Background:

Mutations in Cyclin-dependent kinase-like 5 (CDKL5), a protein that modifies activity of other proteins in the brain¹, result in CDKL5 deficiency disorder (CDD), which presents with early-onset seizures, intellectual impairment, and features of autism spectrum disorder (ASD)². While scientists are beginning to understand the function of CDKL5, it is not understood how mutated CDKL5 causes CDD.

CDKL5 is abundant in the forebrain, where many different brain cell types are found, including neurons that send enhancing signals (excitatory neurons) or dampening signals (inhibitory neurons).  In one study, mice lacking Cdkl5 in excitatory neurons experienced problems with memory and learning, similar to CDD patients⁴. However, these mice did not experience behavioral deficits, including altered socialization with other mice, repetitive movements, or anxiety-related behavior found in mice lacking Cdkl5 in all cells⁴. Consequently, scientists think CDKL5 functions differently in different cell types.

In this paper, researchers removed Cdkl5 from only inhibitory neurons of the forebrain in mice they named Dlx-cKO⁵. They found the mice exhibited autistic-like features, unlike mice that only lost Cdkl5 in excitatory neurons. Following this discovery, the researchers searched for the source of this abnormality, explored a possible treatment, and attempted to replicate these results in mice with a mutation that mimics human CDD.

Behavior vs. Memory

                When presented with a mouse they had never met and a new object, Dlx-cKO mice preferred interaction with the object. Additionally, when only presented with a new mouse, Dlx-cKO mice interacted less often than mice with normal Cdkl5 function. These behaviors indicate reduced socialization and are thought to represent ASD-like features observed in patients⁵. Another ASD-like trait of CDD is repetitive behaviors. Dlx-cKO mice spent more time grooming, considered a metric of repetitive behavior⁵. Through learning and memory tests, researchers found Dlx-cKO mice had typical cognitive capacity, but difficulty adjusting to new situations. This suggests a normal baseline in learning and memory but impaired flexibility, also seen in humans with CDD⁴.

                Together, the data suggests CDKL5 in inhibitory neurons is important for different behaviors than those controlled by CDKL5 in excitatory neurons.

Abnormal Circuit Activity

Increased spontaneous brain activity, or hyperexcitability, is common in epilepsy and ASD⁶. Reversing hyperexcitability in mice alleviates autistic-like behaviors⁷. This suggests that hyperexcitability may be partially responsible for social impairments seen in CDD.

                The researchers examined hyperexcitability in Dlx-cKO mice as a source of behavioral changes. To do this, they compared Dlx-cKO mice, which lack Cdkl5 in inhibitory neurons, to mice lacking Cdkl5 in excitatory neurons and mice that have normal Cdkl5 function. They found hyperexcitability in Dlx-cKO mice, while mice lacking Cdkl5 in excitatory neurons demonstrated hypoexcitability, or reduced spontaneous brain activity. They concluded that hyperexcitability may be linked to the observed ASD-like behaviors in Dlx-cKO mice.

Blocking NMDA Receptors Reverses Autistic-Like Behaviors

One explanation for hyperexcitability in Dlx-cKO mice is changes in receptors on affected neurons. Receptors are protein sensors on neurons that function like coin-operated gates. When coins, or neurotransmitters, interact with the receptor, the gates open and signals are sent through. It is known that CDKL5 regulates the number and function of certain receptors, including NMDA receptors. NMDA receptors regulate neuronal activity and play a role in learning and memory⁹. The researchers hypothesized that blocking NMDA receptors could reduce autistic-like behaviors. They administered a drug (memantine) that blocks neurotransmitters from interacting with NMDA receptors¹⁰. This drug has been shown to improve behavioral deficits in other neurodevelopmental diseases¹⁰. Indeed, in the Dlx-cKO mice, memantine reduced time spent grooming and increased time interacting with new mice, while mice that did not receive memantine spent more time in isolation.

