Dusp8 affects hippocampal size and behavior in mice and humans

Dual-specificity phosphatase 8 (Dusp8) acts as physiological inhibitor for the MAPKs Jnk, Erk and p38 which are involved in regulating multiple CNS processes. While Dusp8 expression levels are high in limbic areas such as the hippocampus, the functional role of Dusp8 in hippocampus morphology, MAPK-signaling, neurogenesis and apoptosis as well as in behavior are still unclear. It is of particular interest whether human carriers of a DUSP8 allelic variant show similar hippocampal alterations to mice. Addressing these questions using Dusp8 WT and KO mouse littermates, we found that KOs suffered from mildly impaired spatial learning, increased locomotor activity and elevated anxiety. Cell proliferation, apoptosis and p38 and Jnk phosphorylation were unaffected, but phospho-Erk levels were higher in hippocampi of the KOs. Consistent with a decreased hippocampus size in Dusp8 KO mice, we found reduced volumes of the hippocampal subregions subiculum and CA4 in humans carrying the DUSP8 allelic variant SNP rs2334499:C > T. Overall, aberrations in morphology and behavior in Dusp8 KO mice and a decrease in hippocampal volume of SNP rs2334499:C > T carriers point to a novel, translationally relevant role of Dusp8 in hippocampus function that warrants further studies on the role of Dusp8 within the limbic network.


Supplement Figures
Suppl. Figure 1 Suppl. Figure 1. Social Discrimination Test in female and male Dusp8 WT and KO mice. (A) Schematic overview of test animals (dark grey) exposed to an unfamiliar ovariectomized female (light grey). Mice were allowed to freely interact for 4 min, followed by a 2-hour inter-trial interval (ITI). Subsequently, the same now familiar mouse and a new unfamiliar ovariectomized female were introduced to the test mouse for 4 min, and the interaction times were recorded. The interaction time of male Dusp8 WT or KO mice with the unfamiliar animals is depicted in the sample phase (B) or the test phase (C). The olfactory interaction times of female Dusp8 WT and KO mice are depicted in the sample (D) and test phase (E). Social investigation times included mounting behavior of male mice in addition to olfactory interactions. (F) shows the social discrimination index of male and (G) female Dusp8 WT and KO mice, respectively, calculated as ratio of interaction time with the unfamiliar mouse and the total interaction time with both mice. Male WT: n=15, male KO: n=14; female WT: n=8, female KO: n=12. Means ± SEM. # p < 0.1. Elements of artwork were designed by PB.

Sample Phase
First exposure Suppl. Figure 2. Reversal Learning Test of female Dusp8 WT and KO mice group-housed in an automated IntelliCage behavioral monitoring system. (A) Schematic of the reversal learning task. In the initial place learning test, one corner (labelled in white) contains a water bottle with free access after a nose poke, the other three (grey) corners contain bottles with blocked access. To induce reversal learning, the accessible bottle in the (white) corner is switched to the opposing corner, which will now grant access to water after a nose poke. Subsequently, (B) the Cognitive Flexibility Adaption of memory number of erroneous nose pokes in the first night phase as measure for the reversal learning performance, and (C-E) the overall activity of the mice in the first night phase, monitored as bi-hourly number of corner visits (C), average number of corner visits (D) and average number of nose pokes (E), were recorded. (F-I) depict the corresponding values from the second night of the reversal learning task. WT: n=8, KO: n=9. Means ± SEM. # p < 0.1; * p < 0.05; ** p < 0.01, *** p < 0.001. Elements of artwork were designed by PB.
Suppl. Figure 3 Suppl. Figure 3. Acoustic Startle Response, Prepulse Inhibition and Virtual Drum Test. The ASR was measured at increasing levels of stimuli varying from 70 dB to 120 dB with a background level of 65 dB. Neither male (A) nor female (C) Dusp8 KO mice showed differences in the ASR compared to their WT littermates. The PPI test was conducted with prepulse stimuli ranging from 67 dB to 81 dB with an inter-stimulus interval of 50 ms. PPI was unperturbed in Dusp8 KO mice compared to WT littermates (B,D   Suppl. Figure 5 Suppl. Figure  Suppl. Figure 7. Original non-cropped blots of Suppl. Figure 4A. Membrane was stripped with stripping buffer between the respective antibody incubation times and detection. Suppl. Figure 8 Suppl. Figure 8. Original non-cropped blots of Suppl. Figure 4B. Membrane was stripped with stripping buffer between the respective antibody incubation times and detection. Suppl. Figure 9 Suppl. Figure 9. Original non-cropped blots of Suppl. Figure 4C. Membrane was stripped with stripping buffer between the respective antibody incubation times and detection. Arrow indicates specific band for p-p38.
Suppl. Figure 10 Suppl. Figure 10. Original non-cropped blots of Suppl. Figure 4D. Membrane was stripped with stripping buffer between the respective antibody incubation times and detection. Arrow indicates specific band for p-p38. Suppl. Figure 11. Original non-cropped blots of Suppl. Figure 4E. Membrane was stripped with stripping buffer between the respective antibody incubation times and detection. Suppl. Figure 12. Original non-cropped blots of Suppl. Figure 4F. Membrane was stripped with stripping buffer between the respective antibody incubation times and detection.

Social discrimination
Social discrimination tests were performed and analyzed as described 1 and conducted between 08.00 am -02.00 pm. Briefly, for the sample phase an ovariectomized female stimulus mouse was exposed to the test mouse in a neutral cage for 4 min. Densitometric analyses were performed with ImageJ (2.0.0-rc-68/1.52f).

Immunohistochemistry
Mouse brains were harvested and processed according to the described method in 4.3 Hippocampus volume calculation. The following antibodies were employed for immunohistochemistry: rabbit anti-Ki67 (1:1000, ab15580, Abcam, Cambridge, UK) or Image-based automated cell counting The stained tissue sections were scanned at 20× objective magnification using an AxioScan.Z1 digital slide scanner (Zeiss, Oberkochen, Germany). Images were evaluated using the commercially available image analysis software Definiens Developer XD 2 (Definiens AG, Munich, Germany) following a previously published procedure 3 . Regions of interest, e.g. specific brain areas, were annotated manually.
The calculated parameters within these defined regions were the ratio of positively immunostained cells per total cell.