Effect of re-expansion after short-period lung collapse on pulmonary capillary permeability and pro-inflammatory cytokine gene expression in isolated rabbit lungs

T. Funakoshi1, Y. Ishibe*,1, N. Okazaki1, K. Miura2, R. Liu3, S. Nagai1 and Y. Minami1

1 Division of Anesthesiology and Critical Care Medicine, Department of Surgery, Tottori University Faculty of Medicine, 36-1 Nishi-cho, Yonago, Tottori, 683-8504, Japan. 2 Malaria Vaccine Development Unit, Laboratory of Parasitic Disease, National Institute of Allergy and Infectious Diseases, National Institutes of Health, Rockville, MD, USA. 3 Department of Anesthesia, University of Pennsylvania, Philadelphia, PA, USA

*Corresponding author. E-mail: ishibe@grape.med.tottori-u.ac.jp

Accepted for publication: December 5, 2003


    Abstract
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
Background. Re-expansion pulmonary oedema is a rare complication caused by rapid re-expansion of a chronically collapsed lung. Several cases of pulmonary oedema associated with one-lung ventilation (OLV) have been reported recently. Elevated levels of pro-inflammatory cytokines in pulmonary oedema fluid are suggested to play important roles in its development. Activation of cytokines after re-expansion of collapsed lung during OLV has not been thoroughly investigated. Here we investigated the effects of re-expansion of the collapsed lung on pulmonary oedema formation and pro-inflammatory cytokine expression.

Methods. Lungs isolated from female white Japanese rabbits were perfused and divided into a basal (BAS) group (n=7, baseline measurement alone), a control (CONT) group (n=9, ventilated without lung collapse for 120 min) and an atelectasis (ATEL) group (n=9, lung collapsed for 55 min followed by re-expansion and ventilation for 65 min). Pulmonary vascular resistance (PVR) and the coefficient of filtration (Kfc) were measured at baseline and 60 and 120 min. At the end of perfusion, bronchoalveolar lavage fluid/plasma protein ratio (B/P), wet/dry lung weight ratio (W/D) and mRNA expressions of tumour necrosis factor (TNF)-{alpha}, interleukin (IL)-1ß and myeloperoxidase (MPO) were determined.

Results. TNF-{alpha} and IL-1ß mRNA were significantly up-regulated in lungs of the ATEL group compared with BAS and CONT, though no significant differences were noted in PVR, Kfc, B/P and W/D within and between groups. MPO increased at 120 min in CONT and ATEL groups.

Conclusion. Pro-inflammatory cytokines were up-regulated upon re-expansion and ventilation after short-period lung collapse, though no changes were noted in pulmonary capillary permeability.

Br J Anaesth 2004; 92: 558–63

Keywords: gene, expression; lung, pulmonary capillary permeability; lung, re-expansion; polypeptides, inflammatory cytokine


    Introduction
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
One-lung ventilation (OLV) has become a common procedure for thoracoscopic surgery and several cases of pulmonary oedema associated with OLV have been reported.13 While the aetiology of re-expansion pulmonary oedema (REPE) after thoracocentesis for sustained lung collapse and pulmonary effusion have been extensively discussed,4 there is little information on the contribution of short-period lung collapse and re-inflation as seen in OLV to the development of REPE. A number of fluid mediators are thought to play important roles in chemotaxis and activation of polymorphonuclear leucocytes in the development of REPE.5 6

Previous animal studies from our laboratory7 demonstrated that 60-min lung collapse combined with perfusion arrest and 60-min follow-up after reperfusion and re-expansion-induced pulmonary oedema was accompanied by increased tumour necrosis factor (TNF)-{alpha} concentrations in the perfusate in rats. However, it is impossible to distinguish a specific aetiology from these studies; i.e. whether perfusion arrest or lung collapse contributes to the induction of pulmonary oedema.

In the present study, we investigated the effects of short-term lung collapse (continuing pulmonary circulation) and re-expansion by measuring pulmonary vascular resistance, pulmonary capillary permeability and especially messenger RNA (mRNA) expression of pro-inflammatory cytokines by using isolated perfused rabbit lungs.


