Hypomorphic nuclear factor-κB essential modulator mutation database and reconstitution system identifies phenotypic and immunologic diversity

Article Outline

• Abstract
• Methods
  • Database
  • Constructs and cell lines
  • Western blot
  • Statistics
  • NF-κB reporter and apoptosis assays
  • Intracellular NEMO fluorescence-activated cell sorting
  • mRNA isolation and analysis
• Results
  • IKBKG hypomorphism and spectrum of disease
  • Immunologic functions in patients with IKBKG hypomorphisms
  • Hypomorphic NEMO complementation
  • Activation of the NF-κB pathway measured by flow cytometry
  • IκB degradation
  • NF-κB directed antiapoptotic function in reconstituted cells
  • TNF-α–induced A20 gene expression
• Discussion
• Methods
  • Generation of cell lines
  • Western blot
  • Phenotype definition
  • Definitions
• References
• Fig E1. 
• Fig E2. 
• Fig E3. 
• Table E1. 
• Table E2. 
• Table E3. 
• Table E4. 
• References
• Copyright

Background

Human hypomorphic nuclear factor-κB essential modulator (NEMO) mutations cause diverse clinical and immunologic phenotypes, but understanding of their scope and mechanistic links to immune function and genotype is incomplete.

Objective

We created and analyzed a database of hypomorphic NEMO mutations to determine the spectrum of phenotypes and their associated genotypes and sought to establish a standardized NEMO reconstitution system to obtain mechanistic insights.

Methods

Phenotypes of 72 individuals with NEMO mutations were compiled. NEMO L153R and C417R were investigated further in a reconstitution system. TNF-α or Toll-like receptor (TLR)–5 signals were evaluated for nuclear factor-κB activation, programmed cell death, and A20 gene expression.

Results

Thirty-two different mutations were identified; 53% affect the zinc finger domain. Seventy-seven percent were associated with ectodermal dysplasia, 86% with serious pyogenic infection, 39% with mycobacterial infection, 19% with serious viral infection, and 23% with inflammatory diseases. Thirty-six percent of individuals died at a mean age of 6.4 years. CD40, IL-1, TNF-α, TLR, and T-cell receptor signals were impaired in 15 of 16 (94%), 6 of 7 (86%), 9 of 11 (82%), 9 of 14 (64%), and 7 of 18 (39%), respectively. Hypomorphism-reconstituted NEMO-deficient cells demonstrated partial restoration of NEMO functions. Although both L153R and C417R impaired TLR and TNF-α–induced NF-κB activation, L153R also increased TNF-α–induced programmed cell death with decreased A20 expression.

Conclusion

Distinct NEMO hypomorphs define specific disease and genetic characteristics. A reconstitution system can identify attributes of hypomorphisms independent of an individual's genetic background. Apoptosis susceptibility in L153R reconstituted cells defines a specific phenotype of this mutation that likely contributes to the excessive inflammation with which it is clinically associated.

Key words: NEMO, immunodeficiency, genetic database, Jurkat reconstitution, NF-κB activation, A20

Abbreviations used: 7-AAD, 7-Amino actinomycin D, DC, Dendritic cell, EDA, Ectodermal dysplasia and anhidrosis, EMSA, Electrophoretic mobility shift assay, FACS, Fluorescence-activated cell sorting, GFP, Green fluorescent protein, IKK, IκB kinase, NEMO, Nuclear factor-κB essential modulator, NF-κB, Nuclear factor-κB, pNEMO, Parental nuclear factor-κB essential modulator, rNEMO, Wild-type reconstituted NEMO(-), TCR, T-cell receptor

Nuclear factor-κB essential modulator (NEMO) is a 419–amino acid regulatory protein encoded by 10 exons on the X chromosome.¹ NEMO participates in the IκB kinase (IKK) complex, which also contains IKKα and IKKβ kinases.² The IKK complex enables nuclear translocation of nuclear factor-κB (NF-κB) dimers by phosphorylating the inhibitor of NF-κB, IκB. This targets IκB for proteosomal degradation releasing its hold on NF-κB in the cytoplasm.

Amorphic NEMO mutations are lethal to males, but hypomorphic mutations can result in ectodermal dysplasia and immunodeficiency. This disease was defined by familial susceptibility to mycobacterial infection, recurrent infection with pyogenic bacteria, and abnormal immunoglobulin production in the setting of variable T-cell and B-cell defects.³, ⁴, ⁵

The ectodermal dysplasia results from an inability of the ectodysplasin A receptor (a TNF receptor family member) to induce NF-κB activation after ligation.⁴ A variety of immunoreceptor functions that depend on NEMO-induced NF-κB activation are similarly defective in patients with NEMO hypomorphisms. The clinical and immunologic phenotypes attributed to NEMO hypomorphs have expanded substantially in recent years. Thus, we have compiled these into a database to define the clinical syndrome further. We have also used a reconstitution system to exploit mechanistic insights derived from the naturally occurring mutations and to test the hypothesis that they are independent of genetic background.

Back to Article Outline

Methods 

Database 

Seventy-two individuals with hypomorphic NEMO mutations were identified using Medline, our own patient evaluations, and conference abstracts. Brothers of index cases having characteristic disease features were assumed to carry the same mutation. Detailed definitions of specific clinical and immunologic categories are provided in this article's Methods in the Online Repository at www.jacionline.org. Patients evaluated through our center were evaluated in accordance with our Institutional Review Board for the protection of human subjects.

Constructs and cell lines 

3T8 is a previously described⁶ Jurkat cell line expressing an NF-κB reporter construct containing the rat Thy-1 gene and is designated parental NEMO (pNEMO). pNEMO was previously mutagenized to generate a NEMO-deficient line, 8321,⁶ herein designated NEMO(-), genomic sequencing of which revealed a hemizygous point mutation 1000G>T leading to a predicted Glu334X (see this article's Fig E1 in the Online Repository at www.jacionline.org). However, direct evaluation of NEMO protein in cells using polyclonal antibodies and mAbs raised against full-length and the leucine zipper of NEMO, respectively, demonstrated negligible specific protein (see this article's Fig E2 in the Online Repository at www.jacionline.org). NEMO cDNA was cloned and transduced⁷ to generate the following cell lines: wild-type reconstituted NEMO(-) (rNEMO); L153R reconstituted NEMO(-) (L153R); C417R reconstituted NEMO(-) (C417R); empty vector reconstituted NEMO(-) [green fluorescent protein (GFP)-NEMO(-)]; and empty vector transduced pNEMO (GFP-pNEMO) (details in Methods in the Online Repository).

