Apoptosis and Proliferation Markers in Diffusely Infiltrating Astrocytomas: Profiling of 17 Molecules (original) (raw)

Abstract

Caspases and inhibitor of apoptosis proteins (IAPs) are antagonizing key apoptosis regulators. Limited studies of a few IAPs indicated their roles in astrocytomas. However, the overall expression status and significance of apoptosis regulators in astrocytomas is not clear. We examined the expression profile of the caspases (CASP3, 6, 7, 8, 9, 10, and 14), APAF1, SMAC, BCL2, the IAPs (BIRC5/survivin, CIAP1, CIAP2, XIAP, and LIVIN), and the proliferation markers Ki67 and PHH3 in 78 diffusely infiltrating astrocytomas and 24 normal brain samples by immunohistochemistry. Western blotting for major caspases and IAPs and reverse transcription-polymerase chain reaction analyses for IAPs were performed on a subset of 27 fresh samples. Our data showed BIRC5 nuclear labeling index (BIRC5-N) was the apoptosis marker most significantly different in World Health Organization grade II to IV astrocytomas and most strongly associated with proliferative activity. Expression level of other apoptosis-related proteins was modest or low in astrocytomas and did not correlate significantly with tumor grade or proliferation. Apoptosis regulators and proliferation markers were not detected in astrocytes of normal brain by immunostaining. This expression profile suggested involvement of apoptosis regulators in astrocytoma tumorigenesis, but tumor progression was more closely associated with proliferative advantages of which BIRC5 nuclear expression appeared to be a manifestation.

Introduction

Diffusely infiltrating astrocytomas are the most common central nervous system neoplasms and comprise 3 clinicopathologic entities: diffuse astrocytoma (DA; World Health Organization [WHO] grade II), anaplastic astrocytoma (AA; WHO grade III), and glioblastoma multiforme (GBM; WHO grade IV) (1). This nomenclature and grading system correlated fairly well with their biologic behavior, although it can be influenced by a variety of factors and is sometimes unpredictable (1).

Proliferative activity is an important factor in evaluating diffusely infiltrating astrocytomas and has been linked to grade or survival in most studies, although not in others (10-12). Activated caspases proteolytically cleave an array of cellular proteins, ultimately resulting in the biochemically and morphologically distinct apoptotic cell death. The inhibitor of apoptosis proteins (IAPs) are the major intracellular antiapoptotic proteins that function by antagonizing caspases or other proapoptotic proteins as well as by participating in cell-cycle regulation (13,14). Apoptotic protease activating factor 1 (APAF1) and second mitochondrial activator of caspases (SMAC) are mitochondrial proteins released on cell death stimuli and are involved in caspase activation and inhibition of IAPs, respectively (15). A few studies examined the roles of these apoptosis regulators in astrocytomas (1). Most patients in this cohort did not receive prior treatment for astrocytoma, with one patient with AA having previous surgery for DA and 6 patients with GBM having previous surgery for GBM (3 of which received radiotherapy after first surgery). For hematoxylin and eosin and immunostaining, the specimens were formalin-fixed and paraffin-embedded. Snap-frozen tissues of 7 NBT samples and 20 astrocytoma samples from this cohort were used for Western blot and RT-PCR analysis. The NBT samples (most from frontal or temporal lobes) included 17 from autopsies (mostly accidental deaths) and 7 fresh nonneoplastic brain biopsies. The age of these control individuals ranged from 1 to 85 years (mean, 51 years) with a male to female ratio of 15:9. NBT samples displayed no obvious neuropathologic changes on routine histologic examination.

Immunohistochemistry

The following primary antibodies at indicated dilutions were used for IHC: CASP3 (rabbit polyclonal, 1:250; Santa Cruz Biotechnology, Santa Cruz, CA), CASP6 (goat polyclonal, 1:100; Santa Cruz), CASP7 (goat polyclonal, 1:100; Santa Cruz), CASP8 (goat polyclonal, 1:100; Santa Cruz), CASP9 (rabbit polyclonal, 1:100; Santa Cruz), CASP10 (goat polyclonal, 1:100; Santa Cruz), CASP14 (goat polyclonal, 1:300; Santa Cruz), CIAP1 (rabbit polyclonal, 1:100; Santa Cruz), CIAP2 (rabbit polyclonal, 1:100; Santa Cruz), X-linked inhibitor of apoptosis protein (XIAP; rabbit polyclonal, 1:300; ProteinTech Group, Inc., Chicago, IL), BIRC5 (Survivin) (rabbit polyclonal, 1:500; R and D Systems, Inc., Minneapolis, MN), LIVIN (goat polyclonal, 1:500; R and D), APAF1 (rabbit polyclonal, 1:100; Santa Cruz), SMAC (goat polyclonal, 1:300; Santa Cruz), BCL2 (mouse monoclonal, 1:200; Zymed Laboratories Inc., San Francisco, CA), PHH3 (rabbit polyclonal, 1:500; Upstate Group LLC, Charlottesville, VA), and Ki67 (MIB1) (mouse monoclonal, 1:100; DakoCytomation, Glostrup, Denmark). Biotinylated secondary antibodies were from Zymed Laboratories.

