Mutations in the gene for neutrophil elastase (ELANE) are the most common cause for cyclic (CyN) and congenital neutropenia (SCN) [1-3]. The majority of patients with these autosomal dominant disorders can be effectively treated with the granulocyte colony-stimulating factor (G-CSF) [4,5]. G-CSF is administered subcutaneously, daily, every other day or several times per week and it has expected adverse effects. These are most commonly bone pain, headache and arthralgias, but for some patients it may predispose to the development of myelodysplasia and or acute myeloid leukemia [6,7]. Although this risk is difficult to quantitate, the concern leads to hesitancy by patients, families and physicians with long-term administration, particularly at high doses [8]. The only other effective treatment currently is hematopoietic stem cell transplantation, and therapy is limited by availability of a suitable donor, the inherent risk of graft versus host disease and the cost and complexity of this treatment [9-11]. Numerous other therapies have been tried without consistent benefit [12].

ELANE, located on chromosome 19, encodes a 29 kd powerful protease which has many potential substrates, including G-CSF [13]. Mutations in this protease were first recognized in CyN and then in SCN because of the overlapping presentations and symptoms for these conditions, largely based on family studies [14]. With availability and wide use of Sanger and exome sequencing, more than 200 different mutations have been identified [1,15]. The clinical phenotypes vary from mild to severe, the more severely affected patients respond poorly to G-CSF and are at greatest risk of evolution to MDS and AML. The mutations G214R, C151Y, and C233X portend the greatest risk [15]. Because the risk of this evolution cannot be clearly separated from G-CSF treatment, we and other investigators have sought alternative therapies.

We have pursued the possibility that inhibitors of neutrophil elastase might be alternatives to G-CSF over several years [16]. We began these investigations based on the hypothesis that the mutant enzyme was the sole cause for these diseases. We found that one inhibitor called MK0339 was consistently the best of the tested inhibitors to promote cell survival and differentiation of HL-60 cells expressing mutant ELANE [17]. Recently we have solidified this hypothesis through gene knockout and gene editing studies [18,19]. We have now expanded on this work comparing several selective inhibitors of NE using CD34⁺ cells from patients with SCN and CyN. We observed that only some of the cell permeable NE inhibitors permit differentiation; others are inactive across a spectrum of CD34⁺ cells from patients. In this report we provide structural modeling data on NE and its interactions with potential therapeutic NE inhibitors to enhance these observations. We believe this work is foundational for advancing the possibility of a novel oral therapy for ELANE associated neutropenia.

IRB approval

The Institutional Review Board of the University of Washington approved these studies. Informed written consent was obtained from all of the subjects. Bone marrow samples were collected in association with an annual follow-up as recommended by the Severe Chronic Neutropenia International Registry.

Patients’ cells: Bone marrow: Separation, CD34 cell preservation and expansion

Freshly collected, anticoagulated bone marrow (3-6 ml) was shipped overnight at an ambient temperature. Hematopoietic stem and progenitor cells (HSPC’s) were initially enriched using RosetteSep Human Bone Marrow Progenitor Cell Pre- Enrichment Cocktail, (Cat. no.15027; All catalog numbers refer to materials from StemCell Technologies unless indicated otherwise) and Lymphoprep (Cat.no. 07801) according to manufacturer’s protocol. The HSC enriched cell population was expanded by culturing for 4 days in CD34⁺ expansion media (StemSpan SFEMII media (Cat.no. 09655) supplemented with 1% Penn Strep (Cat.no 03-031-1B, Biological Industries), 1x StemSpan CD34⁺ Expansion Supplement (10x) (Cat.no. 02691), and 1.0 μM UM729 (Cat.no.72332), at 37°C 5% CO₂. After expansion, CD34⁺ cells were further enriched using EasySep Human CD34 Positive Selection Kit II (Cat.no. 17856) according to manufacturer’s protocol. Cell counts and viability were evaluated using an Invitrogen Countess and trypan blue. Enriched CD34⁺ cells were cryopreserved at 400x10⁵ cells/ml in Cryostor CS10 (Cat.no. 07931). Cells were stored in liquid nitrogen, vapor phase.

Leukapheresis cells

CD34⁺ cells were collected following 5 days of subcutaneous G-CSF administration (10 mcg/kg/day). On day 5 plerixafor was administered once subcutaneously (240 mcg/kg/dose) at Bloodworks Northwest, Seattle WA and peripheral blood stem cells (PBSC) collected by standard institutional procedures. CD34⁺ cell enrichment was performed on the freshly mobilized PBSCs using CliniMACS and cryopreserved following institutional Standard Operating Procedure. For controls, cryopreserved healthy human CD34⁺ progenitor cells from mobilized peripheral blood were obtained from Lonza (Cat no. 4Y-101C).

