Early Exposure of Human Neutrophils to Mycobacteria Triggers Cell Damage and Pro-Inhibitory Molecules, but not Activation

Mycobacterium tuberculosis (Mtb) is responsible for causing tuberculosis (TB) [1], a chronic reemerging infectious disease with a major impact on public health worldwide [2]. Primary TB occurs by inhalation of microdroplets containing Mtb bacilli [3]. If the innate immune response is not efficient to control and eliminate the mycobacteria, then adaptive T cell responses might contain the bacteria resolving the infection [3-5].


Introduction
Mycobacterium tuberculosis (Mtb) is responsible for causing tuberculosis (TB) [1], a chronic reemerging infectious disease with a major impact on public health worldwide [2]. Primary TB occurs by inhalation of microdroplets containing Mtb bacilli [3]. If the innate immune response is not efficient to control and eliminate the mycobacteria, then adaptive T cell responses might contain the bacteria resolving the infection [3][4][5].
Being TB a long-lasting affliction, most research on the immune responses to TB has addressed the chronic stages. In contrast, the earliest contacts of bacilli with cells responsible of the first defenses have been rather neglected, until recent years. Neutrophils (PMNs) are short-lived leukocytes, functioning as the primary defense against microorganisms [6,7], with special ability to kill pathogens by phagocytosis, degranulation and the formation of extracellular traps [8][9][10][11]. The activation of neutrophils is associated with changes in the spatial and temporal expression of certain molecules [12,13] such as CD16b, CD11b and CD66b, molecules responsible for adhesion, degranulation and migration, and are therefore well established as neutrophil activation markers [13][14][15][16][17][18][19]. On the other hand, it is documented that intracellular pathogens causing chronic infections can co-opt the pathway of a pair of molecules, PD-1 and its ligand PD-L1, involved in decreasing crucial T cell responses, provoking a phenomenon called "T cell exhaustion" [20]. In physiological conditions, this pathway is important for peripheral tolerance and the control of adaptive immune responses [21], but it has been barely explored in TB.
In this work we wanted to assess the phenotype of neutrophils during the very early interactions with three different strains from the Mycobacterium tuberculosis-complex.

Ethics
This research was performed on healthy competent volunteers in accordance with the Declaration of Helsinki of the world Medical Association, and the Mexican General Health Law regarding research. The ethics committee of the National School of Biological Sciences ENCB-IPN approved this study (permission number: "Protocolo #CEI-ENCB 011/2013") and informed written consent was obtained from donors.

Statistics
Statistics were performed with Two-way ANOVA using Bonferroni t-test for all multiple pairwise comparisons using GraphPad Prism 5 project (©2013 GraphPad Software).

Neutrophils purity
Blood neutrophils from healthy donors were enriched to about 90% purity, which was checked through staining with Hematoxylin and Eosin (H&E) as well as by flow cytometry parameters such as cell size (FSC) vs. granularity/internal complexity (SSC) (Figure 1

Mycobacterium affects neutrophils integrity at early times of interactions
By means of flow cytometry we analyzed the neutrophils incubated either with culture medium alone, LPS, PMA, Mycobacterium bovis BCG, Mycobacterium canetti or Mycobacterium tuberculosis H37Rv. We observed an increase in neutrophils autofluorescence, i.e., those that were incubated only with mycobacteria but that were not labeled with any fluorescent reagent at all. Autofluorescence was analyzed in a free channel or filter without staining (FL-6=405 nm) (figure 2). In flow cytometry it has long been described that autofluorescence indicates (membrane) cell damage or even cell death [22][23][24][25][26][27]. In this case, the neutrophils incubated with BCG or with Mtb H37Rv increased their autofluorescence emission since early times of exposure to mycobacteria. At 15 min of incubation (figure 2B, D) approximately 60-80% of PMNs were autofluorescent, while for neutrophils interacting with Mycobacterium canetti the autofluorescence increased approx. 15% at 30 minutes (figure 2G). In contrast, neutrophils incubated with culture medium alone showed basal levels, approximately 10% of autofluorescence (figure 2 A, E, I)

Mycobacteria do not activate neutrophils at early times of interaction
Blood neutrophils incubated separately with three different strains of Mycobacteria (BCG, Canetti and Mtb) were evaluated for the expression of CD16b (Figure 3), CD11b and CD66b (Figure 4), at 15, 30 and 60 min incubation. Although these molecules are constitutively expressed on neutrophils, it is well documented that neutrophils modify the expression of these markers upon activation [13,15,17,19,[28][29][30].
Neutrophils incubated with mycobacteria did not display changes in the percentages of expression of CD16b ( Figure

