Diagnostic paradigm for evaluation of

male patients with chronic

granulomatous disease, based on the
dihydrorhodamine 123 assay
Orathai Jirapongsananuruk, MD,a Harry L. Malech, MD,b Douglas B. Kuhns, PhD,c
Julie E. Niemela, MT,a Margaret R. Brown, MT,a Mindy Anderson-Cohen, MD,b and
Thomas A. Fleisher, MDa Bethesda and Frederick, Md

Basic and clinical

immunology

Background: Chronic granulomatous disease (CGD) is a

phagocyte disorder caused by mutations in nicotinamide dinucleotide phosphate (NADPH) oxidase subunits. The dihydrorhodamine 123 (DHR) assay is an effective test for CGD
that for most patients also might help to differentiate between
the 2 most common forms, X-linked (X) gp91phox defect CGD
and autosomal recessive (AR) p47phox defect CGD. However,
some male patients with X-CGD have DHR patterns that overlap the AR-CGD pattern.
Objective: The purpose of this investigation was to develop a
diagnostic paradigm to delineate male patients with X-CGD
expressing a DHR pattern suggestive of p47phox deficiency.
Methods: The DHR assay measured change in fluorescence of
DHR-loaded granulocytes after phorbol myristate acetate
(PMA) stimulation. Western blot analysis measured the presence of NADPH oxidase subunits gp91phox, p47phox, p67phox,
and p22phox. CYBB exonic sequencing was performed on PCRamplified genomic DNA through use of intronic flanking
primers. Ferricytochrome-c assay evaluated specific superoxide production by PMA-stimulated granulocytes.
Results: Although 83% of patients with X-CGD have virtually
no neutrophil DHR activity, we found that 17% of patients,
proven to have X-CGD by other criteria, have modest DHR
activity that is most consistent with p47phox deficiency. We
describe a diagnostic paradigm to deal with such patients, and
we present 2 cases, along with results of additional studies,
including carrier evaluation, protein assessment, and mutation
analysis, that are useful in establishing the genotype under
these circumstances. DHR assays from the 2 patients described
showed a fluorescence shift most characteristic of p47phox-ARCGD; however, each of the patients mothers showed
mosaicism with a bimodal DHR pattern. Patient 1 had some
gp91phox protein with a Y41D mutation and modest superoxide
activity. Patient 2 had a normal level of gp91phox protein with a
C537R mutation without detectable superoxide activity, as
determined by ferricytochrome-c assay, despite the modest
DHR activity.

From athe Department of Laboratory Medicine, Warren G. Magnuson Clinical Center, and bthe National Institute of Allergy and Infectious Diseases,
National Institutes of Health, Bethesda; and cthe Clinical Services Program, SAIC-Frederick, Inc, NCI Frederick, Frederick.
Received for publication March 12, 2002; revised October 10, 2002; accepted for publication October 15, 2002.
Reprint requests: Thomas A. Fleisher, MD, Department of Laboratory Medicine, Warren G. Magnuson Clinical Center, National Institutes of Health,
Building 10, Room 2C-306, 10 Center Drive, MSC 1508, Bethesda, MD
20892.
doi:10.1067/mai.2003.58

374

Conclusions: Evaluation of male patients with CGD with modest DHR activity should initially include evaluation of potential
female carriers for mosaicism with the use of the DHR assay.
In circumstances in which this is uninformative, patients
should be referred to centers capable of additional testing,
including Western blot analysis and CYBB mutation analysis,
to clarify the disease genotype. (J Allergy Clin Immunol
2003;111:374-9.)
Key words: Chronic granulomatous disease, CGD, CYBB, dihydrorhodamine assay, DHR, gp91phox, immune deficiency, mutation
analysis, phagocyte defect

