Inflammatory markers in cystic fibrosis patients with lung ... - NCBI

Research Paper Mediators of Inflammation, 8, 159–167 (1999)

CHR ONIC e ndobro nch ia l in flam m ation and bacte rial in fe ction ar e the m ain caus es of m orbidity and m or tality in cys tic fibro s is (CF), an auto s om al r eces s ive gen etic dis or der as s ociate d w ith im p r op e r fun ction of ch loride ch an ne ls. In flam m ation in CF lun g is gr eatly am p lified afte r Ps e u do mo na s a e r ug ino s a in fe ction . In th is s tudy th e r elation s h ip betw e en P. a e ru g ino s a s tatus an d in flam m atory m arkers h as be en in ve s tigated. Se ven teen CF ch ildr en in acute lun g ex acerbation w e re ex am in e d. CF p atien ts w ith out P. a e r u g ino s a in fe ction w e re ch ar acte rized by ele vated activity o f s p utum elas tas e, r educed r es p o ns e of p erip h eral blood lym ph ocytes to PHA an d s ign ifican t r es is tance to th e antip ro life rative action of glucocor ticoids . Th es e p aram eters w er e no r m alized afte r an tibio tic tr eatm en t. Th e p atien ts w ith pr o lon ged P. a e r u g ino s a in fe ction dem on s trated ex tr em e ly h igh levels of elas tas e activ ity and e le vated am o un ts o f s p utum IL-8 and TNF-a . Alth o ugh an tibiotic tr eatm en t r es ulte d in clin ical im pr ov em en t, it fa iled to s up pr e s s ex ces s ive im m un e r es p o ns e in the lun g. Th e data in dicate that CF patie nts w ith p ro lon ged P. a e r ug ino s a n ee d the m o difie d tr e atm en t, w h ic h s h ould in clude im m un om o dulatin g drugs and p r oteas e in h ib ito rs as w ell as an tibacte rial th erap y.

Inflammatory markers in cystic fibrosis patients with lung Pseudomonas aeruginosa infection A. L. Pukhalsky,1,C A N. I. Kapranov,2 E. A. Kalashnikova,1 G. V. Shmarina,1 L. A. Shabalova,2 S. N. Kokarovtseva,1 D. A. Pukhalskaya ,1 N. J. Kashirskaja2 and O. I. Simonova2

1

Laboratory of Immunogenetics, 2 Department of Cystic Fibrosis, Research Centre for Medical Genetics, 1 Moskvorechie Street, Moscow 115478, Russia

CA

Corresponding Author Fax : (+7) 095 324 0702

Key w o r ds: IL-8, TNF-a , Elastase , PHA re sponse, Ste roid sensitivity, Cystic fibrosis, Ps e u do m o n a s a e ru g in o s a , Ex c essive inflammatory response

Introduction Cystic fibrosis (CF) is a common, se rious, and often fatal autosomal rec essive gene tic disorder, w hich charac te rize d by oversec re tion of p ulmonary mucus, bacte rial infec tions and respiratory congestion. The CF gene c odes for a cellular membrane protein, w hich forms a chloride channel and ap pears to regulate othe r channel proteins. Imprope r function of this channe l results in ele ctrolyte abnormalitie s at the surface s of various sec retory and re sorptive c ells.1,2 The absorption of sodium across airw ay e pithe lia is inc re ased 2–3 fold. As a re sult of the associate d inc re ase in osmotic w ater absorption, the hydration of the airw ay surface liquid is likely to be re duce d, w hich it is thought affec ts the mucocilliary clearance and bac te rial adhe renc e. These abnormalities may be ex plain w hy CF patients suffer from re spiratory trac t infec tions;3 at the same time CF p atie nts have no dete ctable immune defic ie nc y. The y are not more susc eptible to infe ctions, ex c ept those of the the respiratory tract, than normal childre n of the same age.4 The pulmonary infe ctions incite an inte nse host inflammatory re sponse, causing progressive supp urative pulmonary dise ase. This response is characte rized by a marked influx of neutrophils into the

lung, and e le vation in inflammatory mediators such as TNF-a , IL-1b , IL-6, IL-8, and le ukotrie ne B4 .5,6 As a rule CF patie nts suffe r from re current and chronic e ndobronchial Ps e u d o m o n a s a e ru g ino s a infe c tions w hich are ve ry serious in te rms of clinical prognosis.7 ,8 The re are different phase s of P. a e ru g ino s a colonization. Initial pathogen c olonization is characte rized by sele ction of alginate (mucoid) producing mutants and a biofilm formation. The nex t stage is charac te rize d by the appe arance of new mucoid mic roc olonies, dec reased production of ex tracellular virulenc e factors and minimal tissue damage . A te rminal phase is manifeste d in high bac te rial ce ll density, e levated release of ex tracellular virulenc e factors and massive tissue damage .9 In this article w e desc ribe the relationship betw ee n the pe riod of P. a e ru g in o s a infection and several inflammatory markers in CF pediatric patients.

