THP deficiency increases urinary tract UPEC burdens and tissue histopathology. Prior studies have identified the heightened UTI susceptibility of THP-deficient mice at 24 hours after infection (26, 27). To assess the sustained effect of THP deficiency, we used an established model of UTI with cystitis strain UTI89 (37) in THP+/+ (WT) and THP–/– (KO) mice (34). Consistent with previous findings (26, 27), THP-KO mice exhibited persistent increased bacteriuria (Figure 1A) and temporarily elevated bacterial tissue loads (Figure 1, B and C) compared with WT mice throughout the infection course. Bladder and kidney sections were examined by a blinded veterinary pathologist and scored on a 0–4 scale, considering pathologic features such as intraluminal bacteria, submucosal edema, and suppurative (pus-forming) pyelonephritis. UPEC-infected THP-KO mice displayed more severe tissue pathology compared with WT counterparts (Figure 1, D and E). Infected bladders were marked by increased immune infiltration in the urinary epithelium and submucosa (Figure 1F). Suppurative pyelonephritis was the most common renal histopathology and, when evident, tended to occur within, or adjacent to, the renal pelvis (Figure 1G). Hydronephrosis and lymphohistiocytic pyelonephritis were occasionally observed. No differences in tissue histopathology were observed between mock-infected WT and THP-KO groups.

THP deficiency increases urinary tract bacterial burdens and tissue pathology. (A–C) Time course of urine (A), bladder (B), and kidney (C) UPEC burdens from WT and THP-KO mice. Urine samples were collected from mice at multiple time points up until tissue collection. (D and E) Bladder (D) and kidney (E) pathology scores on 1 and 3 dpi. (F and G) Representative H&E images of day 1 bladders (F) and day 3 kidneys (G) from UPEC-infected or mock-infected mice. Scale bars: 110 μm (F) and 210 μm (G). Arrowheads point to polymorphonuclear cell infiltration (black) and polymorphonuclear cell aggregates (blue). Experiments were performed at least twice with data combined. n = 18–46/time point (A), n = 11–31 (B and C), or n = 5–15 (D and E). Box-and-whisker plots show median, all points, and 25–75th percentiles (A–C). Points represent individual samples; lines indicate medians with interquartile ranges (D and E). Data were analyzed by Mann-Whitney U test (A–C) and Fisher’s exact test (D and E). *P < 0.05; **P < 0.01.

THP deficiency alters bladder neutrophil infiltration and the effect of neutrophil depletion during UTI. We evaluated bladder and kidney immune cell infiltration by flow cytometry, surveying the total immune cell fraction (CD45+ [P1]), as well as neutrophils (Ly6G+), nonmyeloid (CD11b–CD11c–), myeloid (CD11b+/–CD11c+/– [P3]), myeloid antigen presenting cells (APCs; MHC-II), and myeloid non-APC subpopulations (gating scheme in Figure 2A). At 3 days postinfection (dpi), THP-KO mice had higher proportions and counts of CD45+ cells compared with WT-infected mice, although counts in the kidneys did not reach significance (P = 0.076). No differences were observed in later time points or mock controls (Figure 2, B–E). Bladder neutrophil proportions and counts were elevated in infected THP-KO mice compared with WT mice at 3 dpi, with no observed differences in kidneys or mock controls (Figure 2, F–I). At 7 dpi, THP-KO bladders showed depressed frequency and counts of myeloid cells with decreased myeloid non-APC and myeloid APC numbers (Supplemental Figure 1; supplemental material available online with this article; https://doi.org/10.1172/jci.insight.180024DS1). Other bladder immune cell subpopulations did not differ between groups. In the kidneys, minimal differences in other subpopulations were noted, including an increased proportion and number of lymphocytes and reduced proportion of myeloid lineages at 3 dpi (Supplemental Figure 2). Under mock-infected conditions, THP-KO mice exhibited a slight but significant decrease in lymphocyte and APC counts and an increased proportion of non-APC myeloid cells (Supplemental Figure 2).

