Circadian changes in granulocyte-macrophage
colony-stimulating factor message in circulating
Ste ´phane Esnault, PhD*†; Yulin Fang, MS‡; Elizabeth A. B. Kelly, PhD‡; Julie B. Sedgwick, PhD‡;
Jason Fine, PhD§; James S. Malter, MD*†; and Nizar N. Jarjour, MD‡
Background: Granulocyte-macrophage colony-stimulating factor (GM-CSF), which stimulates eosinophil recruitment, acti-
vation, and survival, is expressed by activated eosinophils. Although eosinophil recruitment and enhanced survival have been
associated with nocturnal asthma (NA), the contribution of GM-CSF to NA is unknown.
Objective: To determine whether circulating eosinophil GM-CSF expression correlates with the symptoms of NA.
Methods: The GM-CSF messenger RNA (mRNA) expression at 4 PM and 4 AM was determined by reverse-transcriptase
polymerase chain reaction with Southern blot analysis in subjects with and without NA and in controls.
Results: A total of 142 asthma subjects were screened for nocturnal asthma with 1-week home peak expiratory flow rate
(PEFR) monitoring. Eleven subjects had NA (?20% diurnal variation in PEFR on 4 of 7 days), and 6 met the criteria for non-NA
(?10% diurnal variation in PEFR on 7 of 7 days); 8 controls were studied. In subjects with NA, GM-CSF mRNA expression
in circulating eosinophils increased 3-fold at 4 AM compared with 4 PM. Diurnal changes in GM-CSF mRNA expression were not
detected in the non-NA and control groups.
Conclusions: Day-night variation in eosinophil GM-CSF expression is associated with circadian variation in airway function
in asthma, a key manifestation of asthma severity.
Ann Allergy Asthma Immunol. 2007;98:75–82.
Nocturnal asthma (NA) is characterized by wheezing, chest
tightness, breathlessness, and cough typically between 3 and
5 AM in individuals who have a normal sleep pattern (ie,
active in the daytime and asleep at night). These commonly
seen symptoms can be associated with increased asthma
morbidity and decreased quality of life. Nocturnal asthma is
associated with a reduction in pulmonary function, enhanced
bronchial hyperresponsiveness, and increased airway inflam-
mation.1Circadian variations in inflammatory cells in NA
include increased eosinophil numbers at night in bronchoal-
veolar lavage fluid (BALF), alveolar tissue,2–4and sputum.5
In addition, the number and proportion of low-density (ie,
presumably activated) circulating eosinophils are increased at
night in individuals with NA.6
The accumulation of eosinophils is commonly noted in
asthma7and likely reflects the anti-apoptotic effects of gran-
ulocyte-macrophage colony-stimulating factor (GM-CSF) at
the site of inflammation.8We and others have demonstrated
that eosinophils are a significant source of GM-CSF9–11and,
thus, are capable of supporting their own survival by auto-
crine stimulation.12–15Eosinophils produce GM-CSF after
stimulation with a variety of factors, including fibronectin,
hyaluronic acid, and cytokines (such as tumor necrosis factor
? and interleukin 15 [IL-15]), and cross-linking of cell sur-
face molecules (such as CD40, CD9, and CD32).
In addition, we have demonstrated that tumor necrosis
factor ? plus fibronectin or hyaluronic acid alone induces
extracellular signal-regulated kinase (ERK) phosphorylation,
Pin1 activation, GM-CSF messenger RNA (mRNA) stabili-
zation, GM-CSF secretion, and GM-CSF–dependent eosino-
phil survival.12,15–17The translation of GM-CSF mRNA is
regulated by YB-1, a member of the highly conserved cold
shock domain family of proteins, which consists of 3 do-
mains: an N-terminal stretch of undetermined function, a cold
shock domain that mediates nucleic acid binding and contains
a consensus ribonucleoprotein complex 1 RNA-binding mo-
tif, and a C-terminal domain of basic/acidic repeats that has
been implicated in RNA binding and in protein-protein inter-
actions. YB-1 binds in close proximity to the mRNA cap
structure and displaces the initiation factors eukaryotic trans-
lation initiation factor 4E (eIF4E) and eIF4G, thereby causing
mRNA translational silencing.18,19Consistent with its inhibi-
tory role in translation, YB-1 is mainly associated with non-
polysomal inactive messenger ribonucleoprotein complexes,
whereas active messenger ribonucleoprotein complexes de-
* Waisman Center for Developmental Disabilities, Madison, Wisconsin.
† Department of Pathology and Laboratory Medicine, University of Wiscon-
sin School of Medicine, Madison, Wisconsin.
