Cytoplasmic streaming in giant algal cells: A historical survey of experimental approaches

Various methods have been used to study cytoplasmic streaming in giant algal cells during the past three decades. Simple techniques can be used with characean internodal cells to modify the cell constitution in various ways to gain insight into the mechanism of cytoplasmic streaming. Another method involves isolatingin vitro a huge drop of uninjured endoplasm, to examine its physical and dynamic properties. The motive force responsible for streaming has been measured by three different techniques with similar results. Subcortical fibrils consisting of bundles of F-actin with the same polarity are indispensable for streaming. Differential treatment of the endoplasm and ectoplasm has shown that putative characean myosin is localized in the endoplasm. Studies of the roles of ATP, Mg²⁺, Ca²⁺, H⁺ etc. in the streaming have been conducted by cellular perfusion, which allows removal of the tonoplast, or by techniques permeabilizing the protoplasmic membrane. A slow version of the movement can even be artificially reproduced by combining characean actinin situ and exogenous myosin in the presence of Mg-ATP. The findings thus far obtained support the hypothesis that cytoplasmic streaming in characean cells is caused by an active shearing force produced by interaction of the actin filament bundles on the cortex with myosin in the endoplasm.     

Accelerated sliding of pollen tube organelles along Characeae actin bundles regulated by Ca2+

Pollen tubes show active cytoplasmic streaming. We isolated organelles from pollen tubes and tested their ability to slide along actin bundles in characean cell models. Here, we show that sliding of organelles was ATP-dependent and that motility was lost after N-ethylmaleimide or heat treatment of organelles. On the other hand, cytoplasmic streaming in pollen tube was inhibited by either N-ethylmaleimide or heat treatment. These results strongly indicate that cytoplasmic streaming in pollen tubes is supported by the "actomyosin"-ATP system. The velocity of organelle movement along characean actin bundles was much higher than that of the native streaming in pollen tubes. We suggested that pollen tube "myosin" has a capacity to move at a velocity of the same order of magnitude as that of characean myosin. Moreover, the motility was high at Ca2+ concentrations lower than 0.18 microM (pCa 6.8) but was inhibited at concentration higher than 4.5 microM (pCa 5.4). In conclusion, cytoplasmic streaming in pollen tubes is suggested to be regulated by Ca2+ through "myosin" inactivation.     

Myosin and Ca2+-sensitive streaming in the alga Chara: detection of two polypeptides reacting with a monoclonal anti-myosin and their localization in the streaming endoplasm

A monoclonal antibody to the heavy chain of myosin from mouse 3T3 cells was used to detect and localize related proteins in the green alga Chara. Proteins of 200,000 and 110,000 Mr reacted on immunoblots of proteins precipitated rapidly with trichloroacetic acid to minimize proteolysis. Immunofluorescence of whole cells localized these proteins to organelles of the streaming endoplasm, to a system of endoplasmic strands and to the subcortical actin bundles. Except that fewer endoplasmic strands and organelles were found and the strands were tangled, the localization pattern was similar in cells rapidly perfused to remove the bulk of the streaming endoplasm. Actin was confined almost entirely to the system of subcortical actin bundles in both whole and perfused cells. Myosin that was associated with the tangled endoplasmic strands but not that associated with the organelles or actin bundles was removed by concentrations of Ca2+ inhibiting ATP-dependent streaming in perfused cells. ATP extracted both organelles and endoplasmic strands but left a continuous pattern of myosin immunostaining along the actin bundles. The findings are discussed in relation to the possible existence of two forms of myosin and of separate mechanisms moving the bulk endoplasm and individual organelles.     

Microtubule disassembly enhances reversible cytochalasin-dependent disruption of actin bundles in characean internodes