Differentiation assay

Previously frozen CD34⁺ cells thawed and were allowed to recover for 3 days in CD34⁺ expansion media and were subjected to a differentiation protocol adopted from Nasri et al. [20]. In brief, HSCs were cultured for 7 days in RPMI (Cat.no 11875093, Gibco™) supplemented with 1% Glutamax (Cat.no 35050061, Gibco™), 10% FBS (Cat.no 04-001-1A, Biological Industries), 5 ng/ml IL-3 (Cat.no 200-03), SCF (Cat.no 300-07), GM-CSF (Cat.no 300-03) & 10 ng/ml G-CSF (Cat.no 300-23), all from PeproTech, for proliferation and myeloid progenitor differentiation followed by a 7-day culture in RPMI , 1% Glutamax, 10% FBS, 1% Penn Strep (Cat.no 03-031-1B, Biological Industries), 10 ng/ml G-CSF for neutrophil differentiation and maturation. Cell counts and viability were evaluated using an Orflo Moxi V and propidium iodide.

Cytospin staining

8 ´ 10⁴ cells at day 15 of differentiation were spun onto Cytoslide microscope slides (ThermoFisher) using Cytospin 4 low speed cytocentrifuge (Thermo Scientific) and stained with Kwik-Diff staining system (MilliporeSigma, eosin/methylene blue) according to manufacturer’s recommendations. Microphotographs were taken on LEITZ LABORLUX S polarizing light microscope at 400X magnification using Nikon DSLR digital camera.

Flow cytometry

Myeloid maturation of CD34⁺ cells was analyzed at day 14 of the differentiation by flow cytometry utilizing antibodies characterizing the neutrophilic lineage. Pacific Blue-CD66b anti-human (Cat no. 305112, Biolegend), APC-CD14 anti-human (Cat no. 130-110-520, Miltenyi Biotec), APC-CD11b anti-human (Cat no. 130-110-554, Miltenyi Biotec) and Pacific Blue-CD15 anti-human (Cat no. 130-113-488, Miltenyi Biotec) were used. Cell debris was gated out by using a zombie yellow viability kit (Cat no. 423103, Biolegend). Doublet discrimination gating was used for exclusion of doublets. Appropriate isotype controls were used to help distinguish between specific and non-specific antibody bindings. For analyzing cell population of interest, quadrant gate analysis method was used for double positive cells selection.

Molecular modelling

A computational analysis of the neutrophil elastase (NE) protein was conducted using the Molecular Operating Environment 2024 (MOE) software [21]. In this study, Mol2 files of the inhibitors and the PDB file of the wild-type NE, obtained from UniProt, were utilized. MOE was employed to simulate the docking of various inhibitors at specific binding sites on the NE protein. These binding sites were computationally selected based on several factors, including site size, steric hindrance, electrostatic repulsion, and amino acid compatibility. For each inhibitor, 50 docking trials were performed in triplicate, resulting in a total of 150 docking attempts per inhibitor across the four tested inhibitors. The stability of each protein-ligand complex was assessed using the free energy of binding (S score in kcal/mol), providing an approximation of binding stability. The accuracy of this S score was also given as a root mean square deviation (RMSD), with scores <2.0 indicating good conformation, scores 2.0-3.0 indicating acceptable conformation, and >3.0 indicating poor conformation.

Neutrophil elastase proteolytic activity inhibition

The NE proteolytic activity in cell lysates from day 14 differentiated CD34⁺ cells exposed and not exposed to the inhibitors was determined using EnzCheck Elastase assay kit from Thermo Fisher Scientific, according to the manufacturer’s recommendations.

Inhibitors

MK0339 NE inhibitor was provided by Merck & Co. (Kenilworth, NJ, USA). The same inhibitor is also commercially available through MedChemExpress (MCE) as DMP-777 (HY-75957). NE inhibitors: Sivelestat (HY-17443), GW311616 (HY-15891), BAY-678 (HY-111457A) with its inactive control, BAY-677 (HY-111457) and DPP1 inhibitor Brensocatib (HY-101056) were purchased from MedChemExpress (Monmouth Junction, NJ).

Patients and ELANE mutations

5 patients previously enrolled in the Severe Congenital Neutropenia International Registry (SCNIR) participated in the study (Table 1).

Inhibitors, concentrations, proteolytic activity inhibition and toxicity studies

All the inhibitors used in this study are small molecule cell-penetrant chemicals with well characterized chemical and biological properties and have been used or investigated by other groups (Table 2).

