Alu-Repeat–Induced Deletions Within the NCF2 Gene
Causing p67-phox–Deficient Chronic Granulomatous
Marcus Gentsch,1yAneta Kaczmarczyk,2yKarin van Leeuwen,3,13Martin de Boer,3,13Magdalena Kaus-Drobek,2,12
Marie-Claire Dagher,4Petra Kaiser,5Peter D. Arkwright,6Manfred Gahr,1Angela Ro ¨sen-Wolff,1Matthias Bochtler,2,12
Elizabeth Secord,7Pamela Britto-Williams,7Gulam Mustafa Saifi,8Anne Maddalena,8Ghassan Dbaibo,9,14
Jacinta Bustamante,10,11Jean-Laurent Casanova,10,11Dirk Roos,3yand Joachim Roesler,1?y
1Department of Pediatrics, University Hospital Carl Gustav Carus, Dresden, Germany;2Structural Biology Laboratory, International Institute of
Molecular and Cell Biology, Warszawa, Poland;3Sanquin Research, Amsterdam, The Netherlands;4Centre Diagnostic et Recherche CGD, TIM-C
Imag, Centre National de la Recherche Scientifique (CNRS), Universite ´ Joseph Fourier, Grenoble, France;5Professor Hess Kinderklinik, Klinikum
Bremen-Mitte, Bremen, Germany;6Child Health, Division of Translational Medicine, University of Manchester, Manchester, United Kingdom;
7Allergy/Immunology Division, Children’s Hospital of Michigan, Wayne State University, Detroit, Michigan;8GeneDx, Gaithersburg, Maryland;
9Department of Pediatrics, American University of Beirut-Medical Center, Beirut, Lebanon;10Laboratory of Human Genetics of Infectious
Diseases, Institut National de la Sante ´ et de la Recherche Me ´dicale, U550, Paris, France;11Paris Rene ´ Descartes University, Necker Medical
School, Paris, France;12Max-Planck-Institute for Molecular Cell Biology and Genetics, Dresden, Germany;13Karl Landsteiner Laboratory,
Academic Medical Centre, University of Amsterdam, Amsterdam, The Netherlands;14Department of Biochemistry, American University of
Beirut-Medical Center, Beirut, Lebanon.
Communicated by Haig H. Kazazian, Jr.
Received 28 April 2009; accepted revised manuscript 3 November 2009.
Published online 1 December 2009 in Wiley InterScience (www.interscience.wiley.com). DOI 10.1002/humu.21156
ABSTRACT: Mutations that impair expression or function
of the components of the phagocyte NADPH oxidase
complex cause chronic granulomatous disease (CGD),
which is associated with life-threatening infections and
dysregulated granulomatous inflammation. In five CGD
patients from four consanguineous families of two
different ethnic backgrounds, we found similar genomic
homozygous deletions of 1,380bp comprising exon 5 of
NCF2, which could be traced to Alu-mediated recombi-
nation events. cDNA sequencing showed in-frame
deletions of phase zero exon 5, which encodes one of
the tandem repeat motifs in the tetratricopeptide (TPR4)
domain of p67-phox. The resulting shortened protein
(p67D5) had a 10-fold reduced intracellular half-life and
was unable to form a functional NADPH oxidase
complex. No dominant negative inhibition of oxidase
activity by p67D5 was observed. We conclude that Alu-
induced deletion of the TPR4 domain of p67-phox leads
to loss of function and accelerated degradation of the
protein, and thus represents a new mechanism causing
Hum Mutat 31:151–158, 2010. & 2009 Wiley-Liss, Inc.
KEY WORDS: recombination; immunodeficiency; phago-
Chronic granulomatous disease (CGD) is associated with life-
threatening opportunistic infections and dysregulated inflamma-
tion, often accompanied by granuloma formation even in the
absence of detectable infections [Schuetz et al., 2007; Segal et al.,
2000; Winkelstein et al., 2000]. Patients need regular medical
checkups [Roesler et al., 2005], prophylactic and interventional
antimicrobial and/or antiinflammatory treatment [Margolis et al.,
1990; Mouy et al., 1994], and in certain cases bone-marrow
transplantation [Schuetz et al., 2007; Seger et al., 2002]. Gene
therapy may be a future therapeutic option [Ryser et al., 2007].
The phagocyte NADPH oxidase multienzyme complex is needed
for appropriate microbial killing and regulation of inflammation.
CGD is caused by mutations affecting the expression or function of
one out of four components of this enzyme complex [van den Berg
et al., 2009]. These components are gp91-phox (also referred to as
NOX2), p22-phox, p47-phox, and p67-phox, (MIM#s 608515,
233710; -phox, phagocyte oxidase). Rac2, p40-phox, and severe
G6PD deficiency also cause CGD-like diseases, but are accompanied
by different symptoms [Ambruso et al., 2000]. In about two-thirds
of all CGD cases, mutations are found in the X-chromosomal CYBB
gene encoding gp91-phox/NOX2 [Rae et al., 1998; Roesler et al.,
1999]. The genetic aberrations are family-specific and comprise a
wide range of mutation types. Mutations are also family-specific in
p22-phox [Yamada et al., 2000] and—as reported so far—in p67-
phox [Patino etal., 1999]deficiencies, whichare much rarer than the
X-linked form (each ?5% of all CGD cases). In contrast, p47-phox
deficiency (?25% of all CGD cases [Roesler et al., 2000]) is mostly
& 2009 WILEY-LISS, INC.
yMarcus Gentsch, Aneta Kaczmarczyk, Dirk Roos, and Joachim Roesler
contributed equally to this work.