Material and methods Patient assessment Se ve nte en CF children (mean age 12 ye ars) w ere rec ruited for the study. The y w ere lung ex acerbation patients from the De partment of Cystic Fibrosis of the

ISSN 0962-9351 print/ISSN 1466-1861 online/99/030159-09 1999 Taylor & Francis Ltd

159

A. L. Pu kh a ls k y et al.

Research Ce ntre for Medical Ge netic s (Moscow ). The patients show e d variable dise ase severity and different P. a e ru g ino s a status. Five subjects w ere not colonized w ith the pathogen and tw e lve children w ere infected w ith muc oid strains of P. a e ru g in o s a . The last group of patients included four individuals w ho had been colonized w ith P. a e ru g ino s a for le ss than tw o ye ars (short-te rm infec tion) and e ight childre n c olonize d for tw o ye ars or more (prolonged infec tion). The diagnosis of acute pulmonary ex ac erbation w as defined as a marked inc rease in C reactive prote in, by w eight loss, anorex ia, increase d cough, inc re ased sp utum produc tion, feve r w ith and w ithout new lung infiltrates, and de te rioration of ox ygen saturation and pulmonary func tion. The follow ing pulmonary func tion te sts w ere performed: force d vital capac ity; forced ex piratory volume and ox ygen saturation measureme nts after a w alking te st. Patie nts w ith acute pulmonary ex acerbation w e re treate d w ith basic the rap y (microspheric enzyme s, multivitamins, high c alorie diet, mucolytic s) and antibiotics. Antiba c te rial treatment de pende d on the mic robiology analysis of sp utum. In the case of P. a u ro g e n o s a infec tion, c ephalosporins of 3rd gene ration in combination w ith aminoglycosids or c ip roflox ac in w e re presc ribed.

Blood collection and sputum processing Blood w as c ollec te d in tubes w ith heparin (25 IU/ml) by venipuncture . The w eight of e ach sputum sample w as c alc ulated. The same w eight of PBS w ithout Ca 2 + and Mg 2 + w as added to the sputum sample . The mix ture w as plac ed on vortex for 10 se conds and then on the rocker for 30 min. The sample filtere d through a 100 m m filte r to re move muc us. The filtrate has bee n ce ntrifuge d at 400 g for 10 min at 4°C to pellet the c ells. Prote in conc entration w as me asure d using Bradford’s method.1 0,11 The supernatant w as then removed, aliquote d and stored at –60°C.

Assay of human leucocyte elastase The assay method used is base d on the ability of ne utrophil e lastase to interact w ith the spe cific chromogenic BANE (N-t-Boc-L-alanine p-nitrophenyl este r) (Reanal, Hungary) at acidic pH forming p-Nitrophe nol w ith max imum of absorbance at 347.5 nm.1 2,13 The standard assay w as pe rforme d in 0.6 ml of a solution containing an aliquot of sputum sample (20–200 m l), 0.01 M BANE (20 m l) and 0.05 M sodium phosp hate at 24°C and pH 6.5. Probes w ere assaye d for absorbance at 347.5 nm for 12–15 min. Absorbance p er minute w as then acc ounted. The amount of elastase w as calculated using formula: Elastase activity (U/ml) = D3 47 .5 3 109/V 160

Mediators of Inflammation ´ Vol 8 ´ 1999

w here D3 47 .5 = absorbanc e pe r minute; V = volume of the sputum aliquot adde d; 109 is a parameter w hich include s ex tinction coeffic ie nt, the le ngth of light path, and volume of the re ac tion mix ture. Under these conditions one unit of human neutrophil elastase ac tivity w as that amount w hich hydrolyze d 1 nM of BANE per minute. Finally the value of neutrophil elastase activity w as normalized to the protein content in e ach sample of the sputum ex trac t.

