Radiol Oncol 2003; 37(2): 89-99. Cathepsin L in human meningiomas Miha Trinkaus1, Andrej Vranič2, Vincenc V. Dolenc2, Tamara T. Lah1 1Department of Genetic Toxicology and Cancer Biology, National Institute of Biology, Ljubljana, Slovenia, 2Clinical Department of Neurosurgery, University Clinical Centre, Ljubljana, Slovenia Background. Although meningiomas are considered as benign tumours, about 10% comprise a subgroup of atypical meningiomas, classified as WHO grade II, with greater likelihood of recurrences and/or aggressive behaviour, including the possibility of brain tissue invasion. The lysosomal cysteine endopeptidase cathep-sin L plays a role in tumour cell invasion and malignant progression of cancer, and has been suggested as a prognostic marker for certain types of tumours. Results. In our study, we compared the expression of cathepsin L in 30 meningiomas with their clinical in-vasiveness. Cathepsin L was determined by immunohistochemical analysis, quantitative real-time RT-PCR and Northern blot. We showed that expression of cathepsin L protein was significantly higher (p=0.019) in 9 atypical than in 21 benign meningiomas. Within the group of benign meningiomas, expression of cathep-sin L was significantly lower in the transitional histological subtype. We measured the levels of cathepsin L A type of RNA splicing variants: LA, LAI and LAII, but not LAIII and not the LB variant, the latter being several times lower than the L A type. In contrast to protein levels, the levels of cathepsin L A, AI, AII RNA variants did not differ between histological subtypes or between benign and atypical meningiomas. The expression of total measured cathepsin L A, AI, AII RNA variants in the samples, taken from the centre and the periphery of the tumours, also showed no statistically significant differences. Conclusions. These results indicate that cathepsin L protein over-expression may contribute to the development of the aggressive and possibly invasive character of atypical meningiomas and that it may be up regulated at the translational level. Key words: meningioma - pathology; cathepsin L; reverse transcriptase, polymerase chain reaction; neo-plasm invasiveness Introduction Received 25 April 2003 Accepted 20 May 2003 Correspondence to: Tamara T. Lah, Ph.D., Department of Genetic Toxicology and Cancer Biology, National Institute of Biology, Večna pot 111, SI-1000 Ljubljana, Slovenia; Tel.: +386 1 423 5017; Fax: +386 1 423 5038; E-mail: tamara.lah@nib.si Meningiomas are tumours of the central nerv-ous system that derive from the coverings of the brain. Meningiomas are generally slow growing, benign tumours attached to the dura matter and composed of neoplastic meningothelial (arachnoidal) cells.1 Menin-gioma have a wide range of histopathological 90 Trinkaus M et al./Cathepsin L in human meningiomas apperances. However, of the 11 subtypes, the three most common are meningothelial, fi-brous and transitional meningioma.1,2 Most meningiomas are benign and can be graded into WHO grade I. Certain histologi-cal subtypes are associated with a less favourable clinical outcome and correspond to WHO grades II and III classification of his-tological diagnosis of brain tumours. Atypical meningiomas corresponding to grade II have moderately high labelling indices of nuclear proliferation index (PI) and have likely the following features: frequent mitosis, in-creased cellularity, small cells with high nu-clear-cytoplasmic ratio and/or prominent nu-cleoli, uninterrupted pattern-less or sheet-like growth and foci of ‘spontaneous’ or geo-graphic necrosis.1 These criteria have been shown to correlate with higher recurrence rates. Anaplastic malignant meningiomas (grade III) exhibit malignant histological features in excess of the abnormalities noted in atypical meningiomas. The latter constitute about 6%, whereas anaplastic malignant meningiomas account for between 1.0 and 2.8% of all meningiomas.2 Cathepsin L (EC 3.4.22.15) is a cysteine en-dopeptidase belonging to the evolutionarily well conserved clan CA/family CI proteases (http://www.merops.co.uk). It is localised mainly to lysosomes, mediating intracellular protein turnover, but under certain physio-logical conditions it also regulates the func-tion of some cytosolic and extracellular proteins. Cathepsin L gene is