No indication of aberrant neutrophil extracellular trap release in indolent or advanced systemic mastocytosis

In disease states with chronic inflammation, there is a crosstalk between mast cells and neutrophil granulocytes in the inflamed microenvironment, which may be potentiated by tryptase. In systemic mastocytosis (SM), mast cells are constitutively active and tryptase is elevated in blood. Mast cell activation in SM leads to symptoms from various organs depending on where the active mast cells reside, for example, palpitations, flush, allergic symptoms including anaphylactic reactions, and osteoporosis. Whether neutrophil function is altered in SM is not well understood. In the current study, we assessed nucleosomal citrullinated histone H3 (H3Cit‐DNA) as a proxy for neutrophil extracellular trap release in plasma from 55 patients with indolent and advanced SM. We observed a strong trend towards a correlation between leukocyte count, eosinophil count and neutrophil count and H3Cit‐DNA levels in patients with advanced SM but not in indolent SM; however, no differences in H3Cit‐DNA levels in SM patients compared with healthy controls. H3Cit‐DNA levels did not correlate with SM disease burden, tryptase levels, history of anaphylaxis or presence of cutaneous mastocytosis; thus, there is no evidence of a general neutrophil extracellular trap release in SM. Interestingly, H3Cit‐DNA levels and leukocyte counts were elevated in a subgroup of SM patients with aberrant mast cell CD2 expression, which warrants further investigation. In conclusion, we found no evidence of global increase in neutrophil extracellular trap release in SM.


| INTRODUCTION
][4][5][6][7] Tryptase can be stabilized by DNA, 8 and a recent study demonstrated that tryptase may regulate neutrophil extracellular trap (NET) release and increase NET release in PMA activated human neutrophils whereas in mice lacking tryptase, less NETs were formed 9 indicating that under inflammatory conditions, mast cells may alter neutrophil function.
Systemic mastocytosis (SM) is a heterogeneous group of diseases characterized by an activating point mutation D816V in the KIT gene, a mutation that renders mast cells constitutively active.In SM, neoplastic mast cells accumulate in bone marrow, skin and other organs 10 causing symptoms of mast cell activation.SM is most frequently diagnosed by a bone marrow biopsy and the diagnosis is established based on criteria including aggregates of mast cells, aberrant mast cell morphology, aberrant CD2 and/or CD25 expression on the mast cell surface, elevated basal serum tryptase, and the presence of the KIT D816V mutation. 10,11The majority of SM patients have an indolent disease with normal life expectancy, however around 10% will have an advanced SM (AdvSM) and frequently also a concomitant non-mast cell haematological neoplasm such as a myeloproliferative neoplasm (MPN), myelodysplastic syndrome (MDS) or chronic myelomonocytic leukaemia (CMML).In AdvSM, mast cell bone marrow infiltration causes bone marrow failure, and mast cell infiltration in other organs may cause organ impairment and dismal prognosis. 10,12 hallmark of SM and a diagnostic criterion for the disease is constantly elevated serum tryptase levels >20 μg L −1 . 11Elevated basal serum tryptase levels are seen in over 70% of patients 13,14 and reflect a constant activation and granule secretion of mast cells.
Neutrophil extracellular traps are a component of the first line of immune defence. 15,16However, despite known crosstalk between neutrophils and mast cells, neutrophil extracellular trap release in SM has not been investigated.In the current study, we assessed peptidyl arginine deiminase 4 (PAD4) dependent NET release in plasma of patients with indolent and AdvSM by quantifying nucleosomal citrullinated histone H3 (H3Cit-DNA) levels in plasma.

| SM patient and healthy control samples
Patients with SM diagnosed at the Center of Excellence for SM at Karolinska University Hospital (Stockholm, Sweden) according to the WHO criteria 11 were included in the study after oral and written informed consent.Samples were also obtained from healthy controls after oral and written consent.The study was approved by the Regional Ethical Review Board in Stockholm and the Swedish Ethical Review Authority and conducted in accordance with the declaration of Helsinki.
Peripheral blood was drawn into vacutainer EDTA tubes, plasma isolated by centrifugation 15 minutes at 2000 g within 1 hour of sampling and stored in −80°C until time of analysis.