Summary Parallel bundles of actin filaments at the cortex-endoplasm interface provide tracks for myosin-generated cytoplasmic streaming in characean internodes. These bundles resist disassembly or structural modification when exposed to 10 μM cytochalasin D (CD) even though this concentration of CD rapidly (within minutes) but reversibly arrests streaming. Unexpectedly, we discovered that prolonged treatment with lower concentrations of CD could partially disassemble the subcortical actin bundles. Actin bundles became discontinuous following one- to several-day treatment with concentrations (6 μM) that reduced but did not arrest streaming, and the residual fragments mostly remained parallel to the chloroplast files. When microtubules were concurrently disassembled with tubulin-specific drugs, however, low CD concentrations (2.5–3 μM) completely arrested bulk streaming, disrupted the largely 2-dimensional actin bundle array and caused the formation of a coarse, thick-meshed actin network that extended from the cortex to the endoplasm. Despite such massive reconstruction, drug removal enabled cells to recover continuous parallel bundles and streaming. Recovery was possible if both or just one of the drugs were removed. In recovered cells, the streaming pattern frequently redeveloped in new directions that did not follow the chloroplast files, and later, chloroplast files readjusted to the new polarity established by the actin bundles. This first report on the complete and reversible disassembly of characean actin bundles provides new insights into the mechanism of actin bundle assembly and organization and supports the idea of indirect interactions between actin filaments and microtubules.     

Effects of exogenous proteins on cytoplasmic streaming in perfused Chara cells

Cytoplasmic streaming in characean algae is thought to be generated by interaction between subcortical actin bundles and endoplasmic myosin. Most of the existing evidence supporting this hypothesis is of a structural rather than functional nature. To obtain evidence bearing on the possible function of actin and myosin in streaming, we used perfusion techniques to introduce a number of contractile and related proteins into the cytoplasm of streaming Chara cells. Exogenous actin added at concentrations as low as 0.1 mg/ml is a potent inhibitor of streaming. Deoxyribonuclease I (DNase I), an inhibitor of amoeboid movement and fast axonal transport, does not inhibit streaming in Chara. Fluorescein-DNase I stains stress cables and microfilaments in mammalian cells but does not bind to Chara actin bundles, thus suggesting that the lack of effect on streaming is due to a surprising lack of DNase I affinity for Chara actin bundles. Heavy meromyosin (HMM) does not inhibit streaming, but fluorescein-HMM (FL-HMM), having a partially disabled EDTA ATPase, does. Quantitative fluorescence micrography provides evidence that inhibition of streaming by FL-HMM may be due to a tendency for FL-HMM to remain bound to Chara actin bundles even in the presence of MgATP. Perfusion with various control proteins, including tubulin, ovalbumin, bovine serum albumin, and irrelevant antibodies, does not inhibit streaming. These results support the hypothesis that actin and myosin function to generate cytoplasmic streaming in Chara.     

Fluorescence studies on modes of cytochalasin B and phallotoxin action on cytoplasmic streaming in Chara

Various investigations have suggested that cytoplasmic streaming in characean algae is driven by interaction between subcortical actin bundles and endoplasmic myosin. To further test this hypothesis, we have perfused cytotoxic actin-binding drugs and fluorescent actin labels into the cytoplasm of streaming Chara cells. Confirming earlier work, we find that cytochalasin B (CB) reversibly inhibits streaming. In direct contrast to earlier investigators, who have found phalloidin to be a potent inhibitor of movement in amoeba, slime mold, and fibroblastic cells, we find that phalloidin does not inhibit streaming in Chara but does modify the inhibitory effect of CB. Use of two fluorescent actin probes, fluorescein, isothiocyanate-heavy meromyosin (FITC-HMM) and nitrobenzoxadiazole-phallacidin (NBD-Ph), has permitted visualization of the effects of CB and phalloidin on the actin bundles. FITC-HMM labeling in perfused but nonstreaming cells has revealed a previously unobserved alteration of the actin bundles by CB. Phalloidin alone does not perceptibly alter the actin bundles but does block the alteration by CB if applied as a pretreatment, NBD-Ph perfused into the cytoplasm of streaming cells stains actin bundles without inhibiting streaming. NBD-Ph staining of actin bundles is not initially observed in cells inhibited by CB but does appear simultaneously with the recovery of streaming as CB leaks from the cells. The observations reported here are consistent with the established effects of phallotoxins and CB on actin in vitro and support the hypothesis that streaming is generated by actin-myosin interactions.     

