Phosphorylation of the Retinal Ribbon Synapse Specific t-SNARE Protein Syntaxin3B Is Regulated by Light via a Ca2 +-Dependent Pathway

Abstract

Neurotransmitter release at retinal ribbon-style synapses utilizes a specialized t-SNARE protein called syntaxin3B (STX3B). In contrast to other syntaxins, STX3 proteins can be phosphorylated in vitro at T14 by Ca^2+/calmodulin-dependent protein kinase II (CaMKII). This modification has the potential to modulate SNARE complex formation required for neurotransmitter release in an activity-dependent manner. To determine the extent to which T14 phosphorylation occurs in vivo in the mammalian retina and characterize the pathway responsible for the in vivo phosphorylation of T14, we utilized quantitative immunofluorescence to measure the levels of STX3 and STX3 phosphorylated at T14 (pSTX3) in the synaptic terminals of mouse retinal photoreceptors and rod bipolar cells (RBCs). Results demonstrate that STX3B phosphorylation at T14 is light-regulated and dependent upon the elevation of intraterminal Ca²⁺. In rod photoreceptor terminals, the ratio of pSTX3 to STX3 was significantly higher in dark-adapted mice, when rods are active, than in light-exposed mice. By contrast, in RBC terminals, the ratio of pSTX3 to STX3 was higher in light-exposed mice, when these terminals are active, than in dark-adapted mice. These results were recapitulated in the isolated eyecup preparation, but only when Ca²⁺ was included in the external medium. In the absence of external Ca²⁺, pSTX3 levels remained low regardless of light/dark exposure. Using the isolated RBC preparation, we next showed that elevation of intraterminal Ca²⁺ alone was sufficient to increase STX3 phosphorylation at T14. Furthermore, both the non-specific kinase inhibitor staurosporine and the selective CaMKII inhibitor AIP inhibited the Ca²⁺-dependent increase in the pSTX3/STX3 ratio in isolated RBC terminals, while in parallel experiments, AIP suppressed RBC depolarization-evoked exocytosis, measured using membrane capacitance measurements. Our data support a novel, illumination-regulated modulation of retinal ribbon-style synapse function in which activity-dependent Ca²⁺ entry drives the phosphorylation of STX3B at T14 by CaMKII, which in turn, modulates the ability to form SNARE complexes required for exocytosis.

Keywords: SNARE; bipolar cell; exocytosis; modulation; retina; ribbon synapse; syntaxin.

FIGURE 1

In rod terminals, phosphorylation of STX3 at T14 is greater in dark-adapted mice than in light-exposed mice. (A) Top: A representative set of confocal images of a vertical section through the outer plexiform layer (OPL) of the retina of a dark-adapted mouse. Arrows mark the presumptive rod terminals in the image. Note that many of STX3-labeled rod terminals are also pSTX3 positive. Double arrowheads mark the arrestin-positive cone terminals. Bottom: A representative set of confocal images through the OPL of retina from a dark-adapted mouse exposed to a 15-min light stimulus. Note the relatively weak pSTX3 immunoreactivity. For upper and lower panels, retinal sections were triple-labeled for cone-arrestin (blue) to mark the cone terminals, STX3 (red), which labels both rod and cone terminals, and pSTX3 (green). Scale bar = 5 μm. (B) Comparison of the mean pixel intensity of STX3 immunolabeling in rod terminals revealed no light/dark difference. By contrast, the mean pixel intensity of pSTX3 in rod terminals was significantly diminished in the light when compared to the dark (p = 0.0006). (C) Quantification of the pSTX3 to STX3 ratio in rod terminals demonstrates that in the dark-adapted mouse (blue bars) the mean ratio of pSTX3 to STX3 is significantly greater than in the light-treated mouse (white bars). p < 0.0001. (D) There was no light/dark difference in the mean pixel intensity of STX3 immunolabeling in cone terminals and no clear light/dark difference in pSTX3 immunolabeling. (E) Quantification of the pSTX3 to STX3 ratio in cone terminals indicated a mean pSTX3 to STX3 ratio that is not significantly different between lighting conditions. p = 0.465. For (B–E), n = 6,5 (mice). ***Indicates p < 0.001 and ****indicates p < 0.0001.

