Remember me
Research ArticleImmunologyInflammation
Open Access | 
10.1172/jci.insight.198703
1Pulmonary Center and Section of Pulmonary, Allergy, Sleep and Critical Care, Department of Medicine, Boston University Chobanian & Avedisian School of Medicine, Boston, Massachusetts, USA.
2Center for Regenerative Medicine, Boston University and Boston Medical Center, Boston, Massachusetts, USA.
3Department of Biological Sciences, Fordham University, Bronx, New York, USA.
4Department of Pathology & Cell Biology, Columbia University Irving Medical Center, New York, New York, USA.
5Division of Clinical Immunology, Departments of Medicine and Pediatrics, Icahn School of Medicine at Mount Sinai, New York, New York, USA.
Address correspondence to: Paul J. Maglione, Pulmonary Center, Boston University Medical Campus, 72 E Concord St., R304, Boston, Massachusetts, 02118, USA. Phone: 617.358.6587; Email: pmaglion@bu.edu.
Find articles by Hayes, K. in: PubMed | Google Scholar
1Pulmonary Center and Section of Pulmonary, Allergy, Sleep and Critical Care, Department of Medicine, Boston University Chobanian & Avedisian School of Medicine, Boston, Massachusetts, USA.
2Center for Regenerative Medicine, Boston University and Boston Medical Center, Boston, Massachusetts, USA.
3Department of Biological Sciences, Fordham University, Bronx, New York, USA.
4Department of Pathology & Cell Biology, Columbia University Irving Medical Center, New York, New York, USA.
5Division of Clinical Immunology, Departments of Medicine and Pediatrics, Icahn School of Medicine at Mount Sinai, New York, New York, USA.
Address correspondence to: Paul J. Maglione, Pulmonary Center, Boston University Medical Campus, 72 E Concord St., R304, Boston, Massachusetts, 02118, USA. Phone: 617.358.6587; Email: pmaglion@bu.edu.
Find articles by Boldt, K. in: PubMed | Google Scholar
1Pulmonary Center and Section of Pulmonary, Allergy, Sleep and Critical Care, Department of Medicine, Boston University Chobanian & Avedisian School of Medicine, Boston, Massachusetts, USA.
2Center for Regenerative Medicine, Boston University and Boston Medical Center, Boston, Massachusetts, USA.
3Department of Biological Sciences, Fordham University, Bronx, New York, USA.
4Department of Pathology & Cell Biology, Columbia University Irving Medical Center, New York, New York, USA.
5Division of Clinical Immunology, Departments of Medicine and Pediatrics, Icahn School of Medicine at Mount Sinai, New York, New York, USA.
Address correspondence to: Paul J. Maglione, Pulmonary Center, Boston University Medical Campus, 72 E Concord St., R304, Boston, Massachusetts, 02118, USA. Phone: 617.358.6587; Email: pmaglion@bu.edu.
Find articles by Schnorr, P. in: PubMed | Google Scholar
1Pulmonary Center and Section of Pulmonary, Allergy, Sleep and Critical Care, Department of Medicine, Boston University Chobanian & Avedisian School of Medicine, Boston, Massachusetts, USA.
2Center for Regenerative Medicine, Boston University and Boston Medical Center, Boston, Massachusetts, USA.
3Department of Biological Sciences, Fordham University, Bronx, New York, USA.
4Department of Pathology & Cell Biology, Columbia University Irving Medical Center, New York, New York, USA.
5Division of Clinical Immunology, Departments of Medicine and Pediatrics, Icahn School of Medicine at Mount Sinai, New York, New York, USA.
Address correspondence to: Paul J. Maglione, Pulmonary Center, Boston University Medical Campus, 72 E Concord St., R304, Boston, Massachusetts, 02118, USA. Phone: 617.358.6587; Email: pmaglion@bu.edu.
Find articles by
Bawa, P.
in:
PubMed
|
Google Scholar
|
1Pulmonary Center and Section of Pulmonary, Allergy, Sleep and Critical Care, Department of Medicine, Boston University Chobanian & Avedisian School of Medicine, Boston, Massachusetts, USA.
2Center for Regenerative Medicine, Boston University and Boston Medical Center, Boston, Massachusetts, USA.
3Department of Biological Sciences, Fordham University, Bronx, New York, USA.
4Department of Pathology & Cell Biology, Columbia University Irving Medical Center, New York, New York, USA.
5Division of Clinical Immunology, Departments of Medicine and Pediatrics, Icahn School of Medicine at Mount Sinai, New York, New York, USA.
Address correspondence to: Paul J. Maglione, Pulmonary Center, Boston University Medical Campus, 72 E Concord St., R304, Boston, Massachusetts, 02118, USA. Phone: 617.358.6587; Email: pmaglion@bu.edu.