We investigated the potential toxic effects of these compounds on CD34⁺ cells in dose-response studies. Effects on cell counts and viability of CD34⁺ cells were examined in 48 and 72 h culture experiments at 37°C in a standard CO₂ incubator (Figure 1). Based on the results of these studies, subsequent experiments were performed with inhibitors at a concentration of 1.25 uM for MK0339, 1 uM for Sivelestat, BAY-678, BAY-677, and brensocatib and 10uM for GW311616 unless otherwise specified. It is important to note, that in addition to 4 NE inhibitors we have also used brensocatib, a dipeptidyl peptidase 1 (DPP1) inhibitor, which also inhibits NE but indirectly, through inhibiting DPP1 which is an important factor for NE activation.

NE inhibition

We measured NE proteolytic activity and its inhibition with addition of NE inhibitors in the cell lysate of differentiated granulocytes at day 14 of differentiation by utilizing EnzCheck Elastase assay kit from Thermo Fisher Scientific. We determined that all 4 NE inhibitors have similar performance and are blocking the enzyme’s proteolytic activity by 35% to 38% across the line at the concentrations used in the cell culture in this study (Figure 2).

Myeloid differentiation

We obtained about 2 million cells with at least 95% viability from bone marrow samples of 5 different patients with ELANE neutropenia, as well as from healthy volunteers. Purified CD34⁺ cells were pushed towards myeloid differentiation in the presence and absence of MK0339, sivelestat, BAY-678 and GW311616 NE inhibitors and cultured as described above. On day 14, cells were labeled with CD66b/CD14 and CD11b/CD15 myeloid differentiation surface markers and assessed by flow cytometry. The proportion of CD66b⁺/CD14⁺ and CD11b⁺/CD15⁺ cells was 2-to-3-fold lower in all 4 patient cell lines compared to the volunteers (data not shown), showing impairment of myeloid differentiation consistent with our previous report [19]. Addition of sivelestat, BAY-678 and GW311616 NE and brensocatib inhibitors did not reveal any positive effect on the patient derived cell lines. The addition of MK0339 exhibited a restoration effect on the impaired cell differentiation in the cells from 4 of the 5 patients with an increase of CD66b⁺/CD14⁺ and CD11b⁺/CD15⁺ subsets to up to 3-fold (Figure 3A and 3B). Inhibitor treatment of volunteer cells produced no noticeable differences in the proportions expressing these surface markers. Interestingly, GW311616 had a negative effect on both volunteer and patient derived cells myeloid differentiation; it decreased the proportion of cells of both CD66b⁺/CD14⁺ and CD11b⁺/CD15⁺ subsets.

Microscopic examination utilizing Kwik-Diff staining of the differentiated CD34⁺ cells revealed a typical block of differentiation in the promyelocytic stage of all patient derived cell lines compared to the volunteers, which was partially resorted by the addition of MK0339 inhibitor. There were no effects from sivelestat, BAY-678 and GW311616 NE and brensocatib inhibitors (Figure 3C).

Molecular modelling and NE inhibitors docking analysis

To gain deeper insight into the interaction between the inhibitor and NE protein, including their binding specifics, we conducted molecular modeling using the MOE software platform. Molecular docking analysis of 4 small molecule, cell permeable NE inhibitors (MK0339, sivelestat, GW311616, BAY-678) and inactive control, BAY-677, revealed an alternative binding site for MK0339 inhibitor compared to all others.

MOE software analysis showed that sivelestat, GW311616 and BAY-678 have similar binding sites with NE molecule. Sivelestat, BAY-678, and GW311616 inhibitors bind with amino acid chains consisting of amino acids 139-146 and 233-239 for sivelestat, 140-145 and 235-240 for BAY-678, and 139-146 and 233-239 for GW311616. Interestingly, MK0339 binds at the opposite side of the molecule. Compared to the other NE inhibitors used in this study, MK0339 binds to three distinct amino acid chains corresponding to amino acids at 250-255, 122-130, and 98-104 positions, suggesting a different mechanism of inhibition (Figure 4). The calculated stability score of MK0339 inhibitor with the protein complex was -7.76, -7.20, and -8.21 for mutations G214R, P139L, and w-type NE, respectively (RMSD values of 1.08, 2.30, and 2.60, respectively). Interestingly, sivelestat, GW311616, and BAY-678 had higher values that were very similar to BAY-677, the inactive control (stability values were -6.76, -6.57, and -6.75 for G214R, P139L, and w-type NE respectively) (RMSD values of 2.81, 2.68, and 1.74, respectively). Higher values indicate a less stable inhibitor-protein complex (Table 3).

