The time-of-day of myocardial infarction onset affects healing through oscillations in cardiac neutrophil recruitment

Abstract

Myocardial infarction (MI) is the leading cause of death in Western countries. Epidemiological studies show acute MI to be more prevalent in the morning and to be associated with a poorer outcome in terms of mortality and recovery. The mechanisms behind this association are not fully understood. Here, we report that circadian oscillations of neutrophil recruitment to the heart determine infarct size, healing, and cardiac function after MI Preferential cardiac neutrophil recruitment during the active phase (Zeitgeber time, ZT13) was paralleled by enhanced myeloid progenitor production, increased circulating numbers of CXCR2(hi) neutrophils as well as upregulated cardiac adhesion molecule and chemokine expression. MI at ZT13 resulted in significantly higher cardiac neutrophil infiltration compared to ZT5, which was inhibited by CXCR2 antagonism or neutrophil-specific CXCR2 knockout. Limiting exaggerated neutrophilic inflammation at this time point significantly reduced the infarct size and improved cardiac function.

Keywords: circadian rhythm; fibrosis; myocardial infarction healing; neutrophils; progenitors.

Figure 1. Circadian oscillations of neutrophils, adhesion molecules, and chemokines in the mouse heart at steady state

1. Baseline blood counts of granulocytes. One‐way ANOVA; n = 8 mice per ZT; *P = 0.0001 ZT5 versus ZT13.

2. Flow cytometric quantification of neutrophils in digested hearts. The representative dot plots show the gating strategy for cardiac neutrophils (CD45⁺Ly6G⁺ CD11b⁺) at ZT5. One‐way ANOVA; n = 5 mice per ZT; *P = 0.0001 ZT5 versus ZT13.

3. Representative immunostainings for neutrophils in the myocardium (left ventricle), identified as Ly6G positive (20× magnification).

4. Cardiac mRNA expression levels normalized to HPRT. One‐way ANOVA; n = 5 mice per ZT; ZT5 versus ZT17: *P = 0.0064 (CXCL1), *P = 0.0007 (CXCL2), *P = 0.0007 (CXCL5), *P = 0.0001 (ICAM‐1), *P = 0.0009 (VCAM‐1), *P = 0.0181 (CCL3), *P = 0.0360 (CCL5).

5. Mean fluorescence intensity (MFI) of CXCR2 expression by neutrophils in blood and heart. One‐way ANOVA; n = 3 mice for ZT1, ZT17, ZT21 and n = 5 for ZT5, ZT9, ZT13; ZT5 versus ZT13: *P = 0.0425 (blood), *P = 0.0078 (heart).

Data information: All data are expressed as mean ± SEM.

Figure 2. Inflammatory response after MI during active (ZT13) or resting phase (ZT5)

1. Flow cytometric analysis of neutrophils in hearts and representative immunostainings for neutrophils in the infarct area, identified as Ly6G positive (5× and 20× magnifications). Two‐way ANOVA followed by Bonferroni post hoc test; n = 5 mice for no MI, n = 5 for 12 h post‐MI, n = 3 for 24 h post‐MI, and n = 3 for 72 h post‐MI in both ZT groups; ZT5 versus ZT13: *P = 0.0161 (12 h MI), *P = 0.003 (24 h MI).

2. Flow cytometric analysis of neutrophils in bone marrow. Two‐way ANOVA followed by Bonferroni post hoc test; n = 5 mice for no MI, n = 5 for 12 h post‐MI, n = 4 for 24 h post‐MI, and n = 4 for 72 h post‐MI in both ZT groups; ZT5 versus ZT13: *P = 0.0471 (12 h MI), *P = 0.0035 (24 h MI).

3. Blood counts of granulocytes. Two‐way ANOVA followed by Bonferroni post hoc test; n = 9 mice for no MI, n = 10 for 12 h post‐MI, n = 13 for 24 h post‐MI, and n = 10 for 72 h post‐MI in both ZT groups; ZT5 versus ZT13: *P = 0.0364 (24 h MI).

