Aging-Impaired Filamentous Actin Polymerization Signaling Reduces Alveolar Macrophage Phagocytosis of Bacteria

Abstract

In elderly patients, bacterial infection often causes severe complications and sepsis. Compared to younger patients, older patients are more susceptible to sepsis caused by respiratory infection. Macrophage (Mϕ) phagocytosis of bacteria plays a critical role in the clearance of pathogens and the initiation of immune responses. It has been suggested that Mϕ exhibit age-related functional alterations, including reduced chemotaxis, phagocytosis, antibacterial defense, and the ability to generate reactive oxygen species. However, the mechanisms behind these changes remain unclear. The present study sought to determine changes in bacterial phagocytosis in aging alveolar Mϕ (AMϕ) and the underlying mechanisms. We show that bacteria initiate cytoskeleton remodeling in AMϕ through interaction with macrophage receptor with collagenous structure (MARCO), a bacterial scavenger receptor. This remodeling, in turn, promotes enhanced cell surface expression of MARCO and bacterial phagocytosis. We further demonstrate that Rac1-GTP mediates MARCO signaling and activates actin-related protein-2/3 complex, an F-actin nucleator, thereby inducing F-actin polymerization, filopodia formation, and increased cell surface expression of MARCO, all of which are essential for the execution of bacteria phagocytosis. However, AMϕ isolated from aging mice exhibit suppressed Rac1 mRNA and protein expression, which resulted in decreases in Rac1-GTP levels and actin-related protein-2/3 activation, as well as subsequent attenuation of F-actin polymerization, filopodia formation, and cell surface expression of MARCO. As a result, bacterial phagocytosis in aging AMϕ is decreased. This study highlights a previously unidentified mechanism by which aging impairs Mϕ phagocytosis of bacteria. Targeting these pathways may improve outcomes of bacterial infection in elderly patients.

Figure 1. Aging impairs AMϕ phagocytosis of E. coli

(A and B) Uptake of Alexa Fluor 488 conjugated E. coli in AMϕ collected from BALF from young or aging mice at 1 h after E. coli administration (4 × 10⁸ bacteria cells/kg B.W., i.t.). (A) F4/80 positive cells were analyzed by flow cytometry and phagocytosis of E. coli was measured by mean fluorescence intensity (MFI). The graph shows mean ± SEM of MFI, n=3, Statistical analysis using two-way ANOVA, Interaction p = 0.0016. Tukey’s multiple comparisons test shows that E. coli i.t./Young vs. E. coli i.t./Aging, p = 0.0008. (B) Confocal microscopy of Alexa Fluor 488-conjugated E. coli (green) taken up by AMϕ. The graph shows mean ± SEM of percentage (%) of the AMϕ that phagocytosed E. coli, n=3. Statistical analysis using two-way ANOVA, Interaction p = 0.0006. Tukey’s multiple comparisons test shows that E. coli i.t./Young vs. E. coli i.t./Aging, p = 0.0003. (C) Pulmonary E. coli bacterial levels [colony forming units (CFU/ml)]. Young and aging mice were administered E. coli (strain BL21, 4 × 10⁸ bacteria cells/kg B.W., i.t.), and at 2 h after treatment, BALF was collected from the mice. The bacteria CFU/ml of BALF were calculated. The graph shows median with interquartile range of BALF CFU/ml, n=6. Statistical analysis using independent-samples Mann-Whitney U test, E. coli i.t./Young vs. E. coli i.t./Aging p = 0.026. (D) In vitro uptake of Alexa Fluor 488-conjugated E. coli [multiplicity of infection (MOI) =4] by AMϕ for time points up to 120 min harvested from young or aging mice and analyzed by flow cytometry and calculated by MFI. The graph shows mean ± SEM of Alexa Fluor 488-conjugated E. coli MFI, n=3. Statistical analysis using two-way ANOVA, Interaction p = 0.0099. Tukey’s multiple comparisons test shows that 60min/Young vs. 60min/Aging, p = 0.0027; 120min/Young vs. 120min/Aging, p = 0.0254. (E) Confocal images showed young or aging AMϕ treated with Alexa Fluor 488-conjugated E. coli (MOI=4) for 0 to 60 min. The graph shows mean ± SEM of percentage (%) of the AMϕ that phagocytosed E. coli, n=3. Statistical analysis using two-way ANOVA, Interaction p = 0.0002. Tukey’s multiple comparisons test shows that 30min/Young vs. 30min/Aging, p = 0.0010; 60min/Young vs. 60min/Aging, p < 0.0001. The asterisks indicate that *p < 0.05 or **p < 0.01.

