www.nature.com/mi ARTICLE OPEN Deficiency in the glycosyltransferase Gcnt1 increases susceptibility to tuberculosis through a mechanism involving neutrophils Kaori L. Fonseca 1,2,3,4, Ana Raquel Maceiras1,2, Rita Matos1,4,5, Luisa Simoes-Costa1,2, Jeremy Sousa1,2, Baltazar Cá1,2, Leandro Barros1,2, Ana Isabel Fernandes1,2, Stefan Mereiter1,5, Ricardo Reis6, Joana Gomes1,5, Gustavo Tapia7, Paula Rodríguez-Martínez7, Montse Martín-Céspedes7, Sergo Vashakidze8, Shota Gogishvili8, Keti Nikolaishvili8, Rui Appelberg1,2,4, Fátima Gärtner1,4,5, Pedro N. S. Rodrigues1,2,4, Cristina Vilaplana9, Celso A. Reis1,4,5,10, Ana Magalhães1,5 and Margarida Saraiva1,2 Modulation of immunity and disease by glycans is increasingly recognized. However, how host glycosylation shapes and is shaped by tuberculosis remains poorly understood. We show that deficiency in the glucosaminyl (N-acetyl) transferase 1 (Gcnt1), a key enzyme for core-2 O-glycans biosynthesis, drives susceptibility to Mycobacterium tuberculosis infection. The increased susceptibility of Gcnt1 deficient mice was characterized by extensive lung immune pathology, mechanistically related to neutrophils. Uninfected Gcnt1 deficient mice presented bone marrow, blood and lung neutrophilia, which further increased with infection. Blood neutrophilia required Gcnt1 deficiency in the hematopoietic compartment, relating with enhanced granulopoiesis, but normal 1234567890();,: cellular egress from the bone marrow. Interestingly, for the blood neutrophilia to translate into susceptibility to M. tuberculosis infection, Gnct1 deficiency in the stroma was also necessary. Complete Gcnt1 deficiency associated with increased lung expression of the neutrophil chemoattractant CXCL2. Lastly, we demonstrate that the transcript levels of various glycosyltransferase-encoding genes were altered in whole blood of active tuberculosis patients and that sialyl Lewis x, a glycan widely present in human neutrophils, was detected in the lung of tuberculosis patients. Our findings reveal a previously unappreciated link between Gcnt1, neutrophilia and susceptibility to M. tuberculosis infection, uncovering new players balancing the immune response in tuberculosis. Mucosal Immunology (2020) 13:836–848; https://doi.org/10.1038/s41385-020-0277-7 INTRODUCTION glycosyltransferases involved in the biosynthesis of functional The host immune system has evolved to respond to infection selectin ligands, as Galnt1, Gcnt1, B4galt1, Fut4, Fut7 and St3gal4, whilst avoiding tissue damage. Deregulation of protective results in defective leukocyte recruitment7–9 and in homeostatic immune responses to Mycobacterium tuberculosis (Mtb) often alterations, namely pronounced blood neutrophilia.10,11 Selectin associates with tissue pathology.1 For example, an inadequate ligands are commonly decorated with O-glycans bearing a recruitment of neutrophils to Mtb-infected lungs may cause severe terminal sialyl Lewis x (sLex) (sialic acid α2,3Galβ1–4 (Fucα1,3) tissue damage. Exacerbated neutrophilia is thus linked with GlcNAc-R) epitope.6 Extensive O-glycosylation is also characteristic increased severity of infection in mouse2,3 and man.4 Under- of transmembrane or secreted gel-forming mucins,12 important standing the factors that control or disrupt protective immune components of the lung mucus. Mucins play key functions as networks during tuberculosis (TB) is therefore paramount for the surface receptors for the binding of several adhesion molecules, design of more efficient strategies and to inform on potential TB- mediators of the interaction with pathogens, and local modulators associated side effects of immune therapies to other diseases. of inflammatory responses.13,14 Pathogens, as Helicobacter pylori, The importance of glycosylation in innate and acquired immune may modulate the expression, turnover and glycosylation of host responses is emerging.5 Glycosylation regulates leukocyte recruit- mucins.15–18 How O-glycosylation contributes to host protection in ment from the vasculature to most tissues, through the TB is largely unknown. Deficiency in the α1,3-fucosyltransferase modulation of interactions between endothelial selectins Fut7 or Fut4/Fut7 double deficiency impaired T cell homing to the and their ligands in leukocytes.6,7 Deficiency in several lymph nodes during Mtb experimental infection, but the traffic of 1 i3S—Instituto de Investigação e Inovação em Saúde, Universidade do Porto, Porto, Portugal; 2IBMC—Instituto de Biologia Molecular e Celular, Universidade do Porto, Porto, Portugal; 3Programa de Pós-Graduação Ciência para o Desenvolvimento (PGCD), Instituto Gulbenkian de Ciência (IGC), Oeiras, Portugal; 4ICBAS—Instituto de Ciências Biomédicas Abel Salazar, University of Porto, Porto, Portugal; 5IPATIMUP—Instituto de Patologia e Imunologia Molecular da Universidade do Porto, Porto, Portugal; 6CDP-Centro de Diagnóstico Pneumológico do Porto, Porto, Portugal; 7UAB—Pathology Department, Universitat Autònoma de Barcelona, Hospital Universitari Germans Trias i Pujol, Barcelona, Spain; 8National Center for Tuberculosis and Lung Diseases (NCTLD), Tbilisi, Georgia; 9UAB—Experimental Tuberculosis Unit. Universitat Autònoma de Barcelona, CIBER Enfermedades Respiratorias. Fundació Institut d’Investigació en Ciències de la Salut Germans Trias i Pujol, Barcelona, Spain and 10FMUP—Faculdade de Medicina da Universidade do Porto, Porto, Portugal Correspondence: Ana Magalhães (amagalhaes@ipatimup.up.pt) or Margarida Saraiva (margarida.saraiva@ibmc.up.pt) These authors contributed equally: Kaori L. Fonseca, Ana Raquel Maceiras Received: 26 July 2019 Accepted: 12 February 2020 Published online: 13 March 2020 © Society for Mucosal Immunology 2020 Deficiency in the glycosyltransferase Gcnt1 increases susceptibility to. . . KL Fonseca et al. 837 effector T cells to the lung was normal.19 Both Fut7 deficient (−/−) Here, we reveal that core-2 GlcNAc transferase Gcnt1 deficient and Fut4−/−/Fut7−/− mice were initially shown to resist normally (Gcnt1−/−) mice11 display increased susceptibility to Mtb infection, to low dose aerosol infection,19 but a shorter survival of these mechanistically mediated by neutrophils and by Gcnt1 activity in mice upon infection was reported in a later study.20 stromal and hematopoietic compartments. Furthermore, we Mucosal Immunology (2020) 13:836 – 848 Deficiency in the glycosyltransferase Gcnt1 increases susceptibility to. . . KL Fonseca et al. 838 Fig. 1 Deficiency in Gcnt1 associates with increased susceptibility to Mtb infection. C57BL/6 (B6; black circles) or Gcnt1−/− (open circles) mice were infected by aerosol with Mtb strain HN878 with a low (a, c, e–g) or high (b, d, h, i, g) dose of bacteria. a, b The weight of the animals was monitored to determine survival curves, that included 10–13 animals in two independent experiments. Statistical analysis was performed with a log-rank (Mantel-Cox) test for the Kaplan Meyer curve. c, d At the indicated timepoints post-infection, the lungs of infected mice were collected and the bacteria burden determined by CFU enumeration. e, h At the indicated timepoints post-infection, lung pathology defined as the percentage of infiltrate per lobe (Fig. S1A, B), was