Find articles by Abyazi, M. in: PubMed | Google Scholar
1Pulmonary Center and Section of Pulmonary, Allergy, Sleep and Critical Care, Department of Medicine, Boston University Chobanian & Avedisian School of Medicine, Boston, Massachusetts, USA.
2Center for Regenerative Medicine, Boston University and Boston Medical Center, Boston, Massachusetts, USA.
3Department of Biological Sciences, Fordham University, Bronx, New York, USA.
4Department of Pathology & Cell Biology, Columbia University Irving Medical Center, New York, New York, USA.
5Division of Clinical Immunology, Departments of Medicine and Pediatrics, Icahn School of Medicine at Mount Sinai, New York, New York, USA.
Address correspondence to: Paul J. Maglione, Pulmonary Center, Boston University Medical Campus, 72 E Concord St., R304, Boston, Massachusetts, 02118, USA. Phone: 617.358.6587; Email: pmaglion@bu.edu.
Find articles by Ware, M. in: PubMed | Google Scholar
1Pulmonary Center and Section of Pulmonary, Allergy, Sleep and Critical Care, Department of Medicine, Boston University Chobanian & Avedisian School of Medicine, Boston, Massachusetts, USA.
2Center for Regenerative Medicine, Boston University and Boston Medical Center, Boston, Massachusetts, USA.
3Department of Biological Sciences, Fordham University, Bronx, New York, USA.
4Department of Pathology & Cell Biology, Columbia University Irving Medical Center, New York, New York, USA.
5Division of Clinical Immunology, Departments of Medicine and Pediatrics, Icahn School of Medicine at Mount Sinai, New York, New York, USA.
Address correspondence to: Paul J. Maglione, Pulmonary Center, Boston University Medical Campus, 72 E Concord St., R304, Boston, Massachusetts, 02118, USA. Phone: 617.358.6587; Email: pmaglion@bu.edu.
Find articles by Gyimesi, G. in: PubMed | Google Scholar
1Pulmonary Center and Section of Pulmonary, Allergy, Sleep and Critical Care, Department of Medicine, Boston University Chobanian & Avedisian School of Medicine, Boston, Massachusetts, USA.
2Center for Regenerative Medicine, Boston University and Boston Medical Center, Boston, Massachusetts, USA.
3Department of Biological Sciences, Fordham University, Bronx, New York, USA.
4Department of Pathology & Cell Biology, Columbia University Irving Medical Center, New York, New York, USA.
5Division of Clinical Immunology, Departments of Medicine and Pediatrics, Icahn School of Medicine at Mount Sinai, New York, New York, USA.
Address correspondence to: Paul J. Maglione, Pulmonary Center, Boston University Medical Campus, 72 E Concord St., R304, Boston, Massachusetts, 02118, USA. Phone: 617.358.6587; Email: pmaglion@bu.edu.
Find articles by James, M. in: PubMed | Google Scholar
1Pulmonary Center and Section of Pulmonary, Allergy, Sleep and Critical Care, Department of Medicine, Boston University Chobanian & Avedisian School of Medicine, Boston, Massachusetts, USA.
2Center for Regenerative Medicine, Boston University and Boston Medical Center, Boston, Massachusetts, USA.
3Department of Biological Sciences, Fordham University, Bronx, New York, USA.
4Department of Pathology & Cell Biology, Columbia University Irving Medical Center, New York, New York, USA.
5Division of Clinical Immunology, Departments of Medicine and Pediatrics, Icahn School of Medicine at Mount Sinai, New York, New York, USA.
Address correspondence to: Paul J. Maglione, Pulmonary Center, Boston University Medical Campus, 72 E Concord St., R304, Boston, Massachusetts, 02118, USA. Phone: 617.358.6587; Email: pmaglion@bu.edu.
Find articles by Ko, H. in: PubMed | Google Scholar
1Pulmonary Center and Section of Pulmonary, Allergy, Sleep and Critical Care, Department of Medicine, Boston University Chobanian & Avedisian School of Medicine, Boston, Massachusetts, USA.
2Center for Regenerative Medicine, Boston University and Boston Medical Center, Boston, Massachusetts, USA.
3Department of Biological Sciences, Fordham University, Bronx, New York, USA.
4Department of Pathology & Cell Biology, Columbia University Irving Medical Center, New York, New York, USA.
5Division of Clinical Immunology, Departments of Medicine and Pediatrics, Icahn School of Medicine at Mount Sinai, New York, New York, USA.
Address correspondence to: Paul J. Maglione, Pulmonary Center, Boston University Medical Campus, 72 E Concord St., R304, Boston, Massachusetts, 02118, USA. Phone: 617.358.6587; Email: pmaglion@bu.edu.
Find articles by Cunningham-Rundles, C. in: PubMed | Google Scholar
1Pulmonary Center and Section of Pulmonary, Allergy, Sleep and Critical Care, Department of Medicine, Boston University Chobanian & Avedisian School of Medicine, Boston, Massachusetts, USA.
