Purified Cytochrome b from Human Granulocyte Plasma Membrane is
Comprised of Two Polypeptides with Relative Molecular Weights of
91,000 and 22,000
Charles A. Parkos, Rodger A. Allen, Charles G. Cochrane, and Algirdas J. Jesaitis
Research Institute ofScripps Clinic, Department ofImmunology, La Jolla, California 92037
A new method has been developed for purification of cy-
tochrome b from stimulated human granulocytes offering the
advantage of high yields from practical quantities of whole
blood. Polymorphonuclear leukocytes were treated with diiso-
propylfluorophosphate, degranulated and disrupted by nitro-
gen cavitation. Membranes enriched in cytochrome b were
prepared by differential centrifugation. Complete solubiliza-
tion of the cytochrome from the membranes was achieved in
octylglucoside after a 1-M salt wash. Wheat germ agglutinin-
conjugated Sepharose 4B specifically bound the solubilized
cytochrome b and afforded a threefold purification. Eluate
from the immobilized wheat germ agglutinin was further
enriched by chromatography on immobilized heparin. The
final 260-fold purification of the b-type cytochrome with a
20-30% yield was achieved by velocity sedimentation in su-
crose density gradients. Sodium dodecyl sulfate-polyacryl-
amide gel electrophoresis (SDS-PAGE) of the purified prepa-
ration revealed two polypeptides ofM, 91,000 andMr22,000.
Treatment of the "2I-labeled, purified preparation with pep-
tide:N-glycosidase F, which removes N-linked sugars, de-
creased relative molecular weight of the larger species to
-50,000, whereas beta-elimination, which removes O-linked
sugars, had little or no effect on the mobility of the M,-91,000
polypeptide. Neither ofthe deglycosylation conditions had any
effect on electrophoretic mobility of the M,-22,000 polypep-
tide. Disuccinimidyl suberate cross-linked the two polypep-
tides to a new M, of 120,000-135,000 by SDS-PAGE. Anti-
body raised to the purified preparation immunoprecipitated
spectral activity and, on Western blots, bound to the
M,-22,000 polypeptide but not the M,-91,000 polypeptide.
Western blot analysis of granulocytes from patients with X-
linked chronic granulomatous disease revealed a complete ab-
sence of the M,-22,000 polypeptide. These results (a) suggest
that the two polypeptides are in close association and are part
of the cytochrome b, (b) provide explanation for the molecular
weight discrepancies previously reported for the protein, and
(c) further support the involvement of the cytochrome in su-
peroxide production in human neutrophils.
Publication No. 4084 IMM from the Department of Immunology,
Research Institute ofScripps Clinic.
Address correspondence to Algirdas J. Jesaitis, Ph.D., Department
ofImmunology, IMM 12, Research Institute ofScripps Clinic, 10666
North Torrey Pines Road, La Jolla, CA 92037.
Receivedforpublication 18 November 1986 and in revisedform 27
Neutrophils play a crucial role in defending the body against
invading pathogens (1-3). Stimulation of these granulocytes
with bacteria or other chemoattractants results in the release of
microbicidal oxidants into phagolysosomes or the immediate
environment (4-6). The mechanism by which neutrophils
produce oxidants, however, is not fully understood.
Evidence suggests that an inducible electron transport sys-
tem that transfers reducing equivalents from NADPH to oxy-
gen is the source of oxidant production (7, 8). The terminal
component ofthis electron transport system is believed to be a
low-potential b-type cytochrome (9, 10). Support for the in-
volvement of this cytochrome b in oxidant production has
been obtained from (a) its spectrophotometric absence in pa-
tients with certain forms of chronic granulomatous disease
(CGD)' whose neutrophils can not mount a respiratory burst
(11); (b) genetic complementation studies of oxidant produc-
tion by the hybridization of monocytes from cytochrome b-
positive and cytochrome b-negative CGD patients to recon-
stitute the production of oxidants (12); (c) copurification of
cytochrome b along with O-generating activity in detergent
extracts (13, 14); (d) its unusually low (-245 mV) electro-
chemical potential (15, 16); and (e) its anaerobic reduction
upon oxidase activation (10) followed by rapid reoxidation
after introduction of oxygen into the system (15). Recently,
four different groups have reported the purification of cy-
tochrome b from human, bovine, and porcine leukocytes. Rel-
ative molecular weight estimates, obtained from SDS-PAGE
analyses, were diverse, ranging from 11,000 to 14,000 for bo-
vine leukocytes (17), 32,000 for porcine neutrophils (18),
68,000-80,000 for human myelogenous leukemia cells (19),
and 127,000 for purified human neutrophils (20). These prep-
arations also varied significantly in their reported homogene-
ity, specific activity, and methods ofpurification.
In this report, we describe a new method of purification
and partial characterization of cytochrome b obtained from
cytochalasin-treated human neutrophils stimulated by N-for-
myl-Met-Leu-Phe. The purification procedure we describe
yields highly purified cytochrome b (19.2 nmol heme/mg pro-
tein) from readily obtainable quantities of whole blood (3-6
U). The purified preparations were used to produce antibodies
and perform electrophoretic and carbohydrate analyses. Our
1. Abbreviations used in this paper: ABTS, 2,2-azino-di-(3-ethyl-
benzthioline sulfonic acid; CGD, chronic granulomatous disease;
CRB, column running buffer; DFP, diisopropyl fluorophosphate;
DPBS, Dulbecco's phosphate-buffered saline; DPBS+, DPBS plus0.1%
glucose and 0.1% BSA; DSS, disuccinimidyl suberate; MRB, mem-
brane resuspension buffer; PMN, polymorphonuclear neutrophilic
leukocyte; PMSF, phenylmethyl sulfonyl fluoride; WGA-4B, wheat
germ agglutinin-conjugated Sepharose 4B.
C. A. Parkos, R. A. Allen, C. G. Cochrane, and A. J. Jesaitis
J. Clin. Invest.
© The American Society for Clinical Investigation, Inc.
Volume 80, September 1987, 732-742
results suggest that purified cytochrome b is composed oftwo
polypeptides; a heavily glycosylated, Mr-91,000 species and a
Mr-22,000 species which may not be glycosylated. Our finding
of both high- and low-molecular-weight polypeptides in the
purified cytochrome b preparation may help explain the mo-
lecular weight discrepancies reported in the literature.
Lastly, Western blotting studies failed to detect any
Mr-22,000 polypeptide in extracts ofneutrophils from patients
with the X-linked form ofCGD. This result further supports a
central role of cytochrome b in superoxide production. Por-
tions of this work have been previously published in abstract
Reagents. N-Formyl-Met-Leu-Phe, dihydrocytochalasin B, Na2 ATP,
ovalbumin bovine serum albumin, chymostatin, poly-L-lysine (mol
wt 14,000) N-acetylglucosamine, 2,2-azino-di-(3-ethylbenzthioline
sulfonic acid) (ABTS), and fetuin type IV were purchased from Sigma
Chemical Co., St. Louis, MO. Catalase, superoxide dismutase, phen-
ylmethylsulfonyl fluoride (PMSF), octylglucoside, and dithiothreitol
were purchased from Calbiochem Behring Corp., La Jolla, CA. Gelatin
and sodium dithionite were obtained from Fisher Scientific, Pitts-
burgh, PA. N-2-hydroxyethyl piperazine-N-2-ethane sulfonic acid
(Hepes) was purchased from United States Biochemical Corp., Cleve-
land, OH. EDTA and EGTA were obtained from Fluka AG, Haup-
page, NY. Ultrapure Triton X-100 was purchased from Boehringer
Mannheim GmbH, Mannheim, FRG. Wheat germ agglutinin was
purchased from Vector Laboratories Inc., Burlington, CA. Cyanogen
bromide-activated Sepharose 4B and low-molecular-weight protein
standards were purchased from Pharmacia Inc., Upsala, Sweden. SDS,
acrylamide, bis-acrylamide, Tween 20, sodium persulfate, peroxidase
color developer, N,N,N,N,-tetramethylenediamine (TEMED) was
purchased from Bio-Rad Laboratories, Richmond, CA. Prestained
high-molecular-weight protein standards for SDS gels were purchased
from Bethesda Research Laboratories, Gaithersburg, MD. Glycerol,
Coomassie Brilliant Blue G-250, and 2-mercaptoethanol were from J.
T. Baker Chemical Co., Phillipsburg, NJ. Peptide:N-glycosidase F was
purchased from Genzyme Corp., Boston, MA. Sucrose and ultrapure
urea were acquired from Schwarz-Mann, Spring Valley, NY. Chito-
biose was purchased from E. Y. Laboratories Inc., San Mateo, CA.