determined upon H&E staining and morphometric analysis of the right upper lobes of infected lungs. The histopathologic features of the infected lungs were assessed and a relative score attributed (Table 1); the score obtained for necrosis is plotted in f and i. The pictures in g are H&E staining of representative animals within each experimental group. Arrowheads point to intra-alveolar necrotic debris, black diamond point to bronchiolar debris, black arrows point to perivascular lymphocytes, white arrows point to peribronchiolar lymphocytes and asterisks to calcification sites. Scale bar corresponds to 100 µm. In c each dot represents the Mean ± SEM and in d–h each dot represents an individual mouse of 6–12 in at least two independent experiments. In e, f red dots represent moribund Gcnt1−/− mice. Statistical analysis was performed using multiple t-test (c) or unpaired t-test (d–i). *p < 0.05; **p < 0.01; ***p < 0.01; ****p < 0.0001. Table 1. Lung histopathology analysis of Mtb-infected mice. Low dose High dose −/− day 30 day 60 Moribund Gcnt1 (3) day 27 −/− −/− Score B6 (12) Gcnt1 (13) B6 (7) Gcnt1 (6) B6 (7) Gcnt1−/− (10) Perivascular lymphocytes 0 1 8.30 2 91.7 23.1 28.6 50.0 28.6 16.7 3 76.9 71.4 100 50.0 71.4 83.3 Peribronchiolar lymphocytes 0 1 44.4 55.6 57.1 33.3 14.3 2 55.6 33.3 42.9 50 50.0 42.9 66.7 3 11.1 16.7 50.0 42.9 33.3 Bronchiolar debris 0 70.0 66.7 42.9 50 14.3 1 30.0 22.2 57.1 33.3 57.1 10.0 2 11.1 16.7 66.7 14.3 20.0 3 33.3 14.3 70.0 Intra-alveolar necrotic debris 0 1 16.7 46.2 57.1 16.7 33.3 2 66.7 30.8 14.3 50 66.7 28.6 3 16.7 23.1 28.6 33.3 71.4 100 Necrosis 0 100 92.3 100 100 42.9 1 28.6 10.0 2 7.70 28.6 10.0 3 100 20.0 4 60.0 Lung H&E sections of C57BL/6 (B6) or Gcnt1−/− infected animals were assessed and scored for the indicated histopathological features. Indicated is the % of animals with a certain score within the indicated experimental group. Indicated in brackets is the number of animals analysed per group. report for the first time altered expression of glycosyltransferases we used mice lacking Gcnt1, a key enzyme for the biosynthesis of associated with sLex core-2 O-glycan biosynthetic pathway in core-2 O-glycans.11 C57BL/6 and Gcnt1−/− mice were infected via whole blood of TB patients. Our findings support the modulation aerosol with Mtb strain HN878. Upon a low dose infection, 18.7% of O-glycosylation during Mtb infection, showing that alterations of Gcnt1−/− mice succumbed around 30 days post-infection, in this pathway compromise the ability of the host to deal with whereas, as expected, all C57BL/6 mice survived (Fig. 1a). infection in the most competent way. Strikingly, all Gcnt1−/− mice infected with high doses of Mtb succumbed within 40 days, while 90% of C57BL/6 mice survived throughout the 60 days of the experiment (Fig. 1b). Gcnt1−/− RESULTS mice, which survived the low dose of infection, showed a higher Deficiency in Gcnt1 is associated with higher susceptibility to Mtb lung bacterial burden than C57BL/6, particularly on day 60 post- infection infection (Fig. 1c). No significant differences in bacterial burdens of Studies with Fut4−/− and Fut7−/− hinted at a possible role for C57BL/6 or Gcnt1−/− mice infected with high doses of Mtb for the host glycosylation machinery in TB.19,20 As Fut4 and Fut7 27 days were observed (Fig. 1d), despite the different survival catalyse the terminal decoration of glycans, we reasoned that rates. studying the impact of an earlier step of glycan biosynthesis The percentage of lung lesion was quantified to measure lung would better reveal its biological effect during Mtb infection. Thus, pathology (Fig. S1A, B). As compared to C57BL/6, Gcnt1−/− mice Mucosal Immunology (2020) 13:836 – 848 Deficiency in the glycosyltransferase Gcnt1 increases susceptibility to. . . KL Fonseca et al. 839 infected with low doses of Mtb showed increased lung pathology, particularly high for the Gcnt1−/− mice, which succumbed to which was mostly pronounced on day 60 post-infection with infection (Fig. 1e, Fig. S1D). Lung histopathologic analysis of the increased scores for perivascular and peribronchial lymphocytes infected mice revealed an overall exacerbation of pathologic (Table 1, Fig. 1e and Fig. S1C). The extent of lung lesion was features in Gcnt1−/− mice, with the moribund mice presenting Mucosal Immunology (2020) 13:836 – 848 Deficiency in the glycosyltransferase Gcnt1 increases susceptibility to. . . KL Fonseca et al. 840 Fig. 2 Exacerbated neutrophilia drives increased susceptibility of Gcnt1−/− mice to Mtb infection. a–e At the indicated timepoints post- infection, the lungs of C57BL/6 (B6; black circles) or Gcnt1−/− (open circles) mice infected by aerosol with a low or high dose of Mtb strain HN878 were harvested and a cellular suspension prepared. a, b, e The relative expression of the indicated genes was determined by real-time PCR. c Percentages of the indicated immune cell populations were determined by flow cytometry. The gating strategy is shown in Fig. S2A. d Representative images of MPO staining (green) in lung sections of B6 and Gcnt1−/− mice infected with low or high doses of Mtb. Sections were counterstained with DAPI (blue). Scale bars correspond to 1 mm (left panels) and 100μm (right panels). In a, c each dot represents the Mean ± SEM for 10 animals in two independent experiments. Non-infected animals (day 0) were used as controls. b, e Represented are heatmaps of log2 relative expression of the indicated genes in lung samples of B6 and Gcnt1−/− mice 27 days after high dose Mtb infection. f C57BL/6 (B6, black circles) or Gcnt1−/− mice were infected by aerosol with a high dose of Mtb strain HN878. On day 18 post-infection the Gcnt1−/− mice were treated with the neutrophil-depleting anti-Ly6G mAb (black triangles) or an isotype control (open circles). The weight of the animals (5 per group) was monitored to determine survival curves. Statistical analysis was performed per time point with unpaired two- tailed Mann-Whitney test (a, c) or with log-rank (Mantel-Cox) test for the Kaplan Meyer curve (f). * refer to statistic differences between C57BL/ 6 or Gcnt1−/− mice. *p < 0.05; **p < 0.01; ***p < 0.001. severe inflammation, pronounced necrosis and calcification Mtb-infected Gcnt1−/− mice with a neutrophil-depleting anti- (Table 1, Fig. 1e–g). In infections with high doses of bacteria, a body22 or with an isotype control (Fig. S2D). To assess mouse higher frequency of infected Gcnt1−/− mice showed more than survival as the readout of this experiment, a high dose of infection 50% of lung infiltrates as compared to C57BL/6, although globally was performed. Whereas 80% (n = 4) of isotype control-treated no significant differences in relative lesion burden were noted Gcnt1−/− mice died between 25–35 days post-infection, only 20% (Fig. 1h). The histopathological features observed during infection (n = 1) of those treated with the Ly6G depleting antibody did not with low doses of Mtb were recapitulated and exacerbated for survive and all C57BL/6 mice survived infection (Fig. 2f). Therefore, those with high doses, with infected Gcnt1-/- mice presenting depleting neutrophils largely rescued the increased susceptibility extensive lung pathology, particularly