nej mo a 1309199

original article

T h e n e w e ngl a nd j o u r na l o f m e dic i n e

n engl j med 369;26 nejm.org december 26, 20132504

Deficiency of Innate and Acquired Immunity Caused by an IKBKB Mutation

Ulrich Pannicke, Ph.D., Bernd Baumann, Ph.D., Sebastian Fuchs, M.Sc., Philipp Henneke, M.D., Anne Rensing-Ehl, M.D., Marta Rizzi, M.D., Ph.D.,

Ales Janda, M.D., Ph.D., Katrin Hese, Ph.D., Michael Schlesier, Ph.D., Karlheinz Holzmann, Ph.D., Stephan Borte, M.D., Constanze Laux,

Eva-Maria Rump, Alan Rosenberg, M.D., Teresa Zelinski, Ph.D., Hubert Schrezenmeier, M.D., Thomas Wirth, Ph.D., Stephan Ehl, M.D.,

Marlis L. Schroeder, M.D., and Klaus Schwarz, M.D.

From the Institute for Transfusion Medi-cine, University Hospital Ulm (U.P., C.L., H.S., K.S.), the Institute of Physiological Chemistry (B.B., T.W.), the Center for Bio-medical Research, Genomics Core Facility (K. Holzmann), University of Ulm, and the Institute for Clinical Transfusion Medi-cine and Immunogenetics Ulm, German Red Cross Blood Service Baden-Wuert-temberg-Hessen (E.-M.R., H.S., K.S.), Ulm; the Center of Chronic Immunodeficiency, University Medical Center Freiburg (S.F., P.H., A.R.-E., K. Hese, M.R., A.J., M.S., S.E.), the Faculty of Biology, University of Freiburg (S.F.), the Center for Pediatrics and Adolescent Medicine (P.H., S.E.), and the Department of Rheumatology and Clinical Immunology (M.S.), Univer-sity Hospital Freiburg, Freiburg; and the Translational Center for Regenerative Medicine, University of Leipzig, Leipzig (S.B.) — all in Germany; the Division of Clinical Immunology and Transfusion Med-icine, Department of Laboratory Medicine, Karolinska Institute, Karolinska University Hospital Huddinge, Stockholm (S.B.); and the Department of Pediatrics, University of Saskatchewan, Saskatoon (A.R.), and the Departments of Biochemistry and Medi-cal Genetics (T.Z.) and Pediatrics and Child Health (T.Z., M.L.S.), University of Manitoba, Winnipeg — both in Canada. Address reprint requests to Dr. Schwarz at [email protected].

Drs. Pannicke, Baumann, and Henneke and Mr. Fuchs and Drs. Ehl, Schroeder, and Schwarz contributed equally to this article.

N Engl J Med 2013;369:2504-14.DOI: 10.1056/NEJMoa1309199Copyright © 2013 Massachusetts Medical Society.

A BS TR AC T

Background

Severe combined immunodeficiency (SCID) comprises a heterogeneous group of heritable deficiencies of humoral and cell-mediated immunity. Many patients with SCID have lymphocyte-activation defects that remain uncharacterized.

Methods

We performed genetic studies in four patients, from four families of Northern Cree ancestry, who had clinical characteristics of SCID, including early onset of severe viral, bacterial, and fungal infections despite normal B-cell and T-cell counts. Genome-wide homozygosity mapping was used to identify a candidate region, which was found on chromosome 8; all genes within this interval were sequenced. Immune-cell populations, signal transduction on activation, and effector functions were studied.

Results

The patients had hypogammaglobulinemia or agammaglobulinemia, and their peripheral-blood B cells and T cells were almost exclusively of naive phenotype. Regulatory T cells and γδ T cells were absent. All patients carried a homozygous duplication — c.1292dupG in exon 13 of IKBKB, which encodes IκB kinase 2 (IKK2, also known as IKKβ) — leading to loss of expression of IKK2, a component of the IKK–nuclear factor κB (NF-κB) pathway. Immune cells from the patients had im-paired responses to stimulation through T-cell receptors, B-cell receptors, toll-like receptors, inflammatory cytokine receptors, and mitogens.

Conclusions

A form of human SCID is characterized by normal lymphocyte development despite a loss of IKK2 function. IKK2 deficiency results in an impaired response to activa-tion stimuli in a variety of immune cells, leading to clinically relevant impairment of adaptive and innate immunity. Although Ikk2 deficiency is lethal in mouse embryos, our observations suggest a more restricted, unique role of IKK2–NF-κB signaling in humans. (Funded by the German Federal Ministry of Education and Research and others.)

The New England Journal of Medicine Downloaded from nejm.org on November 17, 2015. For personal use only. No other uses without permission.

Copyright © 2013 Massachusetts Medical Society. All rights reserved.

Deficient Immunity Caused by an IKBKB Mutation

n engl j med 369;26 nejm.org december 26, 2013 2505

Severe combined immunodeficiency (SCID) is the most severe primary immuno-deficiency. Affected infants usually present

in the first months of life with Pneumocystis jirovecii pneumonia, bacterial sepsis, chronic cytomegalo-virus or candida infection, or persistent respira-tory or gastrointestinal viral infection, often as-sociated with protracted diarrhea and failure to thrive.1 Impaired T-cell immunity is the main immunologic abnormality in SCID, and most pa-tients have low numbers of T cells or none. How-ever, some patients may have normal T-cell counts with a severe immune-cell activation defect.2