Chronic granulomatous disease (CGD) is an inherited

phagocyte disorder of the reduced nicotinamide dinucleotide phosphate (NADPH) oxidase complex resulting
in defective superoxide generation and intracellular
killing.1 Patients with CGD have recurrent life-threatening bacterial and fungal infections, formation of chronic
granulomas, and poor wound healing.1 Genetic defects of
this disease have been identified, and the most common
presentation (70%2) is caused by mutations in the CYBB
gene (GenBank AF469757-69) on the X chromosome
(Xp21.1) coding for the gp91phox protein. This protein
and p22phox are the 2 membrane subunits of flavocytochrome b558; the protein is an essential component of
the phagocyte NADPH oxidase system. The remainder of
patients with CGD have an autosomal recessive (AR)
form; of these patients, the majority have a deficiency of
p47phox protein, whereas deficiencies of p67phox or
p22phox protein are far less frequent (<3% of cases). The
cytosolic proteins p47phox and p67phox are phosphorylated and bind to the cytochrome on cellular activation.1,2
Diagnosis of CGD is made by demonstrating absent or
markedly reduced oxidase activity in stimulated neutrophils. Screening of CGD has been done by using the
nitroblue tetrazolium (NBT) test. However, it has proved
to be subjective, being based on visual inspection of a
limited number of cells, and can miss the diagnosis of
CGD.1,3 Flow cytometry with the conversion of dihydrorhodamine (DHR) 123 to rhodamine 123 is a rapid
and sensitive assay by which to diagnose CGD. It can
help identify the CGD genotype because in the majority
of cases X-CGD demonstrates virtually no DHR shift,
whereas p47phox-AR-CGD shows a modest DHR shift
with a broad-based histogram.4-6 However, a subset of
male patients who by other criteria have X-CGD have a
modest DHR shift that requires additional studies to con-

Jirapongsananuruk et al 375

J ALLERGY CLIN IMMUNOL

VOLUME 111, NUMBER 2

firm the disease genotype. In this study, 2 such patients

are presented as examples for the laboratory evaluation
of CGD, and a diagnostic paradigm is developed with the
DHR used as the first step, followed by additional testing
as needed to define the CGD genotype.

METHODS
Patients
Studies were conducted with patients, their family members,
and normal control subjects after informed consent was obtained
(IRB approved research protocol 93-I-0119, National Institutes of
Health [NIH]).

Patient 1
An 8-year-old boy followed at the NIH for 1 year had first been
seen at a referral medical center at the age of 6 years for fever of
unknown origin that had persisted for 9 months. Chest radiography
revealed a dense pulmonary infiltrate in the left upper lobe. Open lung
biopsy showed granuloma formation with positive culture for
Aspergillus fumigatus. NBT and myeloperoxidase test results were
reported as normal. The child was treated with amphotericin for 6
weeks; this was followed by treatment with oral itraconazole. The
medical history was positive for BCG infection at the vaccination site,
multiple sinus infections, and chronic cough but no specific infections
suggestive of CGD until he was 6 years old. After antifungal therapy
at the age of 6 years, blood was sent to the NIH for a DHR assay; the
result was abnormal, the stimulation index (SI) being 35.9 (control,
199). The patient was then referred to the NIH at 7 years of age
because of the recurrent fever and persistent infiltrate.

Patient 2
A 34-year-old man followed at the NIH for 5 years had a history
of multiple infections throughout life and inflammatory bowel disease that had necessitated a diverting colostomy. Infections included
cervical abscess at 6 years of age, pneumonia at 7 years, Staphylococcal liver abscess at 11 and 15 years, and perirectal fistulae at 19
years. He was identified as having X-CGD at 7 years, as evidenced
by an abnormal NBT test result in the patient and his X-CGD brother. His mother had mosaicism confirmed by the NBT assay.

forward and side scatterplot were collected, and FL-2 (585 nm) was
analyzed. A sample from a healthy subject was used as a positive
control for each assay. The SI was calculated as the ratio of geometric mean channel fluorescence intensity of PMA-stimulated and
unstimulated granulocytes.4

