Germline IKAROS dimerization haploinsufficiency causes hematologic cytopenias and malignancies

Abstract

IKAROS is a transcription factor forming homo- and heterodimers and regulating lymphocyte development and function. Germline mutations affecting the IKAROS N-terminal DNA binding domain, acting in a haploinsufficient or dominant-negative manner, cause immunodeficiency. Herein, we describe 4 germline heterozygous IKAROS variants affecting its C-terminal dimerization domain, via haploinsufficiency, in 4 unrelated families. Index patients presented with hematologic disease consisting of cytopenias (thrombocytopenia, anemia, neutropenia)/Evans syndrome and malignancies (T-cell acute lymphoblastic leukemia, Burkitt lymphoma). These dimerization defective mutants disrupt homo- and heterodimerization in a complete or partial manner, but they do not affect the wild-type allele function. Moreover, they alter key mechanisms of IKAROS gene regulation, including sumoylation, protein stability, and the recruitment of the nucleosome remodeling and deacetylase complex; none affected in N-terminal DNA binding defects. These C-terminal dimerization mutations are largely associated with hematologic disorders, display dimerization haploinsufficiency and incomplete clinical penetrance, and differ from previously reported allelic variants in their mechanism of action. Dimerization mutants contribute to the growing spectrum of IKAROS-associated diseases displaying a genotype-phenotype correlation.

Graphical abstract

Figure 1.

Genetics and pedigrees of families with IKZF1 mutations. (A) Segregation of the IKZF1 mutations and clinical and immunological phenotypes. Squares and circles indicate male and female family members, respectively. Question mark indicates an unscreened individual. (B) Schematic presentation of the structure of IKAROS isoform 1 (NM_006060). Dark gray indicates ZFs, and arrows indicate the site of the mutations. Numbers indicate amino acid location. Previously reported germline IKAROS mutations in CVID or CID patients are indicated below the domain. Blue color indicates haploinsufficient mutations (DNA binding–defective mutations and loss of 1 IKZF1 allele); DN mutations are shown in red; dimerization defective–mutations are shown in green; and gray color indicates the variants whose functions are unclear. (C) Sequence conservation of ZF5 and ZF6 in IKAROS. The mutated amino acid sites found in families C and D (C467 and R502, respectively) are enclosed in the green box.

Figure 2.

IKAROS transcripts and protein expression in patients. (A) The chromatograms show the sequence of the indicated heterozygous mutation in complementary DNA prepared from peripheral blood mononuclear cells (PBMCs). WT indicates WT of IKAROS’ reference sequence. Arrows indicate the sites of mutations. (B) IKAROS protein expression levels were tested in T and B cells from PBMCs from index patients and a paired HD control. (C) IKAROS expression in nuclear extracts of T-cell blasts. Triangles indicate R213* and S427* mutant proteins. The lower blots were overexposed to visualize the low levels of truncated IKAROS protein (R213* and S427*). Data are representative of 2 independent experiments. (D-E) HEK293T cells were transfected with a vector expressing IKAROS WT or mutant. After 24 hours, cells were treated with cycloheximide (CHX; 20 μg/mL) for an additional 24 hours. (D) Whole-cell lysates were prepared and analyzed by immunoblot analysis using hemagglutinin (HA) antibodies to test IKAROS expression. Vinculin was used as a loading control. (E) The relative IKAROS protein stability was calculated by dividing the CHX-treated sample by the untreated sample (×100) after normalization with vinculin to show the remaining protein amount after CHX treatment. Data are mean ± standard error of the mean of 3 or 4 independent experiments. Data (WT vs each mutant) were analyzed using GraphPad Prism software. *P < .05, unpaired Student t test.

Figure 3.

Impaired interaction of mutant protein with the IKAROS family. (A-C) 293T cells were transfected with Flag-tagged WT IKAROS, WT AIOLOS, WT HELIOS, or hemagglutinin (HA)-tagged WT IKAROS or mutants (R213*, S427*, C467R, R502L, Y462* [a laboratory-generated mutant with deletion of ZF5 and ZF6]). Immunoprecipitation (IP) was performed using anti-Flag antibodies. Western blot analysis of the IP samples with anti-HA and anti-Flag or AIOLOS or HELIOS antibodies is shown. Five percent of the total volumes of the whole-cellular lysates used for each IP reaction were loaded as input controls. Data are representative of 3 independent experiments. Densitometry analysis of immunoprecipitated HA-IKAROS blots was performed using ImageStudio. EV, empty vector control.

