www.elsevier.com/locate/jim

Journal of Immunological Methods 291 (2004) 101–108

Research paper

The induction of eosinophil peroxidase release: improved methods

of measurement and stimulation

Darryl J. Adamkoa,*, Yingqi Wua, Gerald J. Gleichb, Paige Lacyc, Redwan Moqbelc

aThe Pulmonary Research Group, Department of Pediatrics, University of Alberta, Edmonton, Alberta, CanadabDepartment of Dermatology, University of Utah, Salt Lake City, UT, USA

cThe Pulmonary Research Group, Department of Medicine, University of Alberta, Edmonton, Alberta, Canada

Received 21 January 2004; received in revised form 4 May 2004; accepted 12 May 2004

Available online 20 June 2004

Abstract

The production and release of eosinophil peroxidase (EPO) has been associated with human pathology. Degranulation

assays with eosinophils are typically very difficult to do, with very low release values. EPO is unique for its high cationic

charge. As such, it adheres to most extracellular surfaces, rendering it more difficult to measure compared with other released

cellular proteins. Based on the understanding of the sticky nature of EPO, we were concerned that EPO released in vitro cannot

be reproducibly measured in the supernatants of stimulated cells. Instead, we suspected that much of the released EPO was left

adherent to the tube walls. We chose to investigate the measurement of EPO activity using the peroxidase substrate, O-

phenylenediamine (OPD). Unlike other peroxidase substrates, OPD is soluble in aqueous physiological solutions, which do not

lyse cell membranes, thereby allowing us to add OPD directly to eosinophils and exclusively measure extracellular EPO. This

novel approach would remove the concerns of incorrect EPO measurements due to its adhesive nature. In addition, we

developed this method to quantify EPO release in terms of EPO concentration. Finally, using this technique, we have been able

to demonstrate secretory IgA (s-IgA)-induced release of EPO. By using OPD, we have developed a more sensitive and specific

method to analyze the release of extracellular EPO.

D 2004 Elsevier B.V. All rights reserved.

Keywords: Eosinophil; EPO; Methodology; IgA

1. Introduction

Eosinophils are important and prominent cells in

allergic inflammatory conditions affecting a range of

0022-1759/$ - see front matter D 2004 Elsevier B.V. All rights reserved.

doi:10.1016/j.jim.2004.05.003

* Corresponding author. Division of Pediatric Pulmonary

Medicine, School of Medicine and Dentistry, 567B HMRC,

University of Alberta, Edmonton, Alberta, Canada T6G 2S2. Tel.:

+1-780-492-6863; fax: +1-780-492-5329.

E-mail address: dadamko@ualberta.ca (D.J. Adamko).

human body systems. The production and release of a

variety of intracellular products from eosinophils have

been associated with human health and disease (Bous-

quet et al., 1990; Gleich, 2000; Adamko et al., 2003).

One such product, eosinophil peroxidase (EPO), is

stored in the matrix of the secondary crystalloid

granules and is released during degranulation

responses. Unlike other known peroxidases (Cameron

et al., 2000), EPO is unique for its highly charged

cationicity, with an intracellular isoelectric point of

D.J. Adamko et al. / Journal of Immunological Methods 291 (2004) 101–108102

10.8 (Gleich, 2000). As such, it adheres to most

extracellular surfaces, rendering it more difficult to

measure compared with other released intracellular

proteins.

One of the most common techniques for measuring

EPO release, in vitro, involves the use of a peroxidase

substrate, 3,3V,5,5V– tetramethylbenzidine (TMB),

which requires dissolving in the organic solvents

DMSO, ethanol and acetone. Because organic sol-

vents solubilize cell membranes, this substrate solu-

tion fails to distinguish between intracellular and

released extracellular peroxidase activity if the solu-

tion is added directly to stimulated cells. To solve this,

eosinophils must be pelleted out of the cell suspension

to first create a cell-free supernatant. The peroxidase

activity in the supernatants of stimulated cells is

usually measured against the EPO activity from the

fully lysed eosinophil pellets. This supernatant value

is intended to reflect the released EPO as a percentage

of total cellular EPO.

