Medida del consumo de energía en la descodificación de vídeos

Este proyecto se encuentra adscrito a la línea de investigación de optimización de consumo en terminales multimedia móviles que el Grupo de Diseño Electrónico y Microelectrónico (GDEM) de la UPM (Universidad Politécnica de Madrid) está llevando a cabo. Los sistemas empotrados móviles (Smartphone, Tablet,...) están alimentados con baterías. En este tipo de sistemas, una de las aplicaciones que más rápidamente consume energía es la descodificación de secuencias de vídeo. En este trabajo, queremos medir el consumo de energía de distintos descodificadores de vídeo para distintas secuencias, con el objetivo de entender mejor cómo se consume esta energía y poder encontrar diferentes métodos que lo reduzcan. Para ello comenzaremos describiendo nuestro entorno de trabajo, tanto el hardware como el software utilizado. Respeto al hardware cabe destacar el uso de una PandaBoard y un Smart Power de Odroid, entre otros muchos elementos utilizados, los cuales serán debidamente explicados en las siguientes páginas de este proyecto. Mientras que para el software destaca el uso de dos tipos de descodificadores uno CBP y otro PHP, los cuales serán descritos en profundidad en los siguientes capítulos de este documento. Este entorno de trabajo nos servirá para el estudio de las diferentes secuencias de vídeo, cuya codificación ha sido llevada en paralelo con otro proyecto que se está realizando en el grupo de GDEM de la UPM, y cuyo objetivo es el estudio de la calidad subjetiva durante la descodificación del mismo conjunto de secuencias de vídeo. Todas estas secuencias de vídeo han sido codificadas con diferentes parámetros de calidad y diversas estructuras de imágenes, para obtener así un banco de pruebas lo más amplio posible. Gracias a la obtención de estas secuencias de vídeo y utilizando nuestro entorno de trabajo, pasaremos a estudiar el consumo de energía que se produce al descodificar una a una todas las posibles secuencias de vídeo, dependiendo todo esto de su estructura de imágenes, su calidad y por supuesto, el descodificador utilizado en cada caso. Para terminar, se mostrará una comparativa entre los diferentes resultados obtenidos y se hará una discusión de estos, obteniendo en este caso, un resumen de los datos más significativos, así como las conclusiones más importantes obtenidas durante todo este trabajo. Al término de este proyecto y en unión con el estudio que se está llevando a cabo en paralelo sobre la calidad subjetiva, queda como línea futura de investigación encontrar el compromiso entre el consumo de energía de diferentes secuencias de vídeo y la calidad subjetiva de dichas secuencias. ABSTRACT. This project is assigned to the research line on consumption optimization in mobile multimedia terminals carried out by the Group of Electronic Design and Microelectronics (GDEM) of the Polytechnic University of Madrid (UPM). Embedded mobile systems (smartphones, tablets...) are powered by batteries. In such systems, one of the applications that more rapidly consumes power is the decoding of video streams. In this work, we measure the power consumption of different video decoders for different streams, in order to better understand how this energy is consumed and to find different methods to reduce it. To this end, we start by describing our working environment, both hardware and software used. As for the hardware, it is worth mentioning the use of PandaBoard and Smart Power Odroid, among many other elements, which will be duly explained in the following pages of this project. As for the software, we highlight the use of two types of decoders, CBP and PHP, which will be described in detail in the following chapters of this document. This working environment will help us to study different video streams, whose coding has been perfor-med in parallel under another project that is being carried out in the GDEM group of the UPM, and whose objective is the study of subjective quality for decoding the same set of video streams. All these video streams have been encoded with different quality parameters and image structures in order to obtain the widest set of samples. Thanks to the production of these video streams and the use of our working environment, we study the power consumption that occurs when decoding one by one all possible video streams, depending on the image structure, their quality and, of course, the decoder used in each case. Finally, we show a comparison between the different results and a discussion of these, obtaining a sum-mary of the most significant data and the main conclusions obtained during this project. Upon completion of this project and in conjunction with the project on the study of subjective quality that is being carried out in parallel, a future line of research could consist in finding the compromise between power consumption of different video streams and the subjective quality of these.