Figure 4.

Pericentromeric targeting of the mutant IKAROS. NIH3T3 cells were transfected with a hemagglutinin (HA)-tagged WT or mutant expression vector alone (A) or together (B) with Flag-tagged WT IKAROS. Cells were labeled with anti-HA or together with anti-Flag antibodies, followed by Alexa Fluor 488–conjugated (green) and Alexa Fluor 568–conjugated (red) secondary antibodies. Cells were visualized using a ZOE fluorescence microscope (original magnification, ×175). Data are representative of 3 independent experiments. HI and DN indicate DNA binding–defective haploinsufficient mutation and DN mutation, respectively. Scale bars, 25 μm.

Figure 5.

DNA binding and transcription activity of the mutant IKAROS protein. (A-B) EMSAs were performed using nuclear extracts from HEK293T cells transfected with the indicated IKAROS mutation alone (A) or together with WT IKAROS (B). Numbers indicate the ratio of WT and mutant IKAROS DNA used for the cotransfection. The nuclear extracts were allowed to bind to 2 IKAROS probes: IK-bs1, an IKAROS consensus binding sequence, and γ-Sat 8, a sequence from the pericentromeric region of human chromosome 8. IKAROS-containing complexes are indicated with arrows. Triangles indicate mutant IKAROS protein binding to the DNA as a monomer. Data are representative of 3 independent experiments. (C) Four repeats of IKBS1 were inserted into the pGL4.11 vector (pGL4.11-IKBS1) and cotransfected with pcDNA3-HA IKAROS WT or mutants and pRL-TK (Renilla luciferase) as indicated. Twenty hours later, cells were lysed, and luciferase activity was measured using the Dual-Luciferase Reporter Assay System. Firefly luciferase activity was normalized to Renilla luciferase activity, and the results were normalized to the empty vector (EV) control. Each experiment was carried out in duplicate. Data are mean ± standard error of the mean from 3 independent experiments.

Figure 6.

Sumoylation and HDAC1 binding of IKAROS WT and mutant proteins. (A) HEK293T cells were cotransfected with GFP-sumo1 and hemagglutinin (HA)-tagged IKAROS WT or mutant. Protein extracts were immunoprecipitated with anti-HA antibodies and probed with anti-Sumo1 antibodies to see the sumoylation, as well as with anti-HA antibodies to see the immunoprecipitated IKAROS. Arrows near 100 to 200 kDa indicate single or multiple sumoylated IKAROS protein. (B) HEK293T cells were cotransfected with Flag-tagged IKAROS WT or mutant and HA-tagged HDAC1. Protein lysates were immunoprecipitated with antibodies to the Flag epitope. Western blot analysis of the immunoprecipitation (IP) samples with anti-HA (for HDAC1) and anti-Flag (for IKAROS) antibodies is shown. Five percent of the total volume of the whole-cellular lysates used for IP reactions was loaded as input controls. Data shown are representative of 3 independent experiments.

Figure 7.

Correlation analysis of targeted RNASeq data on T-cell blasts from HD controls (n = 5) and patients with DD, HI, or DN mutations. (A) RNASeq was performed using RNA extracted from enriched blasted T cells. Scale bar indicates Pearson correlation coefficient for variance stabilized data for differentially expressed messenger RNA. (B) Venn diagram analysis comparing lists of differentially expressed genes for each IKAROS study group (ie, DD [n = 4], HI (n = 1), DN (n = 1)] vs HD controls (n = 5). (C) Table of differentially expressed genes for the DD study group only as well as differentially expressed genes that are common across groups. Genes overexpressed compared with HD controls are in red text; genes underexpressed compared with HD controls are in blue text. (D) IPA Diseases and Functions Analysis heat map showing a comparison of the predicted effected cellular processes and biological functions based on gene expression. The activation z score represents the bias in gene regulation that predicts whether the biological function exists in an activated (orange) or inactivated (blue) state.