Treatment in Mice with Human-Based Mutations           

                To examine how memantine may function as a treatment when both excitatory and inhibitory neurons are affected by mutations in CDKL5, which is the case in humans with CDD, the researchers created mice with a Cdkl5 mutation identical to a patient mutation, referred to as R59X mice. These mice do not have functional Cdkl5 anywhere in their body. The researchers found an increase in NMDA receptors in R59X mice, a severe deficit in memory based on maze experiments, and demonstrated behavioral issues. This suggests that R59X mice have specific behavioral changes that were likely caused by Cdkl5 deficiency in both excitatory and inhibitory neurons. While the administration of memantine did not rescue the memory deficits, they did find that it reduced time spent engaging in repetitive behaviors and increased socialization.

Note on Seizures

One of the defining characteristics of CDD in humans is spontaneous seizures¹. The researchers did not observe spontaneous seizures in any of the mice. It has been hypothesized that seizures in Cdkl5-deficient mice could be possible under specific developmental conditions¹¹, but has not be widely replicated. These results suggest that mouse and human brains may have inherent differences in neural connections that prevent the development of seizures under CDKL5-deficiency¹².

Limitations & The Future: Mice as a Piece of the Puzzle

Though a promising explanation for behavioral manifestations of CDD, the results presented in this study must be taken with a grain of salt. It must be noted that all instances of behavioral deficits in mice may not be directly translatable to ASD in humans, but rather are referred to as “autistic-like” traits or behaviors. It is important to remember that mice are different from humans in several obvious and crucial ways.

                Mice and humans do not socialize or behave in the same manner. Humans have a much higher cognitive capacity, different social structures, and much more complex social interactions and behaviors, and comparisons to mice only serve as a basis for further experiments. This is not strictly a limitation of this study, but rather a limitation of behavioral mouse studies. While many measurements can be compared across species, behaviors are more difficult to decipher and translate.

This is not to say that results from this study are meaningless. Quite the contrary, this study gives researchers a good idea of next steps. With findings that suggest hyperexcitability in CDD is tied to excess NMDA receptors in inhibitory neurons, researchers can now investigate treatments with NMDA receptor modulators, such as memantine, to alleviate behavioral issues. It would be interesting and important to see if human cells, such as neurons made from patient-derived stem cells, show similar hyperexcitability that can be alleviated by memantine.

A mouse model can produce promising results and provide behavioral readouts, but we must demonstrate caution when applying these results to a human context and validate the efficacy before contemplating a clinical trial, which are financially and temporally expensive. However, therapeutic development is not possible without the groundwork laid by these mouse models.

Conflict of Interest Statement:

The author declares that they have no conflict of interest.

Original Abstract:

                CDKL5 deficiency disorder (CDD) is characterized by epilepsy, intellectual disability, and autistic features, and CDKL5-deficient mice exhibit a constellation of behavioral phenotypes reminiscent of the human disorder. We previously found that CDKL5 dysfunction in forebrain glutamatergic neurons results in deficits in learning and memory. However, the pathogenic origin of the autistic features of CDD remains unknown. Here, we find that selective loss of CDKL5 in GABAergic neurons leads to autistic-like phenotypes in mice accompanied by excessive glutamatergic transmission, hyperexcitability, and increased levels of postsynaptic NMDA receptors. Acute, low-dose inhibition of NMDAR signaling ameliorates autistic-like behaviors in GABAergic knockout mice, as well as a novel mouse model bearing a CDD-associated nonsense mutation, CDKL5 R59X, implicating the translational potential of this mechanism. Together, our findings suggest that enhanced NMDAR signaling and circuit hyperexcitability underlie autistic-like features in mouse models of CDD and provide a new therapeutic avenue to treat CDD-related symptoms.