Long-range signaling within growing neurites mediated by neurotrophin-3

In addition to well established trophic functions, neurotrophins acutely affect neurotransmitter secretion from the presynaptic nerve terminal, influence synaptic development, and may serve as selective retrograde messengers that regulate synaptic efficacy. The crucial question related to the mechanisms of neurotrophin-mediated signaling is whether acute effects of neurotrophins are spatially restricted to the activated synapses. Here we have used a local perfusion technique for local delivery of neurotrophin-3 (NT-3) to various regions of developing Xenopus embryo neurons in culture. Within minutes after a focal exposure of a soma or a small (≈30 μm in length) axonal segment to NT-3, we observed an increase in the spontaneous neurotransmitter secretion from the presynaptic nerve terminals located ≈300–400 μm away from the site of NT-3 application. Secretory activity along the axonal shaft was not affected. Our findings suggest that the NT-3-mediated signal may rapidly travel through neuronal cytoplasm over unexpectedly long distances and modulate neurotransmitter release specifically at the presynaptic nerve terminals.

Long-term potentiation and dual-component quantal signaling in the dentate gyrus

Long-term potentiation (LTP) of excitatory transmission is an important candidate cellular mechanism for the storage of memories in the mammalian brain. The subcellular phenomena that underlie the persistent increase in synaptic strength, however, are incompletely understood. A potentially powerful method to detect a presynaptic increase in glutamate release is to examine the effect of LTP induction on the rate at which the use-dependent blocker MK-801 attenuates successive N-methyl-d-aspartic acid (NMDA) receptor-mediated synaptic signals. This method, however, has given apparently contradictory results when applied in hippocampal CA1. The inconsistency could be explained if NMDA receptors were opened by glutamate not only released from local presynaptic terminals, but also diffusing from synapses on neighboring cells where LTP was not induced. Here we examine the effect of pairing-induced LTP on the MK-801 blocking rate in two afferent inputs to dentate granule cells. LTP in the medial perforant path is associated with a significant increase in the MK-801 blocking rate, implying a presynaptic increase in glutamate release probability. An enhanced MK-801 blocking rate is not seen, however, in the lateral perforant path. This result still could be compatible with a presynaptic contribution to LTP in the lateral perforant path if intersynaptic cross-talk occurred. In support of this hypothesis, we show that NMDA receptors consistently sense more quanta of glutamate than do α-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid receptors. In the medial perforant path, in contrast, there is no significant difference in the number of quanta mediated by the two receptors. These results support a presynaptic contribution to LTP and imply that differences in intersynaptic cross-talk can complicate the interpretation of experiments designed to detect changes in transmitter release.

Differential roles of Ca2+/calmodulin-dependent kinases in posttetanic potentiation at input selective glutamatergic pathways

The electrosensory lateral line lobe (ELL) of the electric fish Apteronotus leptorhynchus is a layered medullary region receiving electroreceptor input that terminates on basal dendrites of interneurons and projection (pyramidal) cells. The molecular layer of the ELL contains two distinct glutamatergic feedback pathways that terminate on the proximal (ventral molecular layer, VML) and distal (dorsal molecular layer) apical dendrites of pyramidal cells. Western blot analysis with an antibody directed against mammalian Ca2+/calmodulin-dependent kinase 2, α subunit (CaMK2α) recognized a protein of identical size in the brain of A. leptorhynchus. Immunohistochemistry demonstrated that CaMK2 α expression in the ELL was restricted to fibers and terminals in the VML. Posttetanic potentiation (PTP) could be readily elicited in pyramidal cells by stimulation of either VML or DML in brain slices of the ELL. PTP in the VML was blocked by extracellular application of a CaMK2 antagonist (KN62) while intracellular application of KN62 or a CaMK2 inhibitory peptide had no effect, consistent with the presynaptic localization of CaMK2 α in VML. PTP in the dorsal molecular layer was not affected by extracellular application of KN62. Anti-Hebbian plasticity has also been demonstrated in the VML, but was not affected by KN62. These results demonstrate that, while PTP can occur independent of CaMK2, it is, in some synapses, dependent on this kinase.

