Ly6G deficiency alters the dynamics of neutrophil recruitment and pathogen capture during Leishmania major skin infection

Neutrophils represent one of the first immune cell types recruited to sites of infection, where they can control pathogens by phagocytosis and cytotoxic mechanisms. Intracellular pathogens such as Leishmania major can hijack neutrophils to establish an efficient infection. However the dynamic interactions of neutrophils with the pathogen and other cells at the site of the infection are incompletely understood. Here, we have investigated the role of Ly6G, a homolog of the human CD177 protein, which has been shown to interact with cell adhesion molecules, and serves as a bona fide marker for neutrophils in mice. We show that Ly6G deficiency decreases the initial infection rate of neutrophils recruited to the site of infection. Although the uptake of L. major by subsequently recruited monocytes was tightly linked with the concomitant uptake of neutrophil material, this process was not altered by Ly6G deficiency of the neutrophils. Instead, we observed by intravital 2-photon microscopy that Ly6G-deficient neutrophils entered the site of infection with delayed initial recruitment kinetics. Thus, we conclude that by promoting neutrophils’ ability to efficiently enter the site of infection, Ly6G contributes to the early engagement of intracellular pathogens by the immune system.

. Equal neutrophil recruitment efficiency under Ly6G-deficiency, but decreased uptake of L. major parasites during early stage of infection. (a) Infection of Ly6G deficient Ly6G cre/cre × Rosa-tdTomato mice and Ly6G proficient Ly6G cre/+ × Rosa-tdTomato control mice with EGFP expressing L. major. Gating strategy to identify neutrophils lacking the Ly6G marker based on the low MHC class II expression and intermediate expression of Ly6C (CD11b hi MHCII lo Ly6C int ) in Ly6G cre/+ × Rosa-tdTomato control and Ly6G-deficient Ly6G cre/ cre × Rosa-tdTomato mice (left and middle panel). Histogram illustrates the quantification of the tdTomato-and Ly6G-signal on gated neutrophils (CD11b hi Ly6C int ) in Ly6G cre/+ × Rosa-tdTomato and Ly6G-deficient Ly6G cre/ cre × Rosa-tdTomato mice (right panel). (b) Proportion of neutrophils among CD45 + cells 2, 7 and 28 days p.i. with 10 6 metacyclic L. major-EGFP parasites in the infected tissue analyzed by flow cytometry. Each dot represents one infected ear. Horizontal bars represent the mean; ns, not significant as determined by two-way ANOVA (time, genotype) with Bonferroni post test for the genotype. (c) Recruitment of monocyte-derived phagocytes to the site of infection. Gating strategy for monocyte-derived dendritic cells (moDC, CD45 + CD11b + CD11c + MHCII + ) and monocyte-derived macrophages (mo-macro, CD45 + CD11b + CD11c -MHCII + ) in heterozygeous Ly6G cre/+ and homozygous Ly6G cre/cre mice. (d,e) Percentage of infected (EGFP containing) mo-macro (d) and mo-DCs (e) among CD45 + cells after quantitative evaluation of flow cytometry. Significant decrease of mo-DCs at late infection phase 28 days p.i. Each dot represents one infected ear. Horizontal bars represent the median; ***p < 0.001; ns not significant as determined by two-way ANOVA (time, genotype) with pairwise Bonferroni post test. (f) Unchanged parasite burden in L. major infected ear determined by limiting dilution 7 and 28 days p.i. Horizontal bars represent mean; ns not significant as determined by two-way ANOVA (time, genotype) with Bonferroni post test for the genotype. (g) Histograms depicting the infection rate of infected neutrophils (left panel), mo-macro (middle panel) and mo-DCs (right panel) after infection with EGFP-expressing L. major 2, 7 and 28 days p.i. in Ly6G cre/+ and Ly6G cre/cre mice. (h-j) Quantitative evaluation of histograms shown in (g). Proportion of infected neutrophils (h), mo-macro (i) and mo-DCs (j) analyzed via EGFP fluorescence of the parasite within the cells at day 2, 7 and 28 p.i. in Ly6G cre/+ and Ly6G cre/cre mice. Analysis revealed that Ly6G-deficient neutrophils phagocytose less parasites at early infection phase. Each dot represents one infected ear. Horizontal bars represent the mean; **p < 0.01; ns not significant as determined by two-way ANOVA (time, genotype) with Bonferroni post test for the genotype. Data are pooled from at least three independent experiments. (k) Mean Ly6C and CD11b fluorescence in tdTomato-expressing neutrophils from infected Ly6G cre/+ × Rosa-tdTomato and Ly6G cre/cre × Rosa-tdTomato mice. The expression of Ly6C surface protein is increased in Ly6G-deficient animals. Each dot represents one infected ear. Horizontal bars represent mean; *p < 0.05; ns not significant as determined by one-way ANOVA with pairwise Bonferroni post test. Data are pooled from at least three independent experiments.  