However, using TMB as a substrate generated

inconsistencies in the measurement of released EPO

in our hands. Based on the understanding of the high

cationic charge and sticky nature of EPO, we were

concerned that released EPO is not reproducibly

measurable in the supernatant. Instead, we suspected

that much of the released EPO was left stuck to the

tube walls. This idea was been previously suggested

by Menegazzi et al. (1992). Furthermore, stimulated

eosinophils were difficult to visualize after pelleting

as they produced a smear on the sides of the tubes.

Thus, there is a possibility that supernatants may

become contaminated with intact eosinophils, result-

ing in falsely high values of measured EPO activity.

We chose to investigate the measurement of EPO

activity using the peroxidase substrate, O-phenylene-

diamine (OPD). OPD is soluble in aqueous physio-

logical solutions, which do not lyse cell membranes,

and thereby allow us to add OPD directly to eosino-

phils and exclusively measure extracellular EPO. This

novel approach would remove the concerns of false

EPO measurements due to its adhesive nature.

In this paper, we have used a simple, yet accurate

and sensitive, measurement to assess the amount of

released EPO. In addition, this method offers the

unique ability to quantify the amount of EPO released

in terms of protein concentration. Finally, as part of

validating the system and optimizing EPO release, we

present data comparing both established and novel

approaches of eosinophil stimulation.

2. Materials and methods

2.1. Eosinophil isolation

Human eosinophils were isolated from peripheral

blood as previously described (Mahmudi-Azer et al.,

1998; Lacy et al., 1998). Briefly, peripheral blood

was obtained from atopic patients who displayed

blood eosinophilia and were not receiving oral cor-

ticosteroids. After red blood cell sedimentation in 6%

dextran (T500), remaining cells were subjected to

density centrifugation on Ficoll (Pharmacia Biotech;

Uppsala, Sweden). Eosinophils were then purified

from the granulocyte pellet by immunomagnetic

selection using the MACS system (Miltenyi Biotec;

Bergisch Gladbach, Germany). Highly purified

CD16� eosinophils (>99%) were obtained by nega-

tive selection, and were depleted of neutrophils using

anti-CD16-conjugated immunomagnetic beads as de-

scribed previously (Lacy et al., 1998). Contamination

by mononuclear cells and lymphocytes was pre-

vented by coincubation with anti-CD14- and anti-

CD3-coated micromagnetic beads (Miltenyi Biotec).

Before use, cells were washed and resuspended in

RMPI-1640 with 0.1% human serum albumin (HSA)

and 25 mM HEPES at a concentration of 1�106

eosinophils/ml.

2.2. Preparation of tubes

To block nonspecific activation of eosinophils,

(Kato et al., 1998) all tubes used in this work for

eosinophil stimulation were precoated with HSA.

Briefly, sterile eppendorf tubes were incubated with

1 ml 2.5% HSA for 2 h at 37 jC and then washed with

1 ml PBS once before aliquots of eosinophils were

added.

2.3. EPO measurement using TMB

3,3V,5,5V-Tetramethylbenzidine (TMB) substrate

solution, which contains the solvents methanol

(20%), acetone (5%) and DMSO (10%), was pur-

chased from Sigma (Oakville, ON). Tubes containing

D.J. Adamko et al. / Journal of Immunological Methods 291 (2004) 101–108 103

eosinophils lysed with 0.2% Triton X-100 were ag-

gressively washed and refilled with color-free RPMI

(100 Al). A total of 300 Al of TMB solution was added

for 20 min at room temperature. The reaction was

stopped with 100-Al 1 M H2SO4. A volume of 200

Al from each tube was transferred to a 96-well plate in

duplicate for reading via a colorimetric assay (Soft-

max; 492-nm wavelength).

2.4. Preparation of OPD

Substrate solution was prepared by adding 800 Al of5 mM O-phenylenediamine HCl (Sigma; Oakville,

ON) to 4-ml 1 M Tris buffer (pH 8.0) and 1.25 Al 30%hydrogen peroxide (Sigma). Distilled water was added

to a total volume of 10 ml. The OPD solution was

protected from light and was made fresh shortly

before each experiment.