Western blot 

Western blotting was performed as previously described⁸ (details in Methods in the Online Repository).

Statistics 

The Student t test was performed to evaluate mean data where indicated.

NF-κB reporter and apoptosis assays 

1 × 10⁶ cells were treated with 10 ng/mL recombinant human TNF-α (R&D Systems, Minneapolis, Minn) or 50 ng/mL recombinant Salmonella flagellin (Invitrogen, San Diego, Calif), after which cells were collected, washed, and incubated with phycoerythrin-conjugated antirat Thy-1 (CD90) antibody (BD Biosciences, San Jose, Calif). Apoptosis was concurrently assayed by resuspension in binding buffer (10 umol/L HEPES, 140 mmol/L NaCl, and 2.5 mmol/L CaCl₂) containing annexinV-Cy5 and 7-amino actinomycin D (7-AAD; BD).

Intracellular NEMO fluorescence-activated cell sorting 

Cells were fixed and permeabilized in cytofix-cytoperm solution (BD) and washed and incubated with mouse anti-NEMO mAb (clone 54; BD) or isotype-matched IgG control (clone MOPC-21; BD) for 1 hour. After washing, cells were incubated for 1 hour with Alexa Fluor 647–conjugated antimouse IgG (Invitrogen) and analyzed by fluorescence-activated cell sorting (FACS).

mRNA isolation and analysis 

RNA was extracted from cells, cDNA generated, and A20 or actin targets amplified as described.⁸

Back to Article Outline

Results 

IKBKG hypomorphism and spectrum of disease 

Because NEMO gene (IKBKG) hypomorphisms cause a variety of phenotypes, a database was compiled to gauge diversity and potentially identify genotype/phenotype correlations. Seventy-two individuals were included (Fig 1). Missense mutations account for 40%, splice-site 21%, frameshift 25%, and nonsense 14%. Eleven mutations were shared by 51 patients; the other 21 mutations were unique. Fifty-three percent of mutations specifically affected the zinc finger domain because of missense, nonsense, or frameshift. Three percent were within the region aa50-120, important for interacting with the other members of the IKK complex,⁹ and 15% were in the region responsible for allowing NEMO oligomerization.¹⁰ Seven percent of mutations affected the NEMO ubiquitin binding domain, important for binding K63-linked polyubiquitin.¹¹ Two patients were female¹², ¹³ but had defective X chromosome lyonization and characteristics of the disease.

• 
  • View Large Image
  • Download to PowerPoint

• Fig 1. 

  Hypomorphic NEMO mutations. Each asterisk represents an individual patient, and mutation types are color-coded. Structural predictions indicate an extended α-helix structure with 2 coiled coils, a leucine zipper, and zinc finger motifs. The minimal oligomerization domain, serine phosphorylation (p-S), ubiquitination (U), sumoylation (S), ubiquitin binding (NUB), and IKK binding/NEMO dimerization regions are shown. αH, Alpha helix; CC, coiled coil; LZ, leucine zipper; ZF, zinc finger.

Patient clinical and immunologic characteristics were compiled according to clinical phenotype, infectious susceptibility, and immune capacity (see Methods in the Online Repository for definitions). Fifty-three categories were defined and considered for each patient (Table E1, Table E2, Table E3, Table E4). For any category in which insufficient details were available, patients were excluded from calculations. A synopsis of key findings is provided in Table I. Seventy-seven percent (40/52) of patients were diagnosed with ectodermal dysplasia and anhidrosis (EDA) or met our definition. Four percent (2/52) of patients had dental abnormalities alone and were not included as having EDA. Three discrete regions of NEMO contained alterations not resulting in an ectodermal phenotype (Fig 2, A). Osteopetrosis has been described in 7.5% (5/65) of patients (Fig 2, B). In 1, bone demonstrated no osteoclasts,¹⁴ but in others, varying severities of pathology were identified.¹⁵, ¹⁶ Ten percent (6/65) of patients had vascular anomalies affecting lymphatic or venous systems⁴, ¹⁵, ¹⁷, ¹⁸, ¹⁹, ²⁰ (Fig 2, B), ranging from transient lower limb edema¹⁷ to persistent defects with abnormal lymphoscintigrams¹⁵ or multiple lymphangiomas.¹⁶

Table I. Clinical and immune function of individuals with hypomorphic NEMO mutation

Functional or clinical category	Observed deficiency	Affected (%)
Ectodermal dysplasia (1)	40/52	77
Osteopetrosis (2)	5/65	8
Lymphedema (3)	5/65	8
Small for gestational age (8)	9/65	14
Autoimmune/inflammatory disease (7)	14/66	23
Dead (10)	24/66	36
Infectious susceptibility (11)	60/61	98
Bacterial infection	45/52	86
Mycobacterial infection	23/52	44
Pneumocystis pneumonia	4/52	8
DNA viral infection	11/52	21
Meningitis	12/61	21
Pneumonia	19/61	31
Sepsis/bacteremia	20/61	33
Abscess	18/61	30
Hyper-IgM (21)	6/40	15
Hypogammaglobulinemia (20)	24/41	59
Hyper-IgA (22)	13/35	37
Hyper-IgD	2/5	40
Specific antibody deficiency (19)	18/28	64
Specific pneumococcal antibody (19)	13/16	81
B-cell costimulation/CD40 signaling (14)	15/16∗	94
TNF response (26)	9/11∗	82
IL-1 response (25)	6/7∗	86
TLR response (27)	9/14∗	64
Natural killer function (23)	10/10∗	100

Numbers in parentheses refer to the phenotype definition number provided in this article's supplemental Methods text in the Online Repository at, www.jacionline.org.

• 
  • View Large Image
  • Download to PowerPoint

• Fig 2. 

  NEMO phenotype maps. The following phenotypes are shown: ectodermal dysplasia (A), lymphedema/osteopetrosis (B), inflammatory disease (C), pyogenic infection (D), mycobacterial infection (E), TNF-α response (F), hyper-IgM phenotype/CD40 (G), IL-1/TLR response (H), TCR response (I), and mortality (J). Each oval represents the reported presence (shaded) or absence (dashed) of the indicated phenotype, and is intended to reflect the protein region affected.

