- Email: [email protected]
Astrocytoma grade IV (glioblastoma multiforme) displays 3 subtypes with unique expression profiles of intermediate filament proteins
Human Pathology (2013) xx, xxx–xxx
www.elsevier.com/locate/humpath
Original contribution
Astrocytoma grade IV (glioblastoma multiforme) displays 3 subtypes with unique expression profiles of intermediate filament proteins☆,☆☆ Omar Skalli PhD a,⁎, Ulrika Wilhelmsson PhD b , Charlotte Örndahl MD, PhD c , Boglarka Fekete MD d , Kristina Malmgren MD, PhD d , Bertil Rydenhag MD, PhD d , Milos Pekny MD, PhD b,⁎ a
Department of Biological Sciences, The University of Memphis, Memphis, TN 38152, USA Center for Brain Repair and Rehabilitation, Department of Clinical Neuroscience and Rehabilitation, Institute of Neuroscience and Physiology, Sahlgrenska Academy, University of Gothenburg, Medicinaregatan 9A, 41390 Göteborg, Sweden c Department of Pathology, Institute of Biomedicine, Sahlgrenska Academy, University of Gothenburg, Medicinaregatan 9A, 41390 Göteborg, Sweden d Department of Clinical Neuroscience and Rehabilitation, Institute of Neuroscience and Physiology, Sahlgrenska Academy, University of Gothenburg, 41390 Göteborg, Sweden b
Received 21 December 2012; revised 25 March 2013; accepted 27 March 2013
Keywords: Astrocytoma grade IV; Glioblastoma multiforme; Intermediate filament proteins; GFAP; Vimentin; Nestin; Synemin
Summary Astrocytoma grade IV (glioblastoma multiforme) is the most common and most malignant tumor of the central nervous system and is currently noncurable. Here, we have examined a populationbased cohort of 47 patients with grade IV astrocytoma, who underwent tumor surgery at Sahlgrenska University Hospital in Sweden and who survived after surgery for less than 200 days (short survivors, 28 patients) and more than 500 days (long survivors, 19 patients). For each tumor, we ascertained information on patient age, sex, tumor location, oncological treatment, and survival after surgery. The analysis of the tumor volume and the extent of tumor resection (incomplete versus complete resection of the macroscopic tumor) was made retrospectively from the preoperative radiological investigations and, when available, also from postoperative radiology. We performed semiquantitative immunohistochemical evaluation of the presence of intermediate filament (nanofilament) proteins glial fibrillary acidic protein, vimentin, nestin, and synemin in tumor cells. The intermediate filament system helps cells and tissues to cope with various types of stress, and thus, it might affect the malignant potential of grade IV astrocytoma. We propose a subclassification of astrocytomas grade IV with respect to the expression of the intermediate filament proteins glial fibrillary acidic protein, vimentin, nestin, and synemin, namely,
☆
Conflict of interests: The authors declare no conflict of interests. Funding: This work was supported by research grants from Swedish Medical Research Council (project 11548), AFA Research Foundation, ALF Göteborg (project 11392), Sten A. Olsson Foundation for Research and Culture, Söderberg Foundations, the Free Mason Foundation, Amlöv's Foundation, Health Medical Care Committee of the Västra Götal and Föreningen Margaretahemmet, NanoNet COST Action (BM1002), the EU FP 7 Program EduGlia (237956), the EU FP 7 Program TargetBraIn (279017), and research grant from the University of Memphis. ⁎ Corresponding authors. Omar Skalli is to be contacted at Department of Life Sciences, University of Memphis, Memphis, TN 38152, USA. Milos Pekny, Center for Brain Repair and Rehabilitation, Department of Clinical Neuroscience and Rehabilitation, Institute of Neuroscience and Physiology, Sahlgrenska Academy at the University of Gothenburg, Medicinaregatan 9A, SE-413 90 Gothenburg, Sweden. E-mail addresses: [email protected] (O. Skalli), [email protected] (M. Pekny). ☆☆
0046-8177/$ – see front matter © 2013 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.humpath.2013.03.013
2
O. Skalli et al. type A, B, and C. Our results suggest that the expression of the intermediate filament proteins glial fibrillary acidic protein, vimentin, nestin, and synemin is coregulated in grade IV astrocytomas. The expression patterns of the intermediate filament proteins in astrocytoma type A, B, and C might have biological and clinical significance. © 2013 Elsevier Inc. All rights reserved.
1. Introduction Astrocytoma grade IV (glioblastoma multiforme) is the most common primary brain tumor with a median survival of only 6 to 12 months. These tumors originate either from astrocytes that have accumulated mutations and de-differentiated or from neural stem/progenitor cells, which, in animal models, were localized to the subventricular zone [1,2]. The cytoplasm of astrocytoma cells contains intermediate filaments (IFs) (known also as nanofilaments), the key cellular system that helps cells and tissues cope with various types of stress [3]. IFs of astrocytes are composed of glial fibrillary acidic protein (GFAP) and vimentin and, in the case of immature or reactive astrocytes, also of nestin and synemin [4-7]. These IF proteins are also present in astrocytomas [4,8,9], and a number of studies addressed a possible connection between astrocytoma cell biology and their IF system. Some in vitro studies have suggested a negative correlation between GFAP protein levels and astrocyte dedifferentiation and transformation. Suppression of GFAP in U-251 human astrocytoma cells by GFAP antisense complementary DNA enhanced cell proliferation and anchorage-independent growth in soft agar [10] and decreased the ability of these cells to extend cellular processes in cocultures with neurons [11]. Expression of GFAP complementary DNA in astrocytoma cells slowed proliferation, increased the size of cellular processes, reduced soft agar colonies growth [12], and reduced tumorigenicity in vivo [13]. We showed that primary astrocytes derived from GFAP−/− mice grew more quickly and reached higher saturation cell densities than astrocytes from wild-type mice [14]. High-grade astrocytomas are often GFAP negative [1517], and in a previous study, we have evaluated the progression of astrocytomas in a GFAP negative context [18]. The results showed that astrocytomas induced in mice of the p53−/− background by prenatal exposure to ethylnitrosourea developed at a comparable rate and showed similar clinical progression and histopathologic features between wild-type and GFAP−/− mice [18], but it remains to be seen whether similar results would be obtained with other mutations prevalent in grade IV astrocytomas, such as those affecting the epidermal growth factor receptor. The role of vimentin in astrocytomas remains unclear, but vimentin is important for the epithelial-mesenchymal transition [19] and it associates with aggressive breast cancer subtypes [20]. It was suggested that nestin contributes to the aggressive behavior of astrocytoma cells as nestin small interfering RNA reduced growth of astrocytoma cells in vitro
[21] and intratumor injection of nestin short hairpin RNA (shRNA) into mouse astrocytomas slowed the tumor growth [22]. We previously demonstrated the presence of synemin in both astrocytoma cell lines and human astrocytomas [4] and suggested that synemin contributes to proliferation, soft agar colony formation, and migratory properties of astrocytoma cells [23,24]. Although many astrocytomas grade IV express the IF proteins GFAP, vimentin, nestin, and synemin, their coexpression and its potential prognostic value have not been studied. Here, we have used tissue from 47 grade IV astrocytomas and evaluated the coexpression of these IF proteins and made the initial assessment whether such a coexpression could be of a prognostic value.
2. Materials and methods 2.1. The patient cohort Between 2004 and 2008, 229 patients underwent neurosurgery at Sahlgrenska University Hospital in Gothenburg, Sweden, with a histopathologic diagnosis of grade IV astrocytoma. In the whole cohort, the median survival was 0.82 years, and the mean survival was 1.06 years. From this cohort, we selected 2 groups of patients (47 patients in total): the patients who survived less than 200 days after the operation (28 patients; “short survivors”) and the patients who survived more than 500 days (19 patients; “long survivors”) and for whom information about the macroscopic completeness of the tumor resection and on additional oncological therapy was available. The tumor tissue was collected in accordance with the ethical guidelines of the Sahlgrenska Academy University Hospital. Diagnosis was based on the criteria defined in the World Health Organization classification of tumors of the central nervous system. The analysis of tumor volume and the extent of tumor resection performed was made retrospectively from the preoperative radiological investigations and, when available, also from the postoperative radiology. No systematic immediate postoperative imaging was performed, and the evaluation of the extent of the resection of the macroscopic tumor was based on the sum of information obtained from the surgeons' visual evaluation and by postoperative imaging when performed. The volume was calculated from the preoperative coronal and axial images as reported before [25]. Data on patient age, sex, and survival time were collected as was information on whether a patient received postoperative oncological treatment.