2Center for Regenerative Medicine, Boston University and Boston Medical Center, Boston, Massachusetts, USA.
3Department of Biological Sciences, Fordham University, Bronx, New York, USA.
4Department of Pathology & Cell Biology, Columbia University Irving Medical Center, New York, New York, USA.
5Division of Clinical Immunology, Departments of Medicine and Pediatrics, Icahn School of Medicine at Mount Sinai, New York, New York, USA.
Address correspondence to: Paul J. Maglione, Pulmonary Center, Boston University Medical Campus, 72 E Concord St., R304, Boston, Massachusetts, 02118, USA. Phone: 617.358.6587; Email: pmaglion@bu.edu.
Find articles by
Mizgerd, J.
in:
PubMed
|
Google Scholar
|
1Pulmonary Center and Section of Pulmonary, Allergy, Sleep and Critical Care, Department of Medicine, Boston University Chobanian & Avedisian School of Medicine, Boston, Massachusetts, USA.
2Center for Regenerative Medicine, Boston University and Boston Medical Center, Boston, Massachusetts, USA.
3Department of Biological Sciences, Fordham University, Bronx, New York, USA.
4Department of Pathology & Cell Biology, Columbia University Irving Medical Center, New York, New York, USA.
5Division of Clinical Immunology, Departments of Medicine and Pediatrics, Icahn School of Medicine at Mount Sinai, New York, New York, USA.
Address correspondence to: Paul J. Maglione, Pulmonary Center, Boston University Medical Campus, 72 E Concord St., R304, Boston, Massachusetts, 02118, USA. Phone: 617.358.6587; Email: pmaglion@bu.edu.
Find articles by Mostoslavsky, G. in: PubMed | Google Scholar
1Pulmonary Center and Section of Pulmonary, Allergy, Sleep and Critical Care, Department of Medicine, Boston University Chobanian & Avedisian School of Medicine, Boston, Massachusetts, USA.
2Center for Regenerative Medicine, Boston University and Boston Medical Center, Boston, Massachusetts, USA.
3Department of Biological Sciences, Fordham University, Bronx, New York, USA.
4Department of Pathology & Cell Biology, Columbia University Irving Medical Center, New York, New York, USA.
5Division of Clinical Immunology, Departments of Medicine and Pediatrics, Icahn School of Medicine at Mount Sinai, New York, New York, USA.
Address correspondence to: Paul J. Maglione, Pulmonary Center, Boston University Medical Campus, 72 E Concord St., R304, Boston, Massachusetts, 02118, USA. Phone: 617.358.6587; Email: pmaglion@bu.edu.
Find articles by
Kotton, D.
in:
PubMed
|
Google Scholar
|
1Pulmonary Center and Section of Pulmonary, Allergy, Sleep and Critical Care, Department of Medicine, Boston University Chobanian & Avedisian School of Medicine, Boston, Massachusetts, USA.
2Center for Regenerative Medicine, Boston University and Boston Medical Center, Boston, Massachusetts, USA.
3Department of Biological Sciences, Fordham University, Bronx, New York, USA.
4Department of Pathology & Cell Biology, Columbia University Irving Medical Center, New York, New York, USA.
5Division of Clinical Immunology, Departments of Medicine and Pediatrics, Icahn School of Medicine at Mount Sinai, New York, New York, USA.
Address correspondence to: Paul J. Maglione, Pulmonary Center, Boston University Medical Campus, 72 E Concord St., R304, Boston, Massachusetts, 02118, USA. Phone: 617.358.6587; Email: pmaglion@bu.edu.
Find articles by
Maglione, P.
in:
PubMed
|
Google Scholar
|
Published March 23, 2026 - More info
Published in Volume 11, Issue 6 on March 23, 2026
AbstractCommon variable immunodeficiency (CVID) is the most prevalent symptomatic primary antibody deficiency. For unclear reasons, inflammatory complications, like gastrointestinal (GI) disease, occur in ~50% of CVID cases, worsening morbidity and mortality. NFKB1 variants are among the most frequent genetic variants in CVID. While effect of NFKB1 variants is not well understood, we previously found frameshift heterozygous NFKB1 variants to increase cytokines, monocytes, and inflammatory complications in CVID. In this report, we used induced pluripotent stem cell–derived (iPSC-derived) monocytes (iMONOs) with CRISPR/Cas9-mediated gene editing to study a heterozygous NFKB1 frameshift found in a patient with CVID with severe GI disease. The heterozygous NFKB1 variant similarly reduced NFKB1 protein in CVID patient– and healthy donor–derived iMONOs, but elevated LPS-induced IL-1β release and expression of inflammatory genes, including IL1B, IL6, TNF, and neutrophil chemoattractants, only in CVID patient iMONOs. CVID patient iMONOs also had elevations of IL-12, CCL4, and CCL12 unaffected by presence or absence of the NFKB1 variant. TNF antagonism improved the patient’s GI disease, diminishing neutrophilic gastritis, circulating neutrophils, and the neutrophil chemoattractant CXCL1 in the blood. While the biology remains complex, our approach found heterozygous NFKB1 variant–induced inflammatory changes intensified in CVID iMONOs, corresponding with clinical response to TNF antagonism.
