Vous êtes sur la page 1sur 4

Clin Exp Med (2004) 4:99–102

DOI 10.1007/s10238-004-0044-1

ORIGINAL

A. Augarten • G. Paret • I. Avneri • H. Akons • M. Aviram • L. Bentur • H. Blau • O. Efrati • A. Szeinberg


A. Barak • E. Kerem • J. Yahav

Systemic inflammatory mediators and cystic fibrosis genotype

Received: 31 March 2004 / Accepted: 6 September 2004

Abstract Morbidity and mortality in cystic fibrosis patients patients who carried two mutations associated with a patho-
is mainly attributed to pulmonary infection and inflamma- logical sweat test and pancreatic insufficiency (∆F508,
tion. Chemokines play a pivotal role in the inflammatory W1282X, G542X, N1303K, S549R). Group B included 11
process. Although genotype-phenotype correlation in cystic compound heterozygote patients who carried one mutation
fibrosis patients has been defined, a clear relationship known to cause mild disease with borderline or normal
between the defect in the cystic fibrosis transmembrane sweat test and pancreatic sufficiency (3849+10kb C to T,
regulator (CFTR) gene and pulmonary inflammation has 5T). Associations between chemokine levels, genotype,
not been established. The aim of this study was to assess pulmonary function, Pseudomonas aeruginosa coloniza-
whether serum chemokines levels in cystic fibrosis patients tion, age, sweat chloride level, and pancreatic and nutri-
correlate with genotype and pulmonary function tests, as tional status were examined. Mean interleukin-8 and mono-
well as with other clinical characteristics. Serum levels of cyte chemoattractant protein-1 levels were significantly
interleukin-8, RANTES, and monocyte chemoattractant higher in group A than group B (11.4±2.1 pg/ml vs. 5±0.9
protein-1 were measured in 36 cystic fibrosis patients pg/ml and 157±16 pg/ml vs. 88.8±16.4 pg/ml, respectively)
grouped according to their genotype. Group A included 25 (P<0.01). No difference in RANTES levels were found
between groups. interleukin-8 levels were inversely related
to forced expiratory volume in 1 s (r=-0.37, P<0.02), while
there was no association between the latter and RANTES
and monocyte chemoattractant protein-1 levels. The
Pseudomonas colonization rate was higher among group A
A. Augarten ()
patients than group B (88% vs. 40%, P<0.01). No relation-
National CF Center,
The Chaim Sheba Medical Center, Tel-Hashomer, Israel 52621
ship was found between measured chemokines and age,
e-mail augarten@post.tau.ac.il sweat chloride levels, and pancreatic and nutritional status.
Tel.: +972-3-5303054 Our study demonstrates an association between interleukin-
Fax: +972-3-5305939 8, forced expiratory volume, and cystic fibrosis genotype.
Hence, elevated interleukin-8 serum levels could serve as
A. Augarten • G. Paret • I. Avneri • H. Akons • O. Efrati
A. Szeinberg • A. Barak • J. Yahav an indicator of an early inflammatory process and encour-
Chaim Sheba Medical Center, Tel-Hashomer, age the initiation of anti-inflammatory treatment.
affiliated to the Sackler School of Medicine,
Tel-Aviv University, Israel Key words Cystic fibrosis • Pulmonary inflammation •

M. Aviram Chemokines • Genotype


Soroka Medical Center, Be’er Sheba, Israel
L. Bentur
Rambam Medical Center, Haifa, Israel
Introduction
H. Blau
Schnieder Medical Center, Petach Tikva, Israel Cystic fibrosis (CF) is a multi-system genetic disease
E. Kerem characterized mainly by progressive lung disease, pancre-
Shaare Zedek Medical Center, Jerusalem, Israel atic dysfunction, and elevated sweat chloride [1]. The
100 A. Augarten et al.: Systemic inflammatory mediators and cystic fibrosis genotype