Inhibition of PHA-induced lymphocyte proliferation by dexamethasone Mononuclear cells w ere isolated from he parinaze d peripheral blood by Fic oll-Verographin de nsity gradient c entrifugation. The cells w ere w ashed tw ic e in RPMI-1640 medium (ICN, USA) supplemente d w ith 10% he at-inac tivated donor horse se rum, 23 10 –3 M HEPES, 2 mM L-glutamine, 2.83 10 –6 M 2-mercaptoethanol, and 20 m g/ml gentamycin. The c ells w ere cultivated in flat-bottome d 96-w e ll plates (Costar, USA), and contained 53 10 4 c ells in each w e ll. The final conce ntration of PHA (Sigma, USA) w as 5m g/ml. Inhibition of PHA stimulation by dex ametasone (Dm) w as e valuated at six differe nt c once ntrations w ithin the dose range of 10 –10 to 10 –6 M. Dm w as not adde d to the control w ells (these c ontained a culture me dium w ith or w ithout PHA). The c ells w e re inc ubate d for 72 h at 37°C in humidified atmosphere containing 5% CO2 . Four hours be fore the end of cultivation, each w ell w as pulsed w ith 40 kBq of [ 3 H]-thymidine (Isotope , Russia). The c ells w ere harveste d w ith a ce ll harve ste r and counte d on a liquid scintillation counte r. Triplicate w e lls of each conc entration w e re assaye d and the counts pe r minute (count/min) w e re averaged. Pe rcentage inhibition w as calculate d by dividing the c ount/min in e ach inhibite d sample by the count/min in the sample containing PHA only, and subtracting the background le vel (c ounts/min in the w e lls w ith culture medium only) from these value s. The inte nsity of suppre ssion w as e stimate d by probit-analysis and ex pre sse d as ED5 0 : a dose of immunosuppre ssive agent at w hich lymphoc yte prolife ration w as 50% of its max imum. Previously the dire ct positive corre lation be tw een the le ve l of PHA-induc ed lymphoc yte prolife ration and of the inhibition de gree of such stimulation by dex amethasone (ED50 ) has be en show n. On the basis of this c orre lation w e propose d the method of e valuation of individual susc eptibility to the antiproliferative effe ct of glucocorticoids by D h -parameter calculation.1 4 D h w as calc ulate d using formula: D

h

= Y – Y9

w here Y = logED5 0 ; Y9 = 0.447X – 4.399; X = ln(count per min).

In fla m m a to ry m a rke rs in cy s tic fibro s is pa tie n ts

Cytokine assays TNF ac tivity w as determine d by the method of Ruff and Gifford 1 5 w ith some modifications. Briefly, L929 ce lls w e re see de d at a density of 33 10 4 ce lls per w ell in 96-w e ll p late s in 100 m l of medium 199, to w hich 10% he at inactivate d calf bovine serum and gentamycin had been adde d. Plate s w ere inc ubate d at 37°C in a humidified atmosphere containing 5% CO2 until monolaye r formation. After the culture me dium e limination, tw o-fold serial dilution of the samples (100 m l of each dilution) and 100 m l fre sh culture me dium w ith 20 m g/ml of ac tinomyc in D (Serva, Ge rmany) w ere adde d, and further incubated for 18 h in the same conditions. Supe rnatants w ere then remove d and cells stained w ith 0.2% crystal violet (Sigma, USA). Afte r w ashing and drying plates w ere finally read at 595 nm on a Titerte k Multiskan microElisa reader. Human re combinant TNF-a (Institute of Bioorganic Che mistry, Mosc ow, Russia) w as used as internal standard. The probitanalysis method w as use d for the comparison of the ex pe rimental and calibrating curves. TNF conte nt in the sample s w as ex pressed in pg/ml and w as normalized to the prote in content in each sputum sample. IL-8 w as de te rmined in the sputum samples using a commerc ially available ELISA (Prote inovyi Kontur, St Pe te rsburg, Russia)

Statistical analysis Statistical analysis w as pe rforme d using Student’s t-te st. Correlation betw e en elastase ac tivity and force d vital capac ity w as analyze d using Spearman rank correlation.

Results Neutrophil elastase activity in the sputum of CF patients During acute pulmonary ex ace rbation all CF patients ex hibited ele vate d le ve ls of neutrophil elastase ac tivity (Fig. 1). The e lastase activity in the sputum of patients w ith short-term P. a e ru g ino s a infe ction appeared to be at or below the prote ase amount in the sputum of uninfec te d subjec ts (8.1 ± 1.7 U/mg prote in and 10.7 ± 1.5 U/mg, respec tive ly, p = 0.3). In contrast, CF patients w ith prolonged P. a e ru g ino s a infec tion de monstrated ex tremely large p rote ase amounts in their sputum sample s. Elastase ac tivity in these p atie nts w as signific antly higher than those in tw o other patient groups (34.7 ± 5.5 U/mg, p< 0.003). Follow ing the antibiotic therapy the measurements of elastase ac tivity re ve aled that prote ase amount w as markedly decre ased in the sputum of P. a e ru g ino s a free p atie nts (2.2 ± 0.6 U/mg, p< 0.03) and in individuals w ith short-te rm infec tion (4.7 ± 1.2 U/mg, p< 0.02). At the same time, antibiotic administration

FIG. 1. Neutrophil elastase activity in the lung of CF patients before and after antibiotic treatment., Sputum samples from CF patients with different P. aeruginosa status were assessed for protease presence as described in Materials and Methods. Results are given as units of elastase activity per mg protein and expressed as mean ± SEM. A two sample Student’s t-test was used to compare the means of elastase activity in separate patient groups before and after antibiotic treatment (p1 ,4 < 0.03; p2,5 < 0.02). Differences in elastase activity levels between patient groups were analyzed using paired, two-tailed Student’s t-test (p1,3 < 0.003; p4 ,6 < 0.02). Mediators of Inflammation ´ Vol 8 ´ 1998