Graphical Abstract
Introduction
Common variable immunodeficiency (CVID) is the most clinically significant of the predominantly antibody deficiency disorders, which collectively are the most numerous inborn errors of immunity overall (1). CVID affects an estimated 1:25,000, and is defined by reduced levels of IgG, IgA, and/or IgM with impaired specific antibody response (2, 3). CVID clinical course varies, with all having increasing susceptibility to infection but about half also developing noninfectious complications, such as inflammatory gastrointestinal (GI) disease, for unclear reasons (4). As noninfectious complications significantly worsen morbidity and mortality of CVID despite currently available treatments, better understanding of underlying pathogenesis is a critical unmet need in CVID that could provide much-needed advancement of therapy (5–8).
Heterozygous NFKB1 variants are among the most frequent CVID-associated genetic findings (9). NFKB1 encodes both the full-length protein p105 and its truncated cleavage product p50, proteins that function in both activation and regulation of canonical NF-κB signaling (10, 11). Notably, NFKB1 variants are more common in those with inflammatory disease than symptomatic antibody deficiency, indicating that they may promote inflammatory disease in CVID (12). Elevated cytokine production was reported in a patient with CVID with an NFKB1 variant and severe GI disease, and a case series study found NFKB1 variants associated with inflammasome activation and other inflammatory dysregulation, though the participants in that report did not satisfy CVID diagnostic criteria (13, 14). Recently, heterozygous NFKB1 variants, particularly those that are frameshift or nonsense, were associated with increased inflammatory complications in a CVID cohort, together with elevated circulating monocytes and increased plasma cytokines (15). In vitro overexpression systems have shown that CVID-associated NFKB1 variants cause protein haploinsufficiency (16, 17). However, in vitro overexpression may not model the variant heterozygosity occurring in patients or provide insight into downstream inflammatory effects that may shape CVID manifestations.
Induced pluripotent stem cells (iPSCs) can be generated from peripheral blood, serving as a self-renewing reservoir to derive cells with donor-specific genetics (18, 19). CRISPR/Cas9-mediated gene editing can introduce or correct genetic variants in iPSCs, which then can be differentiated into disease-relevant cells of interest for further investigation (20–22). An iPSC-based approach offers several advantages over prior studies. First, it can isolate effect of heterozygous NFKB1 variants from other genetic differences through use of targeted gene editing of alternate genetic backgrounds. Second, it achieves heterozygous, endogenous expression as occurs in patients. Third, it can help distinguish genetic effects from epigenetic and chronic inflammatory changes that obscure the basis of an inflammatory phenotype in primary cells taken from patients with active disease because iPSCs have reduced, though not necessarily absent, epigenetic changes (23). Lastly, we are able to isolate a specific cell type to study in the context of CVID inflammatory disease.
Monocyte activation and expansion are features of patients with CVID with inflammatory disease complications, including those with heterozygous NFKB1 variants (15, 24–27). Monocytes are also a significant contributor to cytokine elevation in CVID; phospho-STAT1+ monocytes are a primary source of B cell activating factor elevation found increased in these patients, possibly downstream of the elevated IFN-γ and STAT1 expression that define these patients (28). Given this precedent for monocyte dysregulation in CVID inflammatory complications, we used iPSC-derived monocytes (iMONOs) as our cellular model to elucidate effect of a heterozygous NFKB1 variant.
Using iMONOs together with gene editing, we aimed to elucidate the effect of a heterozygous frameshift NFKB1 variant found in a patient with CVID with a severe inflammatory course, specifically a form of refractory GI disease that had profound clinical response to TNF antagonism. To our knowledge, this is the first application of iPSCs in the study of CVID. We explored effects of the heterozygous NFKB1 variant in different genetic backgrounds, including the effect of gene correction upon immune dysregulation. This approach made it possible to delineate effect of a heterozygous frameshift NFKB1 variant from the other genetic factors in a patient with CVID with severe inflammatory disease and provide rationale for the therapeutic response to TNF antagonism. Our results paint a more complex picture of the relationship between NFKB1 variants and inflammatory disease in CVID than previously thought, exposing influence of the broader genetic background in which the variant is present and demonstrating therapeutic potential of targeting NF-κB–related immune dysregulation.