    Materials and methods
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
Experimental setup
The Ethics Review Committee for Animal Experimentation of Tottori University Faculty of Medicine approved the experimental protocol. The isolated and perfused lung model of female white Japanese rabbits (body weight, 2–3 kg) was set up as described previously79 with minor modifications. Each rabbit was anaesthetized with pentobarbital sodium (25 mg kg–1 i.v.) and ketamine (30 mg kg–1 i.m.). Heparin was injected i.v. at a dose of 500 U kg–1. The rabbit was ventilated with room air (Servo ventilator 900B; Siemens-Elema, Solna, Sweden) via a tracheotomy tube at a rate of 40 min–1 with a tidal volume of 6 ml kg–1and positive end-expiratory pressure of 3 cm H2O until the pulmonary artery and the left atrium had been cannulated. The lungs were then ventilated at the settings described above with a 21% oxygen–5% carbon dioxide–74% nitrogen gas mixture. The left common carotid artery was cannulated and blood was drained and collected. The lungs were perfused through the tubes into the pulmonary artery and the left atrium using a pump (Harvard 681; Harvard Apparatus, Natick, MA, USA) at a constant flow rate (Q) of 40 ml min–1 kg–1 in a recirculating manner. The perfusate consisted of physiological salt solution (containing glucose 1 mg ml–1, bovine serum albumin 30 mg ml–1, insulin 0.2 mU ml–1, NaCl 119 mM, KCl 4.7 mM, MgSO4 1.17 mM, NaHCO3 22.61 mM, KH2PO4 1.18 mM and CaCl2 3.2 mM) and the autologous blood. Haematocrit (Celltac Counter ME-5158; Nihon Koden, Tokyo, Japan) and pH (iSTAT; iSTAT Co., East Windsor, NJ, USA) of the perfusate were adjusted to 15% and 7.40 (SD 0.08) respectively.

The isolated lungs were positioned on an electronic scale (model GX4000; A&D Co., Tokyo, Japan) placed in a box, to allow continuous monitoring of lung weight. After stabilization of perfusion and ventilation for 20 min, baseline pulmonary arterial pressure (Ppa), pulmonary venous pressure (Ppv) and coefficient of filtration (Kfc) were recorded. After baseline measurements, the lung preparations were divided at random into three groups. In the basal group (BAS; n=7), ventilation and perfusion were terminated without further intervention. In the control group (CONT; n=9), ventilation and perfusion were maintained for 120 min. In the atelectasis group (ATEL; n=9), the lungs were continuously perfused but completely collapsed for 55 min, followed by rapid re-inflation, and ventilation using the same conditions for 65 min. In the CONT and ATEL groups, the data were obtained at 60 and 120 min from baseline measurement. The perfusate samples were collected at 55 and 120 min. After termination of perfusion and ventilation, both lungs were used for further measurement.

Northern blot analysis
Experiments were performed in six rabbits in each of the three groups. Just after the end of perfusion, the lung tissue samples were collected from two locations in the right lung; the dorsal medial lobe and ventral superior lobe, and total RNA was obtained. Northern blot analysis was performed as described previously.10 11 In brief, total RNA 10 µg was separated on agarose-formaldehyde gel 1% and subsequently transferred and blotted onto INC nylon membranes (Westborough, MA, USA). Probes were prepared to purify products using the polymerase chain reaction with the primers listed in Table 1 for rabbit TNF-{alpha}, interleukin (IL)-1ß, transforming growth factor (TGF)-ß1 and the housekeeping gene glyceraldehyde-3-phosphate dehydrogenase (GAPDH).1214 Each probe was labelled with [32P]dCTP and gene expression levels were analysed with a laser densitometer (Molecular imager; Bio-Rad GS-525; Japan Bio-Rad-laboratories, Tokyo, Japan). The TNF-{alpha}, IL-1ß and TGF-ß1 densitometer data were corrected for GAPDH and then expressed using the sum total of three groups as 100%. For each group, data obtained from the dorsal middle and ventral superior lobes were combined and analysed statistically.


View this table:
[in this window]
[in a new window]
 
Table 1 PCR primers for probes used in Northern blot analysis
 
Distribution of vascular resistance
The longitudinal distribution of pulmonary vascular resistances was determined according to the following formulae, as described previously:7 8

Rpa=(Ppa–Pdo)Q–1

Rpv=(Pdo–Ppv)Q–1

Rt=(Ppa – Ppv)Q–1

where Rt=total vascular resistance, Rpa=arterial vascular resistance, Rpv=venous vascular resistance and Pdo=double occlusion pressure.