‡ Allergy, Pulmonary, and Critical Care Medicine Section of the Department
of Medicine, University of Wisconsin School of Medicine, Madison, Wis-
§ Department of Statistics and Biostatistics and Medical Information, Uni-
versity of Wisconsin School of Medicine, Madison, Wisconsin.
This work was supported by grants HL64812 and NIH M01 RR03186 from
the National Institutes of Health.
Received for publication May 15, 2006.
Accepted for publication in revised form July 3, 2006.
VOLUME 98, JANUARY, 2007 75
rived from polysomes contain significantly lower YB-1 lev-
In the present study, we compared GM-CSF mRNA accu-
mulation in circulating eosinophils at 4 PM (the predicted peak
of lung function) and 4 AM (the predicted trough of lung
function) to determine GM-CSF mRNA expression in indi-
viduals with NA compared with non-NA (NNA) individuals
Subjects with asthma and healthy volunteers were recruited
for this study (age range, 18–45 years). Smokers and indi-
viduals who had viral illness or asthma exacerbation within 4
weeks, had immunotherapy within 5 years, were pregnant, or
were recently post partum were excluded. All the participants
were corticosteroid naı ¨ve or had not been taking corticoste-
roids for a minimum of 6 weeks. None of the participants
were taking leukotriene inhibitors, theophylline, long-acting
?-agonists, tricyclic antidepressants, anti-inflammatories, or
any other medications that might interfere with study out-
comes. Individuals who worked the night shift or who had
expected disruption of their circadian cycle were also ex-
cluded. All the participants underwent a medical history and
physical examination, skin prick testing, baseline circulating
eosinophil counts, and baseline pulmonary function testing
that included spirometry, reversibility to ?-agonist, and air-
way hyperresponsiveness to inhaled methacholine. Healthy
volunteers had negative allergy skin test results, normal lung
function (forced expiratory volume in 1 second [FEV1]
?90% of predicted, reversibility to ?-agonist ?5%, and
methacholine provocation concentration that caused a 20%
decrease in FEV1[PC20] ?16 mg/mL). Asthma subjects had
positive skin test reactions to 1 or more common aeroaller-
gens, documented airway obstruction, and airway hyperre-
sponsiveness (methacholine PC20?8 mg/mL). To ensure that
adequate mRNA could be extracted for analysis, all the
participants were required to have a circulating eosinophil
count of 100/mm3or greater.
All the participants were characterized by peak flow vari-
ability recorded for 7 consecutive days. Subjects were dis-
pensed a peak flow meter (Mini-Wright; Clement Clarke
International Ltd, Harlow, England), diary card, and instruc-
tions to measure peak expiratory flow rate (PEFR) 4 times per
day (on awakening, at noon, at 4 PM, at bedtime, and during
the night if awakened) for 1 week. Subjects were instructed to
take the readings before exercise or inhaler use to capture
baseline readings. The 1-week PEFR recordings were ob-
tained during a period of stability in the participant’s envi-
ronment and sleep schedule.
Asthma subjects who showed a significant diurnal varia-
tion in peak flow (?20% change in PEFR on 4 of 7 days)
were enrolled in the NA study group. Those who had less
than 10% variation on 7 of 7 consecutive days were included
in the NNA group. Asthma subjects who fell between the 2
groups were excluded. The study was approved by the Uni-
versity of Wisconsin–Madison Center for Health Sciences
Human Subject Committee. Informed consent was obtained
from each participant.
All the participants were admitted to the General Clinical
Research Center at the University of Wisconsin Hospital for
2 separate visits; the order of the visits was random. The
visits were separated by an average of 2 weeks. Before visits,
participants withheld caffeine and ?-agonists for 6 hours. The
timing of the 2 study visits was designed to closely reflect the
expected peak (4 PM) and nadir (4 AM) in pulmonary func-
tion.21For 1 visit, participants were admitted to the research
unit at 3 PM. Baseline pulmonary function testing was per-
formed at 3:30 PM. Spirometry was then repeated 20 minutes
after albuterol treatment (180 ?g administered by metered-
dose inhaler). Phlebotomy (180 mL) was performed, after
which participants were monitored briefly and then dis-
charged. For the early morning study visit, the participant was
admitted to the General Clinical Research Center at 8 PM,
lights were turned out at 10 PM, and the participant was
awakened at 3:30 AM for spirometry, reversibility with ?-ag-
onist, and phlebotomy as previously mentioned. Participants
returned to sleep and were discharged in the morning.
Purification of Human Eosinophils
Eosinophils were purified as previously described.22Briefly,
blood was layered over Percoll (1.090 g/mL) (Pharmacia
Biotech, Piscataway, NJ) and centrifuged for 20 minutes at
700g. The granulocyte pellets were collected, red blood cells
were lysed by hypotonic shock, and neutrophils were de-
pleted by negative anti-CD16–conjugated immunomagnetic
bead selection (AutoMacs; Miltenyi Biotec, Auburn, CA).