Nedd4 mediates control of an epithelial Na+ channel in salivary duct cells by cytosolic Na+

Epithelial Na+ channels are expressed widely in absorptive epithelia such as the renal collecting duct and the colon and play a critical role in fluid and electrolyte homeostasis. Recent studies have shown that these channels interact via PY motifs in the C terminals of their α, β, and γ subunits with the WW domains of the ubiquitin-protein ligase Nedd4. Mutation or deletion of these PY motifs (as occurs, for example, in the heritable form of hypertension known as Liddle’s syndrome) leads to increased Na+ channel activity. Thus, binding of Nedd4 by the PY motifs would appear to be part of a physiological control system for down-regulation of Na+ channel activity. The nature of this control system is, however, unknown. In the present paper, we show that Nedd4 mediates the ubiquitin-dependent down-regulation of Na+ channel activity in response to increased intracellular Na+. We further show that Nedd4 operates downstream of Go in this feedback pathway. We find, however, that Nedd4 is not involved in the feedback control of Na+ channels by intracellular anions. Finally, we show that Nedd4 has no influence on Na+ channel activity when the Na+ and anion feedback systems are inactive. We conclude that Nedd4 normally mediates feedback control of epithelial Na+ channels by intracellular Na+, and we suggest that the increased Na+ channel activity observed in Liddle’s syndrome is attributable to the loss of this regulatory feedback system.

Selective loss of dopaminergic nigro-striatal neurons in brains of Atm-deficient mice

Ataxia-telangiectasia (AT) is a human disease caused by mutations in the ATM gene. The neural phenotype of AT includes progressive cerebellar neurodegeneration, which results in ataxia and eventual motor dysfunction. Surprisingly, mice in which the Atm gene has been inactivated lack distinct behavioral ataxia or pronounced cerebellar degeneration, the hallmarks of the human disease. To determine whether lack of the Atm protein can nonetheless lead to structural abnormalities in the brain, we compared brains from male Atm-deficient mice with male, age-matched controls. Atm-deficient mice exhibited severe degeneration of tyrosine hydroxylase-positive, dopaminergic nigro-striatal neurons, and their terminals in the striatum. This cell loss was accompanied by a large reduction in immunoreactivity for the dopamine transporter in the striatum. A reduction in dopaminergic neurons also was evident in the ventral tegmental area. This effect was selective in that the noradrenergic nucleus locus coeruleus was normal in these mice. Behaviorally, Atm-deficient mice expressed locomotor abnormalities manifested as stride-length asymmetry, which could be corrected by peripheral application of the dopaminergic precursor l-dopa. In addition, these mice were hypersensitive to the dopamine releasing drug d-amphetamine. These results indicate that ATM deficiency can severely affect dopaminergic neurons in the central nervous system and suggest possible strategies for treating this aspect of the disease.

Activation of nicotinic receptor-induced postsynaptic responses to luteinizing hormone-releasing hormone in bullfrog sympathetic ganglia via a Na+-dependent mechanism

Nicotine at very low doses (5–30 nM) induced large amounts of luteinizing hormone-releasing hormone (LHRH) release, which was monitored as slow membrane depolarizations in the ganglionic neurons of bullfrog sympathetic ganglia. A nicotinic antagonist, d-tubocurarine chloride, completely and reversibly blocked the nicotine-induced LHRH release, but it did not block the nerve-firing-evoked LHRH release. Thus, nicotine activated nicotinic acetylcholine receptors and produced LHRH release via a mechanism that is different from the mechanism for evoked release. Moreover, this release was not caused by Ca2+ influx through either the nicotinic receptors or the voltage-gated Ca2+ channels because the release was increased moderately when the extracellular solution was changed into a Ca2+-free solution that also contained Mg2+ (4 mM) and Cd2+ (200 μM). The release did not depend on Ca2+ release from the intraterminal Ca2+ stores either because fura-2 fluorimetry showed extremely low Ca2+ elevation (≈30 nM) in response to nicotine (30 nM). Moreover, nicotine evoked LHRH release when [Ca2+] elevation in the terminals was prevented by loading the terminals with 1,2-bis(2-aminophenoxy)ethane-N,N,N′,N′-tetraacetic acid and fura-2. Instead, the nicotine-induced release required extracellular Na+ because substitution of extracellular NaCl with N-methyl-d-glucamine chloride completely blocked the release. The Na+-dependent mechanism was not via Na+ influx through the voltage-gated Na+ channels because the release was not affected by tetrodotoxin (1–50 μM) plus Cd2+ (200 μM). Thus, nicotine at very low concentrations induced LHRH release via a Na+-dependent, Ca2+-independent mechanism.