Severe congenital neutropenia (SCN) is a rare clinical disorder characterized by extremely low blood neutrophil counts that compromise the acute inflammatory response leading to recurrent and sometimes fatal bacterial infections. Bone marrow examinations usually reveal a focused abnormality in myeloid development with otherwise preserved hematopoiesis. Most marrow examinations show "maturation arrest," with a relative abundance of early myeloid precursors, i.e., promyelocytes, but few cells of this lineage beyond the myelocyte stage of development. The majority of these patients have mutations in ELANE, the gene for NE. This potent protease is synthesized during the promyelocyte and early myelocyte stage in neutrophil development. Most research studies indicate that the generation of the mutant enzyme initiates apoptosis of the developing myeloid cells via the unfolded protein response at this mid-point in neutrophil development. Exactly how this occurs is not yet known. We also do not know why some mutations cause mild and others more severe disease, i.e., the wide spectrum in the genotype-phenotype relationships or why some patients have regular variation in blood neutrophil counts, i.e., high to very low, in a cyclic pattern. However, recent studies indicate that knocking out the mutant gene and some inhibitors of the proteolytic activity of NE can correct the cellular abnormality in laboratory models [18-20].

Results of this study provide encouraging data to support the development of NE inhibitors to treat ELANE-associated neutropenia. To extend our previous work, we examined the effects of several cell-permeable NE inhibitors on NE proteolytic activity and myeloid differentiation of CD34⁺ cells derived from patients with ELANE associated neutropenia, and performed molecular docking studies of the inhibitors with wild-type and mutant NE. We found that only one inhibitor, MK0339, restored neutrophil differentiation in patients’ cells and discovered unique properties which may account for this finding. More specifically, we found that the NE inhibitor MK0339 exhibits a unique mechanism of action compared to other inhibitors. This is demonstrated by its distinct binding site and its interaction with three amino acid chains, whereas other NE inhibitors typically bind to only two. The most important clinical finding was that MK0339 had positive effects on promoting myeloid differentiation in patient-derived CD34⁺ cells, similar to our previous report for MK0339 effects on HL60 cells expressing various ELANE mutations [18]. This result was observed across four out of five patient samples, with increase in the expression of key neutrophil markers, CD66b⁺/CD14⁺ and CD11b⁺/CD15⁺. In contrast, other inhibitors, i.e., sivelestat, GW311616, and BAY-678 had weaker binding and failed to improve differentiation, highlighting the specificity and potential therapeutic advantage of MK0339.

The molecular modeling and docking studies support a distinct activity of MK0339. Unlike the other NE inhibitors, which bind to the same site on the NE protein, MK0339 binds to an alternative site. This alternative binding could be responsible for the observed restoration of differentiation. The greater effectiveness of this inhibitor may also be attributed to a more stable inhibitor-protein complex as shown in the modeling studies for NE expressed by mutants G214R and P139L as well as with the w-type protein. Ultimately, higher stability scores associated with MK0339, compared to other inhibitors, may also influence pharmacokinetics and in vivo effectiveness. We were surprised to observe the negative effect of GW311616 on both patient and healthy donor-derived CD34⁺ cells. The reduction in the proportion of differentiated cells suggests that some inhibitors may have detrimental effects on neutrophil development.

Although the other NE inhibitors tested in this study (sivelestat, BAY-678, GW311616) were effective at inhibiting NE proteolytic activity, their lack of impact on myeloid differentiation raises questions about NE inhibition as a simple explanation for the effects that we observed. In support of this uncertainty, we observed that the DPP1 inhibitor brensocatib, which indirectly inhibits NE activation, also failed to demonstrate efficacy in restoring neutrophil differentiation.

This study represents an important step toward the development of alternative therapies for patients with ELANE mutations, particularly those who respond poorly to G-CSF and are at increased risk of progression to myelodysplastic syndrome (MDS) and acute myeloid leukemia (AML). However, there are substantial limitations to this study. The sample size was limited, and there is a great diversity of the mutations causing ELANE associated neutropenia. So, we cannot infer that all, or even most, patients’ cells will respond similarly. Additionally, while our molecular modeling provided valuable insights into inhibitor binding, further structural studies, including crystallography, could provide a more detailed understanding of how MK0339 interacts with mutant NE. Finally, in vivo studies will be necessary to determine the clinical efficacy and safety of MK0339 and any other NE inhibitors.

In conclusion, our data support further investigations of MK0339 as a viable therapeutic candidate for ELANE-associated neutropenia. By restoring myeloid differentiation in patient-derived cells and exhibiting a stable binding profile with mutant NE, MK0339 presents a promising alternative to G-CSF.

The authors have no conflicts of interest and there are no commercial sponsors.

This work was supported by NIH grants 1 RO1 HL 151629, R54 AI 162637 and gifts to the Ella Jewell Foundation.

We thank Dr. Michael Lenaeus, MD, PhD, for his invaluable assistance with molecular docking modeling experiments, and Jamie Leung for her help in preparing this manuscript. We are also deeply grateful to the patients who generously provided the cells used in this study.