Pulmonary capillary filtration coefficient
Kfc was determined according to the method of Drake et al.15 The venous reservoir was lifted by 6 cm H2O and the change in lung weight was measured continuously with an electronic balance. The change in lung weight per unit time ({Delta}Wt{Delta}t–1) was calculated using data from 120 to 360 s and the initial change was determined as [({Delta}Wt{Delta}t–1)t=0]. {Delta}Pdo represented the difference in Pdo measured before and at 6 cm H2O pressure load. Kfc was calculated per 100 g of lung weight using the following formula:

Kfc=[({Delta}Wt{Delta}t–1)t=0]{Delta}Pdo–1 100 g–1 (ml min–1 mm Hg–1 100 g–1)

Wet to dry lung weight ratio (W/D)
The W/D of the right lung was measured as described previously.7 8 The ratio was calculated as W/D=wet weight/dry weight.

Collection of bronchoalveolar lavage fluid (BALF)
All rabbit lungs were used for analysis of BALF. Briefly, physiological saline solution (5 ml) at 4°C was injected into the left lung and BALF was subsequently collected. This procedure was repeated four times. The recovery rate of BALF was about 60%. Total protein concentrations in the BALF and perfusate were determined with BAC Protein Assay Reagent Kit (Pierce, Rockford, IL, USA) and BALF/plasma total protein ratio was calculated.

Myeloperoxidase (MPO) activity
MPO was measured in the perfusate and BALF at baseline and at 60 and 120 min using the method of o-dianisidine dihydrochloride oxidation.16

Statistical analysis
All parametric data are expressed as mean (SD). Effects of intervention on blood gas data, pulmonary vascular pressure and resistance, Kfc, airway pressure and MPO in the perfusate were analysed with Friedman’s test and subsequent post hoc Student–Newman–Keuls (Stat View 5.0; SAS Institute, NC, USA). Levels of mRNA expression, BALF/plasma total protein ratio, W/D and MPO concentration in BALF were compared using the Kruskal–Wallis H-test and subsequent post hoc Student–Newman–Keuls. A P value less than 0.05 denoted the presence of a statistically significant difference.


    Results
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
Blood gas and acid–base balance analysis
Although PaCO2 and pH were significantly different between CONT and ATEL, they were still within physiological limits (Table 2).


View this table:
[in this window]
[in a new window]
 
Table 2 Changes in blood gas data in CONT and ATEL groups. Data are mean (SD). 55 min=55 min after baseline measurement; 120 min=120 min after baseline measurement; BE=base excess. *P<0.05 vs ATEL group
 
Northern blot analysis
Typical imaging records of gene expression of mRNAs in the lung tissues are shown in Fig. 1A. Relative levels are shown in Fig. 1B. The expression of TNF-{alpha} and IL-1ß mRNAs was weak in the BAS group, whereas TNF-{alpha} and IL-1ß mRNA concentrations were amplified in the CONT group and were more pronounced in the ATEL group. There were significant differences in gene expression levels of TNF-{alpha} and IL-1ß but not TGF-ß1 between the three groups (Fig. 1B).



View larger version (37K):
[in this window]
[in a new window]
 
Fig 1 Expression of cytokines mRNAs. Six lung tissue samples were obtained from two areas of the right lung (dorsal medial and ventral superior lobes) and total RNA was collected. (A) Molecular imager data. The figure is representative of the 12 different samples. (B) Relative densities of TNF-{alpha}, IL-1ß and TGF-ß1 mRNA expression were analysed with the Kruskal–Wallis H-test for unpaired data followed by the Student–Newman–Keuls test as a post hoc analysis. *P<0.05 between groups.

 
Pulmonary vascular pressure, resistance and airway pressure
Ppa, Ppv and Pdo did not change throughout the study in the CONT and ATEL groups. Consequently, the calculated Rt, Rpa and Rpv also remained stable. Airway pressure also did not change throughout the study in the CONT and ATEL groups (Table 3).