Heparin ultrogel was obtained from LKB Instruments Inc., Bromma,
Sweden. Disuccinimidyl suberate, Iodobeads, and BCA protein assay
reagent were purchased from Pierce Chemical Co., Rockford, IL. Fluo-
rescamine was obtained from Roche (Nutley, NJ). Freund's Adjuvant
was from Difco Laboratories Inc., Detroit, MI. Peroxidase-conjugated
goat anti-rabbit IgG was obtained from Tago Inc., Burlingame, CA.
Nitrocellulose sheets were purchased from Millipore/Continental
Water Systems, Bedford, MA.
Buffer composition. Cell resuspension buffer was a modified Dul-
becco's phosphate-buffered saline (DPBS) containing 5 mM KCI, 147
mM NaCI, 1.9 mM KH2PO4, 1.1 mM K2HPO4, 1.5 mM CaC12, 1.1
mM MgCl2 (pH 7.4) (DPBS[-]), to which 0.1% glucose and 0.1%
bovine serum albumin (BSA) was added (DPBS[+]). Nitrogen cavita-
tion buffer consisted of 100mM KCI, 10mM NaCl, 10mM Hepes, 3.5
mM MgCl2, and 1 mM ATP (pH 7.3). Membrane resuspension buffer
(MRB) consisted of 100mM KCG, 10mM NaCl, 10mM Hepes, 1 mM
EDTA, 0.1 mM dithiothreitol, 1 mM PMSF, and 10gg/ml chymosta-
tin (pH 7.3). Column-running buffer (CRB) consisted of 500 mM
NaCl, 50 mM Hepes, 5% (vol/vol) glycerol,
dithiothreitol, 0.02% azide (pH 7.3), and either octyl glucoside or Tri-
ton X-100 in the amounts described in the text. Sucrose solutions
consisted ofthe appropriate amount of sucrose dissolved in 100 mM
KCG, 10 mM NaCl, 10 mM Hepes,
X-100 (pH 7.3).
1 mM EDTA, 0.1 mM
1 mM EDTA, and 0.1% Triton
Preparation ofcytochrome b-enriched neutrophil membranes. Pu-
rified granulocyte polymorphonuclear neutrophilic leukocytes (PMN)
(> 95% PMN) were obtained from peripheral blood by scaling up the
procedure ofHenson and Oades (22) to accommodate 1 U of blood.
Typically, 1 U ofblood yielded between 0.8 and 1.2 X 109neutrophils.
Red cells were lysed by two treatments with isotonic ammonium chlo-
ride. The resultant neutrophil suspensions from different units were
pooled and treated with diisopropylfluorophosphate (DFP) (24) to in-
activate serine esterases. After two washes in DPBS the cells were
resuspended at 108 cells/ml in DPBS' to which catalase (500 U/ml)
and superoxide dismutase (30 U/ml) were added.
The resuspended cells were then treated with 2 isg/ml dihydrocyto-
chalasin B and allowed to warm to 370C for 7 min. Degranulation was
achieved by stimulation with 1 gM N-formyl-Met-Leu-Phe for 3 min,
after which the reaction terminated with 4 vol of ice-cold DPBS and
centrifuged at 1,300 g for 15 min. After washing the cells twice with
DPBS, they were disrupted by nitrogen cavitation (400 psi for 15 min
at 40C) in a modified relaxation buffer originally described by Borre-
gaard and Tauber (25) at 108 cells/ml. The cavitate was collected in a
tube containing one-tenth the starting volume of nitrogen cavitation
buffer plus 12.5 mM EGTA and separated into a low-speed (1,000g, 5
min) supernatant and foam/pellet residue. The foam/pellet residue
was rehomogenized by 10 strokes in a glass Dounce tissue homoge-
nizer and again fractionated into a low-speed supernatant and pellet
The low-speed supernatant fractions were pooled and centrifuged
in a 60 Ti rotor (Beckman Instruments, Inc., Fullerton, CA) for 45 min
at 45,000 rpm and 4°C. The resultant yellowish pellets were carefully
transferred to a glass Dounce homogenizer and resuspended in mem-
brane resuspension buffer (MRB) at a protein concentration of 3-5
mg/ml. The resuspended membranes were stored at -70°C until fur-
Detergent solubilization ofcytochrome b. After thawing on ice, the
membranes were treated with 1 M salt by adding 1 ml of 5 M NaCl per
4 ml of the membrane suspension, thoroughly mixing and then cen-
trifuging the mixture at 100,000 g for 30 min (4°C). The clear yellow
100,000 g supernatant was discarded, and the yellow pellet was resus-
pended in a small volume ofMRB with the aid ofaglassDounce tissue
homogenizer. Complete solubilization of the cytochrome b was
achieved by adjusting the protein concentration to 1-2 mg/ml with
MRB and adding octylglucoside to make a final detergent concentra-
tion of2% (wt/vol). The detergent solution was mechanicallystirredby
a vortex mixer and kept on ice for 30 min and then centrifuged at
100,000 g (ray) for 30 min at 4°C in a 50 Ti or 60 Ti rotor (Beckman
Instruments, Inc.). The 2% octylglucoside 100,000 g supernatantwas
then used for further purification ofcytochrome b.
Affinity chromatography. After detergent solubilization, the cy-
tochrome b was further purified by affinity chromatography utilizing
wheat germ agglutinin-conjugated Sepharose 4B (WGA-4B). Wheat
germ agglutinin (WGA) was conjugated to Sepharose4B at 3 mgof
WGA per gram of dried cyanogen bromide-activated Sepharose as
described by the provider except that 25 mMN-acetyl glucosamine
was included in the conjugation buffer. To optimize bindingofcy-
tochrome b to the immobilizedWGA,thedetergentconcentration had
to be lower than its critical micellar concentration(0.8%).Abinding
and elution protocol was devised as follows: three partsMRB were
mixed with one part ofthe octylglucoside containing 100,000 g super-
natant to reduce the detergent concentration to 0.5% (wt/vol).
WGA-4B was then added at a ratio of 1.7 ml ofaffinitybeadsper4 ml
of diluted extract and rotated end over end at4°C overnight.After
washing the protein-bound affinitymatrix with 0.5%octylglucosidein
MRB, the cytochromeb was eluted with two bed volumes ofMRBplus
0.2% Triton X-100 for 1 h at room temperature.Theaffinitymatrix
was then reincubated for 30 min at room temperaturewith one bed
volume ofthe same buffer. The eluates werepooledand concentrated
10-fold with an Amicon PM 10 ultrafilter(Danvers, MA) (tempera-
Purification ofHumanGranulocyte Cytochromeb
The concentrated WGA-4B eluate was then diluted 10-fold with 10
mM Hepes, 0.1% Triton X-100, pH 7.4, to reduce the salt concentra-
tion to 50 mM. It was then pumped through a 5-ml column ofheparin
ultrogel at 15 ml/h and 40C. After washing with 5 column volumes of
40 mM NaCl, 10 mM KCl, I mM EDTA, 10 mM Hepes, pH 7.4, and
0.2% Triton X-100, the orange-colored cytochrome was eluted at a
flow rate of20 ml/h with a 150-ml linear gradient ofsalt starting with
wash buffer and increasing to 1.5 M NaCl made up in MRB with 0.2%
Triton X-100. 1 mM PMSF and 10 Mg/ml chymostatin were included
in the salt gradient for protease inhibition.
Velocity sedimentation ofcytochrome b in sucrose gradients. Final
purification ofthe cytochrome b was achieved by velocity sedimenta-
tion in sucrose gradients. Linear sucrose gradients (5-20% wt/vol, 5
ml) containing 0.1% (wt/vol) Triton X-100 were constructed and frac-
tionated in a manner similar to that described by Clarke (26). The
heparin ultrogel eluate was concentrated to a final cytochrome con-
centration of 2-4 MM using Centricon 30 microconcentrators (Ami-
con, Inc., Danvers, MA) and 0.325 ml overlaid on each sucrose gra-
dient. The gradients were centrifuged at 45,000 rpm for 20 h in an SW
50.1 rotor (Beckman Instruments, Inc.) at 4°C. Fractions of
0.35-0.375 ml were collected from the top with a densiflow pump
(Buchler Instruments Inc., Fort Lee, NJ).
Electrophoresis. SDS gel electrophoresis was carried out at room
temperature in polyacrylamide slab gels containing 0.1% (wt/vol) SDS
(27). Protein samples were mixed with an equal volume of sample
buffer and boiled for 3-4 min. Sample buffer consisted of 1 part 10%
(wt/vol) SDS in H20, 1 part 0.5 M Tris base, pH 6.8, 1 part glycerol,
and 500 mM 2-mercaptoethanol. In some experiments, sample buffer
contained 8 M urea. The electrophoretic mobility of protein samples
was compared with the mobility of standard proteins. Proteins were
visualized on slab gels by first staining for 30 min with 0.125% Coo-
massie Blue G 250 in 50% methanol, 10% acetic acid. Gels were then
destained in 25% isopropanol and 10% acetic acid and hydrated in
H20 overnight with several changes ofH20. Hydrated gels were then
silver stained under basic conditions as described by Wray et al. (28).
