Myometrial-derived CXCL12 promotes lipopolysaccharide induced preterm labour by regulating macrophage migration, polarization and function in mice

Abstract

Preterm birth is a major contributor to neonatal mortality and morbidity. Infection results in elevation of inflammation-related cytokines followed by infiltration of immune cells into gestational tissue. CXCL12 levels are elevated in preterm birth indicating it may have a role in preterm labour (PTL); however, the pathophysiological correlations between CXCL12/CXCR4 signalling and premature labour are poorly understood. In this study, PTL was induced using lipopolysaccharide (LPS) in a murine model. LPS induced CXCL12 RNA and protein levels significantly and specifically in myometrium compared with controls (3-fold and 3.5-fold respectively). Highest levels were found just before the start of labour. LPS also enhanced the infiltration of neutrophils, macrophages and T cells, and induced macrophage M1 polarization. In vitro studies showed that condition medium from LPS-treated primary smooth muscle cells (SMC) induced macrophage migration, M1 polarization and upregulated inflammation-related cytokines such as interleukin (IL)-1, IL-6 and tumor necrosis factor alpha (TNF-α). AMD3100 treatment in pregnant mice led to a significant decrease in the rate of PTL (70%), prolonged pregnancy duration and suppressed macrophage infiltration into gestation tissue by 2.5-fold. Further, in-vitro treatment of SMC by AMD3100 suppressed the macrophage migration, decreased polarization and downregulated IL-1, IL-6 and TNF-α expression. LPS treatment in pregnant mice induced PTL by increasing myometrial CXCL12, which recruits immune cells that in turn produce inflammation-related cytokines. These effects stimulated by LPS were completely reversed by AMD3100 through blocking of CXCL12/CXCR4 signalling. Thus, the CXCL12/CXCR4 axis presents an excellent target for preventing infection and inflammation-related PTL.

Keywords: AMD3100; CXCL12; LPS; PTL; SMC; macrophage; uterus.

FIGURE 1

Determination of CXCL12 levels in plasma of mice by ELISA. (A) A significant increase in plasma CXCL12 levels with increasing gestational age (at time points 10.5, 15.5, 18.5 days) and a significant reduction post pregnancy (PP) compared with gestational day 18.5 while increased compared with day 0 (n = 6). (B) Plasma CXCL12 levels significantly increased in LPS‐treated mice compared with PBS treated or to untreated controls (n = 6). Each bar represents the mean ± SEM for data from two individual experiments, and each experiment was performed in triplicate. *p < 0.05 versus GD0; **p < 0.05 versus PP and ^# p < 0.05 versus PBS treated or untreated controls. ELISA, enzyme‐linked immunosorbent assay; LPS, lipopolysaccharide; PBS, phosphate buffered saline; SEM, standard error of mean

FIGURE 2

LPS upregulated CXCL12 expression in uterus. (A) Representative IHC images of CXCL12 expression. Arrow (black) shows the CXCL12 expression in myometrium. Original magnification, 100× and 400×. (B) CXCL12 mRNA levels were significantly increased in uterine tissue obtained from mice treated by LPS. Each bar represents the mean ± SEM for data from three individual experiments and each experiment was performed in duplicate. *p < 0.05 versus PBS. (C) Left: haematoxylin and eosin staining showing structural morphology of uterus, decidua, syncytiotrophoblast and cytotrophoblast treated by PBS and LPS. Right: IHC staining for CXCL12 showing representative images of uterus, decidua, syncytiotrophoblast and cytotrophoblast treated with either PBS or LPS. Original magnification (A) 100× and (B) 400×. (D, E) CXCL12 Expression in macrophages. (D) Fluorescence confocal microscopy analyses of CXCL12 expression in macrophages. Representative images of IF showing DAPI, α‐SMA, CXCL12 expression and merged images for macrophages treated with PBS and LPS. CXCL12 expression is high in LPS‐treated cells. Nuclei were stained by DAPI and are shown in blue. Scale bar: 75 µM. (E) CXCL12 mRNA levels were significantly increased in macrophages treated by LPS compared with PBS. Each bar represents the mean ± SEM for data from three individual experiments and each experiment was performed in duplicate. *p < 0.05 versus PBS. α‐SMA, alpha smooth muscle actin; DAPI, 4,6‐diamidino‐2‐phenylindole; ELISA, enzyme‐linked immunosorbent assay; IF, immunofluorescence; IHC, immunohistochemistry; LPS, lipopolysaccharide; PBS, phosphate buffered saline; SEM, standard error of mean

