Bone marrow-derived microglia infiltrate into the paraventricular nucleus of chronic psychological stress-loaded mice

Abstract

Background: Microglia of the central nervous system act as sentinels and rapidly react to infection or inflammation. The pathophysiological role of bone marrow-derived microglia is of particular interest because they affect neurodegenerative disorders and neuropathic pain. The hypothesis of the current study is that chronic psychological stress (chronic PS) induces the infiltration of bone marrow-derived microglia into hypothalamus by means of chemokine axes in brain and bone marrow.

Methods and findings: Here we show that bone marrow-derived microglia specifically infiltrate the paraventricular nucleus (PVN) of mice that received chronic PS. Bone marrow derived-microglia are CX3CR1(low)CCR2(+)CXCR4(high), as distinct from CX3CR1(high)CCR2(-)CXCR4(low) resident microglia, and express higher levels of interleukin-1β (IL-1β) but lower levels of tumor necrosis factor-α (TNF-α). Chronic PS stimulates the expression of monocyte chemotactic protein-1 (MCP-1) in PVN neurons, reduces stromal cell-derived factor-1 (SDF-1) in the bone marrow and increases the frequency of CXCR4(+) monocytes in peripheral circulation. And then a chemokine (C-C motif) receptor 2 (CCR2) or a β3-adrenoceptor blockade prevents infiltration of bone marrow-derived microglia in the PVN.

Conclusion: Chronic PS induces the infiltration of bone marrow-derived microglia into PVN, and it is conceivable that the MCP-1/CCR2 axis in PVN and the SDF-1/CXCR4 axis in bone marrow are involved in this mechanism.

Figure 1. Chronic psychological stress (PS) increases the infiltration of bone marrow-derived microglia into the PVN.

(A, B) After chronic PS-loading for five days, the numbers of GFP-positive cells (green) in the PVN were significantly increased compared with sham-treated mice. Data are expressed as mean ± sem (PS: n = 6−7, sham: n = 5−11). # P < 0.05 with ANOVA followed by Tukey’s multiple comparison. *P < 0.05 with two-tailed Student’s t-test. (C) GFP-positive cells (green) overlapped with Iba-1 (red) in PVN from chronic PS-loaded mice but did not overlap with GFAP (red) in PVN from chronic PS-loaded mice. (D) Photograph of mice received irradiation with head protection and non-irradiated mice. (E, F) After chronic PS-loading for five days, the number of GFP-positive cells (green) in the PVN of mice received the irradiation with head protection was significantly increased compared with sham-treated mice. Data are expressed as mean ± sem. **P < 0.001 for PS (n = 8) to sham (n = 4) with two-tailed Student’s t-test. (G) GFP-positive cells (green) overlapped with Iba-1 (red) in PVN from chronic PS-loaded mice received the irradiation with head protection. PVN shown by dotted line. Scale bars: 50 µm.

Figure 2. Isolation of bone marrow-derived microglia and resident microglia from hypothalamic tissue and comparison of expression of various molecules in chronic PS-loaded and sham-treated mice.

(A) Representative FACS chart in chronic PS-loaded mice with whole body radiation and the number of isolated GFP^-CD45^low (resident microglia) and GFP⁺CD45^low (bone marrow-derived microglia) from mice with whole body radiation and the radiation with head protection (n = 4−6). Total events in FACS were 20000. Data are expressed as mean ± sem. *P < 0.05 with two-tailed Student’s t-test. (B−E) mRNA expression of chemokine receptors: CCR2 (n = 6−8), CX₃CR1 (n = 3−5), CXCR4 (n = 3−5) (B), exiting amino acid transporters: EAAT1 (n = 4−5), EAAT2 (n = 3−4) (C), purinergic P2X receptors: P2X4 (n = 3) and P2X7 (n = 3) and P2Y receptors: P2Y1 (n = 3), P2Y12 (n = 3−4) (D), IL-1β (n = 4−6) and TNF-α (n = 3−5) (E). (F) The length of axis of GFP⁺Iba-1⁺ microglia (bone marrow-derived microglia, BMDM) and GFP^-Iba-1⁺ microglia (resident microglia. RM) in chronic PS-loaded and sham mice (n = 4). Scale bars: 10 µm. Data are expressed as mean ± sem. *P < 0.05, **P < 0.01 with ANOVA followed by Tukey’s multiple comparison.

