- Email: [email protected]
Modulation of radiation-induced tumour necrosis factor α (TNF-α) expression in the lung tissue by pentoxifylline
Radiotherapy and Oncology 64 (2002) 177–187 www.elsevier.com/locate/radonline
Modulation of radiation-induced tumour necrosis factor a (TNF-a) expression in the lung tissue by pentoxifylline Claudia E. Ru¨be a, Falk Wilfert a, Daniela Uthe b, Kurt W. Schmid c, Reinhild Knoop b, Norman Willich b, Andreas Schuck b, Christian Ru¨be a,* a
Department of Radiation Oncology, Saarland University, D-66421 Homburg/Saar, Germany b Department of Radiation Oncology, University of Mu¨nster, D-48129 Mu¨nster, Germany c Department of Pathology, University of Essen, D-45122 Essen, Germany Received 3 August 2001; received in revised form 4 March 2002; accepted 8 April 2002
Abstract Purpose: The lung is the major dose-limiting organ for radiotherapy of cancer in the thoracic region. Immediate cellular damage after irradiation is supposed to result in cytokine-mediated multicellular interactions with induction and progression of inflammatory and fibrotic tissue reactions. Pentoxifylline (PTX) down-regulates the production of proinflammatory cytokines, particularly TNF-a, in response to noxious stimuli and may therefore provide protection against radiation-induced, cytokine-mediated cellular damage. The purpose of this study was to investigate the temporal and spatial release of TNF-a in the lung tissue after thoracic irradiation with 12 Gy. In addition, we evaluated the ability of PTX to reduce the radiation-induced TNF-a release in this animal model of thoracic irradiation. Materials and methods: C57BL/6J mice were exposed to either sham irradiation or single fraction of 12 Gy delivered to the thorax. Four study groups were defined: those that received neither irradiation nor PTX (NT group), those that received PTX but no irradiation (PTX group), those that underwent irradiation without PTX (XRT group) and those that received both PTX and irradiation (PTX/XRT group). Treated and sham-irradiated mice were sacrificed corresponding to the latent period and the pneumonic phase. The TNF-a mRNA expression in the lung tissue was quantified by ‘real-time’ quantitative reverse transcriptase polymerase chain reaction (RT-PCR). Immunohistochemical detection methods (alkaline phosphatase anti-alkaline phosphatase (APAAP)) and automated image analysis were used for objective quantification of TNF-a protein expression. Results: Following thoracic irradiation with a single dose of 12 Gy (XRT group), radiation-induced TNF-a mRNA release in the lung tissue was significantly increased during the acute phase of pneumonitis ðP , 0:05Þ. The elevated levels of TNF-a mRNA during the pneumonic phase correlate with a significant increase of positive inflammatory cells, predominantly macrophages, in the lung parenchyma ðP , 0:05Þ. In contrast to the radiation-only group (XRT-group), the lung tissue of the PTX-treated mice (PTX/XRT group) revealed only a minor radiation-mediated TNF-a response on mRNA and protein level. Conclusions: This study demonstrates a significant radiation-induced increase of TNF-a (on mRNA and protein level) in the lung tissue during the pneumonic phase. The predominant localisation of TNF-a in areas of inflammatory cell infiltrates suggests involvement of this cytokine in the pathogenesis of radiation-induced lung injury. In addition, we observed a pronounced reduction of the TNF-a mRNA and protein production in the study group that received both PTX and radiation (PTX/XRT group) as compared to the radiation-only group (XRT group). Therefore our results indicate that PTX down-regulates the TNF-a mRNA and protein production in the lung tissue in response to radiation. q 2002 Published by Elsevier Science Ireland Ltd. Keywords: TNF-a; Ionising radiation; Pneumonitis; Fibrosis; Pentoxifylline
1. Introduction The radiosensitivity of the lung tissue limits the dose of radiation which can be delivered to tumours in the thoracic region. Radiation-induced lung damage may arise depending on the total dose of radiation, the fractionation schedule, the volume of lung tissue irradiated, the existence of prior * Corresponding author.
lung disease and the use of chemotherapeutic drugs in the treatment of the disease [30,32,40]. Damage to endothelial or epithelial cells is assumed to be the initial step leading to radiation pneumonitis and ultimately to pulmonary fibrosis. But the process of injury and repair initiated by irradiation is also a function of activation of cells to produce important biological mediators, such as cytokines, which modulate diverse aspects of the inflammatory and fibrogenic response. Therefore, the pathophysiological tissue response after lung
0167-8140/02/$ - see front matter q 2002 Published by Elsevier Science Ireland Ltd. PII: S 0167-814 0(02)00077-4
178
C.E. Ru¨ be et al. / Radiotherapy and Oncology 64 (2002) 177–187
irradiation implies the induction of numerous cytokines which form the basis for the multicellular interactions of the inflammatory and fibrogenic process associated with radiation injury [11,12,21,22,34,35,42]. The relative role of cytokine dysregulation versus direct tissue injury from irradiation for the pathogenesis of radiation pneumonitis/ fibrosis remains elusive. TNF-a is thought to be a key mediator for the pathogenesis of radiation pneumonitis, because it shows the following spectrum of biological activities. TNF-a exerts in particular proinflammatory effects by inducing the expression of adhesion molecules that recruit leukocytes into the sites of tissue damage, by priming leukocytes for oxidant production, and by inducing production of prostaglandins and other mediators of inflammation. TNF-a inhibits anticoagulatory mechanisms and therefore promotes thrombotic processes [29]. In addition, TNF-a exerts fibrogenic effects by stimulating the growth of fibroblasts and increasing the collagen deposition. [19]. Therefore, a pharmacological regulation of the TNF-a production at the initial stage could possibly halt the progression of radiation-induced injury. A drug which suppresses the production of TNF-a but lacks the many side effects of glucocorticoids might be useful as an anti-inflammatory agent. Pentoxifylline (PTX) is a xanthine derivate that has generated widespread interest in the field of oncology based on its reported potential ability to ameliorate radiation- and chemotherapy-induced toxicity. PTX has been shown to enhance microvascular blood flow and decrease platelet aggregation, thereby maintaining perfusion in radiated tissues. In addition, PTX down-regulates the production of proinflammatory cytokines, particularly TNF-a, in response to noxious stimuli and inhibits granulocyte-mediated cytotoxicity after TNF-a exposure and may, therefore, provide protection against radiation-induced, cytokine-mediated cellular damage [2,8,38,44,45]. Since PTX has long been used in the treatment of peripheral vascular disease, the drug is widely available and low toxicity has been demonstrated. The purpose of this study was to investigate the temporal and spatial release of TNF-a in the lung tissue after thoracic irradiation with 12 Gy. In addition, we evaluated the ability of PTX to reduce the radiation-induced TNF-a release and the lung toxicity in this animal model of thoracic irradiation.
2. Methods and materials 2.1. Animals C57BL/6J mice were purchased from Charles River Laboratories. Adult female mice, 8 weeks old and approximately 20 g in weight, were housed four to six per cage and allowed to acclimatise from shipping for 1 week prior to treatment. Four study groups were defined: those that received neither radiation nor PTX (NT ¼ no treatment
group: eight animals), those that received PTX (500 mg/l in drinking water) but no irradiation (PTX group: eight animals), those that underwent irradiation without PTX (XRT group: 24 animals) and those that received both PTX and radiation (PTX/XRT group: 24 animals). Pentoxifylline (Sigma, St. Louis, MO, USA) was added to the drinking water of the mice in the PTX and PTX/XRT groups at a concentration of 500 mg/l. This concentration was chosen based on the measured initial water consumption, delivering a predicted dose of 100 mg PTX/kg bodyweight per day. Pentoxifylline administration was initiated 1 week before thoracic radiation. Animal weights and PTX consumption, the latter assessed by average water intake, were monitored during the study period. All mice underwent actual (XRT and PTX/XRT groups) or sham irradiation (NT and PTX groups). 2.2. Radiation schedule A dose of 12 Gy to the midplane of the lungs was delivered in a single fraction via a posterior field using a linear accelerator. A plastic jig was used to restrain the mice without anaesthesia and lead strips were placed to shield the head and abdomen. The irradiation characteristics were as follows: beam energy: 10 MV-photons; dose-rate: 2.4 Gy/ min; source surface distance (SSD): 1 m; size of the radiation field: 18 cm £ 10 cm. A source film was taken to confirm that the entire lung was irradiated. Film dosimetry was used to deduce the relative dose distribution. Dosimetry was performed with a cylindrical ionisation chamber. The depth of the maximum dose of the 10 MVphoton beam could be significantly reduced by the tissueequivalent plastic material (thickness 23 mm) of the restraining jig. Therefore the build-up region of the 10 MV-photon beam was located in the plastic material achieving an acceptable dose uniformity throughout the thorax of individual mice. Following irradiation the mice were maintained four to six per cage in laminar flow hoods in pathogen-free rooms to minimise pulmonary infections and supplied with standard laboratory diet and water ad libitum. Age-matched controls were maintained under identical conditions for the course of the experiment. Treated and control mice were sacrificed by cervical dislocation at times corresponding to the latent period (1, 24, 72 h and 1 week post-irradiation) and the pneumonic phase (2, 4, 8, 16 and 24 weeks post-irradiation). For each treatment modality (XRT and PTX/XRT group) and time point three animals were used, and groups of untreated (NT group) or PTX-treated (PTX group) shamirradiated animals served as experimental controls. The experimental protocol was approved by the Medical Sciences Animal Care and Use Committee of the University of Mu¨ nster. 2.3. Tissue isolation The complete lungs were immediately removed after death without being perfused. The left lobes were placed
C.E. Ru¨be et al. / Radiotherapy and Oncology 64 (2002) 177–187
in fixative for histological and histochemical analysis and the right lobes were quickly frozen in liquid nitrogen for RNA isolation and subsequent polymerase chain reaction (PCR) analyses. The lung tissue from three different mice (XRT and PTX/XRT group) for every time point was analysed by a combination of light microscopy, immunohistochemistry and PCR. 2.4. PCR analysis 2.4.1. RNA extraction Total RNA preparations were performed using the RNeasyeTotal RNA Kit (Qiagen, Hilden, Germany). The lung tissue (a right lobe is equivalent to 75 mg tissue) was homogenised in 1.2 ml lysis buffer using a rotor–stator homogeniser (Ultra-Turrax, IKA, Staufen, Germany). Tissue lysates were kept at 2858C prior to processing or were used directly in the procedure according to the manufacturer’s protocol. Ribonucleic acid concentrations were determined by spectrophotometric absorption at 260 nm. The integrity of the RNA was assessed by denaturing agarose gel electrophoresis. 2.4.2. First-strand complementary DNA (cDNA) synthesis Total RNA was converted to cDNA utilising a first-strand cDNA synthesis kit according to the manufacturer’s protocol (Pharmacia, Biotech, Freiburg, Germany). In this procedure oligo dT primers were used for generating first-strand cDNA in a final reaction mix of 15 ml. 2.4.3. Real-time quantitative reverse transcriptase PCR (RT-PCR) The TNF-a mRNA expression in the lung tissue was quantified by ‘real-time’ quantitative RT-PCR. Using a dual-labelled fluorogenic probe, this method allows direct detection of PCR-product accumulation by measuring the increase in fluorescent emission continuously during the exponential phase of the PCR reaction [16,33]. 2.4.3.1. The principle of the method employed in brief. The assay uses fluorescent Taqman methodology and an instrument for measuring fluorescence in real time (ABI Prism 77-00 Sequence Detector; Perkin Elmer/Applied Biosystems, Foster City, CA, USA). The Taqman reaction requires a hybridisation probe which is labelled with two different fluorescent dyes of which one is a reporter dye (FAM, i.e. 6-carboxyfluorescein) covalently attached at the 5 0 end and the other is a quenching dye (TAMRA, i.e. 6-carboxy-tetramethylrhodamine) covalently attached at the 3 0 end. When the probe is intact, fluorescent energy transfer occurs and the reporter dye (FAM) fluorescent emission is absorbed by the quenching dye (TAMRA). During the extension phase of the PCR cycle, the fluorescent hybridisation probe is cleaved by the 5 0 –3 0 nucleolytic activity of the DNA polymerase. On cleavage of the probe, the reporter dye emission is no longer transferred
179
efficiently to the quenching dye, resulting in an increase of the reporter dye fluorescent emission spectra. The use of the sequence detector (ABI Prism) allows measurement of fluorescent spectra of all 96 wells of the thermal cycler continuously during the PCR amplification. Therefore, the reactions are monitored in real time. 2.4.3.2. Primer and probe design. PCR primers and probes for the murine TNF-a and the housekeeping gene GAPDH were designed using the computer program Primer Express (Perkin Elmer/Applied Biosystems, Foster City, CA, USA). Primers were located in two different exons and either one of the primers was located on an intron–exon junction to prevent co-amplification of genomic DNA. The melting temperature (Tm) of the primers was 58–608C and the probe Tm was at least 108C higher than the primer Tm at approximately 698C. The amplicon lengths were kept very short (see Table 1). The fluorogenic probe contained a reporter dye (FAM) covalently attached at the 5 0 end and a quencher dye (TAMRA) covalently attached at the 3 0 end. Extension from the 3 0 end was blocked by attachment of a 3 0 phosphate group. Fluorogenic probes were highperformance liquid chromatography (HPLC) purified. 2.4.3.3. PCR amplification. PCR reactions were performed in the ABI-prism 7700 sequence detector, which includes a Gene-Amp PCR system 9600 (Perkin Elmer/Applied Biosystems, Foster City, CA, USA). PCR amplifications were performed in a total volume of 25 ml, containing 0.5 ml cDNA sample, 50 mM KCl, 10 mM Tris–HCl (pH 8.3), 60 nM Passive Reference 1 (Rox), 200 mM dATP, dCTP, dGTP and 400 mM dUTP, 5.5 mM MgCl2, 0.05% gelatine, 300 nM of each primer, 1.25 U AmpliTaq Gold and 0.25 U AmpErase Uracil NGlycosylase (TaqMan PCR core reagent kit; Perkin Elmer/Applied Biosystems, Foster City, CA, USA). Each reaction also contained 200 nM of the corresponding detection probe (Table 1). Each PCR amplification was performed in triplicate Table 1 Primer and probe sequences for murine TNF-a and GAPDH a Name
Sequence (5 0 ! 3 0 )
Amplicon length
5 0 -CCA GGC GGT GCC TAT GTC T-3 0 5 0 -GGC CAT TTG GGA ACT TCT CAT-3 0 123 bp 5 0 -AGC CTC TTC TCA TTC CTG GTT GTG GCA-3 0 GAPDH FW 5 0 -CAA CTC ACT CAA GAT TGT CAG CAA-3 0 GAPDH RV 5 0 -GGC ATG GAC TGT GGT CAT GA-3 0 118 bp GAPDH FP 5 0 -CAT CCT GCA CCA CCA ACT GCT TAG CC-3 0 TNF-a FW TNF-a RV TNF-a FP
a Forward (FW) and reverse (RV) primer were always located in different exons. Fluorogenic probes (FP) are FAM-labelled at the 5 0 end and TAMRA-labelled at the 3 0 end. All cDNA sequences were obtained from the Genbank database.