Current address for Magdalena Kaus-Drobek: Cardiff University, Schools of
Chemistry and Biosciences, Park Place, CF10 3AT Cardiff, United Kingdom.
?Correspondence to: Joachim Roesler, MD, PhD, Dept. of Pediatrics, University
Hospital CarlGustav Carus,FetscherStr. 74,01307 Dresden, Germany.
Additional Supporting Information may be found in the online version of this article.
due to recombination events between the NCF1 gene and one
out of two highly homologous pseudogenes, thus leading to
the same GT deletion at the beginning of exon 2 in ?90% of all
p47-phox–deficient CGD patients.
In healthy individuals, the p67-phox protein combines with other
components of the NADPH oxidase to form the fully-functional
reactive oxygen species (ROS)-producing enzyme complex [Grizot
et al., 2001; Mizuki et al., 2005]. The SH3 domain close to the
C-terminal end of p67-phox interacts with the proline-rich region
(PRR) of p47-phox, the PB1 domain links p67-phox to p40-phox,
and the tetratricopeptide repeat (TPR) region of p67-phox domain
binds Rac-GTP [Grizot et al., 2001; Lapouge et al., 2000]. The TPR
domain comprises four repeats, consisting of two antiparallel
a-helices. An additional third a-helix follows the fourth repeat.
A review [Deininger and Batzer, 1999] reports 33 cases of
germline genetic diseases and 16 cases of cancer caused by unequal
homologous recombination between Alu repeats. The authors
estimate that this mode of mutagenesis accounts for 0.3% of
human genetic diseases. Alu sequences are the most abundant
repetitive elements in the human genome (about 1 million copies).
They emerged roughly 65 million years ago and amplified
throughout primate evolution by retrotransposition [Hasler and
Strub, 2006]. The large number of Alu elements within the human
genome provides abundant opportunities for unequal homologous
recombination events. These events often occur intrachromosomally
(Fig. 1A), resulting in deletion (or duplication) of exons in a gene,
but they can also occur interchromosomally (Fig. 1C), causing more
complex chromosomal abnormalities [Deininger and Batzer, 1999].
In five patients of unrelated families, we describe for the first
time Alu-repeat-associated genomic recombinations within the
NCF2 gene that lead to the deletion of part of the TPR domain
and to p67-phox–deficient CGD.
Materials and Methods
The patients and/or their parents consented to blood drawing
and all diagnostic testing, including genetic analysis.
Cell Lines and Culture Conditions
We used a human K562 cell model of p67-phox–deficient CGD
(K562-67def-CGD) that was engineered to contain p47-phox and
gp91-phox (and that naturally expresses p22phox mRNA). Only
when K562-67def-CGD cells are transduced to also produce p67-
phox do these cells become capable of generating superoxide in
response to phorbol 12-myristate 13-acetate (PMA) stimulation
[Leto et al., 2007]. K562 cells and Epstein-Barr virus (EBV)-
immortalized patient B cells were cultured in RPMI-1640 medium
supplemented with 10% (v/v) fetal bovine serum (FBS), 2mM
glutamine, 100U/mL penicillin, and 0.1mg/mL streptomycin.
PCR and Sequencing
Genomic DNA was isolated with the QIAamp DNA Blood kit
(Qiagen, Hilden, Germany). PCR reactions were performed with
AmpliTaq Gold DNA Polymerase (Applied Biosystems, Warring-
ton, UK). DNA analysis was performed with the ABI BigDye
Germany) and an ABI 377 automatic sequencer (Applied Biosystems,
Foster City, CA). Aberrant sequences were confirmed on both
DNA strands and in a second DNA sample to avoid PCR artifacts.
The ENSEMBL NCF2 (ENSG00000116701) sequence was used.
mRNAwas extracted with the DynabeadssmRNAdirect kit (Dynal,
Oslo, Norway) and reverse-transcribed into cDNAwith the SuperScript
First-Strand Synthesis System (Invitrogen, Paisley, UK). To determine
transcriptional levels of NCF2, quantitative RT-PCR was performed on
an ABI 7000 Sequence Detection System (Applied Biosystems,
Warrington, UK). The primer sequences for NCF2 were as follows:
*forward: 50-CCAGAAGCATTAACCGAGACAA-30, cDNA
*reverse: 50-AGAGCATCCCTCGTTGGAAGT-30, position 221–241;
*probe: 50-FAM-CACTTGGCAGTGGCT-TAMRA-30, position
Gene transcription of NCF2 was normalized to transcription of
glyceraldehyde-3-phosphate dehydrogenase (GAPDH) using the
*forward: 50-GAAGGTGAAGGTCGGAGTC-30, position 6–24;
*reverse: 50-GAAGATGGTGATGGGATTTC-30, position 212–231;
Retroviral Vectors and Vector Production
Full-length p67-phox-cDNA was ligated into NdeI/XhoI-opened
pET15b (Novagen; Merck Chemicals, Darmstadt, Germany), to
obtain pET15b-p67. To remove exon 5, we applied overlap extension
PCR and ligated the final PCR product into AvrII/NcoI-opened
pET15b-p67 to obtain pET.p67D5. Full-length p67-phox and
p67-phoxD5 were PCR amplified and cloned into a bicistronic
gammaretroviral transfer vector [Roesler et al., 2002] (MFG-
S-transgene-internal ribosome entry site [IRES]-enhanced green
fluorescent protein [eGFP]). The vectors obtained, pM.p67.iG and
pM.p67D5.iG, enable coexpression of p67-phox or p67D5 and eGFP.