Protein phosphorylation regulates actomyosin-driven vesicle movement in cell extracts isolated from the green algae,Chara corallina

In Characean cells endoplasmic streaming stops upon membrane depolarization accompanied by Ca(2+) entry. We investigated the mechanism of this cessation of endoplasmic streaming by reconstituting the vesicle movement in vitro. In a living cell of Chara corallina, there are a number of vesicles moving along actin cables. Vesicles in the endoplasm squeezed out of the cell into a medium containing Mg-ATP showed directional movements under a dark field microscope. When the extracted endoplasm was treated with 20 nM okadaic acid, vesicles showed only movements like the Brownian motion. When it was treated with 50 nM staurosporine, directional movements of vesicles were activated. These movements were analyzed by image processing of videomicroscopic records. Vesicle movements along F-actin filaments were also observed by merging both images of the same field by dark field microscopy and fluorescence microscopy, indicating that myosin on the vesicle surface was responsible for vesicle movements. We also examined the effects of okadaic acid and staurosporine on in vitro sliding of F-actin on Chara myosin. When Chara myosin was treated with 20 nM okadaic acid in the cell extract, the number of sliding F-actin filaments was greatly reduced. In contrast, it increased when Chara myosin was treated with 50 nM staurosporine. In addition, Chara myosin treated with protein kinase C greatly diminished its motility. These results suggest that inactivation of Chara myosin via its phosphorylation is responsible for cessation of endoplasmic streaming.     

Characean actin bundles as a tool for stydying actomyosin-based motility

At the inner surface of the stagnant chloroplasts of Characeae cells, bundles of actin filaments having uniform polarity are anchored. These bundles are responsible for generating the motive force of cytoplasmic streaming. It is now possible to induce movement of either beads coated with foreign myosin or organelles associated with myosin along the characean actin bundles. The Ca²⁺ sensitivities of the reconstitued movements are consistent with those of the actin-activated myosin ATPases. The use of reconstituted systems is finding wide application in the detection of various myosins in materials from which myosin is not significantly purified. Furthermore, sliding velocities and the Ca²⁺ regulation of myosins bound to organelles are now being determined. Recipient of the Botanical Society Award for Young Scientists, 1987.     

The mechanism of cytoplasmic streaming in characean algal cells: sliding of endoplasmic reticulum along actin filaments

Electron microscopy of directly frozen giant cells of characean algae shows a continuous, tridimensional network of anastomosing tubes and cisternae of rough endoplasmic reticulum which pervade the streaming region of their cytoplasm. Portions of this endoplasmic reticulum contact the parallel bundles of actin filaments at the interface with the stationary cortical cytoplasm. Mitochondria, glycosomes, and other small cytoplasmic organelles enmeshed in the endoplasmic reticulum network display Brownian motion while streaming. The binding and sliding of endoplasmic reticulum membranes along actin cables can also be directly visualized after the cytoplasm of these cells is dissociated in a buffer containing ATP. The shear forces produced at the interface with the dissociated actin cables move large aggregates of endoplasmic reticulum and other organelles. The combination of fast-freezing electron microscopy and video microscopy of living cells and dissociated cytoplasm demonstrates that the cytoplasmic streaming depends on endoplasmic reticulum membranes sliding along the stationary actin cables. Thus, the continuous network of endoplasmic reticulum provides a means of exerting motive forces on cytoplasm deep inside the cell distant from the cortical actin cables where the motive force is generated.     

Actin in living and fixed characean internodal cells: identification of a cortical array of fine actin strands and chloroplast actin rings

Summary We report on the novel features of the actin cytoskeleton and its development in characean internodal cells. Images obtained by confocal laser scanning microscopy after microinjection of living cells with fluorescent derivatives of F-actin-specific phallotoxins, and by modified immunofluorescence methods using fixed cells, were mutually confirmatory at all stages of internodal cell growth. The microinjection method allowed capture of 3-dimensional images of high quality even though photobleaching and apparent loss of the probes through degradation and uptake into the vacuole made it difficult to record phallotoxin-labelled actin over long periods of time. When injected at appropriate concentrations, phallotoxins affected neither the rate of cytoplasmic streaming nor the long-term viability of cells. Recently formed internodal cells have relatively disorganized actin bundles that become oriented in the subcortical cytoplasm approximately parallel to the newly established long axis and traverse the cell through transvacuolar strands. In older cells with central vacuoles not traversed by cytoplasmic strands, subcortical bundles are organized in parallel groups that associate closely with stationary chloroplasts, now in files. The parallel arrangement and continuity of actin bundles is maintained where they pass round nodal regions of the cell, even in the absence of chloroplast files. This study reports on two novel structural features of the characean internodal actin cytoskeleton: a distinct array of actin strands near the plasma membrane that is oriented transversely during cell growth and rings of actin around the chloroplasts bordering the neutral line, the zone that separates opposing flows of endoplasm.     