180
C.E. Ru¨ be et al. / Radiotherapy and Oncology 64 (2002) 177–187
wells, using the following conditions: 2 min at 508C and 10 min at 958C, followed by a total of 40 two-temperature cycles (15 s at 958C and 1 min at 608C). All values for the TNF-a mRNA expression were normalised to the constitutively expressed mRNA of the reference gene GAPDH. 2.5. Histology For histological analysis the left lobes were fixed in 10% neutral buffered formalin, paraffin-embedded, and sectioned at an average thickness of 5 mm. The mounted sections were stained by haematoxylin and eosin as well as Goldner’s stain for collagen. 2.6. Immunohistochemistry 2.6.1. Immunohistochemical staining After dewaxing of the sections in xylene and rehydration in graded alcohols, polyclonal rabbit anti-mouse TNF-a antibody (code: IP-400, Genzyme, Cambridge, MA, USA) was applied overnight in a humid chamber at 48C (dilutions in 0.01 M RPMI containing 0.6% bovine serum albumin: TNF-a antibody 1:100), followed by goat anti-rabbit secondary antibody (1:30 in RPMI; 30 min at room temperature; Dako, Glostrup; Denmark) and alkaline phosphatase anti-alkaline phosphatase (APAAP) complex (1:100 in phosphate buffered saline (PBS); 60 min at room temperature; Dako, Glostrup; Denmark) [5]. The secondary reactions were performed in a semi-automatic immunostaining device (Omnibus, Quartett, Berlin, Germany). Subsequently the enzyme reaction was developed for 25 min at room temperature in a freshly prepared fuchsin solution. Finally, the sections were counterstained with haematoxylin and mounted in Kayser’s glycerine gelatine. Since all lung tissue sections were treated in the same way and stained simultaneously under the same conditions, the intensity of immunostaining could be compared. Omission of primary antibodies as well as the same concentration of rabbit IgG (prepared in the same way) was used as a negative control. 2.6.2. Computerised image analysis TNF-a immunoreactivity was quantified using a computerised image analysis system (microscope/camera Axiocam; image analysis software KS 300, Zeiss, Go¨ ttingen, Germany). Magnification, light intensity and colour bar values remained constant throughout the experiment to allow valid comparisons among specimens. Three variables were measured using computerised image analysis: the number of cells per field which showed definite staining for TNF-a (expressed as positive cell counts), the relative area stained for TNF-a which is a measure of the fraction of cells expressing a detectable amount of TNF-a protein and the mean specific density of the stained area, which is proportional to the concentration of staining. All lung specimens were subjected to anonymous evaluation. To ensure objectivity, a standardised procedure was
established and used consistently throughout the study whereby ten representative, non-contiguous and non-overlapping fields ( £ 20 objective) were systematically selected and analysed (each field covered an area between 40 000– 50 000 mm 2). The sum of the measurements from the ten fields was totalled, and the arithmetic mean was calculated. Positive cell counts were expressed as the average number of cells per field. Mean values from lung tissues of radiated animals (XRT and PTX/XRT group) were compared with the corresponding values obtained from control animals at the respective time points. To assess reproducibility, each slide was analysed twice by an investigator who was unaware of the treatment group and the results of the previous measurement. 2.7. Statistical methods For statistical analysis, the assessment times 2, 4, 8, 16 and 24 weeks after completion of irradiation were summarised to the pneumonic phase. The results of the quantitative assessments of the TNF-a mRNA expression (in the whole lung tissue) as well as the TNF-a protein expression (immunoreactivity in the lung parenchyma, expressed as positive cell counts) were evaluated by Mann–Whitney-U-test to compare the XRT group with the NT group, the PTX group and the PTX/XRT treatment-group. The statistical calculations were performed by SPSS-software (version 9) for Windows NT. The criterion for statistical significance was P , 0:05. 3. Results 3.1. PCR analysis The results of the quantitative assessment of the TNF-a mRNA expression in the lung tissue of the mice treated according to the different study groups are demonstrated in Fig. 1. Non-irradiated lung tissue of control animals (NT and PTX group) exhibited low levels of TNF-a expression (relative mRNA expression: between 0.19 and 0.38) ( ¼ control). The NT and PTX groups showed no significant differences in the level of TNF-a expression. Following thoracic irradiation with a single dose of 12 Gy (XRT group), radiation-induced TNF-a release in the lung tissue was appreciable even within the first hour (relative mRNA expression: 1:34 ^ 0:13) and subsequently decreased to basal levels during the latent period (24 h, 72 h and 1 week post-irradiation). During the complete pneumonic phase, irradiation-mediated TNF-a release was significantly increased and reached maximal values at 8 weeks post-irradiation (relative mRNA expression: 4:05 ^ 1:46). After exposure to thoracic irradiation, the lung tissue of the PTX-treated mice (PTX/XRT group) revealed only a minor radiation-mediated TNF-a mRNA response during the pneumonic phase (relative mRNA expression: up to 1:57 ^ 0:11 at 24 weeks post-irradiation). In contrast to
C.E. Ru¨ be et al. / Radiotherapy and Oncology 64 (2002) 177–187
181
Fig. 1. Time course of the TNF-a mRNA expression in the lung tissue of the mice that underwent thoracic irradiation with 12 Gy ( ¼ XRT group: open diamonds; mean ^ SD) and those that received PTX and thoracic irradiation with 12 Gy ( ¼ PTX/XRT group: closed diamonds; mean ^ SD). Control: TNF-a mRNA expression in non-irradiated lung tissue of control animals (NT and PTX group).
the radiation-only group, the PTX-treated mice revealed a modest up-regulation of TNF-a during the latent phase. Statistical comparison of the TNF-a mRNA expression during the pneumonic phase (2, 4, 8, 16 and 24 weeks postirradiation) revealed significant difference between the XRT group and the NT/PTX-group ðP , 0:05Þ. Due to the data variability and the relatively low number of specimens available (n ¼ 15 per treatment-group), the statistical comparison of the TNF-a mRNA expression between the XRT and PTX/XRT treatment-group was not significant ðP ¼ 0:07Þ.
3.2. Immunohistochemistry Non-irradiated lung tissue (NT and PTX group) exhibited inhomogeneous and variable TNF-a immunoreactivity in the bronchiolar epithelium (Fig. 2). In the lung parenchyma, only a few alveolar macrophages stained positive. In the walls of untreated blood vessels TNF-a could not be detected immunohistochemically. The staining pattern in irradiated lung tissue (XRT and PTX/XRT group) differed from that in non-irradiated specimens in the following respect: as soon as 1 h after thoracic irradiation, the bronchiolar epithelium in the lung tissue of the irradiated mice revealed an intense and homogenous staining for TNF-a. At this time point, between 80 and 90% of the epithelial cells were intensely stained with the anti-TNF-a antibody. In contrast the bronchiolar epithelium revealed variable staining intensity for TNF-a at the later assessment time points, and no clear time-dependent alterations in the number of staining epithelial cells could be observed (between 10 and 80%) (Fig. 3). Interestingly, the subcellular distribution of TNF-a in bronchiolar epithelial
cells often showed a gradient with the highest concentration at the apical pole. Irradiation induces adherence of monocytes to the endothelium within hours, which is followed by monocytes/macrophages invasion of the vessel wall. In the sections analysed, most of the positive macrophages were located in the adventitia of the blood vessels (Fig. 4). The endothelial lining of blood vessels displayed TNF-a immunoreactivity as a function of time after radiation exposure. In the first hours after thoracic irradiation (3 and 6 h postirradiation), TNF-a antibody stained the endothelial cells (Fig. 5), whereas at the beginning of the pneumonic phase (2 and 4 weeks post-irradiation) only a very weak staining of the vascular endothelium was observed. During the later assessment time points of the pneumonic phase (8, 16 and
Fig. 2. Immunohistochemical staining of TNF-a in non-irradiated lung tissue (NT group): variable staining intensity for TNF-a in the bronchiolar epithelium. Magnification £ 200.
182
C.E. Ru¨ be et al. / Radiotherapy and Oncology 64 (2002) 177–187
Fig. 3. Immunohistochemical staining of TNF-a in irradiated lung tissue, 2 weeks after irradiation (XRT group): intense and homogenous staining for TNF-a in the bronchiolar epithelium. Magnification £ 200.
Fig. 5. Immunohistochemical staining of TNF-a in irradiated lung tissue, 6 h after irradiation (XRT group): positive staining of the vascular endothelium. Magnification £ 200.
24 weeks post-irradiation) the endothelium did not react with TNF-a antibody. At day 2 post-irradiation the medial smooth muscle cells were positive for TNF-a and remained positive (with variable staining intensity) throughout the complete observation period of 24 weeks. In spite of the constant anatomical position of the lung tissue samples examined, the histological sections revealed quite variable sections of bronchioli and blood vessels. Therefore, a direct comparison between the XRT and PTX/XRT group regarding the number of staining cells and the staining intensity was not possible. During the stage of interstitial pneumonitis, microscopic examination of the lung tissue revealed inflammation with accumulation of positive inflammatory cells particularly in perivascular and peribronchial areas, as well as in the subpleural region. The lung parenchyma revealed a pronounced increase of positive alveolar macrophages (Figs. 6 and 7). Other cell types of the lung parenchyma
such as type I and II pneumocytes or fibroblasts, did not stain positive for TNF-a. Increased cytokine expression was detected predominantly in regions of histopathologic radiation injury; inflammatory lesions in particular contained detectable levels of TNF-a. In contrast to this no significant immunoreactivity could be observed in areas of normal lung tissue. The results of the quantitative assessment of TNF-a immunoreactivity in the lung parenchyma by the computerised image analysis method are shown in Fig. 8. Nonirradiated lung parenchyma (NT and PTX group) exhibited low levels of TNF-a protein expression with positive cell counts between 10 and 23 ( ¼ control). In the first hours after radiation exposure, the lung parenchyma of the XRT group and the PTX/XRT group revealed no significant increase in the number of stained cells. During the stage of interstitial pneumonitis (2, 4, 8, 16 and 24 weeks postirradiation) TNF-a immunoreactivity in the lung parench-
Fig. 4. Immunohistochemical staining of TNF-a in irradiated lung tissue, 2 weeks after irradiation (XRT group): positive macrophages in perivascular region. Magnification £ 200.