Mycobacteria rapidly induce pro-inhibitory molecules on neutrophils
Mycobacteria have evolved diverse mechanisms to escape, divert or even subvert the immune responses [4,31]. The molecular pair PD1-PD-L1 has been recently shown to induce a phenomenon called T cell exhaustion, and the expression of these two molecules has been shown manipulated by certain pathogens for their advantage [32,33]. The percentage of PD1 expression in neutrophils incubated with mycobacteria increased only with Mtb H37Rv (figure 5A, black arrows) since 15 min, compared with basal expression levels. In contrast, PMNs interacting with the other mycobacterial strains did not modify the percentage of PD-1 expression ( Figure 5A). Regarding MFI we observed a decrease in PD-1 intensity in neutrophils incubated with Mtb H37Rv MOI:1 at 15 and 30 min, but an increase at 60 min ( figure 5B, black arrow). With respect to PD-1 Ligand (PD-L1), the percentage increased in neutrophils incubated with BCG and Mycobacterium canetti at 15 min, compared with basal controls ( Figure 5C). Neutrophils incubated with Mtb H37Rv show a tendency to increase the MFI of PD-L1 at 15 and 60 min of incubation (figure 5D, blue arrow), this increase was more clear after 60 min, compared with basal intensity and with the other mycobacterial strains tested.