2.5. Secretory IgA-coated bead preparation

The coating of secretory IgA (s-IgA) beads were

used as previously described (Abu-Ghazaleh et al.,

1989). Briefly, CNBr–Sepharose 4 MB (Pharmacia)

beads were preswollen in 1 mM HCl. CNBr–Sephar-

ose beads were used to allow covalent binding of s-

IgA. S-IgA (Sigma) dissolved in a borate-containing

buffer was incubated with preswollen, washed beads

at room temperature for 2 h, while mixing end-over-

end. The bead suspension was blocked with 0.1 M

lysine monohydrochloride for 2 h at room temperature

while mixing end-over-end. All beads were washed

three times in borate-coupling buffer and 0.1 M

acetate (pH 4). Before use with eosinophils, the beads

were further washed with RPMI.

2.6. OPD assay for measurement of EPO activity

The assay was standardized using human EPO (Dr.

G.J. Gleich, University of Utah, Salt Lake City, UT).

Stock EPO (10� 6 M) was diluted in serial 1/2

dilutions to a total tube volume of 500 Al. As adaptedfrom Strath et al. (1985), OPD (500 Al) was added for

2 min at 21 jC. The reaction was stopped with the

addition of 1-ml 4 M H2SO4. The EPO–OPD reaction

converts from a clear to a yellow-brown solution; the

optical density of which varies depending on the

initial amount of EPO available. A volume of 200

Al from each tube was transferred to a 96-well plate in

duplicate for reading via a colorimetric assay (Soft-

max; 492-nm wavelength).

2.7. Stimulation of EPO release

In order to develop a reliable method of EPO

release, we tested a combination of cytokines together,

each individually known for an association with

eosinophil activation. Suspensions of eosinophils

(500 Al of 1�106 eosinophils/ml in color-free RPMI

with 0.1% HSA and 25 mM HEPES) were added to

HSA-coated tubes. The eosinophils were incubated

with a cocktail of IL-3 (20 ng/ml), IL-5 (10 ng/ml)

and GM–colony-stimulating factor (CSF ; 20 ng/ml)

at 37 jC for 2 h. This was compared with a less

physiologic, but well-established stimulus, namely the

calcium ionophore, A23187 (5 AM, 30 min at 37 jC;Matsunaga et al., 1994; Kirino et al., 2000). Unstimu-

lated control cells were sham-treated with RPMI

vehicle followed by incubation protocols similar to

the stimulated cells.

2.8. Comparison of EPO measured in tubes,

supernatants or whole cell suspensions

After completion of incubation, the cells (500 Al)were gently pelleted (300 g for 5 min at 4 jC).Supernatants were carefully removed to avoid disturb-

ing pellets. Because the pellets of stimulated cells

were difficult to visualize, we left a small volume at

the bottom of all the tubes (40 Al) to avoid contam-

ination of supernatants with whole eosinophils. OPD

solution (500 Al) was added to both the original tubes

with pelleted cells and the supernatants. Furthermore,

because the OPD solution is not membrane-permeant,

we anticipated that we could measure the extracellular

EPO activity by adding OPD directly to undisturbed

cell suspensions. Thus, in some groups, after stimu-

lation was completed, we simply added OPD 1:1

dilution (500:500 Al) directly to the cell suspensions

and then measured EPO activity.

2.9. Measurement of EPO activity in lysed eosinophils

An eosinophil suspension of 500 Al (500,000 cells/

tube in triplicate) was lysed by adding 0.2% Triton X-

100, followed by vigorous vortexing. Lysed cells were

D.J. Adamko et al. / Journal of Immunological Methods 291 (2004) 101–108104

aliquoted in serial 1/2 dilutions. OPD was added 1:1,

and as described above, EPO activity was measured.

2.10. Statistics

All data are expressed as meanF S.E.M. Compar-

ison of EPO activity between the groups was analyzed

using analysis of variance (ANOVA; Statview 5.0,

SAS Institute; Cary, NC). A p-value < 0.05 was

considered significant. Calculation of values for linear

correlation between optical density and EPO molarity

was performed by Microplate Manager software (Bio-

rad; Hercules, CA).