| Statistical methods
To test for normal and log-normal distribution, Shapiro-Wilk normality test was used.Clinically used biomarkers were not log-transformed.The unpaired student t-test was used for parametric data before or after log transformation, and Mann-Whitney U-test was used for non-parametric data.One-way ANOVA was used if there were more than two independent groups.Correlations between biomarkers were investigated using Spearman correlation coefficient.Statistical analysis was performed using GraphPad Prism 9.
Single-cell transcriptomics data (gene expression counts) of low-density bone marrow cells as deposited on the Gene expression Omnibus database (GEO number GSE222830) was processed. 18The batch-corrected data were re-clustered using the Leiden algorithm (resolution 0.2) to represent the cell types with finer resolution, and reannotated [19][20][21] in addition to previously annotated clusters.

| Patient and healthy control characteristics
A total of 55 systemic mastocytosis patients were included in the study, 13 of which had AdvSM disease, and 42 with indolent SM (ISM).Clinical disease features are further outlined in Table 1.Of the 13 patients with advSM disease, 12 had SM with an associated haematological neoplasm (SM-AHN), of which there were four patients with an associated MDS, three patients with associated CMML, two patients with associated MPN, one with MDS/MPN overlap disease, one each with associated high-grade lymphoma and myeloma.One patient had only aggressive SM and no associated other haematological neoplasm.For comparison, a total of 19 healthy controls were included in the study, median age 52.0 years (range 35-75, mean 52.6 years), of which 9 were males and 10 females.

SM and correlation to disease parameters
NET release measured by H3Cit-DNA levels in plasma revealed overall no difference in H3Cit in SM patients compared to healthy (Figure 1A).Levels of H3Cit-DNA were similar in ISM and AdvSM (Figure 1B), however, varied within the heterogeneous AdvSM group.Of the patients with AdvSM, four patients were untreated as they were sampled at time of diagnosis, two were on antihistamines only.Two patients were on tyrosine kinase inhibitor treatment.Three patients were on either interferon or steroid treatment or both, one was on hydroxyurea, and one on azacitidine.Mean levels were numerically higher in AdvSM patients treated with tyrosine kinase inhibitors, hydroxyurea or azacytidine compared to untreated AdvSM patients (Figure S1).
Levels of H3Cit-DNA did not correlate to serum tryptase or bone marrow mast cell infiltration degree, both established markers of SM disease severity (Figure 2A,B).Similar results were obtained when ISM and AdvSM were analysed separately (Figure S2).In addition, H3Cit-DNA did not correlate to alkaline phosphatase or β2microglobulin, elevated levels of which are associated with advanced SM disease and adverse prognosis (Figure 2C,D).When ISM and AdvSM were analysed separately, there were moderate trends towards positive correlations between H3Cit-DNA and alkaline phosphatase as well as to β2microglobulin in AdvSM but not ISM and (Figure S3).No correlation was seen between H3Cit-DNA and leukocyte, eosinophil count or neutrophil count (Figure 3).However, when the analysed separately, we observed strong trends towards correlations between  leukocyte count, eosinophil count and neutrophil count and H3Cit-DNA levels in patients with advanced SM but not in indolent SM (Figure S4), in line with what has previously been demonstrated in patients with solid cancer 22 and COVID-19. 23Mastocytosis in the skin (cutaneous mastocytosis, CM), anaphylactic reactions and osteoporosis are frequent symptoms of SM and may potentially involve neutrophil activation; however, we found no differences in H3Cit-DNA levels in the SM patients depending on presence or absence of CM, presence of anaphylactic reactions, or of osteoporosis (Figure 4).