The resulting eosinophils were more than 98% pure and more
than 97% viable.
Total RNA Extraction and Reverse Transcription and
Polymerase Chain Reaction
Purified peripheral blood eosinophils (1 ? 106cells) were
pelleted, lysed in a phenol-based solution (TRI Reagent;
Molecular Research Center Inc, Cincinnati, OH), and then
stored at ?80°C. Total RNA was extracted as described by
the manufacturer. Synthesis of first-strand complementary
DNA (cDNA) was performed using the manufacturer’s pro-
tocol (Qiagen Inc, Valencia, CA). Briefly, total RNA from
1 ? 106eosinophils was resuspended in a final volume of 20
?L containing 4 U of the reverse transcriptase (RT) (Omnis-
cript), 2 ?L of 10? RT buffer, 2 ?L of 5-mmol/L 2?-
deoxynucleoside 5?-triphosphate, 20 U of ribonuclease inhib-
itor, and 1 ?L of oligo dT primers (0.5 ?g/mL) (Invitrogen,
Carlsbad, CA) and incubated at 37°C for 60 minutes.
Polymerase chain reaction (PCR) and Southern blots were
performed as previously described.23Briefly, one tenth of the
RT was used for PCR along with 1 U of the DNA polymerase
(HotTaq; Qiagen Inc), 2.5 ?L of 10? PCR buffer, 5 ?L of Q
solution, 1 mL of 10-mmol/L 2?-deoxynucleoside 5?-triphos-
76ANNALS OF ALLERGY, ASTHMA & IMMUNOLOGY
phate, and 1.5 mL of 25-mmol/L magnesium chloride at a
final concentration of 0.2 mmol/L for each forward and
reverse primer in a final volume of 25 ?L. The forward and
reverse primers, respectively, for human GM-CSF were 5?-
CAGGGCCTGCGGGGCAGCCT-3? and 5?-GTCTCACTC-
CTGGACTGG-3?, whereas those for ?-actin were 5?-TCAC-
CAACTGGGACGACATG-3? and 5?-GTACAGGGATAG-
CACAGCCT-3?. Thirty cycles were performed for GM-CSF
or ?-actin. Because the signals obtained in an ethidium bro-
mide gel were generally weak for GM-CSF product, 10 ?L of
PCR product was electrophoresed in a 1% agarose gel before
Southern blotting using a radioactively labeled GM-CSF
The radioactivity or ethidium bromide signals were respec-
tively quantified using a PhosphorImager (model 445SI; Mo-
lecular Dynamics, Sunnyvale, CA) or an AlphaImager
(model 2200 V5.5; Alpha Innotech, San Leantor, CA). The
relative amount of radioactive GM-CSF PCR signal was
normalized to that of ?-actin and is presented as a ratio using
For ?-actin and GM-CSF, we determined that there was a
linear relationship between the signal strength and fold dilu-
tion of RT product that extended beyond 30 cycles. To ensure
that the primers used for RT of ?-actin mRNA did not
co-amplify pseudogenes from the genomic DNA, we per-
formed PCR directly on extracted RNA (without RT). We
also performed RT-PCR on an excess amount (500 ng) of
DNA extracted from the same eosinophil samples. ?-Actin
was not amplified from RNA that underwent PCR in the
absence of RT. The 200–base pair PCR product was not
detected in DNA samples.
As already described,24purified peripheral blood eosinophils
(1 ? 106cells) were snap frozen at ?80°C, and cell pellets
were dissolved in 50-mmol/L Tris (pH 7.4), 150-mmol/L
sodium chloride, 1-mmol/L EDTA, 10-mmol/L sodium flu-
oride, 1-mmol/L orthovanadate, 200 ?g/mL of 4-(2-amino-
ethyl) benzenes sulfonyl fluoride–hydrochloric acid (Pefab-
loc; Roche Biochemicals Inc, Indianapolis, IN), protease
inhibitor cocktail P8340 (Sigma, St Louis, MO), 1-mmol/L
dithiothreitol, 1 U of recombinant ribonuclease inhibitor per
microliter (Promega, Madison, WI), 1% Triton X-100, and
0.1% sodium dodecyl sulfate and 1% NP40 by passing them
through a 29-gauge needle. Cleared lysate was made by
centrifugation at 12,000g for 5 minutes, and 12 ?g of anti-
body was added for 2 hours, with rocking at 4°C. Protein G
agarose beads (Sigma-Aldrich Corp, St Louis, MO) were
added, and incubation was continued overnight. Pellets were
washed 5 times with lysis buffer (without detergent). The last
wash was split, with 40% dissolved in a phenol-based solu-
tion (TRI Reagent) and 60% dissolved in sodium dodecyl
sulfate–polyacrylamide gel electrophoresis loading buffer.
