| Abstract | The serine/threonine protein kinase, ‘Ca2+-calmodulin-dependent protein kinase kinase 2’ (CaMKK2) regulates gene expression and protein activity in response to calcium (Ca2+)-stimuli in brain neurons and thus modulates neuronal function and development, late long-term potentiation, memory formation and cognition (Greenberg et al., 1986; Kang et al., 2001; Llinás et al., 1981; Lynch et al., 1983; Malenka et al., 1989; Mizuno et al., 2007; Morgan & Curran, 1986). Additionally, CaMKK2 has been implicated in the regulation of metabolism (Stork et al., 2022) and emotional behaviours. Thus Camkk2-/- mice display increased locomotor activity, anxiety-like behaviour, and depressive-like behaviour, as well as altered sensorimotor gating and an increased freezing response in the cued fear conditioning paradigm (Scott et al., 2015).
Our laboratory has recently demonstrated a novel mechanism of CaMKK2 activation, whereby the fatty acid metabolite, palmitoyl-coenzyme A (P-CoA), activates CaMKK2 in vitro (Dr John W. Scott, unpublished data). This concentration-dependent activation is abolished in the p.G551L variant of the enzyme. To investigate the role of P-CoA-mediated CaMKK2 activity in vivo, the equivalent mouse model (G551L) was generated and characterised using behavioural assays described in the first results Chapter (Chapter 3). The main finding of this study was that male, but not female, G551L mice lost more body weight during a 24 h fasting period than wildtype (WT) control mice, indicating that P-CoA-mediated CaMKK2 activity might play a role in the regulation of metabolism. Additionally, minor alterations in sensorimotor gating were observed, while G551L mice displayed normal anxiety-and depressive-like behaviours, suggesting that these behaviours are regulated independently of P-CoA-mediated CaMKK2 activity.
Chapters 4−6 describe studies to investigate the role of CaMKK2 in the actions of the mood stabiliser lithium in vivo and in vitro. Lithium is the front-line treatment for bipolar disorder, a lifelong mental illness which requires chronic treatment (American Psychiatric Association, 2013; Goodwin et al., 2016; Yatham et al., 2013). However, lithium has a narrow therapeutic window and long-term use is associated with severe side effects including nephrogenic diabetes, tremor, and gastrointestinal symptoms (Bendz & Aurell, 1999; Lapierre, 1976; Malhi, 2015; McKnight et al., 2012; Schou et al., 1970; Wood et al., 1994). This illustrates the clinical need for improved treatment options with reduced adverse effects. However, while lithium’s mechanistic actions remain somewhat unknown, one identified target is the enzyme, glycogen synthase kinase 3β (GSK3β), which is inhibited by lithium in vitro and in vivo (Beaulieu et al., 2004; Gould et al., 2004; Hong et al., 1997; Sarno et al., 2002) and activated by methamphetamine (B. Xing et al., 2015). Additionally, GSK3β is a negative regulator of CaMKK2 in vitro as it reduces its Ca2+-calmodulin (Ca2+-CaM)-independent activity by phosphorylating the S129, S133, and S137 amino acid residues (Dr John W. Scott, unpublished data).
A major goal of the research described in this thesis was to assess the role of this interaction in vivo. Therefore, mice carrying mutations that mimic the effects of constant GSK3β activity on CaMKK2 and thus block the effects of lithium on GSK3 that are mediated by CaMKK2 have been generated (p.S129D, S133D, S137D, ‘S3D mice’). While short-term lithium treatment did not depend on Ca2+-CaM-independent CaMKK2 activity to exhibit mood-stabilising effects after methamphetamine challenge, long-term lithium treatment altered some behavioural and biochemical responses which depended on normal Ca2+-CaM-independent CaMKK2 activity. Thus, male, but not female S3D mice displayed increased baseline activity in response to lithium treatment compared to WT control mice. In contrast, sex-independent antidepressant effects of lithium in the forced-swim test did not depend on Ca2+-CaM-independent CaMKK2 activity. Furthermore, lithium altered CaMKK2 and GSK3β protein abundance, which was genotype-, sex- and brain region-dependent. Thus, S3D mice displayed increased CaMKK2 protein levels which might be due to compensatory mechanisms caused by the decreased Ca2+-CaM-independent in S3D mice or might be a result of increased CaMKK2 protein stability, since GSK3β-mediated phosphorylation of CaMKK2 increases the enzymatic stability (Green et al., 2011). Additionally, lithium increased GSK3β protein abundance in male and female S3D mice in the hippocampus and the striatum, which suggests a potential interaction between CaMKK2 activity, GSK3β protein abundance, and lithium in vivo. Surprisingly, lithium increased the ratio of S9-phosphorylated to unphosphorylated GSK3β in male, but not in female mice. Together, these data implicate a role of Ca2+-CaM-independent CaMKK2 activity in the actions of lithium in vivo.
Chapter 7 describes in vitro and in vivo studies of the effects of the de-novo mutation c.928A>G p.M310V in the Ca2+-calmodulin-dependent protein kinase 4 (CAMK4) gene, which is clinically associated with developmental delays, intellectual disability, muscle hypotonia, autism spectrum disorder, and epilepsy which in some cases are accompanied by impaired motor function and motor control, feeding difficulties, delayed speech, infantile spasms, and the inability to walk/abnormal gait (Dr Scott D. McLean, The Children's Hospital of San Antonio, San Antonio, Texas, USA, unpublished data).
In vitro studies revealed that the M310V variant in the autoinhibitory domain of CaMK4 displays elevated Ca2+-CaM-independent activity (Dr John W. Scott, unpublished data). In studies described in this Chapter the equivalent mouse model (‘M306V’) was characterised using a range of behavioural assays. Male M306V mice displayed altered spatial memory function in the Barnes maze and the novel object recognition task, slight impairments of motor function, and a trend towards decreased front-paw strength; but no alteration in locomotor activity, or anxiety- and depressive-like behaviours. Together, the behavioural analysis suggests that M306V mice might be a suitable mouse model to study some, but not all clinical symptoms of human M310V carriers. Immunoblot analysis revealed that CaMK4M306V was strongly reduced compared to WT protein, which suggests that the mutation causes a gain-of-function in vitro and a loss-of-protein in vivo. This might explain the similarities between memory deficits observed in M306V mice and Camk4-/- mice (Kang et al., 2001). Additionally, brain-region-specific phosphoproteomics analysis identified abnormalities in the phosphorylation level of postsynaptic density-95 (PSD-95), Ca2+-calmodulin-dependent protein kinase 2γ (CaMK2γ) and Shank in the hippocampus, which are associated with autism spectrum disorder and intellectual disability, as well as in other proteins involved in processes including synaptic function, cellular organisation, neuronal projection development, and guanosine triphosphatase (GTPase) regulator activity. Furthermore, pathway and process enrichment analysis combined with human disease association studies pointed towards an involvement of CaMK4 in pathways and disorders linked to the human phenotype. Together these data are consistent with the idea that the clinical symptoms in human carriers of the M310V mutation are caused by the gain-of-function/loss-of-protein mutation in the Camk4 gene and suggest that the M306V mouse model might be suitable for the study of the neurological symptoms in human patients. Specifically, these studies suggest that the novel M306V mouse model might be suitable to study learning-and memory deficits as well as the underlying molecular mechanisms to gain further insights into how the M310V mutation is linked to human disease. |
|---|