Inflammatory conditions or autoimmunity affected 25% (15/61) of patients (Fig 2, C). The most frequent was inflammatory colitis²¹ and occurred in 21% (13/61). Forty-six percent (6/13) of these individuals had intractable diarrhea, and 30% (4/13) were diagnosed with failure to thrive. Autoantibody-associated disease was described in 1 patient with autoimmune hemolytic anemia.²² Chronic arthritis affected 3% (2/66).²³ Hemophagocytic syndrome after Klebsiella pneumoniae infection was identified in 1 patient.¹⁴ Fourteen percent (9/65) of individuals were small for gestational age, but most were from a single kindred.¹⁸ Pre-eclampsia complicated 3% (2/66) of deliveries.²⁰, ²⁴

The most common infections included pneumonia (31%, 19/61) leading to bronchiectasis in 9%, bacteremia or sepsis (33%, 20/61), skin and deep tissue abscess formation (30%, 18/61), intestinal infection (23%, 14/61), encephalitis or meningitis (20%, 12/61), sinusitis (11%, 7/61), and osteomyelitis (11%, 6/61)—usually with atypical mycobacteria (Table E2). Pyogenic bacterial infection was identified in 87% (45/52) of patients in whom an organism of any kind was identified (Fig 2, D). Pathogens identified in greater than 10% included Streptococcus pneumoniae, Haemophilus influenza, and Staphylococcus aureus. Mycobacterial infection, most commonly caused by Mycobacterium avium intracellulare affected 44% (23/52; Fig 2, E) and included cellulitis, osteomyelitis, lymphadenitis, pneumonia, and disseminated forms. Serious viral infection occurred in 21% (11/52) and included herpes simplex virus encephalitis,²² severe adenoviral gastroenteritis,¹⁶ and cytomegalovirus sepsis.²³ Fungal and opportunistic infections occurred in 10% (6/52) of patients; Pneumocystis and oral candidiasis were predominant.

Intravenous immunoglobulin (IVIG) replacement therapy was documented in 29 of 58 individuals (50%) who survived beyond 6 months. Antibiotic prophylaxis to prevent Pneumocystis and/or mycobacteria was documented as provided to 11 of 58 patients (19%). Additional interventions documented included cytokine therapy to augment immune function,²⁵ IFN-γ as an antimycobacterial,²⁶ and hematopoietic stem-cell transplantation.¹⁵

Immunologic functions in patients with IKBKG hypomorphisms 

Given the range of immunoreceptors that use NEMO, it is possible that specific infectious susceptibilities are defined by the impact of individual mutations on immune signaling. Evaluation of TNFα receptor, CD40, TLR, IL-1 receptor, and T-cell receptor (TCR) signaling, as well as antigen-presenting cell costimulation, antibody repertoire generation, B-cell and T-cell development and memory, natural killer cell function, and monocyte activation have all been recorded. All mutations tested demonstrated some impairment in NF-κB signaling as defined in the Methods in the Online Repository. Defects in the TNFR superfamily functions were common, with 82% (9/11) impairing TNF-α–induced NF-κB activation, but D406V⁵ and C417R⁵ mutations did not (Fig 2, F; Table E3) and R319Q²⁷ showed partial impairment. CD40 signaling impairment was found in 94% (15/16); however, only 27% (4/15) of these had an immunoglobulin class-switch defect in vitro (Fig 2, G; Table E4). Hypogammaglobulinemia occurred in 24 of 41 (59%) but correlated with impaired CD40 signaling only in zinc finger mutations. Defects in specific antibody production occurred in 64% (18/28), and deficits in specific antibodies against S pneumoniae were identified in 72% (13/16) of patients tested. Of patients with specific antibody defects, only 15% (6/40) had low IgG with normal or elevated IgM. Defects in other immune responses and pathways were also common. Eighty-six percent (6/7) of patients had abnormal IL-1 signaling, and 64% (9/14) had abnormal TLR signaling (Fig 2, H). Thirty-nine percent (7/18) of individuals in whom innate signaling pathways were tested had no detectible abnormality in at least 1 test (TNFR, IL-1, TLR4, or other TLR). Lymphocyte quantitation and proliferative function were frequently normal. Sixty-five percent (11/17) and 73% (8/11) had normal or elevated CD4 and CD8 counts, respectively. Mitogen-induced and antigen-induced proliferation was normal in 91% (20/22) and 76% (11/14), respectively (Fig 2, I; Table E3). Delayed-type hypersensitivity testing, however, was normal in only 3 of 7 (43%). Patients with C417R mutation had impaired dendritic cell (DC) IL-12 secretion and failure to upregulate costimulatory molecules.²⁹ Natural killer cell cytotoxicity was globally deficient.¹⁴, ²³, ²⁵, ²⁸

To evaluate consistency of expression of the most common phenotypes, the 11 shared mutations were analyzed (Fig 3). The frequency of EDA was 100% in 9 of these mutations and 0% to 25% in the E315A and R319Q mutations. Inflammatory colitis occurred in 100% of the E391X individuals, but in only 25% to 75% in the 3 other mutations in which it was reported, although it appeared in 71% (5/7) of Δexon 4-6 mutations. This latter mutation was highly correlated with mortality, 10 of 10 (100%), and small for gestational age (SGA), 7 of 7 (100%). Susceptibility to mycobacterial infection was generally absent in individuals with mutations in the first coiled-coil and α-helix and was strongly associated with E315 and E319 mutations. Hypogammaglobulinemia affected patients with Δex4-6, L227P, and C417R substitutions, and zinc finger (ZF) truncations (with the exception of E391X). The hyper-IgM phenotype was particular to individuals with C417 mutations.

• 
  • View Large Image
  • Download to PowerPoint

• Fig 3. 

  Phenotype frequency of shared mutations. Each column represents a mutation that occurred in more than 1 individual. Frequency is depicted by quartile and is color-coded: high (red), intermediate (yellow), and low (green) phenotype presence. IBD, Inflammatory bowel disease; SGA, small for gestational age; Spec. Ab, specific antibody.