Fig. 2. OPD does not cause eosinophil lysis nor does it enter the

cell, thus the EPO activity measured by OPD is extracellular in

nature. OPD mixed with nonlysed, membrane-intact eosinophils

(white bar; n= 29) showed minimal activity compared to the same

concentration of lysed eosinophils (black bar; n= 11). S.E.M. shown

by vertical bars.

3. Results

3.1. EPO adheres to available surfaces

Because of its cationic nature, released extracellu-

lar EPO may not remain free in the supernatant but

may become adherent to available surfaces. This was

confirmed by measuring EPO remaining on the tube

walls after unstimulated eosinophils had been pel-

Fig. 1. Extracellular EPO becomes adherent to surfaces making

measurement in the supernatant difficult. Even after rinsing with

PBS, tubes that had contained eosinophils showed EPO activity as

measured by TMB (n= 3).

leted, lysed and then removed by washing. Even after

aggressive rinsing of the tubes, EPO activity as

measured by the TMB assay was found lining the

tube wall (Fig. 1).

3.2. The majority of EPO activity measured by OPD

is extracellular in nature

To specifically measure released EPO activity

adherent to surfaces, the peroxidase substrate must

be able to discern intracellular vs. extracellular EPO

activity. The use of TMB substrate solution is not

suitable for this because it contains an organic solvent

which lyses cell membranes. The solvent for OPD

does not cause eosinophil lysis nor is it membrane-

permeant. We confirmed this by showing that OPD

mixed with nonlysed, membrane-intact eosinophils

(Fig. 2, white bar, n = 29) has minimal activity com-

pared to lysed eosinophils (black bar, n = 11).

3.3. Quantification of EPO activity

In order to accurately quantify the amount of EPO

in terms of protein concentration, we converted the

Fig. 3. Using purified human EPO, the OPD measurement of

activity was converted from optical density to molarity. The OPD

reading was linear in a specified concentration range. (.) Mean of

at least five samples, with S.E.M. shown by vertical bars.

D.J. Adamko et al. / Journal of Immunolo

optical density reading of the OPD-EPO reaction to a

known amount of EPO protein. A known molarity of

purified human EPO was titrated in serial one half

dilutions to generate the optical density curve of the

OPD-EPO reaction. Within a specified range, the

OPD optical density reading of EPO protein was

linear over two orders of magnitude, with a correlation

coefficient of 1.0 (Fig. 3).

Fig. 4. To quantitate the total available EPO activity, we used lysed

eosinophils. The OPD measurement of total EPO activity in lysed

eosinophils was linear below an eosinophil concentration of

approximately 5� 104 eosinophils/ml. (.) Mean of at least eight

samples, with S.E.M. shown by vertical bars.

3.4. Total EPO available in lysed eosinophils

Because lysed eosinophils represent the total

amount of EPO available in a cell suspension, we used

serial dilutions of lysed eosinophils and measured the

OPD–EPO reaction by optical density. Similar to the

EPO activity of pure protein, there is a range of linearity

but only below an eosinophil concentration of approx-

imately 5� 105 eosinophils/ml (Fig. 4). Thus, the

majority of the EPO activity of lysed eosinophils

remains outside the linear range, making accurate

measurements using percent of total more difficult.

3.5. Measurements using OPD show that the majority

of released EPO is adherent to the tube wall rather

than remaining free in suspension

Cell supernatants removed from pelleted eosino-

phils were compared to their respective unlysed

gical Methods 291 (2004) 101–108 105

Fig. 5. The majority of released EPO remains in the pellet or is stuck

to the tube walls rather than remaining free in suspension. Cell

supernatants from pelleted eosinophils were compared to their

respective nonlysed pellets. EPO release from unstimulated

eosinophils was compared to eosinophils stimulated with a cytokine

cocktail of IL-5, IL-3 and GM–CSF. There is very little EPO free in

the supernatants compared to that found in the tubes with cell pellets

and no difference in the amount of EPO release was seen between

unstimulated supernatant (white bars; n = 27), compared to

respective stimulated supernatants (black bars; n= 30, p= n.s.).