www.nature.com/scientificreports/ sequestration, inhibit the efficient development of an adaptive immune response 58 , but also serve the parasite as a intermediate host cell enabling the infection of further phagocytes by efferocytosis-like mechanisms 59 . Whether Ly6G influences the interactions of neutrophils with Leishmania in the course of an infection has remained unknown.
Here, we show that Ly6G deficiency decreases the initial interaction efficiency and pathogen uptake by neutrophils. While later in the infection, we observed slightly decreased numbers of monocyte-derived dendritic cells in Ly6G-deficient animals, the uptake of infected neutrophils by monocytes, pathogen containment and neutrophil speed at the site of infection remained largely unchanged. Intravital 2-photon microscopy revealed that Ly6G-deficient neutrophils entered the site of infection with delayed initial kinetics, underlining the role of this surface protein in mediating interactions with cell adhesion molecules.

Results
Ly6G-deficient neutrophils take up less L. major during early stages of the infection. Despite the importance of neutrophils for many aspects of L. major infection, it is not known how deficiency in the neutrophil-specific Ly6G protein impacts the course of the immune response against this parasite. We therefore set out to infect Ly6G-deficient Ly6G cre/cre × Rosa-tdTomato mice with EGFP-expressing L. major. Ly6G cre/+ × Rosa-tdTomato littermate control mice served as Ly6G-proficient controls. Rosa-tdTomato mice have been shown to exhibit a bright red fluorescence specifically in neutrophils 48 . In order to identify neutrophils lacking the Ly6G marker in the infected ear tissue, we set up a gating strategy based on the low MHC class II expression and intermediate expression of Ly6C for identifying neutrophils independently of Ly6G. Expression of Ly6G-promoter-dependent cre driven from the knock-in alleles in both Ly6G-deficient and heterozygous control animals enabled us to validate the gating strategy (Fig. 1a). Interestingly, between 2 and 28 days of infection, we did not observe any significant differences in neutrophil content in infected tissue (Fig. 1b). As neutrophils have been shown to influence the recruitment of monocyte-derived macrophage-and dendritic cell-like phagocytes, we also determined the fractions of these cells in the infected ear (Fig. 1c). While no difference could be observed between Ly6G-deficient and control animals at early time points of the infection, we detected a small, but significant decrease in the fraction of monocyte-derived CD11c + MHCII + cells, namely mo-DCs, at 28 days p.i. (Fig. 1d,e). Pathogen burden remained unchanged both at 1 and 4 weeks p.i. (Fig. 1f).
Analysis of the Leishmania-expressed EGFP content of the different phagocyte populations at the site of infection, revealed that at-day 2 post infection, neutrophils in Ly6G-deficient animals had taken up significantly less parasite than cells from control settings (Fig. 1g,h), while the uptake in all other cell types tested was unchanged (Fig. 1g,i,j). Thus, although on the long term perspective, mice lacking Ly6G eventually control L. major infection at a comparable efficiency as wild type animals, the Ly6G-deficient neutrophils phagocytose less parasite at early phases of the infection. As Ly6C and CD11b levels have been reported to change with neutrophil activation 60,61 , we determined the mean fluorescence signal of these markers in Ly6G-deficient and control animals. Of note, while we did not observe any difference in CD11b surface expression, Ly6C was (slightly but) significantly increased in Ly6G-deficient animals (Fig. 1k).