activated by a vari-ety of growth factors and some oncogenes. In NIH3T3 cells (mouse embryonal fibroblasts), cathepsin L gene is activated by platelet de-rived growth factor (PDGF), epidermal growth factor (EGF), tumour promoter phor-bol 12-myristate 13-acetate (PMA), cAMP and oncogenes, including v-ras, v-src and v-mos.4 Enzyme activity is regulated specifically by endogenous inhibitors of the cystatin family, thyropins and p41 fragment of invariant chain (Ii).5 Radiol Oncol 2003; 37(2): 89-99. Gottesman3 first reported an increased synthesis and secretion of the so called »major excreted protein, MEP« in NIH3T3 fibrob-lasts on viral transformation, the protein later being identified as cathepsin L.6 Many subse-quent studies have supported the hypothesis that increased levels and eventual secretion of cathepsin L are related to malignant transformation. Increased expression of RNA, protein and activity of cathepsin L was found in gastric7, colorectal, ovarian, breast, head and neck thyroid, lung endometrial carcinomas and gliomas.8 It was shown that cathepsin L is progressively expressed with increasing tu-mour malignancy in breast9 and brain tu-mours10,11 and has prognostic relevance for some cancers.8,12 The role of cathepsin L in the invasion process was confirmed in in vitro invasion assays by the use of selective inhibitor Click-148, which also reduced in vitro invasion of tumour cells13,14 and reduced bone metastasis in an in vivo mouse model.15 Collectively, this data suggests that cathepsin L plays an important role in tumour invasion. The present study is based on the hypoth-esis that, due to the observed association of cathepsin L with tumour progression and in-vasiveness, this lysosomal endopeptidase may be a biological marker for differentiation between benign and the more aggressive atypical meningioma. This suggested the further question as to whether cathepsin L expression would be increased at the invasive edge of meningiomas. Materials and methods Patients Thirty patients (19 female and 11 male) with diagnosed meningioma were selected ran-domly for the study. They ranged in age from 26-80 years. Immediately after surgical re-moval, samples from the centre and periph-ery of the tumours were removed and snap frozen in liquid nitrogen. The rest of the tu- Trinkaus M et al./Cathepsin L in human meningiomas 91 mour tissue was fixed in formaldehyde and embedded in paraffin. The histological slides were analyzed and categorized according to the WHO classification of brain tumours. The study comprised 22 benign and 8 atypical meningiomas (Table 1). The study was ap-proved by the Ethics Committee at the Ministry of Health of the Republic of Slovenia. Immunohistochemistry Immunohistochemical labelling was per-formed using standard techniques.16 Paraffin-embedded tissue sections were de-paraffinised and rehydrated. The sections were boiled in 100mM EDTA buffer (pH 8.0) for 5 minutes for antigen retrieval. The slides were first blocked with an inhibitor of en-dogenous peroxidase and then incubated with primary anti-cathepsin L mouse mono-clonal antibodies (CatLHLM1, clone N135) purchased by Krka d.d., Novo mesto, Slovenia, at 1:10 dilution of the stock concen-tration of 100µg/ml for 26min at 40°C. Subsequent IHC reaction was performed with biotinylated goat anti-mouse secondary antibodies, avidin with bound horseradish peroxidase and diaminobenzidine as chro-mogen. The immunohistochemistry and counterstaining with haematoxylin, was per-formed in an NexES IHC Staining Module (Ventana Medical Systems, USA). The inten-sity and frequency of immunostaining was considered independently by two patholo-gists for grading (0; 0.5; 1; 1.5; 2; 2.5; 3). Table 1. Histological diagnosis of meningiomas Histological WHO diagnosis grade Benign meningioma I Fibroblastic I Meningothelial I Transitional I Psammomatus I Secretory I Atypical meningioma II Proliferation index: immunolabelling with MIB-1 antibodies Proliferation index (PI) in meningiomas was determined by the IHC procedure, described above, using MIB 1 antibodies (mouse antihuman Ki-67, clone MIB-1, No 7240, Dako Corporation). The MIB-1 solution (200µg/ml) was diluted 1:20prior to use. Positive IHC staining indicated proliferating cells and was counted under a microscope using the computer program Lexica Q