A PDZ scaffolding/CaM‐mediated pathway in Cryptochrome signaling

Abstract

Cryptochromes are cardinal constituents of the circadian clock, which orchestrates daily physiological rhythms in living organisms. A growing body of evidence points to their participation in pathways that have not traditionally been associated with circadian clock regulation, implying that cryptochromes may be subject to modulation by multiple signaling mechanisms. In this study, we demonstrate that human CRY2 (hCRY2) forms a complex with the large, modular scaffolding protein known as Multi‐PDZ Domain Protein 1 (MUPP1). This interaction is facilitated by the calcium‐binding protein Calmodulin (CaM) in a calcium‐dependent manner. Our findings suggest a novel cooperative mechanism for the regulation of mammalian cryptochromes, mediated by calcium ions (Ca²⁺) and CaM. We propose that this Ca²⁺/CaM‐mediated signaling pathway may be an evolutionarily conserved mechanism that has been maintained from Drosophila to mammals, most likely in relation to its potential role in the broader context of cryptochrome function and regulation. Further, the understanding of cryptochrome interactions with other proteins and signaling pathways could lead to a better definition of its role within the intricate network of molecular interactions that govern circadian rhythms.

1. INTRODUCTION

Cryptochromes are flavin‐containing blue light photoreceptors related to photolyases, DNA‐repair enzymes that use blue light to repair UV‐induced DNA damage. Cryptochromes have been described in various animal lineages, including insects, fish, amphibians, and mammals (Deppisch et al., 2022). They have lost the DNA repair activity and have been recruited by the circadian machinery, either as components of the mechanism generating the circadian oscillation and controlling the daily behavioral/physiological rhythms or as photoreceptors mediating the entrainment of the circadian clock to light (Michael et al., 2017). Cryptochromes have acquired an intrinsically unstructured C‐terminal extension that participates, to varying degrees in their circadian functions. In Drosophila, Cryptochrome (CRY) has a well‐established role in mediating the response to light and regulating light‐dependent interactions with target proteins (Busza et al., 2004; Dissel et al., ²⁰⁰⁴; Hemsley et al., ²⁰⁰⁷; Ozturk et al., ²⁰¹¹; Rosato et al., ²⁰⁰¹), while in vertebrate‐like cryptochromes its function is less defined, although several pieces of evidence support the hypothesis that it might be implicated in protein stability, ultimately affecting circadian rhythmicity (Gao et al., 2013; Harada et al., ²⁰⁰⁵). Drosophila CRY, exhibits a wide panoply of functions (reviewed in Damulewicz & Mazzotta, ²⁰²⁰): it acts as the primary blue light photopigment that mediates circadian responses to light (Dolezelova et al., 2007; Emery et al., ^{1998, 2000}; Fogle et al., ^{2011, 2015}; Rieger et al., ²⁰⁰⁶; Stanewsky et al., ¹⁹⁹⁸; Yoshii et al., ²⁰⁰⁴) and evidence is accumulating on its role in light‐dependent magneto‐sensitive responses (Bradlaugh et al., 2023; Fedele et al., ²⁰¹⁴; Gegear et al., ²⁰⁰⁸; Marley et al., ²⁰¹⁴; Yoshii et al., ²⁰⁰⁹). Moreover, CRY appears to be an integral component of the molecular clock in peripheral tissues, including the compound eyes (Collins et al., 2006; Ivanchenko et al., ²⁰⁰¹; Krishnan et al., ²⁰⁰¹), where we have previously shown it to play a role independent on light activation (Schlichting et al., 2018). Vertebrate cryptochromes have apparently lost the ability to sense light and have been recruited as light‐independent negative autoregulators of the circadian clock (Hirayama et al., 2003; Kiyohara et al., ²⁰⁰⁶; McCarthy et al., ²⁰⁰⁹; Sato et al., ²⁰⁰⁶; Shearman et al., ²⁰⁰⁰; van der Horst et al., ¹⁹⁹⁹). There is evidence to suggest the involvement of mammalian CRYs in several additional signaling pathway, to include acting as second messengers between the core clock and other cellular processes, such as maintenance of cellular and genomic integrity and metabolism (Kang et al., 2010; Kang & Leem, ²⁰¹⁴; Lamia et al., ²⁰¹¹; Narasimamurthy et al., ²⁰¹²; Papp et al., ²⁰¹⁵). Nevertheless, the nature of the transduction signaling involving CRYs remains largely unknown. We have previously shown that dCRY acts through Inactivation No Afterpotential D (INAD) in a light‐dependent manner on the Signalplex, a multiprotein complex that includes visual‐signaling molecules (Mazzotta et al., 2013). Furthermore, we have identified and characterized a Calmodulin (CaM) binding motif in the dCRY C‐terminus (Mazzotta et al., 2018). Similarly, we also defined the CaM binding site of the scaffold protein INAD and demonstrated that CaM bridges dCRY and INAD to form a ternary complex in vivo, suggesting a process in which a rapid dCRY light response stimulates an interaction with INAD, that can be consolidated by a novel mechanism regulated by CaM (Mazzotta et al., 2018). The intracellular signaling machinery is often organized around scaffolding proteins localized at the plasma membrane, and multiple PDZ proteins bind to the various constituents of the transduction pathway, bringing them into close proximity and precisely defined stoichiometry, ultimately ensuring a rapid and specific signal transduction (Manjunath et al., 2018). Based on the hypothesis that evolution has conserved in mammals a mechanism similar to the one known for Drosophila, we searched for common components between mammals and insects, to shed light on the mechanisms involved in mammalian clock entrainment. Thus, firstly we performed bioinformatics analyses to identify putative scaffolding proteins able to interact with mammalian CRYs. We then performed in vitro validations to confirm the interaction between the molecular components suggested to be binding partners.