View this table:
[in this window]
[in a new window]
 
Table 3 Changes in pulmonary haemodynamic and respiratory data in the CONT and ATEL groups. Data are mean (SD). There was no significant difference between groups for any variable. 60 min=60 min after baseline measurement; 120 min=120 min after baseline measurement; Ppa=pulmonary arterial pressure; Ppv=pulmonary venous pressure; Pdo=double occlusion pressure; Rt=total pulmonary resistance; Rpa=pulmonary arterial resistance; Rpv=pulmonary venous resistance; Airway P=airway pressure; Kfc=coefficient of filtration
 
Pulmonary vessel permeability
The pulmonary vascular filtration coefficient (Kfc) did not change throughout the study in the CONT and ATEL groups (Table 3). No significant differences in BALF/plasma total protein ratio were noted between the three groups (BAS, 0.88 (0.34); CONT, 0.76 (0.21); ATEL, 0.91 (0.44)). There were no differences in W/D between the three groups (BAS, 5.96 (0.11); CONT, 6.33 (0.72); ATEL, 6.78 (0.88)).

MPO activity
There were no differences in MPO activity in BALF between the three groups (BAS, 0.34 (0.24); CONT, 0.16 (0.04); ATEL, 0.27 (0.06) mol ml–1 min–1). MPO activity increased gradually but the change was similar in the CONT and ATEL groups (Table 4).


View this table:
[in this window]
[in a new window]
 
Table 4 Changes in MPO activity in CONT and ATEL groups. Data are mean (SD). *P<0.05 vs baseline. 55 min=55 min after baseline measurement; 120 min=120 min after baseline measurement; MPO=myeloperoxidase
 

    Discussion
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
The main finding of the present study was that lung collapse for 55 min followed by re-expansion resulted in up-regulation of mRNAs of the pro-inflammatory cytokines TNF-{alpha} and IL-1ß despite no findings of changes in pulmonary capillary permeability and vascular resistance.

Our model has two physiological characteristics. The first is that pulmonary perfusion flow rate is kept constant during lung collapse. If collapsed and non-collapsed areas coexist in the same lung, blood flow should shift from the collapsed to the non-collapsed areas and the collapsed lung should become relatively ischaemic, because of a difference in blood vessel resistance.17 In our model, however, perfusion of the collapsed lungs was maintained, and thus the effect of ischaemia induced by hypoperfusion was excluded. Blood supply to the lung consists of two circulatory systems: the pulmonary and bronchial circulations. In our model, bronchial circulation was completely sacrificed. We considered that this effect on the present study results may have been negligible, as the pulmonary circulation is apparently dominant compared with the bronchial circulation. In addition, the pulmonary circulation flow rate was kept constant by using a pump. The second feature of our experimental design was the short period of alveolar collapse (55 min) and follow-up for 60 min of ventilation after re-expansion. These periods were used in order to compare the data with those of previous studies,7 8 in which pulmonary oedema was observed after 60 min collapse and perfusion arrest followed by 60 min of perfusion and ventilation. The 60-min follow-up time was also based on the finding that 64% of re-expansion pulmonary oedema occurs within 1 h after re-expansion.4 OLV with lung collapse for 1 or 2 h is frequently used in thoracic surgery, such as lobectomy, in the clinical setting. Fifty five-minute lung collapse followed by re-expansion in the present study did not change the physiological indexes of pulmonary oedema (Kfc, W/D, BALF/plasma total protein ratio and PVR). These results are different from those reported by Liu et al.7 8 They demonstrate that short-period lung collapse combined with perfusion arrest resulted in increases in physiological indices of lung oedema.7 8 The discrepancy in the results of the two studies indicates that the effect of short-period lung collapse while maintaining pulmonary circulation is less invasive than that occurring when combined with circulatory arrest, and does not lead to serious deterioration of lung function.

However, it is interesting that our results clearly showed up-regulation of mRNAs of the pro-inflammatory cytokines TNF-{alpha} and IL-1ß only with 1-h collapse with perfusion and 1-h follow-up. In this study we focused on the expression of TNF-{alpha} and IL-1ß because these are typical early phase pro-inflammatory cytokines and are known to be up-regulated within the first 2 h. In addition, TNF-{alpha} and IL-1ß per se not only induce apoptosis and injure organs18 19 but also trigger cytokine networks and further activate additional cytokines, and attack cells.20 21 The present results suggest that TNF-{alpha} and IL-1ß may be involved in pulmonary oedema.13 Another inflammatory cytokine, IL-8, correlates with REPE,22 23 is up-regulated by TNF-{alpha} and IL-1ß 24 25 and shows sustained activity over several hours. As this study was designed to finish within 2 h, we did not measure IL-8. In contrast to TNF-{alpha} and IL-1ß, which were induced immediately after stimulation, TGF-ß1 was not induced in this study.26 We think that, because TGF-ß1 is a relatively late-phase cytokine,27 2 h is not enough time to induce TGF-ß1 in our study model.