Uptake of neutrophil cellular material by monocyte-derived phagocytes is strictly limited to infected cells, but not impaired in Ly6G-deficient neutrophils. Leishmania major infects mainly neutrophils in the first hours of the infection, but are mainly located within macrophage-and dendritic cell-like monocyte-derived phagocytes after a few days of infection 25 . Furthermore, the uptake of neutrophil material by these phagocytes has been shown to occur in the context of their infection with the parasite 59 . Thus, the changed parasite content within neutrophils of Ly6G-deficient mice early after infection might be due to differential cellto-cell transfer into the monocyte-derived cell types recruited to the site of infection.
In order to analyze the concomitant uptake of L. major together with neutrophil cellular material, we set out to measure the tdTomato content in monocyte-derived phagocytes of reporter mice expressing the tdTomato in neutrophils. For this, Ly6G-deficient Ly6G cre/cre × Rosa-tdTomato and Ly6G cre/+ × Rosa-tdTomato control mice were infected with EGFP-expressing L. major and analyzed 2 days p.i. Of note, at this time point, we expected efficient uptake of parasites from neutrophils by monocyte-derived cells to have occurred 25 . Confocal microscopy of the site of infection in Ly6G cre/+ × Rosa-tdTomato mice revealed that indeed, L. major-infected cells containing tdTomato material in subcellular compartments could be detected (Fig. 2a), with some of these cells found positive also for the dendritic cell marker CD11c (Fig. 2b). Flow cytometry analysis showed that the uptake of neturophil-derived tdTomato was nearly exclusively limited to infected cells (Fig. 2c). Ly6G +/+ littermate controls not expressing tdTomato ensured that any spectral overlap of EGFP with the tdTomato detection was not responsible for the higher signal in infected cells. This underlined that L. major transition from neutrophils into different phagocytes early after inoculation occurs together with the uptake of neutrophil material (Fig. 2c-e). However, no significant difference in neutrophil-derived TdTomato content was observable between Ly6G cre/+ controls and Ly6G cre/cre animals (Fig. 2d,e). Thus, Ly6G deficiency does not result in altered uptake of neutrophil material by recruited monocyte-derived phagocytes.
Ly6G-deficient neutrophils enter L. major-infected tissues with different kinetics. Ly6G has been suggested to mediate integrin interaction and extravasation of neutrophils [44][45][46][47] . Upon L. major inoculation, neutrophils are massively recruited within few hours to the site of infection. Thus, we hypothesized that in case the integrin interaction of Ly6G impacts on neutrophil behavior, this would be observable in the kinetics of early recruitment of neutrophils to the parasite. In order to investigate this, Ly6G-deficient Ly6G cre/cre × Rosa-tdTomato and Ly6G cre/+ × Rosa-tdTomato control mice were anesthetized, immediately infected with EGFPexpressing L. major, and imaged by intravital 2-photon microscopy. As reported previously 25  www.nature.com/scientificreports/ depicted. The upper row shows the histograms for quantification of the neutrophil-derived tdTomato signal (omitting the tdTomato high neutrophils) in mo-DCs (CD11b + CD11c + MHCII + ) and mo-macro (CD11b + CD11c -MHCII + ). Middle row: Gating on infected (EGFP + ) and non-infected (EGFP -) population of phagocytes. Lower row: Exclusive uptake of neutrophil-derived tdTomato material in the L. major infected (EGFP + ), but not the uninfected cell population. Ly6G +/+ littermate