References:

1. CDKL5 gene – Genetics Home Reference – NIH. (2020, August 17). Retrieved September 22, 2020, from https://ghr.nlm.nih.gov/gene/CDKL5
2. Weaving, L. S., Christodoulou, J., Williamson, S. L., Friend, K. L., McKenzie, O. L. D., Archer, H., Evans, J., Clarke, A., Pelka, G. J., Tam, P. P. L., Watson, C., Lahooti, H., Ellaway, C. J., Bennetts, B., Leonard, H., Gecz, J. Mutations of CDKL5 cause a severe neurodevelopmental disorder with infantile spasms and mental retardation. Am. J. Hum. Genet. 75, 1079–1093 (2004).
3. Zeisel, A., Munoz-Manchado, A. B., Codeluppi, S., Lonnerberg, P., La Manno, G., Jureus, A., Marques, S., Munguba, H., He, L., Betsholtz, C., Rolny, C., Castelo-Branco, G., Hjerling-Leffler, J., Linnarsson, S. Brain structure. Cell types in the mouse cortex and hippocampus revealed by single-cell RNA-seq. Science 347, 1138–1142 (2015).
4. Tang, S., Wang, I. J., Yue, C., Tanako, H., Terzic, B., Pance, K., Lee, J. Y., Coulter, D. A., Zhou, Z. Loss of CDKL5 in glutamatergic neurons disrupts hippocampal microcircuitry and leads to memory impairment in mice. J. Neurosci. 37, 7420–7437 (2017).
5. Tang, S., Terzic, B., Wang, I. J., Sarmiento, N., Sizov, K., Cui, Y., Takano, H., Marsh, E. D., Zhou, Z., T Coulter, D. A. (2019). Altered NMDAR signaling underlies autistic-like features in mouse models of CDKL5 deficiency disorder. Nature Communications, 10(1). doi:10.1038/s41467-019-10689-w
6. Rubenstein, J. L. R. & Merzenich, M. M. Model of autism: increased ratio of excitation/inhibition in key neural systems. Genes. Brain. Behav. 2, 255–267 (2003).
7. Chung, W., Choi, S. Y., Lee, E., Park, H., Kang, J., Park, H., Choi, Y., Lee, D., Park, S., Kim, R., Cho, Y. S., Choi, J., Kim, M., Lee, J. W., Lee, S., Rhim, I., Jung, M. W., Kim, D., Bae, Y. C., Kim, E. Social deficits in IRSp53 mutant mice improved by NMDAR and mGluR5 suppression. Nat. Neurosci. 18, 435–443 (2015).
8. Kang, J. & Kim, E. Suppression of NMDA receptor function in mice prenatally exposed to valproic acid improves social deficits and repetitive behaviors. Front. Mol. Neurosci. 8, 17 (2015).
9. Okuda, K., Kobayashi, S., Fukaya, M., Watanabe, A., Murakami, T., Hagiwara, M., Sato, T., Ueno, H., Ogonuki, N., Komano-Inoue, S., Manabe, H., Yamaguchi, M., Ogura, A., Asahara, H., Sakagami, H., Mizuguchi, M., Manabe, T., Tanaka, T. CDKL5 controls postsynaptic localization of GluN2Bcontaining NMDA receptors in the hippocampus and regulates seizure susceptibility. Neurobiol. Dis. 106, 158–170 (2017).
10. Figueiredo, C. P., Clarke, J. R., Ledo, J. H., Ribeiro, F. C., Costa, C. V., Melo, H. M., Mota-Sales, A. P., Saraiva, L. M., Klein, W. L., Sebollela, A., De Felice, F. G., Ferreira, S. Memantine rescues transient cognitive impairment caused by high-molecular-weight aβ oligomers but not the persistent impairment induced by low-molecular-weight oligomers. J. Neurosci. 33, 9626–9634 (2013).
11. Amendola, E., ­Zhan, Y., Mattucci, C., Castroflorio, E., Calcagno, E., Fuchs, C., Lonetti, G., Silingardi, D., Vyssotski, A. L., Farley, D., Ciani, E., Pizzorusso, T., Giustetto, M., Gross, C. T. Mapping pathological phenotypes in a mouse model of CDKL5 disorder. PLoS. ONE. 9, e91613–12 (2014).
12. Zhou, A., Han, S. & Zhou, Z. J. Molecular and genetic insights into an infantile epileptic encephalopathy—CDKL5 disorder. Front Biol. (Beijing) 12,1 –6 (2017).