ABSTRACT

   The interactions between a host cell and a pathogen can dictate disease
   outcomes and are important targets for host-directed therapies.
   Mycobacterium abscessus (Mab) is a highly antibiotic resistant, rapidly
   growing nontuberculous mycobacterium that infects patients with chronic
   lung diseases. Mab can infect host immune cells, such as macrophages,
   which contribute to its pathogenesis. However, our understanding of
   initial host-Mab interactions remains unclear. Here, we developed a
   functional genetic approach to define these host-Mab interactions by
   coupling a Mab fluorescent reporter with a genome-wide knockout library
   in murine macrophages. We used this approach to conduct a forward
   genetic screen to define host genes that contribute to the uptake of
   Mab by macrophages. We identified known regulators of phagocytosis,
   such as the integrin ITGB2, and uncovered a key requirement for
   glycosaminoglycan (sGAG) synthesis for macrophages to efficiently take
   up Mab. CRISPR-Cas9 targeting of three key sGAG biosynthesis
   regulators, Ugdh, B3gat3, and B4galt7 resulted in reduced uptake of
   both smooth and rough Mab variants by macrophages. Mechanistic studies
   suggest that sGAGs function upstream of pathogen engulfment and are
   required for the uptake of Mab, but not Escherichia coli or latex
   beads. Further investigation found that the loss of sGAGs reduced the
   surface expression, but not the mRNA expression, of key integrins,
   suggesting an important role for sGAGs in modulating surface receptor
   availability. Together, these studies globally define and characterize
   important regulators of macrophage-Mab interactions and are a first
   step to understanding host genes that contribute to Mab pathogenesis
   and disease.

   IMPORTANCE Pathogen interactions with immune cells like macrophages
   contribute to pathogenesis, yet the mechanisms underlying these
   interactions remain largely undefined. For emerging respiratory
   pathogens, like Mycobacterium abscessus, understanding these
   host-pathogen interactions is important to fully understand disease
   progression. Given that M. abscessus is broadly recalcitrant to
   antibiotic treatments, new therapeutic approaches are needed. Here, we
   leveraged a genome-wide knockout library in murine macrophages to
   globally define host genes required for M. abscessus uptake. We
   identified new macrophage uptake regulators during M. abscessus
   infection, including a subset of integrins and the glycosaminoglycan
   synthesis (sGAG) pathway. While ionic characteristics of sGAGs are
   known to drive pathogen-cell interactions, we discovered a previously
   unrecognized requirement for sGAGs to maintain robust surface
   expression of key uptake receptors. Thus, we developed a flexible
   forward-genetic pipeline to define important interactions during M.
   abscessus infection and more broadly identified a new mechanism by
   which sGAGs control pathogen uptake.

   KEYWORDS: CRISPR-Cas9, functional genetics, Mycobacterium,
   Mycobacterium abscessus, macrophages, phagocytosis

INTRODUCTION

   Mycobacterium abscessus (Mab) is a rapidly growing nontuberculous
   mycobacterium (NTM) that causes opportunistic infections in patients
   with chronic lung diseases, like cystic fibrosis and chronic
   obstructive pulmonary disease (COPD) ([28]1, [29]2). Mab is the second
   most common NTM respiratory pathogen recovered in the United States,
   accounting for a significant number of rapidly growing mycobacterial
   respiratory disease isolates ([30]3). Due to its recalcitrance to many
   antibiotics, current treatment success rates remain below 50%
   ([31]4[32]–[33]6). Treatment is further complicated by the ability of
   Mab to transition from a smooth to rough morphology that is driven by
   mutations that alter the export of the glycolipoproteins (GPLs) to the
   surface of Mab, including mutations in the transporter MmpL4 ([34]7,
   [35]8). This transition to the rough variant increases biofilm
   formation and decreases antibiotic sensitivity ([36]9). Thus, Mab is an
   emerging pathogen of clinical importance and there is a critical need
   to develop new treatment options.

   Developing more effective therapies requires a deeper understanding of
   Mab pathogenesis and host-pathogen interactions that drive infection.
   Recent work suggests that Mab interactions with macrophages are
   critical for disease progression, including the transition of smooth
   Mab variants into rough variants ([37]3, [38]10[39]–[40]12). Both
   smooth and rough Mab variants can survive in macrophages with rough
   variants initiating more rapid cell death cascades ([41]3, [42]10,
   [43]11). Transposon mutagenesis studies and other genetic approaches in
   Mab have identified many genes that are required for effective
   antibiotic killing and intracellular survival, including the ESX-IV
   system ([44]13[45]–[46]15). This virulence determinant contributes to
   the inhibition of lysosomal fusion enabling intracellular survival of
   Mab in macrophages ([47]15). In contrast to what is known about Mab,
   little is known about macrophage genes required for Mab uptake and
   survival. While studies have shown that Mab interacts with TLR2 and
   Dectin-1 to initiate uptake, a systematic characterization of Mab
   uptake by macrophages remains lacking ([48]16, [49]17).

   Uptake of pathogens requires effective interactions between the
   pathogen and macrophages. Macrophages use surface pattern recognition
   receptors (PRRs), including toll-like receptors, mannose receptors and
   scavenger receptors, for the initial recognition of many pathogens
   ([50]18[51]–[52]22). Other modifications to surface proteins, including
   sulfated glycosaminoglycan (sGAG) modifications such as heparin sulfate
   and chondroitin sulfate, affect the efficiency of pathogen attachment
   to the cell surface ([53]23, [54]24). On epithelial cells, sGAGs play
   an important role for the attachment of several pathogens, including
   Coronavirus and Chlamydia trachomatis ([55]25[56]–[57]27). It is
   predicted that these surface modifications alter the charge
   interactions between pathogens and the host cell surface to modulated
   initial attachment ([58]28, [59]29).

   Following adherence, phagocytosis or receptor mediated engulfment of
   pathogens enables internalization. While each phagocytic cargo has
   unique components, the general pathway for phagocytosis is conserved.
   Phagocytosis begins with the generation of a phagosome following a
   ligand binding to macrophage surface receptors ([60]30). This binding
   event activates downstream signaling cascades that remodel the actin
   cytoskeleton to drive the progressive engagement of additional
   receptors around the particle ([61]31, [62]32). After engulfment,
   controlled membrane fusion events result in the formation of the
   nascent phagosome. The phagosome then interacts with various endosome
   components that acidify the compartment and ultimately lead to
   compartment fusion with the lysosome ([63]33[64]–[65]38). Several
   elegant studies have dissected unique pathways that contribute to
   phagocytosis of distinct cargos, defining receptors and signaling
   cascades that are essential ([66]39[67]–[68]41). However, no studies
   have globally examined host genes that control Mab interactions with
   macrophages.