control mice not expressing tdTomato certify that spectral overlap of EGFP and tdTomato is not responsible for the higher tdTomato-signal in infected cells. (d,e) Analysis of the mean fluorescence intensity (MFI) of neutrophil-derived tdTomato signal in control and Ly6G deficient mice for mo-macro (d) and mo-DCs (e) revealed that tdTomato content is Ly6G independent, Thus, the uptake of neutrophil material by recruited monocytederived phagocytes is not altered due to Ly6G deficiency. Infected and non-infected cells are shown from Ly6G cre/+ × Rosa-tdTomato (black symbols), Ly6G cre/cre × Rosa-tdTomato mice (white symbols) and Ly6G +/+ littermates (grey symbols www.nature.com/scientificreports/ recruited to, and accumulating at, the site of infection within less than 2 h, phagocytosing the EGFP-expressing parasites (Fig. 3a). Interestingly, we observed between 1-2 h of infection a substantially higher number of neutrophils entering the site of infection in control mice as compared to Ly6G-deficient mice (Fig. 3a,b). Interestingly, at 2-3 h after infection, the initially lower neutrophil recruitment was compensated by a higher entry rate (Fig. 3b,c). Consequently, while the neutrophil entry rate into the tissue was constant over time in the wild type mice, Ly6G-deficient animals showed a significantly increased neutrophil entry rate between 2 and 3 h versus 1-2 h of infection (Fig. 3c). Of note, the neutrophils accelerated with time in both wild type and control animals, but were unaffected by the Ly6G deficiency (Fig. 3d). Thus, it is conceivable that neutrophil extravasation into the tissue, but not interstitial migration, is influenced by Ly6G. Neutrophil speed is drastically decreased with the uptake of L. major in the tissue 25 . In order to investigate the initial interaction of neutrophils with the parasite, we sought to quantify differential changes in cellular speeds linked to the distance between L. major and recruited neutrophils. For this, we compared the instantaneous speed of neutrophils in close proximity (less than 5 µm) or not contacting Leishmania (more than 5 µm) in wild type and Ly6G-deficient settings (Fig. 3e). As shown in Fig. 3f,g, neutrophil instantaneous speed was affected both by distance from Leishmania and time point of infection, but not by a lack of Ly6G. Therefore, although the recruitment of neutrophils lacking Ly6G is delayed early after infection with L. major, the neutrophil speed within the infected tissue is not affected.

Discussion
Ly6G is among the most important markers used to identify neutrophils in mice 48,62 , however whether it has a function in enabling neutrophils to be recruited to and engage pathogens during infection is unclear. Although Ly6G is present in mice, but not humans, elucidating its role for the dynamics of immune responses might reveal important differences between mouse models and human diseases, enabling the refinement of preclinical experimentation in order to more closely recapitulate clinical settings.
To study the role of Ly6G deficiency in neutrophils, we have employed the well-characterized model of L. major infection in a C57BL/6 mouse background, which is characterized to control the parasite efficiently [63][64][65] . Also, the initial dynamics of neutrophils and early host cell tropism of L. major at the site of infection has been intensely studied in this system 25,56,59 . This is relevant since the role of neutrophils while crucial cells for early defence against infections was shown to depend on the combination of Leishmania species and mouse model used 53,66,67 .