THP deficiency increases bladder neutrophil infiltration and reduces bacterial burdens upon neutrophil depletion. (A) Gating strategy for quantifying immune populations of interest with a focus on neutrophils (Ly6G+) and myeloid lineages (CD11b+/–, CD11c+/–). (B–E) Frequency and counts of CD45+ cells (P1) cells in bladder or kidneys at 3 and 7 dpi. Mock-infected samples from both time points were combined prior to analyses. (F–I) Frequency and counts of neutrophils (Ly6G+) from CD45+ populations infiltrating bladder or kidneys. Mice were administered anti-Ly6G or IgG isotype control prior to bacterial inoculation and on 2 and 4 dpi. (J) Urine sediment scores at 6 dpi. (K) Urine UPEC burdens at 6 dpi. (L and M) Bladder and kidney UPEC burdens at 7 dpi. Experiments were performed at least twice and combined. n = 5–32/group (B–I) and n = 7–10 (J–M). Box-and-whisker plots show median, all points, and 25–75th percentiles (B–I). Data were analyzed by Mann-Whitney U test (B–I, and K–M) and Fisher’s exact test (J). *P < 0.05; **P < 0.01; ***P < 0.001.

Neutrophil depletion exacerbates bacterial burdens and promotes chronic infection depending on the extent of neutrophil reduction (10). To evaluate the effect of neutrophil depletion in THP-deficient mice, we administered anti-Ly6G antibody or isotype IgG i.p. every 48 hours, from 0 to 6 dpi. On 6 dpi, urine sediment was scored for the presence of polymorphonuclear (PMN) cells on a scale of 0–5 as described previously (10). Urine and tissues were collected on 7 dpi to quantify bacterial burden. Anti-Ly6G antibody treatment significantly reduced urine sediment PMN scores in WT mice but had no such effect in THP-KO mice (Figure 2J). Additionally, anti-Ly6G antibody treatment resolved bacterial urine and tissue burdens between WT and THP-KO mice, highlighting the importance of neutrophils in mediating increased susceptibility (Figure 2, K–M). We next probed the crosstalk between cyclooxygenase 2 (COX-2), a critical enzyme that initiates inflammatory cascades, and THP-mediated UTI susceptibility. COX-2 downregulates THP kidney expression, and COX-2–/– mice are hypersusceptible to UTI (38), thus we examined the effect of COX-2 in our model. Mice were treated with diclofenac, a COX-2 inhibitor (39), in drinking water from 0 to 6 dpi, with tissues collected at 7 dpi. We found no differences in tissue burdens between diclofenac-treated and mock-treated WT and THP-KO mice (Supplemental Figure 3), suggesting that enhanced UTI susceptibility in THP-KO mice is independent of COX-2 inflammatory pathways in this model. Together, these findings highlight that elevated neutrophils are a distinctive feature of THP deficiency in UTI, which, paired with enhanced bacterial burdens, suggest impaired neutrophil activity in THP-KO mice.

Murine urinary THP levels and glycosylation change minimally during UTI. Clinical studies have linked UMOD variants (40) and reduced THP production (41, 42) with enhanced UTI risk, although a cross-sectional study found no differences in urinary THP levels between pediatric patients with UTI and controls (43). Similarly, we observed no variations in urinary THP levels between mock-infected and UPEC-infected WT mice (Figure 3A). To delineate the N-glycan profile of murine THP and assess the effect of UTI on THP glycosylation, we collected urine over 96 hours after inoculation and profiled THP glycosylation patterns using Matrix-Assisted Laser Desorption/Ionization TOF/TOF mass spectrometry (MALDI-TOF/TOF MS). Similar to human THP (44–46), murine THP contained multiple bi-, tri-, and tetraantennary sialylated and/or fucosylated complex type N-glycans (Figure 3B). The highest intensity peak (m/z 4,588) represented a tetraantennary, tetrasialylated and fucosylated N-glycan (Figure 3B and Supplemental Table 1), which matches the most abundant N-glycan on human THP (44, 45). Other high-intensity peaks were observed at m/z 2,967; 3,777; and 4,226. In UPEC-infected THP samples, these 4 most abundant structures remained the same, albeit with some proportional differences: the m/z 2,967 peak increased and the m/z 4,588 peak decreased relative to mock-treated spectra (Figure 3C and Supplemental Table 1). We quantified total sialic acids released from murine THP by abeling sialic acids with 1,2-diamino4,5-methylenedioxybenzene (DMB) and measuring via high-performance liquid chromatography (HPLC) with fluorescence detection. No differences in N-glycolylneuraminic acid (Neu5Gc), N-acetylneuraminic acid (Neu5Ac), or total sialic acid levels were observed between mock-infected and UPEC-infected samples (Table 1). Samples from THP-KO mice showed significantly reduced Neu5Ac and total sialic acid compared with WT mock samples, validating that THP was the primary source of sialic acid (Table 1). Together, these data reveal conserved glycosylation patterns, including sialylation by Neu5Ac, in murine and human THP that are retained during murine UTI.