Astrocytoma grade IV subtypes with unique expression profiles
2.2. Antibodies, immunohistochemistry, and light microscopy
with the 2-tailed Fisher's exact test. Comparisons of mean and median were performed with t test and Mann-Whitney rank sum test, respectively. For all these tests, differences were deemed significant when P b .05.
Goat anti-synemin antibodies were raised against bacterially expressed human α-synemin C-terminal domain and purified as described [4]. The other primary antibodies were commercially available and included: mouse anti-human vimentin clone V9 (Sigma-Aldrich Sweden, Stockholm, Sweden), mouse antiGFAP clone GA5 (Sigma-Aldrich), and mouse anti-human nestin clone 10C2 (Merck Millipore, Billerica, MA, USA). For immunohistochemistry, deparaffinized, 4-μm-thick sections of human tissues were rehydrated with phosphatebuffered saline, and endogenous peroxidase activity was blocked for 30 minutes with 0.3% H2O2 in methanol. After a 5-minute phosphate-buffered saline wash, sections were incubated with 10% normal horse serum for 30 minutes and incubated for 30 minutes with the primary antibodies diluted at 1:200 (goat anti-synemin), 1:500 (mouse antinestin), and 1:5000 (mouse anti-vimentin and anti-GFAP). Antibody binding was detected with the avidin-biotinperoxidase method, using the Vectastain reagent kit (Vector Laboratories, Burlingame, CA, USA) and biotinylated pan-specific secondary antibody (Vector Laboratories). The peroxidase activity then was revealed with diaminobenzidine (Vector Laboratories), and nuclear counterstain was performed with Mayer's hematoxylin. For the purpose of quantitation, for each antibody, all 47 cases were stained at the same time using the same solutions and identical conditions. Light microscopic observations were performed with a Nikon Eclipse 800 (Nikon Instruments, Melville, NY, USA).
3. Results 3.1. There are 3 subtypes of astrocytomas grade IV with respect to the expression of the IF proteins GFAP, vimentin, nestin, and synemin To evaluate whether in astrocytomas grade IV the IF proteins GFAP, vimentin, nestin, and synemin are coexpressed and if their individual and/or collective expression patterns could be of prognostic value, we performed histopathologic and immunohistochemical analysis of these 4 IF proteins in 47 grade IV astrocytomas (28 patients surviving b200 days and 19 patients surviving N500 days after operation). Histopathologic assessment of hematoxylineosin–stained sections from the tumor tissue showed the defining characteristics of astrocytoma grade IV in all the patients, including cellular heterogeneity, mitotic figures, abundant necrosis, and/or vascular proliferation. Immunohistochemical analysis on consecutive sections stained with antibodies against GFAP, vimentin, nestin, or synemin demonstrated that all cases examined contained tumor cells positive for each of these proteins with the immunoreactivity being homogenous within the tumor tissue. There were, however, differences in the staining intensities of each IF protein between tumors. Semiquantitative evaluation of the staining intensities of the tumor cells was performed using a scale of 1 (weakest) to 6 (strongest) (Table 1). For the purpose of statistical analysis, grading scores were ultimately divided into 2 categories: weak (score 1-3) and strong (score 4-6). This analysis revealed that most grade IV astrocytomas (62%) were strongly positive for nestin, whereas a smaller proportion of tumors strongly stained for vimentin (19%), GFAP (25%), and synemin (23%), respectively. Two-tailed Fisher's exact test analysis was performed to determine whether there were correlations between the staining intensities for the different IF proteins present in grade IV astrocytomas. The P values obtained ranged from
2.3. Semiquantitative evaluation and statistical analysis For each antibody, slides with the weakest and strongest staining in the tumor cells were selected to define the 2 extremes of the grading scale (1 and 6, respectively). These slides served as reference to define intermediate staining intensities (2-5). All slides were then graded relative to these 6 reference slides. Data were analyzed with the GraphPad Prism 5 software. Association between groups and outcomes was analyzed Table 1 Intensity score
GFAP Vimentin Nestin Synemin
3
Intensity of IF protein immunoreactivity in individual tumors No. of tumors Weak
Mean (median) scores Strong
All tumors (n = 47)
Short survival (n = 15)
Long survival (n = 10)
1
2
3
4
5
6
Score 1-6
Score 1-3
Score 4-6
Score 1-3
Score 4-6
Score 1-3
Score 4-6
3 14 0 0
20 16 7 23
12 8 11 13
8 7 10 5
2 1 11 5
2 1 8 1
2.83 (3.0) 2.32 (2.0) 4.04 (4.0) 2.89 (3.0)
2.25 1.84 2.61 2.36
4.50 (4.0) 4.33 (4.0) 4.93 (5.0) 4.63 (5.0)
2.25 (2.0) 2.08 (2.0) 2.40 (2.0) 2.31 (2.0)
5.00 4.00 4.90 5.00
2.12 (2.0) 1.89 (2.0) 2.50 (2.5) 2.37 (2.0)
4.50 (4.5) 4.00 (4.0) 4.75 (4.5) 4.00 (4.0)
(2.0) (2.0) (3.0) (2.0)
(5.0) (4.0) (5.0) (5.0)
NOTE. Patient survival after surgery: short survival, less than 200 days; long survival, more than 500 days. Patients who did not receive postsurgical adjuvant therapy or had complete tumor removal were not included the short and long survival groups.
4
O. Skalli et al. Table 2 Correlation between the immunostaining intensities (weak or strong) of pairs of IF proteins in astrocytoma grade IV tissue (n = 47 tumors) IF proteins GFAP GFAP GFAP Nestin Nestin Synemin
3.2. Expression of IF proteins GFAP, vimentin, nestin, and synemin by astrocytoma grade IV cells and the patient survival after surgery
P Synemin Vimentin Nestin Synemin Vimentin Vimentin
.0026* .0004* .0167* .0331* .0083* .0225*
NOTE. P values were obtained by analyzing contingency tables with the 2-tailed Fisher's exact test. Asterisks denote significant correlation between the staining intensities of the respective IF proteins.
0.0004 to 0.0331 for each possible pair of IF proteins (Table 2). This demonstrates that there is a high degree of correlation between the staining intensities, of each possible pair of IF proteins present in astrocytoma grade IV tissues, namely, GFAP and synemin, GFAP and vimentin, GFAP and nestin, nestin and synemin, nestin and vimentin, and synemin and vimentin (Table 2). These correlations were further evidenced by the finding that 3 major patterns of IF protein expression exist in the grade IV astrocytomas that we have examined. In the first pattern (30% of the tumors; pattern A), GFAP, synemin, and nestin were all expressed at high levels (Fig. 1). The opposite pattern, that is, weak staining for each of the four IF proteins examined, was observed in 36% of the tumors (pattern B) (Fig. 2). Finally, the third pattern (34% of the tumors; pattern C) was characterized by strong nestin staining but low GFAP, vimentin, and synemin staining (Fig. 3). The average age of patients with pattern B was lower than of those with pattern C (51 ± 3.0 years versus 59 ± 1.6 years, respectively; P b .05). There was no association between pattern A, B, or C and tumor volume (data not shown).