For comparative purposes some polyacrylamide gels were silver
stained under acidic conditions as described by Merril et al. (29).
Deglycosylation experiments. Deglycosylation studies were per-
formed on purified 1251-labeled cytochrome b. Peak sucrose gradient
fractions were iodinated with immobilized chloramine T (Iodobeads)
as described by the provider. Unbound iodine was removed by centrif-
ugation through a 3-ml column of Sephadex G-25 equilibrated with
column running buffer plus 0.1% Triton X-100.
Removal of N-linked carbohydrate was done with peptide:N-gly-
cosidase F as described by the provider except that incubations were
done in column running buffer (pH 7.4) plus 5% (wt/vol) sucrose, 2
mM PMSF, 10 mM 1,10-phenanthroline, and 5 mM EDTA. Degly-
cosylation was complete in 24 h at room temperature. Beta-elimina-
tion experiments were performed to investigate the possibility of 0-
linked carbohydrate (30). '251-labeled cytochrome b was incubated in
the presence of0.1 N NaOH for up to 24 h at 4°C. Samples were then
neutralized with an appropriate volume of 0.3 N HCG. As a positive
control, parallel beta elimination and N-glycosidase F studies were
performed in the presence of 5-10 Mg fetuin (31, 32). The electropho-
retic mobility ofreduced and treated samples was compared with that
ofreduced untreated samples.
Biochemical assays. Cytochrome b was quantitated by reduced
minus oxidized difference spectroscopy on a Cary 219 dual-beam spec-
trophotometer (Varian, Inc., San Jose, CA) assuming an extinction
coefficient of 29.3 mM-' cm-' (20). Samples were reduced by the
addition of 2MAof a 1.0 M solution of sodium dithionite made up in
H20 immediately before use.
Protein was measured by the Bradford method (33), the BCA
method as described by Pierce, Inc., and by fluorescence after conjuga-
tion of free amines with fluorescamine (34) using bovine serum albu-
min as a protein standard.
Cross-linking studies. Samples were treated with nonreversible ho-
mobifunctional cross-linking agent disuccinimidyl suberate (DSS)
(temperature, 20'C) as follows. To 100-200 ,l ofpurified cytochrome
taining 1% (wt/vol) Triton X-100, DSS was added (100 mM in di-
methylsulfoxide [DMSO])to a final concentration of0.25 mM (0.25%
DMSO). At different times aliquots were removed, and the reaction
was terminated by the addition of 0.25 M glycine (pH 7.4) to a final
concentration of 20 mM. Samples were then mixed with an equal
volume of sample buffer and subjected to SDS-PAGE as described in
the previous section. SDS gel profiles of cross-linked samples were
compared with the uncross-linked controls treated with DMSO only.
Production of rabbit antibodies to cytochrome b. Juvenile New
Zealand White rabbits (1.5 kg) were immunized with either cy-
tochrome b-enriched eluate from immobilized heparin or purified
cytochrome b. The initial immunization mixture consisted of20-30Mg
protein in I ml ofsaline plus 1 ml ofComplete Freund's Adjuvant. The
emulsified protein-adjuvant mixture was injected in 10-20 sites intra-
dermally just lateral to the rabbit's spine.
The next two immunizations were at 10-d intervals and consisted
of the same type ofintradermal injection but with an emulsified mix-
ture consisting of 10 Mg of protein in 0.5 ml saline plus 0.5 ml of
Incomplete Freund's Adjuvant. Rabbits were then immunized at 2-wk
intervals with the same mixture of protein and incomplete adjuvant,
and antibody titerswere monitored by enzyme-linkedimmunosorbant
assay (ELISA). Once antibody titers were sufficiently elevated, blood
was obtained from the central artery ofthe ear in 50-75 ml quantities,
depending on the size of the animal. Animals with elevated antibody
titers were maintained on a schedule consisting of reimmunizations
every 2-4 wk followed by bleeding 1 wk after the previous immuniza-
Detection ofantibodies. To monitor antibody levels in the rabbits,
we developed an ELISA. To round bottom microtiter plates we added
50 MAl of a 50-Mg/ml solution of poly-L-lysine (mol wt 14,000) in
DPBS(-) and incubated for 30 min at 37°C. Then we added 50Ml ofa
mixture containing 150-250 ngcytochrome bin 25 mM NaCI, 10mM
Hepes, pH 7.4, and0.02% Triton X-100 to the washed microtiter wells.
The plate was then incubated for at least 60 min at 37°C and washed
with DPBS(-). The microtiter wells were then treated with 100 Ml of
2% bovine serum albumin in DPBS for 15 min at 37°C. After shaking
the wells dry, 50 Ml ofprimary antibody diluted in 2% BSA/DPBS(-)
was added and incubated for either 60 min at 37°C or overnight at
4°C. The wells were then washed four to six times with DPBS(-). The
secondary antibody mixture consisted of 50 ,l of 1 Mg/ml of peroxi-
dase-conjugated goat anti-rabbit IgG in 2% BSA/DPBS(-) and was
incubated for 30 min at 37°C. After washing each well five or six times
with DPBS(-), 50 gl of colorimetric substrate which consisted of 0.5
mM ABTS and 5 mM H202 in 0.1 M citrate buffer, pH 4.2, was added.
After sufficient color development (10-30 min) the reaction was
stopped with 10 Ml of 10% SDS, and the absorbance read at 414 nm in
an automated microtiter plate scanner (Bio-Tek Instruments, Inc.,
Immunoprecipitation studies. The ability ofthe rabbit antibodies to
react with detergent-solubilized cytochrome b was investigated by in-
cubation of concentrated eluate from immobilized heparin with var-
ious amounts ofimmune or preimmune IgG for 1 h at 200C followed
by sedimentation for 10 h in 5-20% (wt/vol) sucrose gradients con-
taining 1% Triton X-100 as described in a previous section. Fractions
of 0.35 ml were collected from sucrose gradients with a densiflow
pump (Buchler Instruments Inc.). Sucrose gradient fractions were then
diluted with 0.25 ml of DPBS(-) and assayed for cytochrome b, as
described in the previous sections.
Western blotting experiments. Electrophoretic transfer of proteins
from SDS-polyacrylamide slab gels onto nitrocellulose was performed
according to Towbin et al. (35). Protein-bound nitrocellulose strips
were first incubated for 1 h in saturating buffer consisting of 10% goat
serum and 3% BSA in 0.5 M NaCl and 10 mM Hepes, pH 7.4. The
nitrocellulose strips were then incubated overnight at4°C with 1 Migper
ml of rabbit IgGin DPBS(-) plus 3% goat serum, 1% BSA, and 0.2%
Tween 20. After rinsing the nitrocellulose five times with wash buffer
1 MM) from concentrated peak sucrose gradient fractions con-
C. A. Parkos, R. A.Allen, C. G. Cochrane, and A. J. Jesaitis
consisting of0.25 M NaCl, 10mM Hepes, 0.2% Tween 20, pH 7.4, the
strips were then incubated for 1 h at 200C with 1 ug/ml ofperoxidase-
conjugated goat anti-rabbit IgG in DPBS(-) plus 3% goat serum, 1%
BSA, and 0.2% Tween 20. Again after rinsing five times with wash
buffer, the nitrocellulose strips were color developed for 5-30 min in a
solution ofdeveloper consisting of 30% methanol, 0.5 mg/mi peroxi-
dase color developer (4-chloro-1-napthol) (Bio-Rad Laboratories), and
5 mM H202 in 0.25 M NaCl, 10 mM Hepes, pH 7.4. The reaction was
terminated by the transfer of nitrocellulose strips to distilled water.
The purification of neutrophil cytochrome b. We designed a
scheme for purification of cytochrome b from human granu-
locytes based on what was known about its colocalization with
both the specific granules and plasma membrane (36-38) and
preliminary evidence about its properties as a putative integral
membrane glycoprotein (17, 23, 37) and ourown unpublished
observations (Parkos, C. A., A. J. Jesaitis, and R. A. Allen).
A summary of the specific activities and recoveries of cy-
tochrome b at various stages ofpurification is shown in Table
I. The membranes used for solubilization offered several ad-
vantages over solubilization ofwhole cells. First, degranulation
and nitrogen cavitation mobilized the internal pool of cy-
tochrome b to the plasma membrane while eliminating soluble
proteins and proteases. Second, using the cavitation buffer de-
scribed, recovery of membranes from the disrupted cells was
90-95%, based on partitioning of cytochrome between the
low-speed supernatants and pellets. Third, treatment ofmem-
branes with 1 M NaCl removed 15-20% ofthe total contami-
nating peripheral membrane protein while increasing the ex-
tractability ofthe cytochrome b by 20%. Therefore, the prepa-
ration of salt-washed membranes from degranulated cells
produced a significantly enriched and more extractable prepa-
ration with high yields ofcytochrome b. This preparation was
depleted of granular proteases and relatively free of cytosolic
and peripheral membrane protein contamination. SDS gel
electrophoresis ofthe preparation at each stage ofpurification
is shown in Fig. 1.