   Here, we defined the host genes that control the uptake of Mab into
   macrophages by coupling a brightly fluorescent Mab reporter with a
   genome-wide CRISPR-Cas9-mediated macrophage knockout library in murine
   immortalized bone marrow-derived macrophages (iBMDMs). We performed a
   forward genetic screen to enrich for macrophages that could or could
   not take up Mab identifying several previously defined phagocytosis
   components. Follow-up studies uncovered a strong requirement for sGAG
   production in macrophages for efficient Mab uptake. While loss of sGAGs
   did not affect general phagocytosis pathways or opsonized uptake of
   Mab, sGAGs were essential for initial interactions of both smooth and
   rough Mab variants with macrophages. Mechanistic studies uncovered a
   role for sGAGs in maintaining high surface expression of the key
   integrins ITGB2 and ITGAL, suggesting a new role for sGAGs in
   maintaining receptor availability on the cell surface. These results
   uncover important host pathways that modulate the interactions between
   Mab and macrophages during infection.

RESULTS

Fluorescent Mab reporter enables dissection of macrophage-Mab interactions.

   While Mab infects macrophages, our understanding of these interactions
   remains limited. To help address this gap in knowledge we developed a
   brightly fluorescent Mab reporter strain. A constitutive mEmerald GFP
   vector was transformed into the smooth variant of Mab strain
   ATCC-19977. Comparisons to nonfluorescent Mab showed that bacterial
   growth and antibiotic mediated killing with rifampicin in liquid media
   were unaffected by fluorescent protein expression ([69]Fig. 1A). We
   next examined the utility of the fluorescent Mab strain to dissect
   host-pathogen interactions with macrophages. We first examined the
   robustness of the fluorescent reporter by infecting immortalized bone
   marrow derived macrophages (iBMDMs) from C57BL/6J mice. One previous
   study using a GFP fluorescent Mab was unable to detect increased
   fluorescence in macrophages until 24 h of infection, when attachment,
   uptake and bacterial growth all contribute to differential fluorescence
   ([70]42). We hypothesized that earlier time points could be examined
   using the brighter mEmerald GFP, allowing us to distinguish uptake from
   intracellular growth. To test this hypothesis, we quantified the number
   of mEmerald positive macrophages by flow cytometry 6 h after infecting
   iBMDMs with fluorescent Mab at increasing multiplicities of infection
   (MOIs). We noted that increasing the MOI resulted in an increase in the
   percentage of mEmerald+ cells and the mean fluorescence intensity of
   these infected cells ([71]Fig. 1B to [72]1D). We next quantified the
   mEmerald+ cells over the first 6 h of infection of iBMDMs at an MOI of
   5. We observed the number of infected cells 2-h postinfection showed a
   minimal increase compared to uninfected controls, while there was a
   significant increase in infected cells at four- and 6-h postinfection
   ([73]Fig. 1E and [74]F). Thus, the mEmerald GFP reporter does not
   affect Mab growth while enabling the dissection of early interactions
   between Mab and macrophages.

FIG 1.

   [75]FIG 1
   [76]Open in a new tab

   Early interactions with fluorescent Mycobacterium abscessus and
   macrophages can be detected by flow cytometry. (A) The optical density
   of control or mEmerald transformed M. abscessus ATCC-19977 was
   monitored over 48 h in 7H9 broth in the presence or absence of
   Rifampicin (128 μg/mL). (B–D) iBMDMs from C57BL6J mice were infected
   with increasing MOI (5–20) of mEmerald M. abscessus for 6 h. Flow
   cytometry was used to measure (B) the percentage of cells that were
   infected (mEmerald+) and (C) the mean fluorescence intensity of
   infected cells. (D) Representative flow cytometry plots gated on live
   and single cells are shown for each MOI. Data are from one of three
   independent experiments. (E–F) iBMDMs from C57BL6/J mice were infected
   with mEmerald M. abscessus at an MOI of 5 for two, four or 6 h. (E)
   Flow cytometry was used to measure the percentage of cells that were
   infected (mEmerald+). (F) Shown are representative flow cytometry plots
   gated on live and single cells for each time point. All results shown
   are representative results from one of three independent experiments.
   ****, P < 0.0001; ***, P < 0.001 by one-way ANOVA with Dunnett’s
   multiple-comparison test.

CRISPR-Cas9 loss-of-function (LOF) screen identifies host genes required for
Mab uptake by macrophages.

   Host pathways required for macrophage uptake of Mab remain almost
   entirely unknown. To test if the flow cytometry-based uptake assay
   could identify host pathways required for Mab uptake, we used a known
   phagocytosis inhibitor, Cytochalasin D, to inhibit actin polymerization
   ([77]43). iBMDMs were pretreated with Cytochalasin D then infected with
   mEmerald Mab. The percentage of cells that were infected were then
   quantified 4 h later by flow cytometry. We observed a nearly 50%
   decrease in mEmerald+ macrophages treated with Cytochalasin D compared
   to vehicle controls ([78]Fig. 2A and [79]B). These data highlight the
   sensitivity of the uptake assay to dissect early interactions between
   Mab and macrophages.

FIG 2.