Hypomorphic NEMO complementation 

We next wanted to determine the effect of particular NEMO mutations on NF-κB–dependent signaling pathways. If we could establish that individual hypomorphisms possessed differential properties in the context of a standardized genetic background, it would support a mechanism of genotype-phenotype correlations, thereby substantiating our central hypothesis. Thus, a NEMO(-)–deficient Jurkat T-cell line stably expressing an NF-κB reporting construct was used. Wild-type or patient-derived hypomorphic sequences were cloned into a retroviral vector preceding internal ribosomal entry site and green fluorescent protein (GFP) sequences. The L153R and C417R mutations were selected because of their similarities and differences. Both result in the originally described syndrome of EDA and immunodeficiency, and both are caused by missense mutations introducing an arginine. Differences included (1) the presence of inflammatory colitis (L153R only), (2) impaired LPS response (L153R only), and (3) hyper-IgM phenotype (C417R only). Recombinant retroviruses encoding wild-type NEMO sequences, L153R, or C417R mutations were therefore generated and used to infect NEMO(-) cells. Nonclonal populations that had stably incorporated the construct were selected by GFP FACS and maintained as stable cultures. These were refined to express equal and physiological levels of NEMO as determined by Western blot (Fig 4, A). The level of reconstituted NEMO in individual cells was also comparable to that in parental Jurkat cells (pNEMO) as demonstrated by intercellular NEMO FACS (Fig 4, B). This correlated with GFP fluorescence in individual reconstituted cells (Fig 4, C), further demonstrating equivalent expression.

• 
  • View Large Image
  • Download to PowerPoint

• Fig 4. 

  Expression levels of reconstituted NEMO are equivalent by anti-NEMO Western blot, intracellular FACS, and GFP FACS. A, Cells from reconstituted lines were lysed and probed with anti-NEMO mAb specific for the NEMO leucine zipper. Actin blotting demonstrates equal loading. FACS to determine GFP expression was performed on NEMO reconstituted cells lines (B), which was evaluated by intracellular staining (n = 2) (C). The gray shaded area demonstrates fluorescence of isotype-control stained cells.

Activation of the NF-κB pathway measured by flow cytometry 

To investigate the effects of NEMO hypomorphism on innate immune signaling in T cells, Jurkat cells were cultured for 8 hours in the presence of flagellin or TNF-α. Surface levels of rat Thy-1 expressed by the NF-κB reporter construct were determined by FACS. Jurkat T cells express TLR5, and exposure to the TLR5 ligand flagellin leads to activation of NF-κB.³⁰, ³¹ Rat Thy-1 expression was not upregulated in NEMO(-) cells after TNF-α or flagellin stimulation but was in rNEMO cells (Fig 5, A). Rat Thy-1 upregulation in rNEMO cells was comparable to that in pNEMO cells (not shown). In contrast, NEMO(-) cells reconstituted with L153R and C417R constructs had reduced NF-κB activation in response to either TNF-α or flagellin. Mean fluorescence intensity of induced rat Thy-1 in repeated experiments was significantly decreased by ∼75% to 90%, respectively, compared with control (Fig 5, A).

• 
  • View Large Image
  • Download to PowerPoint

• Fig 5. 

  Decreased NF-κB reporter expression after stimulation with TNF-α and flagellin in reconstituted NEMO(-) cells and impaired IκB degradation in the L153R but not C417R cell line. A, Cells were stained with rat-Thy-1phycoerytherin and analyzed by FACS. Decreased levels of expression indicate decreased NF-κB activation in response to TNF-α and flagellin in L153R and C417R. Replicates of experiments indicate significant differences compared to rNEMO; means, SDs, and P values are shown in the box above the histogram. ΔMFI denotes the difference in mean fluorescence intensity between stimulated and unstimulated cells. B, Western blot of IκB levels from the various cell lines after TNF-α activation. Densitometry measurements of IκBα/actin normalized to time = 0 for each cell line are indicated below individual bands.

IκB degradation 

To dissect the mechanisms by which each NEMO hypomorphism affects NF-κB activation, and delineate signaling pathways relative to the IKK complex, we initially measured IκBα degradation at different times after TNF-α stimulation. TNF-α failed to induce rapid degradation of IκB in NEMO(-) or L153R cells (Fig 5, B). In C417R-NEMO cells, however, there was initial IκB degradation, and restoration of IκB levels at 60 minutes. Quantitative analysis of IκBα levels relative to actin confirmed these patterns (Fig 5, B) and thus defines differences between hypomorphism-expressing and NEMO(-) cells.

NF-κB directed antiapoptotic function in reconstituted cells 

To determine the effects of NEMO hypomorphism on TNF-α–induced programmed cell death in T cells, the individual cell lines were cultured in the presence of TNF-α for 8 hours. Cell surface binding of annexin-V, which occurs during the early and late phases of apoptosis, and 7-AAD uptake, which occurs only in dead cells, was determined by FACS. After TNF-α activation of NEMO(-) cells, almost all (96%) cells bound annexin-V, of which 29% were in later phases of cell death as determined by 7-AAD retention (Fig 6, A). In rNEMO cells, there was reduced annexin-V binding (30%), and only 9.5% retained 7-AAD, similar to pNEMO cells (not shown). L153R cells bound annexin-V substantially (88%) after TNF-α stimulation and retained 7-AAD similarly to NEMO(-) cells (28%). In contrast, C417R cells demonstrated intermediate annexin-V binding (55%) and 7-AAD retention (11%), more closely resembling rNEMO cells. These differences were confirmed in independently repeated experiments, and annexin-V binding in NEMO(-) and L153R was significantly higher than in rNEMO cells (Fig 6, B). rNEMO and C417R were not statistically different (P = .17). As expected, flagellin did not induce programmed cell death (Fig 6, B).

• 
  • View Large Image
  • Download to PowerPoint

• Fig 6. 

  Apoptosis in TNF-α stimulated cells and A20 expression. A, Cell lines were activated with TNF-α, and apoptosis was measured by annexin-V and 7-AAD. B, Replicates (N = 3) and statistical evaluation of repeated apoptosis assays. C, A20 transcripts were quantified by using real-time PCR, and fold induction of A20 expression is reduced in L153R reconstituted NEMO(-) cells. The result is representative of 2 independently conducted experiments.