Immune-cell activation involves complex sig-naling that regulates transcriptional programs. The nuclear factor κB (NF-κB) transcription fac-tors are key regulators of inflammatory and im-mune responses, mediating cell activation, pro-liferation, survival, and effector functions. The ubiquitously expressed IκB kinase (IKK) complex links these transcription factors to immune re-ceptors, including T-cell and B-cell receptors, toll-like receptors (TLRs), and inflammatory cy-tokine (tumor necrosis factor α [TNF-α] and interleukin-1β) receptors.3-5

The IKK complex encompasses the scaffold protein NF-κB essential modulator (NEMO, also known as IKKγ) and two other IKKs, IKK1 (IKKα) and IKK2 (IKKβ). After activation of the IKK complex, IKK2 primarily phosphorylates inhibi-tors of κB (IκBα, IκBβ, and IκBε), which are degraded by the proteasome after polyubiquiti-nation. The degradation of the inhibitor allows nuclear translocation of NF-κB and regulation of target-gene transcription.6,7 In addition to its cru-cial function in NF-κB signaling, IKK2 phosphor-ylates other substrates and is involved in the acti-vation of the transcription factor activator protein 1 (AP-1) in response to engagement of CD3 and CD28 coreceptors in CD4+ T cells.6,8,9

Several primary immunodeficiencies in hu-mans are caused by an impaired IKK–IκB axis. Hypomorphic mutations in IKBKG, the gene en-coding NEMO, lead to X-linked anhidrotic ecto-dermal dysplasia with immunodeficiency or to isolated immunodeficiency. Anhidrotic ectoder-mal dysplasia with immunodeficiency also re-sults from autosomal dominant mutations of NFKBIA, the gene encoding IκBα.10-16

In this article, we describe four patients who presented with SCID and had homozygous null

mutations of IKBKB. These patients had normal B-cell and T-cell counts but very low levels of immunoglobulins, as well as a severe defect in immune-cell activation that affected both innate and adaptive immune-receptor pathways.

Me thods

Study Patients

The families of the four patients were from Man-itoba and Saskatchewan in Canada and were of Northern Cree ancestry. No close family connec-tion among these patients was known. The par-ents of each patient gave written informed con-sent. The University of Manitoba Health Research Ethics Board approved the study. The clinical characteristics of the patients are summarized in Table S1 in the Supplementary Appendix, avail-able with the full text of this article at NEJM.org.

Patient 1 (from Family 1) became ill at 1 month of age with oral candidiasis and feeding difficul-ties. At 4 months of age, she had Escherichia coli septicemia, parainfluenza virus type 1 pneumonia, and persistent oral candidiasis. Laboratory test-ing showed hypogammaglobulinemia (Table S1 in the Supplementary Appendix), and SCID was diagnosed clinically. Her condition was stabilized with antimicrobial agents and intravenous im-mune globulin. She received a cord-blood trans-plant after reduced-intensity conditioning, but the graft was rejected. She had two episodes of pneumococcal bacteremia associated with osteo-myelitis, from which she recovered, but she died from Mycobacterium avium sepsis. There was no family history of immunodeficiency or death in early infancy.

Patient 2 (from Family 2) first presented at 6 weeks of age, when she required ventilator as-sistance because of upper-airway obstruction caused by severe oropharyngeal candidiasis. She had reduced IgM levels; detectable IgG was prob-ably of maternal origin (Table S1 in the Supple-mentary Appendix). Bacterial septicemia devel-oped, and her neurologic condition deteriorated rapidly; analysis of the cerebrospinal fluid re-vealed Listeria monocytogenes. A computed tomo-graphic (CT) scan of the head showed hemor-rhages in the posterior fossa, frontal subdural effusions, and bilateral middle-cerebral-artery in-farction. She died from this infection 2.5 months after birth. Autopsy revealed a small spleen and





thymus and no lymph nodes in the neck or mes-entery. Microscopic examination showed small periarterial lymphoid aggregates, which were devoid of germinal centers, in the spleen; plasma-cytoid cells were not noted. A scant lymphoid population was found in the thymus, largely lo-cated in the medulla; Hassall’s corpuscles were present, and there were numerous histiocytes in the cortex. The patient had one healthy sibling and no family history of immunodeficiency.

Patient 3 (from Family 3) was hospitalized at 5 months of age with pneumonia, from which he recovered. Oral candidiasis subsequently de-veloped. He was readmitted at 6 months of age with parainfluenza virus type 3 pneumonia, chronic diarrhea, and poor weight gain (Table S1 in the Supplementary Appendix). After myelo-ablative conditioning, he received a bone marrow transplant from his human leukocyte antigen–identical sister. After initial engraftment (97% do-nor CD3+ cells), mixed chimerism with 50% donor CD3+ cells developed; his condition has remained stable over a period of more than 24 months. After tuberculosis was diagnosed in his grand-father, the child’s gastric washings were found to be positive for acid-fast bacilli; he started receiv-ing antituberculosis therapy and has remained clinically well. His three older siblings were healthy; a fourth sibling died at 14 months of age as a result of chronic infections.

Patient 4 (from Family 4) presented at 3 months of age with oral and genitourinary candidiasis. Three months later, Serratia marcescens bacteremia with pancytopenia developed. She had agamma-globulinemia (Table S1 in the Supplementary Appendix) and was given intravenous immune globulin. At 6 months of age, focal seizures de-veloped. Magnetic resonance imaging suggested the presence of infarcts in both hemispheres of the brain. Eye deviation developed; CT scanning showed a right frontal infarct, a deep thalamic hemorrhage, and large subdural collections that were consistent with abscess formation and hemorrhage. Four weeks later, blood cultures were positive for E. coli, klebsiella, and serratia. The bacteremia resolved after surgical drainage of the subdural abscess. At 8 months of age, she received an unrelated-donor cord-blood trans-plant after myeloablative conditioning. She had slow T-cell engraftment. Four months after transplantation, she had mixed chimerism with 85% donor CD3+ cells. She has remained clini-cally well except for neurologic sequelae.