Differential treatment ofAcetabularia with cytochalasin B and N-Ethylmaleimide with special reference to their effects on cytoplasmic streaming

Summary Cytoplasmic streaming in the stalk ofAcetabularia, ryukyuensis at the vegetative stage was reversibly inhibited by cytochalasin B (cB) of 50 g/ml and irreversibly by N-Ethylmaleimide (NEM) above concentrations of 0.25 mM.After the endoplasm and the chloroplasts were pushed forward one end of the stalk by gentle centrifugation at about 500 × g for 3 minutes, numerous ectoplasmic striations remainedin situ in the stalk cortex. The striations ran in parallel with the longitudinal axis of the stalk at unequal intervals. The endoplasm streamed back only along these striations.By combining centrifugation and a double chamber technique, the endoplasm and the cortex of the stalk were treated separately with CB or NEM. CB treatment of the cortex arrested streaming; when treatment was restricted to the endoplasm, streaming continued at an normal rate. NEM treatment restricted to the cortex permitted normal streaming rates. Treatment restricted to the moving endoplasm inhibited streaming.These results suggest that microfilaments and a moiety, possibly myosin, play an active role in the streaming. Microfilaments must reside in the cortex, especially in the ectoplasmic striations, while the putative myosin must reside in the moving endoplasm.     

Regulation of myosin sliding alongChara actin bundles by native skeletal muscle tropomyosin

Summary Native tropomyosin from rabbit skeletal muscle introduced by intracellular perfusion intoChara cells inhibited the cytoplasmic streaming irrespective of the Ca²⁺ concentration. To find the action site of native tropomyosin inChara, the cytoplasmic streaming was reconstituted by introducing isolated endoplasm into actin donorChara cells from which native endoplasm had been removed. The reconstituted streaming was inhibited by pretreatment of the actin donor cells with native tropomyosin but not by that of the endoplasm, suggesting that the native tropomyosin inhibited the cytoplasmic streaming by binding toChara actin bundles. Staining of the actin bundles with FITC-labeled native tropomyosin also showed that the native tropomyosin could bind to the actin bundles. Streaming reconstituted fromChara actin bundles and skeletal muscle myosin was insensitive to Ca²⁺, but became sensitive on application of the native tropomyosin.Abbrevations APW artificial pond water - ATP adenosine 5-triphosphoric acid - BSA bovine serum albumin - EDTA ethylene diamine tetraacetic acid - EGTA ethyleneglycol-bis-(-aminoethylether) N,N,N,N-tetraacetic acid - FITC fluorescein isothiocyanate - FITC-NTM fluorescein isothiocyanate-labeled native tropomyosin - NTM native tropomyosin     

Wide-ranging effects of eight cytochalasins and latrunculin A and B on intracellular motility and actin filament reorganization in characean internodal cells

Numerous forms of cytochalasins have been identified and, although they share common biological activity, they may differ considerably in potency. We investigated the effects of cytochalasins A, B, C, D, E, H and J and dihydrocytochalasin B in an ideal experimental system for cell motility, the giant internodal cells of the characean alga Nitella pseudoflabellata. Cytochalasins D (60 microM) and H (30 microM) were found to be most suited for fast and reversible inhibition of actin-based motility, while cytochalasins A and E arrested streaming at lower concentrations but irreversibly. We observed no clear correlation between the ability of cytochalasins to inhibit motility and the actual disruption of the subcortical actin bundle tracks on which myosin-dependent motility occurs. Indeed, the actin bundles remained intact at the time of streaming cessation and disassembled only after one to several days' treatment. Even when applied at concentrations lower than that required to inhibit cytoplasmic streaming, all of the cytochalasins induced reorganization of the more labile cortical actin filaments into actin patches, swirling clusters or short rods. Latrunculins A and B arrested streaming only after disrupting the subcortical actin bundles, a process requiring relatively high concentrations (200 microM) and very long treatment periods of >1 d. Latrunculins, however, worked synergistically with cytochalasins. A 1 h treatment with 15 nM latrunculin A and 4 microM cytochalasin D induced reversible fragmentation of subcortical actin bundles and arrested cytoplasmic streaming. Our findings provide insights into the mechanisms by which cytochalasins and latrunculins interfere with characean actin to inhibit motility.     