Tamm-Horsfall protein levels and glycosylation patterns change minimally during urinary tract infection in vivo. (A) THP urine levels normalized to creatinine in mock-infected (or baseline day 0) and infected mice (1–4 dpi combined). Box-and-whisker plots show median, all points, and 25-75th percentiles (n = 9–12). Data were collected from 3 experiments. N-glycan MALDI-TOF profiles of THP isolated from WT mice that were either mock- (B) or UPEC-infected (C). Data represent 1 analysis of purified THP harvested from pooled urine from WT mice (n = 15 mock, n = 28 UPEC) collected as a part of 2 independent experiments. Prominent peaks with proportional differences between UPEC-infected and mock samples (m/z 2,967 and 4,588) are highlighted in teal. Data in A were analyzed by Mann-Whitney U test.

Sialic acid concentration in THP purified from mouse urine

Murine neutrophils undergo NETosis during UTI, and THP deficiency alters neutrophil subpopulations. To investigate whether differences in neutrophil abundance corresponded with differences in neutrophil function, we characterized NETosis using several methods. Nucleic acid dyes (e.g., Hoechst) and non–cell permeable SYTOX dyes have distinguished NETosis from other forms of cell death in mixed-cell populations (47–49). Additionally, plasma membrane permeability can be confirmed using a live/dead amine-reactive dye that only labels intracellular amines if the membrane is compromised (50). In classical NETosis, neutrophils permeabilize and expel decondensed chromatin, whereas during nonclassical NETosis, neutrophils release DNA but retain viability and effector functions (51, 52). We subjected mouse urine collected 24 hours after inoculation to flow cytometry. Neutrophils (PMNs) were identified as CD11b+Ly6G+ and were further gated based on presence of extracellular DNA (exDNA; SYTOX Orange [SO)] and plasma membrane permeability (Live/Dead stain) as depicted in Figure 4A. We identified 4 unique populations: viable PMNs (SO–Live/Dead– [Q4]), dead PMNs (SO–Live/Dead+ [Q3]), dead PMNs with exDNA (SO+Live/Dead+ [Q2]), and viable PMNs with exDNA (SO+Live/Dead– [Q1]). WT and THP-KO mice displayed increased urinary neutrophils during infection compared with mock-infected counter parts (Figure 4B). UPEC infection elevated proportions and total counts of exDNA cells (and dead and viable subsets), most notably in WT mice, compared with their mock-infected counterparts (Figure 4, C–H). Uniquely, WT mice showed elevated frequency and counts of viable PMNs with exDNA in response to UPEC infection and compared with UPEC-infected THP-KO mice (Figure 4, E and F). Dead PMNs were observed at a higher frequency in WT mice in response to infection (Figure 4, I and J). In both WT and THP-KO mice, frequency of live PMNs was reduced during infection, but WT sustained higher total live PMN numbers during infection compared with mock-infected WT mice (Figure 4, K and L).