Four-micrometer sections were immunostained by standard labeled streptavidin-biotin protocol (reagents from Zymed Laboratories). Omission of primary antibodies was used as a control. Antigen retrieval was by high-pressure boiling in citrate buffer (pH 6.0) for 3 minutes. A procedure to block potential endogenous avidin binding activity (21) was used.

The first 14 immunomarkers in Table 1 displayed a cytoplasmic staining pattern with CIAP1 showing occasional nuclear staining. Distinct cytoplasmic and nuclear staining of BIRC5 was observed in astrocytomas and separately scored as BIRC5-C and BIRC5-N. Staining of Ki67 and PHH3 was exclusively nuclear. For cytoplasmic staining, a conventional 4-tiered scoring system (score 0-3) integrating intensity and extent of staining was adopted (22) based on thorough examination of the slides (tumor area ranged from around 20 mm2 to over 100 mm2). Nuclear labeling index of BIRC5 (BIRC5-N) (%) and Ki67 (%) was determined by counting positive cells in 1,000 tumor cells starting from the highest labeling region. The PHH3 positivity (positive cells/10 high-power fields at 400x) was determined by counting labeled cells in up to 50 consecutive high-power fields (analogous to mitotic figure [MF] count) starting from the highest labeling region (23). Immunostaining was scored by 2 pathologists independently and crosschecked.

TABLE 1.

Comparison of Apoptosis Regulator Immunostaining and Proliferation Activity in Diffusely Infiltrating Astrocytomas

TABLE 1.

Comparison of Apoptosis Regulator Immunostaining and Proliferation Activity in Diffusely Infiltrating Astrocytomas

Western Blot Analysis

Total proteins were extracted in the presence of protease inhibitor cocktails (Roche Diagnostics, Mannheim, Germany) from snap-frozen tissues quantitated by using the BCA kit (Pierce Biotechnology Inc., Rockford, IL) and resolved by 10% SDS polyacrylamide (Sigma, St. Louis, MO) gel electrophoresis. Proteins were electroblotted to PVDF membrane (Amersham Biosciences Ltd., Little Chalfont, U.K.) in CAPS buffer (pH 11.0) (Amresco, Solon, OH) and then incubated with block solution (5% nonfat milk, 0.1% Tween 20, in 1 × TBS; Sigma) at room temperature for 2 hours.

The primary antibodies to caspases and IAPs are listed in the "Immunohistochemistry" section with the following dilutions used for Western blot analysis: CASP3, 1:800; CASP6, 1:500; CASP7, 1:500; CASP8, 1:50; CASP9, 1:600; CASP10, 1:500; CIAP1, 1:800; CIAP2, 1:800; XIAP, 1:1000; BIRC5/survivin, 1:1000; LIVIN, 1:800. GAPDH (mouse monoclonal, clone 6C5, 1:10,000; Kangcheng, Shanghai, China) was used as an internal control. Horseradish peroxidase secondary antibodies were from Zymed. Primary and secondary antibodies were incubated at room temperature for 2 and 1.5 hours, respectively. Signals were detected by exposure to x-ray films after treatment with the SuperSignal enhanced chemiluminescence kit (Pierce Biotechnology Inc.).

