3.5 hLf inhibits APP phosphorylation via p38 and PP2Ac
interaction-mediated upregulation of PP2A activity
To further unveil the mechanism of p-APP downregulation in the
APP/PS1/Lf mouse brains, the mouse primary astrocytes and N2a-sw cells
were employed in the following studies. Our data revealed that Aβ42
oligomers treatment significantly upregulated the Lf expression in
primary astrocytes from WT mice, and also promote Lf secretion from
astrocytes (Fig. 5A), suggesting that astrocytic Lf may have a role in
regulating Aβ metabolism. Indeed, the conditional astrocytic medium from
Astro-Lf mice significantly repressed the p-APP expression in N2a-sw
cells (Fig. 5B), indicating that the astrocytes may secrete Lf to
directly inhibit the p-APP expression in neurons since the elevated Lf
content was observed in the medium of astrocytes from Astro-Lf mice
(Fig. 2C). To further elucidate our hypothesis, we employed hLf for the
following studies. The hLf was reported to reduce Aβ production via
upregulating ADAM10 in the N2a-sw cells in our previous study (Guo et
al., 2017), therefore, we focused on the effect of hLf on p-APP
expression in this study. As shown in Fig. 5C-D, the p-APP expression
was reduced, which was coincident with the downregulations of sAPPβ
expression and Aβ42 production after hLf treatment in the N2a-sw cells.
Furthermore, the immunostaining data (Fig. 5E) revealed that the p-APP
fluorescent intensity was positively correlated with the Aβ fluorescent
intensity in the control cells (Fig. 5E1) and hLf-treated cells (Fig.
5E1); and the p-APP fluorescent intensity was downregulated (Fig. 5E4),
which was accompanied by the decreased Aβ fluorescent intensity (Fig.
5E3) in the hLf-treated cells, suggesting that hLf effectively reduced
the Aβ production via inhibition of p-APP expression. Importantly, the
p-p38 expression was dramatically upregulated, while the expressions of
p-CDK5, p-GSK3α/β, and PP2Ac were unchanged in the hLf-treated N2a-sw
cells (Fig. 5F). Nevertheless, the PP2A activity was also elevated in
the hLf-treated N2a-sw cells (Fig. 5G), indicating that hLf treatment
may increase the APP dephosphorylation via upregulation of PP2A
activity, and this action may be associated with the activation of p38.
To verify that hLf-induced p-APP downregulation is in a p38
activation-dependent manner, the p38 inhibitor SB203580 was employed to
treat the N2a-sw cells. As shown in Fig. 5H, inhibition of p38 activity
abrogated the hLf treatment-induced p-APP downregulation, suggesting
that hLf treatment inhibited p-APP expression needs p38 activation.
Interestingly, inhibition of p38 considerably increased the p-APP
expression (Fig. 5H), while activation of PP2A by its agonist DT-061
completely reversed this circumstance in N2a-sw cells (Fig. 5I),
indicating that p38 may regulate the p-APP expression in a PP2A
activity-dependent manner. As expected, the PP2A antagonist CA also
clearly rescued the p-APP expression in hLf-treated N2a-sw cells (Fig.
5J). These data suggested that hLf promoted APP dephosphorylation via
enhancing PP2A activity through p38 activation. Next, we asked how the
p38 activation regulates the hLf-induced enhancement of PP2A activity.
As shown in Fig. 5K, hLf treatment increased the interaction of p38 and
PP2Ac, while inhibition of p38 activity significantly abrogated this
effect in N2a-sw cells. Furthermore, the changes in the interaction of
p38 and PP2Ac after hLf and/or SB203580 treatment were highly
coincidence with the changes in the PP2A activities (Fig. 5L),
suggesting that hLf treatment enhanced the PP2A activity via increasing
the interaction of p38 and PP2Ac by promoting p38 activation. To further
elucidate that p38 interacts with PP2Ac to regulate PP2A activity, p38
was introduced into the N2a-sw cells. As shown in Fig. 5M-N,
overexpression of p38 significantly increased the PP2A activity without
changing the expression of PP2Ac. Moreover, overexpression of p38
slightly reduced the APP expression, but sharply inhibited the
phosphorylation of APP at Thr668 and the production of sAPPβ in N2a-sw
cells (Fig. 5O). Together, our data suggested that hLf inhibited APP
phosphorylation at Thr668 via enhancing PP2A activity by activation of
p38.
The actions of astrocyte-derived Lf on neurons may be mediated by the
neuronal receptors of Lf. LRP1 was reported to be a putative receptor of
Lf (Tang et al., 2010), we, therefore, knocked down LRP1 in N2a-sw cells
(Fig. 5P). Our data suggested that LRP1 silence significantly abrogated
the hLf-induced p38 activation, and also rescued the hLf-induced
downregulation of p-APP in N2a-sw cells (Fig. 5Q), indicating that hLf
treatment promoted p38 activation-mediated downregulation of p-APP
expression partly via targeting to its neuronal receptor LRP1.