FIGURE 2

In rod bipolar cell terminals, phosphorylation of STX3 at T14 is greater in light-exposed mice than dark-adapted mice. (A) Top: A representative set of confocal images of a vertical section through the inner plexiform layer (IPL) of a dark-adapted mouse retina. Arrows mark some of the rod bipolar cell synaptic terminals. Bottom: A representative set of confocal images through the IPL of a retina from a dark-adapted mouse exposed to a 15-min light stimulus. Note the obvious double-labeling for pSTX3 and pSTX3 in rod bipolar cell terminals. For all, retinal sections were doubled-labeled for STX3 (red), and pSTX3 (green). The small bracket labeled “RBCT” indicates the region of the IPL where the rod bipolar cell terminals (RBCT) reside. Scale bar = 10 μm. (B) Quantification of the rod bipolar cell terminal mean pixel intensities demonstrates that only the mean intensity of pSTX3 immunolabeling is light-sensitive p = (0.0415). Quantification of the pSTX3 to STX3 ratio in rod bipolar cells terminals demonstrates that in the light-treated mouse (white bars) the ratio of pSTX3 to STX3 is significantly greater in the dark-adapted mouse (blue bars). p < 0.0001. For (B,C), n = 4,4 (mice). *Indicates p < 0.05 and ****indicates p < 0.0001.

FIGURE 3

External Ca²⁺ is required for STX3 phosphorylation at T14 in both rod and cone photoreceptor terminals. (A) Representative confocal images through the OPL show that phosphorylation of STX3 at T14 in photoreceptors terminals in dark-adapted retinal eyecups (top panel) is abolished by the omission of Ca²⁺ from the external bath solution (“O Ca,” bottom panel). Scale bar = 5 μm. (B) In rod spherules, the pSTX3/STX3 ratio in dark-adapted retinal eyecups was greatly reduced in the absence of external Ca²⁺ (p < 0.0001) and indistinguishable from that of light-treated eyecups, with or without external Ca²⁺; p = 0.6035 and 0.6928, respectively. (C) In cone pedicles, the high pSTX3/STX3 ratio in dark-adapted eyecups was reduced in the absence of external Ca²⁺ to levels indistinguishable from that of light-treated eyecups, with or without external Ca²⁺; p = 0.3016; p = 0.998, respectively. For (B,C), n = 4,4 (mice). **indicates p < 0.01, ****Indicates p < 0.0001, and ^†indicates that p approaches significance.

FIGURE 4

External Ca²⁺ is required for STX3 phosphorylation at T14 in rod bipolar cell synaptic terminals. (A,B) A pair of representative confocal images show that phosphorylation of STX3 at T14 in rod bipolar cell synaptic terminals in light-treated retinal eyecups is reduced by the omission of Ca²⁺ from the external bath solution (“O Ca”). The small bracket labeled “RBCT” indicates the region of the IPL where the rod bipolar cell terminals reside. Scale bar = 5 μm. (B) In rod terminals, the pSTX3/STX3 ratio in dark-adapted retinal eyecups was greatly reduced in the absence of external Ca²⁺ (p = 0.0019) and indistinguishable from that of dark-adapted eyecups, with or without external Ca²⁺; p = 0.9043 and 0.2382, respectively. n = 2,2 (mice). **Indicates p < 0.01.