Within the tubes, there was an increase in EPO activity comparing

unstimulated (white bar; n= 26) to stimulated (black bar; n= 29,

p< 0.001). OPD was added to the entire cell suspension. The

presence of extracellular EPO as measured by OPD was increased in

the total suspension of stimulated eosinophils (black bar; n= 71) vs.

unstimulated cells (white bar; n= 63, p< 0.001).

D.J. Adamko et al. / Journal of Immunolo106

pellets. Using OPD, the EPO activity from both

control eosinophils and those eosinophils stimulated

with a cytokine cocktail of IL-5, IL-3 and GM–CSF

was compared. As expected, the majority of EPO

activity was found remaining in the original tube but

not in the supernatant (Fig. 5). While there was no

difference in the amount of EPO activity in the

supernatants of unstimulated (white bars; 0.65 nM)

and stimulated supernatants (black bars; 1.60 nM,

p = n.s.), there was an increase in EPO activity on

the tube walls comparing control (white bar; 9.90 nM)

to stimulated eosinophils (black bar; 22.3 nM,

p < 0.001).

3.6. OPD can be added directly to a cell suspension

to measure released EPO

Because OPD solution does not detect intracellular

EPO, the pelleting step became unnecessary. OPD

was added directly to the entire the cell suspension to

quantitate the amount of extracellular EPO activity

(Fig. 5). Similar to our data from pelleted cells, EPO

protein activity was increased in the total cell suspen-

sion containing stimulated eosinophils (black bar;

17.1 nM) as compared to unstimulated cells (white

bar; 7.5 nM, p < 0.001).

Fig. 6. The calcium ionophore, A32187 (n= 14) caused a large

release of EPO compared to both control eosinophils (n= 86,

p< 0.0001) and to the cocktail (n= 81, p< 0.0002). The ionophore

control, DMSO 0.25%, was similar to the unstimulated eosinophil

EPO activity (n= 8). Secretory IgA-coated beads (n= 9) produced a

significant release of EPO activity compared to control cells (n= 86,

p< 0.001), though still less than the EPO activity from the

ionophore.

3.7. Stimulation with established secretagogues

To determine the physiologic relevance of the

cytokine cocktail, the amount of EPO release after

cocktail stimulation was compared to the effects of the

calcium ionophore, A23187, and the more physiolog-

ical stimulus of secretory IgA (s-IgA) (Fig. 6). Not

surprisingly, there was greater release of EPO triggered

by A23187 (55.3 nM) compared with both control

eosinophils (7.5 nM, p < 0.0001) and the cocktail (17.1

nM, p < 0.0001). The ionophore control, DMSO

0.25%, was similar to the control EPO activity (7.4

nM). Secretory IgA-coated beads (36.3 nM) produced

a significant release of EPO activity compared to

unstimulated control cells ( p < 0.001), though still less

than the EPO activity from A23187 ( p = 0.003).

gical Methods 291 (2004) 101–108

4. Discussion

Degranulation assays with eosinophils are typical-

ly very difficult to do, with very low release values.

In this report, we have demonstrated a reliable

technique for the stimulation and measurement of

extracellular EPO released from peripheral blood

eosinophils. The use of supernatants to measure

EPO release can be inaccurate because, as we have

demonstrated, much of the EPO remains stuck in the

tube where the cells were stimulated. In order to

accurately measure the total released EPO, the per-

oxidase substrate must mix freely with the cells in the

original tube and detect the EPO activity specifically

located in the extracellular space.

Using TMB substrate prepared in an organic sol-

vent, we were not able to clearly discern if increased

EPO activity adhering to the tube wall was due to

released EPO or a contaminant of adherent but intact

eosinophils. For this reason, we investigated the use

of another substrate OPD, which is soluble in aqueous

solutions, and thus, does not disrupt the cell mem-

brane. The OPD solution has the advantage of react-

ing only with extracellular EPO. We have confirmed

this by comparing intact resting eosinophils and lysed

eosinophils. OPD mixed with suspensions of intact

eosinophils showed negligible EPO activity when

compared with lysed eosinophils.

To quantitate EPO activity more accurately, we

used purified human EPO to construct a standard

D.J. Adamko et al. / Journal of Immunological Methods 291 (2004) 101–108 107

curve for the OPD assay. Measurement of OPD

activity at these serial concentrations of EPO protein

permitted conversion from optical density to molar

quantities of EPO. The OPD reading was linear over

a two-log range starting at a low concentration of

10� 9 M.