severity of the disease varies considerably among described standardized pulmonary equations (Polgar and
patients. Since the identification of the CFTR gene, more Promadhat). Sampling of blood for chemokine levels and spirom-
than 1,000 mutations have been recognized. The wide etry were performed during a routine outpatient clinic visit and
spectrum of disease severity and the existence of many not during an exacerbation of the pulmonary disease.
The following clinical parameters were extracted from
mutations led investigators to study the correlation
patients’ charts: age, sweat chloride level , height and weight per-
between genotype and phenotype in CF patients. These centiles, result of the last sputum culture, and pancreatic status.
studies demonstrated a direct linkage between pancreatic The pancreatic function was determined to be either pancreatic
function, nutritional status, sweat chloride levels, and sufficiency or pancreatic insufficiency by measuring pancreatic
longevity and CF genotype. However, such a relationship elastase levels in stool or by fat measurement in 72-h stool col-
has not been established between the genetic defect and lections.
the severity of pulmonary disease [2–4]. The lung tissue
damage has been mainly attributed to infection and sub-
sequent inflammation [5, 6]. The recruitment of leuko-
Statistical analysis
cytes to tissue is essential for inflammation. This process
is controlled by chemokines such as interleukin-8 (IL-8), All measurements are expressed as mean values±standard error.
monocyte chemoattractant protein-1 (MCP1), and Comparisons of chemokine levels and clinical characteristics
RANTES, which are chemotactic cytokines. It is assumed between groups were performed using Student’s t-test. The asso-
that chemokines provide directional cues for the move- ciation between serum chemokine levels and FEV1, sweat chlo-
ment of leukocytes by mediating their interaction with ride levels and age were assessed by Pearson’s coefficient.
vascular endothelium, which leads to their extravasation Analysis of variance (ANOVA) was used to evaluate the associa-
tion between chemokines and nutritional status, as expressed by
[7, 8]. In this study we evaluated the association between
different height and weight percentiles. A P value of less than 0.05
serum chemokines and the CFTR genotype, pulmonary was considered significant. Statistical calculations were per-
function, and other clinical characteristics of our CF formed using SPSS statistical package (SPSS, Chicago, Ill.,
patients. USA). The study was approved by the institutional review board
of Sheba Medical Center.

Materials and methods

The cohort of this study comprised 36 CF patients whose diag- Results


nosis was confirmed by genetic analysis and who were able to
perform spirometry. Patients were grouped according to their Table 1 displays the demographics and clinical characteris-
genotype. Group A included 25 patients who carried two muta- tics of the study cohort. No significant difference was
tions associated with a pathological sweat test and pancreatic found in the mean age between the two groups; the age
insufficiency (∆F508, W1282X, G542X, N1303K, S549R) [2]. range was 1–30 years and 3–37 years for groups A and B,
Group B included 11 compound heterozygote CF patients who respectively. All patients in group A had pancreatic insuffi-
carried one mutation known to cause mild disease with a border- ciency while 4 of 11 patients of group B had pancreatic suf-
line or normal sweat test and pancreatic sufficiency (3849+10kb ficiency. The nutritional status was better in group B
C to T, 5T) [3, 4, 9].
patients, as indicated by significantly higher mean weight
percentiles.
Sputum cultures yielded Pseudomonas aeruginosa at a
Chemokines significantly higher rate in group A patients (88% vs. 40%,
P<0.01). However, no difference in IL-8 levels was found
IL-8, MCP1, and RANTES levels were measured in venous between patients with P. aeruginosa colonization and those
blood samples (2 ml), using a sandwich ELISA (Endogen, with negative sputum cultures (9.76±9.18 vs. 7.93±12.6
Woburn, Mass., USA). The assay was performed according to the pg/ml, P=0.6).
manufacturer’s instructions, and all samples were analyzed at a Mean IL-8 and MCP1 levels were significantly higher
dilution resulting in concentrations within the range of the stan- in group A than group B; RANTES levels were not signif-
dard curve.
icantly higher in group A patients (Table 2). There was a
significant inverse relationship between IL-8 levels and
FEV1 values (Pearson r=-0.37, P<0.02) (Fig. 1). No rela-
Spirometry tionship was found between RANTES and MCP1 levels
and FEV1 values. No correlation was found between all
Forced expiratory volume at 1 s (FEV1) was expressed as a per- measured chemokines and between age, sweat chloride lev-
centage of predicted values for height and sex, using previously els, and pancreatic and nutritional status.
A. Augarten et al.: Systemic inflammatory mediators and cystic fibrosis genotype 101

Table 1 Clinical characteristics of cystic fibrosis (CF) patients

Group Aa (n=25) Group Bb (n=11) P

Age (years) 16.9±7.2 17.7±9.1 NS


Pancreatic sufficiency 0% 0/25 36.3% 4/11 <0.01
Sweat chloride (mmol/l) 105±28 92±18.6 NS
Weight percentiles 19±19.8 57.2±25.2 <0.01
Sputum Pseudomonas 88% 22/24 40% 4/10 <0.01

a Group A CF patients carrying two mutations associated with severe disease presentation (∆F508, W1282X, G542X, N1303K, S549R)
b Group B CF patients carrying mutations associated with mild disease presentation (3849+10kb CT, 5T)