We tested the ability of numerous detergents to solubilize
cytochrome b from membranes and found octylglucoside and
Triton X-100 to be the most efficient. Octylglucoside was
found to extract virtually 100% of the cytochrome under the
conditions described and offered the advantage of having a
high critical micellar concentration. As shown in Table I, the
detergent extract had a mean cytochrome b content of 0.96
nmol/mg (- 0.26AMcytochrome b) with a protein concen-
tration of0.2-0.25 mg/ml.
We performed preliminary tests on whole cells to measure
the ability of lectins to induce cross-linking of cytochrome b
and hence inhibit detergent extractability. It was discovered
that treatment of cells with wheat germ agglutinin caused a
50% reduction in the extractability ofcytochrome b. The solu-
bilized cytochrome b was found to reproducibly bind to
WGA-4B. This binding could be completely inhibited by 100
mM N-acetylglucosamine. To obtain > 90% binding of the
cytochrome to WGA4B, the detergent concentration had to
be less than its critical micellar concentration. For elution of
the bound cytochrome, a combination of competing sugars,
high salt, and greater than critical micellar concentrations of
detergent were necessary to obtain good recoveries (65-85%).
The recovery ofcytochrome b was also sensitive to the density
of wheat germ agglutinin on beads, with
beads being the optimal density ofseveral tested. Elution ofthe
cytochrome in Triton X-100 rather than octylglucoside was
necessary to optimize yields in the subsequent purification
steps. As shown in Table I, the WGA-4B eluate had a cy-
tochrome b content of2.58 nmol per mg protein, which repre-
sents a threefold enrichment over the 100,000-g octyl gluco-
side extract. Elution ofbound cytochrome b with a gradient of
N-acetyl glucosamine or chitobiose offered no advantages be-
cause the protein slowly eluted over the entire gradient, and
the specific activity was the same as that from batch elution.
For optimized binding of the cytochrome to immobilized
heparin, the detergent had to be in excess ofthe critical micel-
lar concentration in a low-ionic strength buffer. The WGA-4B
eluate was concentrated and then diluted because of unsatis-
factory (40%) losses when dialysis was performed. Fig. 2 shows
the elution profile ofcytochrome b from heparin ultrogel. The
cytochrome eluted as a single sharp peak in a gradient ofNaCl,
whereas the protein profile revealed a shoulder comprised
mainly of a higher molecular weight species, Mr 170,000. As
shown in Table I, the heparin eluate was three to fourfold
enriched in cytochrome, and Fig. 1 shows the appearance of
1 mg lectin/ml of
Table I. Purification ofCytochrome b
Recoveries from 6 U ofblood
or -6 X 109 granulocytes
Cytochrome b-enriched membranes
2% octylglucoside 100,000 g supernatant
Wheat germ agglutinin-Sepharose eluate
Heparin ultrogel eluate
Peak sucrose gradient fractions
The specific contents and recoveries ofcytochrome b at various stages ofpurification as described in Methods. Values are reported ±SD. Cy-
tochrome b was quantitated by reduced-minus-oxidized difference spectroscopy as described in Methods assumingan extinction coefficient
(559-540 nm) of29.3 mM-' cm-' (27), Peak sucrose gradient fractions refer to the peak three fractions containing 50-60% ofthe applied cy-
tochrome b. Specific activities were based on BSA as a standard. Recoveries are based on starting cell contents. For comparative purposes,
values in parentheses refer to computations based on an extinction coefficient of21.6 mM-' cm-' (15).
Purification ofHuman Granulocyte Cytochrome b
S Mass (kD)
major protein staining bands with values ofMr 91,000 and Mr
22,000 along with several other protein bands.
As shown in Fig. 3, final purification was achieved by sedi-
mentation in sucrose density gradients. The concentrated hep-
arin eluate was sedimented for 20 h in 5-20% (wt/vol) sucrose
gradients, which separated 50-70% of cytochrome b spectral
activity from a major protein contaminant of 170,000 D. The
results shown in Table I indicate that the pooled peak sucrose
gradient fractions were enriched in cytochrome b
--- SO Concentration
Figure 2. Elution ofcytochrome b from heparin ultrogel. Eluate
from WGA-4B containing 35.8 nmol ofcytochrome b was pumped
through a 5-ml column of heparin ultrogel as described in Methods.
After washing the column with low-ionic strength buffer containing
0.2% Triton X-100, the cytochrome was eluted at a rate of20 ml/h
with a 150-ml gradient ofNaCl increasing from 0.04 to 1.5 M. Nano-
moles cytochrome b (solid circles) and micrograms protein (open cir-
cles) per fraction are plotted against fraction number. Salt concentra-
tion (dashed line) is also plotted against fraction number. Cy-
tochrome b was quantitated as described in Methods and the
recovery was 25.5 nmol or 71.2%. Protein was assayed by the BCA
protein assay kit as described by Pierce Chemical Co., using BSA as a
standard. One ofthree experiments.
Figure 1. Analysis ofthe purification ofcy-
tochrome b by SDS-PAGE. Protein samples from
various stages ofthe purification procedure were
subjected to SDS-PAGE on 8% (A) and 11% (B)
(wt/vol) polyacrylamide gels containing 0.1% SDS
and stained with silver as described in Methods.
Lane 1, 2.5 gg protein in (A) and 1.5 ug protein in
(B) from membranes enriched in cytochrome b.
Lane 2, 2.5 fig protein in (A) and 1.5 gg protein in
(B) from the 2% octyl glucoside, 100,000 g super-
natant. Lane 3, 2.5 tgprotein in (A) and 1.5;igof
protein in (B) ofeluate from WGA-4B. Lane 4,
both panels, 1.5 ug protein from heparin-Ultrogel
eluate. Lane 5, both panels, 0.75 ,g protein from
peak sucrose gradient fractions containing puri-
fied cytochrome b. (A), Molecular mass ofstan-
dard proteins is listed to the right, including myo-
sin (200 kD), phosphorylase B (97.4 kD), BSA (68
kD), ovalbumin (43 kD), and a-chymotrypsino-
gen (25.7 kD). (B) Lane marked S represents 1 Ag
each ofstandard with molecular mass listed to the
right. Standards in (B) include phosphorylase (94
kD), BSA (67 kD), ovalbumin (43 kD), carbonic
anhydrase (30 kD), and soybean trypsin inhibitor
over starting cell material with an overall yield between 20 and
30%. The composition of these fractions was analyzed by
SDS-PAGE. This analysis is shown in Figs.
reveals two protein-staining bands. These two bands were also
present when cytochrome b purification was performed on
purified plasma membranes obtained from unstimulated cells
(not shown), suggesting that their copunfication was not the
result of stimulation of cytochalasin-treated cells with formyl
polypeptides. The unusual staining pattern ofthe smaller pro-
tein component seen in Fig.
our gel system and not due to overlapping polypeptides of
similar molecular weight. Similar staining is seen on the low
molecular weight components ofour protein molecular weight
standards (Fig. 1, lane S). We were unable to resolve two
protein bands on larger SDS polyacrylamide gels.
Relative molecular weight estimates of the larger species
were found to increase with increasing polyacrylamide con-
centration in SDS gels. On 8% (wt/vol) SDS polyacrylamide
slab gels, the relative molecular weight ofthe larger species was
estimated at 91,000+3,000, whereas electrophoresis on 11%
(wt/vol) gels revealed a slightly increased Mr of97,000±3,000.
We have assigned the value of91,000 as the relative molecular
weight of the larger protein species to be consistent. The rela-
tive molecular weight ofthe smaller component was not vari-
able and was determined to be22,000+3,000 on both 11% and
15% (wt/vol) SDS polyacrylamide gels. The reduced minus-
oxidized absorbance difference spectrum of the purified cy-
tochrome b had alpha, beta, and Soret absorbances of 558.7,
528, and 426.5, which are in agreement with those reported by
others (20, 23).
