This Week in The Journal
- Teresa Esch, Ph.D.
Neuroligin Knockout Affects Synapse Function, Not Number
Soham Chanda, W. Dylan Hale, Bo Zhang, Marius Wernig, and Thomas C. Südhof
(see pages 6816–6836)
Neuroligins are postsynaptic adhesion molecules that bind to presynaptic neurexins to enhance the function of excitatory and inhibitory synapses. The most widely expressed neuroligins in the brain are NL1, which is expressed at excitatory synapses, NL2, found primarily at inhibitory synapses, and NL3, expressed at both inhibitory and excitatory synapses. Whether these proteins participate in synaptic assembly has been debated. Expression of neuroligins in non-neuronal cells induces the formation of presynaptic specializations in contacting axons, and in cultured neurons, overexpression of neuroligins increases, whereas knockdown reduces, synaptic density. But constitutive knockout of NL1, NL2, and NL3 has no effect on synapse number in vivo. The reason for this discrepancy is unclear.
To explore the role of neuroligins further, Chanda et al. knocked out NL1, NL2, and NL3—individually or in combination—in cultured hippocampal neurons. Knocking out all three neuroligins reduced the amplitude and frequency of miniature and evoked EPSCs and IPSCs recorded two weeks later. Importantly, these effects were similar regardless of how many neurons lacked neuroligins, indicating the effects were cell-autonomous. Neither the number of presynaptic terminals nor the presynaptic release probability was affected by neuroligin knockout, suggesting reduced PSC frequency resulted from loss of postsynaptic function. The number of dendritic spines and the density of GluA2-containing AMPA receptors were unaffected by loss of neuroligins, however. Instead, GluA2 puncta were smaller and less often colocalized with presynaptic markers in neurons lacking neuroligins.
Consistent with the expression of NL1 at excitatory synapses, knocking out this neuroligin reduced the amplitude of EPSCs but not IPSCs, whereas knocking out NL2, primarily expressed at inhibitory synapses, selectively affected IPSC amplitudes. Interestingly, mPSC frequency was not affected in single or double knockouts, as it was in triple knockouts. And surprisingly, NL3 knockout had no significant effect on excitatory or inhibitory transmission and did not alter the effects of NL1 or NL2 knockout. Nevertheless, overexpressing NL3 in triple-neuroligin-knockout neurons rescued the reduction of AMPA-evoked currents.
These results suggest that neuroligins are not required for the formation of dendritic spines or presynaptic specializations. Instead, neuroligins may help ensure that a high density of postsynaptic receptors is properly aligned with presynaptic release sites. The contribution of NL3 to synaptic function remains unclear.
Tweety Homolog Helps Tumor Cells Invade Brain Tissue
Erik Jung, Matthias Osswald, Jonas Blaes, Benedikt Wiestler, Felix Sahm, et al.
(see pages 6837–6850)
Glioblastomas are the deadliest form of malignant brain tumor: more than 90% of patients die within five years of diagnosis. Glioblastomas are extremely difficult to eliminate because they are diffuse, making complete resection impossible. Moreover, the tumors are highly resistant to radiation and chemotherapy. The tenacity of glioblastoma cells has been linked to their formation of long, thin protrusions called tumor microtubes. These processes extend into surrounding neuropil, facilitating tissue invasion. Furthermore, gap junctions form between microtubes of neighboring tumor cells, and this is thought to allow dissipation of toxic insults, helping the cells tolerate irradiation and chemical treatment. Consistent with these proposed functions, oligodendrocytomas, which rarely form microtubes, are less invasive and more susceptible to treatment than glioblastomas. Therefore, preventing tumor microtube formation might increase the effectiveness of radiation and chemotherapy in glioblastoma patients.
Human glioblastoma cells (green) transplanted into mouse brain extend long processes called tumor microtubes (arrowheads). Red shows blood vessels. See Jung, Osswald, et al. for details.
To identify proteins involved in tumor microtube formation, Jung, Osswald, et al. compared gene expression in human glioblastoma cells cultured in conditions that promoted or limited invasiveness and microtube formation when the cells were implanted into mouse brains. One of the most strongly upregulated genes in microtube-forming cells was Tweety-homolog 1 (Ttyh1), a chloride channel linked to neuritogenesis and filopodial protrusion in neurons. Ttyh1 expression in tumor cells was particularly concentrated at the growth-cone-like tip of extending microtubes. Intriguingly, Ttyh1 levels were higher in tumor cells that extended 1–2 microtubes, which were highly mobile and invasive, than in cells that extended >4 microtubes, which were more stationary. Moreover, knocking down Ttyh1 caused microtube beading and reduced the number of cells extending 1–2 microtubes. Although tumor cells extending >4 microtubes were generally unaffected by Ttyh1 knockdown, knockdown reduced the invasiveness of transplanted glioblastoma cells in vivo, and thus prolonged survival in mice.
These results suggest that Ttyh1 is required for extension of tumor microtubes that invade brain tissue and facilitate colonization, and they provide evidence that impeding microtube extension can limit tumor growth and extend survival. Tthy1 does not appear necessary to maintain the microtubes that interconnect tumor cells and contribute to radiation resistance, however. Whether invading and interconnecting tumor microtubes are fundamentally different or represent different developmental stages should be explored in future work.