Next, we examined whether there was an association between individual IF proteins or IF protein patterns and patient survival after surgery. As mentioned above, 2 groups of patients were considered in our study, one with survival less than 200 days (28 patients) and the other with survival of more than 500 days (19 patients). The 2 groups did not differ with respect to the sex composition, but the average age of patients in the longer survival group was lower than in the shorter survival group (50 ± 2.8 and 57 ± 1.5 years, respectively; P = .0190). The average tumor volume in the longer survival group was smaller than in the shorter survival group (28 cm3 ± 5.5 versus 48 cm3 ± 7.1, respectively; P = .0403). Although most patients received radiotherapy and/or chemotherapy after surgery, 13 patients did not receive any of these treatment modalities, for a variety of reasons. As expected, the absence of postsurgical radiotherapy and/or chemotherapy was associated with poor survival (P = .0005; Table 3): all these 13 patients belonged to the short survival group and, therefore, were excluded from further analysis. Complete resection of the macroscopic tumor was achieved in 11 patients (who also received postsurgical oncological treatment), and it had a profound effect on the survival: 2-tailed Fisher's exact test analysis revealed that patients who received complete resection of the macroscopic tumor were more likely to belong to the long survivor group than those for whom the complete macroscopic resection could not be achieved (36 patients, P = .0034; Table 3). The number of patients in which complete macroscopic tumor removal was achieved was small (11 patients in total) and hindered meaningful statistical analysis; therefore, for the purpose of correlative analysis between survivor groups and IF protein
Fig. 1 Immunohistochemical staining of consecutive sections of a grade IV astrocytoma with GFAP, nestin, and synemin antibodies, as indicated on top of each picture. HE, hematoxylin eosin staining of a parallel section. Tumor cells in this case strongly stain with each of the 3 antibodies used. The figure shows pattern A, which represents 30% of the 47 tumors examined. Bar, 50 μm.
Astrocytoma grade IV subtypes with unique expression profiles
5
Fig. 2 Immunohistochemical staining of consecutive sections of a grade IV astrocytoma with GFAP, vimentin, nestin, and synemin antibodies, as indicated on top of each picture. HE, hematoxylin eosin staining of a parallel section. Tumor cells in this case weakly stain with each of the 4 antibodies used. The picture shows pattern B, which represents 36% of the 47 tumors examined. Note that positive controls for vimentin, nestin, and synemin staining are represented by blood vessels and/or glomeruloids and for GFAP by reactive astrocytes with stellate morphology. Bar, 50 μm.
expression, we focused on patients with incomplete tumor resection excluding those who did not receive postsurgical treatment (25 patients in total). We examined with the 2-tailed Fisher's exact test the correlation between individual IF protein staining intensity (weak or strong; see above for definition) or A, B, and C staining pattern and affiliation to short and long survivor groups. No association with survival was found (Table 4).
4. Discussion 4.1. Three subtypes of astrocytoma grade IV defined by the IF protein expression pattern The presence of individual IF proteins GFAP, vimentin, nestin, or synemin in the tumor cells of astrocytoma grade IV was previously reported [4,8,9,15-17]. This study, however, is the first to examine the coexpression of these 4 IF proteins in grade IV astrocytomas. Our populationbased cohort of 47 patients with astrocytoma grade IV clustered into tumor subtypes A, B, and C based on the expression of the IF proteins GFAP, vimentin, nestin, and
synemin: subtype A with high expression of GFAP, nestin, and synemin; subtype B with low expression of all 4 IF proteins; and subtype C with high expression of nestin and low expression of GFAP, vimentin, and synemin. Interestingly, these findings demonstrate that neither low GFAP nor high nestin is universal characteristic of grade IV astrocytomas because approximately one-third of these tumors exhibited either strong GFAP and/or weak nestin immunostaining. Synemin presence in astrocytomas was reported in an earlier study of 4 grade IV astrocytomas [4]. The present study is the first to examine synemin expression in a larger series of grade IV astrocytomas and suggests that about a fourth of these tumors strongly express this IF protein.
4.2. Coregulation of GFAP, vimentin, nestin, and synemin—from astrocytes to grade IV astrocytoma cells The existence of these 3 astrocytoma grade IV subtypes with a distinct correlation between the staining intensities of individual IF proteins in the tumor cells suggests a
Fig. 3 Immunohistochemical staining of consecutive sections of a grade IV astrocytoma with GFAP, vimentin, nestin, and synemin antibodies, as indicated on top of each picture. HE, hematoxylin eosin staining of a parallel section. Tumor cells in this case are strongly stained with nestin antibodies but weakly stained with vimentin, GFAP, and synemin antibodies. The figure shows pattern C, which represents 34% of the 47 tumors examined. Bar, 50 μm.
6
O. Skalli et al. Table 3 Association between survival and removal of the macroscopic tumor (contingency analysis 1) and survival and postsurgical therapy (contingency analysis 2) Contingency analysis 1 a c
Complete resection Incomplete resection c Total a b c d
N500 9 10 19
b200 2 26 28
Total 11 36 47
Contingency analysis 2 b d
Treatment after surgery No treatment after surgery d Total
4.3. IF proteins and the ability of cells to cope with stress The IF protein up-regulation, in particular, the upregulation of GFAP, the main constituent of astrocyte IFs, is a hallmark of astrocyte activation and reactive gliosis in the whole range of neuropathologies including brain and spinal cord trauma and infections; stroke; epilepsy; some primary and secondary tumors of the brain and spinal cord; and neurodegenerative diseases such as Alzheimer disease, Batten disease, or multiple sclerosis [7,27]. In mice, combined genetic deficiency of GFAP and vimentin leads to attenuated astrocyte activation and reactive gliosis [28,29]. The mice deficient for both GFAP and vimentin exhibit reduced glial scar formation and decreased ability to cope Table 4
19 0 19
15 13 28
34 13 47
4.4. IF proteins and cell motility Astrocytoma grade IV cells show remarkable ability to migrate within the brain tissue. In fact, at the onset of the
Correlation between survival and the immunostaining intensities and survival and A, B, and C staining patterns of IF proteins Pc
IF staining intensity
Outcome a
Vimentin (weak vs strong) GFAP (weak vs strong) Nestin (weak vs strong) Synemin (weak vs strong) IF staining pattern (A vs B+C) b IF staining pattern (B vs A+C) b IF staining pattern (C vs A+B) b
Survival Survival Survival Survival Survival Survival Survival
c
Total
with acute or chronic cellular stress of diverse kind, for example, in acute neurotrauma [28,30], focal brain ischemia [31], retinal injury [32], Batten disease [33], or Alzheimer disease [34]. In this study, we show the existence of 3 subtypes of astrocytoma grade IV, with respect to the immunoreactivity pattern of the IF proteins GFAP, vimentin, nestin, and synemin. Using a rather limited number of astrocytoma grade IV tumors, we have made the first attempt to investigate whether these tumor subtypes associate with different survival and, therefore, could be of prognostic value. For example, it is possible to speculate that—similarly to astrocytes lacking GFAP and vimentin—the tumor cells of the B subtype of astrocytoma grade IV with low expression of IF proteins GFAP, vimentin, nestin, and synemin might be less capable of coping with ischemic/hypoxic, osmotic, and other stresses and, consequently, allow longer survival of the patients after surgery compared with the A or C subtypes. Using a limited cohort of 25 patients, who were comparable with respect to the extent of the removal of the macroscopic tumor and postsurgical oncological treatment, we have not seen any such difference. A larger prospective study might be needed to address this.
Variables compared
b
b200
Association between removal of the macroscopic tumor and patient survival after surgery (P = .0034; 2-tailed Fisher's exact test). Association between additional therapy after surgery (radiotherapy and/or chemotherapy) and patient survival after surgery (P = .0005). Resection of the macroscopic tumor. Radiotherapy and/or chemotherapy treatment after surgery.
coregulation of the expression of GFAP, vimentin, nestin, and synemin. Using gene expression profiling by single cell quantitative real-time polymerase chain reaction, we recently reported that the expression of GFAP, vimentin, and nestin messenger RNA is coregulated in individual astrocytes in primary cultures [26]. Reactive astrocytes and astrocytes in culture (for review see [6]) as well as cells of astrocytoma grade IV examined in this study, all co-express the same IF proteins GFAP, vimentin, nestin, and synemin implying that the expression of these IF proteins is coregulated both in astrocytes and in grade IV astrocytoma cells.
a
N500
(short vs (short vs (short vs (short vs (short vs (short vs (short vs
long) long) long) long) long) long) long)
Patient survival after surgery: short survival, less than 200 days; long survival, more than 500 days. See Results for the description of the staining patterns A, B, and C. P values were obtained by analyzing contingency tables with the 2-tailed Fisher's exact test.