Virus-particle-containing medium was from HEK 293T cells
cotransfected in 10-cm dishes with 10mg of transfer vector
(pM.p67.iG or pM.p67D5.iG), 2mg pMD.G (vesicular stomatitus
virus glycoprotein envelope plasmid), and 6.5mg of pHIT60 (gag-
pol-packaging plasmid) in the presence of 6.75mg/mL polyethy-
leneimine. K562 cells were transduced with virus-containing
media in the presence of protamine sulfate (5mg/mL) and by using
spinocculation (1,200g, 30min). Transduction efficiency was
analyzed by flow cytometry (FACS Calibur; BD Biosciences, San
Jose, CA) and found to be 70% for both vectors.
Assays for NADPH Oxidase Activity
cells, and transduced K562 cells was measured (i.e., three
experiments in duplicate) by chemiluminescence over 45min
after stimulation with 1mg/mL phorbol-12-myristate-13-acetate
(PMA) in a luminol-based chemiluminescence cocktail (Diogenes;
National Diagnostics, Atlanta, GA) in a Mithras LB 940 microplate
reader (Berthold Technology, Bad Wildbad, Germany). Alterna-
tively, hydrogen peroxide release by PMA-stimulated neutrophils
was measured with the Amplex Red assay (Invitrogen, Carlsbad,
CA) in the presence of horseradish peroxidase (HRP).
2) production in patients’ neutrophils, EBV-B-
Determination of Flavocytochrome b558Expression and
H2O2 Production at the Single-Cell Level
The dihydrorhodamine-1,2,3 flow cytometry assay (DHR assay)
for NADPH–oxidase-dependent production of ROS in neutrophils,
HUMAN MUTATION, Vol. 31, No. 2, 151–158, 2010
bp deletion comprising exon 5. A: Schematic representation of a possible loop-forming event. B: The same loop and the Alu alignment in
detail. The sequence that was found in all patients from Lebanon is marked in gray. All sequences not marked, whether depicted in detail or
represented by a line, were found to be deleted. The Alu repeats are underlined, and mismatches between them are marked in bold.
These mismatches narrow down the possible crossover point area. The main religation point (breakpoint, Br) is located within the sequence
marked in italics (however, the breakpoint shown is arbitrary). The ‘‘G’’ marks an accompanying small gene conversion. The sequence found
here had the ‘‘g’’ instead of the ‘‘a.’’ The patient from Pakistan had a breakpoint slightly upstream of Br (double arrow) and no gene conversion.
C: Instead of a recombination event within one NCF2 allele, such an event might have happened between two different alleles (different
chromosomes). Location: The NCF2 sequence can be found on chromosome 1 at location 181,791,321–181,826,339 according to ENSEMBL. The
deletion is found at approximately 181,809,587–181,810, 967 in the Lebanese families and the deletion in the patient from Pakistan is located a
few bp upstream.
An AluSq-repeat upstream of exon 5 of the NCF2 gene recombines with an AluSx-repeat downstream of exon 5, leading to a 1,380-
HUMAN MUTATION, Vol. 31, No. 2, 151–158, 2010
as indicated by H2O2production, was used as described [Mauch
et al., 2007]. Staining with the monoclonal antibody (moAb) 7D5,
which specifically binds to an extracellular epitope of gp91-phox, was
performed according to standard techniques. Samples were analyzed
by flow cytometry (FACS Calibur; BD Biosciences, San Jose, CA).
Sample Preparation and Western Blotting
Neutrophils, K562, or EBV-B cells were washed with PBS, the
pellets were solubilized in radioimmunoprecipitation assay (RIPA)
buffer (Sigma-Aldrich, St. Louis, MO) with 1mM phenylmethylsul-
fonyl fluoride (PMSF) and denatured at 951C after addition of SDS
sample buffer (Rotis-Load, Roth, Karlsruhe, Germany). Probes were
run on 12.5% (w/v) polyacrylamide gels and transferred onto
nitrocellulose membranes. After incubation with the moAb specific
for p67-phox (1:500 diluted, catalog 610912, raised against amino
acids 317–469; Becton Dickinson, Heidelberg, Germany) or for b-
actin (1:5,000 diluted, clone AC-74; Sigma), respectively, and a
secondary rabbit-anti-mouse-immunoglobulin (Ig) polyclonal Ab
conjugated to HRP (1:2,000, P 0260; DakoCytomation, Glostrup,
Denmark), the blots were developed with Amersham enhanced
chemiluminescence (ECL) PlusTM
Reagents (GE Healthcare, Freiburg, Germany).
Western Blotting Detection
EBV-B cells or transduced K562 cells were washed twice with
warm PBS and incubated for 30min at 5?106permL in
methionine and cysteine-free RPMI 1640 medium supplemented
with 2mM L-glutamine, 1mM L-methionine, 25mM HEPES pH
7.5, and 10% (v/v) dialyzed FBS. For labeling, 50mCi [35S]Met-label
(Hartmann Analytic GmbH, Braunschweig, Germany) were added
per milliliter of Met/Cys-free medium, and the cells were incubated
at 371C for 1hr. After labeling, the cells were washed once with PBS
and chased in complete RPMI 1640 medium with 25mM HEPES
pH 7.5. Samples containing 1?107B cells or 5?106K562 cells
were taken at indicated time points, washed twice with ice-cold PBS,
and lysed in 600mL of standard RIPA buffer containing protease
inhibitor cocktail (Sigma P8340; Sigma).