Product (Leica, Germany). PI was calculated as the percent-age of proliferating vs. all cells in the prepa-ration. Quantitative real time RT-PCR A fluorescence-based real-time quantitative RT-PCR method, developed by Perkin Elmer ABI (TaqMan), was used to measure cathep-sin L RNA levels in meningioma tissue. For each sample 1 µg of total RNA was reverse transcribed in a 50µl reaction using a High-Capacity cDNA Archive Kit (Applied Biosystem, USA). After an initial step at room temperature for 10 min, reverse transcription was performed at 37°C for 2h. Quantitative PCR reaction was carried out with a cDNA equivalent of 2ng total RNA per reaction, us-ing the TaqMan Universal PCR Master Mix (Applied Biosystems, USA): 1x TaqMan Universal PCR Master Mix, 200nM of each primer and 100nM TaqMan (final concentra-tion) fluorescent probe in a 10 µl final reac-tion volume. PCR reactions were carried out in an ABI Prism 7900 PCR machine (Applied No of tumours No of tumours analysed by IHC analysed by RT-PCR 66 66 86 11 11 87 Radiol Oncol 2003; 37(2): 89-99. 92 Trinkaus M et al./Cathepsin L in human meningiomas Biosystems, USA), with a step at 50°C for 2min for AmpErase UNG activity, hot start at 95°C for 10min, followed by 40 cycles with denaturation at 95°C for 15s and anneal-ing/elongation at 60°C for 1min. To normal-ize the signal for cathepsin L A type variants, amplification of 18S ribosomal RNA was per-formed as an internal control in a duplex re-action together with amplification of cathep-sin L cDNA. Primers and probes for 18S rRNA were purchased from Applied Biosystem in Pre-Developed TaqMan Assay Reagent Control Kit. We designed probe and primers to meas-ure simultaneously cathepsin LA, cathepsin LAI and cathepsin LAII RNA splicing vari-ants, but not LAIII RNA variant.17 The follow-ing primers were synthesized (Laboratories Eurobio, France): forward primer for cathep-sin L A variants: AGC GTC TAC CCC GAA CTC TG, located at bp 158-178 in exon1 and reverse primer for cathepsin L A variants: TTG TGC ATC GCC TTC CAC T, located at bp 372-387 in exon 2. The nucleotide se-quence for cathepsin L A variants probe, labelled with FAM reporter molecule, ACT CAT CCT TGC TGC CTT TTG CCT GG, is located in exon 2 at bp 372-387. Cathepsin L B variant does not have exon 1, therefore the forward primer was located in intron 1, bp 161-186, extending 4 nucleotides into exon 1, whereas the reverse primers and the probe were the same as for the cathepsin L A variants. Northern blot of selected tumour samples Total RNA was isolated with TRIzol (Gibco BRL, USA) according to the supplier’s in-structions. The RNA concentration was esti-mated by A260nm. 20µg of total RNA per lane was mixed with ethidium bromide and elec-trophoresed on 1% formaldehyde-agarose gel for 3h at 70V. Northern transfer was per-formed overnight in 20x SSC (Standard Saline Citrate, 3M sodium chloride, 0.3M sodium citrate, pH 7.0) on a rapid downward transfer system using the turboblotter appa-Radiol Oncol 2003; 37(2): 89-99. ratus and Nytran membrane (Schleicher and Schuell, Germany). RNA was cross-linked to the membrane by UV irradiation. The membrane was photographed under UV light and the photographs used for signal normaliza-tion. The blots were pre-hybridized for 3h at 50°C in »high SDS« buffer (5x SSC, 2% block-ing solution, 50mM sodium phosphate, 0.1% (w/v) lauroylsarcosine, 7% (w/v) SDS, 50% formamide) as suggested by the supplier of the blocking solution (Roche, Switzerland). Hybridization was performed overnight in the same buffer containing 30ng/ml DIG-la-belled full length cDNA probe. Blots were washed twice with 2x washing buffer (2xSSC, 0.1%, w/v, SDS) and twice with 0.5x washing buffer (0.5xSSC, 0.1% (w/v) SDS). All wash-ing steps were performed for 10min at room T. The signals were detected by chemilumi-nescence using the CDP Star System (Roche, Switzerland) according to the supplier’s in-structions. The signals of cathepsin L RNA on Northern blots were analyzed densitometri-cally with program Lucia G 4.21 (Laboratory Imaging Ltd, GB) and normalized to 18S rRNA signal. Statistics For statistical calculation we used SPSS 10.1 for Windows (SPSS Inc, USA) program. The differences in cathepsin L expression were analyzed by the Mann-Whitney and Kruskal-Wallis non-parametric test. The significance of