Venkatesh et al. provide invaluable insight into HGG, revealing not only a greater mechanistic understanding of the regulation of glioma growth, but also a potential therapeutic target in neuroligin-3. Their observations that neuronal activity promotes the proliferation of multiple glioma types and that neuroligin-3 is mutated in a variety of different types of cancers, combined with recent studies implicating autonomic innervation with cancer progression in other systems (Magnon et al., 2013; Zhao et al., 2014), suggest that this mechanism may be broadly applicable to many cancers.

REFERENCES Buckingham, S.C., Campbell, S.L., Haas, B.R., Montana, V., Robel, S., Ogunrinu, T., and Sontheimer, H. (2011). Nat. Med. 17, 1269–1274. Cuddapah, V.A., Robel, S., Watkins, S., and Sontheimer, H. (2014). Nat. Rev. Neurosci. 15, 455–465. Gibson, E.M., Purger, D., Mount, C.W., Goldstein, A.K., Lin, G.L., Wood, L.S., Inema, I., Miller, S.E., Bieri, G., Zuchero, J.B., et al. (2014). Science 344, 1252304. Khan, T., Akhtar, W., Wotton, C.J., Hart, Y., Turner, M.R., and Goldacre, M.J. (2011). J. Neurol. Neurosurg. Psychiatry 82, 1041–1045. Magnon, C., Hall, S.J., Lin, J., Xue, X., Gerber, L., Freedland, S.J., and Frenette, P.S. (2013). Science 341, 1236361.

Peixoto, R.T., Kunz, P.A., Kwon, H., Mabb, A.M., Sabatini, B.L., Philpot, B.D., and Ehlers, M.D. (2012). Neuron 76, 396–409. Su¨dhof, T.C. (2008). Nature 455, 903–911. Suzuki, K., Hayashi, Y., Nakahara, S., Kumazaki, H., Prox, J., Horiuchi, K., Zeng, M., Tanimura, S., Nishiyama, Y., Osawa, S., et al. (2012). Neuron 76, 410–422. Venkatesh, H.S., Johung, T.B., Caretti, V., Noll, A., Tang, Y., Nagaraja, S., Gibson, E.M., Mount, C.W., Polepalli, J., Mitra, S.S., et al. (2015). Cell 161, this issue, 803–816. Zhao, C.-M., Hayakawa, Y., Kodama, Y., Muthupalani, S., Westphalen, C.B., Andersen, G.T., Flatberg, A., Johannessen, H., Friedman, R.A., Renz, B.W., et al. (2014). Sci. Transl. Med. 6, 250ra115.

Rods Feed Cones to Keep them Alive Jacek Krol1,* and Botond Roska1,2,* 1Neural

Circuit Laboratories, Friedrich Miescher Institute for Biomedical Research, 4058 Basel, Switzerland of Medicine, University of Basel, 4056 Basel, Switzerland *Correspondence: [email protected] (J.K.), [email protected] (B.R.) http://dx.doi.org/10.1016/j.cell.2015.04.031 2Faculty

Cone photoreceptors, responsible for high-resolution and color vision, progressively degenerate following the death of rod photoreceptors in the blinding disease retinitis pigmentosa. Aı¨t-Ali et al. describe a molecular mechanism by which RdCVF, a factor normally released by rods, controls glucose entry into cones, enhancing their survival. The retina is a highly sophisticated biological computer that captures an image with its photoreceptors and extracts different visual features to describe the visual scene to higher brain centers in simple and compact terms. Although photoreceptors, the rods and cones, are only two out of the sixty retinal cell types, they are exceptionally important: all image-forming vision depends on their proper function. Despite the fact that rods outnumber cones 20 to 1, human vision is mostly based on cones. Rods are distributed at the periphery of the retina and are the photosensors for low light levels. Cones are concentrated in the center of the retina and work at higher light levels. Since cones are necessary for the high-resolution color vision that enables us to read, recognize faces, and enjoy the colorful world, in the modern world

we surround ourselves with enough light to turn on the cones. Most of us spend little time in conditions where photons are scarce and, therefore, our dependence on rod function is minor. A study presented in this issue of Cell offers key insight into the interdependence of rods and cones, and how it is disrupted in the genetic disorder retinitis pigmentosa (Aı¨t-Ali et al., 2015) The genes involved in retinitis pigmentosa are primarily expressed only in rods and are important for their function (Hartong et al., 2006). If humans rely mostly on cone vision, why is this disease so severe? The reason stems from the fact that rods and cones are dependent on each other. When rods are dysfunctional but alive, as in another genetic disease called stationary night blindness, cones are functional. Indeed, patients with station-

