Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2020 Nov;34(11):14750-14767.
doi: 10.1096/fj.202000850RR. Epub 2020 Sep 10.

Small ubiquitin-like modifier 2 (SUMO2) is critical for memory processes in mice

Shu Yu  1   2 , Francesca Galeffi  3   4   5   6 , Ramona M Rodriguiz  7 , Zhuoran Wang  1 , Yuntian Shen  1   2 , Jingjun Lyu  1 , Ran Li  1 , Joshua D Bernstock  8 , Kory R Johnson  9 , Shuai Liu  1 , Huaxin Sheng  1 , Dennis A Turner  3   4   5   6 , William C Wetsel  5   7   10 , Wulf Paschen  1 , Wei Yang  1
Affiliations

Small ubiquitin-like modifier 2 (SUMO2) is critical for memory processes in mice

Shu Yu et al. FASEB J. 2020 Nov.

Abstract

Small ubiquitin-like modifier (SUMO1-3) conjugation (SUMOylation), a posttranslational modification, modulates almost all major cellular processes. Mounting evidence indicates that SUMOylation plays a crucial role in maintaining and regulating neural function, and importantly its dysfunction is implicated in cognitive impairment in humans. We have previously shown that simultaneously silencing SUMO1-3 expression in neurons negatively affects cognitive function. However, the roles of the individual SUMOs in modulating cognition and the mechanisms that link SUMOylation to cognitive processes remain unknown. To address these questions, in this study, we have focused on SUMO2 and generated a new conditional Sumo2 knockout mouse line. We found that conditional deletion of Sumo2 predominantly in forebrain neurons resulted in marked impairments in various cognitive tests, including episodic and fear memory. Our data further suggest that these abnormalities are attributable neither to constitutive changes in gene expression nor to alterations in neuronal morphology, but they involve impairment in dynamic SUMOylation processes associated with synaptic plasticity. Finally, we provide evidence that dysfunction on hippocampal-based cognitive tasks was associated with a significant deficit in the maintenance of hippocampal long-term potentiation in Sumo2 knockout mice. Collectively, these data demonstrate that protein conjugation by SUMO2 is critically involved in cognitive processes.

Keywords: LTP; knockout; memory impairment; posttranslational modification.

PubMed Disclaimer

Conflict of interest statement

Conflict of interest.

The authors declare that they have no conflict of interest.