2. RESULTS

2.1. MUPP1 identification and its interaction with hCRY2

In a previous study, we demonstrated that an INAD fragment, including the PDZ2, its upstream region, and PDZ3 domains (residues 207–448) is crucial for its association with dCRY (Mazzotta et al., 2013). To identify an INAD functionally related protein in mammals, we performed a pBlast search against the UniProt database. Initially, the PDZ3 of INAD was employed as the query in our search, focusing on mammalian proteins which led to the identification of multiple MUPP1 orthologue proteins. Subsequently, this search was further refined by employing the same query against the SwissProt database resulting in the isolation of human proteins. Among the outcomes, the highest‐scoring match (Max score: 106, e‐value 3e‐26, Percent Identity: 31.69%) corresponded to the human MUPP1 (UniProt ID: O75970). MUPP1 is a 13 PDZ tandem domains containing protein differentially expressed in several mouse tissues, such as brain, skeletal muscle, heart, liver kidney, and lung (Sitek et al., 2003). It is particularly abundant in the brain, where it localizes at the junctions of epithelial and endothelial cells and at the neuronal synapses (Becamel et al., 2001; Ernkvist et al., ²⁰⁰⁹; Krapivinsky et al., ²⁰⁰⁴; Won et al., ²⁰¹⁷; Zhang et al., ²⁰¹¹; Zhu et al., ²⁰¹⁶). Research in humans and animal models has uncovered several essential neuronal functions for this large modular scaffolding protein, which has been proposed to participate in the N‐methyl‐D‐aspartic acid receptor signaling and to play a role in the regulation of synaptic plasticity (Philipp & Flockerzi, 1997). We then wondered which PDZ domain of hMUPP1 shows the highest sequence similarity with the pair formed by INAD PDZ2 and PDZ3. The best matches were obtained by aligning INAD‐PDZ2 against hMUPP1‐PDZ8 (40% similarity) and INAD‐PDZ3 against hMUPP1‐PDZ9 (30% similarity). High sequence similarity (70%) was also found between the INAD PDZ2 upstream region (residues 207–248) and the corresponding hMUPP1‐PDZ8 upstream region (Figure 1a). Furthermore, we found that both INAD and MUPP1 share a conserved sequence pattern ([DE]‐[DE]‐EDEFGY[TS][MW]‐‐I‐‐RY‐‐[ML]‐‐‐L) localized in a predicted disordered region (Figure 1a), suggesting that MUPP1 may play a role similar to that of INAD. The physical interaction between hCRY2 and MUPP1 was analyzed by a yeast two hybrid assay, in which a full‐length hCRY2, directly fused to LexA (bait), was challenged with a 257 aa fragment of hMUPP1 (aa 1307–1564_ UNIPROT O75970), comprising PDZ domains 8 and 9 and the upstream region, homolog to PDZ 2–3 of INAD. hCRY2 was observed to interact with hMUPP1 independently from light (Figure 1b).

FIGURE 1.

(a) Multiple sequence alignment of drosophilidae INAD‐PDZ2 and mammalian MPDZ‐PDZ8 upstream regions, colored using clustal color scheme. The blue bar on top highlights the predicted disordered region, while the red line on bottom marks the conserved pattern. (b) hCRY2 interacts with MUPP1 both in light and dark. Yeast two‐hybrid assay in which full‐length hCRY2 (bait) was challenged with hMUPP1_1307–1564 (prey). As negative control, full‐length hCRY2 was challenged with the empty prey vector. The mean ± SEM of seven independent clones, three replicates, are reported. ****p < 0.0001.

