4. Discussion
HMGB1 is significantly up-regulated in PD model animals and patients,
and its’ up-regulation is closely related to PD. The use of HMGB1
monoclonal antibody and a variety of compounds that inhibit HMGB1 have
observed protective effects on TH+ neurons in PD model
animals(Huh et al. , 2011; Santoro et al. , 2016; Sasakiet al. , 2016). Deficiency of HMGB1 downstream receptor TLR4 also
protected TH+ neurons in a mouse model of MPTP
(Campolo et al. , 2019). TLR4 are mainly expressed on the surface
of microglia in the CNS and mediate the pro-inflammatory effects of
glial cells (Rahimifard et al. , 2017). Among the three redox
forms of HMGB1, only the dsHMGB1 has the ability to bind to TLRs
(Venereau et al. , 2012; Yang et al. , 2010; Yang et
al. , 2015). The evidence seems to suggest that in PD, dsHMGB1 is the
HMGB1 that plays a deleterious role. However, due to frHMGB1 will always
oxidize into dsHMGB1 in inflammatory state(Palmblad et al. ,
2015), laboratory evidence for the role of frHMGB1 in PD is still few.
Previously, it was shown that injection of frHMGB1 into the brain of
healthy rats was sufficient to induce blood-brain barrier disruption and
IL1-β production(Aucott et al. , 2018). In PD, the passive release
of frHMGB1 in the nucleus of persistently injured neurons likely plays
an important role in the negative effects of HMGB1. Given that knockdown
of neuronal HMGB1 affects neuronal development in vivo , ourin vitro knockdown of midbrain neuron HMGB1 confirmed this.
Recently, Th17 cell subsets have revealed their deleterious effects in
PD in both animal models and patient samples(Chandra et al. ,
2016; Dutta et al. , 2019; Elgueta et al. , 2019; Reynoldset al. , 2010). We previously found that HMGB1 A Box has a
surprising inhibitory effect on T cell infiltration and Th17
differentiation in addition to protecting TH+ neurons
and inhibiting glial activation(Tian et al. , 2020). Although the
stereotactic injection of exogenous frHMGB1 into the rat brain can’t
increase the expression of phagocyte MHC Ⅱ like the rats injected with
dsHMGB1, it is also enough to lead to the destruction of blood-brain
barrier and the increase of CD68+ phagocytes(Aucottet al. , 2018). frHMGB1, passively released from the nucleus
during dopaminergic neuron damage, may play a key role in the
recruitment of peripheral immune cells.
Our results showed that CD45+ F4/80+peripherally derived macrophages were infiltrated in the SN of the MPTP
mouse model and were significantly reduced after HMGB1 A Box
application. Previously, significant nuclear-cytoplasmic HMGB1
translocations from neurons, astrocytes and microglia were observed in
the PD mouse MPTP model(Santoro et al. , 2016). However, the roles
of these different cell-derived HMGB1 on immune cells in the SN of PD
model animals lacked precise definition. We found that in the
MPP+ treated primary cell co-culture system, knockdown
of HMGB1 in midbrain neurons had the most significant inhibitory effect
on macrophage. This suggests that HMGB1 released by sustained injury of
neurons plays a crucial role in the recruitment of peripheral
monocyte/macrophages. The direct inhibitory effect of HMGB1 A Box on
macrophage and T cell was confirmed by experiments with HMGB1 A
Box-treated macrophages transwell experiment and transfusion of HMGB1 A
Box-treated GFP+CD3+ T cells into
wild type mice. Co-immunoprecipitation and immunofluorescence confirmed
that HMGB1 A Box was able to bind to CXCR4 on the surface of these
cells. Mice reinfused with CXCR4-neutralizing antibody-treated
GFP+CD3+ T cells had significantly
fewer GFP+ cells in the SN after 7 day-MPTP induction,
although treatment with neutralizing antibody in vitro did not
inhibit the CXCR4 expression of proliferated cells of
GFP+CD3+ T cells in vivo . But
it was also sufficient to confirm that the migration and infiltration of
T cells into the SN was at least partially dependent on CXCR4. Ans in
our previous study, we found that after the application of HMGB1 A Box,
the levels of HMGB1 and CXCL12 were both reduced, and the infiltration
of CD3+ T cells in the SN was almost disappear(Tianet al. , 2020). This also suggests that the HMGB1/CXCL12-CXCR4
axis plays an important role in T cell infiltration in PD.
Our previous study revealed the important role of microglia in the
differentiation of Th17. In a pure in vitro co-culture system
that excludes the influence of peripheral immune cells, naïve
CD4+ T cells will not be able to differentiate into
Th17 without the presence of microglia. This time, we used an in
vivo model to confirm that peripheral monocytes also have a significant
effect on Th17 differentiation in the SN in MPTP mouse model. Severalin vitro researches reported that the combination of CXCR4/CXCL12
induces the migration of macrophages, DCs and even microglia (Campanaet al. , 2009; Schiraldi et al. , 2012; Tanabe et
al. , 1997). In contrast, CXCL12 needs to form a complex with HMGB1 to
exert its strong migration-inducing effect(Schiraldi et al. ,
2012). In this study, for the first time, we detected the presence of
HMGB1/CXCL12 in an in vitro model of MPP+induced mouse midbrain cells, and also found evidence that HMGB1 A Box
binds to the complex receptor CXCR4.
Previous studies reported an overall increase in CXCR4 and CXCL12 levels
in the SN during PD(Shimoji et al. , 2009). In this study, we
detected increased HMGB1 and CXCL12 in the serum of PD patients.
Excitingly, we also found that the HMGB1/CXCL12 complex also exists in
the serum of PD patients, and observed binding of CXCL12 and HMGB1 to
CXCR4 on patients’ peripheral blood CD14+ monocytes
and CD3+ T cells. Before, it is reported that CXCR4
was up-regulated in peripheral blood mononuclear cells of PD
patients(Bagheri et al. , 2018). This proved that the
HMGB1/CXCL12-CXCR4 axis related mechanism is also involved in human PD.
In addition, in vitro experiments have long confirmed that CXCL12
induces microglial migration(Tanabe et al. , 1997), and recently,
studies have confirmed that CXCR4 on the surface of microglia in the SN
of A53T mice is also up-regulated(Li et al. , 2019). This suggests
that when dopaminergic neurons are damaged, surrounding microglia may
also aggregate to the injury site through the HMGB1/CXCL12-CXCR4 axis,
but this hypothesis still needs more experimental evidence to confirm.
Interestingly, we found that CXCL12 was significantly negatively
correlated with the age of PD patients, suggesting that targeting the
HMGB1/CXCL12-CXCR4 axis in PD could be more suitable for less old
patients.
In conclusions, our results suggest that neuron-derived HMGB1 plays an
important role in the recruitment of peripheral monocyte-macrophages and
T cells in PD model mice, and peripheral monocytes/macrophages
infiltration in the SN also affects Th17 infiltration and
differentiation. The application of HMGB1 A Box significantly inhibited
the infiltration of peripheral monocytes/macrophages and T cells in the
SN by binding CXCR4 (Fig. 7). This effect may make this protein of great
clinical value at least in the treatment of some PD classification,
which is mainly characterized by neuroinflammation.