differences between histological subtypes of meningiomas was given as p values. P<0.05 was considered to be significant. Results Immunohistochemistry Cathepsin L immunostaining was evaluated in 30 meningiomas, characterised histologi-cally (Table 1). Figures 1A and 1B show weak (score 1.0) and strong (score 3.0) IHC staining of cathepsin L in tumour cells of benign tran- Trinkaus M et al./Cathepsin L in human meningiomas 93 A ,--B sitional and an atypical meningioma, respec-tively. No statistically significant differences in cathepsin L levels were observed between meningothelial and fibrous meningioma, but significantly lower levels were observed in transitional meningiomas (Figure 2A). When all benign subtypes were grouped and com-pared with all atypical meningiomas, we found significantly higher (p=0.009) scores of IHC staining in the more invasive atypical meningiomas (Figure 2B). Quantitative real time RT-PCR Cathepsin L is encoded by four major RNA species: cathepsin LA, cathepsin LAI, cathep-sin LAII RNAs, the newly discovered cathep-sin L AIII RNA, and cathepsin LB RNA.19 Figure 1. Immunohistochemical staining for cathepsin L and proliferation index MIB-1?i. (A) Benign menin-gioma (transitional subtype). Cathepsins L is im-munolabelled as dark brown dot, demonstrating its lysosomal location (marked with arrow, 200x magni-fication). (B) Atypical meningioma. Abundant lysoso-mal labelling of cathepsin L, examples of IHC reaction are marked with arrows (200x magnification). The staining was performed as described in Material and methods. (C) PI (MIB-1) IHC staining in atypical meningioma. Nuclear location of the Ki 67 antigen was observed and scored as described in Material and methods. Haematoxylin counter- staining was used to label the nuclei. Cathepsin L RNA levels were analysed by quantitative RT-PCR using specifically de-signed probes, as described in Material and methods. We found that the levels of cathep-sin LB RNA in meningiomas were between 3 and 5 cycles less (average about 8 times low-er) than with cathepsin LA RNA variants (not shown), therefore the LB variant was not de-termined in this study. The primers and the probe which detects all previously mentioned cathepsin L A type RNA variants, but not the recently discovered cathepsin LAIII splice variant, were designed. No statistically signif-icant differences were found in the amount of cathepsin L A, AI, AII type splice variants be-tween different histological subtypes of meningiomas (Figure 3). Also, no statistically Radiol Oncol 2003; 37(2): 89-99. 94 Trinkaus M et al./Cathepsin L in human meningiomas A * B M-B F-B T-B Histological subtypes of bening meningiomas Benign Atypical Groups of meningiomas 2,5 .E 1,5 c o 't 1 | I 0,5 Q. C D lii C 4 ö 1 M-B F-B T-B Benign Atypical Histological subtypes of bening meningiomas Groups of meningiomas Figure 2. Immunohistochemical scoring of cathepsin L and PI in different histologycal subtypes of meningiomas. (A) Comparison of the scores of cathepsin L within benign tumours: M-B meningothelial benign, F-B fibroblastic benign and T-B transitional bening. (B) Comparison of mean IHC scores in all benign and all atypical menin-giomas. IHC scoring was performed according to the frequency and intensity of cathepsin L labeling on a scale from 0 - 3, as described in Methods. (C) Immunohistochemical scoring of PI within benign tumors: M-B meningothelial benign, F-B fibroblastic benign and T-B transitional bening meningioma. (D) Immunohistochemical scoring of PI in all benign and all atypical meningiomas, presented as mean scores. Results are presented as mean ± standard error The stars indicate statistically significant diffferences (p<0.05) between the groups. significant differences were found between the expression of cathepsin LA RNA variants in atypical and benign meningiomas. The comparison between the samples taken from the periphery and the centre of the tumours revealed no difference in the amount of cathepsin L A type variants (not shown). Northern blot of selected tumour samples To determine whether total cathepsin L mRNA expression, using the cDNA probe of cathepsin L, differs between histological sub-types types and between the centre and the periphery of the tumours, Northern blot analysis was performed in two meningothe-lial (samples 2,6), one