.6265 1.0000 .2406 1.0000 1.0000 .2406 .4018
Astrocytoma grade IV subtypes with unique expression profiles clinical symptoms, most patients have already intracerebral metastases of the primary tumor. These tumor “guerrilla cells” are highly resistant to chemotherapy and radiotherapy. The tumors that arise from these cells contribute to the very poor prognosis of grade IV astrocytoma even in those patients in which the macroscopic tumor has been successfully removed by surgery. We have previously demonstrated that astrocytes devoid of GFAP or vimentin show reduced migratory potential and that the cell migration is further reduced in astrocytes devoid of both GFAP and vimentin [35]. Astrocytes devoid of GFAP and vimentin cannot produce IFs because neither of their remaining IF proteins—nestin and synemin—can selfpolymerize or copolymerize into IFs [5,29]. Similarly, we previously demonstrated that synemin contributes to migratory properties of astrocytoma cells [23] and that it is a positive regulator of their proliferative capacity [24]. Thus, it is interesting to speculate that the subclass B astrocytoma grade IV with low expression of all 4 IF proteins might exhibit reduced migratory potential of the tumor cells both before and after surgery. Such a hypothesis could be tested in a prospective study in which the tumor spread can be assessed by magnetic resonance imaging both preoperatively and postoperatively.
[11]
[12]
[13]
[14]
[15]
[16]
[17] [18]
[19]
[20]
Acknowledgments The authors thank Anna Stokowska for her advice on the statistical analysis.
[21]
[22]
[23]
References [1] Chen J, McKay RM, Parada LF. Malignant glioma: lessons from genomics, mouse models, and stem cells. Cell 2012;149:36-47. [2] Watkins S, Sontheimer H. Unique biology of gliomas: challenges and opportunities. Trends Neurosci 2012;35:546-56. [3] Pekny M, Lane EB. Intermediate filaments and stress. Exp Cell Res 2007;313:2244-54. [4] Jing R, Pizzolato G, Robson RM, Gabbiani G, Skalli O. Intermediate filament protein synemin is present in human reactive and malignant astrocytes and associates with ruffled membranes in astrocytoma cells. Glia 2005;50:107-20. [5] Jing R, Wilhelmsson U, Goodwill W, et al. Synemin is expressed in reactive astrocytes in neurotrauma and interacts differentially with vimentin and GFAP intermediate filament networks. J Cell Sci 2007;120:1267-77. [6] Pekny M, Pekna M. Astrocyte intermediate filaments in CNS pathologies and regeneration. J Pathol 2004;204:428-37. [7] Pekny M, Nilsson M. Astrocyte activation and reactive gliosis. Glia 2005;50:427-34. [8] Herpers MJ, Ramaekers FC, Aldeweireldt J, Moesker O, Slooff J. Coexpression of glial fibrillary acidic protein- and vimentin-type intermediate filaments in human astrocytomas. Acta Neuropathol 1986;70:333-9. [9] Yung WK, Luna M, Borit A. Vimentin and glial fibrillary acidic protein in human brain tumors. J Neurooncol 1985;3:35-8. [10] Rutka JT, Hubbard SL, Fukuyama K, Matsuzawa K, Dirks PB, Becker LE. Effects of antisense glial fibrillary acidic protein complementary
[24]
[25] [26]
[27] [28]
[29]
[30]
[31]
[32]
7 DNA on the growth, invasion, and adhesion of human astrocytoma cells. Cancer Res 1994;54:3267-72. Weinstein DE, Shelanski ML, Liem RK. Suppression by antisense mRNA demonstrates a requirement for the glial fibrillary acidic protein in the formation of stable astrocytic processes in response to neurons. J Cell Biol 1991;112:1205-13. Rutka JT, Smith SL. Transfection of human astrocytoma cells with glial fibrillary acidic protein complementary DNA: analysis of expression, proliferation, and tumorigenicity. Cancer Res 1993;53:3624-31. Toda M, Miura M, Asou H, Sugiyama I, Kawase T, Uyemura K. Suppression of glial tumor growth by expression of glial fibrillary acidic protein. Neurochem Res 1999;24:339-43. Pekny M, Eliasson C, Chien CL, et al. GFAP-deficient astrocytes are capable of stellation in vitro when cocultured with neurons and exhibit a reduced amount of intermediate filaments and an increased cell saturation density. Exp Cell Res 1998;239:332-43. Jacque CM, Vinner C, Kujas M, Raoul M, Racadot J, Baumann NA. Determination of glial fibrillary acidic protein (GFAP) in human brain tumors. J Neurol Sci 1978;35:147-55. Tascos NA, Parr J, Gonatas NK. Immunocytochemical study of the glial fibrillary acidic protein in human neoplasms of the central nervous system. HUM PATHOL 1982;13:454-8. van der Meulen JD, Houthoff HJ, Ebels EJ. Glial fibrillary acidic protein in human gliomas. Neuropathol Appl Neurobiol 1978;4:177-90. Wilhelmsson U, Eliasson C, Bjerkvig R, Pekny M. Loss of GFAP expression in high-grade astrocytomas does not contribute to tumor development or progression. Oncogene 2003;22:3407-11. Mendez MG, Kojima S, Goldman RD. Vimentin induces changes in cell shape, motility, and adhesion during the epithelial to mesenchymal transition. FASEB J 2010;24:1838-51. Domagala W, Lasota J, Dukowicz A, et al. Vimentin expression appears to be associated with poor prognosis in node-negative ductal NOS breast carcinomas. Am J Pathol 1990;137:1299-304. Wei LC, Shi M, Cao R, Chen LW, Chan YS. Nestin small interfering RNA (siRNA) reduces cell growth in cultured astrocytoma cells. Brain Res 2008;1196:103-12. Lu WJ, Lan F, He Q, et al. Inducible expression of stem cell associated intermediate filament nestin reveals an important role in glioblastoma carcinogenesis. Int J Cancer 2011;128:343-51. Pan Y, Jing R, Pitre A, Williams BJ, Skalli O. Intermediate filament protein synemin contributes to the migratory properties of astrocytoma cells by influencing the dynamics of the actin cytoskeleton. FASEB J 2008;22:3196-206. Pitre A, Davis N, Paul M, Orr AW, Skalli O. Synemin promotes AKTdependent glioblastoma cell proliferation by antagonizing PP2A. Mol Biol Cell 2012;23:1243-53. Henson JW, Ulmer S, Harris GJ. Brain tumor imaging in clinical trials. AJNR Am J Neuroradiol 2008;29:419-24. Ståhlberg A, Andersson D, Aurelius J, et al. Defining cell populations with single-cell gene expression profiling: correlations and identification of astrocyte subpopulations. Nucleic Acids Res 2011;39:e24. Eng LF, Ghirnikar RS, Lee YL. Glial fibrillary acidic protein: GFAPthirty-one years (1969-2000). Neurochem Res 2000;25:1439-51. Pekny M, Johansson CB, Eliasson C, et al. Abnormal reaction to central nervous system injury in mice lacking glial fibrillary acidic protein and vimentin. J Cell Biol 1999;145:503-14. Eliasson C, Sahlgren C, Berthold CH, et al. Intermediate filament protein partnership in astrocytes. J Biol Chem 1999;274: 23996-4006. Wilhelmsson U, Li L, Pekna M, et al. Absence of glial fibrillary acidic protein and vimentin prevents hypertrophy of astrocytic processes and improves post-traumatic regeneration. J Neurosci 2004;24:5016-21. Li L, Lundkvist A, Andersson D, et al. Protective role of reactive astrocytes in brain ischemia. J Cereb Blood Flow Metab 2008;28: 468-81. Lundkvist A, Reichenbach A, Betsholtz C, Carmeliet P, Wolburg H, Pekny M. Under stress, the absence of intermediate filaments from
8 Muller cells in the retina has structural and functional consequences. J Cell Sci 2004;117:3481-8. [33] Macauley SL, Pekny M, Sands MS. The role of attenuated astrocyte activation in infantile neuronal ceroid lipofuscinosis. J Neurosci 2011;31:15575-85.
