Drosophila CRYPTOCHROME (CRY) is a blue light sensitive protein that conveys photic signals to the circadian clock (Rosato et al., 2001; Busza et al., 2004). The strong hypomorphic mutation cry_b causes aberrant synchronization to light (Emery et al., 1998; Stanewsky et al., 1998), while flies overexpressing cry show increased responsiveness to photic stimuli (Ishikawa et al., 1999; Emery et al., 2000). This suggests that CRY modulates the light-dependent regulation of circadian function. The current model of the clock highlights the direct intervention of CRY on the molecular constituents of the circadian system. Evidence has shown that light generates a conformational change in CRY (Ozturk et al., 2011), enabling it to interact with the core circadian protein TIMELESS (TIM, Ceriani et al., 1999). This event triggers the intervention of kinases and E3-ubiquitin ligases. Thus, TIM is phosphorylated, ubiquitinated and degraded by the proteasome (Naidoo et al., 1999; Peschel et al., 2009), explaining its daily oscillations that are in phase with the light-dark (LD) cycle (Hunter-Ensor et al., 1996). Moreover, CRY directly modulates the firing of neurons and influences the circadian system through processes that are independent from the core components of the clock. These involve the redox-sensor function of the voltage-gated K⁺ channel β-subunit (Kvβ) HYPERKINETIC (HK) and additional signaling mechanisms not yet described (Dissel et al., 2014; Fogle et al., 2011, 2015). In line with this findings, CRY accumulates in the projections of neurons where it is expressed (Klarsfeld et al., 2004), it binds to components of the phototransduction pathway in the retinal photoreceptors (Mazzotta et al., 2013) and is involved in magnetoreception (Gegear et al., 2008; Ritz et al., 2010; Fedele et al., 2014a,b; Bazalova et al., 2016). Moreover, CRY has an essential role in circadian plasticity in the lamina: in fact, in cry-null (cry⁰) mutants the cyclic expression of genes regulating circadian changes in morphology of neurons and synapses is altered (Górska-Andrzejak et al., 2013; Damulewicz et al., 2015).

In our previous work we looked at the rhythmic plasticity of the synapses in the visual system by examining the expression of BRUCHPILOT (BRP) under light/dark (LD 12:12) and constant dark (DD) conditions (Górska-Andrzejak et al., 2013). Using mutants we revealed that the expression of BRP in the distal lamina is under control of both the circadian clock and the light-dark cycle, and that CRY possibly exerts an additional control. In fact, cry-null flies showed a reduction in BRP levels at night that was not found in mutants affecting vision (norp^A7) or the clock (per⁰¹, tim⁰¹). BRP is a prominent constituent of the T-bar, and shows homology in its N-terminus to the mammalian active zone proteins ELKS/CAST/ERC (Kittel et al., 2006; Wagh et al., 2006; Fouquet et al., 2009; Hida and Ohtsuka, 2010). In higher Diptera, the T-bar is an electron dense specialization of the presynaptic active zone, which is the site of neurotransmitter release (Wichmann and Sigrist, 2010). BRP is found as two isoforms of 170 kDa and 190 kDa, respectively. They differ in their N-terminal start but their specific functions are not precisely described. Null mutants for brp (brp⁶⁹) do not produce viable adults but some larvae escape lethality. In those, the model synapse at the larval neuromuscular junction reveals defective active zone membranes, a complete loss of presynaptic specializations and decreased vesicle release (Kittel et al., 2006).

In this study we investigate further the involvement of CRY in the regulation of synaptic plasticity in the visual system, in particular in the lamina, the first optic neuropil.

The results obtained in the present study reveal that CRY forms a complex with the presynaptic scaffolding protein BRP and that it may be involved in the mechanism of BRP degradation in the distal lamina, where the majority of synapses constitute tetrad synapses (Meinertzhagen and O’Neil, 1991).

The discovery of CRY was triggered by its ability to modulate the stability of other proteins. The first mutant cry^b was identified through the loss of rhythmic expression of PER and TIM in peripheral tissues and a dampening of those rhythms in the circadian neurons in the brain (Stanewsky et al., 1998). Subsequently, it emerged that CRY binds to TIM under light (Ceriani et al., 1999) and that such interaction triggers the light-dependent degradation of both proteins (Peschel et al., 2009). This cell-autonomous model of light-induced degradation has dominated the understanding of the circadian function of CRY. However, additional evidence has accumulated showing that CRY must be able to interact with other proteins and to regulate cross-talk among neurons. For instance, antibody staining revealed that CRY accumulates in neuronal projections (Klarsfeld et al., 2004), immune-precipitation of CRY followed by mass spectrometry characterization of co-precipitating proteins has identified components of the phototransduction pathway such as RDGA, INAD and NINAC (Mazzotta et al., 2013), genetic and electrophysiological experiments have associated the voltage-gated Kvβ HK with CRY (Fogle et al., 2015). Moreover, we have shown that CRY affects the expression of BRP, the main constituent of a presynaptic specialization known as the T-bar, in the distal lamina (Górska-Andrzejak et al., 2013).