fibroblast (sample 1), Radiol Oncol 2003; 37(2): 89-99. three transitional (samples 3,7,8) and two atypical meningiomas (samples 4,5). Samples from the centre and periphery of the tumours were loaded separately (Figure 4A). Expression in cathepsin L RNA samples varies, even within the same subtype of tu-mour, for example samples 2 and 6, which are both classified as meningothelial subtype. Densitometric analysis (Figure 4B) showed slightly lower expression in fibroblastic and transitional benign subtypes, but from this we cannot deduce any significant differences between these meningiomas. As observed with cathepsin LA RNA variants, the expression is very similar in the centre and in the periphery of the tumours. Trinkaus M et al./Cathepsin L in human meningiomas 95 Figure 3. Cathepsin LA RNA variants determined by real time RT PCR. The expression of cathepsin LA variants was quantitated and related (normalized) to the levels of 18S rRNA in different histological sub-types of meningioma in the center (c ) and at the pe-riphery (p) of the meningiomas: M-B meningothelial benign, F-B fibroblastic benign, T-B transitional be-nign, At-atypical. Results are shown as mean ± standard error. Correlation of cathepsin L measurements Cathepsin L protein expression (IHC score) and the expression measured by cathepsin L A variants RNA (mRNA/18srRNA) levels cor-related statistically significantly (p<0.02) re-gardless of the location of the sample meas-urements, i.e. tumour centre (r=0.45), tumour periphery (r=0.43) or total tumour (r=0.48). Proliferation index (PI) The proliferation index indicates the relative (%) expression of antibodies labelling a nuclear antigen, which is greatly increased in all phases of the cell cycle, except in Go. It is used as a histopathological marker indicating the degree of cell proliferation (Figure 1C). Our data confirm previous reports 18 that PI discriminates between benign and atypical meningiomas, but not between different his-tologies of benign meningioma (Figure 2 C, D). It has been also reported to be good prog-nostic marker for relapse of meningiomas. In our study, PI correlated moderately (r =0.30), although statistically significantly (p=0.003), with the expression of cathepsin L protein. Discussion Meningiomas are among the most common tumours encountered in neurosurgical prac-tice1,2,19 and the incidence is estimated to constitute between 13% and 26% of primary intracranial tumours, with an annual inci-dence of approximately 6 per 100 000 popu-lation.1,2 Meningiomas are in 90% of cases be-nign (WHO grade I), whereas 6-8 % are atyp-ical, 1-2 % malignant anaplastic menin-giomas, the rest comprising other types of grade II and III meningiomas.1 In spite of the existence of certain morphological, histo-pathological and a few biological parameters, discrimination between clear benign, border benign and atypical meningioma remains the key problem.19,21 Therefore, new markers of malignant progression of atypical menin-gioma, which could lead clinicians towards a more informed treatment of patients, are needed. Proliferation index is such a parameter which has been clearly shown to distinguish between benign and invasive atypical and anaplastic meningioma18, and this has been confirmed in this study. It has been found to correlate with cathepsin L antigen in atypical vs benign, but not within the benign menin-gioma group, suggesting that cathepsin L may not be related to increased proliferation rate. Proteolytic enzymes have long been pro-posed as prognostic factors for disease-free and overall survival of cancer patients (re-cently reviewed in 25), including those with brain tumours26,27,28, due to their role in invasion processes and metastases. The metallo-protease stromelysin was also suggested as a marker for invasiveness of meningioma29 and the same was confirmed for lysosomal cys-teine cathepsins in our previous study.21 We analyzed the expression of cathepsin B and cathepsin L in 67 benign and 21 atypical meningiomas and found that the protein lev-els of both cathepsins were significantly ele-Radiol Oncol 2003; 37(2): 89-99. 