Fig. 6. Immunohistochemical staining of TNF-a in irradiated lung tissue, 6 h after irradiation (XRT group): positive alveolar macrophages in the lung parenchyma. Magnification £ 400.
C.E. Ru¨ be et al. / Radiotherapy and Oncology 64 (2002) 177–187
Fig. 7. Immunohistochemical staining of TNF-a in irradiated lung tissue, 8 weeks after irradiation (XRT group): TNF-a immunoreactivity predominantly in subpleural regions. Magnification £ 200.
yma of the radiation-only group (XRT group) showed a pronounced increase with maximal values up to 80 positive cells (Fig. 8). Statistical comparison of the positive cell counts in the lung parenchyma during the pneumonic phase (2, 4, 8, 16 and 24 weeks post-irradiation) revealed significant difference between the XRT group and the NT/ PTX group ðP , 0:05Þ. The quantitative evaluation of the immunoreactivity in the lung parenchyma of the PTX/XRT group revealed lower positive cell counts with values between 20 and 35 (Fig. 8). Due to the data variability and the relatively low number of specimens available (n ¼ 15 per treatment-group), the statistical comparison between the XRT- and PTX/XRT treatment-group was not significant ðP ¼ 0:07Þ.
183
The wide variability of the positive cell counts in the analysed section areas of the lung parenchyma among the individual mice treated in the same way probably reflects the fact, that at earlier stages the lungs often show differences in the spatial and temporal distribution of inflammatory lesions. Therefore the positive cell counts (expressed as the average number of cells per field) revealed wide variation ranges depending on the fields selected and analysed. Three variables were measured using computerised image analysis: the number of cells which showed definite staining for TNF-a, the relative area stained for TNF-a and the mean specific density of the stained area, which is proportional to the concentration of the stain. While there was a clear increase in the number of stained cells with a maximum at the beginning of the pneumonic phase, no remarkable changes in the specific optical density were observed. Therefore, after irradiation it is not the cytokine production of each cell that changes, but rather the number of cells involved in TNF-a production. These data suggest that with regard to the lung parenchyma, changes in cytokine expression are caused mainly by an increase in the fraction of cells expressing TNF-a, but not by changes in the level of TNF-a expression per cell.
4. Discussion 4.1. ‘Real-time’ quantitative RT-PCR RT-PCR is a powerful method used to quantify the mRNA expression of cytokines, which are often expressed at very low levels. Various methods of (semi-) quantitative RT-PCR have been described for measuring cytokine
Fig. 8. Time course of the TNF-a protein expression (expressed as positive cell counts) in the lung tissue of the mice that underwent thoracic irradiation with 12 Gy ( ¼ XRT group: open diamonds; mean ^ SD) and those that received PTX and thoracic irradiation with 12 Gy ( ¼ PTX/XRT group: closed diamonds; mean ^ SD). Control: TNF-a protein expression (positive cell counts of the lung parenchyma) in non-irradiated lung tissue of control animals (NT and PTX group).
184
C.E. Ru¨ be et al. / Radiotherapy and Oncology 64 (2002) 177–187
mRNA levels. One classical method measures PCR-product accumulation during the exponential phase of the reaction [36]. To obtain semi-quantitative results it is extremely important that the PCR product is measured during the log phase of the reaction, before saturation is reached. This is the major difficulty in using this method, because it necessitates interruption of the PCR reaction after an experimentally determined number of cycles. A drawback is the relatively small linear range, during which samples within a single experimental setup can be analysed. Alternatively, quantitative competitive PCR can be performed [37,39] which requires co-amplification of an internal cDNA or RNA control (competitor) with the unknown sample in the same tube. The amount of mRNA is quantified by titration of an unknown amount of target template against a dilution series of known amounts of the standard. Thus, the same primers are used to co-amplify the target and the competitor, in such a way that they amplify with the same efficiency, although they can be distinguished from each other (by differences in length or restriction sites). Although this method provides a strategy for accurate quantification, the construction of internal standards is technically sophisticated and labour intensive. For the detection of PCR products using either of these methods, several detection techniques can be used, all requiring extensive post-PCR procedures. The most commonly used are agarose gel electrophoresis and EtBr-staining, fluorescent labelling and analysis using polyacrylamide gels, radioactive labelling and Southern blotting or detection by phosphor-imaging. Major drawbacks using these classical detection systems are the use of hazardous chemicals and the potential risk for laboratory contamination. Moreover, all these post-PCR procedures are very time-consuming. In the present study we used the recently developed ‘realtime’ quantitative RT-PCR, a quick and accurate technique for measuring PCR-product accumulation during the exponential phase of the reaction [16,33]. As this method combines PCR amplification and product detection in one single step, the technique is very fast and easy to perform as compared to classical RT-PCR techniques. This method also provides accurate measurements over a wide range of template quantities. Furthermore, the use of an internal probe ( ¼ fluorogenic probe), guarantees the specificity of the PCR product to be measured. 4.2. Immunohistochemistry and quantitative image analysis The immunohistochemical methods for detection of TNF-a used in this study are well established [5], and the specificity of the antibody used has been documented previously [12]. Previous in vivo studies of cytokine protein expression in pathological conditions have largely relied on qualitative assessment of immunoreactivity. The computerassisted image processing and analysis method used in the present study enabled us to assess the quantitative TNF-a immunoreactivity of the lung parenchyma. The image
analysis method yielded highly reproducible results and provided objective, quantitative and comparative assessment of TNF-a protein expression levels. The method thus allowed reliable quantitative estimations of the TNFa immunoreactivity in the lung parenchyma of the different study groups during the different phases of injury. The present work is a detailed study on the temporal release of TNF-a in the lung tissue after thoracic irradiation. Moreover, the cellular source of TNF-a as well as the association between TNF-a immunoreactivity and specific histopathologic alterations could be demonstrated. Secondly, we evaluated the ability of PTX to reduce the radiation-induced TNF-a release on mRNA and protein level using a well-validated in vivo assay of pulmonary injury. Published data [11,12,21,22,34,35,42] and previous studies from our laboratory [37; further data in preparation] suggest a major role of several inflammatory and fibrogenic cytokines in the pathogenesis of radiation-induced lung damage. Of these cytokines studied, TNF-a appears to play a particularly prominent role in the acute phase of radiation-induced damage since it is consistently overexpressed in areas of histopathologic lesions during the course of radiation pneumonitis. Even 1 h after thoracic irradiation, the bronchiolar epithelium in the lung tissue of the irradiated mice revealed an intense and homogenous staining for TNF-a. Similar TNFa mRNA and protein expression pattern by the bronchiolar– alveolar epithelium immediately after asbestos exposure could be observed in a murine model of asbestos-induced lung fibrosis [4]. The normal tracheobronchial epithelium, which has a relatively slow turn-over, may require only low levels of growth factors. However, under conditions of increased growth such as during repair after cell damage, an increase in the synthesis and/or activation of TNF-a may be involved to stimulate cell proliferation and differentiation. Pulmonary epithelial hyperplasia observed in the late stage of various inflammatory lung diseases, is associated with prolonged overexpression of TNF-a [20]. Looking at the cytokine network in radiation-induced lung injury, TNFa which is initially overexpressed in pulmonary epithelial cells, probably not only stimulates epithelial cell proliferation, but also initiates the recruitment of macrophages and other inflammatory cells. Subsequently, TNF-a was overexpressed in the lung parenchyma particularly by pulmonary macrophages. TNF-a enhances the production of different chemokines which are thought to further promote the inflammatory process particularly by monocyte/macrophage recruitment into the sites of tissue damage. Increased cytokine expression was detected prominently in inflammatory lesions, particularly in perivascular, peribronchial and subpleural areas. Other cell types of the lung parenchyma like type I and II pneumocytes or fibroblasts did not stain positive for TNF-a. In the present study, the endothelial lining of blood vessels displayed TNF-a immunoreactivity as a function
C.E. Ru¨ be et al. / Radiotherapy and Oncology 64 (2002) 177–187
of time after radiation exposure. TNF-a induces the expression of cell surface adhesion molecules such as the intercellular adhesion molecule (ICAM-1) and the endothelial cell adhesion molecule (ELAM-1) [3]. Expression of these endothelial cell adhesion molecules is a critical factor in the binding of leukocytes to the vessel wall at the site of inflammation. Furthermore, TNF-a increases the endothelial cell permeability and induces the release of other inflammatory mediators, which may contribute to the increased permeability [29]. From day 2 post-irradiation onwards, the medial smooth muscle cells were positive for TNF-a and remained positive (with variable staining intensity) throughout the complete observation period of 24 weeks. These results correspond to an immunohistochemical study of radiation-induced damage in major blood vessels where a transient dose-dependent TNF-a induction in medial smooth muscle cells after combined intraoperative and external irradiation could be demonstrated [23]. A recent study also showed, that TNF-a is the main factor secreted by injured (but not normal) myocytes which is responsible for induction of additional inflammatory mediators, namely syndecans, in surrounding endothelial cells [46]. Syndecans participate in a variety of biological functions including regulation of blood coagulation, cell adhesion as well as regulation of endothelial cell growth. Therefore, this mechanism may play an important role in myocyte-dependent regulation of endothelial cell activity and the ability of injured tissues to initiate and propagate an angiogenic response. There is currently great interest in the potential value of PTX as an inhibitor of normal tissue injury caused by radiation or chemotherapy. This is due to promising early laboratory and clinical investigations, as well as the drug’s widespread availability and low toxicity profile. In an animal model of brain irradiation, PTX treatment of mice partially suppressed radiation-induced TNF-a expression in cerebral tissue [18]. In pilot human studies, PTX has shown apparent efficacy in promoting healing of chronic radiation muco-cutaneous complications that had previously been unresponsive to conservative management [6,7,13,14]. With respect to the amelioration of radiation-induced pulmonary injury, several observations suggested that PTX might be a useful therapeutic agent. PTX has been reported to significantly reduce radiation-induced pulmonary injury in rats [31]. Radiation pneumonitis is similar in histopathologic appearance to the adult respiratory distress syndrome (ARDS), which may result from a variety of insults such as sepsis, shock or drugs. In a guinea pig model of ARDS, PTX inhibited lung injury induced by TNF-a and Escherichia coli sepsis [15,27]. In various clinical studies, PTX therapy inhibited TNF-a secretion and improved the symptoms related to ARDS with no particular toxic effects [1,28]. In experimental studies of bleomycininduced alveolitis, PTX ameliorated the extent of the inflammatory reaction in the lung tissue [9,10]. Promising results from a clinical study suggest that PTX may improve
185
therapeutic regimens in pulmonary sarcoidosis either by sparing or replacing corticosteroids [43]. Despite its early promise, the ability of PTX to ameliorate radiation-induced injury of normal tissue requires further validation, before the drug can be empirically used on a widespread clinical basis. One intention of the present study was to examine the effects of PTX on the severity of the radiation-induced lung damage. Therefore, the qualitative and quantitative histopathologic alterations of the radiation-induced lung injury were documented (for each individual animal) as follows: after examining the whole lung section and estimating the extent of altered lung parenchyma, the different lesions referring to the respiratory parenchyma, the bronchial and vascular structures and the pleura were documented. The estimates of the overall lung damage were based on the quantitative evaluation of the altered lung parenchyma using computerised image analysis. During the pneumonic phase, the area of altered lung tissue varied from approximately 20% (2 weeks post-irradiation) to 80% (24 weeks post-irradiation). While the effect of PTX on the infiltration of inflammatory cells could be quantified, the other histopathologic alterations like intraalveolar oedema, exudation into air spaces, desquamation of epithelial cells, alterations of the capillaries, thickening of the alveolar septa etc. could only be qualitatively documented. In summary, while we observed a reduction of the inflammatory cell infiltration in the PTX/XRT group as compared to the XRT group, the overall lung damage and the other histopathologic alterations which contribute to the pleomorphic picture of radiation pneumonitis revealed no apparent difference between the two treatment modalities. Other experimental setups including radionuclide lung perfusion scans for functional assessment are possibly more suitable to study the effect of PTX on radiation-related lung injury. However, our data indicate that PTX reduces the TNF-a release on mRNA and protein level as measured by ‘real-time’ RT-PCR and immunohistochemistry. Moreover, the results of PCR analysis and immunohistochemical staining of corresponding lung tissue reveal significant correlation between mRNA expression and immunoreactivity. In the present study we analysed TNF-a on mRNA level by ‘real-time’ RT-PCR and on protein level by immunohistochemical staining at different assessment time points in individual animals; concomitantly the association between cytokine expression and localisation and specific histopathologic alterations were assessed. With limited amounts of RNA template required, ‘real-time’ RT-PCR affords the opportunity to measure transcript levels of different cytokines in the lung tissue, which can be subsequently further analysed by immunohistochemical staining and histological analysis. This strategy facilitates comprehensive analysis of the lung tissue within a single individual mouse and thus allows efficient correlation of the mRNA and protein expression of different cytokines and their cellular location with specific histopathologic alterations. There is no current evidence to suggest that PTX protects
186
C.E. Ru¨ be et al. / Radiotherapy and Oncology 64 (2002) 177–187
tumour cells from the intended effects of ionising radiation. In fact, other investigators have observed that PTX may enhance tumour radiosensitivity, primarily based on its reported ability to increase oxygenation by improving perfusion in otherwise hypoxic tumours [17,24,25,26,41].