Figure 3. MCP-1/CCR2 axis in hypothalamus and peripheral blood, and effects of CCR2 blockade on the infiltration of bone marrow-derived microglia into the PVN and anxiety-like behavior induced by chronic PS.

(A) mRNA expression of chemokines in hypothalamic tissue from chronic PS-loaded and sham-treated mice (n = 4). Data are expressed as mean ± sem. *P < 0.05 with two-tailed Student’s t-test. (B) Immunofluorescence staining with MCP-1 (red) and NeuN (pink) in PVN from chronic PS-loaded and sham-treated mice. Arrows indicate MCP-1⁺NeuN⁺ cells and arrow heads indicate MCP-1^-NeuN⁺ cells. Scale bars: 20 µm. Data are expressed as mean ± sem (n = 7). *P < 0.05 with two-tailed Student’s t-test. (C) Immunofluorescence staining with MCP-1 (red) and GFAP (pink) in PVN from chronic PS-loaded and sham-treated mice. Scale bars: 20µm. Data are expressed as mean ± sem (n = 4−6). *P < 0.05 with two-tailed Student’s t-test. (D) Frequency of GFP⁺CCR2⁺ cells in the peripheral blood obtained by FACS (n = 4−6). Data are expressed as mean ± sem. *P < 0.05 with two-tailed Student’s t-test. (E) The number of CCR2⁺ cells in the hypothalamus obtained by FACS (n = 4). Total events in FACS were 20000. Data are expressed as mean ± sem. *P < 0.05 with ANOVA followed by Tukey’s multiple comparison. (F) Effects of a CCR2 antagonist, RS102895, on infiltration of bone marrow-derived microglia in PVN. PVN is shown by dotted line. Scale bars: 50 µm. Number of GFP-positive cells in a section including PVN is shown. (G) Effect of RS102895 on the anxiety-like behavior induced by chronic PS. Data are expressed as mean ± sem. *P < 0.05 with ANOVA followed by Tukey’s multiple comparison (n = 4).

Figure 4. Expression of SDF-1 in the bone marrow and frequency of GFP⁺CXCR4⁺ cells in peripheral blood.

Effects of β₃-adrenergic blockade on the infiltration of bone marrow-derived microglia in the PVN. (A) Chronic PS decreased the SDF-1 concentration in the bone marrow (n = 3−4) and increased the frequency of CXCR4⁺ monocytes in the peripheral blood (n = 4−6). (B, C) β₃-adrenoceptor antagonist SR59230A blocks the infiltration of bone marrow-derived microglia in the PVN. Number of GFP-positive cells was counted in a section including PVN. Data are expressed as mean ± sem. *P < 0.05 (n = 3−11). PVN shown by dotted line. Scale bars: 50 µm.

Figure 5. Three dimensional image of immunofluorescence staining of IL-1β, pNMDA receptor, and IL-1 receptor in PVN from chronic PS-loaded mouse.

(A) GFP-positive cells (green) overlapped with IL-1β (red). (B, C) Positive reactions for pNMDAR (pink) and IL-1R (pink) detected on the membrane of neurons (red) adjacent to bone marrow-derived microglia (green) in PVN. Scale bars: 10 µm.

Figure 6. Schematic drawing of brain-bone marrow axis in chronic psychological stress conditions.

When brain is exposed to chronic PS (1), the information is mediated to bone marrow through adrenergic nerves (2). Bone marrow niche cells innervated with sympathetic nerves decrease the expression of SDF-1 through β₃-adrenergic receptor (3). CXCR4^high monocytes egress to peripheral circulation from the bone marrow by reduction of SDF-1 at the bone marrow niche (4). Neurons in the PVN express MCP-1 under chronic PS condition, then accelerate infiltration of CCR2⁺ bone marrow-derived microglia into PVN (5).