Western blot analysis

SDS-PAGE immunoblot detection of cytosolic and membrane
NADPH oxidase components was performed as described.8 Diisopropyl fluorophosphatetreated granulocyte pellets were solubilized
in sample buffer (5 106 cells/100 L sample buffer) by sonication.
The lysates (1 106 cell equivalents/lane) were then separated on
either NuPage Novex 10% gels (p47phox and p67phox) or 4% to 12%
gradient gels (p22phox and gp91phox; Invitrogen, Carlsbad, Calif) and
transferred to Immobilon-P (Millipore Corp, Bedford, Mass). The
membranes were probed with goat anti-p22phox, anti-p47phox, antip67phox, and anti-gp91phox heteroantisera. Immunoblots were then
incubated with peroxidase-conjugated rabbit antigoat IgG (Sigma
Chemical Co) and developed with 3,3,5,5 tetramethylbenzidine
(BioFX Laboratories, Owings Mill, Md).

CYBB mutation analysis

Genomic DNA from patients PBMCs was isolated with the use of
the Puregene kit (Gentra systems, Minneapolis, Minn), according to
the manufacturers instructions. The 13 CYBB exons with their adjacent intronic sequences were amplified with flanking primer combinations.9,10 After 35 cycles of thermal cycling, the PCR products were
purified by gel filtration (Edge Biosystems, Gaithersburg, Md) and
directly sequenced by ABI Prism BigDye terminators (Applied
Biosystems, Foster City, Calif) and nested primers specific for each
exon. After 25 cycles of thermal cycling, sequencing extension products were purified by paramagnetic particle separation (RapXtract
Dye Terminator Removal Kit, Prolinx, Bothell, Wash) and analyzed
with a 3100 Genetic Analyzer (Applied Biosystems). To detect potential PCR artifacts, all mutations were confirmed by sequencing a second PCR product amplified from DNA from the original subject or a
carrier relative of which no discrepancies were encountered.

Ferricytochrome-c reduction assay

The specific detection of O2 production was measured by the
superoxide dismutaseinhibitable reduction of ferricytochrome-c.
Briefly, granulocytes (1 106/mL) suspended in HBSS containing 150
L ferricytochrome-c were incubated at 37C for 10 or 60 minutes in
the absence or presence of PMA (100 ng/mL). The reaction was
stopped at 4C and the supernatant fluid was harvested after centrifuging. An identically treated tube containing superoxide dismutase
(100 g/mL) served as the blank for each set of conditions. The reduction of ferricytochrome-c was assayed with an analytic wavelength of
549.5 nm and background wavelengths at the isobestic points, 541 and
556 nm. The data were converted to nanomoles of O2/106 cells per 10
or 60 minutes by using a micromolar extinction coefficient of 0.0211.

DHR assay
DHR assay was performed according to the previously described
methods of Vowells et al.7 In brief, after red blood cell lysis, leukocytes were loaded with DHR (Molecular Probe, Eugene, Ore) at
37C for 5 minutes in the presence of catalase (Sigma Chemical, St
Louis, Mo). After DHR loading, cells were stimulated with phorbol
myristate acetate (PMA; Sigma Chemical) for 14 minutes and
immediately analyzed by flow cytometry.

Flow cytometry
DHR samples were analyzed by using Cell Quest on a FACScan
flow cytometer (Becton Dickinson Immunocytometry Systems, San
Jose, Calif). Twenty thousand viable granulocytes determined by

RESULTS
The DHR flow cytometry assay has proved to be a
rapid and sensitive screening test for CGD. This assay
can differentiate between X-CGD and p47phox-AR-CGD
in the majority of patients, on the basis of distinctive
DHR histogram SI and pattern. Vowells et al7 demonstrated that the geometric mean SI of granulocytes from
normal subjects was 127.9 (range, 85.2 to 264.6), whereas the SIs from patients with CGD with defective
gp91phox and p47phox were 1.3 (range, 0.9 to 2.2) and
13.2 (range, 3.5 to 52.1), respectively.4 Fig 1 (left panel,