Genetic and Other Analyses

The methods for immunologic phenotyping, ge-netic and protein analyses, and functional inves-tigations are described in detail in the Supple-mentary Appendix.

R esult s

Basic Immunologic Phenotype

All four patients had hypogammaglobulinemia or agammaglobulinemia but normal T-cell counts in the absence of maternal engraftment. One pa-tient had slightly reduced B-cell counts, and three patients had reduced natural killer (NK) cell counts. The patients’ immunophenotypes at pre-sentation are summarized in Table S1 in the Supplementary Appendix.

Autosomal Recessive IKBKB Mutation

To elucidate the molecular basis of this unusual immunodeficiency, we assumed autosomal re-cessive inheritance of a founder allele and con-ducted homozygosity mapping in four patients and three healthy siblings (Fig. 1A). Overlapping homozygosity was found in only one candidate region, spanning 11.6 Mbp on chromosome 8 (Fig. 1B). All 40 genes within this region were sequenced with the use of genomic DNA or com-plementary DNA. The only biallelic mutation identified in all four patients was a homozygous duplication, c.1292dupG in exon 13 of the IKBKB gene (Fig. 1C). No other potentially disease- causing mutations were detected. Genotyping of healthy family members revealed carriers of this mutation but no other homozygotes (Fig. 1A). The mutation is not listed in dbSNP (www.ncbi.nlm .nih.gov/projects/SNP) or in the 1000 Genomes database (www.1000genomes.org). This frame-shift mutation leads to a premature stop codon and results in a complete loss of IKK2 protein in fibroblasts and peripheral-blood mononuclear cells (Fig. 1D, 1E, and 1F). Western blot analysis and immunoprecipitation after stimulation with TNF-α or mitogen phorbol 12-myristate 13-acetate (PMA) revealed reduced amounts of NEMO and IKK1 despite normal messenger RNA (mRNA) levels (Fig. 1E and 1F, and Fig. S1 and S7C in the Supplementary Appendix), which suggested that the lack of IKK2 affects the stability of other members of the IKK complex. To determine whether the IKBKB mutation is restricted to pa-tients of Northern Cree ancestry, we sequenced the IKBKB gene in six white patients with a simi-





E Western Blot of Fibroblasts F Western Blot of PBMCs

A Family Pedigrees B Candidate Gene Interval on Chromosome 8

C IKBKB Genomic DNA Sequence

D IKK2 Protein

?

?

?

IKK2 1

15

Kinase LZ HLH NBD

312 458 566 645 745

756

735479

p.GIn432Profs*62AA A A A AGG G GGGGCC C CC CGG GG G GGGGG GG GTTTTTT T TGG

428Lys Val Trp Trp TrpVal ValGlnGly GlyLys

429 430 431 432 433 434 428 429 430 431 432

c. 1292dupG

Pro Gly Leu433 434

100 — 170 —

95 — 72 — 55 —

43 —

34 —

26 —

130 130 95 — 72 —

55 —

43 —

34 —

26 —

170 —

75 —

50 —

37 —

25 —

20 —

15 —

10 —

50 —

37 —

Control

Control

Patien

tContro

l

Patien

t

Control

Patien

t

Patien

t

Control

Patien

t

Untreated TNF-α PMA

IKK2 IKK2IKK2

IKK1 IKK1

NEMO NEMO

ERK2

p65

β-Actinβ-Actin

11.6 Mbp

IKBKB (p11.21)

rs1356754chr8: 50421549chr8: 39816369

rs6984521

Exon 13

Control Patient

Family 1 Family 2 Family 3 Family 4

Patient 1 Patient 2 Patient 4Patient 3

kDkDkD

Figure 1. Genetic Analysis of an IKBKB Mutation in Four Families.

Panel A shows the pedigrees of four families of Northern Cree ancestry with immunodeficient children (Patients 1, 2, 3, and 4). Squares denote male family members, circles female members, and slashes deceased family members. The four patients (solid arrows) and three healthy siblings (open arrows) were included in homozygosity mapping analysis. Solid symbols denote family members who were homozygous for the IKBKB mutation, half-solid symbols members who were heterozygous for the mutation, and open symbols members who did not have the mutation; family members not tested genetically are indicated by question marks. Panel B shows the candidate interval on chromosome 8 (NCBI36, hg18 assembly) encompassing the mutated gene. Delimiting single-nucleotide polymorphisms are indicated. The region spans 40 genes, including IKBKB at p11.21. Panel C depicts the IKBKB sequence analysis in one patient as compared with that in a healthy control; the sequence has a homozygous G duplica-tion in exon 13, leading to a frameshift at the protein level, starting at amino acid 432. Panel D shows the predicted loss of most of the α-helical scaffold dimerization domain encompassing the leucine-zipper (LZ) and helix-loop-helix (HLH) domains and the nuclear factor κB (NF-κB) essential modulator (NEMO)–binding domain (NBD) of IκB kinase 2 (IKK2), with no effects on the kinase domain or the ubiquitin-like domain (amino acids 307 to 384). Panel E shows the Western blot analysis of primary fibroblast–derived protein from Patient 1, which had a complete lack of the IKK2 protein. Amounts of IκB kinase 1 (IKK1), NEMO, and p65 are reduced as compared with control fibroblasts that were untreated or stimulated with TNF-α or phorbol 12-myristate 13-acetate (PMA). β-Actin served as a loading control. Panel F shows the Western blot analysis of peripheral-blood mononuclear cells (PBMCs) from Patient 1, which also lack the IKK2 protein. The extracellular signal-regulated kinase 2 (ERK2) protein level served as a loading control.





lar immunophenotype. No mutations in IKBKB were identified in these patients.