[5]
[6]
5. Conclusions This study demonstrates a significant radiation-induced increase of TNF-a (on mRNA and protein level) in the lung tissue during the pneumonic phase. The predominant localisation of TNF-a in areas of inflammatory cell infiltrates suggests involvement of this cytokine in the pathogenesis of radiation-induced lung injury. In addition, we observed a pronounced (but statistically not significant) reduction of the TNF-a mRNA and protein production in the study group that received both PTX and radiation (PTX/ XRT group) as compared to the radiation-only group (XRT group). Therefore our results indicate that PTX down-regulates the TNF-a mRNA and protein production in the lung tissue in response to radiation. Inhibition of appropriate cytokines such as TNF-a at an early stage of pneumonitis is thought to be important for effective treatment of radiation-induced lung injury. Pharmacological regulation of the TNF-a production may thus ameliorate the severe consequences of lung irradiation. PTX has been shown to improve therapeutic strategies in numerous diseases in which TNF-a represents a causative pathophysiological factor. PTX pharmacologically controls the synthesis and the actions of TNF-a, and may therefore provide protection against radiation-induced, cytokinemediated cellular damage.
[7]
[8]
[9]
[10]
[11]
[12]
[13]
[14]
[15]
[16]
Acknowledgements This investigation was supported by grant no. FI 733/1/1 from the Deutsche Forschungsgemeinschaft (DFG) and by grant no. Fi-1-4-I/98-22; Fi-1-4-II/97-17; Fi-1-1-II/96-29 from the Innovative Medizinische Forschung (IMF) of the University of Mu¨ nster. The authors thank Dr Heinecke for the statistical analysis of these data for publication.
[17]
[18]
[19] [20] [21]
References [22] [1] Ardizzoia A, Lissoni P, Tancini G, et al. Respiratory distress syndrome in patients with advanced cancer treated with pentoxifylline: a randomized study. Support Care Cancer 1993;1:331–333. [2] Balibrea JL, Arias-Diaz J, Garcia C, et al. Effect of pentoxifylline and somatostatin on tumour necrosis factor production by human pulmonary macrophages. Circ Shock 1994;43:51–56. [3] Beck-Schimmer B, Schimmer RC, Warner RL, et al. Expression of lung vascular and airway ICAM-1 after exposure to bacterial lipopolysaccharide. Am J Respir Cell Mol Biol 1997;17:344–352. [4] Brass DM, Hoyle GW, Poovey HG, et al. Reduced tumor necrosis factor-alpha and transforming growth factor-beta 1 expression in the
[23]
[24]
[25]
lungs of inbred mice that fail to develop fibroproliferative lesions consequent to asbestos exposure. Am J Pathol 1999;154:853–862. Cordell JL, Falini B, Erber WN, et al. Immunoenzymatic labeling of monoclonal antibodies using immune complexes of alkaline phosphatase and monoclonal anti-alkaline phosphatase (APAAP complexes). J Histochem Cytochem 1984;32:219–229. Delanian S, Balla-Mekias S, Lefaix JL. Striking regression of chronic radiotherapy damage in a clinical trial of combined pentoxifylline and tocopherol. J Clin Oncol 1999;17:3283–3290. Delanian S. Striking regression of radiation-induced fibrosis by a combination of pentoxifylline and tocopherol. Br J Radiol 1998;71:892–894. Doherty GM, Jensen JC, Alexander HR, et al. Pentoxifylline suppression of tumor necrosis factor gene transcription. Surgery 1991;110:192–198. Entzian P, Gerlach C, Gerdes J, et al. Pentoxifyline inhibits experimental bleomycin-induced fibrosing alveolitis. Pneumologie 1997;51:375–380. Entzian P, Zahringer U, Schlaak M, et al. Comparative study on effects of pentoxifylline, prednisolone and colchicine in experimental alveolitis. Int J Immunopharmacol 1998;20:723–735. Finkelstein JN, Johnston CJ, Baggs R, et al. Early alterations in extracellular matrix and transforming growth factor b gene expression in mouse lung indicative of late radiation fibrosis. Int J Radiat Oncol Biol Phys 1994;28:621–631. Franko AJ, Sharplin J, Ghahary A, et al. Immunohistochemical localization of transforming growth factor b and tumor necrosis factor a in the lungs of fibrosis-prone and ‘non-fibrosing’ mice during the latent period and early phase after irradiation. Radiat Res 1997;147:245–256. Futran ND, Trotti A, Gwede C. Pentoxifylline in the treatment of radiation-related soft tissue injury: preliminary observations. Laryngoscope 1997;107:391–395. Gottlober P, Krahn G, Korting HC, et al. The treatment of cutaneous radiation-induced fibrosis with pentoxifylline and vitamin E. Strahlenther Onkol 1996;172:34–38. Harada H, Ishizaka A, Yonemaru M, et al. The effects of aminophylline and pentoxifylline on multiple organ damage after Escherichia coli sepsis. Am Rev Respir Dis 1989;140:974–980. Heid CA, Stevens J, Livak KJ, et al. Real time quantitative PCR. Genome Methods 1996;6:986–994. Honess DJ, Dennis IF, Bleehen NM. Pentoxifylline: its pharmakokinetics and ability to improve tumour perfusion and radiosensitivity in mice. Radiother Oncol 1993;28:208–218. Hong J-H, Chiang C-S, Campbell IL, et al. Induction of acute phase gene expression by brain irradiation. Int J Radiat Oncol Biol Phys 1995;33:619–626. Ibelgaufts H. TNF-a: tumor necrosis factor alpha, Dictionary of cytokines. Weinheim: VCH Verlagsgesellschaft mbH, 1995 p. 709–17. Jetten AM. Growth and differentiation factors in tracheobronchial epithelium. Am J Physiol 1991;260:L361–L373. Johnston CJ, Piedboeuf B, Baggs R, et al. Differences in correlation of mRNA gene expression in mice sensitive and resistant to radiationinduced pulmonary fibrosis. Radiat Res 1995;142:197–203. Johnston CJ, Piedboeuf B, Rubin F, et al. Early and persistant alterations in the expression of interleukin-1a, interleukin-1b and tumor necrosis factor a mRNA levels in fibrosis-resistant and sensitive mice after thoracic irradiation. Radiat Res 1996;145:762–767. Kallfass E, Kra¨ mling HJ, Schultz-Hector S. Early inflammatory reaction of the rabbit coeliac artery wall after combined intraoperative (IORT) and external (ERT) irradiation. Radiother Oncol 1996;39:167–178. Kim SH, Khil MS, Ryu S, et al. Enhancement of radiation response on human carcinoma cells in culture by pentoxifylline. Int J Radiat Oncol Biol Phys 1993;25:61–65. Kwon H-C, Kim S-K, Chung W-K, et al. Effect of pentoxifylline on
C.E. Ru¨ be et al. / Radiotherapy and Oncology 64 (2002) 177–187
[26]
[27]
[28]
[29]
[30]
[31]
[32] [33]
[34] [35]
radiation response of non-small cell lung cancer: a phase III randomized multicenter trial. Radiother Oncol 2000;56:175–179. Lee I, Levitt SH, Song CW. Improved tumour oxygenation and radiosensitization by combination with nicotinamide and pentoxifylline. Int J Radiat Biol 1993;64:237–244. Lilly CM, Sandhu JS, Ishizaka A, et al. Pentoxifylline prevents tumor necrosis factor-induced lung injury. Am Rev Respir Dis 1989;139: 1361–1368. Lissoni F, Ardizzoia A, Perego MS, et al. Inhibition of tumor necrosis factor-alpha secretion by pentoxifylline in advanced cancer patients with abnormally high blood levels of tumor necrosis factor-alpha. J Biol Regul Homeost Agents 1993;7:73–75. Litwin MS, Gamble JR, Vadas MA. Role of cytokines in endothelial cell function. In: Aggarwal BB, Pun RK, editors. Human cytokines: their role in disease and therapy, Oxford, England: Blackwell Science, 1995. pp. 101–129. McDonald S, Rubin P, Phillips TL, et al. Injury to the lung from cancer therapy: clinical syndromes, measurable endpoints, and potential scoring systems. Int J Radiat Oncol Biol Phys 1995;31:1187– 1203. Molteni A, Moulder JE, Cohen EF, et al. Control of radiation-induced pneumopathy and lung fibrosis by angiotensin-converting enzyme inhibitors and an angiotensin II type 1 receptor blocker. Int J Radiat Biol 2000;76:523–532. Movsas B, Raffin TA, Epstein AH, et al. Pulmonary radiation injury. Chest 1997;111:1061–1076. Overbergh L, Valckx D, Waer M, et al. Quantification of murine cytokine mRNAs using real time quantitative reverse transcriptase PCR. Cytokine 1999;11:305–312. Rodemann HP, Bamberg M. Cellular basis of radiation-induced fibrosis. Radiother Oncol 1995;35:83–90. Rubin P, Johnston CJ, Williams JP, et al. A perpetual cascade of cytokines postirradiation leads to pulmonary fibrosis. Int J Radiat Oncol Biol Phys 1995;33:99–109.
187
[36] Ru¨ be C, Finke C, van Valen F, et al. Ionizing radiation induces production of tumor necrosis factor a in the Ewing’s sarcoma cell line RM 82 in vitro and in vivo. Strahlenther Onkol 1997;173:407– 414. [37] Ru¨ be CE, Uthe D, Schmid KW, et al. Dose-dependent induction of transforming growth factor b (TGF-b) in the lung tissue of fibrosisprone mice after thoracic irradiation. Int J Radiat Oncol Biol Phys 2000;47:1033–1042. [38] Samlaska CP, Winfield EA. Pentoxifylline. J Am Acad Dermatol 1994;30:603–621. [39] Siebert PD, Larrick JW. PCR MIMICS: competitive DNA fragments for use as internal standards in quantitative PCR. Biotechniques 1993;14:244–249. [40] Travis EL, Komaki R. Treatment-related lung damage. In: Pass HI, Mitchell JB, Johnson DH, et al., editors. Lung cancer: principles and practice, Philadelphia, PA: Lippincott-Raven Publishers, 1996. pp. 285–301. [41] Vernimmen F, Verheye-Dua F, du-Toit H, et al. Effect of pentoxifylline on radiation damage and tumor growth. Strahlenther Onkol 1994;170:595–601. [42] Yi ES, Bedoya A, Lee H, et al. Radiation-induced lung injury in vivo: expression of transforming growth factor-beta precedes fibrosis. Inflammation 1996;20:339–351. [43] Zabel P, Entzian F, Dalhoff K, et al. Pentoxifylline in treatment of sarcoidosis. Am J Respir Crit Care Med 1997;155:1665–1669. [44] Zabel P, Schade FU, Schlaak M. Inhibition of endogenous TNF formation by pentoxifylline. Immunobiology 1993;187:447–463. [45] Zabel P, Schade FU, Schlaak M. Pentoxifylline – an inhibitor of the synthesis of tumor necrosis factor alpha. Immun Infekt 1992;20:80– 83. [46] Zhang Y, Pasparakis M, Kollias G, et al. Myocyte-dependent regulation of endothelial cell syndecan-4 expression. Biol Chem 1999;274:14786–14790.
