The ABP CAP1 is a multidomain protein with largely unknown physiological functions. Here we report CAP1 expression throughout postnatal brain development and enrichment in dendritic spine heads. By STED microscopy and live cell imaging in hippocampal neurons from gene-targeted mice, we identified CAP1 as a novel postsynaptic actin regulator relevant for spine density and morphology. Mechanistically, we found the conserved HFD to be essential for CAP1’s function in spines and for CAP1’s interaction with the key synaptic actin regulator cofilin1. Rescue experiments in dKO lacking CAP1 and cofilin1 revealed that both ABP not only cooperated in regulating spine morphology, but demonstrated their mutual functional dependence in the postsynaptic compartment.
The yeast CAP ortholog has been recognized as an ABP three decades ago (for review: Ono, 2013; Rust, 2020; Rust, 2022), but significant progress in its molecular functions has been achieved only recently, predominantly by exploiting recombinant proteins or mutant yeast strains (Jansen, 2014; Johnston, 2015; Kotila, 2018; Kotila, 2019; Shekhar, 2019). In a current model of actin regulation, CAP interacts with cofilin-decorated F-actin via its HFD and thereby accelerates dissociation of the terminal actin subunit (Kotila et al., 2018, Kotila et al., 2019). Subsequently, G-actin-cofilin complexes are passed on to CAP’s CARP domain that, together with its WH2 domain, releases cofilin from this complex and promotes nucleotide (ATP for ADP) exchange on G-actin, which is required for actin polymerization. While these studies provided exciting novel insights into CAP’s molecular activities, the cellular and physiological functions of mammalian CAP largely remained unknown, also because appropriate animal models were lacking. This held true specifically for CAP1, while earlier mouse studies implicated CAP2 in heart physiology and skeletal muscle development (Peche, 2012; Field, 2015; Kepser, 2019; Colpan, 2021). Recently, a TALEN-engineered systemic CAP1-KO mouse model has been generated, but these mice died during embryonic development and analysis of heterozygous mutants solely revealed a CAP1 role in lipoprotein metabolism (Jang, 2019). Moreover, we recently reported impaired neuron connectivity in brain-specific CAP1-KO mice, which was likely caused by compromised growth cone function and delayed neuron differentiation (Schneider, 2021a; Schneider, 2021c). Instead, the function of CAP1 in differentiated neurons or synapses has not been studied to date.
In this study, we demonstrated important functions for CAP1 in postsynaptic actin regulation, in very good agreement with the in vitro studies outlined above. Specifically, FRAP revealed a role for CAP1 in actin turnover and STED nanoscopy in F-actin organization. Interestingly, actin defects were associated with an altered sub-spinous distribution of the PSD proteins PSD-95, Shank3 and Homer, thereby suggesting a role for CAP1 in shaping the postsynaptic machinery, which will be investigated in future studies. While we here report a role for CAP1 in synaptic actin regulation, we earlier showed its relevance for actin dynamics in growth cones (Schneider, 2021a; Schneider, 2021c). We therefore conclude a general requirement for CAP1 in neuronal actin regulation both during differentiation and in differentiated neurons, similar to cofilin1 (Bellenchi, 2007; Gomez, 2014; Omotade, 2017; Rust, 2015a; Rust, 2010; Schneider, 2021b). By PLA we found colocalization (within 40 nm) of CAP1 and cofilin1 in spines, and our coimmunoprecipitation experiments validated their physical interaction in the hippocampus, which required CAP1’s HFD. By rescue experiments in dKO neurons lacking CAP1 and cofilin1, we demonstrated a cooperation of both ABP in regulating spine morphology. Moreover, we showed mutual functional dependence of CAP1 and cofilin1 in spines, similar to their functional interdependence in growth cones (Schneider, 2021a). Hence, an intimate interaction of CAP1 and cofilin1 is likely of general relevance for neuronal actin regulation.