extensive necrosis, and of Gcnt1−/− mice to Mtb infection. Finally, we compared the histological features of bronchopneumonia (Table 1, Fig. 1i, g and response of neutrophils isolated from mice of either genetic Fig. S1E). background to in vitro Mtb infection. Although no statistically significant differences were found, we noted an overall enhanced Neutrophils drive the increased susceptibility of Gcnt1−/− mice to cytokine expression by Gcnt1−/− neutrophils upon Mtb infection Mtb infection (Fig. S2E), but similar production of reactive oxygen and nitrite To investigate the mechanisms underlying the increased suscept- species (Fig. S2F). In all, these findings suggest that Gcnt1 ibility of Gcnt1−/− mice to Mtb, we compared the dynamics of the abrogation did not alter the basic functional response of immune response in C57BL/6 or Gcnt1−/− mice during infection neutrophils to Mtb. with low or high doses of bacteria. No significant differences were noted in the dynamics or levels of expression of genes encoding Gcnt1 activity modulates granulopoiesis and lung neutrophil key inflammatory mediators for the immune response against Mtb recruitment infection with low doses (Fig. 2a). Similarly, no differences were We next investigated the causes underlying Gcnt1−/− mice found between non-infected Gcnt1−/− or C57BL/6 mice (Fig. 2a). neutrophilia. Blood neutrophilia was reported for healthy Gcnt1−/− Upon infection with high doses of Mtb, we noted an overall higher mice.11 We here confirmed this finding and further showed its expression of cytokines in Gcnt1−/− animals, as compared to maintenance throughout infection (Fig. S3A,B). To test whether C57BL/6 ones (Fig. 2b). In all, the lung pathology observed in lack of Gcnt1 may favour granulopoiesis, we analysed the infected Gcnt1−/− mice does not seem to result from an frequency of Lin−Sca1+cKit+ cells (LSK), common myeloid uncontrolled cytokine storm. precursors (CMP), granulocytic-monocytic precursors (GMP) and The dynamics of recruitment of myeloid and lymphoid cells to mature cell populations in the bone marrow (BM) of C57BL/6 or the lung was identical in Gcnt1−/− and C57BL/6 mice infected with Gcnt1−/− mice (Fig. S3C,D). No differences were seen in the a low dose of bacteria, but 30 days after infection all tested frequencies of LSK and CMP progenitor cell populations, but an immune cell populations were significantly higher in the lung of increased frequency of GMPs was noted (Fig. 3a). In line with this Gcnt1−/− as compared to C57BL/6 mice (Fig. 2c and Fig. S2A, B). increase in GMP, the frequencies of neutrophils and monocytes Moreover, increased numbers of lung neutrophils were detected were higher in the BM of Gcnt1−/− mice (Fig. 3b). As for lymphoid in Gcnt1−/− mice, independently of the infection (Fig. 2c). Because populations, a decreased frequency of B cells and a similar lung infiltrating neutrophils have been associated with increased frequency of T cells were observed in Gcnt1−/− BM, as compared bacterial burdens, pathology and TB exacerbation,1–4 we assessed with C57BL/6 ones (Fig. S3E). Thus, the BM and blood neutrophilia the contribution of neutrophils to the increased susceptibility of observed in Gcnt1−/− mice may at least in part result from Gcnt1−/−. For that, we investigated the distribution of neutrophils enhanced BM granulopoiesis. in the lungs of C57BL/6 or Gcnt1−/− mice infected with low or high We next investigated the impact of Gcnt1 deficiency on the doses of bacteria, through detection of myeloperoxidase (MPO) mobilization of the neutrophils between BM and blood. As CXCR4 expression by immunostaining. Although neutrophils were pre- is a key regulator of neutrophil retention versus egress from the sent in both mouse strains upon infection, these cells accumulated BM,23,24 we compared the expression of this receptor in BM in the case of infected Gcnt1−/− mice (Fig. 2d), in line with the neutrophils of each mouse strain. The levels of CXCR4 expressed in cytometry data. An accumulation of neutrophils in the areas of BM neutrophils were comparable in C57BL/6 and Gcnt1−/− mice lesion was particularly observed in infections with high doses of (Fig. 3c). Furthermore, to investigate for differences in the egress Mtb, being more exacerbated in Gcnt1−/− mice (Fig. 2d). A similar of neutrophils from the BM in Gcnt1−/− mice, we intravenously accumulation was found in the necrotic lesions of moribund injected an antibody directed to CD45 and assessed the Gcnt1−/− mice infected with low doses of bacteria (Fig. S2C). percentage of CD45+ BM sinusoidal cells in either mouse strain. Moreover, and in line with the higher presence of neutrophils in No differences were found in the frequency of CD45 labelled the lungs of Gcnt1−/− mice infected with high doses of Mtb, the neutrophils, B cells and T cells when comparing the two mouse expression levels of genes previously associated with a neutrophil- strains (Fig. 3d). Together, these data suggest that Gcnt1 dominated transcriptional module21 were increased in these deficiency does not alter the hematopoietic cell retention in the animals, as compared to C57BL/6 ones (Fig. 2e). We next treated BM or egress from the BM to the blood. Mucosal Immunology (2020) 13:836 – 848 Deficiency in the glycosyltransferase Gcnt1 increases susceptibility to. . . KL Fonseca et al. 841 We then sought to understand the cause for increased lung similar permeability of the lung vasculature in Gcnt1− and C57BL/6 neutrophil frequencies in Gcnt1−/− mice. We started by testing the mice. Next, we questioned whether the lung neutrophilia of permeability of the lung vasculature by measuring the influx of Gcnt1−/− mice depended on altered neutrophil chemoattraction. Evans blue injected intravenously. Abrogation of Gcnt1 did not For this, we measured the expression levels of lung Cxcl1, Cxcl2 impact the diffusion of dye into the lung (Fig. 3e), indicating a and Cxcl5 during infection with low or high doses of Mtb (Fig. 3f). Mucosal Immunology (2020) 13:836 – 848 Deficiency in the glycosyltransferase Gcnt1 increases susceptibility to. . . KL Fonseca et al. 842 Fig. 3 Gcnt1 modulates granulopoiesis and CXCL2 expression. The BM of non-infected C57BL/6 (B6; black circles) or Gcnt1−/− (open circles) mice was harvested and the frequency of a LSK, CMP and GMP populations; b neutrophils and monocytes or (c) neutrophils expressing CXCR4 determined by flow cytometry. All gating strategies are shown in Fig. S3. d Egress of immune cell populations from the BM of in C57BL/B6 (B6; black circles) or Gcnt1−/− (open circles) measured as the percentage of CD45+ BM sinusoidal cells upon intravenous injection of CD45-PE. e Permeability of the lung vasculature in C57BL/B6 (B6; black circles) or Gcnt1−/− (open circles) measured through quantification of the amount of Evans Blue extravasated into the tissue upon intravenous injection. f The expression of Cxcl1, Cxcl2 and Cxcl5 in the lungs of mice infected with low or high doses of Mtb was determined by real-time PCR, for the indicated timepoints post-infection. g Analysis of neutrophils migration through 5 μm transwells towards different concentrations of recombinant CXCL1, CXCL2 and CXCL5. Mean ± SEM of culture