Neutrophil exDNA+ populations are decreased in THP-deficient mice during UTI. (A) Gating strategy for quantifying neutrophil (PMNs, Ly6G+CD11b+ [P4]) subpopulations of interest including viable exDNA+ (extracellular DNA[SYTOX Orange]+Live/Dead–), dead exDNA+ (exDNA+Live/Dead+), dead PMNs (exDNA–Live/Dead+), and live PMNs (exDNA–Live/Dead–). (B) Total PMNs (P4) per mL of urine. (C and D) Frequency out of total PMNs and counts of total exDNA+ cells (Q1+Q2). (E and F) Frequency of and counts of viable exDNA+ cells. (G and H) Frequency of and counts of dead exDNA+ cells. (I and J) Frequency of and counts of dead exDNA– cells. (K and L) Frequency of and counts of live exDNA– cells. Experiments were performed at least twice and combined. n = 11–15/group. Box-and-whisker plots show median, all points, and 25–75th percentiles. Data were analyzed by 2-way ANOVA with uncorrected Fisher’s LSD test. *P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001.

To confirm the presence of NETs, we assessed citrullinated modifications on Histone 3 (citrullinated histone H3 [H3Cit]), catalyzed by the enzyme peptidylarginine deiminase 4 (PAD4), which contributes to chromatin decondensation and NET formation (53). Neutrophils (PMNs) were identified as CD11b+Ly6G+ and were further gated based on staining for presence of H3Cit and plasma membrane permeability (Live/Dead stain) as depicted in Figure 5A. During infection, WT neutrophils displayed higher frequency of NETosis compared with THP-KO mice (Figure 5B). When separated into nonclassical NETosis and classical NETosis subsets based on membrane permeability, WT neutrophils showed greater proportions of nonclassical NETosis and decreased classical NETosis compared with THP KO in both infected and mock-infected states (Figure 5, C and D). The presence of NETs in WT and THP-KO urine samples was visualized by immunofluorescence microscopy using antibodies for neutrophils (MPO), NETosis (H3Cit), and THP (Figure 5, E and F).

Neutrophil NETosis populations are altered in THP-deficient mice during UTI. (A) Gating strategy for quantifying neutrophil (PMNs, Ly6G+CD11b+ [P4]) subpopulations of interest with a focus on NETosis (H3Cit+Hoechst+), and nonclassical (H3Cit+Live/Dead–) and classical (H3Cit+Live/Dead+) subsets. (B) Frequency of NETosis in total PMNs. (C and D) Frequency of nonclassical and classical subsets in total NETosis population. (E and F) Urine samples from UPEC-infected WT and THP-KO mice were mounted on slides and NETs were visualized via immunofluorescence using antibodies against myeloperoxidase (MPO, cyan), citrullinated histone H3 (H3Cit, red), and THP (green). Nucleic acids were stained using Hoechst dye (blue). Yellow arrowheads point to NET structures depicted as strands of DNA dotted with MPO staining. Experiments were performed at least twice and combined. n = 6–17/group. Box-and-whisker plots show median, all points, and 25–75th percentiles. Data were analyzed by 2-way ANOVA with uncorrected Fisher’s LSD test. *P < 0.05; **P < 0.01; ****P < 0.0001.