Reverse Transcription-Polymerase Chain Reaction

Primers for IAPs were designed according to cDNA sequences in GenBank and synthesized by Invitrogen (Carlsbad, CA). The primers and product lengths were: CIAP1: 5′-TTGTCAACTTCAGATACCACTGGAG-3′,5′-CAAGGCAGATTTAACCACAGGTG-3′, 123 bp; CIAP2:5′-AGGGAAGAGGAGAGAGAAAGAGC-3′,5′-CGGCAGTTAGTAGACTATCCAGG-3′, 133 bp; BIRC5: 5′-GCAGTTTGAAGAATTAACCCTTG-3′,5′-CACTTTCTCCGCAGTTTCCTC-3′,121 bp; LIVIN: 5′-CCGTGTCCATCGTCTTTGTGC-3′, 5′-AACACAGTCCAGAACAGGCAGAGAG-3′, 196 bp; XIAP: 5′-GGGTTCAGTTTCAAGGACATTAAG-3′,5′-CGCCTTAGCTGCTCTTCAGTAC-3′, 182 bp. β-Actin was used as an internal control as described previously (24). Primers were designed to span introns to avoid false positivity from genomic DNA contamination. Total RNA from fresh tissue and cultured cells was isolated with the Trizol reagent (Invitrogen, Carlsbad, CA). Reverse transcription was carried out in 20 μL mixture containing 5 μg total RNA, 0.5 μg oligo(dT)18 primer, 2 μL of 10 mmol/L dNTP, 1 μL of 0.1 mol/L DTT, and 1 μL of M-Mulv reverse transcriptase (Fermentas, Hanover, MD) for 60 minutes at 42°C followed by 10 minutes at 72°C. PCR was carried out with Taq DNA polymerase (Takara, Japan) in 30 cycles of amplification (30seconds at 95°C, 30 seconds at appropriate annealing temperature for each primer set, and 35 seconds at 72°C) followed by 10 minutes at 72°C. PCR products were resolved by 2% agarose gel stained with the fluorescent dye GoldView (Beijing SBS Genetech Co., Ltd., Beijing, China) and visualized by scanning with the multiimager Typhoon 8600 (Amersham Pharmacia Biotech Inc., Piscataway, NJ) with excitation at 520 nm and emission at 580 nm.

Data Analysis

Statistical analysis was carried out by using Statistica software (StatSoft, Inc., Tulsa, OK). Spearman rank order correlation was used to examine correlations between variables. Intergroup differences were examined by using the nonparametric Mann-Whitney U test or Fisher exact test as appropriate. Multidimensional scaling (MDS) (25) was used for spatial representation of the astrocytomas to illustrate their relationship.

Results

General Expression Status of Apoptosis Regulators and Proliferation Markers Assessed by Immunohistochemistry

Immunostaining of the apoptosis regulators and proliferation markers varied from negative to strongly positive in the astrocytomas. Figure 1 illustrates representative positive immunostaining of several markers. Thorough examination of the slides did not reveal clearcut patterns of staining, although a general tendency of higher labeling at the invasive front and in tumor cells that were several cell layers away from the center of necrosis (in GBM) was observed. Positivity of apoptotic markers also appeared to be more readily discernible in areas of higher proliferation, but this varied from case to case. Comparison of immunostaining in the small percentage of patients with AA and those with GBM who had prior surgery or irradiation did not reveal a significant difference from those who had not.

FIGURE 1.

(A) Representative immunostaining of CASP3, CIAP1, CIAP2, BIRC5, Ki67, and PHH3 in diffusely infiltrating astrocytomas. Horseradish peroxidase (for first 5 markers with diaminobenzidine as chromogen) or alkaline phosphatase (for PHH3 with AP red [Zymed, San Francisco, CA] as chromogen) labeled streptavidin-biotin immunostaining with hematoxylin counterstain. Original magnification for each panel: 400×. (B) BIRC5 nuclear labeling in astrocytomas. (A) Single nuclear positive cell in a diffuse astrocytoma. (B) High percentage of BIRC5 nuclear labeling in a glioblastoma multiforme. Horseradish peroxidase labeled streptavidin-biotin immunostaining with hematoxylin counterstain. Original magnification for each panel: 1,000×.

The semiquantitative results of immunostaining of the astrocytomas are summarized in Table 1, which shows that most apoptosis regulators were expressed at fairly low levels in astrocytoma (average immunostaining score ≤1) with only a few (CASP3, CASP9, and BIRC5) displaying an average score >1. Figure 2 provides further quantitation of BIRC5-N and the proliferative markers. The apoptotic regulators were not detectable by immunostaining in astrocytes of NBT, although neurons and vascular endothelial cells were immunoreactive to some of these markers to various degrees. MF, Ki67-, or PHH3-labeled nuclei were not observed in NBT.

Box and whisker plots of BIRC5-N and proliferation markers in diffusely infiltrating astrocytomas. The median, lower, and upper quartiles and the minimum-maximum range of each of the 4 parameters (BIRC5-N, Ki67, PHH3, and MF) were plotted against World Health Organization grades (II, III, and IV).