triplicates are presented. In a, f each dot represents a mouse and the Mean±SEM for 6–15 animals in at least two independent experiments are plotted. Statistical analysis was performed with Student’s t-test (a–f) or unpaired two-tailed Mann-Whitney test for each time point (g). * refer to statistic differences between B6 or Gcnt1−/− mice. *p < 0.05; **p < 0.01; ***p < 0.01; ****p < 0.0001. Although the transcription of these chemokines in lungs of C57BL/ Lewis antigens and, among these, sLex is described to decorate 6 or Gcnt1−/− mice was similar upon a low dose of infection, the the surface of human neutrophils (Fig. 5a). Because a neutrophil- expression of Cxcl2 was higher in Gcnt1−/− deficient mice than in driven signature was observed in the whole blood of TB C57BL/6 upon a high dose of infection (Fig. 3f). Finally, we tested patients,25 and since our data link Gcnt1 activity with TB the ability of neutrophils from either genetic background to susceptibility through a mechanism involving neutrophils, we migrate in response to each of these chemokines, using an in vitro next interrogated available human whole blood TB RNA-Seq trans-well system. Independently of the expression of Gcnt1, the datasets26 for transcriptional alterations in glycosyltransferase- migration of neutrophils towards CXCL1, CXCL2 or CXCL5 was encoding genes (Table S1). Several glycosyltransferase-encoding dose dependent and similar (Fig. 3g). Collectively, these findings genes were significantly up- or downregulated in active TB support a higher availability of neutrophils in the absence of patients as compared to control or latently infected individuals Gcnt1, accompanied by an increased recruitment of neutrophils to (Fig. 5b). In particular, significant alterations in genes encoding the lung of Gcnt1−/−-infected mice, possibly mediated by the enzymes of the sLex biosynthetic pathway were detected (Fig. 5b). higher expression of CXCL2. Importantly, a much lower level of variation was observed when latent and controls were compared (Fig. 5b), thus showing that TB Increased susceptibility to Mtb infection requires Gcnt1 deficiency disease impacts the glycosyltransferase transcriptomic signature. in both hematopoietic and non-hematopoietic compartments Lastly, we examined lungs of TB patients for the presence of sLex To investigate if the susceptibility of Gcnt1−/− mice to Mtb infection antigens. TB patients (Table 2) were enrolled in a study in Georgia was fully dependent on the hematopoietic compartment, we and underwent lung surgery as a therapeutic option, after resorted to a mouse BM transplantation model (Fig. 4a and Fig. S4A). antibiotherapy failure.27 Independently of the patient’s gender When comparing all experimental groups, we observed that both and Mtb drug susceptibility profile, lung sLex expression was the Gcnt1−/−→C57BL/6 and the Gcnt1−/−→Gcnt1−/− chimeric detected in epithelial and immune cells (Fig. 5c and Fig. S5). animals presented significantly higher neutrophils in circulation Altogether, we provide, for the first time, evidence for the upon reconstitution and before infection (Fig. 4b), suggesting modulation of core 2-sLex antigens during human TB, through that hematopoietic deficiency of Gcnt1 is sufficient to promote the transcriptional regulation of various glycosyltransferase- neutrophilia. Chimeric mice were then infected with a low dose of encoding genes. Mtb. Mice in the Gcnt1−/−→Gcnt1−/− group presented a different survival curve compared to all other groups, where one animal (representing 14.3%) succumbed to infection after 24 days DISCUSSION (Fig. 4c), and also higher lung bacterial burdens 30 days post- Despite the growing significance of glycosylation in immunity,5,28 infection than any other group (Fig. 4d), paralleling the its impact in TB is poorly understood. To address this hitherto increased susceptibility to infection of Gcnt1−/− mice. Histo- unexplored question, we studied the outcome of Mtb infection in pathological analysis of the infected chimeric animals showed mice lacking Gcnt1, a key glycosyltransferase controlling core 2 O- that although Gcnt1−/−→C57BL/6 mice presented edema and glycans biosynthesis. Gcnt1 deficiency impacted bacterial bur- larger necrotic areas when compared to the C57BL/6→C57BL/6 dens, lung pathology and mouse survival. It also associated with group, Gcnt1−/−→Gcnt1−/− animals presented the most exuber- exacerbated BM, blood and lung neutrophilia, already present in ant pathology with severe inflammation, edema, vast necrosis naïve animals. In keeping with the immune pathological role of and bronchopneumonia (Fig. 4e and Fig. S4B). Lung flow neutrophils in TB,1–4 the accumulation of these cells was strongly cytometry analysis of the infected chimeric mice revealed associated with lung necrotic lesions and it was markedly an overall higher frequency of neutrophils in the lungs of increased in Gcnt1−/− infected mice. Further implicating neutro- Gcnt1−/−→Gcnt1−/− chimeric mice (Fig. 4f) and similar frequen- philia with susceptibility of Gcnt1−/− mice to Mtb infection, cies of all other cell populations (Fig. S4C). Interestingly, in line depleting neutrophils from Gcnt1−/− mice largely rescued their with what we had observed in Gcnt1−/− infected mice, the susceptible phenotype. Interestingly, Fut4−/− mice present normal expression of Cxcl2 upon infection was exclusively increased in neutrophil counts29 and no phenotype upon Mtb infection,19,20 Gcnt1−/−→Gcnt1−/− chimeric mice (Fig. 4g). whereas Fut7−/− mice exhibit blood neutrophilia30 and increased Thus, although Gcnt1 deficiency in the hematopoietic compart- susceptibility to Mtb infection.19,20 Moreover, our preliminary data ment is sufficient to deregulate blood neutrophil frequencies in show that St3gal4−/− mice present normal neutrophil counts in homeostasis, for that to translate into different susceptibilities to homeostasis and respond to Mtb infection similar to C57BL/ Mtb infection, deficiency of Gcnt1 in the stromal compartment is 6 mice. also required. Blood neutrophilia has been described in Gcnt1−/− mice,11 but the underlying mechanisms remain elusive. Here, we show that Mtb infection modulates the host glycosylation machinery at the hematopoietic abrogation of Gcnt1 was sufficient to cause blood gene transcription level neutrophilia, which may result from enhanced BM granulopoiesis. Our findings highlight the importance of a competent core 2 We have excluded other possible regulatory mechanisms, includ- O-glycan pathway in regulating neutrophils and immune pathol- ing a direct modulation of CXCR4 receptor levels in Gcnt1−/− ogy during TB. Core-2 O-glycans are major carriers of terminal neutrophils or a differential egress of immune cells from the BM to Mucosal Immunology (2020) 13:836 – 848 Deficiency in the glycosyltransferase Gcnt1 increases susceptibility to. . . KL Fonseca et al. 843 the blood. We also report an increase in lung neutrophil counts in atypical selectin-independent cascade,9,31 as shown during Strepto- Gcnt1−/− mice, which may seem counterintuitive, as absence of coccus pneumonia infection.32 Also, Gcnt1 activity has been Gcnt1 associates with a partial deficiency of selectin ligands.11 