Modulation of radiation-induced tumour necrosis factor a (TNF-a) expression in the lung tissue by pentoxifylline Claudia E. Ru¨be a, Falk Wilfert a, Daniela Uthe b, Kurt W. Schmid c, Reinhild Knoop b, Norman Willich b, Andreas Schuck b, Christian Ru¨be a,* a
Department of Radiation Oncology, Saarland University, D-66421 Homburg/Saar, Germany b Department of Radiation Oncology, University of Mu¨nster, D-48129 Mu¨nster, Germany c Department of Pathology, University of Essen, D-45122 Essen, Germany Received 3 August 2001; received in revised form 4 March 2002; accepted 8 April 2002
Abstract Purpose: The lung is the major dose-limiting organ for radiotherapy of cancer in the thoracic region. Immediate cellular damage after irradiation is supposed to result in cytokine-mediated multicellular interactions with induction and progression of inflammatory and fibrotic tissue reactions. Pentoxifylline (PTX) down-regulates the production of proinflammatory cytokines, particularly TNF-a, in response to noxious stimuli and may therefore provide protection against radiation-induced, cytokine-mediated cellular damage. The purpose of this study was to investigate the temporal and spatial release of TNF-a in the lung tissue after thoracic irradiation with 12 Gy. In addition, we evaluated the ability of PTX to reduce the radiation-induced TNF-a release in this animal model of thoracic irradiation. Materials and methods: C57BL/6J mice were exposed to either sham irradiation or single fraction of 12 Gy delivered to the thorax. Four study groups were defined: those that received neither irradiation nor PTX (NT group), those that received PTX but no irradiation (PTX group), those that underwent irradiation without PTX (XRT group) and those that received both PTX and irradiation (PTX/XRT group). Treated and sham-irradiated mice were sacrificed corresponding to the latent period and the pneumonic phase. The TNF-a mRNA expression in the lung tissue was quantified by ‘real-time’ quantitative reverse transcriptase polymerase chain reaction (RT-PCR). Immunohistochemical detection methods (alkaline phosphatase anti-alkaline phosphatase (APAAP)) and automated image analysis were used for objective quantification of TNF-a protein expression. Results: Following thoracic irradiation with a single dose of 12 Gy (XRT group), radiation-induced TNF-a mRNA release in the lung tissue was significantly increased during the acute phase of pneumonitis ðP , 0:05Þ. The elevated levels of TNF-a mRNA during the pneumonic phase correlate with a significant increase of positive inflammatory cells, predominantly macrophages, in the lung parenchyma ðP , 0:05Þ. In contrast to the radiation-only group (XRT-group), the lung tissue of the PTX-treated mice (PTX/XRT group) revealed only a minor radiation-mediated TNF-a response on mRNA and protein level. Conclusions: This study demonstrates a significant radiation-induced increase of TNF-a (on mRNA and protein level) in the lung tissue during the pneumonic phase. The predominant localisation of TNF-a in areas of inflammatory cell infiltrates suggests involvement of this cytokine in the pathogenesis of radiation-induced lung injury. In addition, we observed a pronounced reduction of the TNF-a mRNA and protein production in the study group that received both PTX and radiation (PTX/XRT group) as compared to the radiation-only group (XRT group). Therefore our results indicate that PTX down-regulates the TNF-a mRNA and protein production in the lung tissue in response to radiation. q 2002 Published by Elsevier Science Ireland Ltd. Keywords: TNF-a; Ionising radiation; Pneumonitis; Fibrosis; Pentoxifylline
1. Introduction The radiosensitivity of the lung tissue limits the dose of radiation which can be delivered to tumours in the thoracic region. Radiation-induced lung damage may arise depending on the total dose of radiation, the fractionation schedule, the volume of lung tissue irradiated, the existence of prior * Corresponding author.
lung disease and the use of chemotherapeutic drugs in the treatment of the disease [30,32,40]. Damage to endothelial or epithelial cells is assumed to be the initial step leading to radiation pneumonitis and ultimately to pulmonary fibrosis. But the process of injury and repair initiated by irradiation is also a function of activation of cells to produce important biological mediators, such as cytokines, which modulate diverse aspects of the inflammatory and fibrogenic response. Therefore, the pathophysiological tissue response after lung
0167-8140/02/$ - see front matter q 2002 Published by Elsevier Science Ireland Ltd. PII: S 0167-814 0(02)00077-4
178
C.E. Ru¨ be et al. / Radiotherapy and Oncology 64 (2002) 177–187
irradiation implies the induction of numerous cytokines which form the basis for the multicellular interactions of the inflammatory and fibrogenic process associated with radiation injury [11,12,21,22,34,35,42]. The relative role of cytokine dysregulation versus direct tissue injury from irradiation for the pathogenesis of radiation pneumonitis/ fibrosis remains elusive. TNF-a is thought to be a key mediator for the pathogenesis of radiation pneumonitis, because it shows the following spectrum of biological activities. TNF-a exerts in particular proinflammatory effects by inducing the expression of adhesion molecules that recruit leukocytes into the sites of tissue damage, by priming leukocytes for oxidant production, and by inducing production of prostaglandins and other mediators of inflammation. TNF-a inhibits anticoagulatory mechanisms and therefore promotes thrombotic processes [29]. In addition, TNF-a exerts fibrogenic effects by stimulating the growth of fibroblasts and increasing the collagen deposition. [19]. Therefore, a pharmacological regulation of the TNF-a production at the initial stage could possibly halt the progression of radiation-induced injury. A drug which suppresses the production of TNF-a but lacks the many side effects of glucocorticoids might be useful as an anti-inflammatory agent. Pentoxifylline (PTX) is a xanthine derivate that has generated widespread interest in the field of oncology based on its reported potential ability to ameliorate radiation- and chemotherapy-induced toxicity. PTX has been shown to enhance microvascular blood flow and decrease platelet aggregation, thereby maintaining perfusion in radiated tissues. In addition, PTX down-regulates the production of proinflammatory cytokines, particularly TNF-a, in response to noxious stimuli and inhibits granulocyte-mediated cytotoxicity after TNF-a exposure and may, therefore, provide protection against radiation-induced, cytokine-mediated cellular damage [2,8,38,44,45]. Since PTX has long been used in the treatment of peripheral vascular disease, the drug is widely available and low toxicity has been demonstrated. The purpose of this study was to investigate the temporal and spatial release of TNF-a in the lung tissue after thoracic irradiation with 12 Gy. In addition, we evaluated the ability of PTX to reduce the radiation-induced TNF-a release and the lung toxicity in this animal model of thoracic irradiation.
2. Methods and materials 2.1. Animals C57BL/6J mice were purchased from Charles River Laboratories. Adult female mice, 8 weeks old and approximately 20 g in weight, were housed four to six per cage and allowed to acclimatise from shipping for 1 week prior to treatment. Four study groups were defined: those that received neither radiation nor PTX (NT ¼ no treatment
group: eight animals), those that received PTX (500 mg/l in drinking water) but no irradiation (PTX group: eight animals), those that underwent irradiation without PTX (XRT group: 24 animals) and those that received both PTX and radiation (PTX/XRT group: 24 animals). Pentoxifylline (Sigma, St. Louis, MO, USA) was added to the drinking water of the mice in the PTX and PTX/XRT groups at a concentration of 500 mg/l. This concentration was chosen based on the measured initial water consumption, delivering a predicted dose of 100 mg PTX/kg bodyweight per day. Pentoxifylline administration was initiated 1 week before thoracic radiation. Animal weights and PTX consumption, the latter assessed by average water intake, were monitored during the study period. All mice underwent actual (XRT and PTX/XRT groups) or sham irradiation (NT and PTX groups). 2.2. Radiation schedule A dose of 12 Gy to the midplane of the lungs was delivered in a single fraction via a posterior field using a linear accelerator. A plastic jig was used to restrain the mice without anaesthesia and lead strips were placed to shield the head and abdomen. The irradiation characteristics were as follows: beam energy: 10 MV-photons; dose-rate: 2.4 Gy/ min; source surface distance (SSD): 1 m; size of the radiation field: 18 cm £ 10 cm. A source film was taken to confirm that the entire lung was irradiated. Film dosimetry was used to deduce the relative dose distribution. Dosimetry was performed with a cylindrical ionisation chamber. The depth of the maximum dose of the 10 MVphoton beam could be significantly reduced by the tissueequivalent plastic material (thickness 23 mm) of the restraining jig. Therefore the build-up region of the 10 MV-photon beam was located in the plastic material achieving an acceptable dose uniformity throughout the thorax of individual mice. Following irradiation the mice were maintained four to six per cage in laminar flow hoods in pathogen-free rooms to minimise pulmonary infections and supplied with standard laboratory diet and water ad libitum. Age-matched controls were maintained under identical conditions for the course of the experiment. Treated and control mice were sacrificed by cervical dislocation at times corresponding to the latent period (1, 24, 72 h and 1 week post-irradiation) and the pneumonic phase (2, 4, 8, 16 and 24 weeks post-irradiation). For each treatment modality (XRT and PTX/XRT group) and time point three animals were used, and groups of untreated (NT group) or PTX-treated (PTX group) shamirradiated animals served as experimental controls. The experimental protocol was approved by the Medical Sciences Animal Care and Use Committee of the University of Mu¨ nster. 2.3. Tissue isolation The complete lungs were immediately removed after death without being perfused. The left lobes were placed
C.E. Ru¨be et al. / Radiotherapy and Oncology 64 (2002) 177–187
in fixative for histological and histochemical analysis and the right lobes were quickly frozen in liquid nitrogen for RNA isolation and subsequent polymerase chain reaction (PCR) analyses. The lung tissue from three different mice (XRT and PTX/XRT group) for every time point was analysed by a combination of light microscopy, immunohistochemistry and PCR. 2.4. PCR analysis 2.4.1. RNA extraction Total RNA preparations were performed using the RNeasyeTotal RNA Kit (Qiagen, Hilden, Germany). The lung tissue (a right lobe is equivalent to 75 mg tissue) was homogenised in 1.2 ml lysis buffer using a rotor–stator homogeniser (Ultra-Turrax, IKA, Staufen, Germany). Tissue lysates were kept at 2858C prior to processing or were used directly in the procedure according to the manufacturer’s protocol. Ribonucleic acid concentrations were determined by spectrophotometric absorption at 260 nm. The integrity of the RNA was assessed by denaturing agarose gel electrophoresis. 2.4.2. First-strand complementary DNA (cDNA) synthesis Total RNA was converted to cDNA utilising a first-strand cDNA synthesis kit according to the manufacturer’s protocol (Pharmacia, Biotech, Freiburg, Germany). In this procedure oligo dT primers were used for generating first-strand cDNA in a final reaction mix of 15 ml. 2.4.3. Real-time quantitative reverse transcriptase PCR (RT-PCR) The TNF-a mRNA expression in the lung tissue was quantified by ‘real-time’ quantitative RT-PCR. Using a dual-labelled fluorogenic probe, this method allows direct detection of PCR-product accumulation by measuring the increase in fluorescent emission continuously during the exponential phase of the PCR reaction [16,33]. 2.4.3.1. The principle of the method employed in brief. The assay uses fluorescent Taqman methodology and an instrument for measuring fluorescence in real time (ABI Prism 77-00 Sequence Detector; Perkin Elmer/Applied Biosystems, Foster City, CA, USA). The Taqman reaction requires a hybridisation probe which is labelled with two different fluorescent dyes of which one is a reporter dye (FAM, i.e. 6-carboxyfluorescein) covalently attached at the 5 0 end and the other is a quenching dye (TAMRA, i.e. 6-carboxy-tetramethylrhodamine) covalently attached at the 3 0 end. When the probe is intact, fluorescent energy transfer occurs and the reporter dye (FAM) fluorescent emission is absorbed by the quenching dye (TAMRA). During the extension phase of the PCR cycle, the fluorescent hybridisation probe is cleaved by the 5 0 –3 0 nucleolytic activity of the DNA polymerase. On cleavage of the probe, the reporter dye emission is no longer transferred
179
efficiently to the quenching dye, resulting in an increase of the reporter dye fluorescent emission spectra. The use of the sequence detector (ABI Prism) allows measurement of fluorescent spectra of all 96 wells of the thermal cycler continuously during the PCR amplification. Therefore, the reactions are monitored in real time. 2.4.3.2. Primer and probe design. PCR primers and probes for the murine TNF-a and the housekeeping gene GAPDH were designed using the computer program Primer Express (Perkin Elmer/Applied Biosystems, Foster City, CA, USA). Primers were located in two different exons and either one of the primers was located on an intron–exon junction to prevent co-amplification of genomic DNA. The melting temperature (Tm) of the primers was 58–608C and the probe Tm was at least 108C higher than the primer Tm at approximately 698C. The amplicon lengths were kept very short (see Table 1). The fluorogenic probe contained a reporter dye (FAM) covalently attached at the 5 0 end and a quencher dye (TAMRA) covalently attached at the 3 0 end. Extension from the 3 0 end was blocked by attachment of a 3 0 phosphate group. Fluorogenic probes were highperformance liquid chromatography (HPLC) purified. 2.4.3.3. PCR amplification. PCR reactions were performed in the ABI-prism 7700 sequence detector, which includes a Gene-Amp PCR system 9600 (Perkin Elmer/Applied Biosystems, Foster City, CA, USA). PCR amplifications were performed in a total volume of 25 ml, containing 0.5 ml cDNA sample, 50 mM KCl, 10 mM Tris–HCl (pH 8.3), 60 nM Passive Reference 1 (Rox), 200 mM dATP, dCTP, dGTP and 400 mM dUTP, 5.5 mM MgCl2, 0.05% gelatine, 300 nM of each primer, 1.25 U AmpliTaq Gold and 0.25 U AmpErase Uracil NGlycosylase (TaqMan PCR core reagent kit; Perkin Elmer/Applied Biosystems, Foster City, CA, USA). Each reaction also contained 200 nM of the corresponding detection probe (Table 1). Each PCR amplification was performed in triplicate Table 1 Primer and probe sequences for murine TNF-a and GAPDH a Name
Sequence (5 0 ! 3 0 )
Amplicon length
5 0 -CCA GGC GGT GCC TAT GTC T-3 0 5 0 -GGC CAT TTG GGA ACT TCT CAT-3 0 123 bp 5 0 -AGC CTC TTC TCA TTC CTG GTT GTG GCA-3 0 GAPDH FW 5 0 -CAA CTC ACT CAA GAT TGT CAG CAA-3 0 GAPDH RV 5 0 -GGC ATG GAC TGT GGT CAT GA-3 0 118 bp GAPDH FP 5 0 -CAT CCT GCA CCA CCA ACT GCT TAG CC-3 0 TNF-a FW TNF-a RV TNF-a FP
a Forward (FW) and reverse (RV) primer were always located in different exons. Fluorogenic probes (FP) are FAM-labelled at the 5 0 end and TAMRA-labelled at the 3 0 end. All cDNA sequences were obtained from the Genbank database.