FIGURE 2.

Western Blot Analysis of Caspases and Inhibitor of Apoptosis Proteins

The protein expression of caspases and IAPs was further analyzed by Western blotting in the fresh specimens (Fig. 3A, B). Caspase 3, 8, and 9 were detected in various amounts in the astrocytomas (Fig. 3A), with a tendency of higher amount in higher grade tumors, but the difference was not significant across the tumor grades. Caspase 6, 7, and 10 were barely detectable in these samples (not shown).

FIGURE 3.

(A) Western blot analysis of caspases and inhibitor of apoptosis proteins in normal brain tissue (NBT, lanes 1-7), diffuse astrocytoma (lanes 9-15), anaplastic astrocytoma (lanes 17-21), and glioblastoma multiforme (lanes 24-31). Controls included glioma cell line U251 (lanes 8 and 22), melanoma cell line A875 (lane 16), and a carcinoma tissue sample (lane 23), which were known to express BIRC5 and caspase 3. GAPDH was used as internal control. (B) Western blot analysis of BIRC5 in one original blot. Simultaneous probing of BIRC5 and GAPDH illustrate the specificity of the antibodies. Absence of BIRC5 (first 5 lanes from left) in 5 astrocytoma samples and presence of BIRC5 in glioma cell U251 (lane 6) and a carcinoma sample (lane 7). A prestained broad-range (11-170 kDa) protein ladder was used for SDS-PAGE and aligned to the blot on the right.

Under the Western blotting conditions we used, the higher levels of BIRC5 in GBM allowed its ready detection. BIRC5 amount and positivity in AA was much lower, and it was not detectable in DA (Fig. 3A). The result was consistent with the immunohistochemical finding that, overall, the BIRC5 level in astrocytomas was modest, and stronger cytoplasmic immunostaining and/or higher nuclear labeling index was only seen in a fraction of GBMs. A comparison of IHC and Western analysis of BIRC5 for the fresh tissue samples is given in Tables 2 and 3. CIAPs and XIAPs were also present in these astrocytomas with an apparent tendency of higher levels in higher-grade tumors, but the difference was not significant among the 3 grades (Fig. 2). LIVIN was undetectable in these samples (not shown). Although astrocytes in NBT were not immunoreactive to the apoptotic markers on immunostaining, Western blot analysis demonstrated some caspases and IAPs in NBT (Fig. 3A), probably resulting from their presence in nonastrocytic elements.

TABLE 2.

Comparison of IHC and Western Blot for BIRC5 Protein in Fresh Normal Brain Tissue and Astrocytoma Tissue Samples

TABLE 2.

Comparison of IHC and Western Blot for BIRC5 Protein in Fresh Normal Brain Tissue and Astrocytoma Tissue Samples

TABLE 3.

Summary of BIRC5 Assays in Astrocytoma

TABLE 3.

Summary of BIRC5 Assays in Astrocytoma

Inhibitor of Apoptosis Protein mRNA Expression

The mRNA of IAPs in the fresh specimens was examined by RT-PCR (Fig. 4). Consistent with Western and immunohistochemical results, BIRC5 demonstrated differential expression among the 3 tumor types (Fig. 4; Table 3), but CIAP and XIAP mRNA was detected in most of these astrocytoma samples without significant difference across the tumor grades. In agreement with Western analysis, lower mRNA levels of XIAP and CIAP were detected in some NBT samples, but BIRC5 and LIVIN mRNAs were not detected. The positivity rates of the IAPs by RT-PCR, Western blotting, and the immunostaining analysis was compared in the bar chart in Figure 4. Comparison of BIRC5 assessed by these methods was summarized in Table 3.

FIGURE 4.

Reverse transcriptase- polymerase chain reaction (RT-PCR) analysis of inhibitor of apoptosis protein (IAP) mRNA in normal brain tissue (lanes 1-7), diffuse astrocytoma (lanes 8-14), anaplastic astrocytoma (lanes 15-19), and glioblastoma multiforme (lanes 21-28). Controls included glioma cell line U251 (lane 20), melanoma cell line SK-mel-1 (lane 29), and a carcinoma tissue sample (lane 30), which were known to express BIRC5. Lane 31 was a blank control (no templates). The bar chart on the right side represented comparison of positivity rates (%) of IAPs in the astrocytoma samples assessed by RT-PCR (black bars), Western blot (gray bars), and immunohistochemistry (IHC) (white bars). The rates for RT-PCR and Western blotting were based on the 20 fresh astrocytoma samples, whereas that for IHC was based on the staining results of the 78 astrocytoma samples. An IHC score of 1 or greater was counted as positive.

