous environment and related the single particle properties to properties of an ensemble of QDs. As pH was lowered, shorter "on" times and longer "off" times were observed, brightness of single QDs decreased and the number of permanently non-emitting QDs (dark fraction) increased. Based on these results, we proposed a coupled role for H+ ions by which they first reduced the intensity of the emitting state as well as affected probabilities of the QD to switch between the "on" and "off" state / Les nanoparticules semi-conductrices, aussi connues sous le nom de particules quantiques (PQ), suscitent un grand intérêt dans les domaines technologiques et scientifiques en raison de leurs propriétés spectroscopiques uniques et avantageuses. À ce jour, plusieurs études de biophysique ont remplacé succès les sondes fluorescentes organiques traditionnelles par des PQ. / Cette thèse porte principalement sur l'étude de la propriété la plus photophysique des PQ, soit leur capacité à activer ou à inactiver de manière aléatoire l'émission fluorescente (plus connue sous le nom de l'intermittence de l'émission fluorescente, ou « clignotement »), dans le but de mieux comprendre le mécanisme entourant l'émission des particules quantiques et son interaction avec son environnement immédiat. / Nous avons tout d'abord conçu et construit un microscope de fluorescence par réflexion totale interne (TIRFM) qui a la capacité de détecter une seule molécule et de déterminer les conditions idéales afin d'étudier une PQ unique. Nous avons ensuite utilisé des techniques de spectroscopie de corrélation temporelle dans les images afin de montrer que l'intermittence de l'émission fluorescente pouvait être détectée et qu'elle pourrait rendre possible l'interprétation d'analyses plus complexes que les techniques d'analyse traditionnelles qui se fondent sur les clignotements fluorescents comme preuve du mouvement des particules. / De façon plus précise, nous avons démontré que les coefficients de transport obtenus d'analyses utilisant la technique de spectroscopie de corrélation temporelle dans les images (TICS) présentaient une marge d'erreur significative due au « clignotement » des nanoparticules. Ensuite, nous avons procédé à une étude systématique plus approfondie des effets des protons sur la stabilité photochimique des PQ, en modifiant le pH de leur environnement aqueux et en faisant des liens avec les propriétés d'une seule particule dans un ensemble de PQ. Au fur et à mesure que le pH diminuait, nous avons observé une diminution du temps d'émission fluorescente, « ouvert », et une augmentation du temps d'absence d'émission, « éteint ». Nous avons également noté une diminution dans la clarté d'une seule PQ et une augmentation dans le nombre de PQ n'émettant aucune clarté (fraction sombre). / À la lumière de ces résultats, nous avons avancé l'idée d'un rôle double des ions H+, qui réduiraient d'abord l'intensité de l'état d'émission et affecteraient ensuite la probabilité d'une PQ d'alterner entre les états d'émission et d'absence d'émission, pour éventuellement emprisonner la PQ dans un état de fraction sombre permanent. Nous avons analysé et élargi les modèles théoriques concernant le « clignotement » afin de tenir compte des effets des ions H+ et de mieux démontrer que le « clignotement » et la formation de fraction sombre se fondent sur un seul et même mécanisme. / Les résultats présentés dans cette thèse sont particulièrement importants pour l'élaboration d'un modèle théorique complet du « clignotement » d'une PQ, mais peuvent également être importants pour les différents usages des PQ, en particulier dans les applications d'imagerie dans le domaine de la biologie quantique, où on retrouve des variations de pH entre les espaces cytoplasmiques et extracellulaires et dans les différents organes cellulaires.

