The MAP Kinase Phosphatase MKP-1 Modulates Neurogenesis via Effects on BNIP3 and Autophagy

Inherited and acquired defects in neurogenesis contribute to neurodevelopmental disorders, dysfunctional neural plasticity, and may underlie pathology in a range of neurodegenerative conditions. Mitogen-activated protein kinases (MAPKs) regulate the proliferation, survival, and differentiation of neural stem cells. While the balance between MAPKs and the family of MAPK dual-specificity phosphatases (DUSPs) regulates axon branching and synaptic plasticity, the specific role that DUSPs play in neurogenesis remains unexplored. In the current study, we asked whether the canonical DUSP, MAP Kinase Phosphatase-1 (MKP-1), influences neural stem cell differentiation and the extent to which DUSP-dependent autophagy is operational in this context. Under basal conditions, Mkp-1 knockout mice generated fewer doublecortin (DCX) positive neurons within the dentate gyrus (DG) characterized by the accumulation of LC3 puncta. Analyses of wild-type neural stem cell (NSC) differentiation in vitro revealed increased Mkp-1 mRNA expression during the initial 24-h period. Notably, Mkp-1 KO NSC differentiation produced fewer Tuj1-positive neurons and was associated with increased expression of the BCL2/adenovirus E1B 19-kD protein-interacting protein 3 (BNIP3) and levels of autophagy. Conversely, Bnip3 knockdown in differentiated Mkp-1 KO NSCs reduced levels of autophagy and increased neuronal yields. These results indicate that MKP-1 exerts a pro-neurogenic bias during a critical window in NSC differentiation by regulating BNIP3 and basal autophagy levels.


Introduction
Neurogenesis is a tightly controlled process occurring during embryonic development that persists in the adult brain, primarily within the subventricular zone (SVZ) and the hippocampal dentate gyrus (DG). Defects in neurogenesis have been linked to affective disorders and severe intellectual and behavioral deficits. During the initial phase of CNS development, the brain generates up to 250,000 neurons per minute [1]. Adult hippocampal neurogenesis in the DG is highly conserved across mammals [2]. Supported by the permissive environment of the neurovascular niche [3], neurogenesis generates approximately 700 new neurons per day, with a modest decline during aging [4]. Under conditions of mild physiological stress, cell-intrinsic responses and the release of brain-derived neurotrophic factor (BDNF) induces neurogenesis [5]. Conversely, genetic conditions, such as Down Syndrome, and age-related neurodegenerative conditions, such as Parkinson's and Alzheimer's diseases, are associated with a net reduction in neural stem cell (NSC) proliferation and neurogenesis [6][7][8]. Given the potential therapeutic benefit, enhancing induced neurogenesis under these and other pathological conditions remains an active area of investigation [8][9][10]. A more thorough understanding of the cell-intrinsic factors that modulate neurogenesis could inform new approaches to treat inherited and acquired disorders of the central nervous system (CNS).
Initially considered a key pathway for clearing long-lived proteins [11], autophagy also supports cell survival under nutrient starvation and other stress conditions by recycling Millipore Sigma, Burlington, MA, USA) and LC3 (1:50, APG8B, Abgent, San Diego, CA, USA). Coronal sections from matched regions (Bregma −2.46~−1.46 mm, n = 3) were used from each sample to quantify DCX-positive cells and LC3 puncta numbers in the CNS. IHC stained sections were blinded, imaged using Lionheart FX automated microscope (BioTek, Winooski, VT, USA), and evaluated for the numbers of DCX positive cells and LC3 puncta numbers within DG using ImageJ. The sum of DCX counts on both sides of DG was used as the DCX positive cell number for each brain. LC3 puncta were counted in five 100 pixels by 100 pixels regions in each DG, and the sum of the numbers was used as the puncta number of a brain.