BIRC5-N and Proliferative Activity Among Astrocytomas of Different Grades

Data summarized in Table 1 shows that BIRC5-N, MF, PHH3, Ki67, and patient age were most significantly different among the 3 astrocytoma types.

Among the apoptosis regulators examined, only BIRC5-N displayed significant difference across all 3 grades. BIRC5-C was different between GBM and AA or DA, but not between DA and AA. Similarly, when percentage of cases with BIRC5-C immunostaining score ≥2 was compared, the difference between DA (6.7%) or AA (6.7%) and GBM (42%) (Fisher exact test p = 0.0201) was significant, but not between DA and AA.

Expression of other IAP family proteins, as well as the caspases, APAF1, and SMAC, was not significantly different among the 3 grades. On the other hand, all proliferative activity markers (MF, Ki67, and PHH3) were significantly different (Table 1; Fig. 2).

Table 4 lists pairwise correlation of the 5 parameters. Consistent with the previously mentioned analysis, strong correlations were observed between tumor grade and BIRC5-N (p < 0.0001), proliferative markers (p < 0.0001), and patient age (p = 0.0003). A weaker correlation between tumor grade and BIRC5-C was observed (p = 0.028). It is significant that although BIRC5-N was strongly associated with the proliferative markers (p < 0.0001), it was only weakly correlated with BIRC5-C (p = 0.0298). BIRC5-C was not correlated significantly with proliferative activity.

TABLE 4.

Correlation Matrix of Parameters Most Closely Associated With Grade

TABLE 4.

Correlation Matrix of Parameters Most Closely Associated With Grade

Figure 5 is a spatial representation by multidimensional scaling (25) of the cases based on the profile of BIRC5-N, Ki67, PHH3, MF, and patient age and summarizes the relationship of the 3 groups of astrocytomas with respect to these parameters. The illustration visually highlights the better separation of DA from GBM, the greater variation among the GBMs, and the intermediate position of AA, which overlaps with the other 2 groups.

FIGURE 5.

Relationship of 3 types of diffusely infiltrating astrocytomas with respect to the profile of 5 parameters (BIRC5-N, Ki67, PHH3, MF, and patient age). The spatial representation was rendered by multidimensional scaling of Euclidean distance matrix of the cases based on their 5-parameter profiles. Cases were represented by dots and color-coded by grade. The picture visually highlights the similarities and dissimilarities of the 3 groups of astrocytomas with respect to the 5 parameters.

Discussion

The present study of major apoptotic and proliferative markers showed that BIRC5-N was the apoptosis marker most significantly different among WHO grade II to IV astrocytomas and most strongly associated with proliferative activity. BIRC5-C was less strongly associated with tumor grade and not with proliferative activity. Expression level of other apoptosis-related proteins was modest or low in astrocytomas and did not correlate significantly with tumor grade or proliferation. Protein analysis by Western blotting (for caspases and IAPs) and mRNA by RT-PCR (for IAPs) showed general agreement with IHC results in astrocytomas. Apoptosis regulators were virtually undetectable in astrocytes of normal brain by immunostaining.

BIRC5 and Other Inhibitor of Apoptosis Proteins in Astrocytoma

Several studies examined IAP expression in astrocytomas. BIRC5 mRNA was reported in 3 of 8 (37.5%) DA, 13 of 15 (86.7%) AA, and 18 of 20 (90.0%) GBM, which was associated with grade and survival (16). BIRC5 immunoreactivity was reported in 2 of 4 DA, 3 of 3 AA, and ranged from 80% (31 of 39) to 90% (9 of 10) in GBM with expression intensity increasing with grade (17, 18). Western blot analysis showed BIRC5 positivity in 64% (59 of 92) of gliomas in one cohort (19). Kleinschmidt-DeMasters et al reported that BIRC5 nuclear immunostaining and mRNA were detected in most of 25 glioma specimens (16 GBM, 4 AA, and 2 DA) but noticed that BIRC5 staining in GBM was modest compared with epithelial malignancies (20).

Functionally, in addition to suppression of apoptosis (19), BIRC5 was shown to cause radiation resistance through caspase-independent mechanisms in astrocytoma cells (26) or to promote cell growth by counteracting the function of p53 (27). Interestingly, VEGF produced by glioma cells promoted endothelial cell survival by increasing expression of XIAP and BIRC5 (28).