While this study reports an important synaptic function for CAP1, studies of the past two decades recognized cofilin1 as an important regulator of synapse physiology, brain function and behavior (Fukazawa, 2003; Zhou, 2004; Hotulainen, 2009; Rust, 2010; Gu, 2010; Bosch, 2014; Rust, 2015b; Rust, 2015c). In line with these publications, we report an increased spine density and volume in cofilin1-KO neurons. Collectively, these studies let cofilin1 emerge as a key regulator of spine morphology and as a major final point of signaling output for actin regulation in spines (Spence, 2015). This emphasized the necessity for regulatory mechanisms that tightly control cofilin1 activity to ensure proper synapse physiology and brain function. In fact, dysregulation of cofilin1 activity has been linked to synaptic and behavioral deficits associated with ASD or ADHD (Duffney, 2015; Zimmermann, 2015). To date, a plethora of signaling molecules ranging from the Rho GTPases Rac1, Cdc42 and RhoA and their effectors PAK1, ROCK and LIMK1 to the phosphatase calcineurin and its effectors PI3K and slingshot have been implicated in synaptic cofilin1 phosphorylation that controls actin binding (for review: Ben Zablah, 2020; Rust, 2015b; Spence, 2015). Additionally, synaptic cofilin1 activity is regulated by molecules that control its recruitment into spines and by translation within the dendritic compartment (Feuge, 2019; Pelucchi, 2020; Pontrello, 2012). By demonstrating mutual functional dependence of cofilin1 and CAP1 in spines, we report a conceptually novel mechanism of synaptic cofilin1 regulation. Further, our finding that CAP1 was essential for cofilin1 activity in spines opened up a new avenue for the modulation of cofilin1 activity and, hence, actin dynamics in spines. Notably, CAP1 comprises several conserved domains allowing interaction with molecules others than actin and cofilin1. To date, a number of interaction partners have been found for CAP1 or its homologs (for review: Kakurina, 2018; Ono, 2013; Rust, 2020, Rust 2022), including established regulators of spine morphology such as the ABP profilin (Ackermann, 2003; Lamprecht, 2006; Michaelsen-Preusse, 2016; Michaelsen, 2010; Sungur, 2022), the proteinase MMP-9 (Tian, 2007; Wang, 2008), the tyrosine kinases Abl1 and Abl2 (Lin, 2013; Ma, 2014; Omar, 2017), focal adhesion kinase (Moeller, 2006; Shi, 2009) and glycogen synthase kinase 3 (Ochs, 2015; Peineau, 2007). Normalization of spine parameters in CAP1-KO neurons upon expression of a CAP1 variant with a mutated proline-rich motif (CAP1-P1) suggested that its proline-rich domain and, hence, its interaction with profilin, were not relevant in spines. Nevertheless, it will be exciting to test in future studies whether other proteins interact with CAP1 in spines and how these proteins control CAP1-cofilin1 interaction and synaptic actin dynamics.
Apart from CAP1, mammals express a second family member, CAP2, with restricted expression pattern and abundance in striated muscles and brain (Rust, 2020; Rust, 2022). Similar to CAP1, CAP2 is expressed in the postnatal brain and located in spine heads from cortical and hippocampal neurons (Kumar, 2016; Pelucchi, 2020). However, while studies in mutant mice established important CAP2 functions in heart physiology and skeletal muscle development (Peche, 2012; Field, 2015; Kepser, 2019; Colpan, 2021), neuronal CAP2 functions are less clear. Increased spine density has been reported for cerebral cortex neurons lacking CAP2, but this study unfortunately lacked a detailed spine morphometric analysis and did not provide mechanistic insights (Kumar, 2016). Conversely, spine density was unchanged upon shRNA-mediated CAP2 knockdown (CAP2-KD) in hippocampal neurons, and spine length and width were both slightly increased (Pelucchi, 2020). However, compared to roughly 30% increased spine size in CAP1-KO neurons, the effect of CAP2-KD on spine morphology was rather mild, suggesting that CAP1 is the key family member in spines. Interestingly, this study revealed CAP2-mediated cofilin1 recruitment into spines upon induction of long-term potentiation (LTP), which depended on disulfide bond-mediated CAP2 dimerization (Pelucchi, 2020). Remarkably, CAP2-dependent recruitment of cofilin1 into spines was required for LTP-triggered spine remodeling and potentiation of synaptic transmission (Pelucchi, 2020). Although spine changes in CAP2-KD neurons suggested a role for CAP2 in synaptic actin dynamics, this has not been directly tested in this study. Moreover, it remained unknown whether CAP2 and cofilin1 cooperated in regulating spine morphology in basal conditions and whether CAP2 was essential for synaptic cofilin1 activity as both shown in the present study for CAP1. Nevertheless, a model in which both CAP1 and CAP2 cooperate with cofilin1 in synaptic actin dynamics and spine morphology is very appealing, and it will be exciting in future studies to dissect CAP1- vs CAP2-specific mechanisms and to test whether or not CAP1 and CAP2 cooperate and/or are functionally redundant in spines.
In summary, we here identified CAP1 as an essential novel actin regulator in excitatory synapses that is relevant for organization and dynamics of postsynaptic F-actin and thereby controls spine density and morphology. Mechanistically, our data revealed mutual functional dependence of CAP1 and cofilin1 in spine morphology, thereby unravelling a novel synaptic actin regulatory mechanism. Our data let us hypothesize that CAP1 is equally important as cofilin1 for brain function and behavior, and that CAP1 dysregulation may contribute to the pathologies of neuropsychiatric disorders as it has been shown for cofilin1 (Duffney, 2015; Zimmermann, 2015).