The SH3p4/Sh3p8/SH3p13 protein family: Binding partners for synaptojanin and dynamin via a Grb2-like Src homology 3 domain

The GTPase dynamin I and the inositol 5-phosphatase synaptojanin are nerve terminal proteins implicated in synaptic vesicle recycling. Both proteins contain COOH-terminal proline-rich domains that can interact with a variety of Src homology 3 (SH3) domains. A major physiological binding partner for dynamin I and synaptojanin in the nervous system is amphiphysin I, an SH3 domain-containing protein also concentrated in nerve terminals. We have used the proline-rich tail of synaptojanin to screen a rat brain library by the two-hybrid method to identify additional interacting partners of synaptojanin. Three related proteins containing SH3 domains that are closely related to the SH3 domains of Grb2 were isolated: SH3p4, SH3p8, and SH3p13. Further biochemical studies demonstrated that the SH3p4/8/13 proteins bind to both synaptojanin and dynamin I. The SH3p4/8/13 transcripts are differentially expressed in tissues: SH3p4 mRNA was detected only in brain, SH3p13 mRNA was present in brain and testis, and the SH3p8 transcript was detected at similar levels in multiple tissues. Members of the SH3p4/8/13 protein family were found to be concentrated in nerve terminals, and pools of synaptojanin and dynamin I were coprecipitated from brain extracts with antibodies recognizing SH3p4/8/13. These findings underscore the important role of SH3-mediated interactions in synaptic vesicle recycling.

Fe-catalyzed cleavage of the α subunit of Na/K-ATPase: Evidence for conformation-sensitive interactions between cytoplasmic domains

Incubation of Na/K-ATPase with ascorbate plus H2O2 produces specific cleavage of the α subunit. Five fragments with intact C termini and complementary fragments with intact N termini were observed. The β subunit is not cleaved. Cleavages depend on the presence of contaminant or added Fe2+ ions, as inferred by suppression of cleavages with nonspecific metal complexants (histidine, EDTA, phenanthroline) or the Fe3+-specific complexant desferrioxamine, or acceleration of cleavages by addition of low concentrations of Fe2+ but not of other heavy metal ions. Na/K-ATPase is inactivated in addition to cleavage, and both effects are insensitive to OH⋅ radical scavengers. Cleavages are sensitive to conformation. In low ionic strength media (E2) or media containing Rb ions [E2(Rb)], cleavage is much faster than in high ionic strength media (E1) or media containing Na ions (E1Na). N-terminal fragments and two C-terminal fragments (N-terminals E214 and V712) have been identified by amino acid sequencing. Approximate positions of other cleavages were determined with specific antibodies. The results suggest that Fe2+ (or Fe3+) ions bind with high affinity at the cytoplasmic surface and catalyze cleavages of peptide bonds close to the Fe2+ (or Fe3+) ion. Thus, cleavage patterns can provide information on spatial organization of the polypeptide chain. We propose that highly conserved regions of the α subunit, within the minor and major cytoplasmic loops, interact in the E2 or E2(Rb) conformations but move apart in the E1 or E1Na conformations. We discuss implications of domain interactions for the energy transduction mechanism. Fe-catalyzed cleavages may be applicable to other P-type pumps or membrane proteins.