The prognostic value of BIRC5 in gliomas is controversial. BIRC5 protein expression by Western blot analysis (19) or immunoreactivity (8) has been correlated with survival by some groups. However, one group reported that Ki-67 but not BIRC5 nuclear index was correlated with overall survival in GBM (7), although both Ki67 and BIRC5 nuclear staining was correlated with prognosis for ependymoma (29). These authors also demonstrated that BIRC5 nuclear-positive cells represented a fraction (approximately 50-60%) of Ki67-positive cells (29). Our data are consistent with the latter observation, because BIRC5-N is lower than Ki67 index in each grade (Table 1). Future studies aiming at standardization of BIRC5 assay should help reveal its potential use in astrocytoma grading and prognosis.

Differential BIRC5 nuclear and cytoplasmic staining implicated different biologic functions and the mechanisms underlying which may be partly the result of the presence of multiple BIRC5 splice variants (30,31). As we showed, BIRC5-N is strongly correlated with astrocytoma grade and proliferation, whereas BIRC5-C was less strongly associated with grade and not with proliferative activity. Moreover, BIRC5-N and BIRC5-C were only weakly associated with each other. These data indicated that BIRC5-N was more closely associated with proliferation than with apoptosis regulation (14).

The overall expression level of other IAPs did not show dramatic change with increasing tumor grade, suggesting that increased IAP expression in astrocytomas as compared with normal brain tissue might be involved in astrocytoma tumorigenesis, but tumor progression was probably more closely associated with acquisition of additional proliferative advantages, of which BIRC5-N expression appeared to be a manifestation.

Methods using antagonists of IAPs, for example, SMAC (or its mimetics), are attractive approaches to inducing glioma cell death. In chemoresistant human primary glioblastoma cells, downregulation of XIAP by SMAC peptide potentiated TRAIL-mediated apoptosis (32). For any IAP-targeting approaches to succeed, it is necessary to know the IAP levels in individual astrocytomas. More studies are required to evaluate which assays such as IHC or Western blot for proteins, or RT-PCR and in situ hybridization for mRNAs, would be most useful for clinical applications. As we have shown, immunostaining has a particular dimension of application because it readily differentiates the nuclear and cytoplasmic pools of BIRC5, which appeared to be functionally divergent.

Proapoptotic Molecules in Astrocytomas

Expression status of the caspases in diffusely infiltrating astrocytomas has not been evaluated previously. Our data show that immunohistochemically detectable caspases, as well as APAF1 and SMAC, were mostly expressed at modest or low levels in astrocytomas. This is in contrast to epithelial tumors in which some caspases (especially caspase 3) could be expressed at high levels (33,34). In studies using astrocytoma cell lines, the presence of mRNA or protein of functional caspases has been documented, including caspases most commonly involved in apoptotic regulation (caspase 3, 6, 7, 8, 9, and 10) (47).

PHH3 and Other Proliferation Markers

Assessment of proliferative activity is an integral part in evaluating diffusely infiltrating astrocytomas (1,3,4). Most studies reported correlation of higher Ki67 with higher grade (2-5) and unfavorable prognosis (2, 6, 7), providing various cutoff values (ranging from 1.5% for all diffusely infiltrating astrocytomas to 35% for glioblastomas) (3,6,7). The data from our series also supported the usefulness of Ki67 in astrocytoma grading, particularly for separating DA and AA. However, lack of correlation or even favorable correlation of Ki67 with survival was occasionally reported (8,9).

Although MF was used in everyday practice, problems arise when equivocal mitotic figures were encountered. The anti-PHH3 antibody is an attractive solution to such problems (5,23). Histone H3 phosphorylation at serine 10 is a mitosis-specific event in parallel with mitotic chromatin condensation, which is not observed in apoptosis (5,23,49). Application of this marker resulted in more rapid, reliable, and objective grading of meningiomas (23) and in determining cell-cycle phase distribution of tumor cells in astrocytomas (5). Our data favored the usefulness of PHH3 index in grading astrocytomas (Table 1; Fig. 2). In 2 cases of our series originally classified as grade II either because mitotic figures escaped detection or were dismissed as equivocal on hematoxylin and eosin sections, PHH3 staining identified more than occasional unequivocal mitotic figures and resulted in upgrading the tumors to WHO grade III. Because PHH3 labeling highlights mitotic cells and makes it much easier to quickly estimate the mitotic index, it is apparently more sensitive and objective than MF and should be clinically useful.