2.2. MUPP1 and CaM: Details of an interaction

Through bioinformatics analysis, we predicted the existence of a putative CaM binding site within the conserved motif. Specifically, this site is identified as non‐canonical and exhibits characteristics of the “1–12 motif,” characterized by the presence of two conserved bulky hydrophobic residues separated by a series of variable amino acids (https://cam.umassmed.edu). The physical interaction between MUPP1 and CaM was analyzed by using several approaches. In a yeast two‐hybrid assay, the full‐length protein (CaM_1–149) and separate domains (CaM_1–80 and CaM_75–149) were challenged with hMUPP1_1307–1564. The results show an interaction between the two proteins, which is preferentially mediated by the C‐terminal lobe of CaM (Figure 2a). To further elucidate the novel interaction between MUPP1 and CaM, we synthesized a short peptide encompassing residue 1331–1343 of MUPP1 and containing the predicted binding site (MUPP1_1331–1343). MUPP1_1331–1343 addition to CaM in the presence of 5 mM Ca(II) or 5 mM EGTA was followed using isothermal titration calorimetry (ITC). A binding event of the peptide to the protein was revealed to be Ca(II) dependent, as exothermic peaks followed each addition in the presence of the Ca(II) ion, while no heat effect was visible in the presence of EGTA (Figure 2b, top panel). Fits of the integrated heat produced a titration curve with two inflection points, which could not give a good fit with a single binding event model. Thus, a two sets of independent sites model was used (Figure 2b, bottom panel), which produced a good fit (χ ² = 456.2), entailing two binding sites per CaM monomer with affinities that differ by two orders of magnitude: K _A1 = 2.4 ± 0.1 × 10⁶ (K _B1 = 0.43 ± 0.02 μM) and K _A2 = 1.5 ± 0.1 × 10⁴ (K _B2 = 67 ± 4 μM). Both binding events are enthalpically driven (ΔH ₁ = −9.41 ± 0.01 and ΔH ₂ = −7.53 ± 0.01 kcal mol⁻¹) and presented negative entropy (ΔS ₁ = −2.41 and ΔS ₂ = −6.09 kcal mol⁻¹ K⁻¹). We used heteronuclear two‐dimensional nuclear magnetic resonance (2D NMR) to define the portion of CaM involved in the binding with MUPP1. Specifically, we titrated ¹⁵N‐labeled CaM with up to a 2.5‐fold molar excess of unlabeled MUPP1_1331–1343 and we acquired several ¹⁵N‐HSQC at different protein‐peptide ratios. The ¹⁵N‐HSQC spectrum represents a fingerprint of the protein where, as a first approximation, every peak corresponds to a residue. The peaks in the HSQC map are very sensitive to the chemical environment of the corresponding amino acid and therefore this experiment represents a useful tool to study the interaction of a protein with other molecules, at an atomic level: protein residues involved in the interaction will undergo significant changes upon binding of the molecule. To use this tool, the assignment of the ¹⁵N‐HSQC spectrum should be obtained first. The resonance assignment of CaM ¹⁵N‐HSQC had been previously achieved by a set of triple resonance NMR experiments acquired on the ¹³C, ¹⁵N‐CaM (Mazzotta et al., 2018). The ¹⁵N‐HSQC spectra of CaM in the presence of increasing amounts of MUPP1_1331–1343 are reported in Figure 2c, which shows that a large number of peaks are affected upon addition of the peptide. In the first part of the titration, a number of peaks become weaker, and completely disappear upon addition of one equivalent of peptide with respect to CaM. Further, we also compared the spectrum in presence of 0.6 equivalents of MUPP1_1331–1343 (Figure 2c, blue peaks) against the free CaM (red peaks) and concluded that the free protein is significantly decreased upon addition the MUPP1 peptide. Notably, all the signals showing this behavior correspond to the C‐terminal lobe of CaM. At the same time, new signals appear in the spectrum, and they increase in intensity until a 1:1 molar ratio is reached, after which they remain substantially unchanged (Figure 2c, yellow peaks), whereas the same peaks significantly shift upon addition of two equivalents of the peptide (cyan peaks). On the contrary, peaks from the C‐terminal domain of CaM are sharp at one equivalent of MUPP1_1331–1343 and they undergo only minor changes upon further addition of the peptide. When 2 equivalents of peptide are added, peaks become sharper while no further changes were observed over 2.5 equivalents of MUPP1_1331–1343 (Figures S1 and S2). This observation clearly indicates an interaction between CaM and MUPP1_1331–1343 characterized by a slow exchange regime in the NMR timescale and associated therefore to a strong binding between the two macromolecules. In these conditions, it is not possible to derive the assignment of the peptide‐bound CaM from that available for the free form. Anyway, the residues affected by this first binding event can be identified by monitoring the changes in intensity for the peaks of free CaM during the first part of the titration. By comparing the ¹⁵N‐HSQC spectra of CaM in the absence and in the presence of 0.6 equivalents of MUPP1_1331–1343, it is evident that all the peaks of free CaM showing a large decrease of their intensity correspond to residues of the C‐terminal lobe of the protein (Figures 2c and S3). On the contrary, peaks belonging to the N‐terminal domain show only small changes in their position or intensity at this point of the titration (Figures 2c and S3). This observation strongly suggests a preferential interaction of MUPP1 with the C‐terminal domain of CaM as indicated also by yeast two‐hybrid experiments (Figure 2a). By further increasing the amount of MUPP1_1331–1343, also many signals of the N‐terminal domain of CaM were significantly perturbed. In particular, they started shifting and becoming broad in the presence of one equivalent of peptide and then progressively got narrower and moved to a different position of the spectrum in the second part of the titration, until two equivalents of peptide were added (Figure 2c). No significant changes were observed by further increasing the amount of MUPP1_1331–1343 (Figure 2c). This behavior is consistent with a second binding site for MUPP1_1331–1343 localized on the N‐terminal domain of CaM. This second binding event is characterized by an intermediate exchange regime in the NMR timescale and therefore by a lower affinity compared with the one described above for the C‐terminal lobe. The results of these NMR experiments are fully in agreement with those independently obtained by ITC and previously described. An in vitro CaM pulldown assay was performed in the presence or absence of Ca²⁺, to investigate whether binding of hMUPP1 and CaM is Ca²⁺‐dependent. Protein extracts from HEK‐293 T cells overexpressing an MYC‐tagged form of hMUPP1 were incubated with CaM Sepharose beads and the bound proteins were analyzed by western blot. The result shows that hMUPP1 binds CaM in a Ca²⁺‐dependent manner (Figure 2d).