Synapsin I interacts with c-Src and stimulates its tyrosine kinase activity

Synapsin I is a synaptic vesicle-associated phosphoprotein that has been implicated in the formation of presynaptic specializations and in the regulation of neurotransmitter release. The nonreceptor tyrosine kinase c-Src is enriched on synaptic vesicles, where it accounts for most of the vesicle-associated tyrosine kinase activity. Using overlay, affinity chromatography, and coprecipitation assays, we have now shown that synapsin I is the major binding protein for the Src homology 3 (SH3) domain of c-Src in highly purified synaptic vesicle preparations. The interaction was mediated by the proline-rich domain D of synapsin I and was not significantly affected by stoichiometric phosphorylation of synapsin I at any of the known regulatory sites. The interaction of purified c-Src and synapsin I resulted in a severalfold stimulation of tyrosine kinase activity and was antagonized by the purified c-Src-SH3 domain. Depletion of synapsin I from purified synaptic vesicles resulted in a decrease of endogenous tyrosine kinase activity. Portions of the total cellular pools of synapsin I and Src were coprecipitated from detergent extracts of rat brain synaptosomal fractions using antibodies to either protein species. The interaction between synapsin I and c-Src, as well as the synapsin I-induced stimulation of tyrosine kinase activity, may be physiologically important in signal transduction and in the modulation of the function of axon terminals, both during synaptogenesis and at mature synapses.

Calcium-dependent switching of the specificity of phosphoinositide binding to synaptotagmin

The synaptic vesicle membrane protein synaptotagmin (tagmin) is essential for fast, calcium-dependent, neurotransmitter release and is likely to be the calcium sensor for exocytosis, because of its many calcium-dependent properties. Polyphosphoinositides are needed for exocytosis, but it has not been known why. We now provide a possible connection between these observations with the finding that the C2B domain of tagmin I binds phosphatidylinositol-4,5-bisphosphate (PIns-4,5-P2), its isomer phosphatidylinositol-3,4-bisphosphate and phosphatidylinositol-3,4,5-trisphosphate (PIns-3,4,5-P3). Calcium ions switch the specificity of this binding from PIns-3,4,5-P3 (at calcium concentrations found in resting nerve terminals) to PIns-4,5-P2 (at concentration of calcium required for transmitter release). Inositol polyphosphates, known blockers of neurotransmitter release, inhibit the binding of both PIns-4,5-P2 and PIns-3,4,5-P3 to tagmin. Our findings imply that tagmin may operate as a bimodal calcium sensor, switching bound lipids during exocytosis. This connection to polyphosphoinositides, compounds whose levels are physiologically regulated, could be important for long-term memory and learning.

Australian lungfish neurohypophysial hormone genes encode vasotocin and [Phe2]mesotocin precursors homologous to tetrapod-type precursors

In view of the well-established role of neurohypophysial hormones in osmoregulation of terrestrial vertebrates, lungfishes are a key group for study of the molecular and functional evolution of the hypothalamo-neurohypophysial system. Here we report on the primary structure of the precursors encoding vasotocin (VT) and [Phe2]mesotocin ([Phe2]MT) of the Australian lungfish, Neoceratodus forsteri. Genomic sequence analysis and Northern blot analysis confirmed that [Phe2]MT is a native oxytocin family peptide in the Australian lungfish, although it has been reported that the lungfish neurohypophysis contains MT. The VT precursor consists of a signal peptide, VT, that is connected to a neurophysin by a Gly-Lys-Arg sequence, and a copeptin moiety that includes a Leu-rich core segment and a glycosylation site. In contrast, the [Phe2]MT precursor does not contain a copeptin moiety. These structural features of the lungfish precursors are consistent with those in tetrapods, but different from those in teleosts where both VT and isotocin precursors contain a copeptin-like moiety without a glycosylation site at the carboxyl terminals of their neurophysins. Comparison of the exon/intron organization also supports homology of the lungfish [Phe2]MT gene with tetrapod oxytocin/MT genes, rather than with teleost isotocin genes. Moreover, molecular phylogenetic analysis shows that neurohypophysial hormone genes of the lungfish are closely related to those of the toad. The present results along with previous morphological findings indicate that the hypothalamo-neurohypophysial system of the lungfish has evolved along the tetrapod lineage, whereas the teleosts form a separate lineage, both within the class Osteichthyes.