Quantitative PCR
RNA from undifferentiated and differentiated NSCs at different time points were prepared with E.Z.N.A. ® Total RNA Kit (SKU: R6834-01, Omega Bio-Tek, Norcross, GA, USA) and treated with DNase I (M0303S, New England Biolabs, Ipswich, MA, USA) followed by clean up with Monarch ® RNA Cleanup Kit (T2030S, New England Biolabs). cDNA was synthesized using the iScript cDNA Synthesis kit (1708890, Bio-Rad, Hercules, CA, USA). Quantitative PCR (qPCR) was performed with primers and exon spanning Taqman probes for mouse Mkp-1 and β-actin with FAM-MGB and VIC-MGB, respectively (Mm00457274_g1/4331182 and Mm02619580_g1/ 4448489, ThermoFisher Scientific). qPCR was performed on a StepOnePlus Real-Time PCR System (ThermoFisher Scientific), and analyses were performed using β-actin as an internal reference gene for normalization in the ∆∆CT analyses.

Immunocytochemistry
Cells cultured on coated coverslips were washed with PBS and fixed with 4% PFA at room temperature for 15 min. Then cultures were washed three times in PBS with 0.05% Triton X-100, followed by blocking at room temperature for 30 min in TBS-Blotto Biomolecules 2021, 11, 1871 4 of 16 (0.15 M NaCl, 20 mM Tris-HCl, pH 7.5, 4.5% non-fat dry milk) with 0.1% Triton X-100. Cells were incubated with anti Tuj1 (1:1000, T8578, Sigma-Aldrich) and GFAP antibody (1:1000, #PA1-10019, ThermoFisher Scientific) for 1 h at room temperature with shaking, washed three times, and incubated with corresponding Alexa Fluor secondary antibodies (A11029, A11012, ThermoFisher Scientific) covered with foil for 1 h. Cells were rinsed twice, stained with 300 nM DAPI for 5 min, and rinsed twice. Coverslips were mounted with mowiol and imaged using the Lionheart FX automated microscope. Coverslips were coded, and the experimentalist was blind to their assignments. In total, 15-25 fields from 3-5 coverslips were scanned under 20× objective and counted using ImageJ [29].

Flow Cytometry
NSCs were seeded in 6-well plates in 3 mL of complete proliferation or differentiation medium. Prior to flow cytometry, cells were dissociated with Accumax and centrifuged for 5 min at 400× g at room temperature. Cells were then stained with autophagy/cytotoxicity Dual Staining Kit (600140, Cayman Chemical, Ann Arbor, MI, USA). In brief, monodansylcadaverine (MDC, a fluorescence marker of multilamellar autophagic vacuoles [30]) and propidium iodide (PI) were diluted at 1:1000 in Cell-Based Assay Buffer (came with the kit) to make the staining solution. For each sample, 350 µL staining solution was added and incubated at 37 • C for 10 min in the dark. Cells were centrifuged and washed with 500 µL Cell-Based Assay Buffer. Data of forward scatter (FSC), side scatter (SSC), Blue D (PI), and Violet B (MDC) channels were collected on a 3-laser, 12-color BD LSR-II platform (BD Biosciences, San Jose, CA, USA). Data were analyzed with FlowJo software (FLOWJO, Portland, OR, USA). Ten thousand events were collected per sample. Debris, clumped cells, and dead cells were eliminated by gating with FSC, SSC, and PI. The median values for MDC used as an indicator of autophagy were analyzed in single live cells and compared among different groups.

Luciferase Assays
WT or Mkp-1 KO NSCs were dissociated with Accumax and counted. Mouse neural stem cell NucleofectorTM kit (VPG-1004, Lonza AG, Cologne, Germany) was used for electroporation. Briefly, 2 × 10 6 cells were collected and resuspended in 100 µL Mouse Neural Stem Cell Nucleofector ® Solution with supplement 1 and mixed with 2.5 µg pGL3-Basic based luciferase reporter vector and 0.25 µg pRL-TK control vector. The cell/plasmid suspension was transferred into a cuvette and inserted into the Nucleofector ® Cuvette Holder. Program A-033 was applied on a NucleofectorTM II (Amaxa Biosystems, Cologne, Germany). Then, 500 µL pre-equilibrated culture medium was added to the cuvette immediately. Cells were gently transferred into a prepared 24-well plate with complete differentiation medium. Twenty-four hours after transfection, cells were lysed, and luciferase activity measurements were performed with Dual-Luciferase ® Reporter Assay System (E1910, Promega, Madison, WI, USA) on a SpectraMax iD3 Microplate Reader (Molecular Device, San Jose, CA, USA). pGL3-Basic and pRL-TK were purchased from Promega. pGL3-Basic based Bnip3 promoter reporter vector (pGL3-Bnip3) was provided by Dr. Richard Bruick (UT Southwestern) [31].