Analysis ofglycosylation ofcytochrome b. Our preliminary
experiments on whole cells and those on the detergent extracts
of membranes which indicated that cytochrome b bound to
lectins suggested that cytochrome b was a glycoprotein or
closely associated with one. In addition, recent analyses by
Harper et al. (19) indicated that the neutrophil cytochrome b
from human myelogenous leukemia patients (Mr 68,000-
1 B and 3 B and
1 B is most likely an artefact of
C. A. Parkos, R. A. Allen, C. G. Cochrane, and A. J. Jesaitis
'- . .,---
@@ j -97.4
Figure 3. Final purification ofcytochrome b by sucrose density gra-
dient velocity sedimentation. 0.35 ml ofconcentrated eluate from
immobilized heparin with a cytochrome b concentration of 2.5AM
was sedimented for 20 h in 5-20% (wt/vol) linear sucrose density
gradients containing 0.1% Triton X-l00 and then fractionated as de-
scribed in Methods. (A) Percent total recovered cytochrome b spec-
tral activity is plotted against fraction number where fraction 1 repre-
sents the top ofthe gradient. Total recovered spectral activity was
90- 100% ofthat applied to the gradients. (B) 25 Ml ofeach corre-
sponding sucrose gradient fraction shown in (A) was mixed with an
equal volume ofsample buffer and subjected to SDS-PAGE on a 9%
(wt/vol) polyacrylamide gel and silver stained as described in
Methods. Lane S represents standard proteins with their molecular
mass listed to the left. Standards included myosin (200 kD), phos-
phorylase B (97.4 kD), BSA (68 kD), ovalbumin (43 kD), and a-chy-
motrypsinogen (25.7 kD). Profile shown above is typical of> 20 ex-
78,000, reference 23; Mr 72,000-90,000, reference 19) mi-
grated with increased mobility by SDS-PAGE after treatment
with endoglycosidase F. Estimates from carbohydrate analyses
ofpurified preparations indicated that 15% ofthe mass oftheir
preparation was carbohydrate. However, because of uncer-
tainties in the reported SDS-PAGE mobilities and glycopro-
tein contaminants present in their preparation, an indepen-
dent analysis ofthe glycosylation ofcytochrome b was neces-
Glycoproteins often exhibit anomalous behavior when
subjected to SDS-PAGE (39, 40). Thus, to obtain an estimate
of the molecular weight of the polypeptide portion of cy-
tochrome b, the effect ofdeglycosylation on its relative molec-
ular weight was studied. Purified cytochrome b was iodinated
as described in the methods, which by SDS-PAGE and subse-
quent autoradiography revealed two bands of Mr 91,000 and
22,000. As shown in Fig. 4, beta elimination conditions that
decreased the relative molecular weight of fetuin controls had
a minimal effect on the mobility of the Mr-91,000 protein
Figure 4. Autoradiograms demonstrating effect ofdeglycosylation on
Mr of [I25Ilcytochrome b. Deglycosylation experiments were per-
formed on cytochrome b as described in Methods. 10,000 cpm
oflabeled protein was subjected to reduced SDS gel electrophoresis
on 8% (A) and 1% (B) (wt/vol) polyacrylamide gels as described in
Methods. Gels were fixed, dried, and subjected to autoradiography
for 48 h. Lane 1, both panels, untreated control cytochrome b. Lane
2, both panels, cytochrome b after 24 h ofbeta-elimination in 0.1 N
NaOH at 4°C. Lane 3, both panels, cytochrome b after peptide:N-
glycosidase F treatment for 24 h at 20°C. In both panels, the molec-
ular masses in kilodaltons ofprotein standards are included to the
right oflane, including myosin (200 kD), phosphorylase B (97.4 kD),
BSA (68 kD), ovalbumin (43 kD), -chymotrypsinogen (25.7 kD), and
beta-lactoglobulin (18.4 kD).
band. Treatment with peptide:N-glycosidase F, on the other
hand, caused a large increase in the electrophoretic mobility of
the Mr-91,000 band, which suggests the presence ofsignificant
amounts ofN-linked carbohydrate. A single band with anM,
48,000-50,000 appeared after endoglycosidase treatment
along with total disappearance ofthe Mr-9 1,000 band. Neither
treatment had any effect on the electrophoretic mobility ofthe
Cross-linking studies. To investigate whether the two poly-
peptides of Mr 91,000 and 22,000 present in purified cy-
tochrome b preparations are indeed closely associated, cross-
linking studies were performed. Purified cytochrome b at an
approximate concentration of 0.75 uM was exposed to the
homobifunctional cross-linking reagent DSS under conditions
that do not cross-link control standard proteins. As shown in
Fig. 5, this exposure resulted in a time-dependent decrease in
the staining density of the Mr91,000 and Mr-22,000 protein
constituents with the appearance ofa new band having a lower
electrophoretic mobility. This new band had a value of Mr
120,000-135,000 on the 11% (wt/vol) SDS-polyacrylamide gel
shown in Fig. 6. No additional bands were produced either in
the cytochrome-containing sample, or in the more concen-
trated protein standard solution. Because there was no detect-
able effect of cross-linking on spectra or the sedimentation
characteristics shown in Fig. 3, a close association issuggested
between the 91,000- and 22,000-D polypeptides duringtime of
exposure with cross-linkers.
Antibody production. To further obtain specific probes for
cytochrome b, antibodies were produced against our purified
Purification ofHuman Granulocyte Cytochrome b
Figure 5. Cross-linking ofdetergent-solubilized cytochrome b. 1Iug
ofpurified cytochrome b (lanes 1-5) or 1jtgeach ofstandard pro-
teins (lanes S2-S5) were cross-linked with 0.25 mM DSS for various
lengths oftime and then subjected to reduced SDS-PAGE on 11%
(wt/vol) polyacrylamide gel as described in Methods. Standard pro-
teins included phosphorylase B (94 kD), BSA (67 kD), ovalbumin
(43 kD), carbonic anhydrase (30 kD), and soybean trypsin inhibitor
(20 kD). Lanes I andSi,control cytochrome b, standards (no cross-
linking). Lanes 2 and S2, 2 min ofcross-linking. Lanes 3 and S3, 10
min ofcross-linking. Lanes 4 and S4, 20 min ofcross-linking. Lanes
5 and S5, 40 min ofcross-linking. Lane 6, sample buffer only, no
preparation. To assay antibody titers, we developed an ELISA.
Pretreatment of microtiter wells with poly-L-lysine was neces-
sary because the detergent in our cytochrome preparation
completely inhibited its binding to the surface of the wells.
Using '251I-labeled cytochrome b, we found that
total counts in any given microtiter well could be bound pro-
vided the well had been pretreated with poly-lysine and that
the total salt concentration was < 100 mM.
Ofthe six rabbits immunized with cytochrome b, only one
(R3179) responded with favorable antibody titers. The im-
munogen used for R3 179 was eluate from immobilized hepa-
rin, which had a high cytochrome b specific activity of 11.8
nmol heme/mg protein. This rabbit's antibody titers became
maximal at 6 wk with a titer that diluted to 1:32,000. The
antibody titers ofR3179 progressively declined after the 6- and
8-wk peaks. Boosting the animal with purified cytochrome b
did not increase the antibody titers.
Immunoprecipitation studies. We examined the ability of
IgG from R3179 to immunoprecipitate cytochrome b spec-
trum from detergent extracts. It was necessary to use purified
immunogobulin because contaminating hemoglobin in serum
samples interfered with the measurement ofthe cytochrome b
spectrum. To determine if antibody was reacting with cy-
tochrome b, we compared the sucrose density gradient sedi-
mentation profiles ofcytochrome-treated with either immune
or control IgG. The results of the sedimentation studies are
shown in Fig. 7. Increasing the amount ofimmune IgG incu-
bated with eluate from immobilized heparin resulted in a cor-
responding increase in the amount of cytochrome spectrum
sedimented to the bottom of the sucrose gradient. There was
no effect of preimmune IgG on the sedimentation profile of
Western blotting experiments. Determination of the anti-
gen with which the antibody reacts was carried out by Western
- 50% ofthe
Figure 6. Western blots ofR3179 IgG to cytochrome b preparations
and CGD neutrophils. Protein samples were subject to SDS-PAGE
and then transferred to nitrocellulose as described in Methods. The
nitrocellulose strips were incubated with 1 ,ug/ml R3179 IgG fol-
lowed by 1 ;&g/mlofperoxidase-conjugated goat anti-rabbit IgG and
then color developed as described in Methods. (A) Western blot from
a 7-20% (wt/vol) polyacrylamide gradient gel. Lane 1, 30 ,ug mem-
brane protein from degranulated neutrophils; lane 2, 61Ag protein
from the 2% octylglucoside 100,000-g supernatant; lane 3, 6 yg pro-
tein from eluate from immobilized wheat germ agglutinin; lane 4, 3
,ug protein from eluate from heparin ultrogel; lane 5, 0.5 ug ofpuri-
fied cytochrome b. (B) Purified, DFP-treated granulocytes were solu-
bilized at 00C in buffer containing 3 mM MgCl2, 2 mM PMSF, 20
Mg/rml chymostatin, 1% Triton X-100, and 15 mM Hepes, pH 7.4,
and centrifuged at 12,000 g (temperature, 40C) for 2 min to remove
insoluble debris. The supernatants were mixed with an equal volume
ofsample buffer, subjected to SDS-PAGE on a 6-16% (wt/vol) poly-
acrylamide gradient gel and Western blotted as described in
Methods. Lane 1, 0.5 ,g purified cytochrome b; lanes 2, 3, 5, 50 Ag
protein from three different preparations ofnormal granulocytes
C6D granulocytes (- 25 X 106 cell equivalents). The molecular mass
ofprotein standards is included to the right of(A) and (B).