HU-308: A specific agonist for CB2, a peripheral cannabinoid receptor

Two cannabinoid receptors have been identified: CB1, present in the central nervous system (CNS) and to a lesser extent in other tissues, and CB2, present outside the CNS, in peripheral organs. There is evidence for the presence of CB2-like receptors in peripheral nerve terminals. We report now that we have synthesized a CB2-specific agonist, code-named HU-308. This cannabinoid does not bind to CB1 (Ki > 10 μM), but does so efficiently to CB2 (Ki = 22.7 ± 3.9 nM); it inhibits forskolin-stimulated cyclic AMP production in CB2-transfected cells, but does so much less in CB1-transfected cells. HU-308 shows no activity in mice in a tetrad of behavioral tests, which together have been shown to be specific for tetrahydrocannabinol (THC)-type activity in the CNS mediated by CB1. However, HU-308 reduces blood pressure, blocks defecation, and elicits anti-inflammatory and peripheral analgesic activity. The hypotension, the inhibition of defecation, the anti-inflammatory and peripheral analgesic effects produced by HU-308 are blocked (or partially blocked) by the CB2 antagonist SR-144528, but not by the CB1 antagonist SR-141716A. These results demonstrate the feasibility of discovering novel nonpsychotropic cannabinoids that may lead to new therapies for hypertension, inflammation, and pain.

The neuropeptide Y/agouti gene-related protein (AGRP) brain circuitry in normal, anorectic, and monosodium glutamate-treated mice

Neuropeptide Y (NPY) and the endogenous melanocortin receptor antagonist, agouti gene-related protein (AGRP), coexist in the arcuate nucleus, and both exert orexigenic effects. The present study aimed primarily at determining the brain distribution of AGRP. AGRP mRNA-expressing cells were limited to the arcuate nucleus, representing a major subpopulation (95%) of the NPY neurons, which also was confirmed with immunohistochemistry. AGRP-immunoreactive (-ir) terminals all contained NPY and were observed in many brain regions extending from the rostral telencephalon to the pons, including the parabrachial nucleus. NPY-positive, AGRP-negative terminals were observed in many areas. AGRP-ir terminals were reduced dramatically in all brain regions of mice treated neonatally with monosodium glutamate as well as of mice homozygous for the anorexia mutation. Terminals immunoreactive for the melanocortin peptide α-melanocyte-stimulating hormone formed a population separate from, but parallel to, the AGRP-ir terminals. Our results show that arcuate NPY neurons, identified by the presence of AGRP, project more extensively in the brain than previously known and indicate that the feeding regulatory actions of NPY may extend beyond the hypothalamus.

Ablation of P/Q-type Ca2+ channel currents, altered synaptic transmission, and progressive ataxia in mice lacking the α1A-subunit

The Ca2+ channel α1A-subunit is a voltage-gated, pore-forming membrane protein positioned at the intersection of two important lines of research: one exploring the diversity of Ca2+ channels and their physiological roles, and the other pursuing mechanisms of ataxia, dystonia, epilepsy, and migraine. α1A-Subunits are thought to support both P- and Q-type Ca2+ channel currents, but the most direct test, a null mutant, has not been described, nor is it known which changes in neurotransmission might arise from elimination of the predominant Ca2+ delivery system at excitatory nerve terminals. We generated α1A-deficient mice (α1A−/−) and found that they developed a rapidly progressive neurological deficit with specific characteristics of ataxia and dystonia before dying ≈3–4 weeks after birth. P-type currents in Purkinje neurons and P- and Q-type currents in cerebellar granule cells were eliminated completely whereas other Ca2+ channel types, including those involved in triggering transmitter release, also underwent concomitant changes in density. Synaptic transmission in α1A−/− hippocampal slices persisted despite the lack of P/Q-type channels but showed enhanced reliance on N-type and R-type Ca2+ entry. The α1A−/− mice provide a starting point for unraveling neuropathological mechanisms of human diseases generated by mutations in α1A.