Antibody-mediated ligation of Ly6G attenuates beta-2 integrin dependent migration of neutrophils 44 . Furthermore, being a constituent of the MAC-1 integrin, CR3 has been shown to impact on phagocyte activation as well as the interaction with and uptake of pathogens 15,16,68,69 . With respect to the evidence for a crosstalk between Ly6G and integrin activation we hypothesized that the early interaction of immune cells, in particular neutrophils, with L. major might be influenced by Ly6G deficiency. Indeed, analysis of the uptake of EGFP-expressing Leishmania into neutrophils revealed that the early uptake of the parasite is decreased in Ly6G-deficient animals, ns not significant as determined by unpaired t test. (e) Distribution of neutrophil instantaneous speeds in proximity (< 5 µm) and at larger distance from EGFP-expressing L. major. Data are shown separately for the time periods between 60 and 120 versus 120 and 180 min p.i. for Ly6G cre/+ × Rosa-tdTomato (left panels) and Ly6G cre/cre × Rosa-tdTomato (right panels). (f) Analysis of the mean neutrophil distance to EGFP-expressing L. major. Data are shown separately for the time periods between 60 and 120 versus 120 and 180 min p.i. for Ly6G cre/+ × Rosa-tdTomato (closed symbols) and Ly6G cre/cre × Rosa-tdTomato (open symbols). Horizontal bars denote the median; ns, not significant as determined by one-way ANOVA with Bonferroni post test for multiple comparisons. (g) Analysis of neutrophil instantaneous speeds in proximity (< 5 µm) and at larger distance from EGFP-expressing L. major. Data are shown separately for the time periods between 60 and 120 versus 120 and 180 min. p.i. for Ly6G cre/+ × Rosa-tdTomato (closed symbols) and Ly6G cre/cre × Rosa-tdTomato (open symbols). Horizontal bars denote the median. Data were analyzed using a two-way ANOVA yielding a significant contribution of L. major distance and time point of analysis to neutrophil speed, but not of the Ly6G genotype. www.nature.com/scientificreports/ although parasite was found to be unchanged between wildtype and Ly6G-deficient animals eventually. The finding that as late as 4 weeks p.i., we find slight changes in recruited monocyte-derived dendritic cell content in the infected tissue underlines that early changes in neutrophil interaction with the parasite can affect the induction and execution of the adaptive immune response, as shown previously 58,70 . The finding that we do not observe differences in pathogen burden, however, indicates that these changes are compensated by other containment mechanisms, as has been shown for other immunodeficiencies during Leishmania infection [71][72][73] . Neutrophils have the potential to kill phagocytosed intracellular pathogens such as Leishmania. However, these immune cells can also serve as the first host cells for disease-transmitting parasites, with the intracellular killing mechanisms being evaded by the pathogen. Parasitized neutrophils were suggested to function as a 'Trojan horse' , to transfer Leishmania silently to macrophages 25,74,75 . In vivo imaging has contributed a second evasion mechanism called 'Trojan rabbit' strategy, whereby parasites escape apoptotic neutrophils to infect macrophages 76,77 . According to this notion, the dying neutrophil silences host macrophages to enable a productive Leishmania infection. Moreover, the parasite induces neutrophil apoptosis attracting macrophages and dendritic cells, and finally encourages engulfment of the parasite containing neutrophils by these terminal host cell preferred by Leishmania 25,59 . It is also important to mention that it was recently shown that efferocytosis itself is important for subsequent responses by the adaptive immune systems 78 .
Using Ly6G reporter mice in which neutrophil cellular material can be tracked by tdTomato fluorescence, we show that the uptake of neutrophil material by monocyte-derived macrophages and dendritic cells was only efficient in infected cells. Two explanations are conceivable: either L. major itself forces the concomitant uptake of neutrophils by these cells very strongly, or the parasite inhibits the degradation of neutrophil material in infected as compared to uninfected phagocytes. While Ly6G does not influence the uptake by phagocytes, the possibility of an efferocytosis forced by Leishmania might be an important concept to be investigated in future studies.
Using intravital 2-photon microscopy, we found no change in neutrophil speed, but a significant delay in tissue entry for Ly6G-deficient neutrophils as compared to wildtype. These different entry kinetics might also be linked with the slight but significant increase in Ly6C, which has been reported to change rapidly with neutrophil 60 , but not monocyte activation. While the lack of drastic differences in interstitial motility underlines the findings of intergrin independency during this stage 33,34 , we observe a compensation of the initially lower neutrophil entry rate by an increased entry rate 2-3 h p.i. Thus, if the inflammatory signals from the site of infection become sufficiently strong, this might result in efficient neutrophil extravasation also in a Ly6G-deficient setting. The decreased proportion of infected neutrophils at early time points p.i. observed in Ly6G-deficient animals could result from the inefficient initial tissue entry of these cells, which is likely to result from the altered integrin interactions 45 .