Figures

Figure 1.
Figure 1.. Generation and characterization of conditional Sumo2 knockout mice.
A) Targeting strategy for conditional deletion of exon 4 of the Sumo2 gene. The targeting vector was designed to introduce an FRT-flanked neomycin cassette (NEO) and 2 loxP sites into the targeted locus. After removal of the NEO cassette, the Sumo2f/f mouse line was obtained. Gray rectangles, coding exons; red triangle, loxP site; yellow triangle, FRT site; green arrow, primers (F1 and R1 for genotyping; F1 and R2 for verification of exon 4 deletion); TK, thymidine kinase cassette. Not drawn to scale. B,C) Deletion of Sumo2 in the mouse brain. Sumo2f/f mice were crossed with Emx1Cre/Cre mice to generate Sumo2f/f;Emx1-Cre mice (SUMO2-cKO). Genotyping priB) Verification of deletion of Sumo2 exon 4 in the SUMO2-cKO mouse brain using brain DNA samples. C) Quantitative RT-PCR analysis of SUMO1–3 expression in the hippocampus of control and SUMO2-cKO mice. D,E) SUMOylation dynamics in response to ischemic stress in the brain. D) Immunohistochemical analyses of SUMO2/3 and MAP2 expression in CA1 hippocampus and amygdala. Mice were subjected to 8 minutes of global brain ischemia and 1 hour of reperfusion. Brain sections were stained with anti-SUMO2/3 and anti-MAP2 (neuronal marker) antibodies. WT, wild-type. Scale bars: 20 μm. E) Wild-type (WT), SUMO1-KO, SUMO2-cKO, and SUMO3-KO mice were subjected to sham surgery (S) or 10 minutes forebrain ischemia (I) and 1 hour reperfusion. The ischemia-induced changes in global SUMOylation were evaluated by Western blotting. The high-molecular-weight regions, marked by brackets, were used to quantify SUMO1 and SUMO2/3 conjugation. Signal intensities were normalized to β-actin. To calculate fold change, mean values of the WT/sham group were set to 1.0. Data are presented as means ± SEM (n = 3/group). *p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001, vs respective sham group.
Figure 2.
Figure 2.. Open field and elevated zero maze tests.
These tests were conducted with control (Ctrl) and SUMO2-cKO (cKO) mice. A) Open field activity for locomotion (left), rearing (middle), and distance traveled in the center zone (right). No significant genotype effects were found in any of the analyses. Control vs. SUMO2-cKO: locomotion [t(1,15) = −1.799, p = 0.092], rearing [t(1,15) = −0.725, p = 0.480]; and center distance [t(1,15) = −1.846, p = 0.085]. B) Elevated zero maze performance depicting percent time in open areas (left), latency to enter open areas (middle), and numbers of transitions from closed-to-open-to-closed areas (right). Control vs. SUMO2-cKO: percent time [t(1,15) = −1.623, p = 0.125], latency [t(1,15) = 1.127, p = 0.278], and transitions [t(1,15) = −1.562, p = 0.139]. Data are presented as means ± SEM (n = 8–10 mice/genotype).
Figure 3.
Figure 3.. Cognitive tests in SUMO2-cKO mice.
A) Preference scores in the novel object recognition memory (NORM) test depicting training, short-term memory (STM) and long-term memory (LTM) testing. RMANOVA for object preference revealed significant within subjects effects of time [F(2,36)=3.468, p=0.042] and a significant time by genotype interaction [F(2,36)=5.655, p=0.007]. The between subjects effects of genotype was significant [F(1,18)=21.333, p<0.001]. n = 8–10 mice/genotype. B) Preference scores for the displaced object in the spatial object task. The RMANOVA revealed significant trial effects (training and test) [F(1,15) = 32.910, p < 0.001], a trial by genotype interaction [F(1,15) = 24.175, p < 0.001], and a genotype effect [F(1,15) = 4.728, p = 0.046]. C) Preference scores for the novel object in the memory load test. The RMANOVA revealed a significant trials effect (trials 1–7) [F(6,90) = 5.870, p < 0.001], trials by genotype interaction [F(6,90) = 1.888, p = 0.093], and genotype effect [F(1,15) = 23.046, p < 0.001]. D) Fear conditioning. SUMO2-cKO mice were deficient in contextual and cued fear conditioning. (Left) RMANOVA for conditioning revealed a significant time effect [F(2,30) = 12.369, p < 0.001]; however, the time by genotype interaction [F(2,30)=0.208, p=0.813] and the genotype effect [F(1,15)=0.047, p=0.831] were not significant. (Middle) Contextual fear was significant [t(1,15) = 3.382, p = 0.004]. (Right) RMANOVA for cued fear revealed significant time [F(1,15) = 119.492, p < 0.001] and genotype effects [F(1,15) = 8.364, p = 0.011]; the time by genotype interaction [F(1,15)=3.243, p=0.092] was not significant. Results shown as means ± SEM (n = 8–9 mice/genotype for panels B-D). *p < 0.05, compared to the respective controls.
Figure 4.
Figure 4.. Cell numbers and spine densities in CA1 hippocampus of SUMO2-cKO mice.
A) No differences in cell numbers in hippocampus were observed between and control and SUMO2-cKO mouse brains (5 sections from each of 2 mice/group). Scale bars: 50 μm. B) No significant differences in spine densities of CA1 neurons were observed between control and SUMO2-cKO (40–50 neurons per animal; n = 3 per group). Scale bars: 5 μm.
Figure 5.
Figure 5.. RNA-Seq analyses of hippocampal samples from control and SUMO2-cKO mice.
A) RNA-Seq gene expression heat map. RNA hippocampal samples were prepared from control and SUMO2-cKO mice. Differentially regulated genes (SUMO2-cKO vs control; Table S5) were used to generate the heat map. Heat map colors indicate the extent of fold changes between SUMO2-cKO vs control samples. B) Verification of RNA-Seq data by quantitative RT-PCR analysis. All data were normalized to β-actin. To calculate fold change, the mean values of control samples were set to 1.0. Data are presented as means ± SEM (n=3/group); **p < 0.01; ***p < 0.001, cKO versus control. C) Read coverage across the Sumo2 gene in both genotypes. Reads were aligned to multiple genomic positions using the STAR alignment algorithm. Illustration of the Sumo2 gene, including its exons and introns, is shown at the bottom. Note that the number of reads is dramatically reduced in SUMO2-cKO vs control.
Figure 6.
Figure 6.. Western blot analysis of hippocampal samples from control and SUMO2-cKO mice.
Levels of AMPA receptors (GluR1–4), NMDA receptors (NR1, NR2A, and NR2B) and other memory-related proteins, including synaptophysin (SYP) and PSD95, were evaluated using tissue homogenates (A,B) and synaptosome fractions (C,D) from hippocampus. Quantitative data from Western blotting were generated by normalizing the intensities of each band to β-actin. Mean values of the control group were set to 1.0. Data are presented as means ± SEM (n=4/group).
Figure 7.
Figure 7.. CA1 hippocampal synaptic transmission in SUMO2-cKO mice.
A) Input–output curve showing the relationship between pre-synaptic axonal fiber volley (FV) amplitude and fEPSP slope (in the range of 50 to 250 μA stimulation) RMANOVA revealed significant main effects of FV [F(9, 108) = 77.04, p<0.0001], genotype [F(1, 12) = 4.846, p=0.048], and the intensity by genotype interaction [F(9, 108) = 3.52, p=0.0007] (n=6–8 slices/group). B) The mean ratio of the FV-fEPSP slope was significantly reduced in SUMO2-cKO mice compared to littermate controls [t(1,12)=3.166, p=0.008] (n=6–8 slices/group). C) The relationship between FV amplitude and stimulus intensity was similar in SUMO2-cKO and control mice. RMANOVA revealed significant main effects of intensity [F(9, 108) = 47.93, p<0.0001], but genotype was not a significant factor [F(1, 12) = 0.047, p=0.833], and the intensity by genotype interaction was also not significant [F(9, 108) = 0.287, p=0.977] (n=6–8 slices/group). D) fEPSP slope amplitude in response to increasing stimulus intensity. RMANOVA revealed significant main effects of intensity [F(9,117) = 87.04, p<0.0001], genotype [F(1,13) = 5.267, p=0.039], and the intensity by genotype interaction [F(9,117) = 3.747, p=0.0004] (n=6–9 slices/group). E) The mean I/O slopes in SUMO2-cKO were significantly smaller relative to control mice [t(1,13)=2.407, p=0.032]. F) Paired-pulse facilitation was similar between genotypes. Mean responses at different stimulus intervals. G) Examples of traces showing responses to pairs of pulses at a 20 millisecond (ms) intervals. H) The mean time-course of fEPSP before and after induction of long-term potentiation (LTP) at time 0, expressed as the percent fEPSP slope normalized to a 10-minute baseline. Inset: representative traces of fEPSPs before (gray line) and 60 minutes after (black line) LTP induction. I) Percent fEPSP slope potentiation 60 minutes after LTP induction; mean slopes values for each experimental group consisted of 10 consecutive recordings (55–60 minutes post high frequency stimulation) and were significantly different between genotypes [t(1,13)=4.582, p=0.0005] (n = 7–8 slices/group). Data are presented as means ± SEM. *, p < 0.05; **, p < 0.01 ***, p < 0.001. fEPSP, field excitatory postsynaptic potential; presynaptic fiber volley (FV); PPR, paired pulse ratio.
See this image and copyright information in PMC