Impaired NF-κB Signaling

Many receptors involved in innate and adaptive immunity, including antigen receptors, TLRs, and inflammatory cytokine receptors, signal by means of the canonical NF-κB pathway, of which IKK2 is a central component. IKK2-deficient pa-tient fibroblasts showed impaired phosphoryla-tion of IκBα in response to TNF-α stimulation (Fig. 2A). Degradation of IκBα after interleukin-1β stimulation was marginally affected, whereas degradation in response to TLR5 stimulation by flagellin was absent, indicating distinct require-ments for IKK2 (Fig. S2 in the Supplementary Appendix). The reduced but not absent effect of interleukin-1β on IκBα degradation mirrors ob-servations in previous studies of fibroblasts with prominent NEMO reduction.16 We found that NF-κB binding to DNA after TNF-α stimulation was considerably decreased in patient cells but that the NF-κB heterodimer composition (p50–p65) was similar to that in control cells. Similar re-sults were obtained for PMA stimulation (Fig. 2B, and Fig. S3 in the Supplementary Appendix). NF-κB DNA binding in response to TNF-α was reconsti-tuted even by weak IKK2 expression in transfected fibroblasts (Fig. 2C, and Fig. S4 in the Supple-mentary Appendix), confirming that the sig nal-ing impairment in cells from the patients was

due to the specific loss of IKK2. Normal TNF-α–mediated NF-κB DNA binding under conditions of small interfering RNA suppression of IKK2 has been reported.17

We analyzed NF-κB–mediated gene expression in fibroblasts from the patients. Interleukin-6 pro-duction in response to TNF-α or interleukin-1β was normal, whereas it was reduced in response to lipopolysaccharide (which acts through TLR4) or flagellin, suggesting that interleukin-6 pro-duction is variably dependent on IKK2 (Fig. 2D). In NF-κB/AP-1–dependent luciferase reporter as-says, responses to TNF-α, interleukin-1β, lipo-polysaccharide, and flagellin were reduced in IKK2-deficient fibroblasts, as compared with control transfectants, whereas the response to polyinosinic–polycytidylic acid (through TLR3) was not reduced (Fig. S5 in the Supplementary Appendix). The varying responses to TNF-α and interleukin-1β may reflect differences in pro-moter availability, in the number of NF-κB bind-ing sites, or in co–transcription factor depen-dence in the particular assays.

Similar results in studies of protein stability, NF-κB–binding activity induced by TNF-α and PMA, and IκBα phosphorylation were obtained with human IKK2 knockdown fibroblasts (Fig. S6 in the Supplementary Appendix). Inter-leukin-6 production triggered by interleukin-1β was not affected in IKK2 knockdown cells (data not shown).

Figure 2 (facing page). Basal and Induced IκB Kinase (IKK) and NF-κB Activity in Cell Lines Derived from Patient 1 and a Control.

Panel A shows IKK complex activity in primary skin fibroblasts. IKK complex activity was measured before and after treatment for 15 minutes with tumor necrosis factor α (TNF-α; final concentration, 50 ng per milliliter), with the use of anti-NEMO–specific immunoprecipitated IKK complexes of whole-cell lysates (0.5 mg of protein). IKK activity was determined with glutathione S-transferase (GST)–IκBα as a substrate. Basal and TNF-α–induced IKK activity is dramatically reduced in cells from the patient. Panel B shows NF-κB activity in primary skin fibro-blasts, measured with the use of an electrophoretic mobility shift assay (EMSA) for NF-κB and specificity protein 1 (SP1). Whole-cell lysates (10 μg) were isolated from untreated cells, cells treated with TNF-α for 15 minutes (final concentration, 50 ng per milliliter), and cells treated with PMA for 30 minutes (final concentration, 100 ng per milliliter). NF-κB indicates NF-κB–specific DNA binding. NF-κB activation by TNF-α is reduced to a great extent but still present in cells from the patient; basal and PMA-induced NF-κB activity is almost absent. SP1 DNA binding served as a quality control. NS denotes nonspecific band. Panel C shows NF-κB activation reconstituted in primary cells from the patient on reexpression of IKK2. The patient’s cells were transiently transfected with either an IKK2 expression vector or empty vector (Vec). After 48 hours, transfected and nontransfected cells were stimulated with TNF-α for 15 minutes (final concentration, 50 ng per milliliter), and whole-cell lysates were evaluated with the use of EMSA (for NF-κB and SP1) and Western blot analysis (for IKK2, NEMO, and ERK2). NF-κB indicates NF-κB–specific DNA binding. Normal levels of NF-κB activation by TNF-α were restored in cells from the patient that reexpressed IKK2. SP1 DNA binding served as a quality control. Densitometric analysis of NF-κB DNA-binding activity and IKK2 and NEMO protein levels was performed with the use of ImageJ software (National Institutes of Health), with SP1 and ERK2 signals used as loading controls. The num-bers indicate signal intensity, with the signal intensity in untreated samples from the patient set at 1.0. Panel D shows interleukin-6 production in simian vacuolating virus 40 (SV40)–immortalized skin fibroblasts from the patient and a control. The results are shown before treatment and after treatment with TNF-α (final concentration, 30 ng per milliliter), interleukin-1β (final concentration, 5 ng per milliliter), lipopoly-saccharide (LPS) (final concentration, 10 μg per milliliter), or flagellin (final concentration, 2.5 μg per milliliter) for 16 hours. Interleukin-6 was measured with the use of an enzyme-linked immunosorbent assay (ELISA). Data are shown for one representative experiment out of five inde-pendent experiments; duplicate measurements were performed for each assay (intraassay variance, <15%).