demonstrated to be dispensable for lung neutrophil recruitment However, the recruitment of neutrophils to the lung may follow an upon LPS stimulation.33 We show that lung neutrophilia in Gcnt1−/− Mucosal Immunology (2020) 13:836 – 848 Deficiency in the glycosyltransferase Gcnt1 increases susceptibility to. . . KL Fonseca et al. 844 Fig. 4 Blood neutrophilia of Gcnt1−/− mice is promoted by deficiency of this enzyme in hematopoietic cells, but increased susceptibility to Mtb infection also requires the non-hematopoietic compartment. a Schematic representation of the BM transplantation model used and the experimental groups included. b The frequency of neutrophils in the blood of non-infected chimeric mice was determined by flow cytometry, following the gating strategy shown in Fig. S3. c Mice in the different chimeric groups were infected with a low dose of Mtb strain HN878 and the weight of the animals monitored to determine survival curves. d On day 30 post-infection, the bacterial burden in the lungs of the infected mice was determined by CFU enumeration. e H&E staining of one representative animal of each experimental group is represented. Black arrows point to edema, black arrowhead spot bronchopneumonia, and dashed black line limits the necrotic areas. Scale bar corresponds to 100 µm. f The number of neutrophils present in infected lungs was determined on day 30 post-infection, by flow cytometry. g The expression of Cxcl1, Cxcl2 and Cxcl5 in the lungs of infected mice was determined by real-time PCR, as above. b, d, f, g Represented is Mean±SEM, and each symbol represents one mouse. Statistical analysis was performed with a one-way ANOVA using Tukey’s test for multiple comparisons (b, d, f, g) or with log-rank (Mantel-Cox) test for the Kaplan Meyer curve (c). * refer to statistic differences between the indicated chimeric groups. **p < 0.01; ***p < 0.001; ****p < 0.0001. mice involved both hematopoietic and stromal compartments. Hospital Universitari Germans Trias i Pujol (#B9900005) and the Whereas the contribution of the hematopoietic deficiency may be Dept d’Agricultura, Ramaderia, Pesca, Alimentació i Medi Natural related to the increased number of available circulating neutrophils, of the Catalan Government. that of the stroma does not relate with alterations in the permeability of the lung vasculature and may result from an Animals upregulated Cxcl2 expression. Interestingly, it does not rely on C57BL/6 wild-type or Gcnt1−/− mice were bred and housed at i3S differential expression of Cxcl5, a chemokine produced by epithelial and infected under ABSL3 conditions. The Gcnt1−/− mouse was cells that drives neutrophil recruitment, contributing to more severe obtained from the Consortium for Functional Glycomics.11 Mice forms of TB.34 In the same context, Gcnt1−/− neutrophils showed were euthanized with isoflurane or CO2 inhalation with efforts to migratory profiles towards CXCL1, CXCL2 and CXCL5 similar to those minimize suffering. Food and water were ad libitum. of C57BL/6 neutrophils. Neutrophils are widely described as detrimental in TB,1–4 and Bacteria growth and quantification an interferon-inducible neutrophil-driven blood transcriptional Mtb strain HN878 was grown and stored as before.39 Viable signature was defined in human TB.25 Interestingly, human bacteria was determined by serial dilution and colony forming neutrophils are decorated with sialylated antigens, most notably unit (CFU) enumeration after 21–28 days of incubation at 37 °C in sLex, which biosynthetic pathway involves Gcnt1 enzymatic 7H11 agar plates. Bacterial quantification in infected lungs activity.7 We found alterations in the expression of several followed a similar protocol. glycosyltransferases in whole blood of active TB patients as compared to controls. Furthermore, we detected sLex immu- Experimental infection and animal monitoring nostaining in lung sections of TB patients. Of note, previous C57BL/6 and Gcnt1−/− (8–12 week old) mice were aerosol-infected reports showed that the urinary levels of sialic acid, the terminal with Mtb strain HN878 using an inhalation exposure system (Glas- glycan of sLex structure, discriminate patients with active TB Col), as published.39,40 Low and high dose infections delivered from healthy controls or patients with non-tuberculous pul- <200 or >350 bacteria to the lung, respectively, as determined monary diseases.35 Also, the regulation of the gene encoding 3 days post-infection. Infected mice were weighted every week or CMAS, the enzyme that provides the substrate for the addition every two days, and humanely euthanized if 20% of their weight of sialic acid, specifically changed with infection with Mtb.36 or responsiveness to physical stimulation were lost. Whenever Thus, modulation of sialylated antigens seems to occur in TB, as possible, the lungs of moribund animals were harvested for described in several pathologies, such as other infectious histology assessment. diseases and cancer.6,16,17,37 Collectively, this study further strengthens the role of glycosyla- Organ processing tion during immune responses and reveals that regulation of Lungs were aseptically excised and processed as before.39,40 neutrophil homeostasis by host glycosyltranferases impacts TB Single-cell suspensions were used for bacterial burden determina- outcome, a disease connected with high mortality and morbidity. tion, flow cytometry and RNA analysis. Our findings raise awareness for the possible detrimental impact of targeting glycan dependent interactions, as envisaged for example Tissue samples, histology and morphometric analysis in anticancer therapies,38 in the context of TB. Additionally, a better The human lung samples (Table 2) are from the collection understanding of glycans and their functional roles during chronic obtained within the SH-TBL project, led by the Experimental infection, provides increasing opportunities for discovery and Tuberculosis Unit (UTE) and conducted in collaboration with the therapeutic intervention in settings like TB. NCTLD; and registered at the ClinicalTrials.gov database under code NCT02715271. Mouse lungs were fixed in 10% buffered-formalin and paraffin- MATERIAL AND METHODS embedded. Serial sections of 3 µm-thickness were haematoxylin- Ethics statement eosin (H&E) or Ziehl-Neelsen stained or used for immunofluores- The lung samples study was reviewed and approved by the cence (MPO) analysis.39 Morphometric quantification of lung Ethics Committees of the National Center for Tuberculosis and histology images is detailed in Fig. S1. Pathological scoring analysis Lung Diseases of Georgia (IRB00007705-NCTLD-Georgia#1, of H&E-stained histological sections was performed by a veterinary IORG0006411) and of the Germans Trias I Pujol Hospital (EC:PI- pathologist blinded to the treatment groups. The histopathological 16–171). The blood samples study was reviewed and approved features were scored using the following scale: 0 = absent, 1 = focal by the Portuguese Comissão de Ética para a Saúde da ARS Norte 2 = multiple, 3 = severe, 4 = severe and extensive. The % of animals (#T792). All participants signed informed consents, and all data with a certain score within the indicated experimental