TNF-α–induced A20 gene expression 

To evaluate whether differences in programmed cell death observed in L153R correlated with aberrant TNF-α–induced survival gene expression, quantitative real-time PCR was performed. The antiapoptotic A20 gene constitutively expressed in Jurkat cells is strongly induced by TNF-α and requires NEMO function.⁶ In rNEMO and pNEMO cells, A20 expression was induced ∼7 fold after TNF-α stimulation (Fig 6, C). In contrast, induced A20 expression in L153R cells was >50% reduced compared with rNEMO cells. In C417R cells, however, TNF-α induced A20 expression at levels similar to that in rNEMO cells. Thus, hypomorphic NEMO mutations demonstrated differential ability to protect T cells from TNF-α–induced programmed cell death, which may be at least in part a result of impaired expression of A20. This may help explain differences in clinical phenotype in patients with these mutations, because L153R but not C417R has been associated with organ-specific inflammatory disease.

Back to Article Outline

Discussion 

The previous conception of human disease caused by hypomorphic NEMO mutation is one that affects males, is associated with EDA in all but very rare cases, and is characterized by bacterial infection with poor production of specific antibody. We assembled a database of known mutations to discern phenotypic diversity of mutations and discover potential genotype/phenotype correlations. Although many of the previous characteristics of disease are apparent in the 72 patients considered here, the spectrum of disease caused by NEMO hypomorphism is different than what has been based on earlier series.³, ⁴, ²¹, ²³

Although originally described as EDA-ID, only 77% of individuals with NEMO mutation and immunodeficiency in our database had EDA. Because essentially all had immunodeficiency, a more appropriate name for this syndrome might be NEMO mutation with immunodeficiency, or NEMO-ID. Individuals demonstrated susceptibilities to pyogenic bacteria, atypical mycobacteria, viruses, and Pneumocystis, with cases affected by the latter 2 increasing in recent years. Autoinflammatory disease occurred frequently, most commonly affecting the gut.²¹, ²³ Signaling defects were varied, and increasing numbers of mutations that permit partial TNF-α and TLR signaling have been identified (Fig 2, F, H). Early mortality has been increasingly described, because the mean age at death in patients reported over the last 3 years was 2.3 years, compared with 6.4 years for all patients in the database.

A finding consistent with the previous understanding of this disease was the high proportion of patients affected by pyogenic infection. Similarly, CD40 signaling was impaired in most mutations tested. The classic hyper-IgM phenotype, however, affected a minority of patients, most specifically ΔN37,²² R175P,³² C417 alterations,³, ⁵, ²³ or X420 frameshift mutation.⁴ Also, as expected, approximately 2/3 of individuals had defective specific antibody production, with a suggested selective inability to generate pneumococcus-specific antibodies. Hypogammaglobulinemia was still present in the majority (∼60%) and osteopetrosis and lymphedema in the minority (∼7.5%).

Because this was a retrospective investigation of anecdotal reports and case series, both an ascertainment and reporting bias exist because of overrepresentation of severe and extraordinary cases. Generalizations about disease in the native population, therefore, should be made with caution. Recently reported cases have appeared to be more severe, but this is likely skewed because of 1 large kindred.¹⁸ Longitudinal evaluation of patients in prospective studies would address these issues.

Interestingly, some phenotypes were characteristic of mutations in particular NEMO domains, whereas others were private to specific mutations. EDA was attributed to 3 distinct NEMO regions, largely sparing mutations of the leucine zipper and C-terminal portions of the first and second coiled-coil domain (Fig 2, A). The hyper-IgM phenotype occurred with mutations affecting the zinc finger domain (Fig 2, G). The region immediately preceding the leucine zipper is required for signaling by CD40 and TNF-α, but not IL-1/TLR or TCR. Certain mutations, such as zinc finger truncations and Δ4-6 splice mutations, appear to affect function globally (Fig 2; Table E1, Table E2, Table E3, Table E4). Importantly, evaluation of grouped mutations fails to define completely uniform characteristics, thus suggesting some variable penetrance.

To consider genotypic association independently of a patient's genetic background, we established a reconstitution system and studied 2 patient-derived hypomorphisms. These mutations were selected based not on the frequency with which they occurred but on important similarities and differences. This made them suitable candidates to demonstrate the proof-of-principle that functional differences between mutations could be attributed to a specific hypomorphism. Wild-type NEMO reconstitution restored physiologic function, but hypomorphisms did not. The C417R mutation permitted IκB degradation after TNF-α stimulation, in agreement with results obtained in patient-derived cells,⁵ and was accompanied by A20 transcription and protection of cells from TNF-induced apoptosis (Fig 5, Fig 6). Mutation of C417 is known to affect NEMO folding.³³ This may prevent physical interaction between NEMO and proteins required for full signal transduction, but still permit some kinase activity of the complex. A model depicting this possible role for NEMO function is shown in Fig E3 in the Online Repository at www.jacionline.org. In contrast, L153R resulted in full impairment of IKK activity, with no IκB degradation after TNF stimulation, and increased apoptosis after T-cell exposure to TNF and failure to induce A20 expression. This may address mechanisms for NEMO deficiency and inflammation, because the patient with an L153R hypomorphism had severe intestinal inflammation.²³ Complete NEMO deficiency in epithelial cells in mice causes inflammatory colitis and apoptosis, likely because of impaired barrier to intestinal flora.³⁴, ³⁵ However, an additional role in promoting inflammation after exposure of T cells to innate immune signals may contribute to clinical phenotype. It may further explain why not all individuals with NEMO-ID have intestinal inflammation.

Our analysis defines disease caused by hypomorphic NEMO mutations as diverse and complex, but there is suggestion of associations of particular phenotypes with NEMO genotypes, which raise important biological specificities of altered regions of NEMO. The use of reconstitution systems will help further important biological insights that can be derived from the disease. Clinically the list of phenotypes attributed to mutations is expanding and warrants careful consideration of patients with undiagnosed immunodeficiency.