Figure 2. Upregulated MARCO cell surface expression is required for AMϕ phagocytosis of E. coli

(A) Young and aging AMϕ were treated with Alexa Fluor 488-conjugated E. coli (MOI=4) for up to 60 min. Cell surface MARCO stained with PE-MARCO antibody was measured by flow cytometry and expressed as mean fluorescence intensity (MFI). PE-Isotype staining was used as staining control. The graph shows mean ± SEM of MFI, n=3. Statistical analysis using two-way ANOVA, Interaction p = 0.0014. Tukey’s multiple comparisons test shows that 0min/Young vs. 60min/Young, p = 0.0001; 60min/Young vs. 60min/Aging, p = 0.0003. (B) Young and aging AMϕ were treated with E. coli (strain BL21, MOI=4) for up to 60 min. Confocal immunofluorescence images show MARCO staining. The MARCO distribution in AMϕ was analyzed by software Image J. (C) AMϕ were transfected with non-specific small interfering RNA (siNC) or small interfering RNA against MARCO (siMARCO) for 36 h. Total MARCO expression was measured by Western Blot. The graph shows the mean ± SEM of intensity of MARCO/GAPDH, n=3. Statistical analysis using independent-samples two-tailed Student’s t-test. siNC vs. siMARCO, p < 0.01. (D) AMϕ were transfected with siNC or siMARCO for 36 h, and then were treated with E. coli (MOI=4) for 60 min. The cells were stained with PE-Isotype or PE-MARCO. Cell surface MARCO on the AMϕ was determined by flow cytometry. The graph depicts mean ± SEM of MARCO MFI, n=3. Statistical analysis using two-way ANOVA, Interaction p = 0.0365. Tukey’s multiple comparisons test shows that Control/siNC vs. Control/siMARCO, p = 0.0365; E. coli/siNC vs. E. coli/siMARCO, p = 0.0005. (E) AMϕ were transfected with siNC or siMARCO for 36 h, and were then treated with Alexa Fluor 488- conjugated E. coli (MOI=4) for 60 min. The AMϕ phagocytosis of E. coli was measured by flow cytometry. The graph depicts mean ± SEM of intracellular Alexa Fluor 488-conjugated E. coli MFI, n=3. Statistical analysis using two-way ANOVA, Interaction p < 0.0001. Tukey’s multiple comparisons test shows that E. coli/siNC vs. E. coli/siMARCO, p < 0.0001. (F) AMϕ were treated with E. coli (MOI=4) for 60 min. The cells were then stained with PE-Isotype, PE-MARCO or PE-FcγRI, and cell surface MARCO or FcγRI were detected by flow cytometry. The graph depicts mean ± SEM of MFI of MARCO or FcγRI, n=3. Statistical analysis using independent-samples two-tailed Student’s t-test. MARCO/Control vs. MARCO/siMARCO, p < 0.05. (G) AMϕ were transfected with siNC or siMARCO for 36 h, and then were treated with latex beads-rabbit IgG-PE complex in 1/100 (v/v) for up to 2 hours. The intracellular latex beads-rabbit IgG-PE complex were measured by flow cytometry. The graph depicts mean ± SEM of MFI, n=3. Statistical analysis using two-way ANOVA, Interaction p = 0.8631. Tukey’s multiple comparisons test shows that 1h/siNC vs. 1h/siMARCO, p = 0.9784; 2h/siNC vs. 2h/siMARCO, p > 0.9999. The asterisks indicate that *p < 0.05 or **p < 0.01. NS, no significant difference.