Statistical Analysis
Statistical analyses were performed using GraphPad Prism ® v7 (GraphPad Software, San Diego, CA, USA). Data were expressed as mean ± standard deviation and compared among groups, accepting p < 0.05 as significant. Unpaired two-tailed t-testing was used to analyze data from Figures 1, 3A, 4 and 5B,E,F. Data from Figures 2D, 3B,C and 5A were analyzed using two-way ANOVA with Sidak's multiple comparison post hoc test. The qPCR data shown in Figure 2A was analyzed by one-way ANOVA with Tukey's multiple comparison post hoc test.

Loss of MKP-1 Induces Autophagy in Differentiated NSCs
Autophagy influences both the proliferation and differentiation of stem cell populations [35]. There is also ample evidence to indicate that autophagic responses exhibit cellspecific differences [17,18]. Given these observations, we next evaluated MKP-1 effects on autophagy in the CNS, tracking the abundance of intracellular LC3 puncta considered an established marker of autophagy [36]. Mkp-1 KO mice exhibited a robust increase in the number of LC3 puncta present in the dentate gyrus compared to WT controls (WT, 175 ± 38 vs. Mkp-1 KO 346 ± 67, n = 3, p < 0.05) ( Figure 3A). Similarly, LC3-II expression was increased in Mkp-1 KO cultures compared to WT NSCs 24 h post-differentiation. This difference declined between 48 and 72 h ( Figure 3B), mirroring changes in Mkp-1 expression.

Loss of MKP-1 Induces Autophagy in Differentiated NSCs
Autophagy influences both the proliferation and differentiation of stem cell populations [35]. There is also ample evidence to indicate that autophagic responses exhibit cell-specific differences [17,18]. Given these observations, we next evaluated MKP-1 effects on autophagy in the CNS, tracking the abundance of intracellular LC3 puncta considered an established marker of autophagy [36]. Mkp-1 KO mice exhibited a robust increase in the number of LC3 puncta present in the dentate gyrus compared to WT controls (WT, 175 ± 38 vs. Mkp-1 KO 346 ± 67, n = 3, p < 0.05) ( Figure 3A). Similarly, LC3-II expression was increased in Mkp-1 KO cultures compared to WT NSCs 24 h post-differentiation. This difference declined between 48 and 72 h ( Figure 3B), mirroring changes in Mkp-1 expression. We observed similar trends in cultures labeled with monodansylcadaverine (MDC), a fluorescence marker of multilamellar autophagic vacuoles. While differentiation induced MDC levels in both genotypes, MDC levels were higher in Mkp-1 KO cultures relative to WT controls (KO 9841 ± 765 vs. WT 7382 ± 437, n = 3, p < 0.05) ( Figure 3C) 24 h post-differentiation. This effect normalized within 48 h, consistent with observed changes in LC3-II levels. These results indicate that MKP-1 influences levels of autophagy early in the process of NSC differentiation. We observed similar trends in cultures labeled with monodansylcadaverine (MDC), a fluorescence marker of multilamellar autophagic vacuoles. While differentiation induced MDC levels in both genotypes, MDC levels were higher in Mkp-1 KO cultures relative to WT controls (KO 9841 ± 765 vs. WT 7382 ± 437, n = 3, p < 0.05) ( Figure 3C) 24 h postdifferentiation. This effect normalized within 48 h, consistent with observed changes in LC3-II levels. These results indicate that MKP-1 influences levels of autophagy early in the process of NSC differentiation.