   [80]FIG 2
   [81]Open in a new tab

   Genome-wide loss-of-function screen in iBMDMs identifies host genes
   required for uptake of M. abscessus. (A) iBMDMs were treated with DMSO
   or Cytochalasin D (10 μg/mL or 20 μg/mL) for 2 h then infected with
   mEmerald M. abscessus at an MOI of 5 for 4 h. The percent uptake was
   quantified by flow cytometry. Results are normalized to the mean
   percent uptake of the NTC+DMSO condition (B) Shown are representative
   flow cytometry plots gated on live single cells for each treatment. (C)
   A schematic of the genome-wide screen to identify host genes that are
   required for M. abscessus uptake by macrophages. A genome-wide
   CRISPR-Cas9 library generated in Cas9+ iBMDMs with sgRNAs from the Brie
   library (4 sgRNAs per gene) was infected with mEmerald M. abscessus for
   4 h and mEmerald^+ and mEmerald- populations were isolated by FACS. The
   representation of sgRNAs in each population were determined by
   sequencing. Figure made in Biorender. (D) Shown is the score determined
   by the alpha-robust rank algorithm (α-RRA) in MAGeCK for each gene in
   the CRISPR-Cas9 library that passed filtering metrics from three
   independent screen replicates. Highlighted genes represent known host
   factors that contribute to take up of pathogens. (E) Cas9+ iBMDMs
   targeted with the indicated sgRNAs for candidates (2 per candidate
   gene) were infected with mEmerald M. abscessus for 4 h at an MOI of 5.
   The uptake of M. abscessus was quantified by flow cytometry. These data
   are normalized to the mean percentage of uptake of M. abscessus by
   nontargeting control cells. The results are representative of at least
   three independent experiments. ***, P < 0.001; **, P < 0.01 by one-way
   ANOVA with Dunnett’s multiple-comparison test. For the data in panel 2E
   each candidate sgRNA is P < 0.001 compared to the NTC.

   Next, we leveraged the uptake assay to globally identify host
   regulators required for Mab uptake by macrophages ([82]Fig. 2C). We
   previously generated a robust and reproducible pooled genome-wide
   loss-of-function library in Cas9+ iBMDMs ([83]44). Knockouts in this
   pool were generated with sgRNAs from the Brie library which targets
   four independent single guide RNA molecules (sgRNAs) per coding gene
   and over 1,000 nontargeting controls (NTC) ([84]45). To identify
   regulators of Mab uptake, we infected the loss-of-function library of
   macrophages with mEmerald Mab for 4 h. Given our goal to identify host
   genes that specifically modulate Mab uptake, we chose to complete this
   screen at 4 h postinfection since this is the earliest time point with
   a significant increase of mEmerald that is detectable by flow
   cytometry. Fluorescence activated cells sorting (FACS) was then used to
   isolate over 200× sgRNA coverage of the library from mEmerald+ cells
   and mEmerald- cells. Following genomic DNA extraction, sgRNA abundances
   for each sorted bin were quantified by deep sequencing.

   To test for statistical enrichment of sgRNAs and genes, we used the
   modified robust rank algorithm (α-RRA) employed by Model-based Analysis
   of Genome-wide CRISPR/Cas9 Knockout (MAGeCK). This algorithm ranks
   sgRNAs by effect before filtering low ranking sgRNAs to improve
   significance testing. To identify macrophage genes required for Mab
   uptake during early infection, we compared the enrichment of sgRNAs in
   the mEmerald - (Mab uninfected) directly to the mEmerald+ (Mab
   infected) population ([85]Table S1). The α-RRA analysis identified 100
   genes with a P value < 0.01 and a fold change of 2 with at least 2
   independent sgRNAs. Among the top 100 candidates, we identified known
   regulators of macrophage phagocytosis, including Manea, M6PR, Itgb2
   (CD18), Itgam (CD11b), CD46, and Rac1 with each gene showing enrichment
   in the mEmerald- population ([86]Fig. 2D) ([87]46). Guide-level
   analysis showed agreement with all four sgRNAs targeting these genes
   suggesting they are bona fide hits and that our screen was robust
   ([88]Table S1).
   TABLE S1

   Analyzed CRISPR screen data. The gene summary ranking and log fold
   change from the MAGeCK analysis for every gene and individual sgRNA in
   the Brie library from the genome-wide screen for Mab uptake. Download
   [89]Table S1, XLSX file, 3.7 MB^ (3.8MB, xlsx) .

   Copyright © 2023 Gilliland et al.

   This content is distributed under the terms of the [90]Creative Commons
   Attribution 4.0 International license.

   To confirm the screen results, we validated the role of Manea, M6PR,
   Itgb2, and Itgam in controlling Mab uptake by macrophages. We generated
   two independent iBMDM lines targeting two distinct sgRNAs per gene, in
   addition to a nontargeting control. These cells were then infected with
   mEmerald Mab and uptake was quantified 4 h later. In agreement with our
   screen results, we found that targeting each candidate gene resulted in
   a significant decrease in mEmerald+ iBMDMs compared to NTC cells
   ([91]Fig. 2E). Taken together these results show that the genome-wide
   screen identified host genes that contribute to early interactions
   between Mab and macrophages.

Sulfated glycosaminoglycans (sGAGs) are required for macrophage uptake of
Mab.

   To discover new pathways that are required for Mab uptake by
   macrophages, we next used DAVID analysis and gene set enrichment
   analysis (GSEA) to identify functional enrichments from our uptake
   screen data set ([92]Table S2). These analyses identified the sulfated
   glycosaminoglycan (sGAG) synthesis pathway as required for efficient
   uptake of Mab ([93]Fig. 3A and [94]Table S2). This leading-edge
   analysis showed a clear enrichment of 12 sGAG-biosynthesis genes in the
   uninfected population (left side enrichment) compared to Mab infected
   cells. While a variety of pathogens use sGAGs to facilitate attachment
   and invasion of epithelial cells, little is known about their role in
   macrophages. Within the top candidates in the screen were multiple
   genes within the sGAG biosynthesis pathway ([95]Fig. 3B). Three of
   these genes, UGDH, B3GAT3, and B4GALT7 encode UDP-glucose
   6-dehydrogenase, galactosyltransferase I and glucuronosyltransferase I,
   respectively, and were among the top 30 candidates while showing strong
   agreement among the four independent sgRNAs ([96]Table S1). These data
   suggest that sGAGs contribute to Mab uptake by macrophages.

FIG 3.