4. Flow cytometric analysis of neutrophils in hearts and bone marrow under inverted light cycle conditions. Two‐way ANOVA followed by Bonferroni post hoc test; n = 3 mice for no MI in both ZT groups, n = 5 for ZT5 and n = 6 for ZT13 at 24 h post‐MI; ZT5 versus ZT13: *P = 0.0147 (24 h MI).

5. Representative gating strategy for GMPs in the bone marrow, identified as lineage negative (CD11b⁻, Gr1⁻, B220⁻, CD3⁻, and Ter119⁻) and Sca‐1⁻, c‐kit⁺, CD16/32⁺, and CD34⁺. Flow cytometric quantification of GMP in the bone marrow. Two‐way ANOVA followed by Bonferroni post hoc test; n = 4 mice for no MI and 24 h post‐MI in both groups; ZT5 versus ZT13: *P = 0.0077 (no MI), *P = 0.0013 (24 h MI).

6. CXCL12 levels in bone marrow lavage. Two‐way ANOVA followed by Bonferroni post hoc test; n = 7 mice for no MI and n = 6 for 24 h post‐MI in both groups; ZT5 versus ZT13: *P = 0.0077 (no MI), *P = 0.0013 (24 h MI).

7. Plasma levels of pro‐inflammatory cytokines and chemokines. Two‐way ANOVA followed by Bonferroni post hoc test; n = 11 mice for no MI in both ZT groups, n = 7 for ZT5 and n = 8 for ZT13 at 24 h post‐MI; ZT5 versus ZT13: *P = 0.0271 (CXCL12, no MI), *P = 0.0108 (TNF‐α, 24 h MI), *P = 0.001 (G‐CSF, 24 h MI), *P = 0.005 (CXCL1, 24 h MI), *P = 0.0005 (CXCL2, 24 h MI), *P = 0.0016 (CCL3 24 h MI), *P = 0.0144 (CCL5, 24 h MI).

Data information: All data are expressed as mean ± SEM.

Figure 3. MI during active phase (ZT13) leads to infarct expansion and reduced cardiac function

1. Permanent LAD occlusion was performed at ZT5 or ZT13. TTC staining (white, infarct; red, vital myocardium) and quantification of infarct size normalized to the left ventricle (LV). Student's t‐test; n = 4 mice for ZT5 and n = 5 for ZT13 MI; *P = 0.0041.

2. Plasma troponin I levels 24 h after MI. Student's t‐test; n = 4 mice for ZT5 and n = 5 for ZT13 MI; *P = 0.0007.

3. Flow cytometric analysis of dead cardiomyocytes (CD45⁻, Zombie⁺) 24 h after MI. Student's t‐test; n = 3 mice in both groups; *P = 0.0072.

4. Survival rates after MI and cause of death. Log‐rank test; n = 87 mice in both groups; *P = 0.0006.

5. Myofibroblasts within infarcts were quantified by alpha‐smooth muscle actin (αSMA) staining as ratio between stained and total area of random fields. Student's t‐test; n = 4 mice in both groups; *P = 0.0134.

6. Masson's trichome staining of fibrosis (blue, collagen; red, vital myocardium) and quantification relative to total LV (*P = 0.0095) as well as LV anterior wall thickness (*P = 0.0068) 7 days after MI. Student's t‐test; n = 6 mice for ZT5 and n = 7 for ZT13 MI.

7. Analysis of relative collagen type I content identified by Sirius Red staining 7 days after MI. Student's t‐test; n = 4 mice for ZT5 and n = 5 for ZT13 MI; *P = 0.0175.