Basic and clinical

immunology

Abbreviations used
AR: Autosomal recessive
CGD: Chronic granulomatous disease
DHR: Dihydrorhodamine 123
NADPH: Nicotinamide dinucleotide phosphate
NBT: Nitroblue tetrazolium
PMA: Phorbol myristate acetate
SI: Stimulation index

376 Jirapongsananuruk et al

J ALLERGY CLIN IMMUNOL

FEBRUARY 2003

Basic and clinical

immunology
FIG 1. In the left panel, typical DHR histograms and SIs in a normal subject and in p47phox-AR-CGD, XL-CGD
and XL-CGD carriers are shown (top to bottom, respectively). In the right panel, DHR histograms and SIs in
patient 1, patient 1s mother, patient 2, and patient 2s mother are shown (top to bottom, respectively).

Jirapongsananuruk et al 377

J ALLERGY CLIN IMMUNOL

VOLUME 111, NUMBER 2

top to bottom) shows typical DHR histograms and SIs in

normal, p47phox-AR-CGD, X-CGD, and X-CGD carriers. Fig 1 (right panel, top to bottom) shows patient 1, the
mother of patient 1, patient 2, and the mother of patient
2. The 2 patients described in this report had DHR patterns resembling p47phox-AR-CGD with SIs of 44.2 and
17.1, respectively, plus broad histograms. The DHR pattern from the mothers showed bimodal distribution consistent with their obligate X-CGD carrier status. This
mosaic pattern in the mothers is characterized by a
bimodal histogram demonstrating one peak (M1) similar
to their sons DHR results and a second peak (M2) similar to results seen with control cells. Although the results
from the Vowell report are applicable for most patients
with CGD, there is a subset of X-CGD that shows an
overlapping SI and pattern with p47phox-AR-CGD. Our
expanded DHR assay database at the NIH now includes
75 patients with X-CGD, of whom 13 (17%) have an SI
of >4.5. These patients with X-CGD with modest DHR
activity included both patients with and patients without
gp91phox protein production. In contrast, of the 31
patients with AR-CGD who were evaluated, only 2
both malehad SIs of <4.5, and in both cases the broad
histograms, as reflected by a high coefficient of variability (70), distinguishes their results from the typical XCGD DHR pattern. Thus, in male patients with CGD
with modest DHR activity, defined as an SI of >4.5, additional studies are warranted. The simplest means of
establishing the genotype in this group of patients is to
demonstrate mosaicism in a potential female carrier, as
was the case in the 2 cases presented. However, when this
approach is not informative, protein analysis and/or
mutation analysis should be performed. In addition, the
ferricytochrome-c assay provides additional functional
data regarding the in vitro production of reactive oxygen
species. However, it is important to point out that this
assay fails to distinguish between X-CGD and AR-CGD.

In the 2 cases presented, different results were

observed with additional confirmatory testing. Western
blot analysis from patient 1 revealed a small amount of a
normal-sized gp91phox protein (Fig 2); this is unlike what
is seen in most patients with X-CGD, who do not demonstrate any protein. The Western blot also demonstrates
decreased but identifiable p22phox protein. This finding
fits with our observation of a large number of gp91phoxdeficient patients, who consistently show that a small
amount of p22phox protein can be identified by overdeveloping the gel. The only circumstance in which we see
total absence of p22phox protein is AR-CGD related to
documented mutations in p22phox. Superoxide production per 106 neutrophils at 10 and 60 minutes was significantly decreased: 12.6 and 51.9 nmol compared with
control values of 35.8 and 151.7 nmol, respectively (resting cells produced 1.1 nmol at both time points), and
these findings were confirmed on repeat evaluation.
Western blot analysis from patient 2 revealed a normal
amount of a normal-sized gp91phox protein (Fig 2).
Superoxide production per 106 neutrophils at 10 and 60
minutes was absent: 0.5 and 0.9 nmol compared with
control values of 51.2 and 263.1 nmol, respectively (resting cell produced 0.8 nmol at both time points). CYBB
mutation analysis for patient 1 showed a point mutation
in exon 2, yielding an amino acid change 41 Tyr Asp
that was confirmed by the presence of the same mutation
on one X chromosome from his mother (Fig 3). A point
mutation in exon 13 was identified for patient 2, resulting in 537 Cys Arg (data not shown).