RNA was isolated according to the manufacturer’s recom-
mendation. The RT reactions were primed with oligo dT
primers (Gibco BRL, Rockville, MD), and PCR and Southern
blotting were performed as described previously herein. As
already described,24anti–TatYB-1 is a rabbit polyclonal
raised against recombinant TatYB-1 fusion protein (Univer-
sity of Wisconsin Medical School Animal Care Unit Poly-
clonal Antibody Service, Madison) and purified using an IgG
purification kit (ImmunoPure [G]; Pierce, Rockford, IL).
Data were analyzed using a software package (SigmaStat;
Jandel Scientific Software, San Rafael, CA). Pulmonary
function (FEV1) data were normally distributed as indicated
by the Kolmogorov-Smirnov test. These data are expressed as
mean ? SEM, and a paired t test was applied to evaluate the
significance of overnight changes in GM-CSF mRNA expres-
sion (fold increase). The Mann-Whitney rank sum test was
used to compare the difference in GM-CSF mRNA expres-
sion between the NA group and the NNA or control group.
The Spearman test was used to analyze correlations between
the change in GM-CSF mRNA expression and overnight
changes in pulmonary function. P ? .05 was considered
To obtain matched asthma subjects with and without noctur-
nal symptoms, 142 individuals were screened by means of an
initial history and 1 week of home peak flow monitoring.
Twenty-five individuals fulfilled the selection criteria and
were included in the study (Table 1): 11 met the criteria of
NA, 6 had NNA, and 8 were healthy controls. The primary
reason for screen failures was lack of significant diurnal
variation in peak flow. Also, it was difficult to find subjects
who had airway reversibility and a PC20less than 8 mg/mL
and no overnight decline in PEFR on 7 of 7 nights. In the
control group, not having enough circulating eosinophils
(?100/mm3) was the main reason for screen failure.
By definition, the NA group had significantly greater over-
night variation in PEFR during the 7-day screening period
(Table 1). The NA and NNA groups had a similar degree of
airway hyperresponsiveness (methacholine PC20) and revers-
ibility to inhaled ?-agonist during the screening visits (Table
1). During the study, the NA group had lower FEV1at 4 AM
(78% ? 3% vs 92% ? 3% of predicted; P ? .004) and to a
lesser extent at 4 PM (83% ? 3% vs 95% ? 3% of predicted;
P ? .01) compared with the NNA group. Although the NA
group had marked variation during the 1-week PEFR moni-
toring at home, there was only modest variation in FEV1
(5.9% ? 3.4%) at 4 AM compared with 4 PM during their
hospital stay. Finally, the NA and NNA groups tended to
have higher circulating eosinophils at 4 AM vs 4 PM (NA
group: 255 ? 32 vs 203 ? 35/mm2; P ? .10; NNA group:
261 ? 34 vs 221 ? 56/mm2; P ? .48); however, this change
was not statistically significant.
VOLUME 98, JANUARY, 200777
GM-CSF mRNA Expression in Circulating Eosinophils at
4 AM and 4 PM
Using Southern blot analysis with a radioactively labeled
GM-CSF cDNA probe after RT and PCR, steady-state levels
of GM-CSF mRNA from circulating eosinophils were detect-
able in all the subjects (Fig 1). There was an increase in
GM-CSF mRNA from 4 PM to 4 AM in 9 of 11 NA subjects.
Overall, the NA group had a 3.0 ? 0.6-fold increase in
GM-CSF mRNA at 4 AM compared with 4 PM (P ? .005).
This change was not seen in the NNA (ratio of 1.0 ? 0.2) and
control (ratio of 0.9 ? 0.1) groups. The overnight increase in
GM-CSF mRNA expression was significantly higher in the
NA group compared with the NNA (P ? .01) and control
(P ? .01) groups. The 4 PM and 4 AM samples from the same
donor were processed concomitantly for RT-PCR and South-
ern blot, allowing us to compare the 4 PM and 4 AM samples
from a given participant. The GM-CSF mRNA levels from
different donors could not be compared with each other
because they were not determined concurrently. Whereas the
change in GM-CSF mRNA levels was associated with the
presence of NA, the correlation between the magnitude of the
GM-CSF mRNA increase and the fall in pulmonary function
was not statistically significant.