   [97]FIG 3
   [98]Open in a new tab

   Sulfated glycosaminoglycans are required for effective uptake of M.
   abscessus by macrophages. (A) Gene set enrichment analysis was used to
   identify enriched pathways from the ranked forward genetic screen.
   Shown is a leading-edge analysis plot representing a top enriched KEGG
   pathway related to glycosaminoglycan biosynthesis. (B) The α-RRA
   ranking of sGAG biosynthesis related genes identified in the
   genome-wide screen highlighted as blue dots. (C) Total sGAGs were
   quantified using the Blyscan Sulfated Glycosaminoglycan Assay Protocol
   (BioVision) in sgNTC, sgUGDH, sgB3GAT3 and sgB4GALT7 iBMDMs and
   normalized to total protein content. Shown is a representative result
   from four independent experiments. (D) sgNTC, sgUGDH, sgB3GAT3 and
   sgB4GALT7 iBMDMs were infected with mEmerald M. abscessus at increasing
   MOI for 4 h then the percent infected cells were quantified by flow
   cytometry. Shown is the percentage of infected cells normalized to the
   mean of the uptake of NTC at each MOI. These results are representative
   of four independent experiments. (E) Shown is a representative flow
   cytometry plot gated on live and single cells for the indicated
   genotypes at an MOI of 5. ***, P < 0.001; **, P < 0.01 by one-way ANOVA
   with Dunnett’s multiple-comparison test.
   TABLE S2

   Pathway enrichment analysis for genes required for Mab uptake. DAVID
   analysis output of KEGG pathways that are statistically enriched among
   the genes required for Mab uptake in macrophages. Download
   [99]Table S2, XLSX file, 0.01 MB^ (9.4KB, xlsx) .

   Copyright © 2023 Gilliland et al.

   This content is distributed under the terms of the [100]Creative
   Commons Attribution 4.0 International license.

   To validate the importance of macrophage sGAGs for Mab uptake, we
   targeted Ugdh, B3gat3, and B4galt7 directly in iBMDMs. Using an
   approach targeting two independent sgRNAs simultaneously we generated a
   panel of iBMDMs with editing efficiency ranging from 60 to 95% in each
   gene, and one cell line per gene was selected for follow up studies
   ([101]45). We first examined functional differences following gene
   editing by quantifying the total sGAGs in sGAG-targeted and NTC iBMDMs.
   We found that iBMDMs targeted for Ugdh, B3gat3, and B4galt7 each
   resulted in a significant reduction in sGAGs compared to NTC
   macrophages ([102]Fig. 3C). We next used these sGAG-targeted cells to
   examine differences in Mab uptake. NTC and sGAG-targeted iBMDMs were
   infected with mEmerald Mab for 4 h with increasing MOIs and bacterial
   uptake was quantified. At each MOI we observed a significant decrease
   in Mab uptake in sGAG-targeted iBMDMs compared to the NTC cells
   ([103]Fig. 3D and [104]3E). These results confirm that sGAGs contribute
   to the effective uptake of Mab by macrophages, validating our
   genome-wide screen and bioinformatic analysis.

sGAGs are required for macrophage uptake of rough Mab variants.

   While our data suggest sGAGs are needed for uptake of smooth Mab
   variants by macrophages, it remained unclear if this host pathway was
   also required for rough variant uptake. To directly test this question,
   a rough Mab variant derived from ATCC 19977 was transformed with the
   mEmerald reporter. Similar to the smooth variant, we noted no defects
   in growth or antibiotic killing in broth culture ([105]Fig. 4A) and
   robust uptake by iBMDMs in an MOI dependent manner ([106]Fig. 4B to
   [107]D). To test if sGAGs are also required for uptake of rough Mab we
   next infected NTC and sGAG-targeted iBMDMs with the rough mEmerald Mab
   reporter for 4 h at increasing MOI and quantified the percentage of
   infected cells. We observed over a 50% decrease of rough Mab uptake in
   sGAG-targeted macrophages compared to control macrophages at all MOIs
   ([108]Fig. 4E and [109]F). Interestingly, while significant, the
   magnitude of the decrease in uptake in sGAG-targeted iBMDMs was lower
   for rough Mab than smooth Mab. Taken together these data show sGAGs are
   required for macrophage uptake of both smooth and rough Mab variants.

FIG 4.

   [110]FIG 4
   [111]Open in a new tab

   Rough variants of M. abscessus require sGAGs for efficient uptake by
   macrophages. (A) The optical density of control or mEmerald transformed
   rough variant of M. abscessus ATCC-19977 was monitored over 48 h in 7H9
   broth in the presence or absence of Rifampicin (128 μg/mL). (B) iBMDMs
   from C57BL6/J mice were infected with increasing MOI
   ([112]5[113]–[114]20) of mEmerald rough variant of M. abscessus for 4 h
   and flow cytometry was used to measure the percentage of cells that
   were infected (mEmerald+). Shown are representative flow cytometry
   plots gated on live and single cell for an MOI of five. (C) The
   percentage of macrophages infected with rough Mab and (D) the mean
   fluorescence intensity of infected cells at the indicated MOIs. (E)
   sgNTC, sgUGDH, sgB3GAT3 and sgB4GALT7 iBMDMs were infected with
   mEmerald rough variant of M. abscessus at increasing MOIs for 4 h then
   the percent uptake of cells was quantified by flow cytometry. Shown is
   a representative flow cytometry plot gated on live and single cells for
   the indicated genotypes infected at an MOI of 5. (F) Shown is the
   percentage of infected cells normalized to the mean of the percent
   uptake for the NTC at each MOI. Results in (A–F) are all representative
   of at least three independent experiments. ***, P < 0.001; **, P < 0.01
   by one-way ANOVA with Dunnett’s multiple-comparison test.

sGAGs function upstream of Mab internalization.

   We next examined the mechanisms underlying sGAG-mediated control of Mab
   uptake. To understand if sGAGs control specific or general uptake
   pathways, we tested whether sGAGs were necessary for uptake of other
   phagocytic cargo by macrophages. First, we quantified the uptake of
   latex beads by incubating yellow-green-labeled beads at increasing
   concentrations with control or sGAG-targeted iBMDMs. In contrast to our
   results with Mab, we found no differences in the uptake of latex beads
   between control or sGAG-targeted iBMDMs ([115]Fig. 5A and [116]B). We
   next examined whether sGAGs contribute to the uptake of Escherichia
   coli, a Gram-negative bacterium. mEmerald expressing E. coli was
   incubated with NTC or sGAG-targeted iBMDMs for 4 h then uptake was
   quantified. Similar to latex beads, we observed no significant
   differences in E. coli uptake between control or sGAG-targeted
   macrophages ([117]Fig. 5C and [118]D). Taken together, these results
   suggest that sGAG-mediated uptake in macrophages occurs independently
   of general phagocytosis mechanisms and has pathogen specificity.

FIG 5.