Effect of MKP-1 Deficiency on Neurogenesis Is Autophagy-Dependent
Given the observed effects of MKP-1 on neurogenesis and autophagy, we next asked whether MKP-1's effects on neurogenesis were autophagy-dependent. To do this, we treated WT NSC-D cultures with the autophagy-inducer rapamycin (10 nM, in DMSO), while exposing Mkp-1 KO cultures to the autophagy inhibitor 3-MA (2 mM, in medium) during the first 24 h of differentiation ( Figure 4A). Drug effects on levels of autophagy were confirmed in NSC-D cultures using MDC fluorescence and LC3-II expression ( Figure 4B). When WT NSC-D cultures were exposed to rapamycin, both MDC and LC3-II levels were increased relative to DMSO treated controls. Conversely, treatment of Mkp-1 KO cultures with 3-MA treatment decreased both MDC and LC3-II levels when compared against controls ( Figure 4C). treated WT NSC-D cultures with the autophagy-inducer rapamycin (10 nM, in DMSO), while exposing Mkp-1 KO cultures to the autophagy inhibitor 3-MA (2 mM, in medium) during the first 24 h of differentiation ( Figure 4A). Drug effects on levels of autophagy were confirmed in NSC-D cultures using MDC fluorescence and LC3-II expression (Figure 4B). When WT NSC-D cultures were exposed to rapamycin, both MDC and LC3-II levels were increased relative to DMSO treated controls. Conversely, treatment of Mkp-1 KO cultures with 3-MA treatment decreased both MDC and LC3-II levels when compared against controls ( Figure 4C).