1 X 106 cell equivalents). Lane 4, 50;&g protein from X-linked
blotting techniques. Preliminary experiments using '25I-la-
beled cytochrome b revealed that the Mr-91,000 subunit was
difficult to electrophoretically transfer from SDS polyacryl-
amide gels onto nitrocellulose. We found that at a transfer
current and voltage of0.4 A and 115 V,
component was bound to nitrocellulose after 4 h, which in-
C. A. Parkos, R. A. Allen, C. G. Cochrane, andA. J. Jesaitis
VW m Wm IW'MW-
tation ofcytochrome b.
Represented are the sucrose
density gradient profilesof
cytochrome b after incuba-
tion with either R3179 IgG
or control preimmune IgG.
Concentrated eluate from
immobilized heparin with
bated with either immune
or preimmune IgG at the
concentrations shown for
60 min and sedimented in
linear 5-20% sucrose den-
Triton X-100 as described
Methods. Percent of
total recovered cytochrome
b spectrum isplotted
against fraction number,
is the top
of the gradient. (Open cir-
preimmune IgG. (Closed
circles) Treatment with im-
mune IgG. Spectral recov-
eries were 90-100% ofap-
plied material except for
highest immune IgG
treatment profile in (C),
which was 71%.
10 12 14 16
creased to 45% after 18 h. TheMr-22,000 subunit, on the other
hand, was 80% transferred after 4 h. Fig. 6 A shows the West-
ern blotting profiles of R3179 IgG to protein samples taken
from variousstages ofpurification ofcytochrome b and sepa-
rated on a 7-20% (wt/vol) polyacrylamide gradient gel. Note
especiallyin lanes 4 and5, where there is at least 0.5 and 0.2,g
of this species bound to the nitrocellulose, respectively. The
mostlikely explanationfor theblotting profilesshown inFig. 6
A is that R3 179 IgGisreacting primarily with the Mr-22,000
componentofcytochromeb.However,we cannot exclude the
possibilityofthe loss ofantigenicityofthelargersubunitupon
SDS denaturation andelectrophoretic manipulation.
The presence of other labeled protein bands in Fig. 6 A
suggests that the antibody reacts with otherneutrophil pro-
teins. As shown in lane 4, the antibodyhas labeled polypep-
tides withMrvalues of22,000, 35,000, and 170,000.Because
this lane contains the immunogen to which the rabbit was
immunized,thesepolypeptidesareprobably recognized specif-
ically.The other labeled band ofMr 100,000-1 10,000 present
in the more crude protein sample, lanes 2 and 3, is probably
the result ofnonspecific labelingorcross-reactivityofthe anti-
body. Western blots ofpreimmune IgGto the same fractions
resulted in acompleteabsence of labeledproteinbands(data
Fig.6 Bdepictsacomparison of the Western blotting pro-
files of R3179 IgG to granulocytes from normal individuals
and apatient (J.C.)with X-linked CGD. Theneutrophils from
this CGD patient have a spectrophotometric absence of cy-
tochrome b and have no NADPH oxidase activity (41, 42).
Comparison ofthe blotting profile ofX-linked CGD gran-
ulocytes (Fig. 6 B, lane 4) to the profiles of normal granulo-
cytes (lanes 2, 3, 5) reveals a complete absence of the
Mr-22,000 polypeptide in the CGD neutrophils. Since the
blotting profiles ofthe higher molecular weight bands are the
same in the control vs. CGD neutrophil lanes, the absence of
labeling ofthe Mr-22,000 polypeptide in lane 4 is not likely to
be a staining artefact. We have also confirmed the absence of
the Mr-22,000 polypeptide in five other patients with X-linked
CGD. Also shown in Fig. 6 B is the blotting profile ofpurified
cytochrome b (lane 1), which again demonstrates a lack of
reactivity of the antibody with the Mr-91,000 component of
The purpose of this study was to purify cytochrome b with a
sufficiently high yield from human peripheral blood neutro-
phils so as to be practical for biochemical and immunological
analysis. Specific emphasis was placed on the investigation of
the possible glycoprotein nature of cytochrome b and to re-
solve current discrepancies reported in the literature for the
size ofthe cytochrome.
The two protein species ofMr 91,000 and 22,000 identified
in this report may possibly explain these latter discrepancies.
The values reported by Harper et al. (19, 23) and Lutter et al.
(20) of 70-90 kD and 127 kD seem to be in approximate
agreement with our Mr-91,000 protein species. Because the
large polypeptide is heavily glycosylated, a slight technical dif-
ference in electrophoretic systems might be enough to explain
the differences. The lower molecular weight values of 11-14
kD and 32 kD reported by Pember et al. (17) and Bellavite et
al. (18) are in approximate agreement with our lower molecu-
lar weight species with differences probably due to either pro-
teolysis and/or interspecies variation, because bovine and por-
cine granulocytes were used in these latter studies.
We believe that the reason for the lack of reports of both
proteins in purified cytochrome b preparations lies in the vari-
ability in which these two proteins stain on SDS polyacryl-
amide gels. We found that using Coomassie Blue as a primary
method of protein visualization was unsatisfactory because of
the sizeable amounts of protein required (> 20 ,ug) for ade-
quate staining intensity. To get satisfactory protein visualiza-
tion, we found it necessary to silver stain under basic condi-
tions after Coomassie staining. Fig. 8 demonstrates that silver
staining under basic conditions (28) is superior to acidic silver
stains (29) in visualizing both protein bands. Integration ofthe
densitometric scans shown in Fig. 5 demonstrates two points.
First, acidic silver staining is only 30-40% as effective as basic
silver staining in visualizing cytochrome b. Second, at
low staining intensities, the relative staining densities of
the Mr-22,000 and M,-91,000 species exhibit significant vari-
From the lectin binding data in this report and those from
Harper et al. (19), it has become apparent that cytochrome b is
a glycoprotein. Our results suggest that, ofthe Mr 91,000 and
22,000 protein we describe, only the large subunit is signifi-
cantly glycosylated. The deglycosylation experiments in this
report indicate that the Mr91,000 component of cytochrome
Purification ofHuman Granulocyte Cytochrome b
-0.156 mgiml IgG
Figure 8. Visualization of
cytochrome b on SDS
ious protein-staining proce-
dures. Purified cytochrome
to SDS-PAGE on 11% (wt/
vol) polyacrylamide gels
and stained with silver
under basic conditions (A),
silver under acidic condi-
tions (B), or Coomassie
Blue G-250 (C) as de-
scribed in Methods. Lane 2
of each panel represents
SDS-PAGE of control
buffer containing no cy-
tochrome b or protein, and
lanes marked S represent
standard proteins (1
each) with their molecular
mass listed to the left. Stan-
lase B (94 kD), BSA (67
kD), ovalbumin (43 kD),
carbonic anhydrase (30
U)), and soybean trypsin
inhibitor (20.1 kD). The
densitometric scans of
lanes I and 2 of each panel
are shown to the right of
the respective gel lanes.
Mug) shown in lane I
b contains large amounts of N-linked carbohydrate and little
or no O-linked sugars. This result suggests that the polypeptide
core ofthe large subunit has a molecular mass of- 50,000 D.
Although we cannot conclude that the smaller Mr-22,000
component of cytochrome b is not glycosylated, our inability
to detect either N- or O-linked carbohydrate suggests it is not
glycosylated to any substantial extent. Additionally, Western
blotting experiments with peroxidase-conjugated wheat germ
agglutinin have failed to label the M,-22,000 protein under
conditions that label the Mr-9 1,000 component (Parkos, C. A.,
A. J. Jesaitis, and R. A. Allen, unpublished observations).
Because it is unlikely that the Mr91,000 and Mr-22,000
proteins described in this report would copurify randomly, we
hypothesize that these two proteins are closely associated at
least in detergent extracts. Evidence supporting the association
of these two proteins is significant. First, we are unable to
detect carbohydrate on the smaller subunit even though lectin
affinity chromatography was employed as a purification step.
Second, the two proteins copurify identically on gel filtration
columns (Parkos, C. A., A. J. Jesaitis, and R. A. Allen, unpub-
lished observations) and sucrose density gradients even though
their size disparity should have permitted their resolution.