   [119]FIG 5
   [120]Open in a new tab

   sGAGs do not alter general macrophage phagocytosis mechanisms and
   require viable bacteria to contribute to take up. (A)
   Yellow-green-labeled latex beads were added at increasing
   concentrations (Bead:Cell ratio 3–50) to sgNTC, sgUGDH, sgB3GAT3, and
   sgB4GALT7 iBMDMs for 4 h then the percentage of cells with beads was
   quantified by flow cytometry. Shown is the percentage of cells with
   beads normalized to the mean of the percent uptake of NTC at each
   concentration. (B) Shown are representative flow cytometry plots gated
   on live and single cells for the indicated genotypes treated with a
   bead:cell ratio of 5. (C) The indicated genotypes of iBMDMs were
   infected with mEmerald E. coli for 4 h at an MOI of five. The percent
   infected cells were quantified by flow cytometry and were normalized to
   the mean of the percent uptake of NTC. (D) Shown are representative
   flow cytometry plots gated on live and single cells for the indicated
   genotyped infected with E. coli at an MOI of 5 (E) Smooth or (F) rough
   variants of mEmerald M. abscessus were incubated with heat-killed or
   active fetal bovine serum for 30 min then used to infect iBMDMs of the
   indicated genotypes for 4 h at an MOI of five. The percent infected
   cells were quantified by flow cytometry and were normalized to the mean
   of sgNTC uptake in heat-killed serum for either smooth or rough
   variants. (G) iBMDMs of the indicated genotypes were treated with DMSO
   or Cytochalasin D (10 μg/mL or 20 μg/mL) for 2 h then infected with
   mEmerald M. abscessus smooth or (H) rough variants at an MOI of 5 for 4
   h. The percent uptake was quantified by flow cytometry and all samples
   were normalized to the mean of sgNTC uptake in DMSO for either smooth
   or rough variants. All results are representative of at least 3
   independent experiments with similar results. ***, P < 0.001; **, P <
   0.01 by two-way ANOVA with Dunnett’s multiple-comparison test.

   Given that the uptake of pathogens can be influenced by opsonization,
   we next examined whether the sGAG pathway overlaps with
   complement-mediated uptake mechanisms. Both smooth and rough mEmerald
   Mab reporters were incubated in active or heat killed serum for 30 min.
   Following serum incubation, control or sGAG-targeted iBMDMs were
   infected and the percent uptake of Mab was quantified 4 h later.
   Opsonization of the smooth Mab variant with active serum resulted in no
   significant change in uptake compared to heat-inactivated serum
   ([121]Fig. 5E). For the rough Mab variant, while uptake was
   significantly increased following incubation with active serum in all
   cells, the uptake differences between NTC and sGAG-targeted iBMDMs
   remained significant ([122]Fig. 5F). These data show that while smooth
   and rough Mab variants are differentially susceptible to
   complement-mediated phagocytosis, this uptake pathway is independent of
   sGAGs.

   To distinguish if sGAGs are required for internalization or attachment
   of Mab we next treated control or sGAG-targeted macrophages with
   Cytochalasin D. We hypothesized that if sGAGs control initial
   attachment that blocking actin polymerization with Cytochalasin D would
   further inhibit Mab uptake in sGAG-targeted macrophages. To test this
   hypothesis, we pretreated NTC and sGAG-targeted macrophages with DMSO
   or Cytochalasin D for 2 h. Cells were then infected with smooth or
   rough variants of mEmerald Mab and uptake was quantified 4 h later. We
   found that Cytochalasin D treatment significantly reduced uptake of
   both smooth and rough Mab variants compared to vehicle controls in all
   cell lines tested ([123]Fig. 5G and [124]H). These data suggest that
   sGAGs are needed upstream of bacterial internalization as they function
   additively with actin polymerization inhibitors.

Loss of sGAGs reduces the surface integrin expression on macrophages.

   While sGAGs directly modulate ionic interactions at the cellular
   surface, it remained possible that sGAG-modifications also regulate the
   surface levels of the receptors that are required for pathogen uptake
   ([125]29). Given the importance of integrins for Mab uptake from our
   screen, we examined the expression of a subset of integrins, including
   ITGB2, ITGAM, and ITGAL. When we quantified the mRNA levels of these
   integrins we found no changes in the mRNA expression of integrins
   between NTC and sGAG-targeted macrophages ([126]Fig. 6A). In contrast,
   while we observed no difference in the surface expression of ITGAM, we
   observed a significant decrease in ITGB2 and ITGAL expression on the
   surface of sGAG-targeted iBMDMs ([127]Fig. 6B and [128]C). We next
   directly examined surface levels of the ITGB2/ITGAL heterodimer (LFA-1)
   and found significantly reduced expression on the surface of
   sGAG-targeted iBMDMs ([129]Fig. 6D and [130]E). Taken together these
   data suggest that sGAGs modulate the surface expression of key
   integrins that contribute to Mab uptake by macrophages.

FIG 6.

   [131]FIG 6
   [132]Open in a new tab

   Reduced sGAGs results in lower surface expression of integrins on
   macrophages. (A) RNA was isolated from sgNTC, sgUGDH, sgB3GAT3 and
   sgB4GALT7 iBMDMs and the expression of Itgb2, Itgal, and Itgam relative
   to Gapdh was quantified by qRT-PCR. (B) The surface expression of
   ITGB2, ITGAL and ITGAM was quantified in NTC and sGAG-targeted iBMDMs
   by flow cytometry. Shown is the mean fluorescence intensity of each
   integrin and (C) a representative histogram for each sGAG-targeted gene
   for each integrin overlaid with NTC is shown. (D) The surface
   expression of ITGB2-ITGAL heterodimers (LFA-1) were quantified on NTC
   and sGAG-targeted iBMDMs using flow cytometry. Shown is the
   quantification of the mean fluorescence intensity and (F) a
   representative histogram for each iBMDM genotype overlaid with NTC. All
   results are representative of at least 2–3 independent experiments with
   similar results. ***, P < 0.001; **, P < 0.01 by one-way ANOVA with
   Dunnett’s multiple-comparison test.