ResultsSevere and refractory GI disease in a patient with CVID with 1377delT NFKB1 variant. The focus of our report is a 26-year-old female patient (referred to as patient CA01) who was diagnosed with CVID at age 8 years and was receiving immunoglobulin replacement therapy. While her diagnostic IgG value was not available, her current IgA and IgM values were both below the level of detection for the clinical laboratory. Peripheral leukocyte counts demonstrated elevated neutrophils at 11,500 cells/μL (reference range, 1800–7000 cells /μL), monocyte count of 900 cells /μL (200–900 cells /μL), CD4+ T cell count of 1129 cells /μL (430–1185 cells /μL), CD8+ T cells of 552 cells /μL (180–865 cells /μL), and NK cells of 64 cells /μL (48–450 cells /μL) and no detectable B cells. Clinical history is notable for prior GI infections with Campylobacter jejuni, giardia, and recurrent Clostridium difficile infections as well as bronchiectasis, gastritis, and enteropathy. CA01 had severe chronic GI disease with malabsorption and extreme difficulty maintaining a healthy weight. As intraepithelial lymphocytosis of the small bowel was observed on biopsy, enteropathy was presumed to be the etiology of the patient’s failure to thrive, and several immunomodulatory treatments were tried. Vedolizumab, an integrin α4β7 antagonist that blocks GI lymphocyte recruitment, was stopped after a brief course due to drug-induced liver injury. Additionally, oral steroids (prednisone and budesonide) and abatacept, CTLA-4 Ig, which acts as checkpoint repressor of T cells, failed to achieve sustained improvement (Figure 1A). The patient continued to worsen and required total peripheral nutrition (TPN) due to this severe and refractory GI disease.
Figure 1Severe and refractory GI disease in a patient with CVID with heterozygous 1377delT NFKB1 variant (patient CA01). (A) Difficulty to maintain healthy weight persists despite several therapeutic approaches. TPN, total peripheral nutrition. (B) Family pedigree for heterozygous 1377delT NFKB1 variant. (C) Schematic illustration of NFKB1, p50, and p105 coding regions as denoted. Location of 1377delT variant as indicated in red. RHD, Rel homology domain; ARD, ankyrin repeat domain; DD, death domain. (D) NFKB1 sequence from PBMCs by Sanger sequencing for healthy donor (HD) on left and CA01 on right. 1377delT frameshift denoted in red.
Whole exome sequencing was performed, identifying a heterozygous 1377delT variant in NFKB1 not found in genome databases. Her father also had the 1377delT heterozygous NFKB1 variant and lives in Puerto Rico with limited contact with CA01, sharing 1 other child with the patient’s mother, who does not carry the variant (Figure 1B). The father has no known immunodeficiency, though he has had unspecified GI disease requiring medical care. The 1377delT variant results from a deleted thymidine between the p50 and p105 coding regions of NKFB1, resulting in disruption of the downstream reading frame (Figure 1C). This was confirmed via Sanger sequencing of amplicons targeting NKFB1 Exon 14 from gDNA of CA01 peripheral blood mononuclear cells (PBMCs) (Figure 1D). The chromatograms of this sequence introduce mixed, equivalent peaks downstream of the affected thymine, as would be expected of a heterozygous frameshift variant.
Elevation of NF-κB–driven cytokines, including TNF, and clinical response to TNF antagonism in patient with CVID with heterozygous 1377delT NFKB1 variant. CA01, the patient with CVID with a heterozygous 1377delT NFKB1 variant, had elevated levels of the NF-κB–driven cytokines IL-6, IL-12, and TNF in plasma, above the seventy-fifth percentile for each of these cytokines in our CVID cohort and above levels in healthy donors (HD) (Figure 2A). We then tested the effect of LPS, a TLR4 agonist that is a potent inducer of these cytokines via NF-κB, on PBMC cultures, knowing LPS is also elevated in circulation of patients with CVID (29–31). Addition of LPS resulted in elevation of TNF release by CA01 PBMCs (5 replicates from 5 distinct blood draws over the course of 5 months) (Figure 2B). Difference was significant compared with PBMCs from 10 different patients with CVID and 9 different HD, only for LPS-stimulated cultures. Due to TPN-dependent GI disease and evidence of cytokine dysregulation, CA01 was treated with the TNF inhibitor infliximab. In the months while on this therapy, CA01 regained weight, no longer required TPN, and ultimately regained an excellent quality of life, returning to full-time work (Figure 2C).
Figure 2Elevation of NF-κB–driven cytokines, including TNF, and clinical response to TNF antagonism in a patient with CVID with heterozygous 1377delT NFKB1 variant. (A) Levels of TNF, IL-12, and IL-6 measured in plasma by ELISA. P value calculated by 1-way ANOVA and Kruskal-Wallis test for multiple comparisons. (B) TNF levels in 18 hours PBMC culture media with or without 5 ng/mL LPS measured by ELISA. Results from 9 HD donors and 10 CVID donors, and 5 monthly blood draws from CA01. P value calculated by 2-way ANOVA with Holm-Šídák test for multiple comparisons. (C) Effect of infliximab on CA01 weight over time. PBMC, peripheral blood mononuclear cell; TPN, total peripheral nutrition. *P < 0.05, ****P < 0.0001.
