Replication of Salmonella enterica Serovar Typhimurium in Human Monocyte-Derived Macrophages

Abstract

Salmonella enterica serovar Typhimurium is a common cause of food-borne gastrointestinal illness, but additionally it causes potentially fatal bacteremia in some immunocompromised patients. In mice, systemic spread and replication of the bacteria depend upon infection of and replication within macrophages, but replication in human macrophages is not widely reported or well studied. In order to assess the ability of Salmonella Typhimurium to replicate in human macrophages, we infected primary monocyte-derived macrophages (MDM) that had been differentiated under conditions known to generate different phenotypes. We found that replication in MDM depends greatly upon the phenotype of the cells, as M1-skewed macrophages did not allow replication, while M2a macrophages and macrophages differentiated with macrophage colony-stimulating factor (M-CSF) alone (termed M0) did. We describe how additional conditions that alter the macrophage phenotype or the gene expression of the bacteria affect the outcome of infection. In M0 MDM, the temporal expression of representative genes from Salmonella pathogenicity islands 1 and 2 (SPI1 and SPI2) and the importance of the PhoP/Q two-component regulatory system are similar to what has been shown in mouse macrophages. However, in contrast to mouse macrophages, where replication is SPI2 dependent, we observed early SPI2-independent replication in addition to later SPI2-dependent replication in M0 macrophages. Only SPI2-dependent replication was associated with death of the host cell at later time points. Altogether, our results reveal a very nuanced interaction between Salmonella and human macrophages.

FIG 1

Cell phenotype affects the ability of Salmonella Typhimurium to survive and replicate in human MDM. (A) M1, M2, and M0 macrophages were infected with wild-type strain SL1344, fixed, and stained for immunofluorescence microscopy at the indicated times. Infected cells were scored as containing ≥10, 5 to 9, or <5 bacteria. The means ± standard deviations (SD) from 3 independent experiments using three different donors are shown. (B) Representative images of infected M0 macrophages. Green, LPS; red, LAMP1; blue, DAPI. Bar, 10 μm. (C) Gentamicin protection assays were performed using cells infected as described for panel A. At 2 and 18 h p.i., the cells were lysed and recoverable CFU were assessed by plating. Each point represents the fold change in CFU (18 versus 2 h p.i.) obtained in one experiment, and lines represent the means from 3 independent experiments. (D) Cells were infected, fixed, and stained for microscopy as described for panel A. The percentage of cells containing bacteria at each time point is shown. Bars represent means ± SD from 3 independent experiments. ***, P = 0.0003 by two-way ANOVA. (E) Cytokine levels in cell culture supernatants at 12 h p.i. Results are means ± SD from 3 independent experiments. *, P = 0.027 by one-way ANOVA.

FIG 2

Differentiation of monocyte-derived human macrophages in fetal bovine serum versus human serum affects the outcome of Salmonella Typhimurium infection. Cells were cultured in media containing FBS or HuS for 7 days as described in the text and then infected with Salmonella WT strains SL1344 and14028. (A) Phase-contrast images of uninfected cells. Bar, 100 μm. (B) Gentamicin protection assay. Shown is the fold change in CFU (at 18 and 2 h p.i.). Results are means ± SD from 4 independent experiments. *, P < 0.05 by two-way ANOVA. (C) The percentages of infected cells containing <5, 5 to 9, or ≥10 bacteria were assessed by microscopy at 2 and 18 h p.i. Results are means ± SD from 4 independent experiments.

FIG 3

Replication of Salmonella Typhimurium in monocyte-derived human macrophages is dependent on growth phase and supplemental histidine. (A) Gentamicin protection assay. Shown is the fold change in CFU (at 18 and 2 h p.i.) from M0 cells infected with Salmonella SL1344 grown to late-log phase (LL) or to stationary phase (Stat). Stationary-phase bacteria were opsonized with normal human serum. Results are means ± SD from 3 independent experiments. **, P = 0.008 by paired t test. (B) Gentamicin protection assay. Shown is the fold change in CFU (at 18 and 2 h p.i.) of strain SL1344 (mutation in hisG) with and without supplemental histidine (His) or strain 14028 without histidine. Results are means ± SD from 3 independent experiments. P < 0.05 (*) and P < 0.01 (**) by one-way ANOVA. (C) Representative images of macrophages at 18 h p.i., stained for LPS (gray) and LAMP1 (red). Cells were infected with SL1344 in the absence (−his) or presence (+his) of supplemental histidine. On the far right, strain 14028 is shown in the absence of histidine. Bar, 10 μm.

FIG 4

Gene expression in the course of infection. M0 macrophages were infected with SL1344 bacteria carrying a plasmid-borne reporter of the indicated genes. The GFP intensities in individual bacteria were measured as fluorescence units (FU) and the geometric means calculated. Shown are the geometric means ± SD from 3 independent experiments.

FIG 5

Salmonella Typhimurium organisms lacking the PhoP/Q two-component regulatory system are defective for replication in human MDM. M0 macrophages were infected with WT SL1344 or an isogenic ΔphoP strain. A gentamicin assay was performed. Shown is the fold change in recoverable CFU compared to the level at 2 h p.i. (A) and the percentage of cells containing ≥10 bacteria as assessed by microscopy (B). Results are means ± SD from 3 independent experiments. P < 0.05 (*), P < 0.01 (***), and P < 0.0001 (****) by two-way ANOVA of comparisons to 2-h p.i. values.

FIG 6

SPI2-dependent and -independent replication of Salmonella Typhimurium in M0 macrophages. (A) Gentamicin assay. Shown is the fold change in recoverable CFU compared to the values at 2 h p.i. Results are means ± SD from 5 independent experiments. (B) The percentage of infected cells containing the indicated number of bacteria was assessed by microscopy. Results are means ± SD from 3 independent experiments. (C and D) Intracellular replication was assessed by live cell imaging. Macrophages infected with mCherry-expressing WT or ΔSPI2 bacteria were imaged from 1 to 20 h p.i. every 15 min. (C) Cells that contained bacteria at 1 h were assessed for bacterial replication (top) and the appearance of cell death by morphological changes (bottom) over the time course. Macrophages from 3 donors were analyzed for each strain. A total of 62 cells infected with the WT and 59 infected with the ΔSPI2 strain were analyzed. (D) Stills from representative movies showing the replication of mCherry-expressing WT (top) or ΔSPI2 (bottom) bacteria in macrophages. Also see Movies S1 and S2 in the supplemental material. Bar, 10 μm. *, P < 0.05 by two-way ANOVA (A and B) and by paired t test (C).