DISCUSSION
Lf is overexpressed in the glial cells and neurons of the elderly person
and AD patients (Kawamata et al., 1993; Leveugle et al., 1994), while,
to date, the functions of Lf in different cell types remain unexplored.
Our recent study exerted that
astrocytic Lf knockout impaired
the neuronal dendritic complexity and cognitive ability via decreasing
cholesterol production during early life in mice (Xu et al., 2022),
suggesting the important roles of astrocytes in regulating neuronal
functions. Several studies revealed that supplements with Lf effectively
retarded AD progression (Abdelhamid et al., 2020; Guo et al., 2017). In
this study, the human Lf gene was firstly knocked in the astrocytes in
the APP/PS1 mice, and the results showed that overexpression of Lf
reduced the Aβ burden via enhancing PP2A activity by the activation of
p38 in APP/PS1 mice. Hence, promoting astrocytic Lf expression may be a
promising strategy for treating AD.
The current strategies for treating AD are suffering severe challenges
because the clinical trials of Aβ monoclonal antibody drugs and BACE1
inhibitors are almost failed. Therefore, the other targets for
regulating Aβ production have gained great attention in the following
studies. APP was previously demonstrated to be a potential target to
inhibit Aβ production, however, APP is an important membrane receptor
that involves in the signaling transductions in neurons (Deyts et al.,
2016), and APP knockout significantly impairs neuronal excitability and
synaptic plasticity (Lee et al., 2020; Weyer et al., 2011). Therefore,
decreasing neuronal APP expression may lead to serious side effects.
Notably, APP phosphorylation is an important step for Aβ production.
Phosphorylation of APP at Tyr682 by Fyn tyrosine kinase promoted the APP
translocated into acidic neuronal compartments where it is processed to
generate Aβ (Iannuzzi et al., 2020). Phosphorylation of APP at Ser675
also altered the balance of APP-processing through increasing meprin
β-mediated and decreasing α-secretase-mediated APP cleavage at the
plasma membrane (Menon et al., 2019). Additionally, phosphorylation of
APP at Thr668 promoted the APP transfer to the endosome where is cleaved
by the BACE1 and subsequently promoted Aβ generation (Lee et al., 2003).
In this study, astrocytic Lf overexpression inhibited the
phosphorylation of neuronal APP at Thr668, thereby diminishing Aβ burden
in APP/PS1 mice. Our previous study revealed that intranasal treatment
of Lf repressed the Aβ production via upregulating ADAM10 expression in
APP/PS1 mice (Guo et al., 2017). Similar to our previous study, dietary
supplemented with Lf also reduced the Aβ production in J20 transgenic
mice, whereas its mechanism involved the BACE1 inhibition, but not the
ADAM10 upregulation (Abdelhamid et al., 2020). Furthermore, Lf intake
was shown to reduce the contents of Aβ, tau, and oxidative damage
markers in the serum of AD patients, and enhance the cognitive functions
of AD patients (Mohamed et al., 2019). These studies suggested that the
different source of Lf may have different molecular mechanisms against
AD. Surely, our current study showed a significant decrease in APP
phosphorylation, but not ADAM10 and BACE1 in APP/PS1/Lf mice, suggesting
that the astrocyte-derived Lf may preferentially inhibited Aβ production
via decreasing neuronal APP phosphorylation in the condition of AD.
The Aβ burden is also modulated by the capacity of Aβ clearance.
Endothelial LRP1 could interact with Aβ and promote Aβ transcytosis
across the blood-brain barrier, thus enhancing Aβ efflux (Storck et al.,
2016). Furthermore, the neuronal LRP1 could also assist the uptake of Aβ
by neurons to degradation without influencing the Aβ-degrading enzymes
(Kanekiyo et al., 2013). Our study revealed that astrocytic Lf
overexpression rescued the downregulation of LRP1 in APP/PS1 mice,
suggesting the enhancement of Aβ clearance in the APP/PS1/Lf mouse
brains. The IDE expression was also upregulated in APP/PS1/Lf mouse
brains. IDE is an Aβ-degrading enzyme that is predominately secreted by
the astrocytes via an autophagy-based unconventional secretory pathway
(Son et al., 2016). A previous study revealed that Lf promoted the
secretion of astrocytic IDE to degrade Aβ (Abdelhamid et al., 2020), we,
therefore, suggested that astrocytic Lf overexpression may promote Aβ
degradation via directly promoting the astrocytic IDE secretion in
APP/PS1 mice. Notably, LRP1 and IDE were both reduced in the APP/PS1
mice compared to the WT mice, suggesting that Aβ may directly reduce the
expressions of LRP1 and IDE. Therefore, whether the astrocyte-derived Lf
upregulated the expressions of LRP1 and IDE is ascribed to reduced Aβ
burden in APP/PS1/Lf mice merits further investigation. Nevertheless, we
may conclude that astrocytic Lf overexpression could promote Aβ
clearance by enhancing the expressions of LRP1 and IDE in APP/PS1 mice.