FIGURE 5

Elevation of intraterminal Ca²⁺ promotes phosphorylation of STX3B at T14 in rod bipolar cell synaptic terminals. (A) Confocal images of synaptic terminals of isolated rod bipolar cells, identified by their characteristic morphology and PKC immunoreactivity (green), demonstrate an increase in pSTX3 immunoreactivity (red) under conditions of elevated intraterminal Ca²⁺ (“high Ca²⁺,” top panel) when compared to those incubated on a nominally 0 Ca²⁺ external solution (“low Ca²⁺,” bottom panel). Scale bar = 5 μm. (B) The mean pixel intensity of STX3 immunolabeling was not altered by changes in intraterminal Ca²⁺ (p = 0.7238). By contrast, elevation of intraterminal Ca²⁺ led to an increase in pSTX3 immunolabeling (p = 0.0077). (C) Quantification of the pSTX3/STX3 ratio in rod bipolar cell terminals indicates that elevated intraterminal Ca²⁺ (“High”) significantly increased this ratio above that of cells bathed in nominally O Ca²⁺ (“Low”). p < 0.0001. For (B,C), n = 7,7 (mice). (D) Elevated intraterminal Ca²⁺ (“High”) also increased CaMKII activation, as evidenced by an increase in pCaMKII immunolabeling in synaptic terminals of rod bipolar cells relative to terminals bathed in nominally 0 Ca²⁺ external solution (“Low”). (p = 0.0283; n = 6,6 mice). *Indicates p < 0.05, **indicates p < 0.01, and ***indicates p < 0.001.

FIGURE 6

Kinase inhibition prevents the Ca²⁺-dependent phosphorylation of STX3B at T14 in rod bipolar cell synaptic terminals. (A) Confocal images of synaptic terminals of isolated rod bipolar cells, identified by their characteristic morphology and PKC immunoreactivity (green), demonstrate an increase in pSTX3 immunoreactivity (red) under conditions of elevated intraterminal Ca²⁺ (“High Ca²⁺”) when compared to those incubated in a nominally 0 Ca²⁺ external solution (“Low Ca²⁺”). Treatment with staurosporine (100 nM; lower panel) blocked the expected Ca²⁺-evoked increase in pSTX3 immunoreactivity. Scale bar = 5 μm. (B) Quantification of the pSTX3/STX3 ratio in rod bipolar cell terminals shows that elevated intraterminal Ca²⁺ (“High”) significantly increased the pSTX3/STX3 ratio above that of cells bathed in external solution containing no added Ca²⁺ (“Low”; p < 0.0001, n = 4,4 mice). Staurosporine not only blocked this Ca²⁺-evoked rise in the pSTX3/STX3 ratio (p < 0.0001, n = 4,4 mice), it lowered the ratio of the low Ca²⁺ control group (p = 0.0366, n = 4,3 mice). (C) In parallel experiments, treatment with staurosporine suppressed the Ca²⁺-stimulated increase in pCaMKII immunolabeling (p < 0.0001, n = 3,3 mice). *Indicates p < 0.05, ***indicates p < 0.001, and ****indicates p < 0.0001.

FIGURE 7

CaMKII phosphorylates STX3B at T14 in rod bipolar cell synaptic terminals. (A) Confocal images of synaptic terminals of isolated rod bipolar cells, identified by their characteristic morphology and PKC immunoreactivity (green), demonstrate an increase in pSTX3 immunoreactivity (red) under conditions of elevated intraterminal Ca²⁺ (“High Ca²⁺”) compared to those incubated in a nominally 0 Ca²⁺ external solution (“Low Ca²⁺”). Treatment with myristoylated AIP (“mAIP,” 100 nM, and lower panel) blocked the expected Ca²⁺-evoked increase in pSTX3 immunoreactivity. Scale bar = 5 μm. (B,C) Quantification of the pSTX3/STX3 ratio in rod bipolar cell terminals. AIP inhibited the Ca²⁺-evoked rise in the pSTX3/ST3 ratio (p = 0.0171 n = 7,7) without altering the Ca²⁺-evoked rise in pCaMKII immunoreactivity (p = 0.9791, n = 3,3 mice).

FIGURE 8

CaMKII inhibition suppresses exocytosis at rod bipolar cell terminals. (A) Total exocytosis, measured after ten depolarizing stimuli (−65 to −20 mV, 100 ms duration, 4 Hz; Wan et al., 2010), was significantly smaller when AIP (25 μM) was included in the internal recording solution than under control conditions (p = 0.0059). (B) The mean total I_Ca, summed over the pulse train, was not significantly different in the absence or presence of AIP (p = 0.2948). For A, B, control: n = 6 cells/4 mice; AIP: n = 5 cells/3 mice.