706 Cell 161, May 7, 2015 ª2015 Elsevier Inc.

ary night blindness are capable of living an almost normal life. However, when rods die, as happens in retinitis pigmentosa, cones sense this loss and react to it. This reaction is devastating. First, cones lose their outer segments, which serve as light detectors, causing patients to become blind. Second, on a longer timescale, the other parts of the cones progressively degenerate. Due to the importance of cones for human vision, and their dependence on rods, two fundamental questions in retinitis pigmentosa research are why and how do cones react to rod death and how can we prevent cones from degenerating?. There have been several important insights in recent years. One of these insights, originating in Jose´-Alain Sahelʼs laboratory, came from the logic that if rods are necessary for cone

survival, rods may release a One of the most important factor that enhances cone implications of the identificasurvival (Mohand-Said et al., tion of the RdCVF receptor 1998). Indeed, such a Basigin-1 and its binding molecule, named rod-derived partner GLUT1 is the potencone viability factor (RdCVF) tial for developing small molhas been identified by Thierry ecules that could activate Le´veillard and Jose´-Alain Sathem and, as a consehel (Le´veillard et al., 2004). It quence, slow down cone has been shown that, after degeneration in patients. rods die, the resulting loss One may wonder why researchers are focused on of RdCVF production contribprotecting cones, and not utes to cone degeneration, on preventing the death of and that externally supplied rods? There are a number of RdCVF slows down this proreasons. First, since lack of cess (Byrne et al., 2015; Le´vfunction in rods causes few eillard and Sahel, 2010). symptoms, patients often However, the RdCVF recepvisit ophthalmologists when tor in cones has been uncones start to be affected. known, and the mode of acBy this time, however, many tion for protecting cones of the rods have already deremained unclear. generated. Second, rods The present study by the should start to be protected Le´veillard group (Aı¨t-Ali et al., before the disease starts. 2015) identify an RdCVF reHowever, the onset of the ceptor, Basigin-1, and prodisease, even if the affected pose a mechanism, namely Figure 1. Rods Regulate Glucose Entry into Cones members of a family can be an increase in glucose transIn normal retinas, rod photoreceptors secrete rod-derived cone viability factor determined early, is often port via GLUT1 and a con(RdCVF), which is necessary for cone photoreceptor survival. RdCVF binds to Basigin-1, which through the glucose transporter GLUT1, regulates glucose not tractable, complicating comitant increase in aerobic uptake by cones. When rods are lost, the resulting lack of RdCVF leads to cone the design of clinical trials. glycolysis, that could be starvation, which in turn leads to cone degeneration. Despite these problems, responsible for the protection promising new ways of proof cones (Figure 1). The authors identify and verify Basigin-1 as the re- impairs RdCVF-mediated glucose uptake. tecting both rods and cones are ceptor of RdCVF for its trophic function in How does glucose supply improve cone emerging (Byrne et al., 2015). In summary, together with exciting cones using numerous experimental ap- survival? Aı¨t-Ali et al. observe that cones proaches both in vitro and in vivo. exposed to RdCVF have increased intra- new gene therapy approaches to impact After identifying the receptor, Aı¨t-Ali et al. cellular ATP concentrations and propose oxidative stress (Xiong et al., 2015), hissearch for the mechanism leading to that ATP is produced in an unusual form tone deacetylases (Chen and Cepko, enhanced cone survival. Using co-immu- of aerobic glycolysis, in which glucose is 2009), and RdCVF (Byrne et al., 2015; noprecipitation followed by mass spec- converted to lactate in the presence of ox- Le´veillard and Sahel, 2010), small moletrometry and fluorescence resonance ygen. This metabolic process requires cules targeting Basigin-1 or GLUT1 may energy transfer assay, they find a glucose lactate dehydrogenase activity, and its provide ways of slowing down a devastransporter, GLUT1, which interacts with inhibition abolishes RdCVF-mediated tating cause of blindness. Basigin-1. Both Basigin-1 and GLUT1 are cone survival. It has recently been shown that activaexpressed in photoreceptor inner segREFERENCES ments and are essential for increased tion of mTORC1 increases cone survival cone survival mediated by ectopic RdCVF partly by increasing glucose uptake (Ven- Aı¨t-Ali, N., Fridlich, R., Millet-Puel, G., Cle´rin, E., administration. Aı¨t-Ali et al. point out that katesh et al., 2015), suggesting that Delalande, F., Jaillard, C., Blond, F., Perrocheau, cones are highly sensitive to glucose accelerating glucose entry into the cell is L., Reichman, S., Byrne, L.C., et al. (2015). Cell deprivation, suggesting that a glucose a convergence point for different path- 161, this issue, 817–832. uptake-related pathway may underlie ways, such as RdCVF and mTOR, which Byrne, L.C., Dalkara, D., Luna, G., Fisher, S.K., the ability of RdCVF to preserve protect cones. Thus, starvation appears Cle´rin, E., Sahel, J.-A., Le´veillard, T., and Flannery, cones. Consistently, using a non- to be a major contributor to cone degen- J.G. (2015). J. Clin. Invest. 125, 105–116. metabolized glucose analog, the authors eration in retinitis pigmentosa (Punzo Chen, B., and Cepko, C.L. (2009). Science 323, showed that exposure to RdCVF et al., 2009), and feeding cones emerges 256–259. increased glucose entry into cones. Deple- as a central theme to assist in protecting Hartong, D.T., Berson, E.L., and Dryja, T.P. (2006). tion of Basigin-1 and GLUT1 significantly them. Lancet 368, 1795–1809. Cell 161, May 7, 2015 ª2015 Elsevier Inc. 707

Le´veillard, T., and Sahel, J.-A. (2010). Sci. Transl. Med. 2, 26ps16. Le´veillard, T., Mohand-Saı¨d, S., Lorentz, O., Hicks, D., Fintz, A.-C., Cle´rin, E., Simonutti, M., Forster, V., Cavusoglu, N., Chalmel, F., et al. (2004). Nat. Genet. 36, 755–759.