180
C.E. Ru¨ be et al. / Radiotherapy and Oncology 64 (2002) 177–187
wells, using the following conditions: 2 min at 508C and 10 min at 958C, followed by a total of 40 two-temperature cycles (15 s at 958C and 1 min at 608C). All values for the TNF-a mRNA expression were normalised to the constitutively expressed mRNA of the reference gene GAPDH. 2.5. Histology For histological analysis the left lobes were fixed in 10% neutral buffered formalin, paraffin-embedded, and sectioned at an average thickness of 5 mm. The mounted sections were stained by haematoxylin and eosin as well as Goldner’s stain for collagen. 2.6. Immunohistochemistry 2.6.1. Immunohistochemical staining After dewaxing of the sections in xylene and rehydration in graded alcohols, polyclonal rabbit anti-mouse TNF-a antibody (code: IP-400, Genzyme, Cambridge, MA, USA) was applied overnight in a humid chamber at 48C (dilutions in 0.01 M RPMI containing 0.6% bovine serum albumin: TNF-a antibody 1:100), followed by goat anti-rabbit secondary antibody (1:30 in RPMI; 30 min at room temperature; Dako, Glostrup; Denmark) and alkaline phosphatase anti-alkaline phosphatase (APAAP) complex (1:100 in phosphate buffered saline (PBS); 60 min at room temperature; Dako, Glostrup; Denmark) [5]. The secondary reactions were performed in a semi-automatic immunostaining device (Omnibus, Quartett, Berlin, Germany). Subsequently the enzyme reaction was developed for 25 min at room temperature in a freshly prepared fuchsin solution. Finally, the sections were counterstained with haematoxylin and mounted in Kayser’s glycerine gelatine. Since all lung tissue sections were treated in the same way and stained simultaneously under the same conditions, the intensity of immunostaining could be compared. Omission of primary antibodies as well as the same concentration of rabbit IgG (prepared in the same way) was used as a negative control. 2.6.2. Computerised image analysis TNF-a immunoreactivity was quantified using a computerised image analysis system (microscope/camera Axiocam; image analysis software KS 300, Zeiss, Go¨ ttingen, Germany). Magnification, light intensity and colour bar values remained constant throughout the experiment to allow valid comparisons among specimens. Three variables were measured using computerised image analysis: the number of cells per field which showed definite staining for TNF-a (expressed as positive cell counts), the relative area stained for TNF-a which is a measure of the fraction of cells expressing a detectable amount of TNF-a protein and the mean specific density of the stained area, which is proportional to the concentration of staining. All lung specimens were subjected to anonymous evaluation. To ensure objectivity, a standardised procedure was
established and used consistently throughout the study whereby ten representative, non-contiguous and non-overlapping fields ( £ 20 objective) were systematically selected and analysed (each field covered an area between 40 000– 50 000 mm 2). The sum of the measurements from the ten fields was totalled, and the arithmetic mean was calculated. Positive cell counts were expressed as the average number of cells per field. Mean values from lung tissues of radiated animals (XRT and PTX/XRT group) were compared with the corresponding values obtained from control animals at the respective time points. To assess reproducibility, each slide was analysed twice by an investigator who was unaware of the treatment group and the results of the previous measurement. 2.7. Statistical methods For statistical analysis, the assessment times 2, 4, 8, 16 and 24 weeks after completion of irradiation were summarised to the pneumonic phase. The results of the quantitative assessments of the TNF-a mRNA expression (in the whole lung tissue) as well as the TNF-a protein expression (immunoreactivity in the lung parenchyma, expressed as positive cell counts) were evaluated by Mann–Whitney-U-test to compare the XRT group with the NT group, the PTX group and the PTX/XRT treatment-group. The statistical calculations were performed by SPSS-software (version 9) for Windows NT. The criterion for statistical significance was P , 0:05. 3. Results 3.1. PCR analysis The results of the quantitative assessment of the TNF-a mRNA expression in the lung tissue of the mice treated according to the different study groups are demonstrated in Fig. 1. Non-irradiated lung tissue of control animals (NT and PTX group) exhibited low levels of TNF-a expression (relative mRNA expression: between 0.19 and 0.38) ( ¼ control). The NT and PTX groups showed no significant differences in the level of TNF-a expression. Following thoracic irradiation with a single dose of 12 Gy (XRT group), radiation-induced TNF-a release in the lung tissue was appreciable even within the first hour (relative mRNA expression: 1:34 ^ 0:13) and subsequently decreased to basal levels during the latent period (24 h, 72 h and 1 week post-irradiation). During the complete pneumonic phase, irradiation-mediated TNF-a release was significantly increased and reached maximal values at 8 weeks post-irradiation (relative mRNA expression: 4:05 ^ 1:46). After exposure to thoracic irradiation, the lung tissue of the PTX-treated mice (PTX/XRT group) revealed only a minor radiation-mediated TNF-a mRNA response during the pneumonic phase (relative mRNA expression: up to 1:57 ^ 0:11 at 24 weeks post-irradiation). In contrast to
C.E. Ru¨ be et al. / Radiotherapy and Oncology 64 (2002) 177–187
181
Fig. 1. Time course of the TNF-a mRNA expression in the lung tissue of the mice that underwent thoracic irradiation with 12 Gy ( ¼ XRT group: open diamonds; mean ^ SD) and those that received PTX and thoracic irradiation with 12 Gy ( ¼ PTX/XRT group: closed diamonds; mean ^ SD). Control: TNF-a mRNA expression in non-irradiated lung tissue of control animals (NT and PTX group).
the radiation-only group, the PTX-treated mice revealed a modest up-regulation of TNF-a during the latent phase. Statistical comparison of the TNF-a mRNA expression during the pneumonic phase (2, 4, 8, 16 and 24 weeks postirradiation) revealed significant difference between the XRT group and the NT/PTX-group ðP , 0:05Þ. Due to the data variability and the relatively low number of specimens available (n ¼ 15 per treatment-group), the statistical comparison of the TNF-a mRNA expression between the XRT and PTX/XRT treatment-group was not significant ðP ¼ 0:07Þ.
3.2. Immunohistochemistry Non-irradiated lung tissue (NT and PTX group) exhibited inhomogeneous and variable TNF-a immunoreactivity in the bronchiolar epithelium (Fig. 2). In the lung parenchyma, only a few alveolar macrophages stained positive. In the walls of untreated blood vessels TNF-a could not be detected immunohistochemically. The staining pattern in irradiated lung tissue (XRT and PTX/XRT group) differed from that in non-irradiated specimens in the following respect: as soon as 1 h after thoracic irradiation, the bronchiolar epithelium in the lung tissue of the irradiated mice revealed an intense and homogenous staining for TNF-a. At this time point, between 80 and 90% of the epithelial cells were intensely stained with the anti-TNF-a antibody. In contrast the bronchiolar epithelium revealed variable staining intensity for TNF-a at the later assessment time points, and no clear time-dependent alterations in the number of staining epithelial cells could be observed (between 10 and 80%) (Fig. 3). Interestingly, the subcellular distribution of TNF-a in bronchiolar epithelial
cells often showed a gradient with the highest concentration at the apical pole. Irradiation induces adherence of monocytes to the endothelium within hours, which is followed by monocytes/macrophages invasion of the vessel wall. In the sections analysed, most of the positive macrophages were located in the adventitia of the blood vessels (Fig. 4). The endothelial lining of blood vessels displayed TNF-a immunoreactivity as a function of time after radiation exposure. In the first hours after thoracic irradiation (3 and 6 h postirradiation), TNF-a antibody stained the endothelial cells (Fig. 5), whereas at the beginning of the pneumonic phase (2 and 4 weeks post-irradiation) only a very weak staining of the vascular endothelium was observed. During the later assessment time points of the pneumonic phase (8, 16 and
Fig. 2. Immunohistochemical staining of TNF-a in non-irradiated lung tissue (NT group): variable staining intensity for TNF-a in the bronchiolar epithelium. Magnification £ 200.
182
C.E. Ru¨ be et al. / Radiotherapy and Oncology 64 (2002) 177–187
Fig. 3. Immunohistochemical staining of TNF-a in irradiated lung tissue, 2 weeks after irradiation (XRT group): intense and homogenous staining for TNF-a in the bronchiolar epithelium. Magnification £ 200.
Fig. 5. Immunohistochemical staining of TNF-a in irradiated lung tissue, 6 h after irradiation (XRT group): positive staining of the vascular endothelium. Magnification £ 200.
24 weeks post-irradiation) the endothelium did not react with TNF-a antibody. At day 2 post-irradiation the medial smooth muscle cells were positive for TNF-a and remained positive (with variable staining intensity) throughout the complete observation period of 24 weeks. In spite of the constant anatomical position of the lung tissue samples examined, the histological sections revealed quite variable sections of bronchioli and blood vessels. Therefore, a direct comparison between the XRT and PTX/XRT group regarding the number of staining cells and the staining intensity was not possible. During the stage of interstitial pneumonitis, microscopic examination of the lung tissue revealed inflammation with accumulation of positive inflammatory cells particularly in perivascular and peribronchial areas, as well as in the subpleural region. The lung parenchyma revealed a pronounced increase of positive alveolar macrophages (Figs. 6 and 7). Other cell types of the lung parenchyma
such as type I and II pneumocytes or fibroblasts, did not stain positive for TNF-a. Increased cytokine expression was detected predominantly in regions of histopathologic radiation injury; inflammatory lesions in particular contained detectable levels of TNF-a. In contrast to this no significant immunoreactivity could be observed in areas of normal lung tissue. The results of the quantitative assessment of TNF-a immunoreactivity in the lung parenchyma by the computerised image analysis method are shown in Fig. 8. Nonirradiated lung parenchyma (NT and PTX group) exhibited low levels of TNF-a protein expression with positive cell counts between 10 and 23 ( ¼ control). In the first hours after radiation exposure, the lung parenchyma of the XRT group and the PTX/XRT group revealed no significant increase in the number of stained cells. During the stage of interstitial pneumonitis (2, 4, 8, 16 and 24 weeks postirradiation) TNF-a immunoreactivity in the lung parench-
Fig. 4. Immunohistochemical staining of TNF-a in irradiated lung tissue, 2 weeks after irradiation (XRT group): positive macrophages in perivascular region. Magnification £ 200.