Immunoprecipitation and Autoradiography
Cell lysates were precleared three times with 25mL of protein-A
agarose (Roche Applied Science, Mannheim, Germany) rotating the
tubes end-over-end for 3hr or overnight at 41C. Subsequently, the
lysates were incubated with 1mL of the anti-p67-phox antiserum
(Upstate, Lake Placid, NY) for 1hr at 41C, then 10mL of protein A
agarose were added, and the lysates were incubated for 3hr or
overnight at 41C. The beads were washed twice with RIPA buffer and
once with 10mM Tris-HCl pH 8.0. Thereafter, beads were boiled in
SDS sample buffer for 5min and the supernatant was loaded onto
10% (w/v) polyacrylamide gels. The gels were dried and exposed to a
phosphorimaging screen. Quantification was done with the
ImageQuant software (Amersham Pharmacia, Uppsala, Sweden)
and normalized to the initial intensity set at 100%.
Identification of the p67-phox Deficiency
The determination of the defective disease-causing gene is
exemplified in two siblings of consanguineous Lebanese parents, a
boy and a girl (Patients 1 and 2, P1 and P2, respectively, in the figures)
suffering from symptoms typical for CGD (Supp. Information–Case
Reports). DHR testing and measurement of superoxide production of
neutrophils and EBV-B cells from both patients confirmed the
diagnosis of CGD without residual NADPH oxidase activity (Fig. 2A
similar result was obtained with cells from Patient 2) did not produce ROS (DHR assay); Co, healthy control donor; unst, unstimulated cells. B: EBV-B
cells from a healthy control donor (Co) and Patient 1 (P1) were stained with 7D5 anti-gp91-phox antibody and analyzed by flow cytometry. Analogous
data were obtained with cells from Patient 2. C: Failure of superoxide production by EBV-B cells from Patients 1 and 2 (P1 and P2); C, control EBV-B
cells. Superoxide production was detected by a luminol-based chemiluminescence assay of PMA-stimulated cells and is given in AUC (area under the
curve, arbitrary counts in 45min). D: EBV-B cells from the two patients (siblings, P1 and P2), a healthy control donor (Co) and the heterozygous parents
(F, father; M, mother) were analyzed for p67-phox expression by Western blotting with an anti-p67-phox monoclonal Ab (610912). Bands corresponding
to the wt protein and to the mutated protein p67-phox-delta-exon 5 (p67D5, marked D5) are indicated (the same amount of total protein was loaded in
each lane). The missing exon 5 comprises a stretch of 45 amino acids with a molecular weight of 5.3kD, p.123-167delVLYNIAFMYAKKEEWKKAEEQLA-
LATSMKSEPRHSKIDKAMECVW, leaving a truncated protein with a molecular weight of 61.7kD. E: Ribbon model of the structure of the TPR domain of
p67-phox. The coordinates for this figure were taken from a complex of this domain with Rac and GTP (pdb code 1e96) Reference [Lapouge et al., 2000].
The four tetratricopeptide repeats in p67-phox are labeled TPR1 to 4. In the p67D5 patients, TPR4 and the downstream helix (dark gray) are deleted.
Deletions within the p67-phox protein result in failure of ROS production. A: PMA-stimulated granulocytes from Lebanese Patient 1 (P1; a
HUMAN MUTATION, Vol. 31, No. 2, 151–158, 2010
and C). However, staining of the cells with monoclonal 7D5 antibody
specific against human gp91-phox, followed by flow cytometry,
demonstrated the presence of the membrane subunits of the enzyme
(Fig. 2B) and pointed to a cytosolic factor deficiency. Since lack of
p47-phox is usually associated with residual H2O2production seen in
the DHR assay, this combination of findings was highly suspicious of
p67-phox deficiency. Indeed, by Western blotting no full-length p67-
phox was detected in cell extracts from EBV-B cells from the two
children. Instead, a faster migrating band, which was also present in
EBV-B cell extracts from both parents, was recognized by the anti-
p67-phox monoclonal antibody (Fig. 2D). This faster migrating band
(referred to as D5 in Fig. 2D) was not seen in a sample from a patient
with a complete lack of p67-phox (Fig. 3B) and in a control sample.
Moreover, the D5 band was much fainter than that of wild-type (wt)
p67-phox. This was also the case in the parents and thus in cell
extracts that contained both p67-phox protein forms (Figs. 2D and
3B), indicating decreased synthesis or enhanced degradation of the
protein or its mRNA.
Sequencing of the p67-phox (NCF2) cDNA from the patients
revealed that exon 5 of p67-phox was missing without a frameshift
or any additional sequence changes. The consanguineous parents
were both heterozygous for the deletion (Fig. 3A), and their
neutrophils produced normal amounts of ROS (data not shown).
In the other three families, the consanguineous parents were
from Lebanon or Pakistan. No normal or faster migrating p67-
phox bands were identified in Western blot analysis of EBV-B cells
or neutrophil from the patients in these families (Table 1).
Alu Element Recombination Causes the Exon 5 Deletion
We suspected a splice mutation adjacent to or at the 50or 30end
of exon 5 in the NCF2 gene. Instead, PCR reactions with diverse
primers located upstream and downstream of this exon gave no
amplification product, indicating a large deletion. In a next step,
roughly 400-bp-long, overlapping stretches of DNA of introns 4
and 5 were amplified to find the approximate localization of the
deletion. Then, a PCR reaction was performed with a pair of
primers located within the amplimers that could be obtained
upstream and downstream adjacent to the deletion. The PCR
product was sequenced to identify the religation point (breakpoint
or crossover point; Fig. 1).