Discussion
These studies indicate that the canonical nuclear MAPK phosphatase MKP-1 exerts pro-neurogenic bias via effects on BNIP3 expression and autophagic signaling. We found that MKP-1 loss of function is associated with impaired neurogenesis in the dentate gyrus and differentiated NSC cultures under basal conditions. The reduced capacity for neurogenesis in differentiated Mkp-1 KO cultures correlates with increased autophagy and BNIP3 expression. Conversely, Bnip3 knockdown in Mkp-1 KO cultures rescues autophagy and neurogenic activity.
Mounting evidence indicates that MKP-1 serves a protective role in the CNS under a range of disease-relevant conditions. MKP-1 promotes the growth and elaboration of neuronal processes in dopaminergic neurons cultured from E14 rat ventral mesencephalon, protecting them against neurotoxic challenge [38]. MKP-1 also exerts neuroprotective effects following acute cerebral ischemia and in chronic models of neurodegenerative disease [33,39]. For example, MKP-1 loss of function accelerates pathogenic changes in models of Alzheimer's disease, characterized by impaired neurogenesis. Conversely, MKP-1 induction improves long-term potentiation and corrects cognitive deficits in this disease [20]. In light of our observations regarding the pro-neurogenic potential of MKP-1, it is plausible that stress-induced activation of MKP-1 reflects a compensatory mechanism intended to promote adaptive plasticity within hippocampal networks [40].
The existing literature supports a role for MKP-1 in regulating autophagy in nonneuronal contexts. In 2016, Wang et al. [17] reported Mkp-1 shRNA knockdown increased basal and rapamycin-increased autophagic flux in murine embryonic fibroblasts and the human CAOV3 ovarian line. Our finding that loss of Mkp-1 increased autophagy in the dentate gyrus and in differentiated NSCs is consistent with these reports. An important aspect of this study is that MKP-1's effect on autophagy appears operative during the initial 24 h post-differentiation, consistent with transient Mkp-1 mRNA expression in differentiated wild-type NSC cultures. These findings suggest that early induction of MKP-1 and repression of autophagy may be required to exert the observed pro-neurogenic effect. In other studies, transplanted GABAergic interneuron precursors exhibit marked sensitivity to autophagy within the first 24 h following transplantation. Inhibition of autophagy increases the survival and fraction of neurite-bearing GABAergic interneurons within two weeks post-implantation [41]. While our in vitro analyses focused on the initial post-differentiation phase (3-6 days), our in vivo analyses suggest that the pro-neurogenic effects of basal MKP-1 activity in the dentate gyrus persist into adulthood. By extension, these findings argue that targeted augmentation of MKP-1 activity could be used to treat neurodegenerative conditions or acute brain injury by promoting the neurogenic potential of endogenous or transplanted stem progenitors [42,43].
Initially described by Daido et al. [44], BNIP3 -dependent control over autophagy has been demonstrated in a range of cell types, including neurons and myocytes [37,45]. Clues from the existing literature suggested several options through which MKP-1 might influence BNIP3 activity. For example, MKP-1 is associated with changes in JNK-dependent phosphorylation of BNIP3 [19]. However, while we found that λ-phosphatase treatment induced a minor shift in the high molecular weight BNIP3 species in NSC-D cultures (Supplementary Figure S2A), no genotype-specific differences were seen in these two forms under basal conditions. Likewise, loss of MKP-1 in NSCs had no apparent effect on levels of JNK or p38 activation within 24 h of differentiation (Supplementary Figure S2B,C). The lack of MKP-1-dependent effects on these markers may relate to cell-context dependent effects and functional redundancy between the MAPK and DUSP family members [18]. MKP-1 also inhibits the accumulation of HIF-1α protein [46,47], and HIF-1α induced expression of BNIP3 in the context of ischemia-reperfusion injury induces autophagy [48]. Although basal expression of HIF-1α protein remained low in NSC-D cultures irrespective of Mkp-1 status (Supplementary Figure S2D), BNIP3 expression was increased in Mkp-1 KO cultures ( Figure 5A,B). Of note, it remains possible that Mkp-1 deletion increased HIF-1α's transcriptional potency without influencing its stability, which is typically low under normoxic conditions. Nonetheless, we suspect that BNIP3 induction contributed to enhanced autophagy by either interaction with different molecules such as LC-3 and its related molecular receptors [49] or by forcing the release of Beclin-1 from heterodimeric complexes with BCL-2 [50], suppression of the GTPase Rheb, and inhibition of mTOR [51].
Our findings link MKP-1 deficiency with BNIP3-induced autophagy and repressed neurogenesis in differentiated NSCs. Given BNIP3's established role in autophagic cell death [44], it is tempting to posit that BNIP3's anti-neurogenic effects are autophagyrelated, acting via pathways controlling autophagy and apoptosis, which exhibit significant crosstalk. For example, the ATG5-ATG12 conjugation system is required for the process of autophagosome formation, while ATG5 and ATG12 regulate stress-induced apoptosis [52]. It is also plausible that alternate BNIP3 targets may be involved. For example, the prosurvival factors BCL-2 and BCL-XL regulate neural differentiation, support the survival of newborn neurons, and increase rates of adult neurogenesis [53][54][55][56][57]. Considering that BNIP3 induces cell-death by disrupting BCL-2 and BCL-XL interactions [58], elevated BNIP3 levels could titrate out these pro-survival BH3 domain proteins with repressive effects on neurogenesis.
Since our in vitro endpoint analyses focused on changes occurring during the first week, it remains to be seen whether MKP-1 exerts delayed effects on NSC differentiation. For example, manipulation of MKP-1 could influence the engraftment and differentiation of transplanted NSC populations. It is also important to consider that neurogenic effects observed in the Mkp-1 KO background may reflect contributions from non-cell-autonomous effects (i.e., paracrine or cell-cell interactions). For example, since the loss of MKP-1 enhances MAPK-dependent inflammation [59,60], the secretion of anti-neurogenic factors like FGF2 or IL-1β from non-neuronal CNS cell types could suppress neurogenesis in vivo. Implementation of a conditional Mkp-1 knockout model would provide insight regarding this nuanced aspect of MKP-1 function in the CNS [61].

Conclusions
We report the essential role of the dual-specificity phosphatase MKP-1 in neurogenesis during a critical window of neural stem cell (NSC) differentiation. MKP-1 induction within 24 h after NSC differentiation exerts a pro-neurogenic bias via effects on BNIP3 and autophagy. Specifically, MKP-1 deficiency enhances BNIP3 expression and autophagy and impairs neurogenesis in vivo and in differentiated NSCs. Conversely, selective knockdown of BNIP3 is sufficient to reverse these changes. Collectively, these findings argue that transient manipulation of intrinsic MKP-1 activity during the initial phase of NSC differentiation could be used to improve the therapeutic potential of stem cell-based protocols.
Supplementary Materials: The following are available online at https://www.mdpi.com/article/ 10.3390/biom11121871/s1, Figure S1: Genotyping of mice and neural stem cells, Figure S2  Data Availability Statement: Data for this study can be found in the manuscript.

Conflicts of Interest:
The authors declare no conflict of interest.