Figure 3. Actin polymerization signaling mediates MARCO cell surface expression and phagocytosis of E. coli

(A) AMϕ were treated with Alexa Fluor 488-conjugated E. coli (MOI=4) for 0 to 60 min, and phagocytosis of E. coli were measured by flow cytometry and expressed as MFI. The graph illustrates mean ± SEM of MFI of intracellular Alexa Fluor 488-conjugated E. coli, n=3. Statistical analysis using one-way ANOVA, over all ANOVA p = 0.0089. Tukey’s multiple comparisons test shows that E. coli/15min vs. Control, p =0.0121; E. coli/30min vs. Control, p =0.0176; E. coli/60min vs. Control, p =0.0109. (B) AMϕ were pretreated with Vehicle (DMSO) or NSC23766 (Rac1 inhibitor, 30 µM), Wiskostatin (N-WASP inhibitor, 30 µM), or CK666 (Arp2/3 complex inhibitor, 30 µM) for 20 min followed by treatment with E. coli (MOI=4) for 60 min. Cells were then stained with PE-Isotype or PE-MARCO, and cell surface MARCO was measured by flow cytometry and expressed as MFI. The graph shows mean ± SEM of MARCO MFI, n=3. Statistical analysis using one-way ANOVA, over all ANOVA p = 0.0035. Tukey’s multiple comparisons test shows that E. coli/Vehicle vs. Control, p =0.0409; E. coli/NSC23766 vs. E. coli/Vehicle, p =0.0231; E. coli/Wiskostatin vs. E. coli/Vehicle, p =0.0281; E. coli /CK666 vs. E. coli/Vehicle, p =0.0452. (C) AMϕ were pretreated with Vehicle (DMSO) or NSC23766 (Rac1 inhibitor, 30 µM), Wiskostatin (N-WASP inhibitor, 30 µM), or CK666 (Arp2/3 complex inhibitor, 30 µM) for 20 min. The cells were then treated with Alexa Fluor 488-conjugated E. coli (MOI=4) for 60 min followed by the measurement of intracellular Alexa Fluor 488-conjugated E. coli by flow cytometry and expressed as MFI. The graph shows mean ± SEM of intracellular MFI of Alexa Fluor 488-conjugated E. coli, n=3. Statistical analysis using one-way ANOVA, over all ANOVA p = 0.0014. Tukey’s multiple comparisons test shows that E. coli/Vehicle vs. Control, p =0.0118; E. coli/NSC23766 vs. E. coli/Vehicle, p =0.0157; E. coli/Wiskostatin vs. E. coli/Vehicle, p =0.0115, E. coli/CK666 vs. E. coli/Vehicle, p =0.0267. The asterisk indicates that *p < 0.05.

Figure 4. Aging impairs AMϕ filopodia formation and decreases phagocytosis of E. coli