DISCUSSION

   Infections with M. abscessus are on the rise globally and their
   recalcitrance to antibiotic therapy is driving treatment failure
   ([133]47[134]–[135]49). Designing effective treatments against Mab,
   including host-directed therapies, will require a fundamental
   understanding of critical host-Mab interactions that occur during
   infection. Here, we developed a Mab fluorescent reporter-based approach
   to investigate interactions of Mab with macrophages and conducted a
   forward genetic screen to globally identify host genes that are
   required for Mab uptake. Our screen identified many candidates,
   including previously described players in pathogen uptake such as
   Itgb2, Itgam, Manea, and M6pr ([136]50, [137]51), which together with
   our validation highlight the robustness of this data set and the
   strengths of the functional genetic approach.

   As part of our bioinformatic analysis to uncover new pathways required
   for efficient Mab uptake, we identified a strong signature for the sGAG
   biosynthesis pathway. sGAGs are modified glycosaminoglycans, which
   consist of repeating polysaccharides attached to the cell surface and
   surface proteins ([138]29). We targeted three independent genes in the
   sGAG synthesis pathway that all resulted in decreased uptake of both
   smooth and rough Mab variants by macrophages. While sGAGs have
   previously been shown to modulate host-pathogen interactions in
   epithelial cells, very little is known about their role in macrophages
   ([139]25[140]–[141]27). We found that sGAGs did not affect latex bead
   or E. coli uptake, showing that general macrophage uptake mechanisms
   are intact in sGAG-targeted cells. These findings are in line with a
   report that used iterative CRISPR screens to identify genes needed for
   the phagocytosis of distinct cargo, including beads of distinct sizes
   and complement/IgG coated cargo ([142]40). This wide-ranging study did
   not identify a significant enrichment of sGAG biosynthetic genes in any
   of their screens suggesting further specificity of the sGAG-mediated
   uptake pathway. In addition, Cytochalasin D treatment of sGAG-targeted
   macrophages further reduced Mab uptake, suggesting sGAGs contribute
   upstream of the actin polymerization required for Mab internalization.
   Together these data point toward a role for sGAGs in the adherence of
   Mab to macrophages. While many models suggest sGAGs modulate ionic
   interactions between pathogens and the cellular surface, our results
   suggest a parallel mechanism by which sGAGs contribute to pathogen
   uptake ([143]29). In all sGAG-targeted iBMDMs we discovered a
   significant decrease in the surface expression of the integrins ITGB2
   and ITGAL but no change in the mRNA expression. We propose that one
   previously overlooked function of sGAGs in pathogen uptake is
   maintaining the surface expression of critical receptors, including key
   integrins. How sGAGs regulate integrin surface expression remains
   unclear. Two possible models are that sGAGs stabilize interactions
   between surface integrins to maintain high surface expression or that
   sGAGs prevent rapid turnover of surface integrins. Testing these
   possible models will require macrophages lacking one of or combinations
   of sGAG-biosynthesis genes and integrins. These knockout cells will
   further help to delineate the epistatic relationship between sGAGs and
   integrins while also allowing us to investigate the specific roles of
   these pathways in Mab adherence. Beyond these focused studies on Mab, a
   broader understanding of how dynamically sGAGs control the surface
   proteome of cells and if this function of sGAGs contributes to other
   host-pathogen interactions will be of great interest.

   Multiple lines of evidence from our study point toward a central role
   for ITGB2 in the uptake of Mab. Not only was Itgb2 a top candidate from
   the genome-wide screen, but we found the top enriched pathway,
   sGAG-biosynthesis, directly modulates ITGB2 surface expression. ITGB2
   is known to complex with different integrin alpha receptors to serve as
   a key trafficking molecule and complement receptor ([144]52). ITGB2, in
   complex with ITGAM, was previously shown to modulate LC3-associated
   phagocytosis (LAP) during Listeria monocytogenes infections and control
   the uptake of Pseudomonas aeruginosa, suggesting multiple uptake
   pathways intersect with ITGB2 ([145]50, [146]53). The identification of
   both Itgam and Itgb2 among the top candidates in the screen suggests a
   key role for this heterodimer in Mab uptake. ITGB2 also complexes with
   other integrins like ITGAL, yet the contribution of each of these
   distinct receptors in Mab uptake remains unknown and will be examined
   in the future. While human and murine ITGB2 proteins are highly
   similar, it will be important to directly test the role of this
   integrin in Mab uptake in human macrophages ([147]54). In addition,
   investigating how Mab uptake by distinct receptors alters intracellular
   trafficking and/or bacterial control will better define key
   host-pathogen interactions during Mab infections.

   How Mab mediates interactions with distinct macrophage uptake pathways
   remains unknown. Surface glycopeptidolipids (GPLs) are thought to
   contribute to macrophage infections ([148]12, [149]55). However, given
   that both smooth and rough variants of Mab required sGAG-dependent
   pathways for effective uptake, there must be GPL-independent mechanisms
   that contribute during infection. We did observe that rough variants
   were less affected by the loss of sGAGs than smooth variants, which may
   point to multiple functions of sGAGs in mediating Mab uptake. We also
   noted that rough variants are more susceptible to complement-mediated
   uptake than smooth variants. This suggests a model where GPL evade
   complement-mediated uptake which may modulate Mab survival. Several
   other Mab factors have been associated with macrophage infection,
   including the ATPase EccB4 and glycosyl-diacylated-nondecyl-diols
   (GDND), but the direct role of these genes in macrophage uptake remain
   unknown ([150]15, [151]56[152]–[153]58). More broadly dissecting Mab
   determinants of macrophage infection will require unbiased bacterial
   genetic approaches, such as transposon-based approaches that were
   recently developed in Mab ([154]13, [155]15). This approach will link
   macrophage uptake pathways directly with specific Mab-encoded factors.

   In addition to understanding how Mab adheres to macrophages, the role
   of sGAGs during lung infection is important to consider. While Mab is
   mostly associated with monocytes and macrophages in vivo, current mouse
   models to test Mab pathogenesis remain challenging ([156]57, [157]59).
   Sustained Mab infection in mice requires multiple immune deficiencies
   that disrupt normal host-pathogen interactions ([158]59). In addition,
   since sGAGs are essential to development, understanding how sGAGs
   contribute to in vivo disease will require the generation of
   conditional knockout animals in immune knockout backgrounds ([159]60).
   While improved Mab in vivo models are being developed, the approach
   described here highlights a framework for leveraging unbiased genetic
   approaches to glean important mechanistic understanding of host-Mab
   interactions.