The identified sequence is marked in gray in Figure 1. This
mutation (c.36612401_502-527del1380, p.V123_W167del) had
not been described before. Comparison with the standard
ENSEMBL sequence revealed the localization of the religation
point within an Alu repeat that had recombined from two separate
Alu repeats, one upstream and one downstream of exon 5. The
location of this crossover point could be estimated thanks to the
mismatches in the Alu repeats (Figure 1B, marked in bold). In 50
to 30direction, the sequence found in the patients first followed
the AluSq repeat [Price et al., 2004] upstream of exon 5. At the
crossover point, it changed to the AluSx repeat downstream of
exon 5. Interestingly, there is one exception to this general rule: in
the site marked G in Fig. 1, the guanine nucleotide of AluSx is
found instead of the adenine of AluSq, demonstrating a small gene
parents produce normal amounts of p67D5 mRNA, but contain low
amounts of the respective p67D5 protein. A: cDNA was amplified with
primers located in exons 3 and 9; 24, 26, 28, and 30 PCR cycles were
applied. As shown, bands became visible after 26 cycles. The wt and the
p67D5 (D5) bands were equally strong in the father (F) and in the mother
(M), and the intensity of the bands was also similar in the patients (P1 and
P2) compared to the control sample (Co) from a healthy donor. Negative
control sample (Pn): EBV-B cells from a patient with an exon312T4C
mutation and thus without exon 3 and without p67-phox expression. B:
Western blot of EBV-B cells from the same patients with the anti-p67-
phox (610912) and the b-actin (AC-74) monoclonal Ab, respectively. (The
variation in intensity of the D5 band between F, M, P1, and P2 cannot be
explained completely. However, the Western blot has been repeated at
least five times. The variation was much less pronounced in the first two
experiments (Fig. 2D), but increased after the EBV-B cells had been
cultured for a longer period. Therefore, the variation in D5 band intensity
is most probably a matter of EBV-B cell subclone formation, meaning that
the fastest growing subclones express the D5 protein differently.
EBV-B cells from Patients 1 and 2 (P1 and P2) and their
Summary of Results Obtained From the Patients?
Family 1 (Patients 1 and 2)
NADPH oxidase activity in neutrophils
P67-phox expression in neutrophils
P67-phox expression in EBV-B cells
Crossover point in AluSq/Sx-hybrid Alu-repeat
Low amount p67D5
’----------------between-8T/C and –30T/A--------------
(Fig. 2B; Br)
Nil or tracea
Between –30T/A and –50G/T
(Fig. 2B; 2)
No Gene conversion –82A4G in AluSq as part of
recombined Alu (Fig. 2B and G)
The bp numbering refers to the hybrid AluSq/Sx-repeat sequence, arbitrarily counting backward from the 30to the 50end with the last bp as –1 (Fig. 1B).
aThe discrepancy with Family 1 could be due to differences in blotting techniques or to an even faster degradation of p67D5 in cells from Patient 3 than in cells from Patients 1
and 2 (compare to Fig. 4). NT, not tested.
HUMAN MUTATION, Vol. 31, No. 2, 151–158, 2010
The sequences of all patients from Lebanon were all identical to
those marked in gray in Figure 1B, including the small gene
conversion, whereas the crossover point was slightly more
upstream in the patient from Pakistan (Fig. 1B, double arrow),
and no gene conversion was observed. In agreement with these
observations, a 1,380-bp deletion comprising exon 5 was found in
all five patients. The findings are summarized in Table 1.
Stability of p67D5 mRNA and Protein
Next, we evaluated possible explanations for the low amount of
p67D5 protein in the patients’ cells, as exemplified in one of the
Lebanese families (Fig. 3B). The difference in band intensity was not
caused by different binding of the p67-phox monoclonal Ab, since
equal amounts of recombinant p67-phox and p67-phoxD5 gave
similarly strong bands (Supp. Fig. S1). Moreover, the epitope against
which the antibody was raised lies outside the exon-5-encoded region
(epitope amino acids 317–469, deletion amino acids 123–167).
We detected equal amounts of p67-phox and p67-phoxD5
mRNA by gel electrophoresis of low-cycle-number RT-PCR
products (Fig. 3A) and by quantitative RT-PCR (Table 2). Pulse-
chase experiments with EBV-B cells from the mother started out
with equal amounts of the full-length and the truncated p67-phox
protein (Fig. 4B, right side). This demonstrates that up to the
point of protein synthesis there was no quantitative difference
between the two forms. However, once synthesized, the fate of the
two forms differed. The truncated form was degraded more
rapidly (Fig. 4). Transduced K562 cells also showed a vastly
decreased half-life of the p67D5 variant in comparison to the
full-length protein (Fig. 4, left side). This suggests that the p67D5
protein was misfolded and hence quickly degraded.
Influence of p67D5 on ROS Production
The deletion of exon 5 leads to a deletion of 45 amino acids
without frameshift or amino acid change at the exon borders. In
the wt protein the deleted fragment would fold into the three
alpha helices of the TPR4 region (Fig. 2E).
The TPR domain of p67-phox is responsible for the binding of
Rac-GTP, which is needed for activation of the NADPH-oxidase
complex. The residues that interact with Rac-GTP are located in the
loops connecting TPR1 to TPR2 (S37) and TPR2 to TPR3 (D67 and
H69), and in the inserted b-hairpin between TPR3 and TPR4 (R102,
N104, L106, and D108) [Lapouge et al., 2000]. Hence, the TPR4
subdomain and the a-helix adjacent downstream of TPR4 that are
missing from p67D5 are not directly involved in Rac-GTP binding
(Fig. 2E). Therefore, we hypothesized that p67-phoxD5 might have
residual NADPH-oxidase-activating functionality but cannot be
detected due to the fast degradation of this protein.