The lamina is a fantastic model to study synaptic plasticity caused by light input and by the influence of the circadian clock, as synapses and neurons change in size and morphology according to the light regime and endogenous timing (Pyza and Meinertzhagen, 1999; Pyza and Górska-Andrzejak, 2004; Woźnicka et al., 2015).

In the distal lamina BRP levels peak at the beginning of the day (ZT1) and at the beginning of the night (ZT13), which corresponds to an increase in both number and size of tetrad synapses (Górska-Andrzejak et al., 2013; Woźnicka et al., 2015). Interestingly, these changes are paralleled by a rhythm of swelling (morning and evening) and shrinking (middle of the day, middle of the night) of the L1 and L2 monopolar cells (Pyza and Meinertzhagen, 1999).

In the distal lamina of cry⁰ mutants BRP levels are reduced during the night, whereas during the day BRP expression is constantly high (Górska-Andrzejak et al., 2013). In this study we have used several approaches to show that the interaction between cry and brp is not only genetic and that the two proteins physically interact. We have pulled-down BRP and identified endogenous CRY as a co-precipitating protein by Western blot (Figure 1A). Moreover, when we pulled-down overexpressed HA-CRY and probed for the endogenous BRP, we observed that both the 170 kDa and the 190 kDa BRP isoforms co-precipitated with HA-CRY (Figure 1B). We interpret this finding as an indirect confirmation of the weak co-immunoprecipitation results obtained with brp^∆190 and brp^∆170 mutants that express only the 170 kDa and the 190 kDa BRP isoform, respectively (Matkovic et al., 2013). In a Y2H assay we verified that CRY and BRP directly bind to each other and that light promotes this interaction (Figure 1C). This result is in agreement with the co-immunoprecipitation experiment presented in Figure 1B, showing that more BRP co-precipitated together with HA-CRY under light conditions than in darkness. We note that although the Y2H experiment suggests a direct interaction between CRY and BRP, additional proteins might be involved in vivo to stabilize the complex and/or to initiate a signaling cascade.

Using PLA with anti-BRP and anti-CRY antibodies we discovered PLA-positive signal in the retina, in the lamina and in the medulla of Canton S but not of cry⁰¹ flies, showing that the interaction between CRY and BRP occurs in vivo and in situ (Figure 2). The pattern of fluorescence is in agreement with the known distribution of CRY (BRP is found in all photoreceptors and neurons) but it was surprisingly sparse. We interpret this result as a consequence of the complexity of the technique. One possibility is that the PLA-positive signal is limited to those areas where CRY and BRP are expressed at the highest level. It is unlikely that within tissue sections the primary and the secondary antibodies will always bind with the right steric arrangement to allow the best interaction between the antibodies-bound PLA probes and the connectors. Another possibility is that 20 μm sections are quite difficult to penetrate by enzymes such as ligase and rolling circle DNA polymerase. Hence signal may be prevalent in areas that are more exposed and/or have a “looser” structure. Although we did not investigate these possibilities, it is likely that both aspects played a role in determining the tissue distribution of the PLA-signals we revealed.

To test whether BRP may be targeted for degradation following its interaction with CRY, we turned to CRY∆, a C-terminal deletion of CRY. We have previously shown in a Y2H assay that CRY∆ binds to TIM independently from light, and that the overexpression of CRY∆ in flies results in phenotypes suggesting that this form of the protein does not require activation by light (Rosato et al., 2001; Dissel et al., 2004). For instance, in CRY∆ flies TIM is expressed at lower levels than in wild type, which agrees with the idea of a constitutive interaction with an active CRY. Moreover, CRY∆ does not accumulate or marginally so; this is also expected, considering that the interaction between TIM and CRY drives the degradation of both proteins (Peschel et al., 2009). Thus we hypothesized that cry > CRY∆ flies would show reduced immunostaining for BRP at each time point. Indeed, that is what we observed in the distal lamina in comparison to both Canton S and cry⁰¹ flies (Figure 3). This suggests that BRP, like TIM, is targeted for degradation following the formation of a complex with CRY. Although the overall BRP immune-signal was reduced, we could still detect a significant oscillation in BRP levels. At its peak at ZT13, the BRP signal was about twice the size than for the other time points (Figure 3C). Interestingly, this expression profile mimics the temporal distribution of BRP immune-signal in wild type flies maintained under constant darkness (compare the expression profiles of BRP in Figure 3C to Figure 2B of Górska-Andrzejak et al., 2013). This is in spite of the fact that constitutive active CRY, which would arguably simulate constant light, might be expected to produce a non-rhythmic phenotype. When we measured BRP levels in whole heads we observed the reduction of BRP in both cry⁰¹ and cry > CRY∆ in comparing with Canton S; however, these differences were not statistically significant due to a remarkable variability among experiments (Figure 4B). Our interpretation is that although CRY does have an effect on the expression of BRP in general, the mechanisms are complex and possibly tissue/brain area specific. We also note that CRY is not uniformly expressed across the brain or the head. Thus, we might have detected a mixture of direct and indirect effects, and arguably the latter might amplify noise. Thus, we conclude that CRY affects BRP expression in the distal lamina, likely regulating its stability. In addition, we propose that the lamina is a particularly attractive model to investigate the mode of action of CRY.