Quantitative reverse-transcriptase–polymerase-chain-reaction analyses of NF-κB target genes showed similar or only slightly reduced CCL2, ICAM1, NFKB2, RELB, TNF, TNFAIP3, and VIM mRNA levels in stimulated IKK2-deficient fibroblasts, whereas levels of other RNAs — most notably, cer-tain chemokine mRNAs — were markedly lower in response to TNF-α stimulation (BCL2A1, CCL5, CXCL3, CXCL10, TRAF1, and C3) or PMA stimula-tion (CXCL10, TRAF1, and MMP9) (Fig. S7A in the Supplementary Appendix). These results indicate selective dependence of the regulation of NF-κB target genes on IKK2 function.

To exclude the possibility that the reduced cellular responses were caused by inadequate receptor expression, we measured levels of mRNA

encoding the two TNF receptors, the interleukin-1 receptor chains (interleukin-1R1 and interleukin-1RAP), and TLR4 and TLR5, all of which were normal (Fig. S7B in the Supplementary Appendix). Moreover, impaired NF-κB signaling was not due to low NEMO or IKK1 mRNA levels. In con-trast, IKK2 mRNA levels were significantly de-creased, possibly reflecting reduced mRNA sta-bility due to nonsense-mediated decay (Fig. S7C in the Supplementary Appendix).

Lymphocyte Differentiation and Activation Defect

To address the question of whether IKK2 defi-ciency affects T-cell development, we measured T-cell–receptor excision circle (TREC) levels in

D Induction of Endogenous Interleukin-6 Production in Fibroblasts

B NF-κB Binding to DNA

A Kinase Assay C Reconstitution of NF-κB Binding to DNA

Inte

rleu

kin-

6 (n

g/m

l)

30

10

20

0TNF-α Interleukin-1β LPS FlagellinUntreated

TNF-αUntreated

TNF-α

Control Patient

Control

Patien

t

Control

Patien

t

TNF-αUntreated

Control

Patien

t

Control

Patien

t

PMA

Control

Patien

t

GST–IκBα

NF-κB

NF-κB

NS

SP1

EMSA

WesternBlot

SP1

IKK2

NEMO

ERK2

NS

IKK2 Vec Nontransfected

Patient Patient Patient Control

− + − + − +− +

1.5 3.2 0.8 1.9 1.0 3.11.0 1.2

2.8 2.7 1.7 1.6 9.1 11.41.0 1.2

1.3 1.5 1.6 1.5 2.2 3.11.0 1.0





two patients with the use of the dried blood spots from original Guthrie cards used to screen for congenital defects in newborns. TRECs are small pieces of DNA excised from newly pro-duced T cells during the rearrangement of their receptor genes, and TREC levels are a measure of T-cell production. Both patients had normal TREC levels (Table S1 and Fig. S8 in the Supplementary Appendix). Moreover, the T-cell receptor beta-chain variable region repertoire was largely nor-mal, with minor deviations among CD8+ T cells (Fig. S9A in the Supplementary Appendix). How-ever, γδ T cells were absent in the patient ana-lyzed (Fig. S9B in the Supplementary Appendix). Even when patients had a history of serious in-fection, 93% of their CD4+ T cells expressed CD45RA, and 85% of these cells expressed CD27 (data not shown); CD45RA and CD27 are markers of antigen-naive T cells. In addition, CD8+ T cells in these patients were almost exclusively naive (CD45RA+CCR7+), whereas more differentiated CD8+ T cells were readily detectable in age-matched controls (Fig. S9C in the Supplementary Appendix).18,19 CD25highFOXP3+CD4+ regulatory T cells were almost absent in the two patients (Fig. S9D in the Supplementary Appendix). On stimulation with anti-CD3 and anti-CD28 anti-bodies, up-regulation of the T-cell activation mark-ers CD25 and CD69 was reduced but detectable (Fig. 3A). Proliferation responses were assessed with the use of different techniques (Fig. 3B, and Fig. S9E in the Supplementary Appendix). Phyto-hemagglutinin mitogenic responses were moder-ately reduced, and there was little response to soluble or plate-bound anti-CD3 antibodies. However, stimulation with anti-CD3 and anti-CD28 antibody beads produced a surprisingly strong response that was only mildly reduced as compared with the control response. Consistent

with this observation was our finding that the proportion of patient T cells producing interleu-kin-2 in response to stimulation with PMA and ionomycin was similar to that in cells from age-matched controls, whereas interferon-γ produc-tion was relatively low (Fig. 3C).

Stimulation with the erythroleukemia cell line K562 did not induce CD107a degranulation of NK cells from the patients (Fig. S10A in the Supplementary Appendix), an observation that is compatible with the requirement of NF-κB acti-vation for NK-cell cytotoxicity. Moreover, stimu-lation with K562 induced chemokine (C-C mo-tif) ligand 4 (CCL4) expression but did not elicit interferon-γ production (Fig. S10B in the Supple-mentary Appendix).