To test this hypothesis we used CGD model K562 cells that
permanently express all components of the NADPH oxidase with
the exception of p67-phox, and transduced these cells with
retroviral vectors expressing either p67-phoxD5 or wt p67-phox in
the first cistron and EGFP in the second cistron. Transduction
efficiency was determined by flow cytometry and found to be 70%
for both vectors (Supp. Fig. S2). Western blot analyses showed
TaqMan RT-PCR of mRNA From EBV-Transformed
Cycle threshold p67
Patient 1, p67D5
Patient 2, p67D5
?Equal mRNA expression of both the mutated (Patients 1 and 2) and the wild-type
(control) p67-phox (n52, cycle threshold7range).
either full-length p67-phox or mutated p67D5 and EBV-B cells from the Lebanese mother of Patients 1 and 2 (P1 and P2) were labeled with
[35S]methionine for 1hr and chased for the times indicated. The cell lysates were analyzed by immunoprecipitation with p67-phox antiserum,
followed by SDS-PAGE and phosphorimaging. A: Band intensities of at least three independent experiments were quantified and plotted against
the chase times (mean values7SD). B: Representative autoradiographs.
The half-life of p67D5 is approximately 10-fold shorter than that of the wt p67-phox. K562 cells transduced with cDNA encoding
HUMAN MUTATION, Vol. 31, No. 2, 151–158, 2010
equally strong expression of wt p67-phox and p67D5. However,
also with these cells, no detectable ROS production was found in
cells expressing p67-phoxD5 (Fig. 5A). On the other hand, ROS
production by activated K562 cells that contain all components of
the NADPH oxidase including p67-phox was not functionally
inhibited by additional transduction with p67D5 (Fig. 5B). These
results indicate that p67D5 does not significantly interfere with
NADPH oxidase assembly.
Taken together, the p67-phoxD5 protein is nonfunctional, does
not interfere with normal NADPH oxidase assembly if wt p67-
phox is present (it has no dominant negative effect), and is rapidly
Our patients suffered from classical symptoms of CGD, and p67-
phox deficiency was detected as the underlying reason. In all five
patients, an Alu-repeat-induced 1,380-bp deletion comprising exon
5 of the NCF2 gene resulted in the deletion of 45 amino acids from
p67-phox, forming the TPR4 sequence and the downstream helix.
The results of the transduction and p67D5 pulse-chase experiments
suggest misfolding and accelerated degradation of p67D5.
The Alu-repeat induced sequence variation of the Pakistani
family differed from that of the three Lebanese families. The latter
families were not directly related and lived in different countries in
Europe and the United States. Nevertheless, a founder effect is the
most likely reason why they had exactly the same sequence
variation. However, such an effect does not provide a general
explanation for all four families. Rather, the propensity of Alu-
repeats to recombine must be the reason for all deletions found.
This is the first report demonstrating that such recombination
accounts for a fraction of p67-phox–deficient CGD cases.
A number of germline genetic diseases and types of cancer are
caused by unequal homologous recombination between Alu-repeats.
Alu-repeat-mediated interaction between the NCF1 gene and its
pseudogenes has been discussed as a possible reason for such
recombination events in p47-phox–deficient CGD [Roesler et al.,
2000]. In addition, a deletion in the NCF1 gene (c.154-285_4511
821del2860, p.Lys52ThrfsX82) has been found to be flanked by Alu
sequences. Furthermore, Alu repeats may be implicated in large
chromosomal deletions involving the CYBB gene, but they have never
been directly shown to be involved in CGD [Brown et al., 1996].
However, it has been clearly demonstrated that several cases of other
diseases, such as hemophilia [Nakaya et al., 2004], pseudoxanthoma
elasticum [Katona et al., 2005], and familial and sporadic
paraganglioma [Baysal et al., 2004], are caused by homologous
recombination between Alu sequences, mostly leading to deletions.
Alu sequences have also been described to modulate gene expression
at the posttranscriptional level. Furthermore, Alu sequences may
contribute to evolution by providing crossover points for genetic
recombination and gene duplication and have been used to trace
primate evolution [Salem et al., 2003].
As in other diseases resulting from Alu-induced deletions (e.g.,
Rossetti et al. ), we found an additional small gene conversion
close to the breakpoint. This may be explained as follows: the
recombination process probably starts with the formation of a
stretch of heteroduplex DNA before DNA sense and antisense
strands are disrupted and are religated crosswise. The mismatches in
the heteroduplex DNA (Fig. 1B) may then be removed by DNA
repair enzymes that use one strand as a template to which they adapt
the other strand. For some reason, the strand used as a template and
the strand adapted to it may sometimes change. This may also apply
to the region where the crossover point was found, and therefore the
real site(s) where DNA has been physically cut off may be at some
distance from the point indicated (Fig. 1B).
In most other cases of p67-phox–deficient CGD characterized so
far, no p67-phox protein was detected in patient cells [Noack et al.,
1999]. Only a few mutations, such as the deletion of Lys-58,
permit production of an apparently stable protein. However, this
DK58 variant was unable to bind Rac-GTP and to translocate to
the membrane [Leusen et al., 1996]. For the amino-acid
substitutions and small in-frame deletions described by others
[Noack et al., 1999; Patino et al., 1999], the lack or diminished
levels of the protein may be a consequence of structural instability
and increased susceptibility to intracellular proteases, similar to
the case described in our work.
Taken together, p67-phox–deficient CGD adds to the diseases
that can be caused by Alu-repeat-induced genomic deletions. Our
results also indicate that the TPR4 domain of p67-phox is
necessary for proper protein folding, stability and, thus, function.