FIGURE 2.

(a) Yeast two‐hybrid assay in which Calmodulin (CaM; full‐length or fragments) was challenged with hMUPP1(1307–1564). As negative control, CaM (full length or fragments) was challenged with the empty prey vector, and the measured activity, considered as background, was subtracted from that of the samples. The mean ± SEM of seven independent clones, three replicates, are reported. (b) ITC measurements of the interaction between CaM and MUPP1 peptide. Raw titration data are represented in the top panel, representing the thermal effect of 57 × 5 μL injections of a solution of Mupp1 onto a solution of CaM, in the presence of 5 mM EGTA (upper black trace) or 5 mM Ca(II) (bottom blue trace). Normalized heats of reaction derived from the integration of raw data are reported in the bottom panel. The solid line represents the best fit of the data obtained using a nonlinear least‐square fitting procedure using a binding model based on two sets of binding sites. (c) Superpositions of ¹⁵N‐HSQC spectra of ¹⁵N‐labeled calmodulin in the presence of increasing amounts of unlabeled MUPP1_1331–1343 peptide. Different peptide:CaM ratios are color coded as follows: 0:1 in red, 0.6:1 in blue, 1:1 in yellow, 1.3:1 in green, 2:1 in cyan, 2.5:1 in black. Top‐left panel: ¹⁵N‐CaM spectra in the absence and in the presence of 0.6 equivalents of MUPP1_1331–1343 showing that only the C‐terminal domain is affected in the first part of the titration. Blue peaks represent the spectrum in presence of 0.6 equivalents of MUPP1_1331–1343, while red ones correspond to free CaM. Bottom‐left panel: ¹⁵N‐CaM spectra in the absence and in presence of 1 and 2 equivalents of MUPP1_1331–1343. Right panels: enlargement of different regions of the spectra comparing the signals of free CaM with those upon the addition of 1.0, 1.3, 2.0, and 2.5 equivalents of peptide, showing that signals from the C‐terminal domain are not significantly affected by the addition of more than 1 equivalent of MUPP1_1331–1343. (d) CaM pulldown assay and western blot showing the interaction between hMUPP1 and CaM. Protein extract from HEK‐293 T cells overexpressing MYC‐hMUPP1 (input) was divided equally into two parts and the appropriate reagents were added (CaCl₂ or EDTA). Each extract (containing either CaCl₂ or EDTA) was incubated with CaM agarose beads to allow binding: unbound proteins were removed and the bound proteins were detached from the beads. Membranes were probed with an anti‐MYC antibody. Different intensity of the bands is due to the different concentration of the proteins in the different samples.

2.3. CaM forms a ternary complex with hCRY2 and MUPP1

Given that the interaction between hMUPP1 and CaM seems to be mediated by the C‐terminal lobe of CaM, we wondered whether it could bridge hMUPP1 to hCRY2, as we have previously observed with the Drosophila counterparts (Mazzotta et al., 2018). In a yeast two‐hybrid assay, in which we have challenged the hCRY2 with CaM, either as full‐length protein (CaM_1–149) or as separate domains (CaM_1–80 and CaM_75–149), we observed a preferential interaction between hCRY2 and the N‐terminal lobe of CaM, which strengthens our hypothesis (Figure 3a). In other words, this result suggests that CaM may undergo a conformational change in which one of the two partners is bound first, and this interaction then leads to the formation of a ternary complex between hCRY‐CaM and MUPP1. We then performed a coimmunoprecipitation (CoIP) assay, in which protein extracts from Drosophila S2R+ cells overexpressing HA‐CRY2, MUPP1_1307–1564‐HIS, and MYC‐hCAM full‐length (input) were subjected to immunoprecipitation with anti‐HA high affinity (HA) matrix and probed in western blot with anti‐HA, anti‐HIS, and anti‐MYC. A signal in the “BOUND” sample for either anti‐HIS or anti‐MYC revealed that both hMUPP1_1307–1564 and hCaM were bound to hCRY2; however, bands in the “UNBOUND” samples for both constructs indicate that not the total amount of the proteins was involved in the binding. Overall, our results reveal that a ternary complex, hCRY2‐CaM‐MUPP1_1307–1564 is formed (Figure 3b).

FIGURE 3.

Calmodulin (CaM) forms a ternary complex with hCRY2 and MUPP1. (a) Yeast two‐hybrid assay in which full‐length hCRY2 was challenged with CaM (full length or fragments). As negative control, full‐length hCRY2 was challenged with the empty prey vector, and the measured activity, considered background, was subtracted from that of the samples. (b) Coimmunoprecipitation (CoIP) and western blot confirming the interaction between hCRY2, hMUPP1, and hCaM. Protein extract from Drosophila S2R+ cells overexpressing HA‐CRY2, hMUPP1_1307–1564‐HIS, and MYC‐hCAM (input) were incubated with HA‐agarose beads to allow binding: unbound proteins were removed, and the bound proteins were detached from the beads. Membranes were probed with anti‐HIS, anti‐MYC anti‐HA antibodies. Different intensity of the bands is due to the different concentration of the proteins in the different samples. Bands in the “UNBOUND” samples indicate that not the total amount of the proteins was involved in the binding.