THP enhances NETosis in human neutrophils with minimal effects on cellular proteins. To determine if THP’s effect on NETosis extended to human models, we measured NETosis formation in human neutrophils with and without THP exposure at physiologic concentrations. After 2.5 hours of PMA stimulation, NETosis was measured by detection of fluorescently labeled exDNA as described previously (13). THP pretreatment increased exDNA in PMA-stimulated cells but not in unstimulated cells (Figure 6A). To identify cellular processes affected by THP, we performed quantitative proteomics of neutrophils under these same 4 conditions: mock-treated unstimulated (UnTx), THP-treated, PMA-stimulated, and PMA-stimulated and THP-treated (PMA+THP). PMA stimulation was the primary driver of variation between samples as shown by PCA (Figure 6B) and resulted in depleted neutrophil granule and NETs-related proteins likely due to the release of these proteins from activated cells (Supplemental Figure 4). Eight shared proteins were increased in THP and PMA+THP samples compared with their mock-treated counterparts (Figure 6C and Supplemental Tables 2 and 3). These proteins included THP itself (UMOD) and other known urinary proteins: apolipoprotein D, alpha-1-Microglobulin/Bikunin Precursor (AMBP), kininogen (KNG1), and galectin 3 binding protein (LGALS3BP) (54). MS analysis of purified THP confirmed presence of THP (77.43%). Several other proteins were detected at > 0.5%: apolipoprotein D, E3 ubiquitin-protein ligase (MIB2), protein AMBP, Metabolism Of Cobalamin Associated D (MMADHC), LGALS3BP, β-actin (ACTB), and KNG1 (Supplemental Table 4). The remaining 3 shared proteins, not detected in purified THP, were related to cellular metabolism (ACSS2, SLC16A9) and immune signaling (IL-2R-gamma). In unstimulated cells, 10 additional proteins were differentially abundant between THP-treated and mock-treated conditions (Figure 6D and Supplemental Table 2) and included several related to translational regulation and protein turnover (EIF2AK4, POLR3F, UBAC1), second messenger signaling (CD38), cytokine receptor signaling (RNF41), mitochondrial metabolism (GLDC, ALDH5A1), phagosome acidification and fusion (RAB20), and chromatin remodeling (BICRAL). Gene ontology (excluding proteins detected in purified THP) identified mitochondrial respiratory chain complexes as significantly enriched in THP-treated conditions (Figure 6E). In PMA-stimulated cells, 16 unique proteins were differentially abundant between THP-treated and mock-treated conditions (Figure 6F and Supplemental Table 2). These included proteins involved in second messenger and cell signaling (PDE7A, FCSK), transcriptional and translational regulatory proteins (PUM1, E2F3, ZFP36L2, GTPBP6, CCDC86), complement-related and immune related proteins (CD59, CXCL8, C4BPA, CTSW), intracellular trafficking and cytoskeleton arrangement (GIPC2, NCOA4, XIRP2), and DNA/chromatin remodeling proteins (DNASE1L1, BOD1). Gene ontology analyses identified significant enrichment of tertiary granule and primary lysosome pathways in THP-treated conditions, as well as nonsignificant enrichment of specific, secretory, ficolin-1–rich and pigment granules as well as vacuolar and vesicle pathways (Figure 6G). This proteomic profiling suggests that THP induces subtle differential responses related to mitochondrial metabolism in the absence of PMA stimulation and affects multiple nuclear, organelle, and cytoskeletal functions in PMA-stimulated conditions.

THP modestly alters neutrophil responses to PMA stimulation. (A) NETosis was assessed by released dsDNA (detected by PicoGreen dye) and expressed as arbitrary units (AU) of fluorescence normalized to mock-treated, unstimulated controls. Neutrophils were subjected to tandem mass tag-based proteomics profiling. (B) Principal component analysis of neutrophils that were untreated (UnTx), treated with THP (THP), untreated with PMA stimulation (PMA), and THP-treated with PMA stimulation (THP+PMA). Points represent individual samples, colored by treatment, with paired donor samples indicated by matched symbol. (C) Venn diagram of proteins differentially detected in THP-treated samples compared with untreated samples in PMA-stimulated (PMA and THP+PMA) and unstimulated (UnTx and THP) conditions. (D and E) Volcano plot (D) and gene set enrichment analysis (E) of differential proteins in untreated versus THP-treated samples. (F and G) Volcano plot (F) and gene set enrichment analysis (G) of differential proteins in PMA versus THP+PMA samples. Experiments were performed independently 3 times and combined, n = 5 (A), or as one independent experiment, n = 4 (B–G). Bar plots show median, all points, and 95% CI (A). Data were analyzed by 2-way ANOVA with Šídák’s multiple-comparisons test (A). Differential proteins (C–G) were identified via log2 fold change > 1.25 and moderated 2-tailed t test followed by multiple-hypothesis testing correction using the Benjamini-Hochberg procedure with a FDR-adjusted P < 0.05. Individual proteins are listed in Supplemental Table 2.