Table 2 Comparison of serum chemokine levels between groups (IL-8 interleukin-8, MCPI monocyte chemoattractant protein-1)

Chemokine Group Aa Group Bb P

IL-8 (pg/ml) 11.4±2.1 5.0±0.9 0.01


MCP1(pg/ml) 157±16 88.8+16.4 0.01
RANTES (pg/ml) 323±48 287.5±93 NS

a Group A CF patients carrying two mutations associated with severe disease presentation (∆F508, W1282X, G542X, N1303K, S549R)
b Group B CF patients carrying mutations associated with mild disease presentation (3849+10kb CT, 5T)

Fig. 1 Relationship between interleukin-8 (IL-8) levels and


forced expiratory volume in 1 s (FEV1) values

combined mechanism is impaired, Pseudomonas is not


Discussion
eliminated, and a second line of defense is commenced [11].
Pseudomonas and its products stimulate chemokine produc-
Despite an increased understanding of the CFTR protein tion, which provides directional cues for the movement of
function, it is not completely understood how mutations in neutrophils, generating an inflammatory environment [7].
the CFTR gene are related to bacterial infection and inflam- The involvement of chemokines in CF is well estab-
mation of the airways. Recently, two hypotheses were pro- lished and levels of IL-8 are elevated in these patients [8,
posed to explain the link between CFTR function and the 12, 13]. However, the uniqueness of our study is the distri-
host lung defense. Pier et al. [10] suggested that the CFTR bution of serum chemokine levels among CF patients of dif-
protein itself might function as a cellular receptor for bind- ferent genotypes. We found that CF patients who carry
ing, endocytosing and clearing P. aeruginosa from the lung. severe mutations have significantly increased IL-8 levels, as
In addition, Smith and Welsh [11] found that the apical sur- well as a higher rate of Pseudomonas colonization com-
face fluid contains peptides that exhibit local antibacterial pared with patients who carry a mild mutation. In addition,
activity. This bactericidal activity requires a low salt con- we found that IL-8 levels correlate negatively with pul-
centration, which is maintained by the normal function of monary function. Welsh and Smith [14] classified the dif-
CFTR as an ion channel. This action of CFTR is a part of ferent mutations of the CF gene according to the mecha-
the first line of the lung host defense. In CF patients this nisms by which they disrupt CFTR protein function: class I,
102 A. Augarten et al.: Systemic inflammatory mediators and cystic fibrosis genotype