161


FIG. 2. Association betweeen lung elastase activity and forced vital capacity (FVC) in CF patients. The relationships between protease activity and FVC were analyzed by using Spearman rank correlation (R = ± 0.536; p < 0.04).

faile d to suppress e levated elastase ac tivity in the sputum from patie nts w ith prolonged P. a e ru g ino s a infec tion. Prote ase activity w as inc reased 10–20 fold compared w ith patients w ith short-te rm colonization and uninfec te d subje cts. The neutrophil elastase ex pre ssion in CF patie nts w as highly correlated w ith the lung func tion failure (Fig. 2). A significant negative linear assoc iation w as found betw ee n value s of force d vital capac ity and lung elastase activity (R = – 0.536; p< 0.04).

IL-8 concentrations in the sputum of CF patients As can be see n in Fig. 3, a statistically significant ele vation of IL-8 c onc entration in sputum samples from patients w ith prolonged P. a e ru g ino s a infe ction compared w ith uninfec te d patie nts has be en found (7.3 ± 0.6 ng/mg prote in and 3.7 ± 0.7 ng/mg, respective ly; p = 0.02). Patients w ith short-te rm P. a e ru g in o s a infec tion ex hibited IL-8 le ve ls similar to

FIG. 3. IL-8 concentrations in the sputum of CF patients during exacerbation treatment. Sputum samples were obtained from CF patients before and after antibiotic treatment and assessed for IL-8 content by ELISA. Values were normalized to sputum protein concentrations and represent the average ± standard deviation. An umpaired, two-tailed Student’s t-test (p1,3 = 0.02; p4,6 = 0.29). 162



FIG. 4. TNF-a levels in the lungs of CF patients with different P. aeruginosa status. Sputum samples were obtained from CF patients before and after antibiotic treatment. A cytokine concentration was determined using the biological test as described in Material and Methods. Median values are indicated by the lines. Each dot represents one patient. Data reported were compared by unpaired, two-tailed Student’s t-test. There were no statistically significant differences between patient groups.

those observe d in childre n w ith prolonged colonization, but there w e re no marked differenc es betw ee n this group and uninfected patients (6.7 ± 1.8 ng/mg, p = 0.2). Antibiotic treatme nt did not induce significant alteration in IL-8 le ve ls in patients w ith short-te rm (8.5 ± 2.2 ng/mg, p = 0.55) or prolonged colonization (5.13 ± 0.9 ng/mg, p = 0.19), as w e ll as in uninfected patie nts (3.8 ± 0.8 ng/mg, p = 0.97). How e ver, the cytokine conc entrations in the sputum samples from P. a e ru g in o s a -fre e children and patie nts w ith prolonged colonization, be came statistic ally e quivalent.

TNF-a and protein concentrations in the sputum of CF patients Due to large individual variations, the measure ments of TNF-a conc entrations did not reveal statistic ally significant differenc es be tw een patie nt groups (Fig. 4). Prior to antibiotic administration the median cytokine le ve ls dete cted in the sp utum of uninfec te d subjects, patients w ith short-te rm coclonization and childre n w ith prolonged P. a e ru g ino s a infe ction w ere 48 pg/mg prote in, 13 pg/mg and 770 pg/mg, respe c-

FIG. 5. Protein concentrations in the sputum samples from CF patients during exacerbation treatment. Determination of protein amount was performed by Bradford’s method. Data reported were compared by unpaired, two-tailed Student’s t-test. There were no statistically significant differences between patient groups. Symbols as in Fig. 4. Mediators of Inflammation ´ Vol 8 ´ 1998

163

A. L. Pu kh a ls k y et al. Table 1. Effect of antibiotic treatment on PHA-induced peripheral blood lymphocyte proliferation. Data are presented as mean ± SEM Patients

Proliferative response (count per min)

Without P. aeruginosa (n = 5) With short-term P. aeruginosa infection (n = 4) With prolonged p. aeruginosa infection (n = 7)

Before treatment

After treatment

61002 ± 10439 100226 ± 26936 90374 ± 20465

103122 ± 15473* 920023 ± 23292 70269 ± 20712

*P < 0.02, two samples Student’s t-test.