MPO is an enzyme induced by activated neutrophils. Our results showed a 1.5-fold increase in MPO activity in the perfusate of the CONT and ATEL groups at the end of the study compared with baseline values. These findings suggest induction of neutrophil activation by surgical manipulation and subsequent perfusion rather than by lung collapse and re-expansion. On the other hand, there were no differences in BALF concentrations of MPO between the three groups, suggesting that neutrophils had not yet infiltrated into the alveolus.

In the clinical setting, OLV has become a common procedure for thoracoscopic surgery and in most cases it is performed without postoperative pulmonary complications. However, cases of REPE are reported after video-assisted thoracoscopic surgery.13 A possible mechanism is the vascular endothelial damage after reperfusion injury in the areas of previous hypoxic vasoconstriction, with the resulting reactive oxygen species altering the permeability of the vascular endothelium. Therefore, the duration and extent of exposure to hypoxia may play an important role in the development of REPE. In the present study, short-term lung collapse and re-expansion up-regulated the mRNAs of pro-inflammatory cytokines, which may increase cytokine activity and subsequent pulmonary capillary permeability, though lung oedema was not observed. As previously discussed, in the clinical setting blood flow to the collapsed lung decreases and the lungs become more ischaemic than in the present study. The duration of lung collapse in thoracoscopic surgery is usually longer than 60 min. Although there are limitations in extrapolation of the findings obtained from this ex vivo animal study, these considerations suggest that re-expansion of the lungs after OLV makes them susceptible to REPE.

In conclusion, we have demonstrated in the present study that short-period lung collapse and re-expansion up-regulated the pro-inflammatory cytokine expression despite the lack of physiological lung injury in perfused rabbit lungs.


    Acknowledgements
 
The authors are grateful to Professor K. Hirai and Associate Professor S. Fukumoto, Division of Molecular Medical Zoology, Department of Microbiology and Pathology, Faculty of Medicine, Tottori University, for excellent technical assistance and useful comments on the manuscript.


    References
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
1 Iqbal M, Multz AS, Rossoff LJ. Reexpansion pulmonary edema after VATS successfully treated with continuous positive airway pressure. Ann Thorac Surg 2000; 70: 669–71[Abstract/Free Full Text]

2 Wake M, Sanagawa Y, Okamoto Y. A case of anesthetic management for reexpansion pulmonary edema of the dependent lung saved by superimposed HFJV during one lung ventilation for the thoracoscopic operation associated with bilateral pneumothorax. Masui 2000; 49: 643–5[Medline]

3 Yanagidate F, Dohi S, Hamaya Y, Tsujito T. Reexpansion pulmonary edema after thoracoscopic mediastinal tumor resection. Anesth Analg 2001; 92: 1416–7[Free Full Text]

4 Tarver RD, Broderick LS, Conces DJ Jr. Reexpansion pulmonary edema. J Thorac Imaging 1996; 11: 198–209[ISI][Medline]

5 Nakamura H, Ishizaka A, Sawafuji M, et al. Elevated levels of interleukin-8 and leukotriene B4 in pulmonary edema fluid of a patient with reexpansion pulmonary edema. Am J Respir Crit Care Med 1994; 149: 1037–40[Abstract]

6 Sakao Y, Kajikawa O, Martin TR, et al. Association of IL-8 and MCP-1 with the development of reexpansion pulmonary edema in rabbits. Ann Thorac Surg 2001; 71: 1825–32[Abstract/Free Full Text]

7 Liu R, Ishibe Y, Ueda M. Isoflurane–sevoflurane administration before ischemia attenuates ischemia–reperfusion-induced injury in isolated rat lungs. Anesthesiology 2000; 92: 833–40[CrossRef][ISI][Medline]

8 Liu R, Ueda M, Ishibe Y, Hang Y. Isoflurane administration before ischemia and during reperfusion attenuates ischemia–reperfusion-induced injury of isolated rabbit lungs. Anesth Analg 1999; 89: 561–5[Abstract/Free Full Text]

9 Ishibe Y, Liu R, Ueda M, et al. Role of inhaled nitric oxide in ischaemia-reperfusion injury in the perfused rabbit lung. Br J Anaesth 1999; 83: 430–5[Abstract/Free Full Text]