Finally we reasoned that such a reduction in BRP expression in the distal lamina of cry > CRY∆ flies (Figure 3), should give rise to behavioral phenotypes. The optomotor response measures the ability to detect and respond to a moving environment. When the environment moves it generates an apparent self-motion to which a spectator responds with movement to stabilize their apparent course. The optomotor response depends to a large extent on the time of day, with best performances observed between the end of the day and the middle of the night and it reflects the presence of a functional circadian clock in the photoreceptor system (Barth et al., 2010; Mazzotta et al., 2013; Mazzotta and Costa, 2016). In order to study the motion vision of flies, we analyzed their optomotor walking response. As previously reported (Mazzotta et al., 2013) Canton S flies performed better at night (ZT13, 16, 20) than during the day (ZT1, 4, 8) with 70% and 60% correct choices, respectively (Figure 5).

Light adaptation in the retina depends on horizontal migration of screening pigment granules towards the rhabdomeres (Nilsson and Ro, 1994). In Musca domestica screening pigment granules migrate also vertically in photoreceptors, with maximal accumulation in the proximal part of the lamina at the end of day and higher pigment number in the distal lamina at the end of night. This pattern is clock-dependent (Pyza and Meinertzhagen, 1997). Because the absence of screening pigment causes loss of visual acuity (Burnet et al., 1968), the daily changes in optomotor walking response may be correlated with the pattern of pigment granules migration. However, BRP levels peak at ZT1 and ZT13 in the distal lamina of Canton S (Figure 3A), and we would have expected a similar behavioral outcome at these two time points. Thus, the optomotor response did not reflect the daily differences in number and size of tetrad synapses of which BRP levels in the distal lamina are a proxy. Nevertheless, we could see significant differences between the performances of control and of cry > CRY∆ flies (Figure 5). The latter genotype showed 40%–50% of correct choices during the whole day, which is the value expected by chance alone. Thus, these flies either could not detect the movement of the stripes or they were unable to process the information, or may be both. Again we did not see a correlation between the expression profile of BRP and the optomotor response. In an earlier study on the housefly, we found that motor stimulation is more effective than visual stimulation in eliciting morphological changes in the lamina. Thus the lack of correlation between BRP levels and optomotor response is not surprising (Kula and Pyza, 2007). This behavioral assay tests the functioning of the visual system from photoreceptors to higher order motion vision processing neurons in the lobula plate, a region in the optic lobes that is the final destination of visual information (Heisenberg and Wolf, 1984). Our results are in agreement with the previous observation where flies in which CRY lacked its C-terminus tail (cry^M, Busza et al., 2004) showed a reduced performance in the optomotor response (Mazzotta et al., 2013). In addition, flies with brp expression silenced in the visual system (gmr > brpRNAi) showed changes in the electroretinogram (ERG; Wagh et al., 2006), which measures extracellular activity of photoreceptors and interneurons in response to light (Heisenberg, 1971). Lack of BRP in the photoreceptor terminals causes severe defects in synaptic transmission which is visualized in ERG, since light-induced depolarization of photoreceptors is normal but ON/OFF transients originating from the interneurons are absent (Wagh et al., 2006). We speculate that the presence of a constitutively active form of CRY (due to the absence of its regulatory C-terminus) likely impacts on the organization of the visual system beyond the photoreceptors and the lamina and may have similar effect on the retina functionality as brp silencing. This consideration calls for additional studies to dissect the role of CRY in the visual system of Drosophila.

In conclusion, we have identified a physical interaction between CRY and BRP, the main constituent of the presynaptic active zone T bar. We have confirmed this interaction using different techniques such as immune-precipitation, Y2H and PLA. Our data suggest that CRY and BRP can interact directly especially under light, but they do not preclude that CRY and BRP might be part of a larger complex in vivo. We have evidence that CRY regulates the stability of BRP in the distal lamina, possibly mirroring what is known for the CRY-TIM interaction. However, the regulation of BRP in the head is more complex, since not all synapses peak in number at the same time during day and night. Finally, we have presented data showing that the functionality of the visual system is compromised in cry > CRY∆ flies, probably affecting higher order processing neurons. This suggests that the role exerted by CRY on the development and physiology of the visual system is greater than currently appreciated, with important consequences for the interpretation of the effects of CRY on the light entrainment of the circadian clock.

MD, GMM and ES carried out experiments. MD prepared figures. MD, GMM and ER, RC and EMP discussed results and reviewed the manuscript. MD and ER wrote the main manuscript text.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.