tively. Antibiotic tre atment had no significant e ffec t on TNF-a leve ls in all patient groups. Thus, low est TNF-a level (17.3 pg/mg) in the sputum sample s of CF patients w ith short-te rm infe ction has bee n found. Uninfe cte d subjec ts show e d inte rmediate c onc entrations of the cytokine (46.4 pg/mg). Patients w ith prolonge d P. a e ru g ino s a infe ction demonstrate d the highest le vels of the sp utum TNF-a (2539.8 pg/mg) as before tre atment. The prote in levels in the sputum samples from CF patie nts w ith diffe rent P. a e ru g ino s a status have bee n analyzed (Fig. 5). During acute lung ex ace rbation, the patients w ith short-te rm P. a e ru g in o s a infe ction and uninfe cte d CF subje cts show e d

comparable prote in conte nts (me dian 1.02mg/ml and 0.88 mg/ml, respec tive ly). Antibiotic administration did not influenc e on the prote in conce ntration in uninfec te d childre n (0.84 mg/ml) and moderately decre ased protein le ve ls in the sputum of patients w ith short-term infec tion (0.51 mg/ml). In contrast, the most sputum samples from patients w ith prolonged P. a e ru g ino s a infection containe d increase d prote in amount (1.49mg/ml), w hich markedly elevated after antibiotic tre atment (2.77 mg/ml). De spite of the vis ible distinc tions in protein le vels, statistic al comparison did not re ve al significant diffe re nce s betw e en the patients groups.

FIG. 6. Peripheral blood lymphocyte (PBL) susceptibility to antiproliferative effect of dexamethasone in CF patients. Inhibition degree of PHA-induced lymphocyte proliferation by different concentrations of dexamethasone was evaluated. The cell sensitivity is presented as a mean of D h values (see Material and Methods). The asterisk indicates the D h values in P. aeruginosa-free patients before and after antibiotic treatment are significantly different (p < 0.05; two sample Student’s t-test), i.e. after treament PBL became more sensitive to glucocorticoid hormones.

164



Proliferative response to PHA and lymphocyte susceptibility to glucocorticoids There w e re no signific ant differences in the levels of prolife rative re sponse betw ee n all patie nt groups (Table 1). Antibiotic administration did not induce any alteration in T c ell prolife rative response in P. a e ru g in o s a infec te d patients. At the same time PHA-induce d prolife ration of lymphoc ytes from uninfected subje cts, low er at ex ace rbation, w as significantly inc re ased after antibiotic treatment (p< 0.02). To evaluate individual susc eptibility to glucocortic oids, the D h -parameters w ere ac counted as de scribed in ‘Material and Methods’. Patie nts w ere classifie d as ste roid resis tant if their D h -parameters increase d > 0 and steroid sensitive if this parame te r failed to inc re ase > 0 (Fig. 6). During the ac ute lung ex acerbation pe ripheral blood lymp hocyte s obtained from CF patients uninfec te d w ith P. a e ru g ino s a demonstrate d the resis tance to antiproliferative action of Dm (D h = 0.34 ± 0.26). In contrast, the patients w ith short-term and prolonge d P. a e ru g ino s a infe ctions w ere re latively steroid-se nsitive (D h = –0.05 ± 0.36 and D h = –0.02 ± 0.14, re spe ctively). Afte r antibiotic administration, uninfec te d childre n show ed the sw itch from steroid-resistance to steroid-sensitivity (D h = – 0.54 ± 0.22, p< 0.02). Patie nts w ith short-term P. a e ru g ino s a infec tions tended to show highe r responsive ness to Dm (D h = –0.39 ± 0.28). The D h -parame te rs in individuals w ith prolonged P. a e ru g in o s a infe ction w ere of intermediate values and did not significantly differ from those before antibiotic administration (D h = 0.04 ± 0.18, p = 0.78).

Discussion During the first years of life, young childre n w ith CF are c olonize d and de velop pneumonia sec ondary to Sta p hy lo c o c c u s a u re u s , Ha e m o p hilu s influ e n z a e , or less c ommonly, Kle bs ie lla pn e u m o n ia e . These infe ctions are re lative ly e asy to treat w ith the appropriate antibiotic s.16 ,1 7 In our study the ac ute lung ex ace rbation in P. a e ru g ino s a -free CF patie nts w as associate d w ith Sta phy lo c o c c u s a u re u s infe ction and characterize d by ele vated ac tivity of sp utum e lastase , re duc ed re sponse of periphe ral blood lymphoc ytes to PHA and significant re sistanc e to antiprolife rative action of gluc oc ortic oids. Such resis tance is associated w ith ac ute lung inflammation and is ac companied by the large numbers of activated lymphocytes producing e le vated amounts of IL-2. Follow ing suc cessful antibiotic treatme nt, reduction of the sputum e lastase activity, increase d peripheral blood lymphocyte re sponse to PHA, as w ell as the change from ste roid resis tance to steroid se nsitivity of lymp hocyte s have be en observe d. In this patie nt group normalization of laboratory