Measuring protein interactions in subcellular compartments is key to understanding cell signalling mechanisms, but quantitative analysis of these interactions in situ has remained a major challenge. This thesis presents a novel analysis technique, spatial intensity distribution analysis (SpIDA), which may be applied to images obtained using fluorescence microscopy. SpIDA measures fluorescent particle densities and oligomerization states within individual images. The method is based on fitting intensity histograms from single images with super-Poissonian distributions to obtain density maps of fluorescent molecules and their quantal brightness. Since distributions are acquired spatially rather than temporally, this analysis may be applied to both live and chemically fixed cells and tissue. The technique does not rely on spatial correlations, freeing it from biases due to subcellular compartmentalization and heterogeneity within tissue samples. First, we validated the analysis technique evaluating its limits and demonstrating how it can be used to obtain useful information from complex biological samples. Analysis of simulations and heterodimeric GABAB receptors in spinal cord samples shows that the approach yields accurate estimates over a broad range of densities. SpIDA is applicable to sampling within subcell areas and reveals the presence of monomers and multimers with single dye labeling. We show that the substance P receptor (NK-1r) almost exclusively forms homodimers on the membrane and is primarily monomeric in the cytoplasm of dorsal horn neurons. Triggering receptor internalization caused a measurable decrease in homodimer density on the membrane surface. Finally, using GFP-tagged receptor subunits, we show that SpIDA can resolve dynamic changes in receptor oligomerization in live cells and is applicable to detection of high order oligomerization states. We then compared SpIDA results with those obtained from fluorescence lifetime imaging, and used it to extract information on receptor tyrosine kinase (RTK) dimerization at the cell membrane in response to GPCR activation. We show that RTK dimerization can be used as an index of activation or transactivation and then characterize the level of transactivation of many RTK-GPCR pairs, with cell cultures and primary neuron cultures with endogenous levels of RTKs and GPCRs. Dose-response curves were obtained from which pharmalogical parameters can be compared for each GPCR studied. Our data demonstrates that by allowing for time and space quantification of heterogenous oligomeric states, SpIDA enables systematic quantitative mechanistic studies not only of RTK transactivation at the cell membrane, but also of other cell signaling processes involving changes in protein oligomerization, trafficking and activity in different subcellular localizations. Finally, we studied the changes in number of synaptic sites in the neurons of the dorsal horn of the spinal cord of rats after a peripheral nerve injury (PNI), which consists of our model for chronic pain. We show that, after the PNI, there is a general decrease in synaptic sites together with a scaling or increasing of some of the GABAA receptor subunits. This scaling of the GABAA receptors at the postsynaptic sites was replicated by incubating the histological sections in a brain derivative nerve factor. Furthermore, we use SpIDA to obtain stoichiometry information for the GABAA receptor subunits directly at the postsynaptic sites. In short, we observe a switch from receptors containing two alpha1 to receptors containing two alpha2 and alpha3. This general change in subunits will have a direct effect on the cell as it will have different effects on the cell membrane conductance in response to GABA. As demonstrated, the advantages and greater versatility of SpIDA over current techniques opens the door to a new level of quantification for studies of protein interactions in native tissue using standard fluorescence microscopy. / Measuring protein interactions in subcellular compartments is key to understanding cell signalling mechanisms, but quantitative analysis of these interactions in situ has remained a major challenge. This thesis presents a novel analysis technique, spatial intensity distribution analysis (SpIDA), which may be applied to images obtained using fluorescence microscopy. SpIDA measures fluorescent particle densities and oligomerization states within individual images. The method is based on fitting intensity histograms from single images with super-Poissonian distributions to obtain density maps of fluorescent molecules and their quantal brightness. Since distributions are acquired spatially rather than temporally, this analysis may be applied to both live and chemically fixed cells and tissue. The technique does not rely on spatial correlations, freeing it from biases due to subcellular compartmentalization and heterogeneity within tissue samples. First, we validated the analysis technique evaluating its limits and demonstrating how it can be used to obtain useful information from complex biological samples. Analysis of simulations and heterodimeric GABAB receptors in spinal cord samples shows that the approach yields accurate estimates over a broad range of densities. SpIDA is applicable to sampling within subcell areas and reveals the presence of monomers and multimers with single dye labeling. We show that the substance P receptor (NK-1r) almost exclusively forms homodimers on the membrane and is primarily monomeric in the cytoplasm of dorsal horn neurons. Triggering receptor internalization caused a measurable decrease in homodimer density on the membrane surface. Finally, using GFP-tagged receptor subunits, we show that SpIDA can resolve dynamic changes in receptor oligomerization in live cells and is applicable to detection of high order oligomerization states. We then compared SpIDA results with those obtained from fluorescence lifetime imaging, and used it to extract information on receptor tyrosine kinase (RTK) dimerization at the cell membrane in response to GPCR activation. We show that RTK dimerization can be used as an index of activation or transactivation and then characterize the level of transactivation of many RTK-GPCR pairs, with cell cultures and primary neuron cultures with endogenous levels of RTKs and GPCRs. Dose-response curves were obtained from which pharmalogical parameters can be compared for each GPCR studied. Our data demonstrates that by allowing for time and space quantification of heterogenous oligomeric states, SpIDA enables systematic quantitative mechanistic studies not only of RTK transactivation at the cell membrane, but also of other cell signaling processes involving changes in protein oligomerization, trafficking and activity in different subcellular localizations. Finally, we studied the changes in number of synaptic sites in the neurons of the dorsal horn of the spinal cord of rats after a peripheral nerve injury (PNI), which consists of our model for chronic pain. We show that, after the PNI, there is a general decrease in synaptic sites together with a scaling or increasing of some of the GABAA receptor subunits. This scaling of the GABAA receptors at the postsynaptic sites was replicated by incubating the histological sections in a brain derivative nerve factor. Furthermore, we use SpIDA to obtain stoichiometry information for the GABAA receptor subunits directly at the postsynaptic sites. In short, we observe a switch from receptors containing two alpha1 to receptors containing two alpha2 and alpha3. This general change in subunits will have a direct effect on the cell as it will have different effects on the cell membrane conductance in response to GABA. As demonstrated, the advantages and greater versatility of SpIDA over current techniques opens the door to a new level of quantification for studies of protein interactions in native tissue using standard fluorescence microscopy.