Acknowledgments

The authors dedicate this article to Professor Guanghua Yang, mentor to them for decades, who recently died. The authors thank their colleagues in the Department of Neurosurgery at West China Hospital and particularly Dr. Hongwei Chen (current address Department of Neurosurgery, Anhui Medical University) for their assistance. The authors also thank Drs. Y. F. Wang, B. Yang, Ms. Y. L. Deng, Ms. D. Xiao, Mr. K. Yang, Ms. B. L. Liu, and Ms. X. H. Wu for assistance and Dr. Bo Yang for participation in the early phase of this project.

Preliminary results of a portion of the project were submitted in abstract form to the U.S. and Canadian Academy of Pathology Annual Meeting, Atlanta, Georgia, February 2006.

References

eds.

World Health Organization of Tumours. Pathology and Genetics of Tumours of the Nervous System

Lyon

IARC Press

2000

–

et al. .

Prognostication of astrocytoma patient survival by Ki-67 (MIB-1), PCNA, and S-phase fraction using archival paraffin-embedded samples

J Pathol

1994

;

174

275

–

et al. .

Use of MIB-1 (Ki-67) immunoreactivity in differentiating grade II and grade III gliomas

J Neuropathol Exp Neurol

1997

;

857

–

et al. .

Ultrarapid Ki-67 immunostaining in frozen section interpretation of gliomas

J Clin Pathol

2005

;

263

–

et al. .

Immunohistochemical estimation of cell cycle entry and phase distribution in astrocytomas: Applications in diagnostic neuropathology

Neuropathol Appl Neurobiol

2005

;

455

–

et al. .

MIB-1 and DNA topoisomerase II alpha could be helpful for predicting long-term survival of patients with glioblastoma

Am J Clin Pathol

2003

;

119

715

–

et al. .

No prognostic impact of survivin expression in glioblastoma

Acta Neuropathol (Berl)

2005

;

109

534

–

et al. .

Prognostic significance of the immunohistochemical index of survivin in glioma: A comparative study with the MIB-1 index

J Neurooncol

2005

;

231

–

et al. .

High expression of DNA topoisomerase IIalpha and Ki-67 antigen is associated with prolonged survival in glioblastoma patients

Eur J Cancer

2002

;

1343

–

Caspase activation-stepping on the gas or releasing the brakes? Lessons from humans and flies

Oncogene

2004

;

2774

–

Caspase activation, inhibition, and reactivation: A mechanistic view

Protein Sci

2004

;

1979

–

Molecular mechanisms of caspase regulation during apoptosis

Nat Rev Mol Cell Biol

2004

;

897

–

907

IAP proteins: Blocking the road to death's door

Nat Rev Mol Cell Biol

2002

;

401

–

Inhibitor of apoptosis proteins: Translating basic knowledge into clinical practice

Cancer Res

2004

;

7183

–

The expanding role of mitochondria in apoptosis

Genes Dev

2001

;

2922

–

et al. .

Expression of survivin in astrocytic tumors: Correlation with malignant grade and prognosis

Cancer

2003

;

1077

–

et al. .

Expression of survivin, an inhibitor of apoptosis protein, in tumors of the nervous system

Acta Neuropathol (Berl)

2002

;

104

105

–

et al. .

Expression of survivin in primary glioblastomas

J Cancer Res Clin Oncol

2002

;

128

302

–

et al. .

Quantitatively determined survivin expression levels are of prognostic value in human gliomas

J Clin Oncol

2002

;

1063

–

Kleinschmidt-DeMasters

et al. .

Survivin in glioblastomasProtein and messenger RNA expression and comparison with telomerase levels

Arch Pathol Lab Med

2003

;

127

826

–

Blocking of endogenous avidin-binding activity in immunohistochemistry: The use of skim milk as an economical and effective substitute for commercial biotin solutions

Appl Immunohistochem Mol Morphol

1999

;

–

et al. .

Tissue microarray validation of epidermal growth factor receptor and SALL2 in synovial sarcoma with comparison to tumors of similar histology

Am J Pathol

2003

;

163

1449

–

et al. .

The mitosis-specific antibody anti-phosphohistone-H3 (PHH3) facilitates rapid reliable grading of meningiomas according to WHO 2000 criteria

Am J Surg Pathol

2004

;

1532

–

et al. .