We thank Alexander Herr, Department of Human Genetics, for interesting
discussions and helpful search of the literature; Claudia Stihlo, Department
form a functional NADPH oxidase and has no dominant negative
effect. A: Production of ROS was measured by chemiluminescence
(AUC: area under the curve of arbitrary counts in 45min) after PMA
activation of K562 cells that permanently express all components of
the NADPH oxidase except p67-phox. These cells were transduced
with a GFP control vector, with wild-type p67-phox (pM.67.iG) or with
p67D5 (pM.67D5.iG). Western blot analysis (WB) reveals strong
expression of p67-phox and of p67D5 (the same amount of total
protein was loaded in each lane). Effective transduction has also been
ensured by flow cytometric determination of GFP expression from the
second cistron (Supp. Fig. S2). B: ROS production by K562 cells that
already permanently expressed all components of the NADPH oxidase
and that were additionally transduced either with a GFP control vector
or with pM.67D5.iG.
The mutated p67-phox protein (p67D5) cannot interact to
HUMAN MUTATION, Vol. 31, No. 2, 151–158, 2010
of Immunology, MTZ, for anti-p67-phox-antibodies; Katrin Ho ¨hne, Download full-text
University Children’s Hospital, for excellent assistance; Frank Pessler,
University Children’s Hospital, for critical reading of the manuscript and
helpful comments; Andrea Gross (Center for Graphic Design, MRZ), for
expert graphic design (University Hospital, Technical University of
Dresden, Dresden, Germany), and Tom Leto, National Institutes of Health
(NIH), National Institute of Allergy and Infectious Diseases (NIAID),
Bethesda, MD, for kindly providing CGD model K562 cells.
Ambruso DR, Knall C, Abell AN, Panepinto J, Kurkchubasche A, Thurman G,
Gonzalez-Aller C, Hiester A, deBoer M, Harbeck RJ, Oyer R, Johnson GL,
Roos D. 2000. Human neutrophil immunodeficiency syndrome is associated
with an inhibitory Rac2 mutation. Proc Natl Acad Sci USA 97:4654–4659.
Baysal BE, Willett-Brozick JE, Filho PA, Lawrence EC, Myers EN, Ferrell RE. 2004.
An Alu-mediated partial SDHC deletion causes familial and sporadic
paraganglioma. J Med Genet 41:703–709.
Brown J, Dry KL, Edgar AJ, Pryde FE, Hardwick LJ, Aldred MA, Lester DH, Boyle S,
Kaplan J, Dufier JL, Ho MF, Monaco AM, Musarella MA, Wright AF. 1996.
Analysis of three deletion breakpoints in Xp21.1 and the further localization of
RP3. Genomics 37:200–210.
Deininger PL, Batzer MA. 1999. Alu repeats and human disease. Mol Genet Metab
Grizot S, Fieschi F, Dagher MC, Pebay-Peyroula E. 2001. The active N-terminal
region of p67phox. Structure at 1.8 A resolution and biochemical characteriza-
tions of the A128V mutant implicated in chronic granulomatous disease. J Biol
Hasler J, Strub K. 2006. Alu elements as regulators of gene expression. Nucleic Acids
Katona E, Aslanidis C, Remenyik E, Csikos M, Karpati S, Paragh G, Schmitz G. 2005.
Identification of a novel deletion in the ABCC6 gene leading to pseudox-
anthoma elasticum. J Dermatol Sci 40:115–121.
Lapouge K, Smith SJ, Walker PA, Gamblin SJ, Smerdon SJ, Rittinger K. 2000.
Structure of the TPR domain of p67phox in complex with Rac.GTP. Mol Cell
Leto TL, Lavigne MC, Homoyounpour N, Lekstrom K, Linton G, Malech HL, de
Mendez I. 2007. The K-562 cell model for analysis of neutrophil NADPH
oxidase function. Methods Mol Biol 412:365–383.
Leusen JH, de KA, Hilarius PM, Ahlin A, Palmblad J, Smith CI, Diekmann D, Hall A,
Verhoeven AJ, Roos D. 1996. Disturbed interaction of p21-rac with
mutated p67-phox causes chronic granulomatous disease. J Exp Med 184:
Margolis DM, Melnick DA, Alling DW, Gallin JI. 1990. Trimethoprim-sulfamethox-
azole prophylaxis in the management of chronic granulomatous disease. J Infect
Mauch L, Lun A, O’Gorman MR, Harris JS, Schulze I, Zychlinsky A, Fuchs T,
Oelschlagel U, Brenner S, Kutter D, Rosen-Wolff A, Roesler J. 2007. Chronic
granulomatous disease (CGD) and complete myeloperoxidase deficiency both
yield strongly reduced dihydrorhodamine 123 test signals but can be easily
discerned in routine testing for CGD. Clin Chem 53:890–896.
Mizuki K, Takeya R, Kuribayashi F, Nobuhisa I, Kohda D, Nunoi H, Takeshige K,
Sumimoto H. 2005. A region C-terminal to the proline-rich core of
p47phox regulates activation of the phagocyte NADPH oxidase by interacting
with the C-terminal SH3 domain of p67phox. Arch Biochem Biophys 444:
Mouy R, Veber F, Blanche S, Donadieu J, Brauner R, Levron JC, Griscelli C, Fischer A.
1994. Long-term itraconazoleprophylaxis
in thirty-two patients with chronic granulomatous disease. J Pediatr 125:998–1003.
Nakaya SM, Hsu TC, Geraghty SJ, Manco-Johnson MJ, Thompson AR. 2004. Severe
hemophilia A due to a 1.3kb factor VIII gene deletion including exon 24:
homologous recombination between 41 bp within an Alu repeat sequence in
introns 23 and 24. J Thromb Haemost 2:1941–1945.