Based on previous data with in vivo– or in vitro–activated
eosinophils,15–17we investigated whether the mitogen-acti-
vated protein kinase ERK and the peptidyl propyl isomerase
Pin1 played a role in the increase in GM-CSF mRNA ex-
pression at night in the NA group. Therefore, eosinophils
were cultured in RPMI plus 10% fetal bovine serum with or
without the ERK inhibitor (PD98059) or the Pin1 inhibitor
(juglone) for 1 hour. As observed in Figure 2, ERK and Pin1
inhibition reduced the abundance of GM-CSF mRNA at 4 AM
compared with that seen at 4 PM. This suggests that the
intracellular signaling in eosinophils from asthmatic individ-
Table 1. Participant Characteristics
SexAge, y FEV1, % of predicted
ß-Agonist reversibility, %
Nocturnal Asthma Group
Subtotal*24 ? 281 ? 2†1.2 ? 0.3†16 ? 2†30 ? 3†‡
Nonnocturnal Asthma Group
Subtotal* 21 ? 189 ? 5†1.7 ? 1.0†
20 ? 5†5 ? 1
Subtotal* 27 ? 3 101 ? 32 ? 14 ? 1
Abbreviations: FEV1, forced expiratory volume in 1 second; PC20, provocation concentration that caused a 20% decrease in FEV1; PEFR, peak
expiratory flow rate.
* Subtotals are given as mean ? SE.
† P ? .05 vs the control group.
‡ P ? .05 vs the nonnocturnal asthma group.
78 ANNALS OF ALLERGY, ASTHMA & IMMUNOLOGY
uals is similar to that observed after in vivo segmental aller-
gen challenge or after in vitro activation with hyaluronic acid,
fibronectin, and cytokines.15,16
YB-1 Binding Activity to GM-CSF mRNA in Circulating
Eosinophils at 4 AM and 4 PM
Whereas the change in GM-CSF mRNA was associated with
the presence of NA, the correlation between the magnitude of
GM-CSF mRNA and changes in pulmonary function were
not statistically significant. This suggested that other molec-
ular mechanisms might control GM-CSF production. YB-1 is
a multifunctional nucleic acid binding protein that has been
implicated in cytokine mRNA translation.20,25It is known to
bind in close proximity to the mRNA cap structure and to
displace eIF4E and eIF4G, thereby silencing mRNA transla-
tion.18,19YB-1 binding to GM-CSF mRNA at 4 PM and 4 AM
was evaluated in a pilot experiment performed on 3 NA
subjects. Thus, YB-1 was immunoprecipitated from eosino-
phil cell lysates, followed by GM-CSF RT-PCR. We ob-
served that YB-1 binding to GM-CSF mRNA was nearly
absent at 4 AM compared with 4 PM despite abundant GM-CSF
mRNA accumulation (Figs 1 and 3). Therefore, in addition to
GM-CSF mRNA accumulation, the observed circadian
changes in YB-1 interaction with GM-CSF mRNA in eosin-
ophils may modulate GM-CSF production and eosinophil
The cellular and molecular mechanisms underlying NA re-
main unknown. In this study, we showed that subjects with
NA had increased GM-CSF mRNA in circulating eosinophils
at 4 AM compared with 4 PM, a change that was not seen in the
NNA and control groups. Whereas the significance of the
3-fold increase in eosinophil GM-CSF mRNA in these sub-
jects cannot be established, it was previously demonstrated
that a 2-fold increase in GM-CSF mRNA stabilization leads
to a 10-fold increase in the generation of GM-CSF by T
Figure 1. Diurnal variation in granulocyte-macrophage colony-stimulating factor (GM-CSF) messenger RNA (mRNA) in circulating eosinophils from
nocturnal asthmatic (NA) subjects. A, Steady-state levels of GM-CSF (Southern blot) and ?-actin (ethidium bromide gel) mRNA at 4 AM and 4 PM in peripheral
blood eosinophils from representative donors. B, The relative amount of GM-CSF signal was normalized to that of ?-actin and presented as a ratio. Graphs depict
data points from individual participants expressed as fold increases at 4 AM compared with 4 PM. The values of the latter are arbitrarily fixed at 1. The statistics
were performed as described in the “Methods” section. NS indicates not significant; NNA, non-NA.
VOLUME 98, JANUARY, 200779
lymphocytes.26In the absence of exogenous ex vivo stimuli,
release of GM-CSF protein by eosinophils is at or below the
level of detection by enzyme-linked immunosorbent assay.
However, we showed that very small amounts (multiple ad-
ministrations of 0.1 pg/mL) of GM-CSF prolong ex vivo
survival of eosinophils.13Based on these observations, we
hypothesized that even minute quantities of autocrine GM-
CSF may be adequate to counteract the effect of various
factors that induce eosinophil apoptosis, such as corticoste-
roids, IL-4, and sialic acid–binding immunoglobulin-like lec-
tin.27–29Therefore, we propose that enhanced GM-CSF
mRNA expression in eosinophils at night is a contributing
factor to nocturnal worsening of asthma.
