GATA2 Genetic Variant-Induced Pathogenesis: Loss-of-Function and Gain-of-Function Mechanisms

Genetic variation in genes encoding hematopoietic transcription factors disrupts physiological mechanisms, thus causing human transcriptopathies that impact hematopoiesis and broader processes. Pathogenic variants in GATA2, encoding a master regulator of hematopoiesis, create a predisposition to immunodeficiency, myelodysplastic syndromes, and acute myeloid leukemia (Calvo et al., 2023). Although certain disease-linked GATA2 variants were considered to be loss-of-function, our initial work provided evidence that the R307W variant has activity exceeding that of GATA2 to induce granulocytic colonies from primary Lin^- cells (Katsumura et al., 2018). However, the relative importance of gain-of-function versus loss-of-function activities to GATA2-linked pathogenic mechanisms is unresolved.

We developed genetic rescue systems using primary and immortalized progenitor cells from mice lacking the Gata2 -77 enhancer, which is critical for Gata2 expression in mouse and man, to quantify GATA2 variant activities (Hi-77^-/- cells) (Jung et al., 2023; Katsumura et al., 2024). While -77^-/- Lin^- cells preferentially undergo monocytic differentiation, at levels resembling endogenous GATA2, T354M and R307W decreased monocytic differentiation and induced granulocytes, including eosinophils. Contrasting with GATA2, the variants failed to strongly induce mast cell markers. RNA-Seq analysis revealed that T354M and R307W are loss-of-function with select gene expression readouts, yet they retained activity to regulate a GATA2 target gene cohort, and a distinct gene cohort that was not GATA2-regulated. Using Hi-77^-/- cells, we dissected mechanisms by which T354M regulates transcription. GATA2 and T354M occupied chromatin at loci containing WGATAR motifs. GATA2 and T354M activity to regulate target genes required the DNA binding C-finger. As CRISPR-mediated deletion of enhancers at Ms4a3, Cebpe, and Cpa3 loci abrogated T354M-mediated activation of these genes, T354M retains the capacity to occupy WGATAR motif-containing chromatin at select target genes. Surprisingly, T354M disproportionately required the N-finger versus GATA2. While disrupting the N-finger (C295A) did not affect GATA2-mediated activation of Cpa3 and Hdc and granulocytic differentiation, T354M/C295A failed to regulate these genes and exhibited reduced differentiation-inducing activity. Thus, despite the deleterious impact of T354M on the DNA-binding C-finger, the N-finger endows T354M with context-dependent transcriptional regulatory activity.

The rescue system was utilized to discover transcription factors and coregulators that cooperate with GATA2 and T354M. By using CRISPR-mediated depletion, we demonstrated that TAL1 and SMARCD2 contribute to GATA2- and T354M-mediated regulation of select target genes. Deletion of Cebpe+6 enhancer abrogated GATA2- and T354M-mediated activation of Cebpe and impaired GATA2- and T354M-mediated induction of Prg2, Prg3, Epx, and Ms4a3. Expressing C/EBPε in -77^-/- Lin^- cells partially normalized the excessive monocytic differentiation of -77^-/- Lin^- cells. Thus, multiple transcription factors and coregulators cooperate with GATA2 and T354M, and C/EBPε forms a feed-forward loop with GATA2 and T354M to regulate other target genes and myeloid differentiation.

We asked if the principles that emerged from T354M analyses can be extrapolated to other GATA2 clinical variants. P304H, A318T, R330Q, T355Δ, L359V, R362P, R362Q, and 9aa-ins were analyzed. Among the C-finger variants, R362Q closely resembled T354M. N-finger variants shared activity to induce Ces1d expression. SMARCD2 was required for Ctsg activation by C-finger variants and Ces1d activation by N-finger variants. These results indicate that certain mechanistic principles are shared between GATA2 and disease variants, whereas others differ.

This study unveiled mechanisms by which GATA2 disease variants lose certain molecular and cellular activities while retaining others. We propose that loss-of-function and gain-of-function consequences of genetic variation, which may occur concomitantly, yield pathogenesis by disrupting the establishment and/or maintenance of tightly controlled networks required for normal hematopoiesis.