O. Skalli et al. [34] Kraft AW, Hu X, Yoon H, et al. Attenuating astrocyte activation accelerates plaque pathogenesis in APP/PS1 mice. FASEB J 2013;27:187-98. [35] Lepekhin EA, Eliasson C, Berthold CH, Berezin V, Bock E, Pekny M. Intermediate filaments regulate astrocyte motility. J Neurochem 2001;79:617-25.
www.elsevier.com/locate/humpath
Original contribution
Astrocytoma grade IV (glioblastoma multiforme) displays 3 subtypes with unique expression profiles of intermediate filament proteins☆,☆☆ Omar Skalli PhD a,⁎, Ulrika Wilhelmsson PhD b , Charlotte Örndahl MD, PhD c , Boglarka Fekete MD d , Kristina Malmgren MD, PhD d , Bertil Rydenhag MD, PhD d , Milos Pekny MD, PhD b,⁎ a
Department of Biological Sciences, The University of Memphis, Memphis, TN 38152, USA Center for Brain Repair and Rehabilitation, Department of Clinical Neuroscience and Rehabilitation, Institute of Neuroscience and Physiology, Sahlgrenska Academy, University of Gothenburg, Medicinaregatan 9A, 41390 Göteborg, Sweden c Department of Pathology, Institute of Biomedicine, Sahlgrenska Academy, University of Gothenburg, Medicinaregatan 9A, 41390 Göteborg, Sweden d Department of Clinical Neuroscience and Rehabilitation, Institute of Neuroscience and Physiology, Sahlgrenska Academy, University of Gothenburg, 41390 Göteborg, Sweden b
Received 21 December 2012; revised 25 March 2013; accepted 27 March 2013
Keywords: Astrocytoma grade IV; Glioblastoma multiforme; Intermediate filament proteins; GFAP; Vimentin; Nestin; Synemin
Summary Astrocytoma grade IV (glioblastoma multiforme) is the most common and most malignant tumor of the central nervous system and is currently noncurable. Here, we have examined a populationbased cohort of 47 patients with grade IV astrocytoma, who underwent tumor surgery at Sahlgrenska University Hospital in Sweden and who survived after surgery for less than 200 days (short survivors, 28 patients) and more than 500 days (long survivors, 19 patients). For each tumor, we ascertained information on patient age, sex, tumor location, oncological treatment, and survival after surgery. The analysis of the tumor volume and the extent of tumor resection (incomplete versus complete resection of the macroscopic tumor) was made retrospectively from the preoperative radiological investigations and, when available, also from postoperative radiology. We performed semiquantitative immunohistochemical evaluation of the presence of intermediate filament (nanofilament) proteins glial fibrillary acidic protein, vimentin, nestin, and synemin in tumor cells. The intermediate filament system helps cells and tissues to cope with various types of stress, and thus, it might affect the malignant potential of grade IV astrocytoma. We propose a subclassification of astrocytomas grade IV with respect to the expression of the intermediate filament proteins glial fibrillary acidic protein, vimentin, nestin, and synemin, namely,
☆
Conflict of interests: The authors declare no conflict of interests. Funding: This work was supported by research grants from Swedish Medical Research Council (project 11548), AFA Research Foundation, ALF Göteborg (project 11392), Sten A. Olsson Foundation for Research and Culture, Söderberg Foundations, the Free Mason Foundation, Amlöv's Foundation, Health Medical Care Committee of the Västra Götal and Föreningen Margaretahemmet, NanoNet COST Action (BM1002), the EU FP 7 Program EduGlia (237956), the EU FP 7 Program TargetBraIn (279017), and research grant from the University of Memphis. ⁎ Corresponding authors. Omar Skalli is to be contacted at Department of Life Sciences, University of Memphis, Memphis, TN 38152, USA. Milos Pekny, Center for Brain Repair and Rehabilitation, Department of Clinical Neuroscience and Rehabilitation, Institute of Neuroscience and Physiology, Sahlgrenska Academy at the University of Gothenburg, Medicinaregatan 9A, SE-413 90 Gothenburg, Sweden. E-mail addresses: [email protected] (O. Skalli), [email protected] (M. Pekny). ☆☆
0046-8177/$ – see front matter © 2013 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.humpath.2013.03.013
2
O. Skalli et al. type A, B, and C. Our results suggest that the expression of the intermediate filament proteins glial fibrillary acidic protein, vimentin, nestin, and synemin is coregulated in grade IV astrocytomas. The expression patterns of the intermediate filament proteins in astrocytoma type A, B, and C might have biological and clinical significance. © 2013 Elsevier Inc. All rights reserved.
1. Introduction Astrocytoma grade IV (glioblastoma multiforme) is the most common primary brain tumor with a median survival of only 6 to 12 months. These tumors originate either from astrocytes that have accumulated mutations and de-differentiated or from neural stem/progenitor cells, which, in animal models, were localized to the subventricular zone [1,2]. The cytoplasm of astrocytoma cells contains intermediate filaments (IFs) (known also as nanofilaments), the key cellular system that helps cells and tissues cope with various types of stress [3]. IFs of astrocytes are composed of glial fibrillary acidic protein (GFAP) and vimentin and, in the case of immature or reactive astrocytes, also of nestin and synemin [4-7]. These IF proteins are also present in astrocytomas [4,8,9], and a number of studies addressed a possible connection between astrocytoma cell biology and their IF system. Some in vitro studies have suggested a negative correlation between GFAP protein levels and astrocyte dedifferentiation and transformation. Suppression of GFAP in U-251 human astrocytoma cells by GFAP antisense complementary DNA enhanced cell proliferation and anchorage-independent growth in soft agar [10] and decreased the ability of these cells to extend cellular processes in cocultures with neurons [11]. Expression of GFAP complementary DNA in astrocytoma cells slowed proliferation, increased the size of cellular processes, reduced soft agar colonies growth [12], and reduced tumorigenicity in vivo [13]. We showed that primary astrocytes derived from GFAP−/− mice grew more quickly and reached higher saturation cell densities than astrocytes from wild-type mice [14]. High-grade astrocytomas are often GFAP negative [1517], and in a previous study, we have evaluated the progression of astrocytomas in a GFAP negative context [18]. The results showed that astrocytomas induced in mice of the p53−/− background by prenatal exposure to ethylnitrosourea developed at a comparable rate and showed similar clinical progression and histopathologic features between wild-type and GFAP−/− mice [18], but it remains to be seen whether similar results would be obtained with other mutations prevalent in grade IV astrocytomas, such as those affecting the epidermal growth factor receptor. The role of vimentin in astrocytomas remains unclear, but vimentin is important for the epithelial-mesenchymal transition [19] and it associates with aggressive breast cancer subtypes [20]. It was suggested that nestin contributes to the aggressive behavior of astrocytoma cells as nestin small interfering RNA reduced growth of astrocytoma cells in vitro
[21] and intratumor injection of nestin short hairpin RNA (shRNA) into mouse astrocytomas slowed the tumor growth [22]. We previously demonstrated the presence of synemin in both astrocytoma cell lines and human astrocytomas [4] and suggested that synemin contributes to proliferation, soft agar colony formation, and migratory properties of astrocytoma cells [23,24]. Although many astrocytomas grade IV express the IF proteins GFAP, vimentin, nestin, and synemin, their coexpression and its potential prognostic value have not been studied. Here, we have used tissue from 47 grade IV astrocytomas and evaluated the coexpression of these IF proteins and made the initial assessment whether such a coexpression could be of a prognostic value.
2. Materials and methods 2.1. The patient cohort Between 2004 and 2008, 229 patients underwent neurosurgery at Sahlgrenska University Hospital in Gothenburg, Sweden, with a histopathologic diagnosis of grade IV astrocytoma. In the whole cohort, the median survival was 0.82 years, and the mean survival was 1.06 years. From this cohort, we selected 2 groups of patients (47 patients in total): the patients who survived less than 200 days after the operation (28 patients; “short survivors”) and the patients who survived more than 500 days (19 patients; “long survivors”) and for whom information about the macroscopic completeness of the tumor resection and on additional oncological therapy was available. The tumor tissue was collected in accordance with the ethical guidelines of the Sahlgrenska Academy University Hospital. Diagnosis was based on the criteria defined in the World Health Organization classification of tumors of the central nervous system. The analysis of tumor volume and the extent of tumor resection performed was made retrospectively from the preoperative radiological investigations and, when available, also from the postoperative radiology. No systematic immediate postoperative imaging was performed, and the evaluation of the extent of the resection of the macroscopic tumor was based on the sum of information obtained from the surgeons' visual evaluation and by postoperative imaging when performed. The volume was calculated from the preoperative coronal and axial images as reported before [25]. Data on patient age, sex, and survival time were collected as was information on whether a patient received postoperative oncological treatment.