Cloning and characterization of a myosin from characean alga, the fastest motor protein in the world

In characean algae, very rapid cytoplasmic streaming is generated by sliding movement of an unconventional myosin on fixed actin cables. The speed of this sliding movement is the fastest among many molecular motors known so far. We have cloned a set of overlapping cDNAs encoding the heavy chain of this myosin by immunoscreening with antibody raised against characean myosin. The molecular mass of this heavy chain is 248 kDa, and the protein has a conserved motor domain, six IQ motifs, an extensive alpha-helical coiled-coil domain, and a C-terminal globular domain. Phylogenetic analysis suggested that this myosin belongs to class XI.     

Effects of Ca(2+) and calmodulin on the motile activity of characean myosin in vitro

It is well known that the cytoplasmic streaming of characean cells is readily inhibited by Ca(2+). However, neither the actin-activated MgATPase nor the in vitro motile activity of purified characean myosin were inhibited by Ca(2+). Recently, amino acid sequence of characean myosin was determined in our laboratory and the sequence revealed that characean myosin contains six calmodulin binding sites in the neck region. We also detected calmodulin in quickly prepared characean myosin fraction. It is, therefore, possible that the insensitivity of characean myosin to Ca(2+) is due to the dissociation of some calmodulin molecules from the neck region during the course of protein purification. To determine strictly the Ca(2+) sensitivity of characean myosin, we intentionally used crude preparation of characean myosin to reduce the possibility of calmodulin dissociation and examined the motile activity of characean myosin in vitro in the presence of excess characean calmodulin. We could not observe any drastic inhibition of characean myosin activity by Ca(2+). The results suggest that the brief cessation of cytoplasmic streaming is not caused by the direct inhibition of myosin activity by Ca(2+).     

Myosin heavy chains: Detection by immuno-blotting in higher plants and localization by immunofluorescence in the alga chara

A monoclonal antibody to the heavy chain of myosin from mouse 3T3 cells was used to identify myosin heavy chains in four flowering plants and to identify and localize them in the green alga Chara. The M_r of the immunoreactive bands varied from ca 200 000 in Chara and Arabidopsis to 170 000 in mung beans, peas and wheat. An additional band of 158 000-M_r was resolved in roots and shoots of mung beans. Chara contained a second, immunoreactive band of 110 000-M_r whose possible relationship to the tail-less myosin I enzymes is discussed.Immunofluorescence of giant internodal cells of Chara showed that myosin was almost entirely confined to the streaming endoplasm. Individual organelles and beaded endoplasmic strands were heavily labelled as were the sub-cortical filament bundles. Actin, in contrast, was confined to the sub-cortical bundles. It is proposed that force is generated by interaction of the actin in the subcortical bundles with myosin on individual organelles and on the beaded endoplasmic strands. By ramifying through the endoplasm, the strands may ensure the cohesive movement of the whole mass of endoplasm.     

The Stepping Pattern of Myosin X Is Adapted for Processive Motility on Bundled Actin

Myosin X is a molecular motor that is adapted to select bundled actin filaments over single actin filaments for processive motility. Its unique form of motility suggests that myosin X's stepping mechanism takes advantage of the arrangement of actin filaments and the additional target binding sites found within a bundle. Here we use fluorescence imaging with one-nanometer accuracy to show that myosin X takes steps of ∼18 nm along a fascin-actin bundle. This step-size is well short of the 36-nm step-size observed in myosin V and myosin VI that corresponds to the actin pseudohelical repeat distance. Myosin X is able to walk along bundles with this step-size if it straddles two actin filaments, but would be quickly forced to spiral into the constrained interior of the bundle if it were to use only a single actin filament. We also demonstrate that myosin X takes many sideways steps as it walks along a bundle, suggesting that it can switch actin filament pairs within the bundle as it walks. Sideways steps to the left or the right occur on bundles with equal frequency, suggesting a degree of lateral flexibility such that the motor's working stroke does not bias it to the left or to the right. On single actin filaments, we find a broad mixture of 10-20-nm steps, which again falls short of the 36-nm actin repeat. Moreover, the motor leans to the right as it walks along single filaments, which may require myosin X to adopt strained configurations. As a control, we also tracked myosin V stepping along actin filaments and fascin-actin bundles. We find that myosin V follows a narrower path on both structures, walking primarily along one surface of an actin filament and following a single filament within a bundle while occasionally switching to neighboring filaments. Together, these results delineate some of the structural features of the motor and the track that allow myosin X to recognize actin filament bundles.     

