SPI-ing on human B-cell development Free

PU.1 is critical for generating myeloid and lymphoid cells. The first cases of PU.1 haploinsufficiency were reported in 2021. By identifying an additional 33 individuals, as reported in this issue of Blood, Knox et al have broadened the clinical and cellular phenotype of PU.1 deficiency and established the importance of PU.1 in B cells for protection against specific pathogens.¹ 

Hematopoiesis involves the progression of hematopoietic stem cells (HSCs) into lineage-specific precursors, which subsequently commit to lymphoid or myeloid fates, yielding B, T, and natural killer cells or erythrocytes, megakaryocytes, or myeloid cells, with the latter then generating granulocytes, monocytes, macrophages, and dendritic cells (DCs).² HSC fate specification is governed by transcription factors (TFs), which promote or repress commitment to specific leukocyte lineages.² PU.1, encoded by SPI1, is an ETS family TF highly expressed by HSC, myeloid cells, and lymphocytes.^3,4 Germ line targeting of Spi1 in mice severely disrupts hematopoiesis, evidenced by a loss of all leukocyte lineages (B cells, T cells, granulocytes, macrophages).^3,4 Strikingly, deletion of Spi1 from progenitor B cells does not affect subsequent B-cell development due to compensatory functions of Spi-B, a related TF. Indeed, deletion of both Spi1 and Spib from murine B-cell progenitors blocks development at the pre-B-cell stage.^3,4 PU.1 deficiency in mice also causes various hematological malignancies, such as B-cell acute lymphocytic leukemia and acute myeloid leukemia.^4,5 Thus, PU.1 is essential for generating lymphoid and myeloid precursors, but is redundant for B-cell development from B-cell precursors. PU.1 can also function as a tumor suppressor in murine myeloid and lymphoid cells.

Inborn errors of immunity (IEI) have shed significant light on nonredundant molecular requirements for the generation, maintenance, and function of human immune cells.^6,7 Studies of these rare individuals have defined TFs and signaling pathways necessary for lymphocyte development and differentiation and led to therapies for several diseases.^6,7 Importantly, IEI also define fundamental differences between humans and mice, which has relevance for understanding not only basic mechanisms of human immunology but also disease pathogenesis.^6,7 In 2021, Le Coz and colleagues reported 6 individuals with heterozygous variants in SPI1 causing an IEI characterized by frank B-cell deficiency, agammaglobulinemia, and recurrent sinopulmonary bacterial infections.⁸ All variants affected PU.1 expression, nuclear localization, or binding to target genes; however, mutant PU.1 had no effect on wild-type (WT) PU.1 function, indicating that disease resulted from PU.1 haploinsufficiency.⁸ 

As with all discoveries, it is critical to confirm and extend the clinical and cellular phenotypes of novel IEI by identifying additional individuals with mutations in the causative gene. Here, Knox and colleagues have undertaken a large functional screen of SPI1 variants identified in molecularly undiagnosed patients with agammaglobulinemia, as well as public and institutional genomic databases. Starting with 134 SPI1 variants, covering most functional domains of PU.1, 14 encoded an unstable protein, and a further 11 were loss of function (LOF), based on in vitro transcriptional activity. Consistent with their original study,⁸ none of the novel variants interfered with WT PU.1 transcriptional activity, confirming loss of expression or LOF from 1 SPI1 allele is sufficient to cause disease (ie, haploinsufficiency). Clinical information was available for 22 individuals from 21 families bearing 1 of 20 pathogenic variants. All presented with recurrent bacterial respiratory infections, agammaglobulinemia, and severe deficits in circulating B cells and conventional and plasmacytoid DCs; all other leukocyte populations examined were unaffected, at least numerically. Many patients also experienced enteroviral infections and noninfectious complications (gastrointestinal disease, autoimmunity, cognitive impairment). Remarkably, familial genotyping identified 11 additional individuals carrying the pathogenic SPI1 variant: 6 were unaffected, consistent with normal B-cell numbers and serum immunoglobulin levels, and 5 had mildly impaired humoral immunity, for example, reduced B-cell numbers, hypogammaglobulinemia, and immunoglobulin A deficiency. Overall, it was concluded that PU.1 haploinsufficiency causes immune deficiency with a clinical penetrance of ∼80% (27 of 33 cases). However, it was unclear whether the 5 individuals with features of immunodeficiency experienced mild disease or were also clinically unaffected. Thus, it is possible that penetrance, at least in terms of the most severe clinical features, is lower, at ∼66%. Interestingly, incomplete penetrance could not be explained by different variants, as age at disease onset in individuals with null SPI1 mutations ranged from <6 months to the fourth decade of life, even for families harboring the same variant.

This study highlights the approaches necessary to demonstrate pathogenicity of candidate genetic variants, including the requirement for robust in vitro functional testing, and the importance of pursuing family genetics and medical histories to determine genotype and phenotype relationships and disease penetrance. This resulted in several salient observations regarding human genetics and disease, pathogen-specific immunity, and mouse-human similarities and differences. First, although in silico tools can predict the impact of some variants, it was notable that beyond the SPI1 LOF variants, >20 benign variants had combined annotation dependent depletion scores exceeding 25, underscoring the need for experimental screening. Second, the infectious phenotype of PU.1 haploinsufficiency was more reminiscent of B-cell deficiency and agammaglobulinemia (due to mutations in BTK, BLNK, IGHM, etc)⁷ than IEI characterized by a lack of DCs and other immune cells (eg, GATA2, SPPL2A, IRF8 mutations), which cause mycobacterial and some viral (human papillomavirus) infections.⁶ Thus, B-cell-mediated immunity is required for protection against sinopulmonary bacterial infections, with DCs being less critical. Third, as mice heterozygous for Spi1 or engineered to delete Spi1 in B-lineage cells generate normal numbers of B cells,^3,5,9 it was surprising that heterozygous germ line SPI1 variants have such a profound effect on human B-cell development. This suggests SPI-B may not compensate for reduced PU.1 expression and questions the shared and unique role(s) of PU.1 and SPI-B in mice and humans.

Studies on IEI highlight the power and elegance of establishing the impact of single-gene defects on basic, clinical, and translational immunology. IEI also provide opportunities to further illuminate specific functions of key genes in human immunity, for instance, determining the role of PU.1 in human CD4⁺ T cells, particularly Th9, and tumor suppression,^4,5 elucidating potential differences between thresholds of PU.1 expression and function required for B-cell development in mice and humans, and leveraging these findings to improve diagnosis, treatment, and outcomes for individuals with PU.1 deficiency.

Conflict-of-interest disclosure: S.G.T. declares no competing financial interests.