Astrocytoma grade IV subtypes with unique expression profiles
2.2. Antibodies, immunohistochemistry, and light microscopy
with the 2-tailed Fisher's exact test. Comparisons of mean and median were performed with t test and Mann-Whitney rank sum test, respectively. For all these tests, differences were deemed significant when P b .05.
Goat anti-synemin antibodies were raised against bacterially expressed human α-synemin C-terminal domain and purified as described [4]. The other primary antibodies were commercially available and included: mouse anti-human vimentin clone V9 (Sigma-Aldrich Sweden, Stockholm, Sweden), mouse antiGFAP clone GA5 (Sigma-Aldrich), and mouse anti-human nestin clone 10C2 (Merck Millipore, Billerica, MA, USA). For immunohistochemistry, deparaffinized, 4-μm-thick sections of human tissues were rehydrated with phosphatebuffered saline, and endogenous peroxidase activity was blocked for 30 minutes with 0.3% H2O2 in methanol. After a 5-minute phosphate-buffered saline wash, sections were incubated with 10% normal horse serum for 30 minutes and incubated for 30 minutes with the primary antibodies diluted at 1:200 (goat anti-synemin), 1:500 (mouse antinestin), and 1:5000 (mouse anti-vimentin and anti-GFAP). Antibody binding was detected with the avidin-biotinperoxidase method, using the Vectastain reagent kit (Vector Laboratories, Burlingame, CA, USA) and biotinylated pan-specific secondary antibody (Vector Laboratories). The peroxidase activity then was revealed with diaminobenzidine (Vector Laboratories), and nuclear counterstain was performed with Mayer's hematoxylin. For the purpose of quantitation, for each antibody, all 47 cases were stained at the same time using the same solutions and identical conditions. Light microscopic observations were performed with a Nikon Eclipse 800 (Nikon Instruments, Melville, NY, USA).
3. Results 3.1. There are 3 subtypes of astrocytomas grade IV with respect to the expression of the IF proteins GFAP, vimentin, nestin, and synemin To evaluate whether in astrocytomas grade IV the IF proteins GFAP, vimentin, nestin, and synemin are coexpressed and if their individual and/or collective expression patterns could be of prognostic value, we performed histopathologic and immunohistochemical analysis of these 4 IF proteins in 47 grade IV astrocytomas (28 patients surviving b200 days and 19 patients surviving N500 days after operation). Histopathologic assessment of hematoxylineosin–stained sections from the tumor tissue showed the defining characteristics of astrocytoma grade IV in all the patients, including cellular heterogeneity, mitotic figures, abundant necrosis, and/or vascular proliferation. Immunohistochemical analysis on consecutive sections stained with antibodies against GFAP, vimentin, nestin, or synemin demonstrated that all cases examined contained tumor cells positive for each of these proteins with the immunoreactivity being homogenous within the tumor tissue. There were, however, differences in the staining intensities of each IF protein between tumors. Semiquantitative evaluation of the staining intensities of the tumor cells was performed using a scale of 1 (weakest) to 6 (strongest) (Table 1). For the purpose of statistical analysis, grading scores were ultimately divided into 2 categories: weak (score 1-3) and strong (score 4-6). This analysis revealed that most grade IV astrocytomas (62%) were strongly positive for nestin, whereas a smaller proportion of tumors strongly stained for vimentin (19%), GFAP (25%), and synemin (23%), respectively. Two-tailed Fisher's exact test analysis was performed to determine whether there were correlations between the staining intensities for the different IF proteins present in grade IV astrocytomas. The P values obtained ranged from
2.3. Semiquantitative evaluation and statistical analysis For each antibody, slides with the weakest and strongest staining in the tumor cells were selected to define the 2 extremes of the grading scale (1 and 6, respectively). These slides served as reference to define intermediate staining intensities (2-5). All slides were then graded relative to these 6 reference slides. Data were analyzed with the GraphPad Prism 5 software. Association between groups and outcomes was analyzed Table 1 Intensity score
GFAP Vimentin Nestin Synemin
3
Intensity of IF protein immunoreactivity in individual tumors No. of tumors Weak
Mean (median) scores Strong
All tumors (n = 47)
Short survival (n = 15)
Long survival (n = 10)
1
2
3
4
5
6
Score 1-6
Score 1-3
Score 4-6
Score 1-3
Score 4-6
Score 1-3
Score 4-6
3 14 0 0
20 16 7 23
12 8 11 13
8 7 10 5
2 1 11 5
2 1 8 1
2.83 (3.0) 2.32 (2.0) 4.04 (4.0) 2.89 (3.0)
2.25 1.84 2.61 2.36
4.50 (4.0) 4.33 (4.0) 4.93 (5.0) 4.63 (5.0)
2.25 (2.0) 2.08 (2.0) 2.40 (2.0) 2.31 (2.0)
5.00 4.00 4.90 5.00
2.12 (2.0) 1.89 (2.0) 2.50 (2.5) 2.37 (2.0)
4.50 (4.5) 4.00 (4.0) 4.75 (4.5) 4.00 (4.0)
(2.0) (2.0) (3.0) (2.0)
(5.0) (4.0) (5.0) (5.0)
NOTE. Patient survival after surgery: short survival, less than 200 days; long survival, more than 500 days. Patients who did not receive postsurgical adjuvant therapy or had complete tumor removal were not included the short and long survival groups.
4
O. Skalli et al. Table 2 Correlation between the immunostaining intensities (weak or strong) of pairs of IF proteins in astrocytoma grade IV tissue (n = 47 tumors) IF proteins GFAP GFAP GFAP Nestin Nestin Synemin
3.2. Expression of IF proteins GFAP, vimentin, nestin, and synemin by astrocytoma grade IV cells and the patient survival after surgery
P Synemin Vimentin Nestin Synemin Vimentin Vimentin
.0026* .0004* .0167* .0331* .0083* .0225*
NOTE. P values were obtained by analyzing contingency tables with the 2-tailed Fisher's exact test. Asterisks denote significant correlation between the staining intensities of the respective IF proteins.
0.0004 to 0.0331 for each possible pair of IF proteins (Table 2). This demonstrates that there is a high degree of correlation between the staining intensities, of each possible pair of IF proteins present in astrocytoma grade IV tissues, namely, GFAP and synemin, GFAP and vimentin, GFAP and nestin, nestin and synemin, nestin and vimentin, and synemin and vimentin (Table 2). These correlations were further evidenced by the finding that 3 major patterns of IF protein expression exist in the grade IV astrocytomas that we have examined. In the first pattern (30% of the tumors; pattern A), GFAP, synemin, and nestin were all expressed at high levels (Fig. 1). The opposite pattern, that is, weak staining for each of the four IF proteins examined, was observed in 36% of the tumors (pattern B) (Fig. 2). Finally, the third pattern (34% of the tumors; pattern C) was characterized by strong nestin staining but low GFAP, vimentin, and synemin staining (Fig. 3). The average age of patients with pattern B was lower than of those with pattern C (51 ± 3.0 years versus 59 ± 1.6 years, respectively; P b .05). There was no association between pattern A, B, or C and tumor volume (data not shown).