Endoplasmic filaments generate the motive force for rotational streaming in Nitella

The streaming endoplasm of characean cells has been shown to contain previously unreported endoplasmic filaments along which bending waves are observed under the light microscope using special techniques. The bending waves are similar to those propagated along sperm tails causing propulsion of sperm. In Nitella there is reason to believe that nearly all of the filaments are anchored in the cortex and that their beating propels the endoplasm in which they are suspended. This hypothesis is supported by calculations in which typical and average wave parameters have been inserted into the classical hydrodynamic equations derived for sperm tail bending waves. These calculations come within an order of magnitude of predicting the velocity of streaming and they show that waves of the character described, propagated along an estimated 52 m of endoplasmic filaments per cell, must generate a total motive force per cell within less than an order of magnitude of the forces measured experimentally by others. If we assume that undulating filaments produce the force driving the endoplasm, then the method described for measuring the motive force could lead to a lower than actual value for the motive force, since both centrifugation and vacuolar perfusion would reverse the orientation of some filaments. Observations of the initiation of particle translation in association with the filaments suggest that particle transport and wave propagation, which occur at the same velocity, may both be dependent on the same process. The possibility that some form of contractility provides the motive force for filament flection and particle transport is discussed.     

In vitro models of tail contraction and cytoplasmic streaming in amoeboid cells

We have developed a reconstituted gel-sol and contractile model system that mimics the structure and dynamics found at the ectoplasm/endoplasm interface in the tails of many amoeboid cells. We tested the role of gel-sol transformations of the actin-based cytoskeleton in the regulation of contraction and in the generation of endoplasm from ectoplasm. In a model system with fully phosphorylated myosin II, we demonstrated that either decreasing the actin filament length distribution or decreasing the extent of actin filament cross-linking initiated both a weakening of the gel strength and contraction. However, streaming of the solated gel components occurred only under conditions where the length distribution of actin was decreased, causing a self-destruct process of continued solation and contraction of the gel. These results offer significant support that gel strength plays an important role in the regulation of actin/myosin II-based contractions of the tail cortex in many amoeboid cells as defined by the solation-contraction coupling hypothesis (Taylor, D. L., and M. Fechheimer. 1982. Phil. Trans. Soc. Lond. B. 299:185-197). The competing processes of solation and contraction of the gel would appear to be mutually exclusive. However, it is the temporal-spatial balance of the rate and extent of two stages of solation, coupled to contraction, that can explain the conversion of gelled ectoplasm in the tail to a solated endoplasm within the same small volume, generation of a force for the retraction of tails, maintenance of cell polarity, and creation of a positive hydrostatic pressure to push against the newly formed endoplasm. The mechanism of solation-contraction of cortical cytoplasm may be a general component of the normal movement of a variety of amoeboid cells and may also be a component of other contractile events such as cytokinesis.     

Filaments associated with the endoplasmic reticulum in the streaming cytoplasm of Chara corallina

Perfused Chara cells capable of resuming ATP-dependent cytoplasmic streaming in low free Ca++ solutions have been examined by electron microscopy for myosin-like filaments. Filaments 44 nm in diameter and up to 3 micron in length have been found associated with the endoplasmic reticulum that along with mitochondria, microbodies and dictyosomes from the endoplasm becomes immobilised around the sub-cortical actin bundles when ATP is depleted. Such endoplasmic filaments have not been detected in association with mitochondria or microbodies and they have not been found in the stationary cortex. These filaments are extracted from the perfused cell by ATP unless motility-inhibiting levels of cytochalasin B are present. The filaments are not detectable in cells inactivated in solutions containing high (10(-4) M) Ca++ concentrations even when the Ca++ level is subsequently lowered. Consistent with their being required for motility, cytoplasmic streaming cannot be effeiciently reactivated by ATP in such filament-depleted cells. The possibility is discussed that the filaments contain myosin and that the endoplasmic reticulum with which they are associated has a major role in generating and transmitting the motive force for streaming.