Back to Article Outline

Methods 

Generation of cell lines 

NEMO cDNA was cloned from human PBMCs into the pCDNA3 vector system. Specific primer sets were used for site-directed mutagenesis to introduce predicted L153R and C417R mutations (available on request) and were confirmed by sequencing. NEMO constructs were amplified with primers adding 5' XhoI and 5'EcoRI restriction sites, and ligated into the Topo cloning system (Invitrogen) and subcloned into the MIGR1-IRES-GFP retroviral vector (a kind gift of Dr Warren Pear) after confirmation by DNA sequencing. Recombinant vectors were lipofected using Fugene (Roche, Basel, Switzerland) into amphotropic retroviral packaging cells^E7 and supernatants used to infect NEMO(-) cells. Subsequently, cells were sorted for GFP expression by FACS and grown as stable cultures, which were further sorted for uniformity of GFP expression among all cell lines. Equivalent GFP expression was regularly monitored. Cell lines generated included rNEMO, L153R reconstituted NEMO(-) (L153R), C417R reconstituted NEMO(-) (C417R), empty vector reconstituted NEMO(-) [GFP-NEMO(-)], and empty vector transduced pNEMO (GFP-pNEMO). All cell cultures were maintained in RPMI 1640 (Invitrogen) with 10% FCS (Atlanta Biologicals, Lawrenceville, Ga), HEPES, essential amino acids, L-glutamine, sodium pyruvate, penicillin-streptomycin, and 500 mg/mL G418 (to maintain selection of reporter constructs).

Western blot 

Nuclear factor-κB essential modulator, IκB, and actin immunoblotting was performed by using 1.5 × 10⁶ cells per condition. Cells were treated as described, lysed in NuPAGE LDS sample buffer (Invitrogen), and boiled for 5 minutes before loading equal cell equivalents per lane and separating lysates on 4% to 12% Bis-Tris density gradient gels (Invitrogen) in MOPS SDS running buffer. Separated proteins were transferred to PVDF membranes (Invitrogen), which were blocked with 3% BSA at room temperature for 1 hour. Blocked membranes were then incubated in 1% BSA and 0.1% Tween-20 with mouse monoclonal anti-NEMO, clone 54 (BD Biosciences), rabbit polyclonal anti-NEMO, SC-8330 (Santa Cruz Biotechnology, Santa Cruz, Calif), or rabbit polyclonal anti-IκBα, C-21 (Santa Cruz Biotechnology). Bound antibody was detected by using horseradish peroxidase–conjugated donkey antirabbit or sheep antimouse (Amersham Biosciences, Piscataway, NJ) and ECL plus detection system (Amersham Biosciences). Where specified, membranes were stripped in 0.2 mol/L glycine (pH 2.5), 0.05% Tween-20, and 140 mmol/L NaCl in TRIS-buffered saline at 50°C for 30 minutes, blocked with 3% BSA, and reprobed with rabbit antiactin polyclonal antibody 20–33 (Sigma-Aldrich, St Louis, Mo). Densitometry was performed using ImageJ software (http://rsbweb.nih.gov/ij/).

Phenotype definition 

To compile the database of clinical and immunologic characteristics of patients with hypomorphic NEMO mutations, specific definitions were used for each database component. In some cases, the definitions are purposefully flexible to allow evaluation of clinical and immunologic characteristics that were not uniformly repeated. In some cases, direct evidence within a given source was available, but in others, the evidence may have been only referred to within the source material. Individual definitions are provided.

Definitions 

Back to Article Outline

References 

Back to Article Outline

Fig E1. 

• 
  • View Large Image
  • Download to PowerPoint

• Electrophoregram of NEMO G1000T after sequencing of the gel-purified NEMO gene–specific long range PCR product, demonstrating hemizygous presence in the IKBKG gene-specific sequence of the male karyotype Jurkat cell line, resulting in stop codon and predicted L334X protein.

Back to Article Outline

Fig E2. 

• 
  • View Large Image
  • Download to PowerPoint

• Western blot of the parental Jurkat cell line (pNEMO) and NEMO-deficient line (NEMO[-]). Membranes were probed with rabbit polyclonal (left) and mouse monoclonal (right) antibodies, indicating the presence of only a specific band at the expected molecular weight. P, polyclonal; M, monoclonal.

Back to Article Outline

Fig E3. 

• 
  • View Large Image
  • Download to PowerPoint

• Model of innate immune signaling through NEMO. Signaling occurs through different groups of functionally related proteins downstream of TNFR and TLR5, which activate the IKK complex and lead to IκB degradation (1). IκB processing leads to nuclear NF-κB translocation (2) and gene transcription (3). Simultaneously, programmed cell death pathways are activated (4) and suppressed (5) by NF-κB–dependent (right orange line) and classical NF-κB–independent (left orange line) gene transcription, such as A20. Thus, a NEMO-dependent but NF-κB–independent pathway uncovered by NEMO-C417R may exist (6). IRAK, IL-1 receptor–associated kinase; TRAF, TNF receptor–associated family of proteins; TAB, TAK-binding protein; TAK, TGF-β activated kinase; MyD, myeloid differentiation; FADD, Fas-associated death domain-containing protein; TRADD, TNF receptor–associated death domain-containing protein; RIP, receptor-interacting protein; cyt. C, cytochrome C.

Back to Article Outline

Table E1. 

Clinical phenotypes in 72 patients with NEMO hypomorphism

A, Arthritis; a(d) C reactive protein (Crp), appropriate (decreased) Crp with infection; AD, atopic dermatitis; aH, alpha helix; AIHA, autoimmune hemolytic anemia; C, colitis; CC, coiled coil; CI, conical incisors; D, delayed, dia, persistent diarrhea; EDA, ectodermal dysplasia with anhidrosis; FTT, failure to thrive; HD, hypodontia; HPS, hemophagocytic syndrome; I Ab, induced abortion; ISD, immune system dysregulation, autoimmune; ↓LF, decreased or absent lymphoid follicles; LZ, leucine zipper; PLE, protein losing enteropathy; Ost, osteopetrosis; SD, seborrheic dermatitis with progression to erythroderma; SGA, small for gestational age; UC, ulcerative colitis pathology; ZF, zinc finger domain.

Back to Article Outline

Table E2. 

Infections disease phenotypes in 72 patients with NEMO polymorphism

AF, Aspergillus fumigatus; a MB, atypical mycobacterial species; B, bacteremia; CA, Candida albicans; C/h, colitis/hepatitis; CMV, cytomegalovirus; D, diarrhea; EC, Enterococcus spp; ENT, Enterobacter sp; GE, Giardia enteritidis; H flu, Hemophilus influenzae; HSV-1, herpes simplex virus 1; K, Klebsiella pneumoniae; M, meningitis; M Ab, Mycobacterium abscessus; MAI, Mycobacterium avium intracellulare; MCV, molluscum contagiosum virus; M. Tb, Mycobacterium tuberculosis; NS, not specified; PCP, Pneumocystis jirovecii; PS A, Pseudomonas aeruginosa; r, recurrent; SA, Staphylococcus aureus; SE, Salmonella enteritidis; SM, Streptococcus mitis; SP, Streptococcus pneumoniae; SV, Streptococcus vestibularis; VRE, vancomycin-resistant Enterococcus.