   Altogether, we optimized a host genetic platform to define important
   macrophage pathways during Mab infection. By coupling a brightly
   fluorescent Mab reporter with a genome-wide knockout macrophage
   library, we are positioned to rapidly screen for diverse phenotypes
   related to macrophage-Mab interactions. Here, we leveraged this
   approach to dissect Mab uptake during early infection. Yet, this same
   approach can be used to understand host pathways that control Mab
   growth or contribute to macrophage cell death. Thus, our host-focused
   functional genetic approach will broadly identify potential
   host-directed therapeutic targets, with the goal of improving treatment
   success during Mab infection.

MATERIALS AND METHODS

Mab culture conditions, generation of fluorescent reporter, and growth
assays.

   Isogenic pairs of smooth and rough Mab (ATTCC 19977) were used
   throughout this study. The Mab or E. coli mEmerald strains were built
   by transforming pmV261 hsp60::mEmerald into either the smooth or rough
   Mab variant by electroporation or DH5α E. coli by heat shock followed
   by selection on zeocin. All Mab cultures were grown aerobically in
   Middlebrook 7H9 medium supplemented with 10% Middlebrook OADC (oleic
   acid, dextrose, catalase, and bovine albumin) at 37°C. mEmerald
   GFP-expressing strains were propagated in medium containing 5 μg/mL
   zeocin (Invivogen).

   To quantify bacterial growth, nonfluorescent and mEmerald GFP-Mab
   single cell suspensions were inoculated in 7H9 at starting
   concentration of 5x10^5 bacteria/mL. To examine antibiotic killing,
   rifampicin was included at a final concentration of 128 μg/mL. Optical
   density was quantified by measuring the absorbance at 600 nm on a Tecan
   Spark 20 M plate reader at the indicated time points.

Cell culture.

   J2-virus immortalized murine bone marrow derived macrophages (iBMDMs)
   and Cas9+ iBMDMs were maintained in Dulbecco’s Modified Eagle Medium
   (DMEM; HyClone) supplemented with 10% heat-inactivated fetal bovine
   serum (FBS; Seradigm) as previously described ([160]45). All macrophage
   cell lines were incubated in 5% CO[2] at 37°C.

Macrophage infections and uptake assays.

   Single cell Mab or E. coli suspensions were prepared by resuspending
   logarithmic phase bacteria in DMEM with 10% heat-inactivated FBS
   (Seradigm) followed by a soft spin (800g) to pellet large bacterial
   clumps. The supernatant was then used to infect macrophages, seeded at
   5x10^5/well in a 12-well plate, at the indicated MOIs for the indicated
   time points (2 to 6 h). Following infection, macrophages were washed
   with PBS, lifted from plates by scrapping, then fixed in 4%
   paraformaldehyde. Macrophage uptake was quantified using the BD LSRII
   flow cytometer (BD Biosciences) at the Michigan State University Flow
   Cytometry Core. Live and single macrophages were identified using
   forward and side scatter and the number of infected cells was
   determined by the fluorescence in the GFP channel. All experiments
   include an uninfected control to set gates for uptake quantification
   during analysis that was performed using FlowJo v10. For
   complement-mediated uptake experiments mEmerald Mab was first incubated
   in heat-inactivated or active FBS for 30 min prior to macrophage
   infections. For actin polymerization inhibitor experiments, cells were
   treated with DMSO or Cytochalasin D (Cayman Chemicals) for 2 h prior to
   infection.

   For the bead uptake assay, 1 μm carboxylate-modified polystyrene latex
   beads expressing yellow-green fluorescence (Sigma-Aldrich) were diluted
   in PBS to 5 × 10^8 beads per mL. Beads were incubated with iBMDMs at
   the indicated number of latex beads per macrophage for 4 h. Following
   bead exposure, macrophages were washed with PBS and fixed in 4%
   paraformaldehyde solution. The percent uptake of yellow-green
   fluorescence latex beads was quantified by flow cytometry as described
   above.

   For surface expression assays, plated cells were washed with PBS and
   lifted from plates by scrapping. Cells were then incubated with an Ab
   mixture of APC ITGAM (CD11b) and PE ITGB2 (CD18) or PE ITGB2-ITGAL
   (LFA1) (BioLegend; 1:400 in PBS) for 20 min at room temperature in
   light-free conditions. Labeled cells were then washed two times with
   PBS and fixed in 2% paraformaldehyde solution before relative surface
   expression was quantified by flow cytometry. The mean fluorescence
   intensity was quantified from samples using FlowJo software (BD).

CRISPR screen and analysis.

   Mouse Brie CRISPR knockout pooled library was a gift from David Root
   and John Doench (Addgene number 73633) ([161]45) and infected with our
   smooth mEmerald-GFP Mab reporter strain at an MOI of five for 4 h.
   Following the infection, macrophages were washed with PBS then fixed
   with 4% paraformaldehyde. Infected library was then sorted using a
   Bio-Rad S3e cell sorter to isolate mEmerald+ and mEmerald-
   macrophages.1.5 to 2.5 × 10^7 cells were sorted into each bin from
   triplicate experiments. Genomic DNA was isolated from each sorted
   population using Qiagen DNeasy kits after reversing DNA cross-links
   following a 55°C incubation overnight. Amplification of sgRNAs by PCR
   was performed as previously described using Illumina compatible primers
   from IDT ([162]45) and amplicons were sequenced on an Illumina NextSeq
   500 at the Genomics Core at Michigan State University.

   Sequenced reads were first trimmed to remove any adapter sequence and
   to adjust for the p5 primer stagger. We used model-based analysis of
   genome-wide CRISPR-Cas9 knockout (MAGeCK) to map reads to the sgRNA
   library index without allowing for any mismatch. Subsequent sgRNA
   counts were median normalized to control sgRNAs in MAGeCK to account
   for variable sequencing depth. To test for sgRNA and gene enrichment,
   we used the “test” command in MAGeCK to compare the distribution of
   sgRNAs in the GFP^+ and GFP- bins. The MAGeCK output Gene Summary
   provided a ranked list of genes significantly underrepresented in the
   mEmerald+ bin based on 4 independent sgRNAs that was used to curate the
   candidate gene list.

CRISPR-targeted knockouts.