As a surrogate marker for cytokine secretion, we analyzed
YB-1/GM-CSF interactions at various times. Although the
YB-1 findings are preliminary, the suggestion of YB-1 re-
lease from GM-CSF mRNA at 4 AM points toward rebound
increases in the GM-CSF translation rate and downstream
eosinophil activation. This, in addition to the observed circa-
dian variation in GM-CSF mRNA levels, could amplify GM-
CSF protein production and in part explain the lack of a
statistically significant correlation between GM-CSF mRNA
changes (4 AM compared with 4 PM) and changes in FEV1
(R ? ?0.599; P ? .10; not shown).
The mechanisms of the circadian variations in GM-CSF
mRNA expression in eosinophils are unknown. A central
circadian pacemaker is known to reside in the hypothalamic
suprachiasmatic nucleus, but it is not known whether the
eosinophil has a peripheral circadian oscillator that affects the
expression of other genes (such as GM-CSF) and eosinophil
functions. Circadian changes in circulating hormones are
probably a more likely cause of eosinophil activation. The
nadir of the plasma epinephrine level (4 AM)21correlates with,
whereas the decrease in cortisol level30precedes, the GM-
CSF mRNA increase observed herein. Fibronectin is an ex-
tracellular matrix protein that is increased in the airway of
asthmatic individuals31,32and is further increased in BALF
after allergen challenge.33Our group12recently showed that
fibronectin induces GM-CSF mRNA expression in eosino-
phils. However, we did not observe circadian variation in
circulating fibronectin levels (data not shown). The IL-1?
levels increase at night in NA,34and in vitro, IL-1? induces
GM-CSF production in airway smooth muscle cells.35To our
knowledge, the effect of IL-1? on GM-CSF mRNA expres-
sion in eosinophils remains unknown.
We recognize that there are inherent limitations to these
types of studies. First, circulating, not airway, eosinophils
were evaluated. We have extensive experience with broncho-
scopic studies in asthma including subjects with NA.33,34,36,37
However, owing to the low number of eosinophils in the
BALF of asthmatic individuals (even during nocturnal exac-
erbations),2–4,36it is not technically feasible to separate suf-
ficient numbers of airway eosinophils with adequate purity
for molecular biologic studies. Nonetheless, although the
present investigation focused on the blood compartment, we
demonstrated significant day and night differences within the
NA group and nighttime differences between the NA and
NNA groups. Second, as noted in previous studies,37although
the NA group had a history of recurrent NA symptoms and
showed circadian variation in the pulmonary peak flow rate
measured for more than 1 week at home (Table 1), they had
only modest changes in FEV1at 4 AM compared with 4 PM
during the study. A variety of factors could contribute to the
lack of demonstrable circadian variation in FEV1. Because
relatively large amounts of blood were necessary to obtain
adequate numbers of purified eosinophils, the 4 AM and 4 PM
blood collections were performed at least 5 days apart. Thus,
the FEV1measurements were not performed in the same
24-hour period. In addition, the fact that participants were
hospitalized and, therefore, removed from their external en-
Figure 2. Extracellular signal-regulated kinase and Pin1 regulate granu-
locyte-macrophage colony-stimulating factor (GM-CSF) messenger RNA
accumulation at night. Peripheral blood eosinophils were cultured for 1 hour
with or without PD05895 (PD) or juglone (Jug). Total RNA was extracted
and underwent GM-CSF reverse-transcriptase polymerase chain reaction
plus Southern blot analysis. The figure is representative of 2 subjects with
Figure 3. YB-1 binding activity with granulocyte-macrophage colony-
stimulating factor (GM-CSF) messenger RNA (mRNA) decreases at night in
circulating eosinophils from nocturnal asthmatic subjects. Eosinophils were
snap frozen at ?80°C. Ten percent of the lysate was used for total RNA
isolation, and the remaining 90% was used for YB-1 immunoprecipitation.
Reverse-transcriptase polymerase chain reaction and Southern blotting with
radiolabeled GM-CSF complementary DNA probes were performed for
GM-CSF mRNA co-immunoprecipitating with YB-1 or an irrelevant IgG
antibody. Three subjects with nocturnal asthma are presented.