Establishment of an isogenic iMONO model of NFKB1-variant immune dysregulation. We employed an iPSC-based approach that used CRISPR/Cas9-mediated gene editing to prepare multiple syngeneic lines with versus without the NFKB1 variant in order to profile the effects of the variant while controlling for genetic backgrounds. Specifically, we engineered 2 paired clones from each independent genetic background: (a) iPSCs derived from the patient carrying the variant with parallel iPSCs corrected by editing, and (b) normal HD iPSCs with parallel HD iPSCs edited to introduce the variant. This approach is anticipated to remove the potential confounding factors of epigenetic alterations and chronic inflammatory changes that can complicate analysis of primary cells from a patient with CVID with active inflammatory disease, like CA01 (32).
We reprogrammed iPSCs from CA01 and used previously published HD iPSCs (clone BU3) (19). Next, we repaired the 1377delT NKFB1 variant to WT in CA01 iPSCs using CRISPR/Cas9 gene editing. In parallel, we introduced the heterozygous 1377delT NFKB1 variant in HD iPSCs by gene editing, testing sufficiency of this variant to affect observed phenotypes into other genetic backgrounds. All 4 iPSC lines (CA01-derived iPSCs with and without heterozygous 1377delT NFKB1 corrected to WT, HD-derived iPSCs with and without 1377delT NFKB1 heterozygous introduction) exhibited normal growth, morphologies, and karyotypes as well as expected NFKB1 sequences, including 50% with the intended variant sequence as consistent with a heterozygous variant (Supplemental Figure 1; supplemental material available online with this article; https://doi.org/10.1172/jci.insight.198703DS1).
To generate monocytes from iPSCs, we used the method of directed differentiation, adapting the protocol of van Wilgenburg et al. to first differentiate each line into embryoid bodies (EB), which were patterned into lateral plate mesoderm using BMP4, VEGF, and SCF (33) (Figure 3A). After transfer of the resulting cells onto gelatin and treatment with M-CSF and IL-3, putative monocyte-like cells (iMONOs) emerged in suspension culture from their adherent precursors (Figure 3, A and B). We harvested putative iMONOs from each line for profiling, finding they retained the desired NFKB1 genotypes and exhibited similar morphologies (Figure 3B). We monitored iMONO cell counts and viabilities by trypan blue staining and an automated cell counter across several weeks of differentiation. Variation in this data was not statistically significant between genetic backgrounds or NFKB1 genotypes at any point. In addition, iMONOS across all lines produced similar populations of cells expressing monocyte marker CD14 together with hematopoietic marker, CD45 (>90% double positive) (Figure 3C). We did not find CD14 mean fluorescence intensity to differ between groups. CD45+CD14+ iMONOs also expressed similar levels of CD163 (monocyte/macrophage marker) and had low CD16 expression, consistent with the monocyte subset found elevated in blood of patients with CVID with inflammatory complications (24). Our gating strategy for iMONOs is outlined in Supplemental Figure 2.
Figure 3Establishment of an iPSC-derived monocyte (iMONO) model of NFKB1-variant immune dysregulation. (A) Schematic of protocol used to generate monocytes from iPSCs. (B) Morphology and DNA sequence in region of 1377delT NFKB1 variant (highlighted in yellow) for each iPSC line used in this study. (C) Flow cytometry of CD45/CD14 and CD16/CD163 on iMONOs generated from each iPSC line used in this study. Data representative of more than 12 iMONO differentiations. Genetic background of iPSC line and presence of heterozygous NFKB1 variant as denoted.
Heterozygous NFKB1 1377delT variant is necessary and sufficient to cause p105/p50 haploinsufficiency in iMONOs. We next measured levels of the NFKB1-encoded proteins, p105 and its processed form p50, by Western blot in iMONOs after culture with and without LPS (Figure 4A). Protein bands measured corresponded with the 105 kDA and 50 kDA molecular weights for p105 and p50, respectively (Figure 4B). Comparing band density normalized to total protein loaded, we observed that expression of heterozygous NKFB1 1377delT was necessary to reduce the NFKB1 protein by ~50% in the CVID patient–derived iMONOs and was sufficient to also reduce NFKB1 protein by ~50% in HD-derived iMONOs (Figure 4C). This effect was similar for both p50 and p105 and present under both unstimulated and LPS-stimulated conditions. In order to corroborate results from the iMONOs, we also examined PBMCs from the patient with CVID with the heterozygous 1377delT NFKB1 variant along with 2 other patients with CVID with heterozygous NFKB1 variants. We found reduced levels of p105 in the cytosolic fraction and p50 in the nuclear fraction both with and without LPS stimulation compared with HD (Supplemental Figure 3A). We also analyzed protein expression of the other NF-κB family members in iMONOs but found no significant differences under unstimulated or LPS stimulated conditions (Supplemental Figure 3B).