Next, we examined whether there was an association between individual IF proteins or IF protein patterns and patient survival after surgery. As mentioned above, 2 groups of patients were considered in our study, one with survival less than 200 days (28 patients) and the other with survival of more than 500 days (19 patients). The 2 groups did not differ with respect to the sex composition, but the average age of patients in the longer survival group was lower than in the shorter survival group (50 ± 2.8 and 57 ± 1.5 years, respectively; P = .0190). The average tumor volume in the longer survival group was smaller than in the shorter survival group (28 cm3 ± 5.5 versus 48 cm3 ± 7.1, respectively; P = .0403). Although most patients received radiotherapy and/or chemotherapy after surgery, 13 patients did not receive any of these treatment modalities, for a variety of reasons. As expected, the absence of postsurgical radiotherapy and/or chemotherapy was associated with poor survival (P = .0005; Table 3): all these 13 patients belonged to the short survival group and, therefore, were excluded from further analysis. Complete resection of the macroscopic tumor was achieved in 11 patients (who also received postsurgical oncological treatment), and it had a profound effect on the survival: 2-tailed Fisher's exact test analysis revealed that patients who received complete resection of the macroscopic tumor were more likely to belong to the long survivor group than those for whom the complete macroscopic resection could not be achieved (36 patients, P = .0034; Table 3). The number of patients in which complete macroscopic tumor removal was achieved was small (11 patients in total) and hindered meaningful statistical analysis; therefore, for the purpose of correlative analysis between survivor groups and IF protein
Fig. 1 Immunohistochemical staining of consecutive sections of a grade IV astrocytoma with GFAP, nestin, and synemin antibodies, as indicated on top of each picture. HE, hematoxylin eosin staining of a parallel section. Tumor cells in this case strongly stain with each of the 3 antibodies used. The figure shows pattern A, which represents 30% of the 47 tumors examined. Bar, 50 μm.
Astrocytoma grade IV subtypes with unique expression profiles
5
Fig. 2 Immunohistochemical staining of consecutive sections of a grade IV astrocytoma with GFAP, vimentin, nestin, and synemin antibodies, as indicated on top of each picture. HE, hematoxylin eosin staining of a parallel section. Tumor cells in this case weakly stain with each of the 4 antibodies used. The picture shows pattern B, which represents 36% of the 47 tumors examined. Note that positive controls for vimentin, nestin, and synemin staining are represented by blood vessels and/or glomeruloids and for GFAP by reactive astrocytes with stellate morphology. Bar, 50 μm.
expression, we focused on patients with incomplete tumor resection excluding those who did not receive postsurgical treatment (25 patients in total). We examined with the 2-tailed Fisher's exact test the correlation between individual IF protein staining intensity (weak or strong; see above for definition) or A, B, and C staining pattern and affiliation to short and long survivor groups. No association with survival was found (Table 4).
4. Discussion 4.1. Three subtypes of astrocytoma grade IV defined by the IF protein expression pattern The presence of individual IF proteins GFAP, vimentin, nestin, or synemin in the tumor cells of astrocytoma grade IV was previously reported [4,8,9,15-17]. This study, however, is the first to examine the coexpression of these 4 IF proteins in grade IV astrocytomas. Our populationbased cohort of 47 patients with astrocytoma grade IV clustered into tumor subtypes A, B, and C based on the expression of the IF proteins GFAP, vimentin, nestin, and
synemin: subtype A with high expression of GFAP, nestin, and synemin; subtype B with low expression of all 4 IF proteins; and subtype C with high expression of nestin and low expression of GFAP, vimentin, and synemin. Interestingly, these findings demonstrate that neither low GFAP nor high nestin is universal characteristic of grade IV astrocytomas because approximately one-third of these tumors exhibited either strong GFAP and/or weak nestin immunostaining. Synemin presence in astrocytomas was reported in an earlier study of 4 grade IV astrocytomas [4]. The present study is the first to examine synemin expression in a larger series of grade IV astrocytomas and suggests that about a fourth of these tumors strongly express this IF protein.
4.2. Coregulation of GFAP, vimentin, nestin, and synemin—from astrocytes to grade IV astrocytoma cells The existence of these 3 astrocytoma grade IV subtypes with a distinct correlation between the staining intensities of individual IF proteins in the tumor cells suggests a
Fig. 3 Immunohistochemical staining of consecutive sections of a grade IV astrocytoma with GFAP, vimentin, nestin, and synemin antibodies, as indicated on top of each picture. HE, hematoxylin eosin staining of a parallel section. Tumor cells in this case are strongly stained with nestin antibodies but weakly stained with vimentin, GFAP, and synemin antibodies. The figure shows pattern C, which represents 34% of the 47 tumors examined. Bar, 50 μm.
6
O. Skalli et al. Table 3 Association between survival and removal of the macroscopic tumor (contingency analysis 1) and survival and postsurgical therapy (contingency analysis 2) Contingency analysis 1 a c
Complete resection Incomplete resection c Total a b c d
N500 9 10 19
b200 2 26 28
Total 11 36 47
Contingency analysis 2 b d
Treatment after surgery No treatment after surgery d Total
4.3. IF proteins and the ability of cells to cope with stress The IF protein up-regulation, in particular, the upregulation of GFAP, the main constituent of astrocyte IFs, is a hallmark of astrocyte activation and reactive gliosis in the whole range of neuropathologies including brain and spinal cord trauma and infections; stroke; epilepsy; some primary and secondary tumors of the brain and spinal cord; and neurodegenerative diseases such as Alzheimer disease, Batten disease, or multiple sclerosis [7,27]. In mice, combined genetic deficiency of GFAP and vimentin leads to attenuated astrocyte activation and reactive gliosis [28,29]. The mice deficient for both GFAP and vimentin exhibit reduced glial scar formation and decreased ability to cope Table 4
19 0 19
15 13 28
34 13 47
4.4. IF proteins and cell motility Astrocytoma grade IV cells show remarkable ability to migrate within the brain tissue. In fact, at the onset of the
Correlation between survival and the immunostaining intensities and survival and A, B, and C staining patterns of IF proteins Pc
IF staining intensity
Outcome a
Vimentin (weak vs strong) GFAP (weak vs strong) Nestin (weak vs strong) Synemin (weak vs strong) IF staining pattern (A vs B+C) b IF staining pattern (B vs A+C) b IF staining pattern (C vs A+B) b
Survival Survival Survival Survival Survival Survival Survival
c
Total
with acute or chronic cellular stress of diverse kind, for example, in acute neurotrauma [28,30], focal brain ischemia [31], retinal injury [32], Batten disease [33], or Alzheimer disease [34]. In this study, we show the existence of 3 subtypes of astrocytoma grade IV, with respect to the immunoreactivity pattern of the IF proteins GFAP, vimentin, nestin, and synemin. Using a rather limited number of astrocytoma grade IV tumors, we have made the first attempt to investigate whether these tumor subtypes associate with different survival and, therefore, could be of prognostic value. For example, it is possible to speculate that—similarly to astrocytes lacking GFAP and vimentin—the tumor cells of the B subtype of astrocytoma grade IV with low expression of IF proteins GFAP, vimentin, nestin, and synemin might be less capable of coping with ischemic/hypoxic, osmotic, and other stresses and, consequently, allow longer survival of the patients after surgery compared with the A or C subtypes. Using a limited cohort of 25 patients, who were comparable with respect to the extent of the removal of the macroscopic tumor and postsurgical oncological treatment, we have not seen any such difference. A larger prospective study might be needed to address this.
Variables compared
b
b200
Association between removal of the macroscopic tumor and patient survival after surgery (P = .0034; 2-tailed Fisher's exact test). Association between additional therapy after surgery (radiotherapy and/or chemotherapy) and patient survival after surgery (P = .0005). Resection of the macroscopic tumor. Radiotherapy and/or chemotherapy treatment after surgery.
coregulation of the expression of GFAP, vimentin, nestin, and synemin. Using gene expression profiling by single cell quantitative real-time polymerase chain reaction, we recently reported that the expression of GFAP, vimentin, and nestin messenger RNA is coregulated in individual astrocytes in primary cultures [26]. Reactive astrocytes and astrocytes in culture (for review see [6]) as well as cells of astrocytoma grade IV examined in this study, all co-express the same IF proteins GFAP, vimentin, nestin, and synemin implying that the expression of these IF proteins is coregulated both in astrocytes and in grade IV astrocytoma cells.
a
N500
(short vs (short vs (short vs (short vs (short vs (short vs (short vs
long) long) long) long) long) long) long)
Patient survival after surgery: short survival, less than 200 days; long survival, more than 500 days. See Results for the description of the staining patterns A, B, and C. P values were obtained by analyzing contingency tables with the 2-tailed Fisher's exact test.