defective protein production; class II, defective protein pro- ed IL-8 serum levels could serve as an indicator of an early
cessing; class III, defective chloride channel regulation; and inflammatory process and encourage the initiation of anti-
class IV, defective chloride channel conduction. The two inflammatory or antibiotic therapy.
mutations carried by group B patients, namely 3849+10 kb
C  T and 5T, are splice mutations and were classified later
by Zielenski and Tsui [15] as class V mutation. In the 3849-
kb mutation, a C to T base substitution in intron 19 was References
found to create a new alternative splicing site, resulting in
the insertion of 84 base pairs into the CFTR region [4]. In 1. Mickle JE, Cutting GR (1998) Clinical implication of cystic
the 5T mutation the sequence of five thymines in intron 8 is fibrosis transmembrane conductance regulator mutations.
a DNA variant that causes skipping of exon 9 in the tran- Clin Chest Med 3:443–455
scription process. mRNA analysis of patients bearing these 2. The Cystic Fibrosis Genotype Phenotype Consortium (1993)
Correlation between genotype and phenotype in patients with
two mutations demonstrated two different transcripts, indi-
cystic fibrosis. N Engl J Med 329:1308–1313
cating synthesis of a protein with normal CFTR function 3. Augarten A, Kerem BS, Yahav Y et al (1993) Mild cystic
together with a defective non-functioning protein [4, 16]. fibrosis and normal or borderline sweat test in patients with
The association between IL-8 levels, CF genotype, and the 3849+10kb CT mutation. Lancet 342:25–26
pulmonary function demonstrated in our study, as well as the 4. Highsmith WE, Burch LH, Zhou Z et al (1994) A novel muta-
association between the CF genotype and Pseudomonas colo- tion in the cystic fibrosis gene in patients with pulmonary dis-
nization, indirectly supports both the recently reported ease but normal sweat chloride concentrations. N Engl J Med
hypotheses of Pier et al. [10] and Smith and Welsh [11]. In 331:974–980
group A patients who carry class I, II, and III mutations, the 5. Wood RE, Boat TF, Doershuk CF (1976) Cystic fibrosis. Am
Rev Respir Dis 113:833–878
CFTR is completely inactive, both as a receptor for
6. Armstrong DS, Groimwood K, Carlin JB et al (1997) Lower
Pseudomonas and as an ion channel. Therefore, Pseudomonas airway inflammation in infants and young children with cys-
binding and elimination is disrupted and the bactericidal activ- tic fibrosis. Am J Respir Crit Care Med 156:1197–1204
ity of the apical surface fluid is severely impaired due to fail- 7. Epstein FH (1998) Chemokines – chemotactic cytokines that
ure in maintaining a low salt concentration, leading to bacter- mediated inflammation. N Engl J Med 338:436–445
ial colonization with a consequent initiation of an inflamma- 8. Dean TP, Dai Y, Shute JK, Church MK, Warner JO (1993)
tory response, as represented by the elevated chemokine lev- Interleukin-8 concentrations are elevated in bronchoalveolar
els. However, in group B patients who carry class V mutation lavage, sputum, and sera of children with cystic fibrosis.
the CFTR is partially active both in Pseudomonas binding and Pediatr Res 34:159–161
in maintaining a lower salt concentration, leading to a partial 9. Kerem E, Harel NR, Augarten A et al (1997) A cystic fibrosis
transmembrane conductance regulator splice variant with par-
eradication of Pseudomonas. As a result, the bacterial burden
tial penetrance associated with variable cystic fibrosis presen-
is diminished, resulting in a milder inflammatory response and tation. Am J Respir Crit Care Med 155:1914
relatively preserved pulmonary function. However, the fact 10. Pier GB, Grout M, Zaidi TS (1997) Cystic fibrosis transmem-
that IL-8 levels were not significantly increased in patients brane conductance regulator is an epithelial cell receptor for
colonized with Pseudomonas may indicate that the inflamma- clearance of Pseudomonas aeruginosa from the lung. Proc
tory process is more complicated and not explained by solely Natl Acad Sci U S A 94:12088–12093
the hypothesis of Piers et al. [10]. 11. Smith JJ, Welsh MJ (1998) Cystic fibrosis airway epithelia
We found the IL-8 level to be related to the severity of fail to kill bacteria because of abnormal airway surface fluid.
the lung disease, while no correlation was found with the Cell 85:229–236
12. Segal SD, Sontag MK, Wagener JS et al (2002) Induced spu-
other examined clinical characteristics, such as sweat chlo-
tum inflammatory measures correlate with lung function in
ride levels and pancreatic and nutritional status. This may children with cystic fibrosis. J Pediatr 141:811–817
indicate that the measured chemokine originated in the lung 13. Dakin CJ, Numa AH, Wang HE et al (2002) Inflammation,
and that in the other organs affected by CF, the pathogenesis infection, and pulmonary function in infants and young children
does not involve infection and inflammation. Our work is with cystic fibrosis. Am J Respir Crit Care Med 165:904–910
also distinctive as we measured the chemokine levels in the 14. Welsh MJ, Smith AE (1993) Molecular mechanism of CFTR
serum, which is much easier and less invasive than measure- chloride channel dysfunction in cystic fibrosis. Cell
ment in the broncho-alveolar fluid. The high serum concen- 73:1251–1254
tration of IL-8 in CF patients probably represents a spillover 15. Zielenski J, Tsui LC (1995) Cystic fibrosis: genotypic and
phenotypic variations. Annu Rev Genet 29:777–807
from the localized inflammatory response in the lung [8].
16. Chu CS, Trapnell BC, Curristin S et al (1993) Genetic basis of
It was shown that airway infection and subsequent variable exon 9 skipping in cystic fibrosis transmembrane
inflammation may be present in the first weeks of life [6] conductance regulator mRNA. Nat Genet 3:151–156
and that early treatment of asymptomatic CF children 17. Wang SS, O’eary LA, FitzSimmons SC et al (2002) The
resulted in better pulmonary function [17]. Therefore, our impact of early cystic fibrosis diagnosis on pulmonary func-
findings may be of practical clinical importance, as elevat- tion in children. J Pediatr 141:804–810

Vous aimerez peut-être aussi