RECON gene disruption enhances host resistance to enable genome-wide evaluation of intracellular pathogen fitness during infection

ABSTRACT Transposon sequencing (Tn-seq) is a powerful genome-wide technique to assess bacterial fitness under varying growth conditions. However, screening via Tn-seq in vivo is challenging. Dose limitations and host restrictions create bottlenecks that diminish the transposon mutant pool being screened. Here, we have developed a murine model with a disruption in Akr1c13 that renders the resulting RECON−/− mouse resistant to high-dose infection. We leveraged this model to perform a Tn-seq screen of the human pathogen Listeria monocytogenes in vivo. We identified 135 genes which were required for L. monocytogenes growth in mice including novel genes not previously identified for host survival. We identified organ-specific requirements for L. monocytogenes survival and investigated the role of the folate enzyme FolD in L. monocytogenes liver pathogenesis. A mutant lacking folD was impaired for growth in murine livers by 2.5-log10 compared to wild type and failed to spread cell-to-cell in fibroblasts. In contrast, a mutant in alsR, which encodes a transcription factor that represses an operon involved in D-allose catabolism, was attenuated in both livers and spleens of mice by 4-log10 and 3-log10, respectively, but showed modest phenotypes in in vitro models. We confirmed that dysregulation of the D-allose catabolism operon is responsible for the in vivo growth defect, as deletion of the operon in the ∆alsR background rescued virulence. By undertaking an unbiased, genome-wide screen in mice, we have identified novel fitness determinants for L. monocytogenes host infection, which highlights the utility of the RECON−/− mouse model for future screening efforts. IMPORTANCE Listeria monocytogenes is the gram-positive bacterium responsible for the food-borne disease listeriosis. Although infections with L. monocytogenes are limiting in healthy hosts, vulnerable populations, including pregnant and elderly people, can experience high rates of mortality. Thus, understanding the breadth of genetic requirements for L. monocytogenes in vivo survival will present new opportunities for treatment and prevention of listeriosis. We developed a murine model of infection using a RECON−/− mouse that is restrictive to systemic L. monocytogenes infection. We utilized this model to screen for L. monocytogenes genes required in vivo via transposon sequencing. We identified the liver-specific gene folD and a repressor, alsR, that only exhibits an in vivo growth defect. AlsR controls the expression of the D-allose operon which is a marker in diagnostic techniques to identify pathogenic Listeria. A better understanding of the role of the D-allose operon in human disease may further inform diagnostic and prevention measures.


SNP PCR genotyping of mice
Mouse genotyping was performed using a custom multiplex Akr1c13 SNP-based genotyping assay.An end-point PCR was first performed to amplify Akr1c13 exon 6.The product was diluted 1:1,000 in nuclease-free water, and 1.0 μL was used in a 20 μL qRT-PCR reaction, along with TaqMan Master Mix, primers that amplify Akr1c13 exon 6 (500 nM final concentration, Table S4) and FAM probes that detect the WT allele mixed with HEX probes that detect the mutant allele (250 nM final probe concentrations, Table S4).

Bacterial growth curves
Overnight cultures of Lm were back diluted to OD600=0.05 and grown at 37˚C with shaking until OD600 reached 0.4.Bacteria were resuspended in 1X PBS to OD600=1.0, diluted 1:100 in BHI or MM and 200 µL of this suspension was distributed to a 96-well plate.The plate was covered in Breathe Easy film (Diversified Biotech #BEM-1) and incubated at 37˚C, 237 CPM in a BioTek Synergy HTX plate reader with OD600 reads every 30 minutes.