96 Trinkaus M et al./Cathepsin L in human meningiomas 1c 1p 2c 2p 3c 3p 4c 4p 5c 5p 6c 6p 7c 7p 8c 8p RNA wieght marker 6948 b -— 4742 — , 2661 __ 1821 --- 1517 --- 1049 --- 575 --- 28 S rR NA 18 S rR NA I ««M»« vated in atypical tumours. Moreover, among 67 benign tumours, nine had certain features of malignancy, classified as borderline benign meningioma, and significantly higher expression of cathepsin B was also found in the borderline benign tumours compared with clear benign tumours. However, cathepsin L could only discriminate between atypical and clear benign tumours. This suggested that cathep-sin L was a less selective diagnostic marker for distinguishing between histomorphologi-cally benign but invasive, and histomorpho-logically clear benign tumours. However, sev-eral recent in vitro studies have confirmed a Radiol Oncol 2003; 37(2): 89-99. 1. 6 kb Figure 4. Detection of total cathepsin L RNA by Northern blot. The upper part of the figure indicates the total RNA (20µg) loaded on the gel and separated by electrophoresis in 28S and 18S rRNA bands. Below, 1.6 kb cathepsin L hybridised with DIG labelled full length cathepsin L cDNA probe, as de-scribed in Material and methods. The samples repre-sent: 1, fibroblastic benign; 2 and 6, meningothelial benign; 3, 7 and 8, transitional benign; 4 and 5 atypi-cal meningioma. Samples were loaded in pairs of tis-sue from center (c) and periphery (p) of the tumors. (B) The mean signal intensity of cathepsin L RNA from center and periphery of the tumor normalized to 18S rRNA signal in histological subtypes of meningothe-lial (M-B), fibroblastic (F-B), transitional (T-B) and atypical (At) of 8 individual meningiomas, which were analyzed by Northern blot experiment above (A). close link between the invasive potential of tumour cells and increased cathepsin L expression. In brain tumours, first reports that cathepsin L may be related to brain tumour invasion came from Sivaparvathi et al.22, who found that cathepsin L expression paralleled increased malignancy of astrocytoma and glioblastoma cell lines. Moreover, the authors prevented glioblastoma cells from invading Matrigel by using specific cathepsin L anti-bodies. We have confirmed these results in an astrocytoma cell line model, using differ-ent extracellular matrices and a variety of synthetic cathepsin B and L inhibitors.11 Trinkaus M et al./Cathepsin L in human meningiomas 97 Recently, we have also demonstrated stable transfection of an IPTP glioblastoma cell line with whole length cDNA of cathepsin L, which resulted in significant, 80%, inhibition of cell invasion in Matrigel compared with control cells transfected with the vector-GFP.23 This was confirmed in other cell lines by others.24 In this study, protein expression of cathep-sin L was significantly lower in transitional benign meningiomas than in other benign types of meningiomas. At present we cannot explain this, since the transitional subtype has features transitional between those of meningothelial and fibrous meningiomas. Strojnik et al.21 also found no variations in the subtypes of benign meningioma. However, the levels of cathepsins were more related to their intracranial localisation and were the highest in parasagittal and convexity menin-gioma, which were also clinically more aggressive. In our previous study of glioblastoma invasion, we detected high concentrations of cathepsin B at the invasive edges of invading tumours27, confirming earlier reports that in-tra-tumour and intracellular localisation of cathepsins at the plasma membrane may be important in facilitating lytic interactions be-tween the tumours and surrounding stro-ma.25 In this study we failed to observe dif-ferent expression of cathepsin L between the centre and the periphery of the tumour, ei-ther by immunohistochemistry, Northern blot or RT-PCR methods. This suggests that cathepsin L may not be directly involved in degradation of extracellular matrix proteins at the cell (tumour surface), but may play a different role in the proteolytic events leading either to tumour cell invasion and/or other features of malignancy. In the present study we found significantly higher levels of cathepsin L protein in atypi-cal (grade II) compared with benign menin-gioma (grade I), indicating possible correla-tion of cathepsin L with invasiveness and/or clinical aggresiveness of grade II menin-giomas. However, we have not succeeded in demonstrating increased cathepsin L RNA in atypical meningioma. Although total cathep-sin L RNA was