3. DISCUSSION

In this study, we have identified the mammalian protein MUPP1 as having a functional connection to INAD, a protein involved in the CRY mediated light signaling to the circadian clock in Drosophila and discovered that MUPP1 has a high sequence similarity with INAD in PDZ domains 2 and 3 and the upstream region of PDZ2. MUPP1 (multi‐PDZ domain protein 1), also known as DLG3 (discs large homolog 3), is a protein that plays a crucial role in the organization and function of protein complexes at the plasma membrane (Krapivinsky et al., 2004; Sitek et al., ²⁰⁰³). It contains multiple PDZ domains, which are protein–protein interaction domains that bind to specific short amino acid sequences, and it can interact with a variety of proteins including ion channels, receptors, and cytoskeletal proteins (Becamel et al., 2001; Krapivinsky et al., ²⁰⁰⁴). PDZ domain proteins are involved in a wide range of cellular processes, including signal transduction, cell adhesion, and protein trafficking. They are found in all eukaryotes and are particularly abundant in the nervous system, where they play important roles in the regulation of synaptic signaling and plasticity. PDZ domains and CaM are two distinct protein domains that can interact with each other in various cellular processes. CaM has been shown to interact with several PDZ domain‐containing proteins, including the MAGUK family of proteins (membrane‐associated guanylate kinase) such as PSD‐95 and SAP97 (Becamel et al., 2001; Zhu et al., ²⁰¹⁶). These interactions can regulate the localization and activity of these proteins at the plasma membrane, and thereby regulate various signaling pathways. For example, the interaction between CaM and PSD‐95 (Zhang et al., 2011) can regulate the clustering of NMDA receptors at synapses, which is important for synaptic plasticity and learning and memory. Overall, the interaction between PDZ domains and CaM is an important mechanism for regulating protein–protein interactions and signaling pathways in various cellular processes. Identifying MUPP1 as functionally related to INAD offers insights into the conservation of circadian clock mechanisms across species. We explored the physical interaction between MUPP1 and hCRY2, finding that they interact in a light‐independent manner. We also demonstrated a Ca²⁺‐dependent MUPP1 interaction with CaM using biochemical and biophysical techniques. ITC showed that two independent binding sites on CaM are involved, with affinities differing by two orders of magnitude (K _D = 0.43–67 μM). Differently, INAD was found to bind CaM with a 1:1 stoichiometry and lower affinity (K _D = 164 μM). This result was fully confirmed by heteronuclear 2D NMR, which also demonstrated that the association with MUPP1 is primarily mediated by CaM C‐terminal lobe. This domain binds MUPP1 with HF and only when it is largely saturated, the interaction occurs also with the N‐terminal lobe of CaM. This observation suggests that when MUPP1 is bound to CaM through its C‐terminal domain, the N‐terminal lobe remains available for an interaction with a different protein partner. We experimentally showed that CaM forms a ternary complex with hCRY2 and MUPP1 in vitro, suggesting that CaM may serve as a bridge between the proteins (Figure 4), analogously to what we have observed for the Drosophila orthologues (Mazzotta et al., 2018). The interactions between hCRY2 and MUPP1 and the ternary complex formation with CaM imply that MUPP1 may contribute to circadian clock regulation by modulating hCRY2 activity. Identifying MUPP1 as a factor potentially involved in the regulation of the mammalian circadian clock implies that this mechanism may be conserved, despite protein differences, and that the interaction between hCRY2, CaM, and MUPP1 is part of the intricate network of intracellular signals that regulate and are regulated by the clock. Mammalian circadian clocks are primarily synchronized by light, detected by retinal photoreceptors and transmitted to the brain's master clock. However, several pieces of evidence indicate that they can be modulated by multiple other signaling pathways. The key role of Ca²⁺ signaling in the modulation of circadian clocks is established, with the calcium ion (Ca²⁺) being known to maintain molecular rhythmicity by regulating the expression of clock genes as Per1, Per2, and Bmal1 (Kon et al., 2014; Lundkvist et al., ²⁰⁰⁵; Nahm et al., ²⁰⁰⁵) and the stability and cellular localization of clock proteins such as PER2 (Jakubcakova et al., 2007). This study offers valuable insight into the interactions between proteins involved in circadian signaling and proposes a Ca²⁺/CaM‐mediated mechanism regulating mammalian CRYs (Michael et al., 2017).

FIGURE 4.

CaM serves as a bridge between hCRY2 and MUPP1. In the presence of Ca2+, MUPP1 interacts with CaM's C‐terminal lobe with high affinity. This observation suggests that the N‐terminal lobe remains available for the interaction with hCRY2, thus forming a ternary complex. Created with BioRender.com (https://www.biorender.com).

4. CONCLUSIONS

This study has identified a functional connection between the mammalian protein MUPP1 and INAD, a protein involved in clock signaling in Drosophila. The study shows that MUPP1 interacts with hCRY2 and CaM in a light‐independent manner, suggesting that MUPP1 may contribute to circadian clock activity by modulating hCRY2. Although future research is needed to examine their functional outcomes, the discovery of these interactions identifies a potential cooperative mechanism explaining how intracellular signals transmit time information to the host and, in turn, how they can be affected by the host's physiology.

