Expression of granule-associated proteins in neutrophils from patients with severe congenital neutropenia

To the editor:

In a recent publication in Blood, Donini et al¹  conclude that granulocyte colony-stimulating factor (G-CSF) treatment of severe congenital neutropenia (SCN) results in abnormal expression of granule-associated proteins. We have previously published this concept,²  and showed that neutrophils from patients with SCN lack the antimicrobial peptide LL-37 and have reduced levels of defensins (HNP1–3), but are functionally capable of both phagocytosis and generating oxidative radicals. Our recent results (Figure 1) differ in part to those of Donini et al, and we would like to address this discrepancy.

SCN is a heterogeneous disease that is characterized by bone marrow failure to produce normal numbers of mature neutrophils. The mechanisms behind this arrest involve apoptosis and 2 genes that are frequently mutated, HAX1 and ELA2, in autosomal recessive and autosomal dominant/sporadic SCN, respectively (reviewed in Skokowa et al⁴ ). Without correction of neutrophil levels, the patients are at risk for life-threatening infections. Despite treatment with G-CSF that normally reverses neutropenia, many patients still are at risk for infections, notably in the oral cavity.⁵ 

We have recently analyzed circulating neutrophils from 4 patients with SCN and compared with 3 control participants and 1 patient who underwent bone marrow transplantation. Lysozyme, myeloperoxidase (MPO), and lactoferrin are produced at similar levels, while neutrophils from some but not all patients with SCN have lower levels of gelatinase and HNP1–3 (Figure 1). Compared with the results of Donini et al, our patients are not deficient in lactoferrin, HNP1–3 (mature), or MPO. In addition, Donini et al noted normal levels of ELA2 mRNA, while Skokowa et al⁴  present patients with reduced ELA2 mRNA levels.

What could be the reason for this variation in expression, and is it of functional significance? One important issue may be the responsiveness to G-CSF treatment. Although the authors state that G-CSF reverses neutropenia (which commonly is the case), 2 of their patients in fact respond poorly to the treatment, and thus a proper increase of neutrophil protein synthesis may not occur. A second issue is the cohort chosen. It is estimated that about 60% of patients with SCN harbor mutations in ELA2.⁶  However, more than 40 different sites for mutations are known, and thus the impact of a specific mutation may vary.⁷  Donini et al investigated 3 patients with 2 different ELA2 mutations (G185R, P176fsX182), while our study included 2 patients with ELA2 mutations at other locations (L92H, C26S).³ 

Besides abnormal protein expression, neutrophils from patients with SCN are known to respond poorly to fMLP with diminished chemotaxis and O2⁻ generation,⁸  the latter in accordance to findings by Donini et al. A previous report concludes that neutrophils from patients with SCN kill Staphylococcus aureus,⁹  while Donini et al reported reduced killing of Escherichia coli and Candida albicans.

Taken together, the results presented by Donini et al, interesting as they are, may be restricted to certain ELA2 phenotypes, especially G185R, and not a general phenomenon of patients with SCN. Additional clinical details of the patients presented by Donini et al would have been useful for comparison to earlier work, but also to understand the pathophysiology of SCN.T1