Mohand-Said, S., Deudon-Combe, A., Hicks, D., Simonutti, M., Forster, V., Fintz, A.C., Le´veillard, T., Dreyfus, H., and Sahel, J.A. (1998). Proc. Natl. Acad. Sci. USA 95, 8357–8362. Punzo, C., Kornacker, K., and Cepko, C.L. (2009). Nat. Neurosci. 12, 44–52.

Venkatesh, A., Ma, S., Le, Y.Z., Hall, M.N., Ru¨egg, M.A., and Punzo, C. (2015). J. Clin. Invest. 125, 1446–1458. Xiong, W., MacColl Garfinkel, A.E., Li, Y., Benowitz, L.I., and Cepko, C.L. (2015). J. Clin. Invest. 125, 1433–1445.

Three Cell Fusions during Double Fertilization Stefanie Sprunck1 and Thomas Dresselhaus1,* 1Cell Biology and Plant Biochemistry, Biochemie-Zentrum Regensburg, University of Regensburg, 93053 Regensburg, Germany *Correspondence: [email protected] http://dx.doi.org/10.1016/j.cell.2015.04.032

Fertilization of both egg and central cell is a major distinguishing feature of flowering plants. Now, Maruyama et al. report a third cell fusion event between the persistent synergid and the fertilized central cell shortly after double fertilization in Arabidopsis. This causes rapid dilution of pollen tube attractant(s), preventing polytubey. Almost 120 years ago, Sergei Gavrilovich Navashin (1898) and Le´on Guignard (1899) described independently for the first time that two fertilization events occur in lily, a major model plant at that time. The universality of this observation was confirmed in numerous flowering plant species (angiosperms) and is now widely considered as a major feature distinguishing angiosperms from all other organisms. During the double-fertilization event, one sperm cell fuses with the egg cell, forming the embryo, and a second sperm cell fertilizes the central cell, which develops into the endosperm. This sounds simple, but fertilization in angiosperms is a very complex process: the two genetically identical and immobile sperm cells are transported via the pollen tube over long distances (e.g., up to 30 cm in maize) through the maternal tissues of the flower in order to deliver them to the ovule. Many hurdles have to be taken before the pollen tube finally arrives at the embryo sac harboring the two female gametes, egg and central cell, as well as a number of accessory cells, including two synergids (Figure 1). The synergids are known as gland cells playing a leading role in pollen tube attraction and sperm release (for review, see Dresselhaus and Franklin-Tong, 2013). In species like the model plant

Arabidopsis, usually only one pollen tube arrives at the embryo sac and communicates with the synergids until the tube tip bursts simultaneously with the first synergid, termed receptive synergid. A block to polytubey (arrival of excess pollen tubes) is established soon after fertilization and minimizes the risk of polyspermy (fusion of a female gamete with multiple sperms). However, plants are capable of attracting multiple pollen tubes—for example, when gamete fusion fails—to maximize reproductive success. In Arabidopsis, the second synergid, persistent synergid, was shown to be responsible for polytubey in the case of fertilization failure and continues to attract pollen tubes until it degenerates (Beale et al., 2012; Kasahara et al., 2012). It was further indicated that successful double fertilization induces a block to polytubey and thus avoids the delivery of additional sperm cells to the embryo sac. But how is this block to polytubey established? Nature has an astonishingly simple solution for this problem, which is now reported in this issue of Cell by Maruyama et al. (2015): the persistent synergid fuses with the huge fertilized central cell (about 20 times larger volume, which even quickly increases after fertilization), and thereby pollen tube attractants are rapidly diluted. This peculiar

708 Cell 161, May 7, 2015 ª2015 Elsevier Inc.

phenomenon was named as synergidendosperm fusion (SE fusion; Figure 1). Using various fluorescent markers to label the cytosol, mitochondria, and endoplasmic reticulum, Maruyama et al. show by time-lapse imaging the mixing of persistent synergid and endosperm cytoplasm about 5 hr after fertilization, when the fertilized central cell or primary endosperm nucleus starts to divide. They further show fusion of plasma membranes of both cells, which was never observed in unfertilized ovules. Even more important and significant are the experiments in which they investigate the quick dilution of the pollen tube attractant AtLURE1 (Takeuchi and Higashiyama, 2012). AtLURE1 signals quickly decrease in the degenerated receptive synergid after sperm release but remain high in the persistent synergid. A rapid decrease of AtLURE1 signals was observed to coincide with the measurements of the dilution of cytoplasmic components. The attractant disappears almost completely within 36 min after initiation of SE fusion, a time point when the primary endosperm nucleus divides. The induction of SE fusion and thus rapid dilution of AtLURE1 into the early developing endosperm is sensed by fertilization success of the central cell, but not by