group was were anonymized. calculated for group comparison. Animal experiments followed the 2010/63/EU Directive, and were approved by the i3S Animal Ethics Committee and the MPO immunostaining Portuguese National Authority for Animal Health (#014811/2016- Lung sections were stained with a goat anti-mouse antibody 07-13) or by the Animal Experimentation Ethics Committee of the against MPO (AF3667, R&D systems; 1:40) overnight at 4 °C Mucosal Immunology (2020) 13:836 – 848 Deficiency in the glycosyltransferase Gcnt1 increases susceptibility to. . . KL Fonseca et al. 845 followed by a secondary donkey anti-goat IgG antibody con- RNA extraction, cDNA synthesis and real-time PCR analysis jugated with Alexa Fluor 568 (A11057, Invitrogen; 1:1000), and Total RNA was extracted from mouse lungs or in-vitro-infected DAPI (Biolegend; 1:1000) for 2 h in the dark at RT. Glass coverslips neutrophils, using TRI-Reagent (Sigma), converted to cDNA (NE were mounted onto the slides using Vectashield (Sigma). Slides Biolabs), and subjected to real-time PCR using SYBR green were acquired using the IN Cell Analyzer 2000 (GE Healthcare) and (Applied Biosystems) and specific oligonucleotides (Table S2). analysed using IN Cell Developer software and ImageJ.41,42 Melting and standard curves and RQ values were determined for Mucosal Immunology (2020) 13:836 – 848 Deficiency in the glycosyltransferase Gcnt1 increases susceptibility to. . . KL Fonseca et al. 846 Fig. 5 Mtb infection impacts the expression of several glycosyltransferase-encoding genes in humans. a Neutrophils are the main cells expressing sLex in human peripheral blood. Vein blood was collected and stained for CD3, CD14, CD16, CD19 and sLex by flow cytometry. Data shown represent one donor out of 25 analysed. b Volcano plots displaying the comparisons between Active TB vs Control, Active TB vs Latent TB and Latent TB vs Control regarding the differential expression of annotated coding genes of glycosyltransferases (obtained from Glyco-Enzyme Repository) on the Berry London, Berry Leicester (progressor) and Berry South Africa datasets.26 Labelled genes encode for glycosyltransferases associated with the sLex pathway. Dot color indicates if the corresponding gene is upregulated (red, log-fold-change >1 and p-value < 0.05), downregulated (blue, log-fold-change < -1 and p-value < 0.05), significant (black, log-fold-change >−1 to <1 and p-value < 0.05) and not significant (grey). c Lung sections of TB patients who underwent therapeutic surgery were stained for sLex as above. Black and white arrowheads point to positive epithelial or immune cells, respectively. Scale bars correspond to 100 µm. Table 2. Overall characteristics of the participants originating lung Evans Blue (1 mg per 30 g mouse, prepared in PBS). Thirty minutes samples. after injection mice were sacrificed and lungs collected. After weight determination, samples were incubated in formamide for Patient Gender Lesion Lesion TB Drug ~36 h at 55 °C. The formamide solution was recovered localizationa size (mm) symptoms sensitivity and absorbance was measured at 620 nm and 740 nm to profileb determine the amount of Evans Blue present in the tissue. The presence of heme groups was corrected with the 1 Male RUL 21, 30 Yes DS formula: A620(corrected) = A620 − (1.426 × A740 + 0.030). 2 Male RUL 35 Yes DS 3 Female LLL 30 No MDR In vitro infection of neutrophils and chemoattraction assays 4 Female LLL 30 No DS For BM neutrophils isolation, BM cells were stained with anti- 5 Female RUL 30 No DS mouse Ly6G (1A8; Biolegend) followed by purification using anti- biotin magnetic beads and LS columns of MACS cell separation 6 Female RLL 42 No MDR system (both Miltenyi Biotec). For in vitro infection, 105 7 Male RUL 30, 21 No MDR neutrophils were seeded per well and infected with Mtb HN878 8 Female LUL 40 No DS at MOI 2 or 10. After 3 h of infection, cells were recovered for RNA 9 Male LUL 38 No MDR analysis or reactive oxygen species detection with DHE and DHR 10 Male LUL 29 No MDR probes. Cells were incubated with DHE or DHR at 37 °C for 10 or 30 min, respectively. Samples were then acquired on a BD FACS All patients were antibiotic-treated before surgery. Highlighted in grey are Canto II to detect red-fluorescence (DHE) or green fluorescence the patients represented in Fig. 5. (DHR) due to probes oxidization. For the chemotaxis assay, 105 RUL right upper lobe, LLL left lower lobe, RLL right lower lobe, LUL left neutrophils were seeded on the upper chamber of the 5 μm-pore upper lobe, DS drug sensitive, MDR multi-drug resistant. transwells (96 well plate; Corning) and allowed to migrate for 3 h towards the lower chamber containing media with CXCL1, CXCL2 and CXCL5 (all from Peprotech) at 500, 50, 5 and 0.5 ng/ml or each gene. Ubiquitin was used as a reference gene for normal- without cytokines. As control for the amount of cells seeded, cells ization of target gene abundance. were also seeded in the lower chamber. Cells were then recovered and stained with SAV PE-Cy7 (Biolegend). To quantify the number Flow cytometry of migrated cells, counting beads (Biolegend) were added to the Human or mouse blood, mouse BM and lung were stained for cells prior to acquisition in a BD FACS Canto II. All cytometry data surface antigens (30 min; 4 °C) and fixed for 20 min in 2% were analysed using FlowJo version 10.1.r7. paraformaldehyde-PBS after erythrocyte lysis. For the analysis of BM precursors, the mature lineage (Lin) was depleted (Miltenyi Generation of BM chimeric mice Biotec magnetic cell separation system), prior to Linneg staining with Recipient mice were irradiated with doses of 750 rad and specific antibodies for progenitor markers. Dead cells were always reconstituted with 5 × 106 CD3-TCRβ-TCRγ/δ-depleted BM cells. excluded using a viability dye. Cells were acquired on a BD FACS Bactrim (80 mg sulfametoxadol and 16 mg trimetoprim/250 mL of Canto II. Data were analyzed using FlowJo version 10.1.r7. Gating water, for an estimated dose of 50 mg antibiotic/Kg/day) was strategies are in Fig. S2 and S3 and the antibodies used in Table S3. administered in drinking water for the first 3 weeks post- reconstitution. A control group, in which BM cells from CD45.1- Neutrophil depletion C57BL/6 mice were transferred into irradiated CD45.2-Gcnt1−/− In vivo depletion of neutrophils from infected Gcnt1-/- mice was mice, was included to monitor the efficiency of chimerism, as described:22 0.2 mg anti-Ly6G mAb (clone 1A8; BioXCell) or 6 weeks post-transplantation (Supplemental Fig. 3A). Chimeras isotype control (clone GL117; R&D Systems) were administered were infected at this stage. i.p. from day 18 post-infection, every other day for 20 days. To control for the protocol, neutrophils in the lungs of mice RNA-seq data analysis treated with anti-Ly6G mAb or isotype control were detected at Raw paired-end RNA-seq data from whole-blood cohorts Berry the end of the experiment by MPO staining (Fig. S2D). London, Berry South Africa and Berry Leicester progressor26 (GSE107995) were processed separately. The reads were quality Analysis of BM egress controlled through FastQC [v0.11.7] (Babraham Bioinformatics) Mice were intravenously injected with 0.6 μg of anti-mouse CD45 and MultiQC.46 Filtering below 20 quality score and below 50 conjugated with Phycoerythrin (PE) (30-F11; Biolegend) in PBS, 2 base-pairs, and adapters exclusion was performed using Trimmo- min prior to euthanasia, to label BM sinusoidal cells.43 Blood and matic [v0.38].47 Filtered reads were aligned to the human BM cells were collected and analysed by flow cytometry. reference genome GRCh38 using HISAT2 [v2.1.0].48 StringTie [v1.3.4] was used for the assembly and raw counts matrix Analysis of lung vasculature permeability generated by prepDE.py.49 Differentially expressed genes were Vasculature permeability was assessed using Evans Blue dye as analysed in R [v3.5.1] with Bioconductor packages edgeR50,51 and described previously.44,45 Mice were intravenously injected with limma.52 Genes with less than 15 raw counts in all samples were Mucosal Immunology (2020) 13:836 – 848 Deficiency in the glycosyltransferase Gcnt1 increases susceptibility to. . . KL Fonseca et al. 847 excluded. 