8. Echocardiographic assessment of ejection fraction (EF), end‐systolic volume (ESV), and end‐diastolic volume (EDV). Two‐way ANOVA; n = 6 mice for no MI for both groups, n = 6 for ZT5 and n = 9 for ZT13 at 72 h post‐MI, n = 8 for both groups at 7 days post‐MI, and n = 7 for ZT5 and n = 5 for ZT13 at 14 days post‐MI; ZT5 versus ZT13: *P = 0.0001 (EF, 3 days), *P = 0.0042 (EF, 7 days), *P = 0.0121 (EF, 14 days); *P = 0.0253 (ESV, 3 days), *P = 0.0421 (ESV, 7 days), *P = 0.0005 (ESV, 14 days); *P = 0.0053 (EDV, 14 days).

Data information: All data are expressed as mean ± SEM.

Figure 4. Limiting neutrophilic inflammation during active phase (ZT13) reduces MI damage

1. Permanent LAD occlusion was performed at ZT13 followed by injection of isotype or Ly6G antibody 45 min after surgery and then every 24 h.

2. Flow cytometric analysis of cardiac neutrophils 24 h after ZT13 MI. The dotted line indicates cardiac neutrophil counts 24 h after ZT5 MI, as shown in Fig 2A. Student's t‐test; n = 3 mice for isotype and n = 5 for Ly6G injected mice; *P = 0.0004.

3. Plasma TNF‐α levels 24 h after ZT13 MI. Student's t‐test. n = 8 independent samples for isotype and n = 6 for Ly6G injected mice; *P = 0.0019.

4. Infarct size relative to left ventricular area (LV) and plasma troponin I levels 24 h after ZT5 or ZT13 MI. Student's t‐test; for infarct size, n = 3 mice in both groups for ZT5 and n = 4 mice for isotype and n = 3 mice for Ly6G at ZT13; isotype versus Ly6G: *P = 0.0158 (ZT13). For troponin levels, n = 8 mice for isotype and n = 6 mice for Ly6G with ZT13 MI; *P = 0.0012.

5. Masson's trichrome staining of fibrosis (blue, collagen; red, vital myocardium) and quantification relative to total area of the LV (*P = 0.0084); morphometric quantification of the LV anterior wall thickness (*P = 0.0027) 7 days after ZT13 MI. Student's t‐test; n = 5 mice for isotype and n = 4 mice for Ly6G.

6. Analysis of collagen type I fibers within infarcts identified by Sirius Red staining 7 days after ZT13 MI. Student's t‐test; n = 5 mice for isotype and n = 4 mice for Ly6G; *P = 0.0375.

7. Echocardiographic measurement of ejection fraction (EF), end‐systolic volume (ESV) and end‐diastolic volume (EDV) before and after ZT13 MI. Two‐way ANOVA; n = 6 mice for no MI in both groups, n = 9 for isotype and n = 5 for Ly6G at 72 h post‐MI, and n = 8 for isotype and n = 4 for Ly6G at 7 days post‐MI. Isotype versus Ly6G: *P = 0.0001 (EF, 3 days), *P = 0.0227 (EF, 7 days); MI *P = 0.0133 (ESV, 3 days), *P = 0.0004 (ESV, 7 days); *P = 0.0017 (EDV, 7 days).

Data information: All data are expressed as mean ± SEM.

Figure 5. Antagonism or deficiency of CXCR2 inhibits enhanced cardiac neutrophil accumulation during active phase (ZT13)

1. Representative flow cytometric analysis and quantification of CXCR2^high neutrophils in the blood after ZT5 and ZT13 MI. Two‐way ANOVA followed by Bonferroni post hoc test; n = 4 mice for all time points in both groups; ZT5 versus ZT13: *P = 0.0001 (4 h), *P = 0.0001 (8 h), *P = 0.0447 (12 h).

2. Percentage and mean fluorescence intensity (MFI) of CXCX2 expression by cardiac neutrophils 12 h post‐MI after ZT5 and ZT13 MI. Student's t‐test; n = 4 mice in both groups; *P = 0.0001.