DISCUSSION
The most common form of X-CGD is caused by large
deletions, nonsense mutations, and splice-junction mutations in the CYBB gene, leading to an absence of
gp91phox protein (X910).1 Fewer than 10% of X-CGD

Basic and clinical

immunology

FIG 2. Western immunoblot analysis measuring gp 91phox, p67phox, p47phox, and p22phox protein in the granulocyte fractions of patient 1 (left panel) and patient 2 (right panel) compared with control subjects and a patient
with typical XL-CGD.

378 Jirapongsananuruk et al

J ALLERGY CLIN IMMUNOL

FEBRUARY 2003

FIG 3. CYBB mutation analysis in patient 1 revealed a point mutation T G in exon 2 at nucleotide 133
(nucleotide number in accordance with cDNA sequence described by Orkin,10 in which the start of translation is +1 and the A of the ATG start codon is 13; see full genomic DNA sequence in GenBank files AF
469757-69). This results in alteration of amino acid 41 Tyr Asp. Patients mothers sequence showed heterozygosity in this mutation.

Basic and clinical

immunology

cases have a gp91phox protein level ranging from 1% to

25% of normal (X91).1 X91 phagocytes typically generate reduced but measurable amounts of superoxide. In
some X-CGD cases, normal amounts of nonfunctional
gp91phox protein are present (X91+). The X91 and X91+
forms are both due to a missense mutations or small
inframe deletions.1,11 These patients often have variable
clinical symptoms and might present at an older age. The
phenotype of patient 1 is most compatible with X91
with decreased quantity of gp91phox protein expression
but, interestingly, with approximately one third of control
superoxide production. The clinical history in this patient
is puzzling, in that this relatively high level of in vitro
superoxide activity appears to be insufficient for normal
host defense. CYBB mutation analysis demonstrated a
mutation in exon 2 on the N-terminal domain. This
domain serves multiple functions, including binding of
gp91phox protein in the membrane, interaction with
p22phox protein, and binding of the heme groups.12 There
are a number of potential explanations for diminished
gp91phox protein in this patient, including decreased stability or impaired folding of the protein as well as
impaired membrane integration.11 However, we could
not find decreased DHR activity or intracellular staining
for gp91phox protein when the patients blood was held
overnight for additional testing. Thus, the functional
basis of the defect for patient 1 remains to be defined.
The phenotype of patient 2 is most compatible with
X91+, given the normal quantity of gp91phox protein
expression and the absence of superoxide production.
Interestingly, despite the lack of superoxide generation,
there must be electron leakage that reduces DHR, yielding an increased SI compared with the typical X-CGD
DHR histogram. This patients mutation is in exon 13
within the NADPH-binding domain. Mutations in this
domain previously have been reported to cause the X91+
phenotype as a result of decreased NADPH binding or

lack of interaction with the cytosolic NADPH-oxidase

components.12,13 This probably is the explanation for the
normal quantity of gp91phox protein observed in the
absence of superoxide production. We noted previously
that granulocytes from several other patients with X-CGD
evaluated in our laboratory were found to have a DHR
pattern typical for p47phox-AR-CGD, and their cells were
also observed to express some gp91phox protein. However, we have also seen patients with X91 CGD whose
granulocytes have modest DHR activity typical for ARCGD. This suggests that some electron flow to DHR
occurs in the absence of NADPH oxidase activity, but the
nature of the electron flow to DHR in the absence of
superoxide production is not known. Furthermore, the
fact that the 2 patients had similar DHR patterns but
showed different results with the ferricytochrome-c assay
points out that these 2 assays measure different end products. In addition, these findings indicate that even patients
with CGD who have some level of in vitro superoxide
production, as reflected by the ferricytochrome-c assay,
might have recurrent severe infections.
We conclude that the majority (83%) of patients with
X-CGD show a unique DHR assay pattern that is diagnostic for this genotype. The remainder (17%) of patients
with X-CGD have a DHR assay histogram that overlaps
with that of p47phox-AR-CGD; they require additional
testing to identify the genotype. Evaluation of a male
patient with modest DHR activity (SI, >4.5) should
include DHR assay of obligate and/or potential carriers
to establish mosaicism. If this is unrevealing, it might be
necessary to refer the patient to a specialized center capable of specialized protein and CYBB mutation analysis.
We thank Pablo Patino for his excellent technical assistance.