Having established an accurate and sensitive

method to quantitate extracellular EPO activity, we

used this technique to confirm our hypothesis, that

the majority of released extracellular EPO becomes

adherent to tube walls rather than remaining free in

suspension. To stimulate EPO release, we used a

cytokine cocktail of IL-5, IL-3 and GM–CSF. Indi-

vidually, these cytokines play important roles in the

eosinophil lifecycle from the bone marrow to the

tissues (Adamko et al., 2002), as well as eosinophil

adhesion and activation (Wong et al., 2003). While

each cytokine has been demonstrated to have an

ability to induce release of an eosinophil mediator

(Horie et al., 1996), there is no reported induction

specifically of EPO release using these cytokines.

Using the cytokine cocktail stimulus, cell super-

natants from pelleted eosinophils were compared with

unlysed pellets. Most of the EPO activity adhered to

the original tube with pelleted cells and was absent

from the supernatant. Furthermore, using OPD, we

confirmed that relying on the data from supernatants

alone would not detect differences between the con-

trol and cocktail-stimulated groups. Because OPD

substrate solution does not cross the cell membrane,

we concluded that the increased activity measured in

the tubes of pelleted stimulated cells represents extra-

cellular EPO release.

Using OPD, pelleting eosinophils becomes unnec-

essary and could increase nonspecific EPO release

(data not shown). The slightly increased amount of

EPO release in the tube compared to cell suspension

likely reflects such damage. Therefore, without pellet-

ing the cells, we added OPD directly to the suspension

of cells. Again, we measured increased EPO activity

by OPD in the cell suspension containing cocktail-

stimulated eosinophils. This not only confirmed the

effect of cocktail stimulation but it also demonstrated

that the OPD solution could be used to distinguish

extracellular from intracellular EPO.

To determine the relevance of this cytokine cock-

tail and to validate the OPD solution for measurement,

the amount of EPO release after cocktail stimulation

was compared to some established stimuli. The calci-

um ionophore, A23187, induced a significantly great-

er release of EPO from A32187 stimulation compared

with the cytokine cocktail. To check our method

against a more potent, physiological stimulus, we

used secretory IgA (s-IgA). While it too induced a

greater release of EPO compared with the cytokine

cocktail, it remained less than that seen with the

calcium ionophore. This is the first report of s-IgA-

stimulated EPO release.

Total available EPO release is represented by lysed

eosinophils. The majority of the EPO measured in

lysed cells is not in the linear range which demon-

strates the difficulty in relying on optical density. We

have been using purified human EPO as our standard

of EPO activity and have been able to keep our EPO

release in the linear range. If EPO protein were not

available, the linear range of the eosinophil lysis curve

could be adapted to convert optical density to a unit of

measure (eosinophils/ml).

We have demonstrated that when EPO is released

from stimulated eosinophils, in vitro, a significant

amount cannot be measured in the cell supernatant

but remains adherent to the tube surface where

stimulation occurred. This idea had previously been

suggested (Menegazzi et al., 1992); however, the

researchers continued to use TMB. To more accu-

rately measure the amount of EPO that is released

extracellularly and present amongst the intact eosi-

nophils, we have used OPD, a peroxidase substrate

mixed in an aqueous solvent. Using the OPD solu-

tion, we have been able to mix substrate directly

with the cells of interest. Because OPD cannot react

with intracellular peroxidase, the EPO activity mea-

sured in the cell suspension reflects the specific

presence of extracellular EPO. This has led to

increased sensitivity for smaller differences as that

of the cytokine cocktail while still appreciating the

larger magnitude of difference as that with the

calcium ionophore.

Using OPD, we have developed a more sensitive

and specific method to analyze the release of extra-

cellular EPO. We believe that, because this simple

method requires less cell handling, it is less prone to

nonspecific cell secretion or damage. To our knowl-

edge, this is the first report to quantify EPO release in

terms of its concentration, as well as to show IgA-

induced release of EPO.