5. MATERIALS AND METHODS

5.1. Bioinformatic analysis

MUPP1 was obtained through p‐BLAST search from the nonredundant database. Conservation analysis was performed using Jalview (Waterhouse et al., 2009) with T‐Coffee (Notredame et al., 2000) and a BLOSUM62 matrix. Intrinsic‐disordered regions of the hMUPP1 and INAD were predicted with Cspritz (Walsh et al., 2011). ELM (Dinkel et al., 2012) was used to identify PDZ interaction motifs in mammalian CRY2 CTTs. A sequence logo was built with Weblogo (Crooks, 2004). Protein sequences used for these analyses were retrieved from Uniprot (UniProt Consortium, 2019) and presented as follows: Uniprot ID – INAD: D. melanogaster (Q24008), Drosophila sechellia (B4I8C5), Drosophila simulans (B4QI19), Drosophila yakuba (B4P8L2), Drosophila erecta (B3NNX5), Drosophila persimilis (B4H514), Drosophila mojavensis (B4KSL1), Drosophila virilis (B4LMW1), Drosophila willistoni (B4MJM0), Drosophila grimshawi (B4J763), Drosophila pseudoobscura (Q28X50), Drosophila ananassae (B3MEK4). MUPP1: Homo sapiens (O75970), Mus musculus (Q8VBX6), Rattus norvegicus (055164), Pan troglodytes (Chimpanzee; H2R430), Bos taurus (Bovine; F1MMT3), Macaca mulatta (Rhesus macaque; F6T1W3), Equus caballus (Horse; F6S559), Ailuropoda melanoleuca (Giant panda; G1L4U6), Canis familiaris (Dog; F1PGB9), Callithrix jacchus (White‐tufted‐ear marmoset; F7FYD6), Oryctolagus cuniculus (Rabbit; G1SLR4), Mustela putorius furo (ferret; M3XYK5), Sus scrofa (Pig; F1SMP5), Myotis lucifugus (Little brown bat; G1P3E1), Felis catus (Cat; M3WF73), Otolemur garnettii (Small‐eared galago; H0WTI1), Pongo abelii (Sumatran orangutan; H2PS57). CRY2: M. musculus (Q9R194), H. sapiens (Q49AN0), R. norvegicus (Q92318), P. troglodytes (Chimpanzee; H2Q3H0), B. taurus (Bovine; F1N0J9), M. mulatta (Rhesus macaque; F7HH05), E. caballus (Horse; F6Y7Z4), A. melanoleuca (Giant panda; G1L1K6), C. familiaris (Dog; F1PNC9), C. jacchus (White‐tufted‐ear marmoset; F6RC52), O. cuniculus (Rabbit; G1THJ9), M. putorius furo (ferret; M3Y337), S. scrofa (Pig; F1SHJ9), M lucifugus (Little brown bat; G1PQN8), F. catus (Cat; M3VZW3), O. garnettii (Small‐eared galago; H0X769), P. abelii (Sumatran orangutan; H2NDL2). The presence of putative CaM binding sites was investigated with the CaM target Databases (Yap et al., 2000).

5.2. Yeast two‐hybrid assays

The experiments were performed in the EGY48 yeast strain (MATα, ura3, trp1, his3, 3LexA‐operator‐LEU). Baits were prepared by cloning the sequence of interest fused to the LexA moiety in the bait vector (pEG202), while preys contained the desired proteins fused to the “acid‐blob” portion of the prey vector (pJG4‐5; Bartel & Fields, ¹⁹⁹⁷).

The full‐length hCry2 bait was described in (Schlichting et al., 2018). The CaM constructs (bait or prey) were described in (Mazzotta et al., 2018). The hMUPP1 fragment comprising aa 1307–1620 was amplified from pCMV6_entry_MPDZ_Myc_DDK tagged (plasmid no. RC219952 from Origene_ Rockville, MD, USA) with the primers pJG_hMUPP_F (5′‐GTGCCAGATTATGCCTCTCCCGAATTCATGGAAATGGGTAGTGATCACACACAG‐3′) and pJG_hMUPP_R (5′‐CGAAGAAGTCCAAAGCTTCTCGAGCATATGTCAGCAGGTTGCAGGATCAGAAGC‐3′). The cloning was performed by using the In‐Fusion® HD Cloning Kit (Clontech, Mountain View, CA, USA) and the construct fully sequenced to assess the in‐frame insertion of the cDNA and to control for unwanted mutations. Quantification of β‐galactosidase activity was performed in liquid culture by the Miller assay, as in (Ausbel, 1998). One Miller unit corresponds to a standardized amount of β‐Gal activity. A student's t‐test was used to perform single‐group comparisons.