Noack D, Rae J, Cross AR, Munoz J, Salmen S, Mendoza JA, Rossi N, Curnutte JT,
Heyworth PG. 1999. Autosomal recessive chronic granulomatous disease caused
by novel mutations in NCF-2, the gene encoding the p67-phox component of
phagocyte NADPH oxidase. Hum Genet 105:460–467.
Patino PJ, Rae J, Noack D, Erickson R, Ding J, de Olarte DG, Curnutte JT. 1999.
Molecular characterization of autosomal recessive chronic granulomatous
disease caused by a defect of the nicotinamide adenine dinucleotide phosphate
(reduced form) oxidase component p67-phox. Blood 94:2505–2514.
Price AL, Eskin E, Pevzner PA. 2004. Whole-genome analysis of Alu repeat elements
reveals complex evolutionary history. Genome Res 14:2245–2252.
Rae J, Newburger PE, Dinauer MC, Noack D, Hopkins PJ, Kuruto R, Curnutte JT.
1998. X-Linked chronic granulomatous disease: mutations in the CYBB gene
encoding the gp91-phox component of respiratory-burst oxidase. Am J Hum
Roesler J, Heyden S, Burdelski M, Schafer H, Kreth HW, Lehmann R, Paul D,
Marzahn J, Gahr M, Rosen-Wolff A. 1999. Uncommon missense and splice
mutations and resulting biochemical phenotypes in German patients with
X-linked chronic granulomatous disease. Exp Hematol 27:505–511.
Roesler J, Curnutte JT, Rae J, Barrett D, Patino P, Chanock SJ, Goerlach A. 2000.
Recombination events between the p47-phox gene and its highly homologous
pseudogenes are the main cause of autosomal recessive chronic granulomatous
disease. Blood 95:2150–2156.
Roesler J, Brenner S, Bukovsky AA, Whiting-Theobald N, Dull T, Kelly M, Civin CI,
Malech HL. 2002. Third-generation, self-inactivating gp91(phox) lentivector
corrects the oxidase defect in NOD/SCID mouse-repopulating peripheral blood-
mobilized CD341 cells from patients with X-linked chronic granulomatous
disease. Blood 100:4381–4390.
Roesler J, Koch A, Porksen G, von BH, Brenner S, Hahn G, Fischer R, Lorenz N, Gahr M,
Rosen-Wolff A. 2005. Benefit assessment of preventive medical check-ups in patients
suffering from chronic granulomatous disease (CGD). J Eval Clin Pract 11:
Rossetti LC, Goodeve A, Larripa IB, De Brasi CD. 2004. Homeologous recombination
between AluSx-sequences as a cause of hemophilia. Hum Mutat 24:440.
Ryser MF, Roesler J, Gentsch M, Brenner S. 2007. Gene therapy for chronic
granulomatous disease. Expert Opin Biol Ther 7:1799–1809.
Salem AH, Ray DA, Xing J, Callinan PA, Myers JS, Hedges DJ, Garber RK,
Witherspoon DJ, Jorde LB, Batzer MA. 2003. Alu elements and hominid
phylogenetics. Proc Natl Acad Sci USA 100:12787–12791.
Schuetz C, Hoenig M, Schulz A, Lee-Kirsch MA, Roesler J, Friedrich W, von Bernuth H.
2007. Successful unrelated bone marrow transplantation in a child with chronic
granulomatous disease complicated by pulmonary and cerebral granuloma
formation. Eur J Pediatr 166:785–788.
Segal BH, Leto TL, Gallin JI, Malech HL, Holland SM. 2000. Genetic, biochemical,
and clinical features of chronic granulomatous disease. Medicine (Baltimore)
Seger RA, Gungor T, Belohradsky BH, Blanche S, Bordigoni P, Di Bartolomeo P,
Flood T, Landais P, Mu ¨ller S, Ozsahin H, Passwell JH, Porta F, Slavin S,
Wulffraat N, Zintl F, Nagler A, Cant A, Fischer A. 2002. Treatment of chronic
granulomatous disease with myeloablative conditioning and an unmodified
hemopoietic allograft: a survey of the European experience, 1985–2000. Blood
van den Berg JM, van Koppen E, Ahlin A, Belohradsky BH, Bernatowska E, Corbeel L,
Espan ˜ol T, Fischer A, Kurenko-Deptuch M, Mouy R, Petropoulou T, Roesler J,
Seger R, Stasia MJ, Valerius NH, Weening RS, Wolach B, Roos D, Kuijpers TW.
2009. Chronic granulomatous disease: the European experience. PLoS One 4:e5234.
Winkelstein JA, Marino MC, Johnston Jr RB, Boyle J, Curnutte J, Gallin JI,
Malech HL, Holland SM, Ochs H, Quie P, Buckley RH, Foster CB, Chanock SJ,
Dickler H. 2000. Chronic granulomatous disease. Report on a national registry
of 368 patients. Medicine (Baltimore) 79:155–169.
Yamada M, Ariga T, Kawamura N, Ohtsu M, Imajoh-Ohmi S, Ohshika E,
Tatsuzawa O, Kobayashi K, Sakiyama Y. 2000. Genetic studies of three Japanese
patients with p22-phox-deficient chronic granulomatous disease: detection of a
possible common mutant CYBA allele in Japan and a genotype-phenotype
correlation in these patients. Br J Haematol 108:511–517.
HUMAN MUTATION, Vol. 31, No. 2, 151–158, 2010