202 genes encoding glycosyltransferases was extracted 3. Marzo, E. et al. Damaging role of neutrophilic infiltration in a mouse model of from the Glyco-Enzyme Repository (http://glycoenzymes.ccrc.uga. progressive tuberculosis. Tuberculosis (Edinb.) 94, 55–64 (2014). edu/). Differentially expressed genes were determined through 4. Lowe, D. M. et al. Neutrophilia independently predicts death in tuberculosis. Eur. linear model fitting. Empirical Bayes moderated t-statistics test Respir. J. 42, 1752–1757 (2013). 5. van Kooyk, Y. & Rabinovich, G. A. Protein-glycan interactions in the control of was performed and genes with p-value < 0.05 and log2 fold- innate and adaptive immune responses. Nat. Immunol. 9, 593–601 (2008). change < −1 or >1 were considered significant. 6. Pinho, S. S. & Reis, C. A. Glycosylation in cancer: mechanisms and clinical impli- cations. Nat. Rev. Cancer 15, 540–555 (2015). Immunohistochemistry and image analysis 7. Sperandio, M., Gleissner, C. A. & Ley, K. Glycosylation in immune cell trafficking. Human lung sections were immunolabeled for sLex as described,53 Immunol. Rev. 230, 97–113 (2009). with an antibody directed against CD15s (BD Pharmingen; 8. Maas, S. L., Soehnlein, O. & Viola, J. R. Organ-specific mechanisms of transen- CSLEX1) and ultraView DAB (Ventana Medical Systems).54 dothelial neutrophil migration in the lung, liver, kidney, and aorta. Front. Immunol. 9, 2739 (2018). Statistical analysis 9. Rossaint, J. & Zarbock, A. Tissue-specific neutrophil recruitment into the lung, liver, and kidney. J. Innate Immun. 5, 348–357 (2013). Data were analysed using GraphPad Prism (version 8.1.0). 10. Bartunkova, J. et al. Reduced phagocytic activity of polymorphonuclear leuko- Student’s t-test was used to compare two groups and one-way cytes in alpha(1,3) fucosyltransferase VII-deficient mice. APMIS 108, 409–416 ANOVA for more than two groups, with post-tests as in Figure (2000). legends. Data were checked for normality and log normality. 11. Ellies, L. G. et al. Core 2 oligosaccharide biosynthesis distinguishes between Survival curves were analysed using Log-rank (Mantel-Cox) test. selectin ligands essential for leukocyte homing and inflammation. Immunity 9, Significant differences are as follows: *p ≤ 0.05; **p ≤ 0.01; ***p ≤ 881–890 (1998). 0.001 and ****p ≤ 0.0001. 12. Duarte, H. O. et al. Mucin-type O-glycosylation in gastric carcinogenesis. Biomo- lecules. https://doi.org/10.3390/biom6030033 (2016). 13. Symmes, B. A., Stefanski, A. L., Magin, C. M. & Evans, C. M. Role of mucins in lung homeostasis: regulated expression and biosynthesis in health and disease. Bio- ACKNOWLEDGEMENTS chem. Soc. Trans. 46, 707–719 (2018). The authors thank the excellent support from the i3S scientific platforms, namely 14. Tan, F. Y., Tang, C. M. & Exley, R. M. Sugar coating: bacterial protein glycosylation Animal facility, Translational Cytometry, Advanced Light Microscopy and BioSciences and host-microbe interactions. Trends Biochem. Sci. 40, 342–350 (2015). Screening, member of the national infrastructure PPBI—Portuguese Platform of 15. Linden, S. et al. Role of ABO secretor status in mucosal innate immunity and H. Bioimaging (PPBI-POCI-01-0145-FEDER-022122). This work was financed by grants pylori infection. PLoS Pathog. 4, e2 (2008). POCI-01-0145-FEDER-028955 (to M.S.), POCI-01-0145-FEDER-028489 (to A.M.), and 16. Magalhaes, A. et al. Helicobacter pylori chronic infection and mucosal inflam- POCI-01-0145-FEDER-016585 (to C.A.R.) and by FEDER - Fundo Europeu de mation switches the human gastric glycosylation pathways. Biochim. Biophys. Desenvolvimento Regional funds through the COMPETE 2020 - Operacional Acta 1852, 1928–1939 (2015). Programme for Competitiveness and Internationalization (POCI), Portugal 2020, and 17. Marcos, N. T. et al. Helicobacter pylori induces beta3GnT5 in human gastric cell by Portuguese funds through FCT - Fundação para a Ciência e a Tecnologia/ lines, modulating expression of the SabA ligand sialyl-Lewis x. J. Clin. Invest. 118, Ministério da Ciência, Tecnologia e Inovação in the framework of the project 2325–2336 (2008). “Institute for Research and Innovation in Health Sciences” (POCI-01-0145-FEDER- 18. Navabi, N., Johansson, M. E., Raghavan, S. & Linden, S. K. Helicobacter pylori 000012). C.A.R. acknowledges the Consortium for Functional Glycomics funded by infection impairs the mucin production rate and turnover in the murine gastric NIGMS-GM62116. This study was funded by the Spanish Government-FEDER Funds mucosa. Infect. Immun. 81, 829–837 (2013). through CP13/00174, CPII18/00031 and PI16/01511 grants, and the CIBER Enferme- 19. Schreiber, T. et al. Selectin ligand-independent priming and maintenance of T cell dades Respiratorias Network; and by the Spanish Society of Pneumology and immunity during airborne tuberculosis. J. Immunol. 176, 1131–1140 (2006). Thoracic Surgery (SEPAR) through grant 16/023 (to C.V.). K.L.F. and R.M. are funded by 20. Ehlers, S., Schreiber, T., Dunzendorfer, A., Lowe, J. B. & Holscher, C. FCT PhD scholarships SFRH/BD/114405/2016 and SFRH/BD/131159/2017, respec- Fucosyltransferase IV and VII-directed selectin ligand function determines tively; A.M. is funded by FCT; MS is funded by FCT through Estimulo Individual ao long-term survival in experimental tuberculosis. Immunobiology 214, 674–682 Emprego Científico. The authors thank Anne O’Garra for insightful discussions. (2009). 21. Singhania, A. et al. Transcriptional profiling unveils type I and II interferon net- works in blood and tissues across diseases. Nat. Commun. 10, 2887 (2019). AUTHOR CONTRIBUTIONS 22. Nandi, B. & Behar, S. M. Regulation of neutrophils by interferon-gamma limits K.L.F., A.R.M., R.M.: performed experiments, analysed data and wrote the manuscript. A.R. lung inflammation during tuberculosis infection. J. Exp. Med. 208, 2251–2262 M., L.S.C.: analysed the RNA-Seq data. J.S., B.C., L.B., A.I.F., S.M., J.G., G.T., P.R.M., M.M.C.: (2011). performed experiments. R.R., S.V., S.G., K.N.: provided the human samples used in the 23. Eash, K. J., Means, J. M., White, D. W. & Link, D. C. CXCR4 is a key regulator of study. R.A., F.G.: analysed the mouse histology samples. C.V.: coordinated the human neutrophil release from the bone marrow under basal and stress granulopoiesis study. R.A., P.R., C.V., C.A.R.: critically discussed the experimental design and data and conditions. Blood 113, 4711–4719 (2009). wrote the manuscript. A.M., M.S.: coordinated the study, designed the experiments, 24. Martin, C. et al. Chemokines acting via CXCR2 and CXCR4 control the release of analysed data and wrote the manuscript. All authors read and approved the manuscript. neutrophils from the bone marrow and their return following senescence. Immunity 19, 583–593 (2003). 