Nucleotide sequence
The nucleotide sequence data used in this report have been submitted to GenBank with the accession number AF 469757-69.

REFERENCES
1. Segal BH, Leto TL, Gallin JI, Malech HL, Holland SM. Genetic, biochemical, and clinical features of chronic granulomatous disease. Medicine 2000;79:170-200.
2. Winkelstein JA, Marino MC, Johnston RB Jr, Boyle J, Curnutte J, Gallin
JI, et al. Chronic granulomatous disease: report on a national registry of
368 patients. Medicine 2000;79:155-69.
3. Vowells SJ, Fleisher TA, Malech HL. Testing for chronic granulomatous
disease. Lancet 1996;347:1048-9.
4. Vowells SJ, Fleisher TA, Sekhsaria S, Alling DW, Maguire TE, Malech
HL. Genotype-dependent variability in flow cytometric evaluation of
reduced nicotinamide adenine dinucleotide phosphate oxidase function in
patients with chronic granulomatous disease. J Pediatr 1996;128:104-7.
5. Crockard AD, Thompson JM, Boyd NAM, Haughton DJ, McCluskey
DR, Turner CP. Diagnosis and carrier detection of chronic granulomatous
disease in five families by flow cytometry. Int Arch Allergy Immunol
1997;114:144-52.
6. Roesler J, Hecht M, Freihorst J, Lohmann-Matthes ML, Emmendorffer
A. Diagnosis of chronic granulomatous disease and its mode of inheritance by dihydrorhodamine 123 and flow microcytofluorometry. Eur J
Pediatr 1991;150:161-5.

Jirapongsananuruk et al 379

7. Vowells SJ, Sekhsaria S, Malech HL, Shalit M, Fleisher TA. Flow cytometric analysis of the granulocyte respiratory burst: a comparison study
of fluorescent probes. J Immunol Methods 1995;178:89-97.
8. Levy R, Rotrosen D, Nagauker O, Leto T, Malech H. Induction of the respiratory burst in HL-60 cells, correlation of function and protein expression. J Immunol 1990;145:2595-601.
9. Jirapongsananuruk O, Niemela JE, Malech HL, Fleisher TA. CYBB mutation analysis in X-linked chronic granulomatous disease. Clin Immunol
2002;104:73-6.
10. Orkin SH. Molecular genetics of chronic granulomatous disease. Ann
Rev Immunol 1989;7:277-307.
11. Roesler J, Heyden S, Burdelski M, Schafer H, Kreth HW, Lehmann R, et
al. Uncommon missense and splice mutations and resulting biochemical
phenotypes inn German patients with X-linked chronic granulomatous
disease. Exp Hematol 1999;27:505-11.
12. Roos D, Curnutte JT, Hossle JP, Lau YL, Ariga T, Nunoi H, et al. XCGDbase: a database of X-CGD-causing mutations. Immunol Today
1996;17:517-21.
13. Heyworth PG, Curnutte JT, Rae J, Noack D, Roos D, Van Koppen E, et
al. Hematologically important mutations: X-linked chronic granulomatous disease (second update). Blood Cells Mol Dis 2001;27:16-26.

Basic and clinical

immunology

J ALLERGY CLIN IMMUNOL

VOLUME 111, NUMBER 2