BIOCHEMICAL STUDIES OF HUMAN MICAL1, THE FLAVOENZYME CONTROLLING CYTOSKELETON DYNAMICS

MICAL from the “Molecule Interacting with CasL” indicates a family of cytoplasmic multidomain proteins conserved from insects to humans, which participates in the control of cytoskeleton dynamics. A unique feature of MICAL proteins is the presence of a catalytic N-terminal flavoprotein monoonoxygenase-like (MO) domain that is followed by several protein interaction domains, namely: a calponin homology (CH) domain, a LIM domain and a C-terminal region containing potential coiled-coil motifs. In neurons MICAL1 is an essential component for the transduction of the semaphorin signaling downstream of plexin by catalyzing a NADPH-dependent F-actin depolymerization through the N-terminal flavoprotein domain. But it may also control integrin pathway and microtubules assembly through interaction with CasL and CRMP, respectively and apoptosis through interaction with NDR1. The MICAL2 and MICAL3 isoforms are implicated in vesicles trafficking and gene transcription. The aim of this project is to contribute to define the function of human MICAL1 by characterizing the catalytic properties of the MO domain and how they are modulated by its CH, LIM and C-terminal domains and, eventually, by its interacting proteins. The human MICAL1 form containing the isolated MO domain (55.1 kDa; pI 9) and the form comprising both the MO and the CH domains (MOCH; 68.5 kDa; pI 7.7) have been produced according to the procedures available in the laboratory, which were further optimized. The full-length MICAL1 (MICAL; 119 kDa; pI, 6.2) and the form lacking the C-terminal region (MOCHLIM; 86.4 kDa; pI, 6.7) were produced in E.coli cells and their purification protocols were set-up exploiting the engineered C-terminal His6-tag. All the purified MICAL forms are stable and greater than 95% homogeneous. MICAL forms are isolated with the correct complement of FAD bound to the MO domain and zinc ions bound to the LIM domain, which predicts the formation of two zinc fingers. The absorption spectra of all MICAL forms are similar to each other, with an extinction coefficient at 458 nm of ≈8.1 mM-1cm-1 similar to that previously determined for the isolated MO domain, indicating that the CH, LIM and C-terminal regions do not alter the conformation of the catalytic domain. However, the LIM domain causes MOCHLIM to oligomerize to yield dimers, trimers and higher order aggregates, while the full-length protein yielded stable dimers as opposed to the monomeric state of MO and MOCH forms. All MICAL forms catalyze a NADPH oxidase (H2O2-producing) activity, which is associated with the MO domain. By combining steady-state kinetic measurements of the reaction as a function of pH and of solvent viscosity we concluded that the CH, LIM and C-terminal domains lead to a progressive lowering of the catalytic efficiency (MO, ≈165 s-1mM-1; MOCH, ≈18.5 s-1mM-1; MOCHLIM, ≈15 s-1mM-1; MICAL, ≈0.75 s-1mM-1) of the reaction due to an increase of Km for NADPH from ≈20 μM (MO), to ≈130 μM (MOCH), ≈230 μM (MOCHLIM) and ≈370 μM (MICAL). The 200-fold drop of the catalytic efficiency of the full-length MICAL compared to that of MO is also due to a ≈10-fold decrease of kcat. The study of the pH and viscosity dependence of the NADPH oxidase reaction of MICAL forms led us to conclude that the observed changes in the values of the kinetic parameters are not due to changes in rate determining steps of the reaction taking place within MO. The increase of KNADPH correlates with a decrease of the positive charge of the protein due to the acidity of the CH, LIM and C-terminal domains. The 10-fold drop of kcat observed with full-length MICAL is consistent with the proposal of an autoinhibitory role of the C-terminal region on MICAL catalytic activity. Our experiments reveal that is likely due to a conformational equilibrium between an inactive and an active conformation of MICAL, which is shifted 9:1 toward the inactive conformation in solution. The position of the C-terminal region is crucial to make the protein fully inactive . All MICAL forms are able to depolymerize F-actin in the presence of NADPH as observed by monitoring the decrease of fluorescence of pyrenyl-actin and of the average radius of F-actin solution by dynamic light scattering. F-actin leads to an approximately 5-10-fold increase of the maximum rate of the NADPH consumption for all MICAL forms (MO, 20 s-1; MOCH, 26 s-1; MOCHLIM, 15 s-1; MICAL, 8 s-1) compared to that measured in its absence. The presence of F-actin leads to a lowering of the Km for NADPH to similar values (11-50 μM) for all the four forms. The apparent Km for F-actin is similar (≈4 μM) for MO, MOCH and MOCHLIM, but 10-fold higher for MICAL (≈30 μM). Also this effect can beexplained by the conformational equilibrium between an inactive and an active conformation of the full-length MICAL. The actin binding site on the MO domain may be physically accessible to F-actin only when MICAL is in the active conformation, subtracting it from the inactive/active conformational equilibrium. In this case, the high apparent Km for actin of MICAL would reflect the coupled equilibria between inactive/active conformations and complex formation between actin and MICAL in the active conformation. In all cases, the amount of NADPH oxidized during the reaction exceeds that of total actin present, suggesting a case of substrate recycling or an enhancement of the NADPH oxidase activity of MICAL forms when actin is bound. This finding is in contrast with the proposal that MICAL forms catalyze a (slow) NADPH oxidase activity when isolated but a fast and specific hydroxylation of actin Met44 (and Met 47) in the presence of F-actin, which leads to filament depolymerization (Hung et al. (2013) Nat Cell Biol., 15, 1445-1454). Mass spectrometry analyses of actin samples treated with MICAL forms and NADPH support the hypothesis that actin depolymerization is mediated by the H2O2 released by the enhanced NADPH oxidase activity of MICAL in complex with actin. A maximum of two residues are oxidized per actin molecule among which are Met44 and Met47 but also several other Met and Trp residues are oxidized with similar probability. Overall it appears that MICAL in the free state can catalyze a basal NADPH oxidase reaction even in the cell. This activity is enhanced when the fraction of MICAL in the catalytically active state binds to actin leading to its depolymerization. Interaction of MICAL C-terminus with the cytoplasmic side of plexin upon semaphorin signaling would shift the equilibrium of inactive/active conformations toward the active form favoring the interaction with actin in proximity of plexin and its local depolymerization. With purified MICAL forms now available, it will possible to identify its interactors and to study their effect on its activities. As a first step in this direction we produced and purified a CRMP1 form that was found to cooperate with MICAL1 in determining COS7 cells collapse in response to semaphorin signaling. We demonstrated that CRMP1 and MICAL1 MO and MOCH interact, but in a complex way: at low ionic strength the interaction, which is largely non specific due to the strong opposite charges of the proteins, leads to inhibition of the MO and MOCH NADPH oxidase activity; at higher ionic strength, where non specific interactions are removed, CRMP stimulates the NADPH oxidase activity of MO and MOCH and competes with actin for binding MICAL, but the effect is mild. These preliminary experiments support the hypothesis that actin and microtubules dynamics may be linked through MICAL-CRMP interaction. More importantly, they show the feasibility of detailed in vitro studies for the identification of MICAL interacting proteins and their mutual modulatory effect. This approach, along with the detailed mechanistic studies of MICAL reactions thanks to the availability of various MICAL forms, will also open the way to the identification of molecules able to inhibit MICAL that could be beneficial to combat diseases in which MICAL has been implicated such as neurodegeneration, cancer and even vasculo- and cardiogenesis.