Back to Article Outline

Table E3. 

Immunologic function phenotypes in 72 patients with NEMO hypomorphism, part 1

ADCC, Antibody-dependent cellular cytotoxicity; DTH, delayed-type hypersensitivity;↑, high level; ↓, low level; nl, level within normal range; NK, natural killer; PMA/I, phorbol myristate acetate/ionomycin; tet, tetanus toxoid; dipth, diptheria toxoid; CONA, concanavalin A; PWM, pokeweed mitogen; CA, Candida albicans.

Back to Article Outline

Table E4. 

Immunologic function phenotypes in 72 patients with NEMO hypomorphism, part 2

CSR, Class switch recombination; ↑, high level; ↓, level below normal; nl, level within normal range; tet, tatanus toxoid; dipth, diptheria toxoid; HiB, Haemophilus influenzae type B; IHA, isohemagglutinins; PCV, pneumoccocal conjugate vaccine; 2/7 conj, antibodies present to 2/7 conjugate antigens; CMV, cytomegalovirus; AHA, allohemagglutinin; MMR, measles, mumps, rubella; VZV, varicella zoster virus; OPV, oral poliovirus; H flu, Hemophilus influenzae; MGUS, monoclonal gammopathy of undetermined significance.

Back to Article Outline

References 

1. Aradhya S, Bardaro T, Galgoczy P, Yamagata T, Esposito T, Patlan H, et al. Multiple pathogenic and benign genomic rearrangements occur at a 35 kb duplication involving the NEMO and LAGE2 genes. Hum Mol Genet. 2001;10:2557–2567
   • View In Article
   • MEDLINE
   • CrossRef
2. Gilmore TD. Introduction to NF-kappaB: players, pathways, perspectives. Oncogene. 2006;25:6680–6684
   • View In Article
   • MEDLINE
3. Zonana J, Elder ME, Schneider LC, Orlow SJ, Moss C, Golabi M, et al. A novel X-linked disorder of immune deficiency and hypohidrotic ectodermal dysplasia is allelic to incontinentia pigmenti and due to mutations in IKK-gamma (NEMO). Am J Hum Genet. 2000;67:1555–1562
   • View In Article
   • MEDLINE
   • CrossRef
4. Doffinger R, Smahi A, Bessia C, Geissmann F, Feinberg J, Durandy A, et al. X-linked anhidrotic ectodermal dysplasia with immunodeficiency is caused by impaired NF-kappaB signaling. Nat Genet. 2001;27:277–285
   • View In Article
   • MEDLINE
   • CrossRef
5. Jain A, Ma CA, Liu S, Brown M, Cohen J, Strober W. Specific missense mutations in NEMO result in hyper-IgM syndrome with hypohydrotic ectodermal dysplasia. Nat Immunol. 2001;2:223–228
   • View In Article
   • MEDLINE
   • CrossRef
6. He KL, Ting AT. A20 inhibits tumor necrosis factor (TNF) α-induced apoptosis by disrupting recruitment of TRADD and RIP to the TNF receptor 1 complex in Jurkat T cells. Mol Cell Biol. 2002;22:6034–6045
   • View In Article
   • MEDLINE
   • CrossRef
7. Swift S, Lorens J, Achacoso P, Nolan GP. Rapid production of retroviruses for efficient gene delivery to mammalian cells using 293T cell-based systems. Curr Protoc Immunol. 2001;31:14–29
   • View In Article
8. Pandey R, Destephan CM, Madge LA, May MJ, Orange JS. NKp30 ligation induces rapid activation of the canonical NF-κB pathway in NK cells. J Immunol. 2007;179:7385–7396
   • View In Article
9. Marienfeld RB, Palkowitsch L, Ghosh S. Dimerization of the I kappa B kinase-binding domain of NEMO is required for tumor necrosis factor alpha-induced NF-kappa B activity. Mol Cell Biol. 2006;26:9209–9219
   • View In Article
   • MEDLINE
   • CrossRef
10. Agou F, Traincard F, Vinolo E, Courtois G, Yamaoka S, Israel A, et al. The trimerization domain of NEMO is composed of the interacting C-terminal CC2 and LZ coiled-coil subdomains. J Biol Chem. 2004;279:27861–27869
    • View In Article
    • MEDLINE
    • CrossRef
11. Wu CJ, Conze DB, Li T, Srinivasula SM, Ashwell JD. Sensing of Lys 63-linked polyubiquitination by NEMO is a key event in NF-kB activation. [corrected] Nat Cell Biol. 2006;8:398–406
    • View In Article
    • MEDLINE
    • CrossRef
12. Martinez-Pomar N, Munoz-Saa I, Heine-Suner D, Martin A, Smahi A, Matamoros N. A new mutation in exon 7 of NEMO gene: late skewed X-chromosome inactivation in an incontinentia pigmenti female patient with immunodeficiency. Hum Genet. 2005;118:458–465
    • View In Article
    • MEDLINE
    • CrossRef
13. Kosaki K, Shimasaki N, Fukushima H, Hara M, Ogata T, Matsuo N. Female patient showing hypohidrotic ectodermal dysplasia and immunodeficiency (HED-ID). Am J Hum Genet. 2001;69:664–666
    • View In Article
    • MEDLINE
    • CrossRef
14. Schmid JM, Junge SA, Hossle JP, Schneider EM, Roosnek E, Seger RA, et al. Transient hemophagocytosis with deficient cellular cytotoxicity, monoclonal immunoglobulin M gammopathy, increased T-cell numbers, and hypomorphic NEMO mutation. Pediatrics. 2006;117:e1049–e1056
    • View In Article
    • CrossRef
15. Dupuis-Girod S, Corradini N, Hadj-Rabia S, Fournet JC, Faivre L, Le Deist F, et al. Osteopetrosis, lymphedema, anhidrotic ectodermal dysplasia, and immunodeficiency in a boy and incontinentia pigmenti in his mother. Pediatrics. 2002;109:e97
    • View In Article
    • CrossRef
16. Mansour S, Woffendin H, Mitton S, Jeffery I, Jakins T, Kenwrick S, et al. Incontinentia pigmenti in a surviving male is accompanied by hypohidrotic ectodermal dysplasia and recurrent infection. Am J Med Genet. 2001;99:172–177
    • View In Article
    • MEDLINE
    • CrossRef