Melanoma inhibitor of apoptosis protein is expressed differentially in melanoma and melanocytic naevus, but similarly in primary and metastatic melanomas

J Clin Pathol

2005

;

1081

–

Analyzing Multivariate Data

Pacific Cove

Brooks/Cole Publishing Co

2003

206

–

et al. .

Survivin enhances radiation resistance in primary human glioblastoma cells via caspase-independent mechanisms

Oncogene

2004

;

7494

–

506

et al. .

Survivin inhibits anti-growth effect of p53 activated by aurora B

Biochem Biophys Res Commun

2005

;

336

1164

–

et al. .

Glioma cells under hypoxic conditions block the brain microvascular endothelial cell death induced by serum starvation

J Neurochem

2005

;

–

110

et al. .

Survivin expression in intracranial ependymomas and its correlation with tumor cell proliferation and patient outcome

Am J Clin Pathol

2005

;

124

543

–

Molecular circuits of apoptosis regulation and cell division control: The survivin paradigm

J Cell Biochem

2004

;

656

–

et al. .

Survivin study: What is the next wave?

J Cell Physiol

2003

;

197

–

et al. .

X-linked inhibitor regulating TRAIL-induced apoptosis in chemoresistant human primary glioblastoma cells

Clin Invest Med

2003

;

231

–

et al. .

Apoptosis regulators in superficial urothelial carcinoma: An analysis of eleven proteins of the caspase family and the inhibitor of apoptosis protein (IAP) family [Abstract]

Mod Pathol

2006

;

169A

et al. .

Overexpression of caspase-3 in hepatocellular carcinomas

Mod Pathol

2004

;

861

–

et al. .

Proteasome inhibitor-induced apoptosis of glioma cells involves the processing of multiple caspases and cytochrome c release

J Neurochem

2000

;

2288

–

et al. .

Erucylphosph180ocholine-induced apoptosis in chemoresistant glioblastoma cell lines: Involvement of caspase activation and mitochondrial alterations

Anticancer Res

2001

;

3389

–

et al. .

Protein kinase C-eta regulates resistance to UV- and gamma-irradiation-induced apoptosis in glioblastoma cells by preventing caspase-8 activation

Neuro-oncol

2002

;

–

et al. .

Molecular evidence of apoptotic death in malignant brain tumors including glioblastoma multiforme: Upregulation of calpain and caspase-3

J Neurosci Res

2002

;

197

–

206

et al. .

Tumor necrosis factor-alpha-induced cell death in U373 cells overexpressing alpha-synuclein

J Neurosci Res

2003

;

334

–

et al. .

A key role for caspase-2 and caspase-3 in the apoptosis induced by 2-chloro-2′-deoxy-adenosine (cladribine) and 2-chloro-adenosine in human astrocytoma cells

Mol Pharmacol

2003

;

1437

–

Caspase-1 mediates Fas-induced apoptosis and is up-regulated by interferon-gamma in human astrocytoma cells

J Neurooncol

2004

;

167

–

et al. .

Ubiquitin-proteasome pathway as a primary defender against TRAIL-mediated cell death

Cell Mol Life Sci

2004

;

1075

–

Antiangiogenic and proapoptotic effects of dietary restriction on experimental mouse and human brain tumors

Clin Cancer Res

2004

;

5622

–

et al. .

Alkylphosphocholine-induced glioma cell death is BCL-X(L)-sensitive, caspase-independent and characterized by massive cytoplasmic vacuole formation

Cell Death Differ

2004

;

1326

–

Resistance of human astrocytoma cells to apoptosis induced by mitochondria-damaging agents: Possible implications for anticancer therapy

J Pharmacol Exp Ther

2005

;

314

825

–

et al. .

Sensitization for gamma-irradiation-induced apoptosis by second mitochondria-derived activator of caspase

Cancer Res

2005

;

10502

–

The role of caspases 9 and 9-short (9S) in death ligand- and drug-induced apoptosis in human astrocytoma cells

Brain Res Mol Brain Res

2002

;

106

–

et al. .

Tumor necrosis factor-related apoptosis-inducing ligand induces caspase-dependent interleukin-8 expression and apoptosis in human astroglioma cells

Mol Cell Biol

2002

;

724

–

et al. .

Chromatin condensation is not associated with apoptosis

J Biol Chem

1998

;

273

24470

–

Author notes

Drs. Liu and Chen shared the first authorship of this article.

This work was supported by grants to Dr. Zhou from the Natural Science Foundation of China (NSFC, 30125023, 30570692, 30221001) and the Ministry of Science and Technology (2002CCA01400) of China.