   Individual sgRNAs were cloned as previously described using sgOPTI that
   was a gift from Eric Lander (Addgene plasmid no. 85681) ([163]45). In
   short, annealed oligonucleotides containing the sgRNA targeting
   sequence were phosphorylated, then cloned into a dephosphorylated and
   BsmBI (New England Biolabs) digested sgOPTI. To facilitate rapid and
   efficient generation of sgRNA plasmids with different selectable
   markers, we used sgOPTI-Blasticidin-Zeocin (BZ) and
   sgOPTI-Puromycin-Ampicillin (PA) as previously described ([164]45). The
   simultaneous use of sgOPTI-BZ and sgOPTI-PA plasmids allowed for pooled
   cloning in which a given sgRNA was ligated into a mixture of
   BsmBI-digested plasmids. Successful transformants for each plasmid were
   selected by plating on ampicillin or zeocin in parallel. Next, two
   sgRNA-plasmid constructs per gene were packaged into lentivirus as
   previously described ([165]45), then used to transduce Cas9+ iBMDMs.
   Transductants were selected with blasticidin and/or puromycin then
   genomic DNA was isolated and PCR was used to amplify edited regions and
   sent for Sanger Sequencing (Genewiz). The resultant ABI files were used
   for Tracking of Idels by Decomposition (TIDE) analysis to assess the
   frequency and size of indels in each population compared to control
   macrophages ([166]61). Three sgRNAs were targeted per gene and one
   targeted line was selected for follow-up studies with editing
   efficiency 60% to 95% for each gene. sgRNA sequences and primers for
   Tide analysis are included in [167]Table S3.
   TABLE S3

   sgRNA and Primer sequences. The sgRNAs used to target candidate genes
   and the primers used to amplify the edited regions for TIDE analysis.
   Download [168]Table S3, XLSX file, 0.01 MB^ (10.2KB, xlsx) .

   Copyright © 2023 Gilliland et al.

   This content is distributed under the terms of the [169]Creative
   Commons Attribution 4.0 International license.

Quantification of sulfated glycosaminoglycans.

   Sulfated Glycosaminoglycans (sGAGs) were isolated following the Blyscan
   Sulfated Glycosaminoglycan Assay (BioVision). In short, macrophages
   were washed with PBS, then 1 mL of Papain Extraction Reagent containing
   1 mg/mL of Papain (Sigma-Aldrich) was added. Plates were incubated in
   5% CO[2] at 37°C for 1 min. Cell Suspensions were transferred to 1.5 mL
   microcentrifuge tubes and immediately placed in an ice-water bath. To
   extract sGAGs, 200 μL of cell suspension was transferred to a 1.5 mL
   microcentrifuge tube and incubated at 65°C for 3 h.

   sGAG were then quantified using the Blyscan Sulfated Glycosaminoglycan
   Assay Protocol (BioVision). Briefly, 25 μL of test sample was
   transferred into a 1.5 mL microcentrifuge tube. Reagent blanks, sGAG
   standards and test samples were prepared according to the Blyscan
   Sulfated Glycosaminoglycan Assay General Protocol and adjusted to a
   100 μL final volume. To stain sGAGs, 1 mL of Blyscan Dye Reagent
   (BioVision) was added to each tube, then tubes were placed in a gentle
   mechanical shaker for 30 min at room temperature. To pellet stained
   sGAGs, tubes were centrifuged at 13,000 × g for 10 min, then
   supernatant was discarded. To release the recovered sGAGs, 500 μL of
   Dissociation Reagent (BioVision) was added to each tube and vortexed
   until pellets were dissolved. 200 μL of reagent blanks, sGAG standards
   and test samples were transferred to a 96-well plate and absorbance
   readings at 656 nm were obtained from an Agilent BioTek Synergy H1
   Hybrid Multi-Mode Plate Reader.

RNA isolation and quantitative real-time PCR.

   Macrophages were resuspended in 500 μL of TRIzol reagent (Life
   Technologies) and incubated for 5 min at room temperature. One
   hundred μL of chloroform was added to the homogenate, vortexed and
   centrifuged at 10,000 × g for 18 min at 4°C to separate nucleic acids.
   The clear, RNA containing layer was removed and combined with equal
   parts ethanol. This mixture was placed into a collection tube and
   protocols provided by the Zymo Research Direct-zol RNA extraction kit
   were followed. Quantity and purity of the RNA were checked using a
   NanoDrop and diluted to 5 ng/μL in nuclease-free water.

   PCR amplification of the RNA was completed using the One-step Syber
   Green RT-PCR kit (Qiagen). 25 ng of total RNA was added to a master mix
   reaction of the provided RT mix, Syber green, gene specific primers (5
   μM of forward and reverse primer), and nuclease-free water. For each
   biological replicate (triplicate), reactions were conducted in
   technical triplicates in 96-well plates. PCR product was monitored
   using the QuantStudio3 (ThermoFisher). The number of cycles needed to
   reach the threshold of detection (Ct) was determined for all reactions.
   The mean CT of each experimental sample in triplicate was determined.
   The average mean of glyceraldehyde 3-phosphate dehydrogenase (GAPDH)
   was subtracted from the experimental sample mean CT for each gene of
   interest (dCT).

Bioinformatics analysis.

   Gene set enrichment analysis (GSEA) was used to identify enriched
   pathways in the MAGeCK analyzed data set based on the full ranked list
   as done previously. For GSEA analysis, the “GSEA Preranked” function
   was used to complete functional enrichment using default settings for
   Kyoto Encyclopedia of Genes and Genomes (KEGG) and Gene Ontology terms.

   DAVID analysis was used to identify enriched pathways within the
   positive regulators of Mab uptake. Using the robust rank aggregation
   (RRA) the top enriched positive regulators (2-fold change with at least
   3 sgRNAs) were used as a “candidate list.” Functional analysis and
   functional annotation analysis were completed, and top enriched
   pathways and protein families were identified.

Statistical analysis.

   Statistical analysis and data visualization were performed using Prism
   version 9 (GraphPad Software) or Biorender, as indicated in figure
   legends. Data are presented, unless otherwise indicated, as the mean ±
   SD and individual data points are shown for each experiment. For
   parametric data, one-way or two-way ANOVA followed by Tukey’s or
   Dunnett’s multiple-comparison test were used to identify significant
   differences between multiple groups.

Data availability.

   Raw sequencing data in FASTQ and processed formats are available for
   download from NCBI Gene Expression Omnibus (GEO) accession number
   [170]GSE221413.

ACKNOWLEDGMENTS