Third, the two proteins can be cross-linked under conditions
that do not cross-link standard control proteins. Furthermore,
cross-linking does not influence the sedimentation behavior of
the cytochrome in sucrose density gradients. Lastly, antibody
that immunoprecipitates spectral activity from detergent ex-
tracts binds to the Mr-22,000 species and not the Mr91,000
species on Western blots.
The nature of the apparent association between the
Mr-9 1,000 and Mr-22,000 proteins is not clear. It is not due to
disulfide linkage because no sensitivity to reduction is ob-
served when comparing SDS-PAGE analysis under nonreduc-
ing and reducing conditions. In addition it is not dependent on
the activation state of the cells on which the purification was
performed. Although we cannot absolutely rule out proteolysis
as the origin ofthe smaller protein, every precaution was taken
to minimize proteolysis. Cells were treated with DFP before
degranulation and cavitation. Degranulation before lysis re-
duced the content of granule hydrolytic enzymes in the cavi-
tate. Nitrogen cavitation was performed under conditions that
minimize the breakage ofany remaining granules, and all sub-
sequent purification steps contained phenylmethylsulfonic
acid and chymostatin, which have been shown to inhibit pro-
teolysis in whole neutrophil detergent extracts by greater than
85% (43). The measures taken to avoid proteolysis appear to
have been successful because the SDS polyacrylamide gel
banding patterns at the various stages ofpurification are stable
for months at 4VC with no apparent changes.
Our finding of the absence of the Mr-22,000 species in
granulocytes from patients with X-linked CGD provides
strong evidence that this protein is a functional component of
cytochrome b. Therefore, ifwe assume that the copurification
is not the result ofsome nonspecific association occurring after
cell lysis or membrane solubilization, then our results would
suggest that the human neutrophil cytochrome b is a hetero-
dimer ofa heavily glycosylated polypeptide (Mr 91,000) with a
polypeptide core of
polypeptide. In addition to the evidence in this report, this
hypothesis is supported by our hydrodynamic analysis of the
size of the detergent-solubilized cytochrome showing that the
molecular mass ofthe cytochrome is
rough correspondence of staining density ofthe two polypep-
tides separated on silver-stained SDS-PAGE shown in Fig. 8 A
is also not inconsistent with this view.
It is not known which ofthe two polypeptides described in
this report contains the heme prosthetic group. In our hands,
separation ofthe two species has only resulted in denaturation
of the heme spectrum. Comparison of the properties of our
protein preparation with those of other nonmitochondrial
membrane-bound b-cytochromes such as cytochrome b56, (44,
45) and cytochrome b5 (46) would suggest that the Mr22,000
species is more likely to contain heme due to its size and
apparent lack of carbohydrate. The absence ofthe Mr-22,000
species in CGD granulocytes with a spectrophotometric ab-
sence ofcytochrome b supports this contention.
- 50 kD and an unglycosylated 22-kD
100-135 kD (21). The
C. A. Parkos, R. A. Allen, C. G. Cochrane, and A. J. Jesaitis
If the smaller polypeptide we describe contains the heme
prosthetic group then our findings are compatible with the
recent results of Royer-Pokora et al. (47) who have reported
the sequence of the transcript for the gene responsible for X-
linked, cytochrome b-negative CGD. The transcript they re-
port does not show any significant homology to previously
sequenced cytochromes and suggests the absence of a heme-
binding region. In addition, the amino acid composition is
different than that reported by Harper et al. (23) and Lutter et
al. (20) for purified cytochrome b. However, because the
amino analyses reported by these authors was not performed
on the individual polypeptides resolved by SDS-PAGE they
cannot be expected to match those reported by Royer-Pokora.
This problem is further compounded by other glycoprotein
contaminants in the preparation acknowledged by Harper et
al. yet still used for the amino acid analyses.
The size and sequence ofthe transcript reported by Royer-
Pokora et al. is, however, consistent with our view of the size
and physicochemical characteristics of the Mr-9 1,000 species.
This putative subunit is heavily glycosylated and may not
carry the heme group. Furthermore, because the Mr-22,000
polypeptide is absent in patients with X-linked CGD, it is
possible that the presence of functional Mr-91,000 subunit is
required for the cellular processing of the Mr-22,000 compo-
Note Added in Proof
During the review of this manuscript, we engaged in a collaboration
with M. Dinauer and S. Orkin and confirmed that the large subunit of
purified cytochrome b was indeed the product of the X-linked gene
whose deletion is responsible for CGD (48). This confirmation was
based on the finding that antibodies made against cDNA-derived syn-
thetic peptides and fusion proteins cross-reacted with the Mr91,000
polypeptide in its intact and deglycosylated form. Subsequently, Segal
(49) published a revision of his earlier purification procedure which
included an undocumented velocity sedimentation step and a second
revision for the molecular weight of cytochrome b. Also reported was
the existence of a copurifying 23-kD protein and the absence of both
proteins in granulocytes from X-linked CGD patients. The latter con-
clusion was based on the assumptions that (a) the cytochrome purifies
identically from CGD as from normal cells, and (b) the antibody used
to detect the 23-kD protein was specific for the cytochrome in spite of
its inability to immunoprecipitate the inferred "a-#" heterodimeric
complex. In addition, because the amino acid composition for the
large protein determined by him and co-workers did not match that
predicted by sequence of the X-CGD gene, Segal also concluded that
the cytochrome was not coded forby this gene. Most recently however,
Segal and co-workers reversed their conclusion and published the
amino acid sequence for the 90-kD polypeptide which matched that
predicted by Royer-Pokora et al. (47) with an additional piece at the
NH2-terminalend (50). This latter observation now fully confirms our
purification and results. In addition, our physicochemical studies on
the hydrodynamic and cross-linking properties ofthis cytochrome (21,
51) present a much stronger case for its heterodimeric nature than a
Special thanks to Dr. J. Curnutte for important discussion and for
generously supplying CGD neutrophils. We, wish to thank Velda
Comstock and Dian Caudebec for excellent secretarial and editorial
This work was supported by United States Public Health Service
grants AI-17354, ROI AI-22735, and RR-00833. Charles A. Parkos is
the recipient of National Institute of General Medicine Sciences Na-
tional Research Award PHFGM07198 from the University ofCalifor-
nia School ofMedicine, San Diego, CA. A. J. Jesaitis is the recipient of
an American Heart Association (AHA) Established Investigator
Award, with funds contributed in part by the California Affiliate of
1. McRipley, R. J., and A. J. Sbarra. 1967. The role of the phago-
cytein host-parasiteinteractions. J. Bacteriol. 94:1417-1424.
2. Mandell, G. L. 1974. Bactericidal activityofaerobic and anaer-
obic polymorphonuclear neutrophils. Infect.Immun. 9:337-341.
3. Lehrer, R. I., and M. J. Cline. 1969. Interaction of Candida
albicans with humanleukocytesand serum. J. Bacteriol. 98:996-1004.
4. Babior, B. M., R. S. Kipnes,and J. T. Curnutte. 1973.Biological
defense mechanisms: production by leukocytes ofsuperoxide.Apo-
tential bactericidal agent.J. Clin. Invest. 52:741-744.
5. Briggs, R. T., D. B. Drath, M. L. Karnovsky, and M. J. Kar-
novsky. 1975. Localization ofNADH oxidase on the surface ofhuman
polymorphonuclear leukocytes bya newcytochemicalmethod. J. Cell
6. Iyer, G. Y.N.,M. F.Islam,and J. H.Quastel.1961. Biochemical
aspects ofphagocytosis.Nature (Lond.). 192:535-541.
7. Badwey, J. A., and M. L. Karnovsky. 1980. Activeoxygen spe-
cies and the functions ofphagocytic leukocytes.Annu. Rev. Biochem.
8. Karnovsky,M. L.,and J. A.Badwey. 1983. Determinants ofthe
productionofactiveoxygen species by granulocytesandmacrophages.
J. Clin. Chem. Clin. Biochem. 21:545-553.
9. Segal, A. W., and 0. T. G. Jones. 1978. Novelcytochromeb
10. Segal, A. W., and 0. T. G. Jones. 1979. Reduction and subse-
quentoxidation ofacytochromeb ofhumanneutrophilsafter stimula-
tion with phorbol myristateacetate. Biochem. Biophys.Res. Comm.
11. Segal, A. W., 0. T. G. Jones, D. Webster,and A. C. Allison.
1978. Absence of a newlydescribedcytochromeb fromneutrophilsof
patientswith chronic granulomatousdisease. Lancet. ii:446-449.
12. Hamers, M. N., M. deBoer,L. J. Meerhof,R. S. Weening,and
D. Roos. 1984. Complementationinmonocyte hybrids revealing ge-
netic heterogeneityin chronicgranulomatousdisease. Nature(Lond.).