only assayed in a few menin-giomas, the cathepsin L A splicing variants (LA, LAI and LAII) measured in all 30 pa-tients, also did not differ significantly be-tween benign and atypical meningiomas. Chauhan et al.30 first reported two splicing variants of the single cathepsin L gene, LA and LB variants, expressed concurrently at a similar ratio in several different cell lines, LB variant being higher than LA. This is in con-trast to our results in meningioma tissue, where we found about 8 times lower L B variant and therefore did not quantify this variant. The discrepancy between messenger RNA (L A variants) and protein levels of cathepsin L can be explained either by an in-creased rate of translation of cathepsin L protein in atypical meningioma or by the fact, that we have not determined the LAIII splic-ing variant. Abula et al.31 has recently demon-strated that LAIII is predominant in all tis-sues, including malignant tumours, and showed the highest expression and transla-tional rate in vitro and in vivo.17,31 It is possi-ble, that this variant is also the most active in atypical and possibly anaplastic malignant meningioma, resulting in higher protein lev-els of cathepsin L, as observed in this study and previously.21 This needs to be further in-vestigated. Brain invasion may occur in histologically benign, atypical or anaplastic malignant meningiomas.1 The presence of brain invasion connotes a greater likelihood of recur-rence with brain-invasive histologically be-nign meningioma having clinical courses sim-ilar to atypical meningioma.1 This clearly in-dicates that new markers of invasiveness, such as cathepsin L, are needed to predict malignant clinical behaviour of histologically classified benign tumours. In conclusion, we have demonstrated sig-Radiol Oncol 2003; 37(2): 89-99. 98 Trinkaus M et al./Cathepsin L in human meningiomas nificantly higher levels of cathepsin L protein in atypical meningiomas. As cathepsin L is a marker of invasion but not of cell proliferation of malignant tumours, we may hypothe-size that atypical meningioma acquire inva-sive behaviour. In contrast to protein, the lev-els of cathepsin L (LA, LAI and LAII) splicing RNA variants were not higher in atypical meningioma, strongly suggesting that the rate of RNA translation increased in atypical compared to benign meningioma. Parti-cularly, the level and translational rate of LAI-II splicing variant, may be responsible for in-creased protein concentration of cathepsin L in atypical meningioma, what needs to be further investigated. We confirmed our hypoth-esis that cathepsin L is a biological marker for differentiation between benign and the more aggressive atypical meningioma and may be used to predict clinically observed aggressive behaviour of meningiomas. Acknowledgements This study was supported by the Project J3-2409 (granted to V.V.Dolenc,P.I.) and Programme 105/509 (granted to T.T.Lah) from the Ministry of Education, Science and Sports of Republic of Slovenia. We thank Prof. Dr. Roger Pain for fruitful discussion and careful reading the paper. References 1. Louis DN, Scheithauer BW, Budka H, von Deimling A, Kepes JJ. Meningeal tumors. In: Kleihaus P, Cavenee WK, editors. Pathology and ge-netics of tumours of the nervous system. Lyon: IARC press; 2000. p. 175-84. 2. Kleihues P, Burger PC, Scheithauer BWm, editors. Histological typing of tumours of the central nerv-ous system. New York,: Springer-Verlag; 1993, pp 33-40. 3. Gottesman MM. Transformation-dependent secre-tion of a low molecular weight protein by murine fi-broblasts. Proc Natl Acad Sci USA 1978; 75: 2767-71. 4. Kane SE, Gottesman MM. The role of cathepsin L in malignant transformation. SeminCancer Biol 1990; 1: 127-36. 5. Turk B, Turk D, Salvesen GS. Regulating cysteine protease activity: Essential role of protease inhibitors as guardians and regulators. Curr Pharm Design 2002; 8: 1623-37. 6. Portnoy DA, Erickson AH, Kochan J, Ravetch JV, Unkeless JC. Cloning and characterization of a mouse cysteine proteinase. J Biol Chem 1986; 261: 14697-703. 7. Chauhan SS, Goldstein LJ, Gottesman MM. Expression of cathepsin L in human tumors. Cancer Res 1991; 51: 1478-81. 8. Kos J, Lah T. Cysteine proteinases and their en-dogenous inhibitors: Target proteins for progno-sis, diagnosis and therapy in cancer. Oncol Rep 1998; 5: 1349-61. 