.6265 1.0000 .2406 1.0000 1.0000 .2406 .4018
Astrocytoma grade IV subtypes with unique expression profiles clinical symptoms, most patients have already intracerebral metastases of the primary tumor. These tumor “guerrilla cells” are highly resistant to chemotherapy and radiotherapy. The tumors that arise from these cells contribute to the very poor prognosis of grade IV astrocytoma even in those patients in which the macroscopic tumor has been successfully removed by surgery. We have previously demonstrated that astrocytes devoid of GFAP or vimentin show reduced migratory potential and that the cell migration is further reduced in astrocytes devoid of both GFAP and vimentin [35]. Astrocytes devoid of GFAP and vimentin cannot produce IFs because neither of their remaining IF proteins—nestin and synemin—can selfpolymerize or copolymerize into IFs [5,29]. Similarly, we previously demonstrated that synemin contributes to migratory properties of astrocytoma cells [23] and that it is a positive regulator of their proliferative capacity [24]. Thus, it is interesting to speculate that the subclass B astrocytoma grade IV with low expression of all 4 IF proteins might exhibit reduced migratory potential of the tumor cells both before and after surgery. Such a hypothesis could be tested in a prospective study in which the tumor spread can be assessed by magnetic resonance imaging both preoperatively and postoperatively.
[11]
[12]
[13]
[14]
[15]
[16]
[17] [18]
[19]
[20]
Acknowledgments The authors thank Anna Stokowska for her advice on the statistical analysis.
[21]
[22]
[23]
References [1] Chen J, McKay RM, Parada LF. Malignant glioma: lessons from genomics, mouse models, and stem cells. Cell 2012;149:36-47. [2] Watkins S, Sontheimer H. Unique biology of gliomas: challenges and opportunities. Trends Neurosci 2012;35:546-56. [3] Pekny M, Lane EB. Intermediate filaments and stress. Exp Cell Res 2007;313:2244-54. [4] Jing R, Pizzolato G, Robson RM, Gabbiani G, Skalli O. Intermediate filament protein synemin is present in human reactive and malignant astrocytes and associates with ruffled membranes in astrocytoma cells. Glia 2005;50:107-20. [5] Jing R, Wilhelmsson U, Goodwill W, et al. Synemin is expressed in reactive astrocytes in neurotrauma and interacts differentially with vimentin and GFAP intermediate filament networks. J Cell Sci 2007;120:1267-77. [6] Pekny M, Pekna M. Astrocyte intermediate filaments in CNS pathologies and regeneration. J Pathol 2004;204:428-37. [7] Pekny M, Nilsson M. Astrocyte activation and reactive gliosis. Glia 2005;50:427-34. [8] Herpers MJ, Ramaekers FC, Aldeweireldt J, Moesker O, Slooff J. Coexpression of glial fibrillary acidic protein- and vimentin-type intermediate filaments in human astrocytomas. Acta Neuropathol 1986;70:333-9. [9] Yung WK, Luna M, Borit A. Vimentin and glial fibrillary acidic protein in human brain tumors. J Neurooncol 1985;3:35-8. [10] Rutka JT, Hubbard SL, Fukuyama K, Matsuzawa K, Dirks PB, Becker LE. Effects of antisense glial fibrillary acidic protein complementary
[24]
[25] [26]
[27] [28]
[29]
[30]
[31]
[32]
7 DNA on the growth, invasion, and adhesion of human astrocytoma cells. Cancer Res 1994;54:3267-72. Weinstein DE, Shelanski ML, Liem RK. Suppression by antisense mRNA demonstrates a requirement for the glial fibrillary acidic protein in the formation of stable astrocytic processes in response to neurons. J Cell Biol 1991;112:1205-13. Rutka JT, Smith SL. Transfection of human astrocytoma cells with glial fibrillary acidic protein complementary DNA: analysis of expression, proliferation, and tumorigenicity. Cancer Res 1993;53:3624-31. Toda M, Miura M, Asou H, Sugiyama I, Kawase T, Uyemura K. Suppression of glial tumor growth by expression of glial fibrillary acidic protein. Neurochem Res 1999;24:339-43. Pekny M, Eliasson C, Chien CL, et al. GFAP-deficient astrocytes are capable of stellation in vitro when cocultured with neurons and exhibit a reduced amount of intermediate filaments and an increased cell saturation density. Exp Cell Res 1998;239:332-43. Jacque CM, Vinner C, Kujas M, Raoul M, Racadot J, Baumann NA. Determination of glial fibrillary acidic protein (GFAP) in human brain tumors. J Neurol Sci 1978;35:147-55. Tascos NA, Parr J, Gonatas NK. Immunocytochemical study of the glial fibrillary acidic protein in human neoplasms of the central nervous system. HUM PATHOL 1982;13:454-8. van der Meulen JD, Houthoff HJ, Ebels EJ. Glial fibrillary acidic protein in human gliomas. Neuropathol Appl Neurobiol 1978;4:177-90. Wilhelmsson U, Eliasson C, Bjerkvig R, Pekny M. Loss of GFAP expression in high-grade astrocytomas does not contribute to tumor development or progression. Oncogene 2003;22:3407-11. Mendez MG, Kojima S, Goldman RD. Vimentin induces changes in cell shape, motility, and adhesion during the epithelial to mesenchymal transition. FASEB J 2010;24:1838-51. Domagala W, Lasota J, Dukowicz A, et al. Vimentin expression appears to be associated with poor prognosis in node-negative ductal NOS breast carcinomas. Am J Pathol 1990;137:1299-304. Wei LC, Shi M, Cao R, Chen LW, Chan YS. Nestin small interfering RNA (siRNA) reduces cell growth in cultured astrocytoma cells. Brain Res 2008;1196:103-12. Lu WJ, Lan F, He Q, et al. Inducible expression of stem cell associated intermediate filament nestin reveals an important role in glioblastoma carcinogenesis. Int J Cancer 2011;128:343-51. Pan Y, Jing R, Pitre A, Williams BJ, Skalli O. Intermediate filament protein synemin contributes to the migratory properties of astrocytoma cells by influencing the dynamics of the actin cytoskeleton. FASEB J 2008;22:3196-206. Pitre A, Davis N, Paul M, Orr AW, Skalli O. Synemin promotes AKTdependent glioblastoma cell proliferation by antagonizing PP2A. Mol Biol Cell 2012;23:1243-53. Henson JW, Ulmer S, Harris GJ. Brain tumor imaging in clinical trials. AJNR Am J Neuroradiol 2008;29:419-24. Ståhlberg A, Andersson D, Aurelius J, et al. Defining cell populations with single-cell gene expression profiling: correlations and identification of astrocyte subpopulations. Nucleic Acids Res 2011;39:e24. Eng LF, Ghirnikar RS, Lee YL. Glial fibrillary acidic protein: GFAPthirty-one years (1969-2000). Neurochem Res 2000;25:1439-51. Pekny M, Johansson CB, Eliasson C, et al. Abnormal reaction to central nervous system injury in mice lacking glial fibrillary acidic protein and vimentin. J Cell Biol 1999;145:503-14. Eliasson C, Sahlgren C, Berthold CH, et al. Intermediate filament protein partnership in astrocytes. J Biol Chem 1999;274: 23996-4006. Wilhelmsson U, Li L, Pekna M, et al. Absence of glial fibrillary acidic protein and vimentin prevents hypertrophy of astrocytic processes and improves post-traumatic regeneration. J Neurosci 2004;24:5016-21. Li L, Lundkvist A, Andersson D, et al. Protective role of reactive astrocytes in brain ischemia. J Cereb Blood Flow Metab 2008;28: 468-81. Lundkvist A, Reichenbach A, Betsholtz C, Carmeliet P, Wolburg H, Pekny M. Under stress, the absence of intermediate filaments from
8 Muller cells in the retina has structural and functional consequences. J Cell Sci 2004;117:3481-8. [33] Macauley SL, Pekny M, Sands MS. The role of attenuated astrocyte activation in infantile neuronal ceroid lipofuscinosis. J Neurosci 2011;31:15575-85.
O. Skalli et al. [34] Kraft AW, Hu X, Yoon H, et al. Attenuating astrocyte activation accelerates plaque pathogenesis in APP/PS1 mice. FASEB J 2013;27:187-98. [35] Lepekhin EA, Eliasson C, Berthold CH, Berezin V, Bock E, Pekny M. Intermediate filaments regulate astrocyte motility. J Neurochem 2001;79:617-25.