| Elevated H3Cit-DNA levels in patients with aberrant mast cell CD2 surface expression
Interestingly, H3Cit-DNA levels were significantly elevated in SM patients that had aberrant mast cell surface CD2 expression (Figure 5A).The CD2± patient populations also showed significant differences in leukocyte count, neutrophil count, eosinophil granulocyte count with increased levels in the CD2+ population, but no difference in serum tryptase levels (Figure 5B-E).

| Analysis of PAD4 and CD58 expression by single-cell RNA seq
To assess whether the PAD4 dependent H3Cit-DNA assay used in this study measured only neutrophil derived extracellular traps, we assessed PAD4 mRNA (PADI4 gene) expression in a single-cell RNA-seq dataset of three ISM patients.PAD4 was expressed in the neutrophil population and essentially not in any other cell type (Figure 6) thus in the current assay there is no contribution of potential mast cell derived extracellular traps.Although the CD2 gene expression levels by single-cell RNA seq were low in mast cells compared to T/NK cells (Figure S5), all three patients had aberrant mast cell surface CD2 expression according to the clinical routine diagnostic workup.The CD2 ligand CD58 was highly expressed in neutrophils  as expected (Figure 6), possibly indicating that aberrant CD2+ mast cells in SM may activate neutrophils via CD2-CD58 interaction.

| DISCUSSION
In the majority of SM patients, serum tryptase is elevated as a consequence of chronic mast cell activation.Based on the findings that in chronic inflammation activated mast cells recruit neutrophils to the site of inflammation, [2][3][4][5][6][7] and the recent finding that in vitro stimulation of healthy neutrophils with tryptase potentiated NET release, 9 we set out to assess whether NETs are increased in SM patients.
To our knowledge, altered NET release has not been previously investigated in the setting of SM patients.In the current study, we used an H3Cit-DNA ELISA to assess NETs in the plasma of ISM and patients.Of note, as citrullination of histones is PAD4-dependent, 24 this assay only quantifies PAD4-dependent NET release.In the current study, we found overall similar H3Cit-DNA levels in ISM compared to healthy, thus no signs of NET release as measured by H3Cit-DNA.This is perhaps not surprising as we 18 and others 25,26 have shown that most cytokine/chemokine levels are similar in ISM and healthy, which is also why it has been difficult to establish biomarkers for SM disease. 18However, to our surprise, mean H3Cit-DNA levels in AdvSM, which are patients with an expected overall survival of 2-4 years 10,27 were also similar to healthy, although a great variation was seen.Recently, increased NET release has been demonstrated in patients with MPN 28,29 as well as when ex vivo stimulating chronic myeloid leukaemia patient cells with ionomycin or PMA, 30 and impaired NET release in peripheral blood of patients with MDS. 31 In our AdvSM cohort, we could see a similar trend as our SM-AHN-MDS cases had lower NET release and our myeloproliferative cases with MPN or CMML had increased NET release; however, the numbers are low  and requires further investigation in a larger SM-cohort.Similarly, we observed higher levels of H3Cit-DNA in AdvSM patients treated with tyrosine kinase inhibitors, hydroxyurea or azacytidine compared to untreated AdvSM patients, however we cannot draw any conclusions regarding whether different treatments affect NET release from this study.Blood sampling before and after treatment initiation will help determine how different treatment options impact NET release.In our study, there were trends towards correlation between H3Cit-DNA levels and alkaline phosphatase, β2microglobulin, white blood cell count, eosinophil count and neutrophil count in patients with advSM but not in patients with ISM.However, H3Cit-DNA levels were not associated with SM disease burden, presence of anaphylaxis or presence of cutaneous mastocytosis.Further, we observed no correlation of H3Cit-DNA levels to tryptase levels.Thus, chronic elevation of tryptase does not seem to generate any general increase in NETs, at least not detectable in plasma.However, it is known that mast cells may produce extracellular traps 32 and that extracellular traps can be formed locally in skin tissue. 33,34In the current study, we cannot exclude a local extracellular trap release in skin or other organs, however we could not detect elevated H3Cit-DNA levels in SM patients with cutaneous mastocytosis lesions compared to patients without.Of note, the current study only measured PAD4-dependent NET release and we could demonstrate by single-cell RNA sequencing analysis that neutrophil granulocytes are the main producers of PAD4 whereas mast cells virtually do not produce PAD4 transcripts.Thus, SM mast cells are unlikely to contribute to H3Cit-DNA generation.
F I G U R E 5 Mast cell CD2 expression and levels.When comparing SM patients with aberrant mast cell surface CD2 expression to patients without, levels of (A) H3Cit-DNA were significantly increased in CD2+ SM patients, (B) white blood cell count was significantly increased in CD2+ SM patients, (C) neutrophil count was significantly increased; however, (D) no change in eosinophil count and (E) no change in serum tryptase levels were found.H3Cit-DNA levels are log2-transformed.Interestingly, 64% (35/55) of the SM patients presented with aberrant mast cell surface CD2 expression.Aberrant CD2 surface expression is a known phenomenon in SM; however, the functional consequences of aberrant mast cell CD2 have not been investigated in detail.In SM patients with aberrant mast cell CD2 expression, total white blood cell count was increased, and NET release was elevated compared to patients with normal non-CD2 expressing mast cells.This indicates a possible crosstalk between aberrant CD2+ SM disease mast cells and its receptor CD58 that we found was strongly expressed in SM neutrophils, to local NET release.
In conclusion, we found no evidence of global increase in NET release in SM; however, a possible functional role of aberrant mast cell surface CD2 expression in inducing NET release.This warrants further investigation of the functional role of aberrant mast cell CD2 in SM.