(A) Confocal images show AMϕ filopodia formation. AMϕ were treated with E. coli (MOI=4), BLP (1 µg/ml), or LPS (1 µg/ml) for 1 h, and stained with phalloidin (for F-actin) and Hoechst (nuclear stain). The phalloidin staining was outlined by using software Image J. The arrows indicate the filopodia on AMϕ. The graph depicts mean ± SEM of number of the filopodia in AMϕ, n=5. Statistical analysis using one-way ANOVA, over all ANOVA p = 0.0028. Tukey’s multiple comparisons test shows that E. coli vs. Control, p =0.0061; BLP vs. Control, p =0.0019; LPS vs. Control, p =0.0089. (B) Confocal images show AMϕ filopodia formation in young and aging mice. Young and aging mice were treated with Alexa Fluor 488-conjugated E. coli (4 × 10⁸ bacteria cells/kg B.W. i.t.) for 30 min, and AMϕ were then collected from BALF from the mice, and stained with phalloidin and Hoechst. The phalloidin staining was outlined by using software Image J. The arrows indicate the filopodia in AMϕ. The graph depicts mean ± SEM of the number of filopodia in AMϕ. n=5. Statistical analysis using two-way ANOVA, Interaction p < 0.0001. Tukey’s multiple comparisons test shows that Young/E. coli vs. Sham, p < 0.0001; Young/E. coli vs. Age/E. coli, p < 0.0001. (C) Confocal images showing dynamic changes in filopodia formation in young and aging AMϕ following the treatment of Alexa Fluor 488 conjugated E. coli (MOI=4). The AMϕ were stained with phalloidin and Hoechst, and the phalloidin staining was outlined by using software Image J. The arrows indicate the filopodia. The graph depicts mean ± SEM of the numbers of filopodia on the AMϕ that phagocytosed E. coli, n=5. Statistical analysis using two-way ANOVA, Interaction p = 0.0013. Tukey’s multiple comparisons test shows that Young/15min vs. Age/15min, p = 0.0104; Young/30min vs. Age/30min, p = 0.0006; Young/60min vs. Age/60min, p = 0.0002. (D) Confocal images show AMϕ filopodia formation in young and aging AMϕ. Young and aging AMϕ were treated with 0 to 1000 ng/ml LPS for 1 h, and stained with phalloidin and Hoechst. The phalloidin staining was outlined by using software Image J. The arrows indicate the filopodia in AMϕ. The graph depicts mean ± SEM of the number of filopodia in AMϕ. n=5. Statistical analysis using two-way ANOVA, Interaction p =0.9686. Tukey’s multiple comparisons test shows that 10/Young vs. 10/Aging, p > 0.9999; 100/Young vs. 100/Aging, p > 0.9999; 1000/Young vs. 1000/Aging, p > 0.9999. (E and F) AMϕ were pretreated with Vehicle or Fascin inhibitor (25 µM) for 20 min followed by stimulation with Alexa Fluor 488-conjugated E. coli (MOI=4) for 60 min. (E) AMϕ were then stained with phalloidin and Hoechst, and phalloidin staining was outlined by using software Image J. The arrows indicate the filopodia. The graph depicts mean ± SEM of the numbers of filopodia on the AMϕ that phagocytosed E. coli, n=5. Statistical analysis using one-way ANOVA, over all ANOVA p = 0.0021. Tukey’s multiple comparisons test shows that E. coli/Vehicle vs. Control, p = 0.003; E. coli/Vehicle vs. E. coli/Fascin inhibitor, p = 0.0134. (F) Phagocytosis of Alexa Fluor 488- conjugated E. coli determined by flow cytometry. The graph depicts mean ± SEM of the MFI of Alexa Fluor 488-conjugated E. coli, n=3. Statistical analysis using one-way ANOVA, over all ANOVA p = 0.0083. Tukey’s multiple comparisons test shows that E. coli/Vehicle vs. Control, p = 0.0162; E. coli/Vehicle vs. E. coli/Fascin inhibitor, p = 0.0090. (G) AMϕ were pretreated with Vehicle (DMSO) or NSC23766 (30 µM), Wiskostatin (30 µM), or CK666 (30 µM) for 20 min, then the cells were stimulated with Alexa Fluor 488-conjugated E. coli (MOI=4) for 30 min. The AMϕ were stained with phalloidin and Hoechst. The phalloidin staining was outlined by using software Image J. The arrows indicated the filopodia in AMϕ. The graph depicts mean ± SEM of number of filopodia in AMϕ n=5. Statistical analysis using one-way ANOVA, over all ANOVA p = 0.0004. Tukey’s multiple comparisons test shows that E. coli/Vehicle vs. Control, p =0.0108; E. coli/NSC23766 vs. E. coli/Vehicle, p =0.0093; E. coli/Wiskostatin vs. E. coli/Vehicle, p =0.0172; E. coli/CK666 vs. E. coli/Vehicle, p =0.0097. The asterisks indicate that *p < 0.05 or **p < 0.01. NS, no significant difference.