9. Lah TT, Čerček M, Blejec A, Kos J, Gorodetsky E, Somers R, Daskal I. Cathepsin B, a prognostic in-dicator in lymph node negative breast carcinoma patients: comparison with cathepsin D, cathepsin L and other clinical indicators. Clin Cancer Res 2000; 6: 578 -84. 10. Levičar N, Strojnik T, Kos J, Dewey AR, Pilkington GJ, Lah TT. Lysosomal enzymes, cathepsins, in brain tumor invasion. J Neuro-Oncol 2002; 58: 21-32. 11. Lah TT, Strojnik T, Levičar N, Bervar A, Zajc I, Pilkington G, Kos J. Clinical and experimental studies of cysteine cathepsins and their inhibitors in human brain tumors. Int J Biol Markers 2000; 15: 90-2. 12. Harbeck N, Alt U, Berger U, Kruger A, Thomssen C, Janicke F, Hofler H, Kates RE, Schmitt M. Prognostic impact of proteolytic factors (uroki-nase-type plasminogen activator, plasminogen ac-tivator inhibitor 1, and cathepsins B, D, and L) in primary breast cancer reflects effects of adjuvant systemic therapy. Clin Cancer Res 2001; 7: 2757-64. 13. Premzl A, Puizdar V, Zavasnik-Bergant V, Kopitar-Jerala N, Lah TT, Katunuma N, Sloane BF, Turk V, Kos J. Invasion of ras-transformed breast epithe-lial cells depends on the proteolytic activity of cys-teine and aspartic proteinases. Biol Chem 2001; 382: 853-7. 14. Yagel S, Warner AH, Nellans HN, Lala PK, Waghorne C, Denhardt DT. Suppression by cathepsin L inhibitors of the invasion of amnion membranes by murine cancer cells. Cancer Res 1989; 49: 3553-7. Radiol Oncol 2003; 37(2): 89-99. Trinkaus M et al./Cathepsin L in human meningiomas 99 15. Katunuma N, Tsuge H, Nukatsuka M, Asao T, Fukushima M. Structure-based design of specific cathepsin inhibitors and their application to protection of bone metastasis of cancer cells. Arch Biochem Biophys 2002, 397; 2: 305-11. 16. Perentes E, Rubinstein LJ. Recent applications of immunoperoxidase histochemistry in human neu-ro-oncology: An update. Arch Patol Lab Med 1987; 111: 796-812. 17. Arora S, Chauhan SS. Identification and charac-terization of a novel human cathepsin L splice variant. Gene 2002; 293: 123-31. 18. Perry A, Stafford LS, Scheithauer BW, Suman VJ, Lohse CM. The prognostic significance of MIB-1, p53 and DNA flow cytometry in completely re-sected primary meningiomas. Cancer 2002; 94: 765-72. 19. Nakasu S, Hirano A, Llena JF, Shimra T, Handa J. Interface between the meningioma and the brain. Surg Neurol 1989; 32: 206 - 12. 21. Strojnik T, Židanik B, Kos J, Lah TT. Cathepsins B and L are markers for clinically invasive types of meningiomas. Neurosurgery 2001, 48: 598-605. 22. Sivaparvathi M, Yamamoto M, Nicolson GL, Gokaslan ZL, Fuller GN, Liotta LA, Sawaya R, Rao JS. Expression and immunohistochemical localiza-tion of cathepsin L during the progression of human gliomas. Clin Exp Metastas 1996; 14: 27-34. 23. Levičar N, Dewey R, Daley E, Bates TE, Davies D, Kos J, Pilkington GJ, Lah TT. Selective suppres-sion of cathepsin L by antisense cDNA impairs human brain tumor cell invasion in vitro and pro-motes apoptosis. Cancer Gene Therapy 2003; 10: 141-51. 24. Krueger S, Kellner U, Buehling F, et al. Cathepsin L antisense oligonucleotides in a human osteosar-coma cell line: Effects on the invasive phenotypes. Cancer Gene Therapy 2001; 8: 522-8. 25. Koblinski JE, Ahram M, Sloane BF. Unraveling the role of proteases in cancer. Clin Chim Acta 2000; 291: 113-35. 26. Frosch BA, Sloane BF. The role of proteolytic en-zymes in brain tumor invasion. In: Mikkelsen T, Bjerkvig R, Laerum OD, Rosenblum MI editors. Brain tumor invasion: Biological, Clinical and Therapeutic Considerations. Wiley-Liss Inc. 1998. p. 275-99. 27. Strojnik T, Kos J, Židanik B, Golouh R, Lah TT. Cathepsin B immunohistochemical staining in tumor and endothelial cells is new prognostic factor in patients with brain tumors. Clin Cancer Res 1999; 5: 559-67. 29. Perret SG, Duthel R, Fotso MJ, Brunon J, Mosnier JF. Stromelysin-3 is expressed by aggresive menin-giomas. Cancer 2002; 94: 765-72. 30. Chauhan SS, Popescus NC, Ray D, Fleischmann R, Gottesman MM, Troen BR. Cloning, genomic organisation and chromosomal localisation of human cathepsin L. J Biol Chem 1993: 268, 2: 1039-45. 31. Abudula A, Rommerskirsch W, Weber E, Gunther D, Wiederanders B. Splice variants of human cathepsin L mRNA show different expression rates. Biol Chem 2001; 382: 1583-91. Radiol Oncol 2003; 37(2): 89-99.