THP increases NETosis in human neutrophils in a ROS-dependent manner. To determine whether human neutrophils were similarly affected by THP, we modified our flow cytometry strategy for human neutrophils. Peripheral human neutrophils were treated with human THP, stimulated with PMA, and analyzed via flow cytometry. Single cells were gated for the presence of extracellular/surface neutrophil granule content (MPO) and exDNA (SO) to identify double positive cells (MPO+SO+ [P3]) (Figure 7A). P3 cells were further separated based on Hoechst and Live/Dead staining into viable exDNA+ (HoechstloLive/Dead–) and dead exDNA+ (HoechsthiLive/Dead+) subsets. Consistent with the fluorescence-based NETosis assay (Figure 6A), THP significantly increased total exDNA+ cells in PMA-stimulated conditions but not in unstimulated cells (Figure 7B). Furthermore, THP enhanced viable exDNA+ (Figure 7C) and dead exDNA+ (Figure 7D) subsets specifically in PMA-stimulated conditions. Classical and nonclassical NETosis are dependent on NOX 2–mediated (NOX2-mediated) production of ROS (21, 55, 56), and hydrogen peroxide (H2O2), as an exogenous source of ROS, is sufficient to stimulate NETosis (19). To examine the importance of ROS on THP-mediated effects, we compared viable and dead exDNA+ subsets in the presence of PMA, H2O2, or PMA and a NOX inhibitor diphenyleneiodonium (DPI). THP-mediated increase in viable exDNA+ cells occurred in the presence of both PMA and H2O2 but was abrogated with the addition of DPI (Figure 7E). In contrast, no significant differences were observed in dead exDNA+ subsets under these same conditions (Figure 7F). Together, these data suggest that THP-mediated effects are in part dependent on ROS, specifically in viable exDNA+ cells.

THP exposure increases viable exDNA+ human neutrophils. (A) Gating strategy for quantifying neutrophil exDNA+ (SYTOX Orange+MPO+ [P3]) subpopulations of interest focusing on viable (HoechstloLive/Dead–) and dead (HoechsthiLive/Dead+) populations. (B–D) Cell counts for total exDNA+ (B), viable exDNA+ (C), and dead exDNA+ (D) across treatment groups. Human neutrophils were pretreated with THP or were mock treated and stimulated with either PMA, H2O2, or PMA + DPI (ROS inhibitor). (E and F) Frequency of viable exDNA+ cells (E) or dead exDNA+ cells (F) normalized to frequency of unstimulated cells from the same donor. Experiments were performed in at least 4 independent experiments and combined, n = 9 (B–D), or n = 5 (E and F). Box-and-whisker plots show median, all points, and 25–75th percentiles (B–F). Data were analyzed by 2-way ANOVA with Šídák’s multiple-comparisons test (B–F). *P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001.

To determine if THP effects were specific to cells undergoing NETosis, we assessed the presence of H3Cit via flow cytometry. Neutrophils identified as MPO+SO+ were further gated based on H3Cit and plasma membrane permeability (Live/Dead stain) as depicted in Figure 8A. We included a caspase inhibitor (YVAD) to test whether blocking other forms of cell death was a potential mechanism of NETosis enhancement. Total NETosis was increased in all PMA-stimulated groups; however, THP further increased proportions of NETosis over PMA and PMA+YVAD groups (Figure 8B). When separated into classical and nonclassical NETosis based on Live/Dead staining, PMA stimulation across all groups significantly increased both forms over mock-stimulated counterparts (Figure 8, C and D). Uniquely, THP further increased nonclassical NETosis over other stimulated groups (Figure 8C). THP-treated, PMA-stimulated cells also displayed higher median fluorescence intensity of H3Cit staining compared with PMA and PMA+YVAD groups (Figure 8, E and F). Inhibition of caspases (YVAD) under these experimental conditions did not alter frequency of NETosis, in line with prior findings (57).

THP exposure increases NETosis in human neutrophils. (A) Gating strategy for quantifying neutrophil exDNA+ (SYTOX Orange+MPO+ [P4]) subpopulations of interest focusing on NETosis (HoechstvarH3Cit+ [P5]) populations that were viable (Live/Dead–) or dead (Live/Dead+). (B–D) Frequency for total NETosis (B), nonclassical NETosis (C), and classical NETosis (D). (E) Histogram plots depicting H3Cit staining. (F) Median fluorescence intensity of H3Cit. Experiments were performed in at least 4 independent experiments and combined, n = 7–8 (B–F). Bar plots show median, all points, and 95% CI (B–D and F). Data were analyzed by 2-way ANOVA with Tukey’s multiple-comparisons test (B–D and F). **P < 0.01; ***P < 0.001; ****P < 0.0001.