The object of this thesis is to develop a new extension of Image Correlation Spectroscopy (ICS) that can measure velocity vectors for flowing protein populations in living cells. This new technique, called Spatio-Temporal Image Correlation Spectroscopy (STICS), allows measurement of both diffusion coefficients and velocity vectors (magnitude and direction) from fluorescence microscopy image time series of fluorescently labeled cellular proteins via monitoring of the time evolution of the full space-time correlation function of the intensity fluctuations. By using filtering in Fourier space to remove frequencies associated with immobile or slow components, it is possible to measure the protein transport even in the presence of a large fraction of immobile species that are static in the image series. The STICS method can generate complete transport maps of proteins within sub-regions of the basal membrane even if the protein concentration is too high to perform single particle tracking measurements, and it can be applied to any type of fluorescence microscopy image time series. This thesis presents the background theory, computer simulations, and analysis of measurements on fluorescent microspheres and fixed cell samples to demonstrate proof of principle, capabilities, and limitations of the method. Visible fluorescent proteins (VFPs) were used to label a variety of the proteins involved in cell-to-extra-cellular-matrix adhesions, including focal adhesion kinase, paxillin, alpha-actinin, alpha5-integrin, talin, vinculin and actin. Various fusion protein pairs were transfected in living cells and imaged using both laser scanning microscopy and total internal reflection microscopes. Using STICS analysis, co-transport maps of proteins were generated within protruding sub-regions of the basal membrane. The new space time image correlation method can probe the mechanistic details of the hypothesized molecular clutch that regulates the extra cellular matrix/cytoskeletal interactions during migration. The technique was also applied to mapping fluid flow in migrating keratocytes in order to elucidate the role that fluid flow plays in migrating cells.

Sensory neurons respond to a stimulus, such as light or sound with temporal sequences of electric pulses known as action potentials. Action potentials carry information about a stimulus in temporal sequences. Little is known, however, about how groups of neurons convey information, in part because of the complexity of natural stimuli and also because of neurons' complex dynamics that make them respond with spike patterns. Spike patterns, a tightly packed group of action potentials followed by a quiet period, are also known as a burst. Another kind of pattern is alternating time intervals between action potentials: a long interval followed by a short one, followed by a long one, knownas a nonrenewal pattern. Here we hypothesize that both bursts and nonrenewalpatterns may optimize information transmission in populations of neurons. To testthis, we used a combination of numerical neuron models, mathematical theory andelectrophysiology experiments. We investigated burst firing at the single neuronlevel as well as correlations between the activity of bursting neurons. In addition, weexplored information transmission by nonrenewal patterns in populations of neurons.We found that bursts of action potentials can act as single units of information.Also, bursts regulate correlated activity in neurons. Moreover, nonrenewal patternsincrease information transmission in groups of neurons coupled with excitation. Ourresults have implications for information coding by neural populations. In particular,our results suggest that spike patterns may optimize information transmission inpopulations of neurons. / En présence d'un stimulus, tel que la lumière ou le son, les neurones sensoriels répondent par des séquences temporelles d'impulsions électriques, appelées également des potentiels d'action. Il est généralement accepté que ces séquences temporelles de potentiels d'action acheminent des informations concernant le stimulus, cependant, la façon dont les neurones transmettent ces informations est difficile à comprendre, car les neurones agissent selon une dynamique complexe. C'est le cas, par exemple, lorsqu'ils répondent par des groupes de potentiels d'action serrés suivis d'intervalles de calme -- phénomène connu sous le nom de bouffée -- ou lorsque les potentiels d'actions sont alternativement entrecoupés d'intervalles de temps courts et longs -- ce que l'on appelle motifs de mémoire. Quoiqu'on comprenne les mécanismes impliqués dans la production de ces séquences temporelles (modèles temporels), leur rôle fonctionnel est moins bien compris.Dans ce texte, nous avançons l'hypothèse que ces deux types de séquences de potentiels d'action pourraient optimiser la transmission d'information dans des populations de neurones. Pour vérifier cela, nous avons eu recours à des modèles numêriques de neurones, à la théorie mathématique et à des expériences électrophysiologiques. Nous avons étudié les bouffées au niveau des neurones uniques. Ensuite, nous avons comparé la transmission d'information dans des groupes de neurones qui démontrent des motifs de mémoire avec celle dans des groupes qui n'en démontrent pas. Nous avons constaté que la transmission d'information peut être régulée par des séquences de potentiels d'action : dans des réseaux de neurones couplés, l'addition de motifs de mémoire peut augmenter la transmission d'information lorsque les neurones sont couplés avec de l'excitation. Nous avons également constaté que les bouffées régulaient l'activité corrélée de neurones qui reçoivent un stimulus commun avec un contraste qui varie selon le temps. Nos résultats suggèrent que les séquences de pic pourraient jouer un rôle important dans la modulation de la transmission d'information dans des populations de neurones.