10 Miura K, Fukumoto S, Dirgahayu P, Hirai K. Excretory/secretory products from plerocercoids of Spirometra erinaceieuropaei suppress gene expressions and production of tumor necrosis factor-alpha in murine macrophages stimulated with lipopolysaccharide or lipoteichoic acid. Int J Parasitol 2001; 31: 39–47[CrossRef][ISI][Medline]

11 Dirgahayu P, Fukumoto S, Miura K, Hirai K. Excretory/secretory products from plerocercoids of Spirometra erinaceieuropaei suppress the TNF-alpha gene expression by reducing phosphorylation of ERK1/2 and p38 MAPK in macrophages. Int J Parasitol 2002; 32: 1155–62[CrossRef][ISI][Medline]

12 Yamawaki H, Sato K, Hori M, et al. Impairment of EDR by a long-term PDGF treatment in organ-cultured rabbit mesenteric artery. Am J Physiol 1999; 277: H318–23

13 Hart DA, Boykiw R, Sciore P, Reno C. Complex alterations in gene expression occur in the knee ligaments of the skeletally mature multiparous rabbit during pregnancy. Biochim Biophys Acta 1998; 1397: 331–41[ISI][Medline]

14 Hellio Le Graverand MP, Reno C, Hart DA. Gene expression in menisci from the knees of skeletally immature and mature female rabbits. J Orthop Res 1999; 17: 738–44[ISI][Medline]

15 Drake R, Gaar KA, Taylor AE. Estimation of the filtration coefficient of pulmonary exchange vessels. Am J Physiol 1978; 234: H266–74[ISI][Medline]

16 Henson PM, Zanolari B, Schwartzman NA. Hong SR. Intracellular control of human neutrophil secretion. I. C5a-induced stimulus-specific desensitization and the effects of cytochalasin B. J Immunol 1978; 121: 851–5[Abstract]

17 Glasser SA, Domino KB, Lindgren L, et al. Pulmonary blood pressure and flow during atelectasis in the dog. Anesthesiology 1983; 58: 225–31[ISI][Medline]

18 Ashkenazi A, Dixit VM. Death receptors: signaling and modulation. Science 1998; 281: 1305–8[Abstract/Free Full Text]

19 Nagata S. Apoptosis by death factor. Cell 1997; 88: 355–65[ISI][Medline]

20 Okusawa S, Gelfand JA, Ikejima T, Connolly RJ, Dinarello CA. Interleukin 1 induces a shock-like state in rabbits. Synergism with tumor necrosis factor and the effect of cyclooxygenase inhibition. J Clin Invest 1988; 81: 1162–72[ISI][Medline]

21 Ward PA. Role of complement, chemokines, and regulatory cytokines in acute lung injury. Ann N Y Acad Sci 1996; 796: 104–12[Abstract]

22 Nakamura M, Fujishima S, Sawafuji M, et al. importance of interleukin-8 in the development of reexpansion lung injury in rabbits. Am J Respir Crit Care Med 2000; 161: 1030–6[Abstract/Free Full Text]

23 Sekido N, Mukaida N, Harada A, et al. Prevention of lung reperfusion injury in rabbits by a monoclonal antibody against interleukin-8. Nature 1993; 365: 654–7[CrossRef][ISI][Medline]

24 Fujishima S, Hoffman AR, Vu T, et al. Regulation of neutrophil interleukin 8 gene expression and protein secretion by LPS, TNF-alpha, and IL-1 beta. J Cell Physiol 1993; 154: 478–85[ISI][Medline]

25 Standiford TJ, Kunkel SL, Basha MA, et al. Interleukin-8 gene expression by a pulmonary epithelial cell line. A model for cytokine networks in the lung. J Clin Invest 1990; 86: 1945–53[ISI][Medline]

26 Jiang Q, Detolla L, Singh IS, et al. Exposure to febrile temperature upregulates expression of pyrogenic cytokines in endotoxin-challenged mice. Am J Physiol 1999; 276: R1653–6[ISI][Medline]

27 Xing Z, Jordana M, Kirpalani H, et al. Cytokine expression by neutrophils and macrophages in vivo: endotoxin induces tumor necrosis factor-alpha, macrophage inflammatory protein-2, interleukin-1 beta, and interleukin-6 but not rantes or transforming growth factor-beta 1 mRNA expression in acute lung inflammation. Am J Respir Cell Mol Biol 1994; 10: 148–53