3. Percentage of CXCR2^high neutrophils in the blood 24 h after ZT5 or ZT13 MI in mice receiving CXCR2 antagonist SB225002 or vehicle. Two‐way ANOVA followed by Bonferroni post hoc test; n = 4 mice in both groups; DMSO versus SB225002: *P = 0.0005 (ZT13).

4. Flow cytometric quantification of neutrophils in hearts 24 h after ZT5 or ZT13 MI in mice receiving CXCR2 antagonist SB225002 or vehicle. Two‐way ANOVA followed by Bonferroni post hoc test; n = 4 mice in both groups; DMSO versus SB225002: *P = 0.0079 (ZT13).

5. Flow cytometric quantification of neutrophils in bone marrow 24 h after ZT5 or ZT13 MI in mice receiving CXCR2 antagonist SB225002 or vehicle. Two‐way ANOVA followed by Bonferroni post hoc test; n = 4 mice in both groups; DMSO versus SB225002: *P = 0.0002 (ZT13).

6. MFI of CXCR2 expression at ZT5 in the blood 24 h after MI in wild‐type (WT) and CXCR2 KO mice. Student's t‐test. n = 6 mice in both groups; *P = 0.0001.

7. Percentage of CXCR2^hi blood neutrophils 24 h after ZT5 or ZT13 MI in WT and CXCR2 KO mice. Two‐way ANOVA followed by Bonferroni post hoc test; for ZT5 n = 6 mice in both groups, for ZT13 n = 7 WT mice and n = 5 CXCR2 KO mice; WT versus CXCR2 KO: *P = 0.0001 (ZT13).

8. Flow cytometric quantification of neutrophils in hearts 24 h after MI in ZT5 and ZT13‐operated WT and CXCR2 KO mice. Two‐way ANOVA followed by Bonferroni post hoc test; for ZT5 n = 6 mice in both groups, for ZT13 n = 7 WT mice and n = 5 CXCR2 KO mice; WT versus CXCR2 KO: *P = 0.0461 (ZT13), ns = not significant.

Data information: All data are expressed as mean ± SEM.

Figure 6. Antagonism of CXCR2 inhibits enhanced cardiac neutrophil accumulation after ischemia/reperfusion during active phase (ZT13)

1. Schematic representation of transient ischemia and reperfusion protocol, performed at ZT5 and ZT13. The CXCR2 antagonist SB225002 or vehicle was injected 5 min before reopening the LAD.

2. Percentage of CXCR2^hi neutrophils in the blood 24 h after ZT5 or ZT13 MI in mice receiving CXCR2 antagonist SB225002 or vehicle. Two‐way ANOVA followed by Bonferroni post hoc test; n = 5 mice in both groups at ZT5, n = 5 mice for vehicle, and n = 6 mice for SB225002 at ZT13; DMSO versus SB225002: *P = 0.0046 (ZT13); ZT5 versus ZT13: *P = 0.0153 (DMSO).

3. Flow cytometric quantification of cardiac neutrophils 24 h after ZT5 or ZT13 MI in mice receiving CXCR2 antagonist SB225002 or vehicle. Two‐way ANOVA; n = 5 mice in both groups at ZT5, n = 5 mice for vehicle, and n = 6 mice for SB225002 at ZT13; DMSO versus SB225002: *P = 0.0006 (ZT13); ZT5 versus ZT13: *P = 0.0001 (DMSO).

4. Flow cytometric quantification of neutrophils in bone marrow 24 h after ZT5 or ZT13 MI in mice receiving CXCR2 antagonist SB225002 or vehicle. Two‐way ANOVA followed by Bonferroni post hoc test; n = 5 mice in both groups at ZT5, n = 5 mice for vehicle, and n = 6 mice for SB225002 at ZT13; DMSO versus SB225002: *P = 0,0486 (ZT13).

Data information: All data are expressed as mean ± SEM.