B-cell development in two patients, assessed as levels of kappa-deleting recombination exci-sion circles (KRECs, indicating production of new B cells) in dried blood spots from Guthrie cards, was within the range of values in controls (Table S1 and Fig. S8 in the Supplementary Ap-pendix). B cells were almost exclusively naive, with a normal proportion of transitional cells (CD38+IGM+) and a lack of class-switched mem-ory B cells (CD27+IGD−) and plasmablasts (CD38+CD20−) (Fig. S11A in the Supplementary Appendix), which were present in age-matched controls.18,19

Germinal-center B-cell activation is depen-dent on costimulatory signals and cytokines such as CD40L and interleukin-21, provided by follicular helper T cells.20,21 In line with the de-fect observed in peripheral blood from our pa-tients, their B cells did not differentiate into plasmablasts when stimulated with CD40L and interleukin-21, whereas cord-blood B cells, simi-larly composed of transitional and naive B cells, generated 5 to 10% CD27highCD38high plasmablast

Figure 3 (facing page). Phenotypic and Functional Characterization of T Cells, Natural Killer Cells, and B Cells from Patients 3 and 4.

Panel A shows surface expression of CD25 and CD69 on CD4+ T cells from Patient 4 before and after stimulation with anti-CD3 and anti-CD28 antibody beads. One control assay was performed with adult cells in parallel to the patient assay (parallel control), whereas age-matched control assays were performed on separate occasions. Panel B shows proliferative T-cell responses in Patients 3 and 4, measured with the use of bromodeoxyuridine incorporation after stimulation with phytohemagglutinin (PHA), soluble anti-CD3 (sCD3), or plate-bound anti-CD3 antibody (pbCD3). Panel C shows interleukin-2 and interferon-γ expression of CD45RO+CD4+ T cells on stimulation with PMA and iono-mycin, as determined with the use of flow cytometry. The percentage of CD45RO+ cells among CD4+ T cells was 7.5% in Patient 4, 23% in the parallel control, and 1.4% in the 1-month-old control. Panel D shows in vitro B-cell differentia-tion in Patient 4 and in a control. Magnetically isolated B cells from the patient and from cord blood obtained for the control were stimulated in vitro with CD40 ligand (CD40L) with or without interleukin-21 for 6 days. The gate indi-cates newly formed plasmablasts.





CD

25+

CD

69+

(%)

100

50

0

Inte

rleu

kin

2+ (%

)

80

40

0

Inte

rfer

on γ

+ (%

)

20

10

0

60

20

15

5

Untreated

Relative Europium Fluorescence

sCD3

pbCD3

0 10 20 30 40

C T-Cell Cytokine Expression

D In Vitro B-Cell Differentiation

A Early T-Cell Activation B T-Cell Proliferation

Patient 3 (6 mo)

Adult parallel control


Untreated

Relative Europium Fluorescence

sCD3

PHA

pbCD3

0 20 40 60 80

Patient 4 (6 mo)



Untreated

CD25

Patient 4(7 mo)

ParallelControl(adult)

Control (1 mo)

CD3–CD28

Gate: CD45+CD4+

Patient 4 (7 mo)

CD

690.9% 0.5%

96.0% 2.7%

100100

104

104 100100

104

104

100100

104

104 100100

104

104

100100

104

104

100100

104

104 100100

104

104 100100

104

104

100100

104

104

8.5% 12.3%

76.8% 2.3%

0.6% 0.1%

93.9% 5.5%

17.3% 65.7%

14.2% 2.8%

0.6% 0.1%

93.7% 5.7%

7.1% 82.8%

9.3% 0.8%

Interferon-γ Gate: CD45RO+CD4+

Inte

rleu

kin-

2

CD

38+

38% 3.9%

CD40L

CD27+

CD40L+Interleukin-21

Parallel Control (adult)

44% 23%

Control (4 mo)

42% 12%

Patient 4 (7 mo)

Control (cord)

Gate: CD19+

0.0% 0.23%

0.0% 5.54%

Age-matchedcontrols

Patient 4

Patient 4 (fresh cells)

Patient 3 (frozen cells)

Age-matched control(frozen cells)

Age-matched controls(fresh cells)

105

104

103

102

102 103 104 105 102 103 104 105

102 103 104 105 102 103 104 105

105

104

103

102

105

104

103

102

105

104

103

102





cells after 9 days of cultivation (Fig. 3D). In B cells, proliferation is required for differentiation.22 B cells from our patients did not proliferate in response to CD40L and interleukin-21 (Fig. S11B in the Supplementary Appendix), because they did not up-regulate interleukin-21 receptor in response to CD40L stimulation (Fig. S11C in the Supplementary Appendix). In contrast, prolifera-tion was observed in the B cells when they were stimulated by means of the B-cell antigen recep-tor and TLR9. For TLR9 stimulation, CpG, a cytosine–guanine dinucleotide polymer, was used (Fig. S11B in the Supplementary Appendix). B cells from the patients produced no immunoglobulins on stimulation with CD40L and interleukin-21 or anti-IgM antibody and CpG (Fig. S11D in the Supplementary Appendix). Furthermore, activa-tion markers correlating with B-cell function, such as CD69 (location in the germinal center), CD95 (selection in the germinal center), and CD86 (antigen presentation), were differentially expressed on activation in one of the patients, as compared with a control (Fig. S11E in the Sup-plementary Appendix).

Discussion

In this study, we examined four immunodefi-cient infants with an autosomal recessive IKBKB mutation resulting in a complete loss of IKK2 protein expression. All patients presented with early-onset, life-threatening bacterial, fungal, and viral infections and failure to thrive, conditions that are consistent with a clinical diagnosis of SCID.