Figure 4Heterozygous NFKB1 1377delT variant is necessary and sufficient to cause p105/p50 haploinsufficiency in iMONOs. (A) Schematic of protocol for iMONO assay. (B) Whole cell iMONO protein lysate (20 μg) analyzed for p105/p50 protein level by Western blot. Source of iPSC and present of NFKB1 variant as indicated. (C) Densitometry analysis of Western blot bands from 3 blots generated from 3 independent experiments. P value was calculated by 1-way ANOVA with Holm-Šídák test for multiple comparisons. *P < 0.05.
iMONO cytokines are shaped by both heterozygous NFKB1 variant and broader CVID genetic background. To measure cytokine protein release from iMONOs we used the same culture conditions as with Western blot (Figure 4A). Using multiplex cytokine measurement, we found that IL-1β, CXCL1, and CXCL2 levels were significantly altered by the presence of the heterozygous 1377delT NFKB1 variant in the CA01 background (Figure 5A). Presence of the heterozygous 1377delT NFKB1 variant was necessary for IL-1β alteration in CA01 iMONOs but not sufficient to induce a significant IL-1β elevation in HD iMONOs. Presence of the heterozygous 1377delT NFKB1 variant reduced levels of CXCL1 and CXCL2 in the CA01 background, with CXCL2 altered in both CA01 and HD backgrounds. All other cytokines measured were not significantly altered by introduction or correction of the 1377delT NFKB1 variant into iMONOs from CA01 or HD, except CXCL5, which was decreased by the 1377delT variant in HD, but not CA01, iMONOs (Supplemental Figure 4).
Figure 5iMONO cytokines are shaped by both heterozygous NFKB1 variant and broader genetic background. (A) IL-1β, CXCL1, and CXCL2 measured in supernatant of iMONO cultures with and without LPS. (B) IL-12, CCL4, and CCL22 measured in supernatant of iMONO cultures with and without LPS. Parental iPSC line (CA01 or HD) and presence of WT or 1377delT NFKB1 for both alleles as noted. Each data point represents an experimental replicate. P value was calculated by 2-way ANOVA with Holm-Šídák test for multiple comparisons. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.
Regarding broader genetic effects (i.e., those other than due to the NFKB1 variant), we found that LPS-stimulated iMONO release of IL-12, as well as the chemokines CCL4 and CCL22, were elevated in cultures of CA01-derived iMONOs compared with those derived from HD (Figure 5B). However, levels of IL-12, CCL4, and CCL22 were unchanged after correction of the heterozygous 1377delT NFKB1 variant in CA01 iMONOs or introduction of the variant into HD iMONOs. We confirmed this finding using a second HD iPSC line (data not shown). Elevation of CCL22 was also noted in CA01 relative to HD iMONO cultures with and without LPS stimulation but was not affected by correction or introduction of 1377delT NFKB1. CCL1 was elevated in unstimulated CA01 relative to HD iMONOs only, and CXCL5 was reduced in LPS-stimulated CA01 versus HD iMONOs and with introduction of 1377delT NFKB1 into HD iMONOs only (Supplemental Figure 4). These results indicate that some aspects of CVID inflammatory dysregulation, including IL-12 elevation, may result from factors independent of the heterozygous NFKB1 variant.
Inflammatory pathway gene expression changes due to 1377delT NFKB1 variant in iMONOs derived from CA01 and HD. To understand the transcriptional dynamics preceding cytokine release and based on qPCR analysis of several marker genes of interest over time after culture with LPS, 4 and 6 hours were selected as time points for bulk RNA-seq (Supplemental Figure 5). All 4 iPSC lines were included in 0 hours, 4 hours, and 6 hours iMONO cultures with LPS, with 4 biological replicates, defined as separate differentiations of iMONOs. A principal component analysis (PCA) plot of our RNA samples demonstrated clusters along the 2 major sources of variance in our data set, culture condition (with or without LPS), and iPSC genetic background (CA01 or HD) (Figure 6A). Thus, like cytokine and chemokine protein levels measured in media of iMONO cultures, iMONO mRNA expression is influenced by iPSC genetic differences and inflammatory conditions, in addition to presence of the heterozygous NFKB1 variant.
Figure 6Inflammatory pathway gene expression changes due to 1377delT NFKB1 variant in iMONOs derived from CA01 and HD. (A) PCA plot of RNA-seq from iMONO cultures for CA01 and HD with and without heterozygous 1377delT NFKB1 variant. (B and C) Volcano plots illustrating distribution of RNA-seq changes from iMONO cultures with and without heterozygous 1377delT NFKB1 variant in CA01 (B) and HD backgrounds (C). (D) Top 10 affected pathways by P value from HALLMARK pathway clustergrams of gene upregulated by heterozygous 1377delT NFKB1 variant in iMONOs from CA01 and HD genetic backgrounds 0, 4, and 6 hours culture with LPS. EMT, epithelial mesenchymal transition. (E) Gene expression increased by 1377delT NFKB1 in CA01, but not HD, iMONOs. (F) Genes from TNF signaling and inflammatory response gene sets with increased expression due to heterozygous 1377delT NFKB1 in CA01, but not HD, iMONOs. Significant pathways in red. PCA, principal component analysis.