THP alters NETosis and other forms of cell death as measured by imaging flow cytometry. Prior to DNA release, cells destined for NETosis undergo cellular remodeling, including cytoskeletal and endoplasmic reticulum disassembly (58), vacuolization, autophagy, superoxide production (57), and lastly, chromatin swelling and nuclear envelope rupture (59). Live cell imaging or imaging flow cytometry techniques have revealed predictable morphologic changes that delineate NETosis from other forms of cell activation and death (22, 47, 58, 60, 61). To assess whether THP altered neutrophil morphology, we adapted an imaging flow cytometry method and algorithm from prior work (60) to identify NETs, NET precursors, and other forms of cell death. Using this method, cells are distinguished into 6 types: healthy (Type I), live cells with decondensed nuclei (NET precursors, Type II), NETs (Type III), DNA NET fragments (Type IV), dead cell condensed nuclei (Type V), and dead cell diffuse nuclei (Type VI). Neutrophils were pretreated with THP or an estimated equivalent amount of sialic acid, stimulated with PMA; stained with α-MPO-FITC, SO, Hoechst 33342, and Live/Dead Near I/R; subjected to imaging flow cytometry; and gated as shown in Figure 9A. Cells were separated from debris based on bright-field (BF) area and Hoechst+ staining. Within populations with higher extracellular DNA (SO staining beyond the BF cell margins) area, NETs and DNA NET fragments were distinguished by higher or lower Hoechst intensity, respectively. Remaining cells were further gated to collect focused, single cells and separated based on SO intensity (indicating membrane permeability) and Hoechst area (indicating nuclear area). Dead cells (higher SO intensity) with condensed nuclei or diffuse nuclei were delineated by lower and higher Hoechst area, respectively. Healthy cells and live cells with decondensed nuclei were demarcated by Hoechst area. Representative images of each cell type are shown in Figure 9B with dead cell types (Type V and VI) confirmed by staining Live/Dead+. The sum of NET precursors and NETs (Type II and Type III) were higher in THP-treated cells compared with PMA alone, although there were no differences in NET fragments (Type IV; Figure 9, C and D). Additionally, the PMA+THP group exhibited decreased dead cell (Type V and VI) frequencies compared with PMA controls (Figure 9, E and F). Live cell and Hoechst+ populations were not different across groups (Figure 9, G and H). We also identified a circularity feature, which gives higher scores to features closely resembling a circle, as significantly higher in PMA+THP compared with PMA alone (Figure 9I). No effects were seen with sialic acid treatment (Figure 9), and no effects of THP were observed in the absence of PMA stimulation, with the exception of a slight but significant decrease in circularity (Supplemental Figure 5, A–G). Overall, these analyses reveal that THP enhances the frequency of NETs and NET precursors in the presence of a NETosis stimulus.

THP exposure alters proportions of NETs, NET precursors, and other cellular morphologies. Neutrophils were stained with anti–MPO-FITC, SYTOX Orange, Hoechst, and Live/Dead stain and visualized for fluorescence and bright-field (BF). (A and B) Gating strategy (A) of neutrophils subpopulations with representative images shown (B). (C–G) Frequency of NETs and NET precursors (Type III and II) (C), NET fragments (Type IV) (D), dead cells with condensed nuclei (Type V) (E), dead cells with decondensed nuclei (Type VI) (F), and live cells (Type I) (G). (H) Frequency of Hoechst+ cells. (I) Circularity scores across groups. Experiments were performed in 8 experiments and combined, n = 8. Box-and-whisker plots show median, all points, and 25–75th percentiles (C–I). Data were analyzed by 1-way ANOVA with Holm-Šídák’s multiple-comparisons test (C–I). *P < 0.05. Sia, sialic acid.