Our immunologic and functional investigations revealed that IKK2 is mostly expendable for the development of B and T lymphocytes but is nec-essary for the differentiation of regulatory and γδ T cells and possibly also NK cells. A lack of memory and regulatory T cells has been re-ported in conditional mouse models of Ikk2 deficiency, but the absence of γδ T cells was not anticipated from these studies.23,24 The observa-tion that γδ T cells are absent in humans with autosomal dominant mutations in IκBα corrob-orates our findings.12 Impairment of intrathy-mic lymphotoxin β receptor and CD27-mediated activation of NF-κB during the development of γδ T cells may account for their absence.25 It is notable that in Ikk2 conditional knockout mice in which Ikk2 was deleted only in B cells, the relative frequency of B cells was significantly reduced in peripheral blood.26,27

IKK2 is indispensable for transmitting sig-nals by various surface receptors. Thus, the acti-vation of T, B, and NK cells in our patients was significantly reduced, which accounted for the defective immunoglobulin production. The re-sidual proliferative and cytokine response of T cells was higher than expected for infants with such a severe and early clinical presentation. Moreover, although all four patients had severe oral candidiasis compatible with a T-cell deficien-cy, characteristic pneumocystis infections were not observed in our four patients. Our finding of an impaired response to TNF-α, as well as to TLR4 or TLR5 stimulation, indicates an additional in-nate immunologic defect in these patients. In fact, bacterial sepsis and mycobacterial infection are seen in patients with hypomorphic muta-tions in IKBKG, the X-linked gene encoding NEMO.10 Although the clinical presentation in the context of the B-cell and T-cell defects justi-fies the classification of IKK2 deficiency as SCID, it is likely that the innate immunologic defect contributes substantially to the severity of the phenotype. Overall, our findings suggest that IKK2 deficiency, in addition to deficiencies in ORAI1 (a calcium-channel protein) and STIM1 (a stromal interaction molecule that is a calcium sensor), is a molecular cause of SCID character-ized by a normal number of T cells that fail to be activated.2

Mutations in genes encoding the two other components of the human IKK complex have been described. Mutations in CHUK, encoding IKK1, result in the autosomal-recessive lethal cocoon syndrome, characterized by fetal encase-ment and multiple malformations.28 These de-velopmental defects were absent in our patients. The immunologic phenotypes of patients with NEMO deficiency are accompanied in about 90% of cases by anhidrotic ectodermal dysplasia and occasionally by osteopetrosis and lymphedema.

Patients with IκBα mutations always have anhi-drotic ectodermal dysplasia.10 In contrast, we have not observed any of the developmental symptoms associated with deficiencies in IKK1, NEMO, and IκBα in our IKK2-deficient patients. An impaired NEMO–IKK2 axis is therefore not the cause of these nonimmunologic symptoms of NEMO defi-ciency, suggesting that ectodysplasin A–receptor signaling may not depend on IKK2.

In contrast to the effects we observed in hu-mans, a complete Ikbkb knockout in mice causes embryonic death at approximately day 13 as a





result of extensive TNF-α–triggered apoptotic liver damage.29,30 Ikk2−/− mice can be rescued by inactivation of the TNF receptor 1 gene. In our patients, results of sequence and expression analy-ses of TNF-α and TNF receptor chain genes were normal, except for a heterozygous TNFRSF1B mu-tation (p.Arg41Pro) in Patient 1 (data not shown). This suggests that the survival of IKK2-deficient patients is independent of a disruption of the TNF-α–TNF receptor signaling pathway. Severe liver damage was not present in our patients.

Canonical IKK2–NF-κB signaling plays a cru-cial role in the regulation of inflammation and oncogenesis. In the past, efforts have been made to detect and develop IKK2 inhibitors for thera-peutic purposes.31,32 The fact that complete Ikk2 deficiency is lethal in mouse embryos challenged this strategy. The cases of immunodeficiency that we describe suggest that a complete loss of IKK2 is not lethal in all humans. However, on

the basis of these cases, it remains possible that the consequences of IKK2 inhibition would be severe.

Supported by grants from the German Federal Ministry of Education and Research (BMBF 01GM1111F, to Dr. Schwarz; BMBF 01EO0803, to Drs. Ehl, Henneke, and Schwarz; and BMBF 1315883, to Dr. Borte, for the newborn screening assay), an op-erating grant from the Winnipeg Rh Institute Foundation (to Dr. Zelinski), an unrestricted fellowship grant from the European Society for Immunodeficiencies provided by Baxter (to Dr. Janda), and a grant from the Helmholtz Alliance Preclinical Compre-hensive Cancer Center (to Dr. Wirth).

Disclosure forms provided by the authors are available with the full text of this article at NEJM.org.

We thank the family members for participating in this study; Drs. Cheryl Greenberg and Margaret Fast, Department of Pedi-atrics and Child Health, University of Manitoba, for their sup-port and assistance in the studies of the patients and for their review of an earlier draft of the manuscript; the medical and nursing staff for their assistance in the care of the patients; Ute Leschik and Monika Haeffner for technical assistance; Gail Coghlan for preliminary SNP analysis; and Ralf Marienfeld for providing the short hairpin RNA plasmids.

We dedicate this article to the memory of Dr. Walter H. Hitzig, one of the founders of European pediatric immunology.