Besides the reported impact on integrin activation, recent studies suggest further, so far unrecognized functions of Ly6G: For example, single cell RNA sequencing revealed that Ly6G ligation might not, as previously suggested, mainly result in depletion of the neutrophils, but in shifts in neutrophil populations with respect to circadian rhythm-related neutrophil turnover 79,80 . Our observation that Ly6G deficiency impaired early neutrophil engagement of pathogens through delayed entry to the site of infection, therefore also opens interesting questions on the role of different neutrophil subpopulations in the control of intracellular pathogens.

Materials and methods
Pathogen. EGFP-expressing parasites (kindly provided by T. Aebischer, Robert Koch Institute, Berlin, Germany) were generated by stable integration of the EGFP coding sequence 81 into a small subunit (18S) rRNAcoding locus of L. major LRC-L137 V121 82 . Parasites were grown in M119 medium supplemented with 10% heat-inactivated fetal calf serum, 10 mM adenine, 1 mg/ml biotin, 5 mg/ml hemin, 2 mg/ml biopterin (all from Sigma) and 7.5% (m/v) NaHCO 3 at 26 °C for a maximum of 6 passages. Maintenance of EGFP transgene in transgenic strains is associated with co-inserted hygromycin B resistance gene, hence hygromycin B was routinely added to parasite culture. www.nature.com/scientificreports/ mania culture medium and again centrifuged at 1000×g for 10 min at RT. After discarding the supernatant, the pellet containing the parasites was subsequently taken up in fresh, 4 °C cold PBS (Dulbecco without Calcium and Magnesium, L 1820), spun down at 1000×g for 5 min at RT and taken up in PBS for a final concentration of 2 × 10 4 L. major per µl, of which 0.5 µl were injected into the ear dermis.

Mice
Cell isolation. Infected  Intravital two-photon imaging. The mice were injected with 100 mg/kg Ketamin and 10 mg/kg Xylazin i.p. and supplemented with 3 mg/kg Acepromazin i.p. after entering anaesthesia. The ventral side of the ear was prepared for intravital 2-photon microscopy as described previously 83,84 . In brief, the animals were placed on a heating stage adjusted to 37 °C and the ear was fixed to a metal platform using double-sided tape and covered with Vidisic carbomer gel (Bausch + Lomb). A coverslip sealed to a surrounding parafilm blanket was placed onto the ear and fixed above the ear. Intravital 2-photon imaging was performed using Zeiss LSM 700 upright microscope equipped with a Mai Tai DeepSee Ti:Sa laser (Spectra-Physics) tuned to 980 nm. The emitted signal was sequentially split by 625 nm (reflected), 490 nm and 560 nm long pass dichroic mirrors and filtered with 485 nm SP (second harmonics), 525/50 nm (EGFP), 589-654 nm (tdTomato), filters before collection with nondescanned detectors.
The analysis was carried out using the Imaris analysis software using the Isosurface tool. Neutrophils could be marked and assigned to a position with x, y and z coordinates using the tracking function. Calculation and analysis of neutrophil speeds and distances to L. major was performed from datasets extracted from the Imaris software using ImageJ (http:// imageJ. nij. gov/ ij) and the DISCit software 85 .
Statistical analysis. One-way analysis of variance (ANOVA) with appropriate multiple comparison posttests (indicated in the respective figure legends) was employed to compare multiple samples or pairwise analysis within datasets with more than two experimental groups. For time course experiments, two-way ANOVA with genotype and time point of infection was employed as indicated in the figure legends. Two-group comparisons were made by two-sided, unpaired t tests. Data are always presented as mean with each individual sample. Statistical analysis, including Shapiro-Wilk normality testing of the acquired data, was performed using the Prism software (Version 8.0, GraphPad).

Data availability
The raw data that support the findings of this study are available from the corresponding author upon request.