Hyperphosphorylation of tau is also a prominent hallmark of AD, it can
impede the microtube assembly and impair the cellular cargo transport,
subsequently leading to neuronal dysfunction and neuronal loss (Xia et
al., 2021). Activations of CDK5 and GSK3α/β promote tau phosphorylation,
while PP2A own the opposite effect through its phosphatase activity
(Castro-Alvarez et al., 2014). Astrocytic Lf overexpression inhibited
the activities of CDK5 and GSK3α/β and increase the PP2A activity, thus
decreasing tau phosphorylation in APP/PS1 mice. Of note, PP2A is also a
critical regulator of APP phosphorylation, activation of PP2A activity
effectively reduces the Aβ production via dephosphorylation of APP at
Thr668 (Hu et al., 2022). The PP2A activities were elevated in the
APP/PS1/Lf mouse brains and hLf-treated N2a-sw cells, suggesting that Lf
may inhibit Aβ production via promoting PP2A-mediated APP
dephosphorylation. Increasing evidence exerted that p38 activation may
play an important role in regulating PP2A activity (Chiou et al., 2022;
Grethe and Porn-Ares, 2006). Activation of p38 was reported to enhance
the PP2A activity in tumor necrosis factor α (TNFα)-treated endothelial
cells (Grethe and Porn-Ares, 2006). In addition, activation of p38 was
also reported to stimulate the PP2A-mediated dephosphorylation of cAMP
response element binding protein (CREB) (Chiou et al., 2022), and the
mechanism may be ascribed to the p38 activation-induced phosphorylation
of PP2Ac at Tyr307 (Hsiao et al., 2020). However, the PP2A activity
evaluated using the PP2A (Tyr307) monoclonal antibodies is suffering
great challenges since those antibodies are not specific for
phosphorylated Tyr307 but instead are hampered by PP2Ac phosphorylation
at Thr304 or methylation at Leu309 (Frohner et al., 2020; Mazhar et al.,
2020). Therefore, the mechanism regarding p38 activation-induced
enhancement of PP2A activity should be re-interpreted. Interestingly,
p38 was verified as a partner of PP2A, and the interaction of p38 and
PP2A involved in the stem cell factor-induced cardiac stem cell
migration (Wang et al., 2017) and oxidative stress-induced DNA damage
response (Guillonneau et al., 2016). The p38 activation was found in the
APP/PS1/Lf mice and hLf-treated N2a-sw cells; furthermore, inhibition of
p38 activity significantly reduced the interaction of p38 and PP2Ac, and
also abrogated the hLf-induced interaction of p38 and PP2Ac; in
addition, the alternations in the interaction of p38 and PP2Ac after hLf
and/or SB203580 treatment were coincident with changes in the activities
of PP2A, indicating that p38 activity regulated the interaction of p38
and PP2Ac, thus directly influenced the PP2A activity. Indeed,
overexpression of p38 effectively upregulated the PP2A activity, further
reinforcing our conclusion that p38 interacts with PP2Ac to enhance the
PP2A activity. Taken together, our study firstly revealed that
astrocyte-derived Lf inhibited APP phosphorylation via enhancing PP2A
activity by promoting p38 activation-induced interaction of p38 and
PP2Ac.
The p38 was proposed to be a downstream target of Lf since Lf treatment
could effectively promote osteoblast cell proliferation and
differentiation by stimulating the activation of p38 (Liu et al., 2018;
Zhang et al., 2014). However, as the putative receptor of Lf, LRP1 did
not implicate the Lf-induced p38 activation in osteoblast cells (Tang et
al., 2010; Zhang et al., 2014). In contrast, our study revealed that
LRP1 knockdown largely abrogated the hLf-induced p38 activation, and
subsequently reversed hLf-mediated p-APP downregulation in N2a-sw cells,
indicating that the hLf-indued p38 activation was partly in a
LRP1-dependent manner in N2a-sw cells. Consistent with our results, the
p38 activation was also observed in a LRP1-dependent manner in the
Aβ42-treated neurons and astrocytes (Ma et al., 2016; Yang et al.,
2015), suggesting the LRP1-p38 axis may be a conserved signaling pathway
in neurons. Collectively, we may conclude that astrocyte-derived Lf may
stimulate p38 activation via targeting neuronal LRP1 and subsequently
reduce the neuronal p-APP expression.
CONCLUSION
our data firstly confirmed that overexpression of Lf in astrocytes
promoted the secretion of Lf from astrocytes, which subsequently bound
to its neuronal receptor LRP1 and trigged the activation of p38,
followed by elevation of PP2A activity via stimulating the interaction
of p38 and PP2Ac, and finally diminished Aβ production through
dephosphorylation of APP at Thr668 in neurons (Fig. 6). Hence, promoting
astrocytic Lf expression may be an attractive strategy to alleviate AD
progression.