25. Berry, M. P. et al. An interferon-inducible neutrophil-driven blood transcriptional ADDITIONAL INFORMATION signature in human tuberculosis. Nature 466, 973–977 (2010). The online version of this article (https://doi.org/10.1038/s41385-020-0277-7) 26. Singhania, A. et al. A modular transcriptional signature identifies contains supplementary material, which is available to authorized users. phenotypic heterogeneity of human tuberculosis infection. Nat. Commun. 9, 2308 (2018). Competing interests: The authors declare no competing interests. 27. Vashakidze, S. et al. Retrospective study of clinical and lesion characteristics of patients undergoing surgical treatment for Pulmonary Tuberculosis in Georgia. Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims Int. J. Infect. Dis. 56, 200–207 (2017). in published maps and institutional affiliations. 28. Johnson, J. L., Jones, M. B., Ryan, S. O. & Cobb, B. A. The regulatory power of glycans and their binding partners in immunity. Trends Immunol. 34, 290–298 (2013). 29. Weninger, W. et al. Specialized contributions by alpha(1,3)-fucosyltransferase-IV REFERENCES and FucT-VII during leukocyte rolling in dermal microvessels. Immunity 12, 1. Dorhoi, A. & Kaufmann, S. H. Pathology and immune reactivity: understanding 665–676 (2000). multidimensionality in pulmonary tuberculosis. Semin. Immunopathol. 38, 30. Maly, P. et al. The alpha(1,3)fucosyltransferase Fuc-TVII controls leukocyte traf- 153–166 (2016). ficking through an essential role in L-, E-, and P-selectin ligand biosynthesis. Cell 2. Eruslanov, E. B. et al. Neutrophil responses to Mycobacterium tuberculosis 86, 643–653 (1996). infection in genetically susceptible and resistant mice. Infect. Immun. 73, 31. Kolaczkowska, E. & Kubes, P. Neutrophil recruitment and function in health and 1744–1753 (2005). inflammation. Nat. Rev. Immunol. 13, 159–175 (2013). Mucosal Immunology (2020) 13:836 – 848 Deficiency in the glycosyltransferase Gcnt1 increases susceptibility to. . . KL Fonseca et al. 848 32. Mizgerd, J. P. et al. Selectins and neutrophil traffic: margination and Strepto- 46. Ewels, P., Magnusson, M., Lundin, S. & Kaller, M. MultiQC: summarize analysis coccus pneumoniae-induced emigration in murine lungs. J. Exp. Med. 184, results for multiple tools and samples in a single report. Bioinformatics 32, 639–645 (1996). 3047–3048 (2016). 33. Broide, D. H. et al. Core 2 oligosaccharides mediate eosinophil and neutrophil 47. Bolger, A. M., Lohse, M. & Usadel, B. Trimmomatic: a flexible trimmer for Illumina peritoneal but not lung recruitment. Am. J. Physiol. Lung Cell Mol. Physiol. 282, sequence data. Bioinformatics 30, 2114–2120 (2014). L259–L266 (2002). 48. Kim, D., Langmead, B. & Salzberg, S. L. HISAT: a fast spliced aligner with low 34. Nouailles, G. et al. CXCL5-secreting pulmonary epithelial cells drive destructive memory requirements. Nat. Methods 12, 357–360 (2015). neutrophilic inflammation in tuberculosis. J. Clin. Invest. 124, 1268–1282 49. Pertea, M. et al. StringTie enables improved reconstruction of a transcriptome (2014). from RNA-seq reads. Nat. Biotechnol. 33, 290–295 (2015). 35. Isa, F. et al. Mass spectrometric identification of urinary biomarkers of pulmonary 50. McCarthy, D. J., Chen, Y. & Smyth, G. K. Differential expression analysis of mul- tuberculosis. EBioMedicine 31, 157–165 (2018). tifactor RNA-Seq experiments with respect to biological variation. Nucleic Acids 36. Blischak, J. D., Tailleux, L., Mitrano, A., Barreiro, L. B. & Gilad, Y. Mycobacterial Res. 40, 4288–4297 (2012). infection induces a specific human innate immune response. Sci. Rep. 5, 16882 51. Robinson, M. D., McCarthy, D. J. & Smyth, G. K. edgeR: a Bioconductor package for (2015). differential expression analysis of digital gene expression data. Bioinformatics 26, 37. Mahdavi, J. et al. Helicobacter pylori SabA adhesin in persistent infection and 139–140 (2010). chronic inflammation. Science 297, 573–578 (2002). 52. Ritchie, M. E. et al. limma powers differential expression analyses for RNA- 38. Cagnoni, A. J., Perez Saez, J. M., Rabinovich, G. A. & Marino, K. V. Turning-off sequencing and microarray studies. Nucleic Acids Res. 43, e47 (2015). signaling by siglecs, selectins, and galectins: chemical inhibition of glycan- 53. Magalhaes, A. et al. Fut2-null mice display an altered glycosylation profile and dependent interactions in cancer. Front. Oncol. 6, 109 (2016). impaired BabA-mediated Helicobacter pylori adhesion to gastric mucosa. Gly- 39. Bhatt, K. et al. A nonribosomal peptide synthase gene driving virulence in cobiology 19, 1525–1536 (2009). Mycobacterium tuberculosis. mSphere. https://doi.org/10.1128/mSphere.00352-18 54. Kroesen, V. M. et al. A beneficial effect of low-dose aspirin in a murine model of (2018). active tuberculosis. Front. Immunol. 9, 798 (2018). 40. Moreira-Teixeira, L. et al. Type I IFN inhibits alternative macrophage activation during Mycobacterium tuberculosis infection and leads to enhanced protection in the absence of IFN-gamma signaling. J. Immunol. 197, 4714–4726 Open Access This article is licensed under a Creative Commons (2016). Attribution 4.0 International License, which permits use, sharing, 41. Schneider, C. A., Rasband, W. S. & Eliceiri, K. W. NIH Image to ImageJ: 25 years of adaptation, distribution and reproduction in any medium or format, as long as you give image analysis. Nat. Methods 9, 671–675 (2012). appropriate credit to the original author(s) and the source, provide a link to the Creative 42. Schindelin, J. et al. Fiji: an open-source platform for biological-image analysis. Nat. Commons license, and indicate if changes were made. The images or other third party Methods 9, 676–682 (2012). material in this article are included in the article’s Creative Commons license, unless 43. Beck, T. C., Gomes, A. C., Cyster, J. G. & Pereira, J. P. CXCR4 and a cell-extrinsic indicated otherwise in a credit line to the material. If material is not included in the article’s mechanism control immature B lymphocyte egress from bone marrow. J. Exp. Creative Commons license and your intended use is not permitted by statutory regulation or Med. 211, 2567–2581 (2014). exceeds the permitted use, you will need to obtain permission directly from the copyright 44. Reutershan, J. et al. Critical role of endothelial CXCR2 in LPS-induced neutrophil holder. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/. migration into the lung. J. Clin. Invest. 116, 695–702 (2006). 45. Radu, M. & Chernoff, J. An in vivo assay to test blood vessel permeability. J. Vis. Exp. https://doi.org/10.3791/50062 (2013). © The Author(s) 2020 Mucosal Immunology (2020) 13:836 – 848