Fig. 6. Immunohistochemical staining of TNF-a in irradiated lung tissue, 6 h after irradiation (XRT group): positive alveolar macrophages in the lung parenchyma. Magnification £ 400.
C.E. Ru¨ be et al. / Radiotherapy and Oncology 64 (2002) 177–187
Fig. 7. Immunohistochemical staining of TNF-a in irradiated lung tissue, 8 weeks after irradiation (XRT group): TNF-a immunoreactivity predominantly in subpleural regions. Magnification £ 200.
yma of the radiation-only group (XRT group) showed a pronounced increase with maximal values up to 80 positive cells (Fig. 8). Statistical comparison of the positive cell counts in the lung parenchyma during the pneumonic phase (2, 4, 8, 16 and 24 weeks post-irradiation) revealed significant difference between the XRT group and the NT/ PTX group ðP , 0:05Þ. The quantitative evaluation of the immunoreactivity in the lung parenchyma of the PTX/XRT group revealed lower positive cell counts with values between 20 and 35 (Fig. 8). Due to the data variability and the relatively low number of specimens available (n ¼ 15 per treatment-group), the statistical comparison between the XRT- and PTX/XRT treatment-group was not significant ðP ¼ 0:07Þ.
183
The wide variability of the positive cell counts in the analysed section areas of the lung parenchyma among the individual mice treated in the same way probably reflects the fact, that at earlier stages the lungs often show differences in the spatial and temporal distribution of inflammatory lesions. Therefore the positive cell counts (expressed as the average number of cells per field) revealed wide variation ranges depending on the fields selected and analysed. Three variables were measured using computerised image analysis: the number of cells which showed definite staining for TNF-a, the relative area stained for TNF-a and the mean specific density of the stained area, which is proportional to the concentration of the stain. While there was a clear increase in the number of stained cells with a maximum at the beginning of the pneumonic phase, no remarkable changes in the specific optical density were observed. Therefore, after irradiation it is not the cytokine production of each cell that changes, but rather the number of cells involved in TNF-a production. These data suggest that with regard to the lung parenchyma, changes in cytokine expression are caused mainly by an increase in the fraction of cells expressing TNF-a, but not by changes in the level of TNF-a expression per cell.
4. Discussion 4.1. ‘Real-time’ quantitative RT-PCR RT-PCR is a powerful method used to quantify the mRNA expression of cytokines, which are often expressed at very low levels. Various methods of (semi-) quantitative RT-PCR have been described for measuring cytokine
Fig. 8. Time course of the TNF-a protein expression (expressed as positive cell counts) in the lung tissue of the mice that underwent thoracic irradiation with 12 Gy ( ¼ XRT group: open diamonds; mean ^ SD) and those that received PTX and thoracic irradiation with 12 Gy ( ¼ PTX/XRT group: closed diamonds; mean ^ SD). Control: TNF-a protein expression (positive cell counts of the lung parenchyma) in non-irradiated lung tissue of control animals (NT and PTX group).
184
C.E. Ru¨ be et al. / Radiotherapy and Oncology 64 (2002) 177–187
mRNA levels. One classical method measures PCR-product accumulation during the exponential phase of the reaction [36]. To obtain semi-quantitative results it is extremely important that the PCR product is measured during the log phase of the reaction, before saturation is reached. This is the major difficulty in using this method, because it necessitates interruption of the PCR reaction after an experimentally determined number of cycles. A drawback is the relatively small linear range, during which samples within a single experimental setup can be analysed. Alternatively, quantitative competitive PCR can be performed [37,39] which requires co-amplification of an internal cDNA or RNA control (competitor) with the unknown sample in the same tube. The amount of mRNA is quantified by titration of an unknown amount of target template against a dilution series of known amounts of the standard. Thus, the same primers are used to co-amplify the target and the competitor, in such a way that they amplify with the same efficiency, although they can be distinguished from each other (by differences in length or restriction sites). Although this method provides a strategy for accurate quantification, the construction of internal standards is technically sophisticated and labour intensive. For the detection of PCR products using either of these methods, several detection techniques can be used, all requiring extensive post-PCR procedures. The most commonly used are agarose gel electrophoresis and EtBr-staining, fluorescent labelling and analysis using polyacrylamide gels, radioactive labelling and Southern blotting or detection by phosphor-imaging. Major drawbacks using these classical detection systems are the use of hazardous chemicals and the potential risk for laboratory contamination. Moreover, all these post-PCR procedures are very time-consuming. In the present study we used the recently developed ‘realtime’ quantitative RT-PCR, a quick and accurate technique for measuring PCR-product accumulation during the exponential phase of the reaction [16,33]. As this method combines PCR amplification and product detection in one single step, the technique is very fast and easy to perform as compared to classical RT-PCR techniques. This method also provides accurate measurements over a wide range of template quantities. Furthermore, the use of an internal probe ( ¼ fluorogenic probe), guarantees the specificity of the PCR product to be measured. 4.2. Immunohistochemistry and quantitative image analysis The immunohistochemical methods for detection of TNF-a used in this study are well established [5], and the specificity of the antibody used has been documented previously [12]. Previous in vivo studies of cytokine protein expression in pathological conditions have largely relied on qualitative assessment of immunoreactivity. The computerassisted image processing and analysis method used in the present study enabled us to assess the quantitative TNF-a immunoreactivity of the lung parenchyma. The image
analysis method yielded highly reproducible results and provided objective, quantitative and comparative assessment of TNF-a protein expression levels. The method thus allowed reliable quantitative estimations of the TNFa immunoreactivity in the lung parenchyma of the different study groups during the different phases of injury. The present work is a detailed study on the temporal release of TNF-a in the lung tissue after thoracic irradiation. Moreover, the cellular source of TNF-a as well as the association between TNF-a immunoreactivity and specific histopathologic alterations could be demonstrated. Secondly, we evaluated the ability of PTX to reduce the radiation-induced TNF-a release on mRNA and protein level using a well-validated in vivo assay of pulmonary injury. Published data [11,12,21,22,34,35,42] and previous studies from our laboratory [37; further data in preparation] suggest a major role of several inflammatory and fibrogenic cytokines in the pathogenesis of radiation-induced lung damage. Of these cytokines studied, TNF-a appears to play a particularly prominent role in the acute phase of radiation-induced damage since it is consistently overexpressed in areas of histopathologic lesions during the course of radiation pneumonitis. Even 1 h after thoracic irradiation, the bronchiolar epithelium in the lung tissue of the irradiated mice revealed an intense and homogenous staining for TNF-a. Similar TNFa mRNA and protein expression pattern by the bronchiolar– alveolar epithelium immediately after asbestos exposure could be observed in a murine model of asbestos-induced lung fibrosis [4]. The normal tracheobronchial epithelium, which has a relatively slow turn-over, may require only low levels of growth factors. However, under conditions of increased growth such as during repair after cell damage, an increase in the synthesis and/or activation of TNF-a may be involved to stimulate cell proliferation and differentiation. Pulmonary epithelial hyperplasia observed in the late stage of various inflammatory lung diseases, is associated with prolonged overexpression of TNF-a [20]. Looking at the cytokine network in radiation-induced lung injury, TNFa which is initially overexpressed in pulmonary epithelial cells, probably not only stimulates epithelial cell proliferation, but also initiates the recruitment of macrophages and other inflammatory cells. Subsequently, TNF-a was overexpressed in the lung parenchyma particularly by pulmonary macrophages. TNF-a enhances the production of different chemokines which are thought to further promote the inflammatory process particularly by monocyte/macrophage recruitment into the sites of tissue damage. Increased cytokine expression was detected prominently in inflammatory lesions, particularly in perivascular, peribronchial and subpleural areas. Other cell types of the lung parenchyma like type I and II pneumocytes or fibroblasts did not stain positive for TNF-a. In the present study, the endothelial lining of blood vessels displayed TNF-a immunoreactivity as a function
C.E. Ru¨ be et al. / Radiotherapy and Oncology 64 (2002) 177–187
of time after radiation exposure. TNF-a induces the expression of cell surface adhesion molecules such as the intercellular adhesion molecule (ICAM-1) and the endothelial cell adhesion molecule (ELAM-1) [3]. Expression of these endothelial cell adhesion molecules is a critical factor in the binding of leukocytes to the vessel wall at the site of inflammation. Furthermore, TNF-a increases the endothelial cell permeability and induces the release of other inflammatory mediators, which may contribute to the increased permeability [29]. From day 2 post-irradiation onwards, the medial smooth muscle cells were positive for TNF-a and remained positive (with variable staining intensity) throughout the complete observation period of 24 weeks. These results correspond to an immunohistochemical study of radiation-induced damage in major blood vessels where a transient dose-dependent TNF-a induction in medial smooth muscle cells after combined intraoperative and external irradiation could be demonstrated [23]. A recent study also showed, that TNF-a is the main factor secreted by injured (but not normal) myocytes which is responsible for induction of additional inflammatory mediators, namely syndecans, in surrounding endothelial cells [46]. Syndecans participate in a variety of biological functions including regulation of blood coagulation, cell adhesion as well as regulation of endothelial cell growth. Therefore, this mechanism may play an important role in myocyte-dependent regulation of endothelial cell activity and the ability of injured tissues to initiate and propagate an angiogenic response. There is currently great interest in the potential value of PTX as an inhibitor of normal tissue injury caused by radiation or chemotherapy. This is due to promising early laboratory and clinical investigations, as well as the drug’s widespread availability and low toxicity profile. In an animal model of brain irradiation, PTX treatment of mice partially suppressed radiation-induced TNF-a expression in cerebral tissue [18]. In pilot human studies, PTX has shown apparent efficacy in promoting healing of chronic radiation muco-cutaneous complications that had previously been unresponsive to conservative management [6,7,13,14]. With respect to the amelioration of radiation-induced pulmonary injury, several observations suggested that PTX might be a useful therapeutic agent. PTX has been reported to significantly reduce radiation-induced pulmonary injury in rats [31]. Radiation pneumonitis is similar in histopathologic appearance to the adult respiratory distress syndrome (ARDS), which may result from a variety of insults such as sepsis, shock or drugs. In a guinea pig model of ARDS, PTX inhibited lung injury induced by TNF-a and Escherichia coli sepsis [15,27]. In various clinical studies, PTX therapy inhibited TNF-a secretion and improved the symptoms related to ARDS with no particular toxic effects [1,28]. In experimental studies of bleomycininduced alveolitis, PTX ameliorated the extent of the inflammatory reaction in the lung tissue [9,10]. Promising results from a clinical study suggest that PTX may improve
185
therapeutic regimens in pulmonary sarcoidosis either by sparing or replacing corticosteroids [43]. Despite its early promise, the ability of PTX to ameliorate radiation-induced injury of normal tissue requires further validation, before the drug can be empirically used on a widespread clinical basis. One intention of the present study was to examine the effects of PTX on the severity of the radiation-induced lung damage. Therefore, the qualitative and quantitative histopathologic alterations of the radiation-induced lung injury were documented (for each individual animal) as follows: after examining the whole lung section and estimating the extent of altered lung parenchyma, the different lesions referring to the respiratory parenchyma, the bronchial and vascular structures and the pleura were documented. The estimates of the overall lung damage were based on the quantitative evaluation of the altered lung parenchyma using computerised image analysis. During the pneumonic phase, the area of altered lung tissue varied from approximately 20% (2 weeks post-irradiation) to 80% (24 weeks post-irradiation). While the effect of PTX on the infiltration of inflammatory cells could be quantified, the other histopathologic alterations like intraalveolar oedema, exudation into air spaces, desquamation of epithelial cells, alterations of the capillaries, thickening of the alveolar septa etc. could only be qualitatively documented. In summary, while we observed a reduction of the inflammatory cell infiltration in the PTX/XRT group as compared to the XRT group, the overall lung damage and the other histopathologic alterations which contribute to the pleomorphic picture of radiation pneumonitis revealed no apparent difference between the two treatment modalities. Other experimental setups including radionuclide lung perfusion scans for functional assessment are possibly more suitable to study the effect of PTX on radiation-related lung injury. However, our data indicate that PTX reduces the TNF-a release on mRNA and protein level as measured by ‘real-time’ RT-PCR and immunohistochemistry. Moreover, the results of PCR analysis and immunohistochemical staining of corresponding lung tissue reveal significant correlation between mRNA expression and immunoreactivity. In the present study we analysed TNF-a on mRNA level by ‘real-time’ RT-PCR and on protein level by immunohistochemical staining at different assessment time points in individual animals; concomitantly the association between cytokine expression and localisation and specific histopathologic alterations were assessed. With limited amounts of RNA template required, ‘real-time’ RT-PCR affords the opportunity to measure transcript levels of different cytokines in the lung tissue, which can be subsequently further analysed by immunohistochemical staining and histological analysis. This strategy facilitates comprehensive analysis of the lung tissue within a single individual mouse and thus allows efficient correlation of the mRNA and protein expression of different cytokines and their cellular location with specific histopathologic alterations. There is no current evidence to suggest that PTX protects
186
C.E. Ru¨ be et al. / Radiotherapy and Oncology 64 (2002) 177–187
tumour cells from the intended effects of ionising radiation. In fact, other investigators have observed that PTX may enhance tumour radiosensitivity, primarily based on its reported ability to increase oxygenation by improving perfusion in otherwise hypoxic tumours [17,24,25,26,41].