Figure 5. Aging suppresses Rac1 expression and decreases Rac1-GTP level

(A) Young and aging AMϕ were treated with E. coli (MOI=4) for 30 min. The Rac1 mRNA level was measured by RT-QPCR. The graph depicts mean ± SEM of Rac1 mRNA fold induction, n=3. Statistical analysis using two-way ANOVA, Interaction p = 0.7593. Tukey’s multiple comparisons test shows that Young/Control vs. Age/Control, p = 0.0457; Young/E. coli vs. Age/E. coli, p = 0.0363. (B) Young and aging AMϕ were treated with or without E. coli (MOI=4) for 30 min, the Rac1-GTP was enriched via Rac1/cdc42 activation magnetic beads and measured by Western Blot. The left graph shows mean ± SEM of Rac1/GAPDH intensity fold induction, n=3. Statistical analysis using two-way ANOVA, Interaction p = 0.4492. Tukey’s multiple comparisons test shows that Young/Control vs. Age/Control, p = 0.0015; Young/E. coli vs. Age/E. coli, p = 0.0005. The right graph shows mean ± SEM of Rac1-GTP/GAPDH intensity fold induction, n=3. Statistical analysis using two-way ANOVA, Interaction p = 0.0225. Tukey’s multiple comparisons test shows that Young/Control vs. Young/E. coli, p = 0.0256, Young/E. coli vs. Age/E. coli, p = 0.0142. (C) Young and aging AMϕ were treated with or without E. coli (MOI=4) for 30 min, the total Arp2 and its phosphorylated form were measured by Western Blot. The graph depicts mean ± SEM of p-Arp2/t-Arp2 intensity fold induction, n=3. Statistical analysis using two-way ANOVA, Interaction p = 0.0003. Tukey’s multiple comparisons test shows that Young/Control vs. Young/E. coli, p = 0.0006, Young/E. coli vs. Age/E. coli, p = 0.0007. (D) Young and aging AMϕ were treated with or without E. coli (MOI=4) for 30 min, the N-WASP was detected by Western Blot. The graph shows mean ± SEM of N-WASP/GAPDH intensity fold induction, n=3. Statistical analysis using two-way ANOVA, Interaction p = 0.5214. (E) Confocal immunofluorescence images showing colocalization of p-Arp2 (cyan), phalloidin (red), and Hoechst (blue) in young and aging AMϕ, which were treated with Alexa Fluor 488-conjugated E. coli (MOI=4) for 0 to 30 min. The graph depicts mean ± SEM of Pearson’s Coefficient of p-Arp2 and phalloidin, n=3. Statistical analysis using two-way ANOVA, Interaction p = 0.0089. Tukey’s multiple comparisons test shows that 15min/Young vs. 15min/Aging, p = 0.0039; 30 min/Young vs. 30 min/Aging, p = 0.0058. (F) Young and aging mice were administered with Alexa Fluor 488-conjugated E. coli (4 × 10⁸ bacteria/kg B.W., i.t) for 30 min. The images show the colocalization of p-Arp2 (cyan), phalloidin (red), and Hoechst (blue) in young and aging AMϕ. The graphs show mean ± SEM of Pearson’s Coefficient of p-Arp2 and phalloidin, n=3. Statistical analysis using two-way ANOVA, Interaction p = 0.0016. Tukey’s multiple comparisons test shows that E. coli/Young vs. E. coli/Aging, p = 0.0007. The asterisks indicate that *p < 0.05 or **p < 0.01.