FIGURE 3

AMD3100 rescued PTL induced by LPS. (A) AMD3100 alone had no effect on the length of gestation. LPS significantly decreased gestational length and this effect was rescued by AMD3100. (B) AMD3100 significantly reduced the rate of PTL induced by LPS. Results shown as mean ± SE. *p < 0.05 versus PBS or AMD3100 while ^# p < 0.05 versus LPS. LPS, lipopolysaccharide; PBS, phosphatecbuffered saline; PTL, preterm labour; SE, standard error

FIGURE 4

LPS‐stimulated engraftment of immune cells into the uterus while AMD3100 inhibited macrophage engraftment. Representative IHC images showing that the number of immune cells engrafted were increased by LPS treatment compared to PBS treated controls. Macrophages (anti‐F4/80); T cells (anti‐CD3) and leukocytes (anti‐elastase) are all increased by LPS treatment, however, only macrophages are decreased after AMD3100 treatment. On the right quantitative analyses demonstrates the significant increase in F4/80^+ve, CD3^+ve and elastase^+ve cells in LPS‐treated mice. AMD3100 inhibited the macrophage (F4/80^+ve) engraftment induced by LPS. Results shown as mean ± SE. *p < 0.05 versus PBS while ^# p < 0.05 versus LPS. Original magnification 400×. IHC, immunohistochemistry; LPS, lipopolysaccharide; PBS, phosphate buffered saline; SE, standard error

FIGURE 5

In vivo: AMD3100 inhibited LPS‐stimulated M1 macrophage polarization. (A, B) Fluorescence confocal microscopy analysis of M1 macrophage polarization. Representative images of IF showing DAPI, F4/80^+ve, iNOS^+ve and merged images. Nuclei were stained by DAPI and are shown in blue. F4/80 marks macrophages while iNOS is a marker of M1 macrophages. (A) Uterus tissue from mice treated with PBS, LPS or LPS + AMD3100 respectively. (B) Macrophages treated with PBS, LPS, CXCL12, LPS‐SMC‐CM and LPS‐SMC‐CM + AMD3100 respectively. (C) Quantitative analyses of macrophage polarization (iNOS^+ve cells) in uterus showing significant increase in polarization in LPS‐treated macrophages compared to PBS. Treatment with AMD3100 inhibited the polarization stimulated by LPS. (D) Quantitative analyses of macrophage polarization (iNOS^+ve cells) in vitro showing significant increase in polarization in LPS, CXCL12 and LPS‐SMC‐CM treated macrophages compared with PBS treatment. AMD3100 inhibited the polarization stimulated by LPS. Results shown as mean ± SE. *p < 0.05 versus PBS while ^# p < 0.05 versus LPS‐SMC‐CM. Scale bar: 75 μm. DAPI, 4,6‐diamidino‐2‐phenylindole; IF, immunofluorescence; IHC, immunohistochemistry; LPS, lipopolysaccharide; PBS, phosphate buffered saline; SE, standard error; SMC‐CM, smooth muscle cell conditioned media

FIGURE 6

AMD3100 inhibited USMCs‐CM induced macrophage migration through CXCL12/CXCR4 signalling. (A) Macrophage attraction was determined by migration assay. Macrophages migrated significantly when treated with CXCL12 but not LPS. Conditioned medium from SMCs increased migration. Conditioned media from SMCs treated with LPS further increased migration. AMD3100 significantly inhibited macrophage migration stimulated by LPS‐treated SMC‐conditioned medium. Data are shown as a number of macrophages migrated. Each bar represents the mean ± SEM for data from two individual experiments and each experiment was performed in triplicate. *p < 0.05 versus PBS, **p < 0.05 versus PBS‐SMC‐CM and ^# p < 0.05 versus LPS‐SMC‐CM. (B, C) AMD3100 inhibited LPS stimulation of proinflammatory cytokines in M1 macrophage. (B) qRT‐PCR results showing significant increase in mRNA levels of IL‐1, IL‐6, TNF‐α induced by LPS or CXCL12 treatments compared with PBS. (C) qRT‐PCR results showing significant induction of IL‐1, IL‐6, TNF‐α expression stimulated by conditioned medium from smooth muscle cells treated with LPS. The effect of conditioned media on cytokine induction is blocked by AMD3100. Each bar represents the mean ± SEM for data from three individual experiments, and each experiment was performed in duplicate. *p < 0.05 versus PBS or PBS‐SMC‐CM. IHC, immunohistochemistry; IL, inyterleukin; LPS, lipopolysaccharide; PBS, phosphate buffered saline; SEM, standard error of mean; SMC‐CM, smooth muscle cell conditioned media; qRT‐PCR, quantitative real‐time polymerase chain reaction; TNF‐α, tumor necrosis factor alpha; USMC‐CM, uterine smooth muscle cell conditioned media