13. Gabig,T. G.,E. W. Schervish,and J. T.Santinga.1982. Func-
tional relationship of the cytochromeb to thesuperoxide-generating
oxidase ofhumanneutrophils.J. Biol. Chem. 257:4114-4119.
14. Bellavite, P., A. R. Cross,M. C. Serra, 0. T. G. Jones,and F.
Rossi. 1983. Thecytochromeband flavin contentandpropertiesofthe
O° formingNADPH oxidase solubilized from activated neutrophils.
Biochim. Biophys.Acta. 746:40-47.
15. Cross, A. R., F. K. Higson, and 0. T. G. Jones. 1982. The
enzymic reduction and kinetics of oxidation ofcytochrome b245Of
neutrophils.Biochem. J. 204:479-485.
16. Cross, A. R., 0. T. G. Jones,A. M. Harper,and A. W.Segal.
1981. Oxidation-reductionpropertiesofthecytochromeb found in the
plasma membrane fraction of human neutrophils.Biochem. J.
17. Pember, S. O.,B. L.Heyl,J. M. Kinkade, Jr.,and J. D. Lam-
beth. 1984. Cytochrome b558 from (bovine) granulocytes.J. Biol.
18. Bellavite, P.,E.Papini,L.Zeni,V. DellaBianca,and F. Rossi.
1985. Studies on the nature and activation ofO-formingNADPH
oxidase ofleukocytes.Identification of aphosphorylated component
of the active enzyme. Free Radical Research Communications.
Purification ofHuman Granulocyte Cytochromeb
19. Harper, A. M., M. J. Dunne, and A. W. Segal. 1984. Purifica-
tion of cytochrome b-245 from human neutrophils. Biochem. J.
20. Lutter, R., M. L. J. van Schaik, R. V. van Zwieten, R. Wever,
D. Roos, and M. N. Hamers. 1985. Purification and partial character-
ization of the b-type cytochrome from human polymorphonuclear
leukocytes. J. Biol. Chem. 260:2237-2244.
21. Parkos, C. A., R. A. Allen, C. G. Cochrane, and A. J. Jesaitis.
1986. Characterization of purified cytochrome b559 from the plasma
membrane of stimulated human granulocytes. J. Cell Biol. 103:51Oa.
22. Henson, P. M., and Z. G. Oades. 1975. Stimulation ofhuman
neutrophils by soluble and insoluble aggregates. J. Clin. Invest.
23. Harper, A. M., M. F. Chaplin, and A. W. Segal. 1985. Cy-
tochrome b-245 from human neutrophils is a glycoprotein. Biochem.
24. Amrein, P. C., and T. P. Stossel. 1980. Prevention ofdegrada-
tion ofhuman polymorphonuclear leukocyte proteins by diisopropyl
fluorophosphate. Blood. 56:442-447.
25. Borregaard, N., and A. I. Tauber. 1984. Subcellular localization
ofthe human neutrophil NADPH oxidase. J. Biol. Chem. 259:47-52.
26. Clarke, S. 1975. The size and detergent binding of membrane
proteins. J. Biol. Chem. 250:5459-5469.
27. Laemmli, U. K. 1970. Cleavage of structural proteins during
the assembly of the head of bacteriophage T4. Nature (Lond.).
28. Wray, W., T. Boulikas, V. Wray, and R. J. Hancock. 1981.
Silver staining of proteins in polyacrylamide gels. Anal. Biochem.
29. Merril, C. R., D. Goldman, S. A. Sedman, and M. H. Ebert.
1981. Ultrasensitive stain for proteins in polyacrylamide gels shows
regional variation in cerebral spinal fluid proteins. Science (Wash.
30. Spiro, R. G. 1972. The carbohydrates of glycoproteins.
Methods Enzymol. 28:31-43.
31. Edge, A. S., C. R. Faltynek, L. Hof, L. E. Reichert, Jr., and P.
Weber. 1981. Deglycosylation of glycoproteins by trifluoromethane-
sulfonic acid. Anal. Biochem. 118:131-137.
32. Spiro, R. G., and V. D. Bhoyroo. 1974. Structure ofthe O-gly-
cosidically linked carbohydrate units of fetuin. J. Biol. Chem.
33. Bradford, M. 1976. A rapid and sensitive method for the quan-
titation of microgram quantities of protein utilizing the principle of
protein-dye binding. Anal. Biochem. 72:248-254.
34. Undenfreind, S., S. Stein, P. Bohlen, W. Dairman, W. Leim-
graber, and M. Weigle. 1972. Fluorescamine: a reagent for assay of
amino acids, polypeptides, proteins and primary amines in the pico-
mole range. Science (Wash. DC). 178:871-872.
35. Towbin, H., T. Staehelin, and J. Gordon. 1979. Electrophoretic
transfer of proteins from polyacrylamide gels to nitrocellulose sheets:
proceedure and some applications. Proc. Natl. Acad. Sci. USA.
36. Segal, A. W., and 0. T. G. Jones. 1979. The subcellular distri-
bution and some properties of the cytochrome b component of the
microbicidal oxidase system of human neutrophils. Biochem. J.
37. Borregaard, N., J. M. Heiple, E. R. Simons, and R. A. Clark.
1983. Subcellular localization ofthe b-cytochrome component of the
human neutrophl microbicidal oxidase: translocation during activa-
tion. J. Cell Biol. 97:52-61.
38. Parkos, C. A., C. G. Cochrane, M. Schmitt, and A. J. Jesaitis.
1985. Regulation of the oxidative response ofhuman granulocytes to
chemoattractants: no evidence for stimulated traffic ofredox enzymes
between endo and plasma membranes. J. Biol. Chem. 260:6541-6547.
39. Frank, R. N., and D. Rodbard. 1975. Precision of sodium
dodecyl sulfate-polyacrylamide gel electrophoresis for the molecular
weight estimation ofa membrane glycoprotein: studies on bovine rho-
dopsin. Arch. Biochem. Biophys. 171:1-13.
40. Segrest, J. P., and R. L. Jackson. 1972. SDS gel electrophoresis
ofglycoproteins. Methods Enzymol. 28:54-63.
41. Gabig, T. G., and B. A. Lefker. 1984. Deficient flavoprotien
component of the NADPH-dependent Q2 generating oxidase in the
neutrophils from three male patients with chronic granulomatous dis-
ease. J. Clin. Invest. 73:701-705.
42. Curnutte, J. T., R. Kuver, and P. J. Scott. 1987. Activation of
neutrophil NADPH oxidase in a cell-free system: partial purification of
components and characterization of the activation process. J. Biol.
Chem. In press.
43. Sheterline, P., and C. R. Hopkins. 1981. Transmembrane link-
age between surface glycoproteins and components ofthe cytoplasm in
neutrophil leukocytes. J. Cell Biol. 90:743-754.
44. Apps, D. K., J. G. Pryde, and J. H. Phillips. 1980. Cytochrome
b561 is identical with chromomembrin B, a major polypeptide ofchro-
maffin granule membranes. Neuroscience. 5:2279-2287.
45. Doung, L. T., and P. J. Fleming. 1982. Isolation and properties
ofcytochrome b561 from bovine adrenal chromaffin granules. J. Biol.
46. Spatz, L., and P. Strittmaher. 1971. A form of cytochrome b5
that contains an additional hydrophobic sequence of 40 amino acid
residues. P.N.A.S. 68:1042-1046.
47. Royer-Pokora, B., L. M. Kunkel, A. P. Monaco, S. C. Goff,
P. E. Newburger, R. L. Baehner, F. S. Cole, J. T. Curnutte, and S. H.
Orkin. 1986. Cloning the gene for an inherited human disorder-
chronic granulomatous disease-on the basis of its chromosomal lo-
cation. Nature (Lond.). 322:32-38.
48. Dinauer, M. C., S. H. Orkin, R. Brown, A. J. Jesaitis, and C. A.
Parkos. 1987. The glycoprotein encoded by the X-linked chronic
granulomatous disease locus is a component of the neutrophil cy-
tochrome b complex. Nature (Lond.). 327:717-720.
49. Segal, A. W. 1987. Absence of both cytochrome b245 subunits
from neutrophils in X-linked chronic granulomatous disease. Nature
50. Teahan, C., P. Rowe, P. Parker, N. Totty, and A. W. Segal.
1987. The X-linked chronic granulomatous disease gene codes for the
13 chain ofcytochrome b245. Nature (Lond.). 327:720-721.
51. Parkos, C. A., R. A. Allen, C. G. Cochrane, and A. J. Jesaitis.
1987. The b-cytochrome from human granulocytes purifies as an
M,= 91,000/Mr = 22,000 heterodimer. J. Clin. Invest. 35:652A.
C. A. Parkos, R. A. Allen, C. G. Cochrane, andA. J. Jesaitis