parame te rs w as strongly re lated to e vident clinic al improve ment. Although sputum amounts of ac tive elastase decre ased five-fold afte r antibiotic tre atment, complete inhibition of prote ase ac tivity did not occur. In addition, antibiotics did not affec t the elevate d conc entrations of IL-8 in the sputum. It is possible that the stable ele vation of the c ytokine is associate d w ith CF pathoge nesis. It is know n that the NaCl c once ntration in CF bronchial sec retion liquids is highe r than that found in normal subje cts.18 – 2 0 High NaCl c once ntrations may c ontribute to a prolonged inflammatory state by re le asing high amounts of IL-8 from CF airw ay submucosal glands. Consequently, neutrophils activated by loc ally se creted IL-8 rele ase elastase and ox idants that stimulate airw ay surface e pithe lial cells to produc e othe r che motac tic fac tors, such as TNF-a , IL-1b , IL6, and IL-8,21 ,22 w hich may gene rate and perp etuate an inflammatory vic ious cycle in CF airw ays. This inflammation c an be amplified after P. a e ru g ino s a infe c tion.23 In most CF patients P. a e ru g ino s a colonization is initiate d by nonmucoid strains. Hype rosmolar environment of the CF airw ays as w ell as ox ygen radicals produce d by the inflammatory response induc e the phenotypic change from non-alginate producing to alginate (muc oid) produc ing phenotyp es of P a e ru g ino s a .24 – 2 6 Alginate , unbranched line ar hete rop olysaccharide , is a signific ant virule nce fac tor. Initially, a muc oid c oat is produc ed around single bacte ria and late r surrounds several micro-colonies of bacterial c ells. It has be en show n in v itro , that P. a e ru g in o s a grow ing in alginate p roduc ing mic rocolonies (biofilms) is highly re sistant to opsonic and nonopsonic phagocytosis,2 7 c omp le ment,28 re ac tive ox ygen inte rmediates 2 9,3 0 and antibiotic s.3 1,3 2 Biofilm bacteria behave as a p opulation inste ad of individual c ells.32 ,3 3 The y are able to estimate their ow n cell de nsity, inte ract w ith e ach othe r and react appropriate ly to environme nt change s. This phenomenon is te rmed quorum-sensing or ce ll-to-c ell signaling.34 ,3 5 P. a e ru g in o s a biofilms appear to coordinate ex pre ssion of virulence ge nes and c ontrol the p roduc tion of many ex trac ellular virulence fac tors by quorum-se nsing systems. A spe cific transc riptional activator of virule nce gene ex pression (R-protein) is not active w ithout the corresp onding autoinducer. 3 4 The latte r is synthe size d at basal le ve ls and diffuses into the surrounding me dia. At low cell density, the intrace llular c onc entration of autoinduc er is not enough to activate a spe cific R-prote in. So the inte nsity of virule nce ge ne transc ription and, as a consequenc e, produc tion of ex trac ellular virulenc e factors are also low. With inc reasing c ell density the intrace llular amount of autoinduc er reaches critic al conc entration for a spec ific R-protein activation. The resulting increase in ex pre ssion of virule nce genes Mediators of Inflammation ´ Vol 8 ´ 1998

165


can reach 1,000-fold. Elevated ex pre ssion of bacterial ex oproducts initiate s ex uberant host immune response, including ex c essive production of proinflammatory cytokines and ne utrophil recruitme nt w ith large amounts of elastase being re le ased.9 It is prec isely this fac t that ex plains se rious pulmonary destruc tion in CF patients c olonized w ith P. a e ru g in o s a .36 No w onder the appropriate antibiotic the rapie s are ofte n unable to stop this c ourse. This situation w as clearly demonstrate d in our study. Thus, ex treme ly high e lastase ac tivity, elevate d amount of IL-8 and TNF-a , as w ell as increase d prote in conce ntration s, have bee n observe d in the sputum samples from ex amined CF patients w ith prolonge d P. a e ru g in o s a infection. It appears that entire biofilm in airw ays of these patients produce s enough amount of autoinduc er to initiate elevate d production of virulence fac tors. Although antibiotic tre atment resulte d in clinical improve ment, it faile d to suppress ex ce ssive immune re sponse in the lung. By c ontrast, an immature P. a e ru g in o s a biofilm in patients w ith short-te rm infe ction is not a sourc e of virulenc e fac tors. In addition, the alginate c oat preve nts c ontacts of the host’s defe nse syste ms w ith bacte ria. For this re ason inflammatory response in the patients w as c omparable to that in uninfec te d childre n. Lung ex ace rbation in patients w ith shortte rm colonization and P. a e ru g ino s a -fre e subjec ts w as assoc iate d w ith acute S. a u re u s infec tion, w hich w as succ essfully treated w ith antibiotic s. All CF patients w ith P. a e ru g ino s a infec tion (prolonge d and short-te rm) demonstrate d high IL-8 le ve ls in comparison w ith uninfected individuals (see Fig. 3) that may be ex plained by influenc e of P. a e ru g ino s a products (e .g. alginate ) on airw ay e pithe lial c ells.22 Re lative ly low lymphocyte sensitivity to the antiprolife rative effe cts of gluc oc orticoids (D h about 0) in P. a e ru g ino s a -infec te d patie nts is evide nce of permane nt inflammation in airw ays. The sw itch from resis tance to sensitivity in patients w ith short-term colonization, show s that the antibiotic treatme nt is able to suppress the inflammation. The same suppression is impossible in the subjects w ith prolonged P. a e ru g in o s a infection (se e Fig. 6). Our data suggest that CF patie nts chronic ally infec te d w ith P. a e ru g in o s a , e spec ially the subjec ts w ith prolonged colonization, ne ed the modifie d approach to their treatment. This approach has to include , be sides antibacte rial the rap y, immunomodulating drugs and protease inhibitors. The first are inte nded for the inhibition of ex c essive immune response and the sec ond are nece ssary to pre ve nt ne utrophil elastase -mediated pulmonary destruction. The novel strate gy of CF treatme nt also require s permane nt immunological monitoring w ith evaluation of ne utrophil e lastase activity and proinflammatory c ytokine le ve ls in the sputum sample s of CF patie nts. 166