Focusing on gene expression changes due to the 1377delT NFKB1 variant, we identified differentially expressed genes in iMONOs between isogenic lines at 0, 4, and 6 hours of culture with LPS using bulk RNA-seq. Isogenic comparison with NFKB1 variant–corrected iMONOs in the CA01 background found 3,195 differentially expressed genes before culture, 3,629 genes after 4 hours of culture with LPS, and 5,806 genes after 6 hours of culture with LPS due to presence of the heterozygous 1377delT NFKB1 variant (Figure 6B). Introduction of heterozygous 1377delT NFKB1 in HD iMONOs resulted in significant differential expression of 5,966 genes before culture, 7,608 genes after 4 hours culture with LPS, and 5,327 genes after 6 hours of culture with LPS (Figure 6C). By isolating a list of genes with similar responses in both of these isogenic comparisons, we identified gene set pathways affected by the 1377delT NFKB1 variant in both CA01 and HD background iMONOs using ENRICHR and analyzed against the HALLMARK pathway gene sets (Figure 6D). Without culture with LPS, gene sets upregulated by heterozygous 1377delT NFKB1 across different genetic backgrounds include those relating to hypoxia, glycolysis, myogenesis, and epithelial to mesenchymal transition. After 4 hours of culture with LPS, additional pathways like WNT and TGF-β signaling are upregulated. Consistent with the observation that longer cultures of our iMONO system exposes greater differences in inflammatory response, 6 hours of culture with LPS revealed differences in gene sets representing inflammatory response, TNF signaling by NF-κB, and IFN-γ response (log10P value of 1.33 × 10–7, 7.54 × 10–7, and 1.38 × 10–3, respectively).
Regarding gene expression changes due to the 1377delT NFKB1 variant found in isogenic comparisons of CA01 but not HD iMONOs, we found 1,059 genes upregulated after 0 hours, 1,967 genes up after 4 hours, and 2,111 genes up after 6 hours (Figure 6E). Again, of the 3 time points we examined, 6 hours of culture with LPS caused upregulation of the most number of genes relating to inflammatory gene expression, including increases in gene sets representing inflammatory response and TNF signaling via NF-κB. Notably, genes upregulated by heterozygous 1377delT NFKB1 variant in CA01, but not HD, iMONOs included the NF-κB–driven cytokines IL-1α, IL-6, and TNF as well as the neutrophil chemoattractants CXCL2 and CXCL3 (Figure 6F). These results indicate key inflammatory genes and pathways altered by the heterozygous 1377delT NFKB1 variant more in the patient with CVID than HD.
Clinical response to TNF antagonism in patient CA01 corresponds with resolution of ulcerative neutrophilic gastritis. Repeat endoscopic biopsy found no change in intraepithelial lymphocytosis of the small bowel despite marked clinical improvement after TNF antagonism in patient CA01. Notably, severe gastritis with mucosal ulcerations, absence of parietal cells, and neutrophil-predominant inflammation within the epithelium, forming crypt abscesses and cryptitis, was present prior to TNF antagonist therapy (Figure 7, A and B). Gastric biopsy had negative gastrin and Helicobacter pylori immunostain (data not shown). This ulcerative neutrophilic gastritis improved with infliximab treatment, including resolution of gastric ulcerations (Figure 7C), cryptitis, and crypt abscesses, as well as fewer neutrophils in the lamina propria (Figure 7D). Convalescence of neutrophilic gastritis corresponded with reduction of neutrophils in the peripheral blood (Figure 7E), while levels of circulating lymphocytes and monocytes were unchanged (data not shown). Plasma levels of the neutrophil chemoattractant chemokine CXCL1 was also reduced by infliximab (Figure 7F). However, there was no significant reduction in other plasma proteins measured, included the systemic inflammatory marker soluble CD14 (data not shown). These results suggest that efficacy of infliximab in patient CA01 may be related to the role of TNF in neutrophil recruitment and/or NF-κB–related inflammation potentiated by the heterozygous 1377delT NFKB1 variant in the GI tract.
Figure 7Clinical response to TNF antagonism in patient CA01 corresponds with resolution of ulcerative neutrophilic gastritis. (A) Endoscopic image of stomach with diffuse ulcerations. (B) Biopsy from the gastric body shows active chronic gastritis, atrophy, and absence of parietal cells (H&E stain at 200× and 400× [inset] magnification). Abundant neutrophils are seen in the lamina propria (black arrows) and within the epithelium, forming crypt abscesses and cryptitis (red arrows). (C) Endoscop
Comments (0)