References

1. Dvorak CC, Cowan MJ, Logan BR, et al. The natural history of children with severe combined immunodeficiency: base-line features of the first fifty patients of the Primary Immune Deficiency Treatment Consortium prospective study 6901. J Clin Immunol 2013 July 12 (Epub ahead of print).2. Notarangelo LD. Functional T cell im-munodeficiencies (with T cells present). Annu Rev Immunol 2013;31:195-225.3. Israël A. The IKK complex, a central regulator of NF-kappaB activation. Cold Spring Harb Perspect Biol 2010;2(3): a000158.4. Oeckinghaus A, Ghosh S. The NF-kappaB family of transcription factors and its regulation. Cold Spring Harb Per-spect Biol 2009;1(4):a000034.5. Oeckinghaus A, Hayden MS, Ghosh S. Crosstalk in NF-κB signaling pathways. Nat Immunol 2011;12:695-708.6. Perkins ND. Integrating cell-signalling pathways with NF-kappaB and IKK func-tion. Nat Rev Mol Cell Biol 2007;8:49-62.7. Hayden MS, Ghosh S. Shared princi-ples in NF-kappaB signaling. Cell 2008; 132:344-62.8. Scheidereit C. IkappaB kinase com-plexes: gateways to NF-kappaB activation and transcription. Oncogene 2006;25: 6685-705.9. Lupino E, Ramondetti C, Piccinini M. IκB kinase β is required for activation of NF-κB and AP-1 in CD3/CD28-stimulated primary CD4(+) T cells. J Immunol 2012; 188:2545-55.10. Picard C, Casanova JL, Puel A. Infec-tious diseases in patients with IRAK-4, MyD88, NEMO, or IκBα deficiency. Clin Microbiol Rev 2011;24:490-7.11. Zonana J, Elder ME, Schneider LC, et

al. A novel X-linked disorder of immune deficiency and hypohidrotic ectodermal dysplasia is allelic to incontinentia pig-menti and due to mutations in IKK-gamma (NEMO). Am J Hum Genet 2000;67:1555-62.12. Courtois G, Smahi A, Reichenbach J, et al. A hypermorphic IkappaBalpha mu-tation is associated with autosomal domi-nant anhidrotic ectodermal dysplasia and T cell immunodeficiency. J Clin Invest 2003;112:1108-15.13. Janssen R, van Wengen A, Hoeve MA, et al. The same IkappaBalpha mutation in two related individuals leads to completely different clinical syndromes. J Exp Med 2004;200:559-68.14. McDonald DR, Mooster JL, Reddy M, Bawle E, Secord E, Geha RS. Heterozygous N-terminal deletion of IkappaBalpha re-sults in functional nuclear factor kappaB haploinsufficiency, ectodermal dysplasia, and immune deficiency. J Allergy Clin Im-munol 2007;120:900-7.15. Lopez-Granados E, Keenan JE, Kinney MC, et al. A novel mutation in NFKBIA/IKBA results in a degradation-resistant N-truncated protein and is associated with ectodermal dysplasia with immuno-deficiency. Hum Mutat 2008;29:861-8.16. Mooster JL, Cancrini C, Simonetti A, et al. Immune deficiency caused by impaired expression of nuclear factor-kappaB essen-tial modifier (NEMO) because of a muta-tion in the 5′ untranslated region of the NEMO gene. J Allergy Clin Immunol 2010;126(1):127.e7-132.e7.17. Jain G, Voogdt C, Tobias A, et al. IκB kinases modulate the activity of the an-drogen receptor in prostate carcinoma cell lines. Neoplasia 2012;14:178-89.

18. Morbach H, Eichhorn EM, Liese JG, Girschick HJ. Reference values for B cell subpopulations from infancy to adult-hood. Clin Exp Immunol 2010;162:271-9.19. van Gent R, van Tilburg CM, Nibbelke EE, et al. Refined characterization and reference values of the pediatric T- and B-cell compartments. Clin Immunol 2009; 133:95-107. [Erratum, Clin Immunol 2010; 136:307-8.]20. Shapiro-Shelef M, Calame K. Regula-tion of plasma-cell development. Nat Rev Immunol 2005;5:230-42.21. Avery DT, Deenick EK, Ma CS, et al. B cell-intrinsic signaling through IL-21 receptor and STAT3 is required for estab-lishing long-lived antibody responses in humans. J Exp Med 2010;207:155-71.22. Hodgkin PD, Lee JH, Lyons AB. B cell differentiation and isotype switching is related to division cycle number. J Exp Med 1996;184:277-81.23. Schmidt-Supprian M, Tian J, Ji H, et al. I kappa B kinase 2 deficiency in T cells leads to defects in priming, B cell help, germinal center reactions, and homeo-static expansion. J Immunol 2004;173: 1612-9.24. Schmidt-Supprian M, Courtois G, Tian J, et al. Mature T cells depend on signaling through the IKK complex. Im-munity 2003;19:377-89.25. Gerondakis S, Banerjee A, Grigoriadis G, et al. NF-κB subunit specificity in he-mopoiesis. Immunol Rev 2012;246:272-85.26. Li ZW, Omori SA, Labuda T, Karin M, Rickert RC. IKK beta is required for pe-ripheral B cell survival and proliferation. J Immunol 2003;170:4630-7.27. Pasparakis M, Schmidt-Supprian M, Rajewsky K. IkappaB kinase signaling is





essential for maintenance of mature B cells. J Exp Med 2002;196:743-52.28. Lahtela J, Nousiainen HO, Stefanovic V, et al. Mutant CHUK and severe fetal en-casement malformation. N Engl J Med 2010;363:1631-7.29. Li Q, Van Antwerp D, Mercurio F, Lee KF, Verma IM. Severe liver degeneration in

mice lacking the IkappaB kinase 2 gene. Science 1999;284:321-5.30. Tanaka M, Fuentes ME, Yamaguchi K, et al. Embryonic lethality, liver degenera-tion, and impaired NF-kappa B activation in IKK-beta-deficient mice. Immunity 1999;10:421-9.31. Schmid JA, Birbach A. IkappaB kinase

beta (IKKbeta/IKK2/IKBKB) — a key mol-ecule in signaling to the transcription fac-tor NF-kappaB. Cytokine Growth Factor Rev 2008;19:157-65.32. Gilmore TD, Herscovitch M. Inhibitors of NF-kappaB signaling: 785 and counting. Oncogene 2006;25:6887-99.Copyright © 2013 Massachusetts Medical Society.



nej mo a 1309199

Documents