5.3. CoIP and western blot

The full‐length hCry2 and hCaM coding sequences and the hMUPP1 fragment comprising aa 1307–1620 were cloned into S2 expression vector pAC5.1/V5‐His (Thermo Fisher Scientific, Waltham, MA, USA). hCry2 coding sequence was amplified from pSO2002 plasmid (Addgene plasmid no. 25842,_Ozgur & Sancar, 2003) by pAc_HA_hCRY2F (5′‐CAGTGTGGTGGAATTCATGTACCCATACGATGTTCCAGATTACGCTGCGGCAACTGTGGCAACGGC‐3′), that adds a HA tag at the 5′ of the coding sequence, and pAc_HA_hCRY2R (5′‐GACTCGAGCGGCCGCTCAGGCATCCTTGCTCGGCAG‐3′). hCaM coding sequence was amplified from Universal Human Reference RNA (Thermo Fisher Scientific, Waltham, MA, USA), by pAct‐hCaM‐MYC‐F (5′ CAGTGTGGTGGAATTCATGGAACAAAAACTCATCTCAGAAGAGGATCTGGCTGATCAGCTGACCGAAGAAC‐3′) that adds an MYC tag at the 5′ of the coding sequence and pAct‐hCaM‐R (5′‐TAGACTCGAGCGGCCGCTCATTTTGCAGTCATCATCTGTAC‐3′). The hMUPP1(1307–1620) was amplified from plasmid no. RC219952 with the primers pAc_hMUPP1F (5′‐GTGCCAGATTATGCCTCTCCCGAATTCATGGAAATGGGTAGTGATCACACACAG‐3′) and pAc_hMUPP1R (5′‐GACTCGAGCGGCCGCAAGCAGGTTGCAGGATCAGAAG‐3′), in frame with the HIS‐tag present in the vector downstream the cloning site. The cloning was performed by using the In‐Fusion® HD Cloning Kit (Clontech, Mountain View, CA, USA) and the constructs fully sequenced to assess the in‐frame insertion of the cDNA and to control for unwanted mutations. Drosophila S2R+ cells, maintained at 25°C in Schneider's Drosophila medium (Thermo Fisher Scientific, Waltham, MA, USA) added with 10% FBS (Gibco_Thermo Fisher Scientific, Waltham, MA, USA) were cotransfected by using using the Effectene Transfection Reagent (Qiagen, Gaithersburg, MD, USA), following manufacturer instruction. Cell extracts were subjected to CoIP by using an anti‐HA Affinity Matrix (Roche, Basel, CH), following manufacturer instructions. SDS‐PAGE was performed as previously described in (Mazzotta et al., 2013) and immunocomplexes were analyzed using a mouse anti‐HIS (1:2000; Qiagen, Gaithersburg, MD, USA), a mouse anti‐HA (1:2000; Sigma Aldrich, St. Louis, MO, USA) and a mouse anti‐MYC (1:2000; Sigma Aldrich, St. Louis, MO, USA) as primary antibodies and an antimouse IgG HRP (1:5.000; Sigma Aldrich, St. Louis, MO, USA) as secondary antibody.

5.4. CaM pulldown assay

Human embryonic‐kidney 293 T (HEK‐293 T) cells, maintained at 37°C in Dulbecco's Modified Eagle's Medium (DMEM; Thermo Fisher Scientific, Waltham, MA, USA) added with 10% FBS (Gibco_Thermo Fisher Scientific, Waltham, MA, USA), were transfected with 2 mg of pCMV6_entry_MPDZ_Myc_DDK (hMUPP1) using Lipofectamine‐2000 (Thermo Fisher Scientific, Waltham, MA, USA) following manufacturer instruction. Briefly, DNA was precomplexed with Lipofectamine‐2000 at a ratio of 1:3, in serum‐free medium for 15 min at room temperature. The complex medium was added to the cells and incubated for 4 h at 37°C before being replaced with DMEM plus 10% FBS. Cells were collected after 24 h. Protein extraction and CaM pulldown assay were performed as described in (Mazzotta et al., 2018). CaM‐bound proteins were detached from the beads by the addition of loading buffer (LDS_Thermo Fisher Scientific, Waltham, MA, USA) and heating at 75°C for 10 min and analyzed by SDS‐PAGE on 3%–8% NuPAGE Tris‐Acetate Gels (Thermo Fisher Scientific, Waltham, MA, USA). Western blot was performed with monoclonal anti‐MYC antibody (1:2.000, Clontech, Mountain View, CA, USA) and antimouse IgG HRP secondary antibody (1:5.000; Sigma Aldrich, St. Louis, MO, USA).

5.5. CaM expression and purification

Unlabeled and ¹⁵N‐isotopically labeled samples of CaM from Xenopus laevis (with identical amino acid sequence of human CaM) were recombinantly produced in Escherichia coli BL21(DE3) cells grown in LB (unlabeled protein) or in M9 minimal medium containing 4.4 g/L glucose monohydrate and 1 g/L ¹⁵NH₄Cl (¹⁵N‐labeled protein) with 50 μg/mL kanamycin for selection. Cells were grown at 37°C and induced overnight at 20°C. Bacteria were suspended in 50 mM Tris–HCl pH 7.5, lysed and the soluble fraction was mixed with 5 mL of Phenyl FF resin (GE Healthcare, Chicago, IL, USA) in ice for 1 h and 15 min to remove highly hydrophobic proteins. The resin was centrifuged for 15 min, at 5000g, 4°C, the supernatant was filtered and CaCl2 was added at a final concentration of 5 mM before loading onto into a series of two Phenyl Sepharose HP, 5 mL columns (GE Healthcare, Chicago, IL, USA) equilibrated with 50 mM Tris–HCl buffer, 1 mM CaCl2, pH 6.5. CaM was eluted with 50 mM Tris–HCl buffer, 1 mM EGTA, pH 6.5 and further purified by means of a Superdex 75 prep‐grade 16/60 (GE Healthcare, Chicago, IL, USA) SEC column equilibrated with 50 mM Tris–HCl, 150 mM NaCl, pH 7.5. Pooled samples were concentrated at 20–30 mg/mL, and flash‐frozen in liquid nitrogen for further analysis.