BMDM growth curves with Lm
BMDM were seeded in 24-well tissue culture treated plates at a density of 2.5 x 10 5 cells/well with the addition of 100ng/µL murine IFN-g when indicated and incubated overnight at 37˚C 5% CO2.Cells were washed once in 1X PBS prior to infection.
Overnight cultures of Lm were resuspended in 1X PBS to OD600=1.0 and diluted 1:1,000 in DMEM +10% FBS +10% CSF and 500 µL of this infection medium was added to each well (MOI 3).Cells were incubated for 30 minutes to allow phagocytosis, then were washed twice in 1X PBS and replaced with fresh media containing 100µg/mL gentamicin for the remaining time points.At each time point the cells were washed twice in 1X PBS and lysed in 1 mL (0.5, 4, 6 hpi) or 500 µL (2 hpi) ice cold nanopure water.Lysates were diluted in PBS and plated for CFU enumeration on BHI + Streptomycin.

Supplemental Figure 2 .
Identification of novel genes required for in vivo growth only.(A) CFU harvested from livers and spleens of WT mice infected via IV with 1x10 5 CFU of WT (n=11) or ∆eslAB (n=11) Lm for 72 hours.Data are combined from two independent experiments.****, P<0.0001 by Mann Whitney test.(B) Naïve BMDM were infected with WT or ∆eslAB Lm at an MOI 1. Intracellular CFU were collected at the indicated time points.(C) Plaque area measured in fibroblasts infected with WT or ∆elsAB Lm for 48 hours and stained with neutral red.(D) CFU harvested from livers and spleens of WT mice infected via IV with 1x10 5 CFU of WT (n=11) or ∆lmo0897 (n=10) Lm for 72 hours.Data are combined from two independent experiments.(E) Naïve BMDM were infected with WT or ∆lmo0897 Lm at an MOI 1. Intracellular CFU were collected at the indicated time points.(F) Plaque area measured in fibroblasts infected with WT or ∆lmo0897 Lm for 48 hours and stained with neutral red.Supplemental Figure 3.The folate cycle gene folD is not required for Lm in vitro growth or growth in macrophages.(A-B) Bacterial growth as measured by OD600 over time in BHI (A) or minimal media (B).(C-D) BMDMs were left naïve (C) or stimulated with IFN-g (D) for 18 hours prior to infection with Lm.Intracellular CFU were collected at the indicated time points.Supplemental Figure 4. Dysregulation of the D-allose utilization operon impairs Lm growth in vivo but not in vitro.(A) CFU harvested from spleens of WT mice infected via IV with 1x10 5 CFU of WT, ∆alsR or ∆alsR::alsR Lm (n=4/group) for 72 hours.(B) CFU harvested from spleens of WT mice infected via IV with 1x10 5 CFU of WT or ∆alsR for 4 (n=6/group) or 25 (n=5/group) hours.(C) Mean CFU of WT or ∆alsR from panels Figures 4B and 4C plotted over time.(D) IFN-g-stimulated BMDMs were infected with Lm for the indicated time points and intracellular CFU were enumerated.(E) Plaque area measured in fibroblasts infected with WT, ∆alsR or ∆alsR::alsR Lm for 48 hours and stained with neutral red.(F) Bacterial growth as measured by OD600 over time in BHI.(G-H) CFU harvested from livers (G) and spleens (H) of WT mice infected via IV with 1x10 5 CFU of WT (n=5), ∆alsR (n=4), PrfA* (n=5) or ∆alsR PrfA* (n=5) Lm for 72 hours.(I).CFU harvested from spleens of WT mice infected via IV with 1x10 5 CFU of WT (n=6), ∆alsR (n=5) or ∆alsR∆alsOP (n=6) Lm for 72 hours.(J-N) CFU harvested from livers of WT mice infected via IV with 1x10 5 CFU with the indicated strains for 72 hours.J, n=5/group.K, n=6/group.L, n=4-5/group.M, n=3/group.N, n=4-5/group.Experiments in B, G-H, and I-N were performed once.Plaque area in E and CFU in A, G-N were analyzed by Kruskal-Wallis test with * p<0.05, ** p< 0.005 and **** p< 0.0001.CFU in B was analyzed by Mann-Whitney analysis with *p<0.05,**p<0.01.