80 ANNALS OF ALLERGY, ASTHMA & IMMUNOLOGY
vironment for 8 hours before the 4 AM blood collection may
have affected the measurement at 4 AM. Finally, the 4 AM and
4 PM spirometry measurements may not coincide with the
actual peak and trough for pulmonary function in each indi-
vidual. These times were chosen based on published studies38
and our own observations.39
In summary, we show that in subjects with NA, GM-CSF
mRNA levels in circulating eosinophils are significantly in-
creased at night. This change was not seen in asthma subjects
without nocturnal exacerbations or in controls. The factors
responsible for this circadian change remain unknown but
involve ERK/Pin1 intracellular signaling pathways. In addi-
tion, in the 3 NA subjects tested, YB1/GM-CSF mRNA
interaction decreased at night, which may account for auto-
crine GM-CSF production and eosinophil activation. Future
studies will need to address the potential mechanisms for this
change and directly evaluate its contribution to airway eosin-
We thank the nursing staff; Ann Dodge, BS, BSN; Mary Jo
Jackson, BS, BSN; and Andrea Tweedie, BS, BSN, for their
assistance with participant recruitment and screening.
1. Sutherland ER. Nocturnal asthma: underlying mechanisms and
treatment. Curr Allergy Asthma Rep. 2005;5:161–167.
2. Mackay TW, Wallace WA, Howie SE, et al. Role of inflam-
mation in nocturnal asthma. Thorax. 1994;49:257–262.
3. Martin RJ, Cicutto LC, Smith HR, Ballard RD, Szefler SJ.
Airway inflammation in nocturnal asthma. Am Rev Respir Dis.
4. Kraft M, Djukanovic R, Wilson S, Holgate ST, Martin RJ.
Alveolar tissue inflammation in asthma. Am J Respir Crit Care
5. Panzer SE, Dodge AM, Kelly EA, Jarjour NN. Circadian vari-
ation of sputum inflammatory cells in mild asthma. J Allergy
Clin Immunol. 2003;111:308–312.
6. Calhoun WJ, Bates ME, Schrader L, Sedgwick JB, Busse WW.
Characteristics of peripheral blood eosinophils in patients with
nocturnal asthma. Am Rev Respir Dis. 1992;145:577–581.
7. Bousquet J, Chanez P, Lacoste JY, et al. Eosinophilic inflam-
mation in asthma. N Engl J Med. 1990;323:1033–1039.
8. Simon HU, Yousefi S, Schranz C, Schapowal A, Bachert C,
Blaser K. Direct demonstration of delayed eosinophil apoptosis
as a mechanism causing tissue eosinophilia. J Immunol. 1997;
9. Broide DH, Firestein GS. Endobronchial allergen challenge in
asthma: demonstration of cellular source of granulocyte macro-
phage colony-stimulating factor by in situ hybridization. J Clin
10. Kita H, Ohnishi T, Okubo Y, Weiler D, Abrams JS, Gleich GJ.
Granulocyte/macrophage colony-stimulating factor and inter-
leukin 3 release from human peripheral blood eosinophils and
neutrophils. J Exp Med. 1991;174:745–748.
11. Moqbel R, Hamid Q, Ying S, et al. Expression of mRNA and
immunoreactivity for the granulocyte/macrophage colony-
stimulating factor in activated human eosinophils. J Exp Med.
12. Esnault S, Malter JS. Granulocyte macrophage-colony-
stimulating factor mRNA is stabilized in airway eosinophils and
peripheral blood eosinophils activated by TNF-? plus fibronec-
tin. J Immunol. 2001;166:4658–4663.
13. Esnault S, Malter JS. Minute quantities of granulocyte-
macrophage colony-stimulating factor prolong eosinophil sur-
vival. J Interferon Cytokine Res. 2001;21:117–124.
14. Hoontrakoon R, Chu HW, Gardai SJ, et al. Interleukin-15
inhibits spontaneous apoptosis in human eosinophils via auto-
crine production of granulocyte macrophage-colony stimulating
factor and nuclear factor-?B activation. Am J Respir Cell Mol
15. Shen ZJ, Esnault S, Malter JS. The peptidyl-prolyl isomerase
Pin1 regulates the stability of granulocyte-macrophage colony-
stimulating factor mRNA in activated eosinophils. Nat Immu-
16. Esnault S, Malter JS. Hyaluronic acid or TNF-? plus fibronectin
triggers granulocyte macrophage-colony-stimulating factor
mRNA stabilization in eosinophils yet engages differential in-
tracellular pathways and mRNA binding proteins. J Immunol.