5.6. Peptide synthesis

The MUPP1_1331–1343 peptide (purity >99%) was purchased from Thermo Fisher Scientific (Waltham, MA, USA). The peptide was acetylated at the N‐terminus and amidated at the C‐terminus, to mimic the protein environment and remove extra charges.

5.7. NMR experiments

All NMR experiments were performed with a Bruker DMX 600 MHz spectrometer with a room temperature probe, at 303 K. ¹⁵N‐HSQC experiments were collected with eight scans, 2048 complex data points and a spectral width of 14 ppm in the ¹H dimension, and 200 increments and a spectral width of 25 ppm in the ¹⁵N dimension. Samples of 340–380 μM uniformly ¹⁵N‐labeled CaM were titrated with peptide stock solutions (3.2 mM MUPP1_1331–1343). The protein and the peptide were dissolved in the same buffer consisting of 50 mM Tris‐Cl, 100 mM NaCl, 5 mM CaCl₂ at pH 6.5. The pH of the solutions was checked and adjusted after dissolving CaM and the peptides, to avoid pH changes during the titration. Deuterated water (10% v/v) was added to the NMR tube. Resonance assignment of the ¹⁵N‐HSQC for the Ca²⁺ loaded apo‐CaM spectrum in these experimental conditions was achieved as previously described (Mazzotta et al., 2018). Data were processed with TOPSPIN 3.1 (Bruker BioSpin GmbH, Rheinstetten, DE) and analyzed using CARA 1.9 (Keller) and NMRFAM‐Sparky (Lee et al., 2015).

5.8. Isothermal titration calorimetry

ITC experiments were performed using a high‐sensitivity VP ITC microcalorimeter (MicroCal LLC, Northampton, MA, USA). The reference cell was filled with deionized water. Protein and peptide solutions were prepared by diluting concentrated stock solutions in the reaction buffer (50 mM TrisHCl, pH 7.5, 150 mM NaCl), in the absence or in the presence of 5 mM CaCl₂ or EGTA. Each experiment started with a small injection of 1–2 μL, which was discarded from the analysis of the integrated data. Care was taken to start the first addition after baseline stability had been achieved. In each individual titration, 5 μL of a 850 μM MUPP1 peptide solution was injected into a solution of 50 μM CaM, using a computer‐controlled 310‐μL microsyringe. To allow the system to reach equilibrium, an interval of 240 s was applied between each ligand injection. Integrated heat data obtained for each titration were fitted using a nonlinear least‐squares minimization algorithm to a theoretical titration curve, using the Origin package from the manufacturer. A model involving two sets of independent sites was used. ΔH (reaction enthalpy change, cal mol − 1), K _A (binding constant, M − 1), and N (stoichiometry) were the thermodynamic fitting parameters. K _D (dissociation constant) was calculated as the inverse of the association constant. The reaction entropy was calculated using the relationships ΔG = − RTlnKa (T = 298 K) and ΔG = ΔH − TΔS.

AUTHOR CONTRIBUTIONS

Gabriella M. Mazzotta: Writing—review and editing; conceptualization; investigation; funding acquisition; writing—original draft; methodology; validation; visualization; data curation; supervision. Massimo Bellanda: Investigation; funding acquisition; writing—original draft; data curation; writing—review and editing. Milena Damulewicz: Methodology; validation; investigation; data curation. Barbara Zambelli: Investigation; methodology. Elisa Costanzi: Investigation. Francesco Gregoris: Investigation. Stefano Mammi: Formal analysis; investigation. Silvio C. E. Tosatto: Writing—original draft; conceptualization; data curation. Rodolfo Costa: Funding acquisition; formal analysis; investigation; conceptualization. Giovanni Minervini: Supervision; conceptualization; investigation; writing—original draft; methodology; validation; visualization; data curation; writing—review and editing.

FUNDING INFORMATION

This work was funded by grants from the Università degli Studi di Padova (P‐DiSC no. 01BIRD2018 to MB and BIRD213814/21 to GMM), Fondazione Cassa di Risparmio di Padova e Rovigo (Progetti di Eccellenza 2011–2012), National Research Council of Italy and Ministero dell'Istruzione, dell'Universitá e della Ricerca (MIUR; EPIGEN Flagship project—Subproject 4), INsecTIME FP7 People: Marie‐Curie Actions Initial Training Network (grant PITN‐GA‐2012‐316790) to RC.

Supporting information

Data S1 Supporting information.

Bellanda M, Damulewicz M, Zambelli B, Costanzi E, Gregoris F, Mammi S, et al. A PDZ scaffolding/CaM‐mediated pathway in Cryptochrome signaling. Protein Science. 2024;33(3):e4914. 10.1002/pro.4914