References 1. Graham A, Hasani A, Alton EW, e t a l. No added benefit from nebulize d amiloride in patients w ith Cystic Fibrosis. Eu r Re s p J 1993; 6: 1243–1248. 2. Høiby N, Koch C. Ps u do m o m a s a e r u g ino s a infection in Cystic Fibrosis and its management. Tho ra x 1990; 45: 881– 884. 3. Koch C, Høiby N. Pathoge nesis of cystic fibrosis. La nc e t 1993; 341: 1065–1069. 4. Pedersen SS. Lung infec tion w ith alginate-producing, mucoid Ps u do m o n a s a e ru g in is a in cystic fibrosis. APMIS 1992; 100 (Suppl. 28): 5–79. 5. Konstan MW, Be rger M. Infection and inflammation of the lung in cystic fibrosis. In: Davis PB, ed. Cy s tic Fibro s is . Ne w York: Marc el De cke r, 1993; 219 –276. 6. Konstan MW, Hillard KA, Norvell TM, Be rge r M. Bronchoalveolar lavage findings in cystic fibrosis patients w ith stable, clinically mild lung disease suggest ongoing infe ction and inflammation. Am J Re s p ir Crit Ca re Med 1994; 150: 448–454. 7. Fick RBJ, Sonoda F, Hornick DB. Eme rge nce and pe rsistence of Ps u do m o n a s a e ru g inis a in the cystic fibrosis airw ay. Se m in Re s pir In fe ct 1992; 7: 168 –178. 8. Govan JRW, De re tic V. Mic robial pathogenesis in c ystic fibrosis: muc oid Ps u do m o n a s a e ru g inis a and Burkho lde ria cepa cia . Micro bio l Re v 1996; 60: 539 –574. 9. Van De lden C, Iglew ski BH. Cell-to-ce ll signaling and Ps e u do m o n a s a e ru g in o s a infec tions. Em e rging In fe c tio u s Dis e a s e s 1998; 4:551–560. 10. Bradford MM. An n Bio c he m 1976; 72: 248 –254. 11. Brogdon WG, Dickinson CM. An n Bio c he m 1983; 131: 499–503. 12. Visser L, Blout ER. The use of p-nitrophenyl-N-tetrabutoyc arbonylL-alaninate for elastase. Bich e m Bip hy s Acta 1972; 268: 257–260. 13. Kaminskaya GA, Zhukova NL, Stepanyan IE. Comparison of tw o methods for the study of the sputum elastolytic ac tivity and assessment of the re sults. La bo ra torn o y e Delo 1984; 2: 110 –113 (in Russian). 14. Pukhalsky AL, Kalashnikova EA, Lyashko VN, Pe vnitsky LA. Inhibition of phytohemagglutinin-induc ed lymphocyte proliferation by dex ame thasone : me chanisms of individual susce ptibility. Int J Im m u no p ha rm a c 1990; 12: 657–663. 15. Ruff MR, Gifford GE. Tumor ne crosis fac tor. In: Pick E, ed. Ly m pho k in e s . Ne w York: Academic Press. 1981; 235 –241. 16. Whe ele r WB, Colten HR. Cystic fibrosis: Curre nt approach to diagnosis and management. Pe dia tr Re v 1988; 9: 241– 248. 17. Saiman L. Tre atment of infections in patients w ith cystic fibrosis. In fe ct Med 1993; 10: 37– 43. 18. Van Hee ckern A, Walenga R, Konstan MW, Bonfield T, Davis PB, Fe rkol T. Ex c essive inflammatory response of cystic fibrosis mic e to bronchopulmonary infec tion w iyh Ps e u do m o n a s a e ru g ino s a . J Clin Inve s t 1997; 100: 2810 –2815. 19. Joris L, Dab I, Quinton P. Elemental composition of human airw ay surface fluid in healthy and dise ased airw ays. Am Re v Re s p ir Dis 1993; 148: 1633–1637. 20. Smith JJ, Travis SM, Gre enbe rg EP, Welsh MJ. Cystic fibrosis airw ay epithe lial fail to kill bacte ria because of abnormal airw ay surfac e fluid. Ce ll 1996; 85: 229–236. 21. Massion PP, Inou H, Richman-Eise nstat J, e t a l. A. Novel Ps e u do m o n a s product stimulates interleukin-8 production in airw ay epithelial c ells in vitro . J Clin Inve s t 1994; 93: 26–32. 22. Rue f C, Je fferson DM, Schlegel-Hauter SE, Suete r S. Re gulation of cytokine secretion by cystic fibrosis airw ay epithelial cells. Eur Re s p ir J 1993; 6: 1429 –1436. 23. Van Hee ckern A, Walenga R, Konstan MW, Bonfield T, Davis PB, Fe rkol T. Ex c essive inflammatory response of cystic fibrosis mic e to bronchopulmonary infec tion w iyh Ps e u do m o n a s a e ru g ino s a . J Clin Inve s t 1997; 100: 2810 –2815. 24. Pie r G. Pulmonary dise ase associate d w ith Ps e u do m o na s a e r ug in o s a in cystic fibrosis: Current status of the host-bacterium interaction. J In fe c t Dis 1985; 4: 575–580. 25. Govan IRW, Harris GS. Ps e u do m o na s a e ru g in o s a and cystic fibrosis: Unusual bacterial adaptetion and pathigenesis. Micro bio l Sc i 1986; 3: 302–308. 26. Mathe K, Sternbe rg C, Ciofu O, e t a l. Ox yge n radic al induc tion phe notypic change from non-alginate producing to alginate -producing form of Ps e u d o m o na s a e ru g in o s a in biofilm. In: Lore nzo V, ed. Ps e u do m o na s ’97. VI In tern a tio na l Co n g res s o n Ps e ud o m o na : Mo le cu lar bio lo g y a nd bio te chn o lo g y, Madrid, Spain, 1997: 116. 27. Spee rt DP, Ps e u do m o n a s a e ru g in o s a infe ctions in patients w ith cystic fibrosis. In: Baltch AL, Smith RP e ds. Ps e u do m o n a s a e r u g in o s a in fe c tio ns a nd tre a tm e nt.Ne w York, Marc el De kker, 1994: 183 –236. 28. Anw ar H, Strap JL, Coste rton JW. Susc eptibility of biofilm ce lls of Ps e u do m o na s a e ru g in o s a to bactericidal action of w hole blood and serum. FEMS Micro b io l Le tt 1992; 92: 235– 242. 29. Le arn DB, Bre ste l EP, Se etharama S. Hypochlorite scavenging by Ps e u do m o na s a e r ug in o s a alginate. Infe ct Im m u n 1987; 55: 1813–1818. 30. Simpson JA, Smith SE, De an RT. Scavenging by alginate of free radic als re leased by mac rophage s. Fre e Ra dica l Bio l Med 1989; 6: 347–353.