17. Esnault S, Malter JS. Extracellular signal-regulated kinase me-
diates granulocyte-macrophage colony-stimulating factor mes-
senger RNA stabilization in tumor necrosis factor-? plus fi-
bronectin-activated peripheral blood eosinophils. Blood. 2002;
18. Nekrasov MP, Ivshina MP, Chernov KG, et al. The mRNA-
binding protein YB-1 (p50) prevents association of the eukary-
otic initiation factor eIF4G with mRNA and inhibits protein
synthesis at the initiation stage. J Biol Chem. 2003;278:
19. Evdokimova V, Ruzanov P, Imataka H, et al. The major
mRNA-associated protein YB-1 is a potent 5? cap-dependent
mRNA stabilizer. EMBO J. 2001;20:5491–5502.
20. Minich WB, Ovchinnikov LP. Role of cytoplasmic mRNP
proteins in translation. Biochimie. 1992;74:477–483.
21. Barnes P, Fitzgerald G, Brown M, Dolley C. Nocturnal asthma
and changes in circulating epinephrine, histamine, and cortisol.
N Engl J Med. 1980;303:263–267.
22. Yamamoto H, Sedgwick JB, Busse WW. Differential regulation
of eosinophil adhesion and transmigration by pulmonary micro-
vascular endothelial cells. J Immunol. 1998;161:971–977.
23. Esnault S, Jarzembowski JA, Malter JS. Stabilization of gran-
ulocyte-macrophage colony-stimulating factor RNA in a human
eosinophil-like cell line requires the AUUUA motifs. Proc
Assoc Am Physicians. 1998;110:575–584.
24. Capowski EE, Esnault S, Bhattacharya S, Malter JS. Y box-
binding factor promotes eosinophil survival by stabilizing gran-
ulocyte-macrophage colony-stimulating factor mRNA. J Immu-
25. Chen CY, Gherzi R, Andersen JS, et al. Nucleolin and YB-1 are
required for JNK-mediated interleukin-2 mRNA stabilization
during T-cell activation. Genes Dev. 2000;14:1236–1248.
26. Rajagopalan LE, Malter JS. Turnover and translation of in vitro
synthesized messenger RNAs in transfected, normal cells. J Biol
27. Wallen N, Kita H, Weiler D, Gleich GJ. Glucocorticoids inhibit
cytokine-mediated eosinophil survival. J Immunol. 1991;147:
28. Wedi B, Raap U, Lewrick H, Kapp A. IL-4–induced apoptosis
in peripheral blood eosinophils. J Allergy Clin Immunol. 1998;
VOLUME 98, JANUARY, 2007 81
102:1013–1020. Download full-text
29. Nutku E, Aizawa H, Hudson SA, Bochner BS. Ligation of
Siglec-8: a selective mechanism for induction of human eosin-
ophil apoptosis. Blood. 2003;101:5014–5020.
30. Szefler SJ, Ando R, Cicutto LC, Surs W, Hill MR, Martin RJ.
Plasma histamine, epinephrine, cortisol, and leukocyte ?-adren-
ergic receptors in nocturnal asthma. Clin Pharmacol Ther.
31. Roberts CR. Is asthma a fibrotic disease? Chest. 1995;107:
32. Roche WR, Beasley R, Williams JH, Holgate ST. Subepithelial
fibrosis in the bronchi of asthmatics. Lancet. 1989;1:520–523.
33. Meerschaert JA, Kelly EA, Mosher DF, Busse WW, Jarjour
NN. Segmental antigen challenge increases fibronectin in bron-
choalveolar lavage fluid. Am J Respir Crit Care Med. 1999;
34. Jarjour NN, Busse WW. Cytokines in bronchoalveolar lavage
fluid of patients with nocturnal asthma. Am J Respir Crit Care
35. Hallsworth MP, Soh CP, Twort CH, Lee TH, Hirst SJ. Cultured
human airway smooth muscle cells stimulated by interleukin-1?
enhance eosinophil survival. Am J Respir Cell Mol Biol. 1998;
36. Jarjour NN, Busse WW, Calhoun WJ. Enhanced production of
oxygen radicals in nocturnal asthma. Am Rev Respir Dis. 1992;
37. Kelly EAB, Houtman JJ, Jarjour NN. Inflammatory changes
associated with circadian variation in pulmonary function in
subjects with mild to moderate asthma. Clin Exp Allergy. 2004;
38. Martin RJ, Cicutto LC, Ackerson LM. Stability of the circadian
alteration in lung function in asthma. J Allergy Clin Immunol.
39. Bates ME, Clayton M, Calhoun WJ, et al. Relationship of
plasma epinephrine and circulating eosinophils to nocturnal
asthma. Am J Respir Crit Care Med. 1994;149:667–672.
Requests for reprints should be addressed to:
Nizar N. Jarjour, MD
Section of Allergy Pulmonary and Critical Care Medicine
University of Wisconsin School of Medicine
600 Highland Ave
Madison, WI 53792-9988
82 ANNALS OF ALLERGY, ASTHMA & IMMUNOLOGY