D.J. Adamko et al. / Journal of Immunological Methods 291 (2004) 101–108108

Acknowledgements

Funding for this study was provided by The

Hospital for Sick Children Foundation, AHFMR and

by the CIHR. DJA is an Alberta Heritage Clinical

Investigator, RM is an Alberta Heritage Medical

Scientist R. Moqbel and PL is a CLA/CIHR Young

Investigator.

References

Abu-Ghazaleh, R.I., Fujisawa, T., Mestecky, J., Kyle, R.A.,

Gleich, G.J., 1989. IgA-induced eosinonophil degranulation.

J. Immunol., 2393–2400.

Adamko, D., Lacy, P., Moqbel, R., 2002. Mechanisms of eosinophil

recruitment and activation. Curr. Allergy Asthma Rep., 107.

Adamko, D., Odemuyiwa, S.O., Moqbel, R., 2003. The eosinophil

as a therapeutic target in asthma: beginning of the end, or end of

the beginning? Curr. Opin. Pharmacol., 227.

Bousquet, J., Chanez, P., Lacoste, J.Y., Barneon, G., Ghavanian, N.,

Enander, I., Venge, P., Ahlstedt, S., Simony-Lafontaine, J., God-

ard, P., et al., 1990. Eosinophilic inflammation in asthma. N.

Engl. J. Med., 1033.

Cameron, L., Christodoulopoulos, P., Lavigne, F., Nakamura, Y.,

Eidelman, D., McEuen, A., Walls, A., Tavernier, J., Minshall,

R., Moqbel, R., Hamid, Q., 2000. Evidence for local eosinophil

differentiation within allergic nasal mucosa: inhibition with sol-

uble IL-5 receptor. J. Immunol., 1538.

Gleich, G.J., 2000. Mechanisms of eosinophil-associated inflamma-

tion. J. Allergy Clin. Immunol., 651.

Horie, S., Gleich, G.J., Kita, H., 1996. Cytokines directly induce

degranulation and superoxide production from human eosino-

phils. J. Allergy Clin. Immunol, 371.

Kato, M., Abraham, R.T., Okada, S., Kita, H., 1998. Ligation of the

h2 integrin triggers activation and degranulation of human eosi-

nophils. Am. J. Respir. Cell Mol. Biol, 675.

Kirino, Y., Mio, M., Kamei, C., 2000. Regulatory mechanism of

eosinophil peroxidase release from guinea pig eosinophils. Jpn.

J. Pharmacol., 293.

Lacy, P., Levi-Schaffer, F., Mahmudi-Azer, S., Bablitz, B., Hagen,

S.C., Velazquez, J., Kay, A.B., Moqbel, R., 1998. Intracellular

localization of interleukin-6 in eosinophils from atopic asth-

matics and effects of interferon gamma. Blood, 2508.

Mahmudi-Azer, S., Lacy, P., Bablitz, B., Moqbel, R., 1998. Inhibi-

tion of nonspecific binding of fluorescent-labelled antibodies to

human eosinophils. J. Immunol. Methods, 113.

Matsunaga, Y., Kido, H., Kawaji, K., Kamoshita, K., Katunuma,

N., Ogura, T., 1994. Inhibitors of chymotrypsin-like proteases

inhibit eosinophil peroxidase release from activated human

eosinophils. Arch. Biochem. Biophys., 67.

Menegazzi, R., Zabucchi, G., Knowles, A., Cramer, R., Patriarca,

P., 1992. A new, one-step assay on whole cell suspensions

for peroxidase secretion by human neutrophils and eosino-

phils. J. Leukoc. Biol., 619.

Strath, M., Warren, D.J., Sanderson, C.J., 1985. Detection of

eosinophils using an eosinophil peroxidase assay. Its use as

an assay for eosinophil differentiation factors. J. Immunol.

Methods, 209.

Wong, C.K., Ip, W.K., Lam, C.W., 2003. Inter-3, -5, and granulo-

cyte macrophage colony-stimulating factor-induced adhesion

molecule expression on eosinophils by p38 mitogen-activated

kinase and nuclear factor-[kappa] B. Am. J. Respir. Cell Mol.

Biol., 133.