17. Nishikomori R, Akutagawa H, Maruyama K, Nakata-Hizume M, Ohmori K, Mizuno K, et al. X-linked ectodermal dysplasia and immunodeficiency caused by reversion mosaicism of NEMO reveals a critical role for NEMO in human T-cell development and/or survival. Blood. 2004;103:4565–4572
    • View In Article
    • MEDLINE
    • CrossRef
18. Orstavik KH, Kristiansen M, Knudsen GP, Storhaug K, Vege A, Eiklid K, et al. Novel splicing mutation in the NEMO (IKK-g) gene with severe immunodeficiency and heterogeneity of X-chromosome inactivation. Am J Med Genet A. 2006;140:31–39
    • View In Article
    • MEDLINE
19. Smahi A, Courtois G, Vabres P, Yamaoka S, Heuertz S, Munnich A, et al. Genomic rearrangement in NEMO impairs NF-kappaB activation and is a cause of incontinentia pigmenti. The International Incontinentia Pigmenti (IP) Consortium. Nature. 2000;405:466–472
    • View In Article
    • MEDLINE
    • CrossRef
20. Risma K, Deering R, Monaco-Shawver L, Heltzer M, Burnham J, Niemela J, et al. Ectodermal dysplasia with immunodeficiency and lymphedema, but not osteopetrosis, is associated with a unique NF-κB essential modulator (NEMO) mutation. Clin Immunol. 2005;116:300–301
    • View In Article
21. Orange JS, Levy O, Geha RS. Human disease resulting from gene mutations that interfere with appropriate nuclear factor-kappaB activation. Immunol Rev. 2005;203:21–37
    • View In Article
    • MEDLINE
    • CrossRef
22. Niehues T, Reichenbach J, Neubert J, Gudowius S, Puel A, Horneff G, et al. Nuclear factor kappaB essential modulator-deficient child with immunodeficiency yet without anhidrotic ectodermal dysplasia. J Allergy Clin Immunol. 2004;114:1456–1462
    • View In Article
    • Abstract
    • Full Text
    • Full-Text PDF (197 KB)
    • CrossRef
23. Orange JS, Jain A, Ballas ZK, Schneider LC, Geha RS, Bonilla FA. The presentation and natural history of immunodeficiency caused by nuclear factor κB essential modulator mutation. J Allergy Clin Immunol. 2004;113:725–733
    • View In Article
    • Abstract
    • Full Text
    • Full-Text PDF (613 KB)
    • CrossRef
24. Salt BH, Niemela JE, Pandey R, Hanson EP, Deering RP, Quinones R, et al. IKBKG (nuclear factor-κB essential modulator) mutation can be associated with opportunistic infection without impairing TLR function. J Allergy Clin Immunol. 2007;121:976–982
    • View In Article
    • Abstract
    • Full Text
    • Full-Text PDF (362 KB)
    • CrossRef
25. Orange JS, Brodeur SR, Jain A, Bonilla FA, Schneider LC, Kretschmer R, et al. Deficient natural killer cell cytotoxicity in patients with IKK-gamma/NEMO mutations. J Clin Invest. 2002;109:1501–1509
    • View In Article
    • MEDLINE
    • CrossRef
26. Dai YS, Liang MG, Gellis SE, Bonilla FA, Scheider LC, Geha RS, et al. Characteristics of mycobacterial infection in patients with immunodeficiency and nuclear factor-kappaB essential modulator mutation, with or without ectodermal dysplasia. J Am Acad Dermatol. 2004;51:718–722
    • View In Article
    • Abstract
    • Full Text
    • Full-Text PDF (465 KB)
    • CrossRef
27. Filipe-Santos O, Bustamante J, Haverkamp MH, Vinolo E, Ku CL, Puel A, et al. X-linked susceptibility to mycobacteria is caused by mutations in NEMO impairing CD40-dependent IL-12 production. J Exp Med. 2006;203:1745–1759
    • View In Article
    • MEDLINE
    • CrossRef
28. Orange JS, Levy O, Brodeur SR, Krzewski K, Roy RM, Niemela JE, et al. Human nuclear factor kappa B essential modulator mutation can result in immunodeficiency without ectodermal dysplasia. J Allergy Clin Immunol. 2004;114:650–656
    • View In Article
    • Abstract
    • Full Text
    • Full-Text PDF (661 KB)
    • CrossRef
29. Temmerman ST, Ma CA, Borges L, Kubin M, Liu S, Derry JM, et al. Impaired dendritic-cell function in ectodermal dysplasia with immune deficiency is linked to defective NEMO ubiquitination. Blood. 2006;108:2324–2331
    • View In Article
    • MEDLINE
    • CrossRef
30. Kabelitz D, Medzhitov R. Innate immunity–cross-talk with adaptive immunity through pattern recognition receptors and cytokines. Curr Opin Immunol. 2007;19:1–3
    • View In Article
    • MEDLINE
    • CrossRef
31. Kabelitz D. Expression and function of Toll-like receptors in T lymphocytes. Curr Opin Immunol. 2007;19:39–45
    • View In Article
    • MEDLINE
    • CrossRef
32. Abinun M, Spickett G, Appleton AL, Flood T, Cant AJ. Anhidrotic ectodermal dysplasia associated with specific antibody deficiency. Eur J Pediatr. 1996;155:146–147
    • View In Article
    • MEDLINE
    • CrossRef
33. Cordier F, Vinolo E, Veron M, Delepierre M, Agou F. Solution structure of NEMO zinc finger and impact of an anhidrotic ectodermal dysplasia with immunodeficiency-related point mutation. J Mol Biol. 2008;377:1419–1432
    • View In Article
    • CrossRef
34. Nenci A, Becker C, Wullaert A, Gareus R, van Loo G, Danese S, et al. Epithelial NEMO links innate immunity to chronic intestinal inflammation. Nature. 2007;446:557–561
    • View In Article
    • CrossRef
35. Zaph C, Troy AE, Taylor BC, Berman-Booty LD, Guild KJ, Du Y, et al. Epithelial-cell-intrinsic IKK-beta expression regulates intestinal immune homeostasis. Nature. 2007;446:552–556
    • View In Article
    • CrossRef