In fla m m a to ry m a rke rs in cy s tic fibro s is pa tie n ts 31. Giw ersman B, Je nsen ET, Høiby N, Kharazmi A, Costerton JW. Induc tion of b lactamase production in Ps e u do m o n a s a e ru g in o s a biofilm. An tim icro b Age nts Ch e m o the r 1991; 35: 1008 –1010. 32. Fuqua WC, Winas SC, Greenberg EP. Census and consensus in bacte rial coosystems: the Lux R-Lux I family of quorum sensing transcriptional re gulators. An nu Re v Micro bio l 1996; 50: 727–751. 33. Gray KM. Interce llular c ommunication and group behaviour in bac teria. Tren ds Micro b io l 1997; 5: 184–188. 34. Greenberg EP. Quorum sensing in gram-negative bacteria. ASM New s 1997; 63: 371–377.

35. Kle ere bezem M, Quadri LE, Kuipers OP, de Vos VM. Quorum sensing by peptide pheromones and two-component signal-transduction syste ms in gram-positive bac teria. Mo l Micro bio l 1997; 24: 895–904. 36. Sute r S, Schad JM, Roux I. Granulocyte neutral prote ases and Ps e u do m o n a s e lastase as possible c auses of airw ay damage in patients w ith cystic fibrosis. J In fe ct Dis 1984; 149: 523– 531.

Accepted 18 May 1999


167