Similar articles

See all similar articles

Cited by

See all "Cited by" articles

References

    1. Ho VM, Lee JA, and Martin KC (2011) The cell biology of synaptic plasticity. Science 334, 623–628 - PMC - PubMed
    1. Wang L, Rodriguiz RM, Wetsel WC, Sheng H, Zhao S, Liu X, Paschen W, and Yang W (2014) Neuron-specific Sumo1–3 knockdown in mice impairs episodic and fear memories. J Psychiatr Neurosci 39, 259–266 - PMC - PubMed
    1. Henley JM, Craig TJ, and Wilkinson KA (2014) Neuronal SUMOylation: mechanisms, physiology, and roles in neuronal dysfunction. Physiol Rev 94, 1249–1285 - PMC - PubMed
    1. Matsuzaki S, Lee L, Knock E, Srikumar T, Sakurai M, Hazrati LN, Katayama T, Staniszewski A, Raught B, Arancio O, and Fraser PE (2015) SUMO1 Affects Synaptic Function, Spine Density and Memory. Sci Rep 5, 10730. - PMC - PubMed
    1. Ripamonti S, Shomroni O, Rhee JS, Chowdhury K, Jahn O, Hellmann KP, Bonn S, Brose N, and Tirard M (2020) SUMOylation controls the neurodevelopmental function of the transcription factor Zbtb20. J Neurochem. - PubMed

Publication types

Substances