Discussion
Mycobacterium tuberculosis is the causative agent of tuberculosis (TB) [1]. According to the World Health Organization (WHO), TB caused the death of 1.3 million people worldwide in 2012 [2]. Despite the vast amount of TB research in more than 100 years, there are still many aspects of the disease that remain poorly known, for instance the host's innate response during the earliest stages of the infection. The involvement of neutrophils in the immune response against Mtb has been rather neglected until recently. Conceivably, this might be due to the short life span of these cells, the quickness of their mechanisms of action and also to the chronicity of TB.
We aimed at assessing activation as well as pro-inhibitory molecules on neutrophils from healthy donors using flow cytometry, thus PMNs were incubated for short periods with three different strains of mycobacteria. As positive controls for PMN activation at the time points evaluated, we used two standard components, a microbial one such as lipopolysaccharide (LPS), and a chemical one such as phorbol myristate acetate (PMA). At least two main FACS criteria were evaluated, the percentage of cells and the median fluorescence intensity (MFI). The latter directly reflects the intensity of expression of a given molecule in the cells evaluated.
In flow cytometry the emission of autofluorescence in both animal and plants cells has been indicative of cell damage or even cell death. In addition, the reduction of NADH induces intracellular fluorescence, also indicating cell damage [22][23][24][25][26][27]. When we cultured mycobacteria with neutrophils, these increased their autofluorescence emission since 15 min, compared with neutrophils in medium alone. Previously, Perskvist et al. (2002) demonstrated that Mtb H37Rv induced apoptosis in neutrophils, but this was evaluated at 5h of interaction [34,35]. However other studies of neutrophils with Leishmania major showed increased IL-8 production and inhibition of apoptosis at 24h [36]. Corleis et al. (2012) described that at 20 min, neutrophils can kill Mycobacterium smegmatis but not Mycobacterium tuberculosis [37]. In addition, Ramos-Kichik et al. (2009) showed that Mtb induced neutrophil extracellular traps (NETs) at 180 min [38], compared with other microorganisms such as C. albicans that do it in shorter time [39].
Few and recent studies have focused on the effects that mycobacteria might have on neutrophils [40][41][42], for instance whether mycobacteria can induce neutrophil extracellular traps, as seen when PMNs interact with other microorganisms [38]. Of note, even the phenotype of neutrophils during and after their interaction with mycobacteria remains poorly addressed.
From the set of molecules that we measured, some (CD16b, CD11b, CD66b) are well established neutrophil activation markers, while the others (PD1/PD1-L) are very important pro-inhibitory molecules. Interestingly, however, even at the longest time of incubation with mycobacteria (60 min), we did not observe statistically significant differences in the percentages of PMNs expressing CD16b, CD11b or CD66b, compared to neutrophils incubated with the control substances LPS, PMA or with culture medium alone. In contrast, it is known that certain components from microorganisms such as E. coli, S. thyphimurium. S. flexneri and Y. enterocolitica induce activation of neutrophils at early times (30 min) of exposure [43]. Interestingly, even at 15 min incubation the percentage and the MFI of CD11b, CD66b, CD18 or CD62L increased in human healthy neutrophils incubated with DNA of E. coli, LPS, fMLP, C5a and three different strains of Staphylococcus aureus [19,44].
On the other hand, decrease in surface CD16b in neutrophils is also related to the activation of these cells [29,30]. When we analyzed the median fluorescence intensity (MFI) of CD16b on mycobacteria-incubated neutrophils, we did not find significant differences compared to the controls, although there was a slight decrease at 15 minutes in cells incubated with Mtb H37Rv compared to freshly isolated neutrophils. Regarding the expression of CD11b and CD66b, both LPS and PMA (used as positive controls for stimulation), increased the expression of these two markers since 15 minutes compared to neutrophils in medium alone, indicative of the intact capacity of neutrophils to respond. Although there was a subtle decrease in CD66b expression at 15 minutes of incubation with bacilli, compared to freshly isolated neutrophils.
Previous studies have evaluated the expression of CD16b and CD66b as activation markers in neutrophils from patients with active TB [45]. Likewise, in healthy neutrophils incubated with clinical isolates of Mtb, the expression of CD16, CD69 and CXCR2 was evaluated, but only at 3 and 18 h [46]. Hilda et al. in 2012 demonstrated that the virulent strain Mtb H37Rv can modulate the expression of FcγRII (CD32), FcγRI (CD64), TLR-4 and CXCR3 in neutrophils, while BCG increased CD32 expression only, and a vaccine strain, Mycobacterium indicus pranii, did not produce any apparent effect with MOI: 3, all at 4h of incubation [47]. We did not see differences in neutrophil activation in response to the three strains of mycobacteria, regardless of their different virulence.
The PD-1/PD-L1 pathway is important for the establishment and maintenance of peripheral tolerance as well as to modulate immune and inflammatory responses to pathogens [21,48]. This pathway has been found altered during chronic viral [49,50], parasitic [51], fungal [52] and bacterial [32] infections, provoking a phenomenon known as "exhaustion" in CD8+T cells, exerting an inhibitory effect related to pathogens evasion of immune responses [20]. Apparently, Mtb H37Rv can also modulate and manipulate the PD/PD-L1 path not only in T cells, but also in antigen presenting cells and even in elements of the innate immune response, such as NK cells [53]. However, this pair of molecules has been poorly explored in neutrophils, especially in the interplay with microorganisms. Therefore, we sought to evaluate also the neutrophils expression of PD-1 and PD-L1 in the early response to mycobacteria. There was an apparent increase in the proportion of neutrophils expressing surface PD-1 following incubation with Mtb H37Rv (MOI 1), which was more evident at 60 minutes; although not statistically significant. These results are interesting and opposed to data by Yao et al. (2009) where they reported that neither NK cells, macrophages nor neutrophils express PD-1 during infection with Listeria monocytogenes [54]. On the other hand, we did not observe changes in the proportion of neutrophils expressing PD-L1 after incubation with the different mycobacteria strains at the time points we evaluated. Regarding the mean fluorescence intensity (MFI) of PD-1, we observed that at 15 and 30 min there were no differences in its expression in neutrophils incubated with mycobacteria. However, at 60 min the expression of PD-1 increased importantly in neutrophils but only with Mtb H37Rv at MOI 1. When evaluating the MFI for PD-L1 we did not find significant differences among neutrophils incubated with mycobacteria or with culture medium alone. McNab et al. (2011) also evaluated these pro-inhibitory molecules but at the mRNA level in the blood from tuberculosis patients [55]. They found that the polymorphonuclear population overexpressed the PD-1 ligand in patients with active TB, in comparison to patients with latent TB and healthy donors.
Our results indicate that, while neutrophils are readily stimulated upon short time of incubation with either LPS or PMA, they do not get activated during early interactions with three mycobacteria of different virulence. This is revealed by the PMNs inability to modify activation molecules which are involved in crucial processes such as migration, adhesion, phagocytosis and degranulation. However, at these early times of interactions with mycobacteria, there are indeed alterations occurring in other important molecules in PMNs, such as PD1/PD1-L.
It is conceivable that early during a mycobacterial infection neutrophils might encounter and interact with other recruited cells of the immune system while still expressing inhibitory molecules such as PD-1 or PD-L1, thus limiting the intensity of the response at the site of infection. The fact that neutrophils infected with Mtb express both the receptor and the ligand of PD-1 could be related to an autocrine regulation of this population. Further experiments are required to understand the role of the early innate immune response against Mycobacterium tuberculosis. Since there are few studies focused on the neutrophils phenotype during the very early interactions with mycobacteria, we consider our results of interest and we expect that this work will instigate more research on the participation of neutrophils during the early stages of mycobacterial infection.