DATA AVAILABILITY STATEMENT
Aggregated data are presented in the article.The patient datasets generated for the current study are not publicly available due to privacy concerns and limitations from the ethical review board.Single-cell transcriptomics data (gene expression counts) of low-density bone marrow cells are available through the Gene expression Omnibus database (GEO number GSE222830).

T A B L E 1
Patient characteristics.

F I G U R E 1 23 F I G U R E 2
H3Cit-DNA levels in SM patients and healthy volunteers.(A) Levels of neutrophil extracellular trap marker H3Cit-DNA in plasma were similar in SM patients and healthy volunteers, and (B) with no difference between AdvSM and ISM patient groups.H3Cit-DNA levels are log2-transformed.Correlation between H3Cit-DNA and SM disease severity markers.Correlation between H3Cit-DNA and (A) serum tryptase, (B) bone marrow (BM) mast cell infiltration degree, (C) alkaline phosphatase (ALP), and (D) β2 microglobulin in the SM patient cohort reveal no correlation between known SM disease severity markers and H3Cit-DNA.BM mast cell infiltration was unknown in 2 patients, ALP levels in 2 patients and β2 microglobulin levels in 8 patients.H3Cit-DNA levels are log2-transformed.

F I G U R E 3
Correlation between H3Cit-DNA and blood cell counts.(A) No correlation was seen between H3Cit-DNA and white blood cell count (WBC), (B) eosinophil count, or (C) neutrophil count.Eosinophil count was unknown in 6 patients.H3Cit-DNA levels are log2-transformed.

4
H3Cit-DNA levels and SM clinical presentations.No differences were seen in H3Cit-DNA levels with regard to symptoms of (A) cutaneous mastocytosis (CM), (B) anaphylaxis, or (C) presence of osteoporosis (OPo) or osteopenia (OPe).H3Cit-DNA levels are log2-transformed.

F
I G U R E 6 Single-cell RNA sequencing of bone marrow mononuclear cells enriched for mast cells, from three ISM patients.(A) Cell type annotation, (B) We observe PAD4 expression almost exclusively in neutrophil granulocytes, CD58 expression mainly in neutrophils but also in NK/T-cells and mast cells, and neutrophil markers KCNJ15 and LCN2 in neutrophils.(C) Heatmap of cluster-wise gene expression of selected genes.Values have been normalized by gene, and plotting values set between 0 and 1.