[5]
[6]
5. Conclusions This study demonstrates a significant radiation-induced increase of TNF-a (on mRNA and protein level) in the lung tissue during the pneumonic phase. The predominant localisation of TNF-a in areas of inflammatory cell infiltrates suggests involvement of this cytokine in the pathogenesis of radiation-induced lung injury. In addition, we observed a pronounced (but statistically not significant) reduction of the TNF-a mRNA and protein production in the study group that received both PTX and radiation (PTX/ XRT group) as compared to the radiation-only group (XRT group). Therefore our results indicate that PTX down-regulates the TNF-a mRNA and protein production in the lung tissue in response to radiation. Inhibition of appropriate cytokines such as TNF-a at an early stage of pneumonitis is thought to be important for effective treatment of radiation-induced lung injury. Pharmacological regulation of the TNF-a production may thus ameliorate the severe consequences of lung irradiation. PTX has been shown to improve therapeutic strategies in numerous diseases in which TNF-a represents a causative pathophysiological factor. PTX pharmacologically controls the synthesis and the actions of TNF-a, and may therefore provide protection against radiation-induced, cytokinemediated cellular damage.
[7]
[8]
[9]
[10]
[11]
[12]
[13]
[14]
[15]
[16]
Acknowledgements This investigation was supported by grant no. FI 733/1/1 from the Deutsche Forschungsgemeinschaft (DFG) and by grant no. Fi-1-4-I/98-22; Fi-1-4-II/97-17; Fi-1-1-II/96-29 from the Innovative Medizinische Forschung (IMF) of the University of Mu¨ nster. The authors thank Dr Heinecke for the statistical analysis of these data for publication.
[17]
[18]
[19] [20] [21]
References [22] [1] Ardizzoia A, Lissoni P, Tancini G, et al. Respiratory distress syndrome in patients with advanced cancer treated with pentoxifylline: a randomized study. Support Care Cancer 1993;1:331–333. [2] Balibrea JL, Arias-Diaz J, Garcia C, et al. Effect of pentoxifylline and somatostatin on tumour necrosis factor production by human pulmonary macrophages. Circ Shock 1994;43:51–56. [3] Beck-Schimmer B, Schimmer RC, Warner RL, et al. Expression of lung vascular and airway ICAM-1 after exposure to bacterial lipopolysaccharide. Am J Respir Cell Mol Biol 1997;17:344–352. [4] Brass DM, Hoyle GW, Poovey HG, et al. Reduced tumor necrosis factor-alpha and transforming growth factor-beta 1 expression in the
[23]
[24]
[25]
lungs of inbred mice that fail to develop fibroproliferative lesions consequent to asbestos exposure. Am J Pathol 1999;154:853–862. Cordell JL, Falini B, Erber WN, et al. Immunoenzymatic labeling of monoclonal antibodies using immune complexes of alkaline phosphatase and monoclonal anti-alkaline phosphatase (APAAP complexes). J Histochem Cytochem 1984;32:219–229. Delanian S, Balla-Mekias S, Lefaix JL. Striking regression of chronic radiotherapy damage in a clinical trial of combined pentoxifylline and tocopherol. J Clin Oncol 1999;17:3283–3290. Delanian S. Striking regression of radiation-induced fibrosis by a combination of pentoxifylline and tocopherol. Br J Radiol 1998;71:892–894. Doherty GM, Jensen JC, Alexander HR, et al. Pentoxifylline suppression of tumor necrosis factor gene transcription. Surgery 1991;110:192–198. Entzian P, Gerlach C, Gerdes J, et al. Pentoxifyline inhibits experimental bleomycin-induced fibrosing alveolitis. Pneumologie 1997;51:375–380. Entzian P, Zahringer U, Schlaak M, et al. Comparative study on effects of pentoxifylline, prednisolone and colchicine in experimental alveolitis. Int J Immunopharmacol 1998;20:723–735. Finkelstein JN, Johnston CJ, Baggs R, et al. Early alterations in extracellular matrix and transforming growth factor b gene expression in mouse lung indicative of late radiation fibrosis. Int J Radiat Oncol Biol Phys 1994;28:621–631. Franko AJ, Sharplin J, Ghahary A, et al. Immunohistochemical localization of transforming growth factor b and tumor necrosis factor a in the lungs of fibrosis-prone and ‘non-fibrosing’ mice during the latent period and early phase after irradiation. Radiat Res 1997;147:245–256. Futran ND, Trotti A, Gwede C. Pentoxifylline in the treatment of radiation-related soft tissue injury: preliminary observations. Laryngoscope 1997;107:391–395. Gottlober P, Krahn G, Korting HC, et al. The treatment of cutaneous radiation-induced fibrosis with pentoxifylline and vitamin E. Strahlenther Onkol 1996;172:34–38. Harada H, Ishizaka A, Yonemaru M, et al. The effects of aminophylline and pentoxifylline on multiple organ damage after Escherichia coli sepsis. Am Rev Respir Dis 1989;140:974–980. Heid CA, Stevens J, Livak KJ, et al. Real time quantitative PCR. Genome Methods 1996;6:986–994. Honess DJ, Dennis IF, Bleehen NM. Pentoxifylline: its pharmakokinetics and ability to improve tumour perfusion and radiosensitivity in mice. Radiother Oncol 1993;28:208–218. Hong J-H, Chiang C-S, Campbell IL, et al. Induction of acute phase gene expression by brain irradiation. Int J Radiat Oncol Biol Phys 1995;33:619–626. Ibelgaufts H. TNF-a: tumor necrosis factor alpha, Dictionary of cytokines. Weinheim: VCH Verlagsgesellschaft mbH, 1995 p. 709–17. Jetten AM. Growth and differentiation factors in tracheobronchial epithelium. Am J Physiol 1991;260:L361–L373. Johnston CJ, Piedboeuf B, Baggs R, et al. Differences in correlation of mRNA gene expression in mice sensitive and resistant to radiationinduced pulmonary fibrosis. Radiat Res 1995;142:197–203. Johnston CJ, Piedboeuf B, Rubin F, et al. Early and persistant alterations in the expression of interleukin-1a, interleukin-1b and tumor necrosis factor a mRNA levels in fibrosis-resistant and sensitive mice after thoracic irradiation. Radiat Res 1996;145:762–767. Kallfass E, Kra¨ mling HJ, Schultz-Hector S. Early inflammatory reaction of the rabbit coeliac artery wall after combined intraoperative (IORT) and external (ERT) irradiation. Radiother Oncol 1996;39:167–178. Kim SH, Khil MS, Ryu S, et al. Enhancement of radiation response on human carcinoma cells in culture by pentoxifylline. Int J Radiat Oncol Biol Phys 1993;25:61–65. Kwon H-C, Kim S-K, Chung W-K, et al. Effect of pentoxifylline on
C.E. Ru¨ be et al. / Radiotherapy and Oncology 64 (2002) 177–187
[26]
[27]
[28]
[29]
[30]
[31]
[32] [33]
[34] [35]
radiation response of non-small cell lung cancer: a phase III randomized multicenter trial. Radiother Oncol 2000;56:175–179. Lee I, Levitt SH, Song CW. Improved tumour oxygenation and radiosensitization by combination with nicotinamide and pentoxifylline. Int J Radiat Biol 1993;64:237–244. Lilly CM, Sandhu JS, Ishizaka A, et al. Pentoxifylline prevents tumor necrosis factor-induced lung injury. Am Rev Respir Dis 1989;139: 1361–1368. Lissoni F, Ardizzoia A, Perego MS, et al. Inhibition of tumor necrosis factor-alpha secretion by pentoxifylline in advanced cancer patients with abnormally high blood levels of tumor necrosis factor-alpha. J Biol Regul Homeost Agents 1993;7:73–75. Litwin MS, Gamble JR, Vadas MA. Role of cytokines in endothelial cell function. In: Aggarwal BB, Pun RK, editors. Human cytokines: their role in disease and therapy, Oxford, England: Blackwell Science, 1995. pp. 101–129. McDonald S, Rubin P, Phillips TL, et al. Injury to the lung from cancer therapy: clinical syndromes, measurable endpoints, and potential scoring systems. Int J Radiat Oncol Biol Phys 1995;31:1187– 1203. Molteni A, Moulder JE, Cohen EF, et al. Control of radiation-induced pneumopathy and lung fibrosis by angiotensin-converting enzyme inhibitors and an angiotensin II type 1 receptor blocker. Int J Radiat Biol 2000;76:523–532. Movsas B, Raffin TA, Epstein AH, et al. Pulmonary radiation injury. Chest 1997;111:1061–1076. Overbergh L, Valckx D, Waer M, et al. Quantification of murine cytokine mRNAs using real time quantitative reverse transcriptase PCR. Cytokine 1999;11:305–312. Rodemann HP, Bamberg M. Cellular basis of radiation-induced fibrosis. Radiother Oncol 1995;35:83–90. Rubin P, Johnston CJ, Williams JP, et al. A perpetual cascade of cytokines postirradiation leads to pulmonary fibrosis. Int J Radiat Oncol Biol Phys 1995;33:99–109.
187
[36] Ru¨ be C, Finke C, van Valen F, et al. Ionizing radiation induces production of tumor necrosis factor a in the Ewing’s sarcoma cell line RM 82 in vitro and in vivo. Strahlenther Onkol 1997;173:407– 414. [37] Ru¨ be CE, Uthe D, Schmid KW, et al. Dose-dependent induction of transforming growth factor b (TGF-b) in the lung tissue of fibrosisprone mice after thoracic irradiation. Int J Radiat Oncol Biol Phys 2000;47:1033–1042. [38] Samlaska CP, Winfield EA. Pentoxifylline. J Am Acad Dermatol 1994;30:603–621. [39] Siebert PD, Larrick JW. PCR MIMICS: competitive DNA fragments for use as internal standards in quantitative PCR. Biotechniques 1993;14:244–249. [40] Travis EL, Komaki R. Treatment-related lung damage. In: Pass HI, Mitchell JB, Johnson DH, et al., editors. Lung cancer: principles and practice, Philadelphia, PA: Lippincott-Raven Publishers, 1996. pp. 285–301. [41] Vernimmen F, Verheye-Dua F, du-Toit H, et al. Effect of pentoxifylline on radiation damage and tumor growth. Strahlenther Onkol 1994;170:595–601. [42] Yi ES, Bedoya A, Lee H, et al. Radiation-induced lung injury in vivo: expression of transforming growth factor-beta precedes fibrosis. Inflammation 1996;20:339–351. [43] Zabel P, Entzian F, Dalhoff K, et al. Pentoxifylline in treatment of sarcoidosis. Am J Respir Crit Care Med 1997;155:1665–1669. [44] Zabel P, Schade FU, Schlaak M. Inhibition of endogenous TNF formation by pentoxifylline. Immunobiology 1993;187:447–463. [45] Zabel P, Schade FU, Schlaak M. Pentoxifylline – an inhibitor of the synthesis of tumor necrosis factor alpha. Immun Infekt 1992;20:80– 83. [46] Zhang Y, Pasparakis M, Kollias G, et al. Myocyte-dependent regulation of endothelial cell syndecan-4 expression. Biol Chem 1999;274:14786–14790.