Characterization of an IDH1 R132H Rabbit Monoclonal Antibody, MRQ-67, and Its Applications in the Identification of Diffuse Gliomas

Abstract

The current diagnosis of diffuse glioma involves isocitrate dehydrogenase (IDH) mutation testing. Most IDH mutant gliomas carry a G-to-A mutation at IDH1 position 395, resulting in the R132H mutant. R132H immunohistochemistry (IHC), therefore, is used to screen for the IDH1 mutation. In this study, the performance of MRQ-67, a recently generated IDH1 R132H antibody, was characterized in comparison with H09, a frequently used clone. Selective binding was demonstrated by an enzyme-linked immunosorbent assay for MRQ-67 to the R132H mutant, with an affinity higher than that for H09. By Western and dot immunoassays, MRQ-67 was found to bind specifically to the IDH1 R1322H, with a higher capacity than H09. IHC testing with MRQ-67 demonstrated a positive signal in most diffuse astrocytomas (16/22), oligodendrogliomas (9/15), and secondary glioblastomas tested (3/3), but not in primary glioblastomas (0/24). While both clones demonstrated a positive signal with similar patterns and equivalent intensities, H09 exhibited a background stain more frequently. DNA sequencing on 18 samples showed the R132H mutation in all IHC positive cases (5/5), but not in negative cases (0/13). These results demonstrate that MRQ-67 is a high-affinity antibody suitable for specific detection of the IDH1 R132H mutant by IHC and with less background as compared with H09.

Keywords: IDH1 R132H mutant; dot immunoassay; glioma; immunohistochemistry; isocitrate dehydrogenase 1; monoclonal antibody.

Figure 1

Liquid-phase binding of the antibodies MRQ-67 (A,B) and H09 (C,D) to wild-type (wt) and mutant IDH1 (A,C) and IDH2 peptides (B,D), as measured by ELISA and expressed as optical densities at 450 nm (OD450).

Figure 2

Solid-phase binding of the antibodies MRQ-67 and H09 to endogenous IDH1 R132H mutant protein on Western blots (A,B) and the synthetic R132H mutant peptide on dot blots (C,D). (A,B) Cell lysates of the glioma cell line BT142 mut/-, which harbors a homozygous IDH1 R132H mutation, and HepG2, a liver epithelial tumor cell line expressing only wt IDH1, were loaded onto a 4–12% Bis-Tris mini gel and electro-transferred onto PVDF membranes. The membranes were probed with MRQ-67 (A) or H09 (B). The presence of IDH1 protein, in both the wt and mutant forms, was confirmed by probing the membranes with a polyclonal antibody to IDH1 (28206). (C,D) Synthetic IDH1/2 wt and mutant peptides were dotted to PVDF membranes, 2 µL each, at different concentrations. The blots were probed with MRQ-67 (C) or H09 (D). Immunoreactions were demonstrated by incubation with an AP-conjugated secondary antibody and visualized in an NBT/BCIP solution.

Figure 3

Binding of the antibodies MRQ-67 (A,C) and H09 (B,D) with endogenous IDH1 R132H mutant, but not with the IDH1 wt protein, as demonstrated on cell pellet samples of cell lines BT142 mut/- (A,B) and HepG2 (C,D) by IHC. UltraView, visualized with DAB, and counterstained with hematoxylin. Scale bar = 100 µm.

Figure 4

Immunoreactivity for IDH1 R132H in IDH1-muted gliomas including anaplastic astrocytoma (AST-19; A,B), oligodendroglioma (AST-18; C,D), and secondary glioblastoma (GLB-9; E,F) as stained separately with MRQ-67 (A,C,E) and H09 (B,D,F) on serial FFPE tissue sections. All of the tumor cells were labelled in a cytoplasmic and nuclear pattern, but blood vessels (arrow) were not. UltraView, visualized with DAB, and counterstained with hematoxylin. Scale bar = 100 µm.

Figure 5

Immunohistochemical demonstration of glioma growth fronts, either with a pushing border (ODG-5; A,B) or in an infiltrating fashion (ODG-2; C), and cryptically infiltrating low-grade glioma cells from surrounding brain tissue (BRN-24; D) using MRQ-67 (A,C,D) or H09 (B). UltraView, visualized with DAB, and counterstained with hematoxylin. (A,B,D) scale bar = 100 µm; (C) scale bar = 200 µm.

Figure 6

Intensities of background stain for MRQ-67 (A) and H09 (B) as encountered in astrocytoma (AST), oligodendroglioma (ODG), glioblastoma (GLB), meningioma (MGM), and surrounding nonneoplastic brain samples (BRN), with the statistical results indicated on each bar in B (*, p < 0.05; **, p < 0.01; ∆, p > 0.05).

Figure 7

Representative meningioma samples, MGM-2 (A,B) and MGM-7 (C,D), were stained separately with MRQ-67 (A,C) and H09 (B,D), with some disturbing nonspecific stain observed with H09 (scores of 2 and 2.5 as shown in (B,D), respectively), but not with MRQ-67 (scores of 0 and 0.5 in (A,C), respectively), mainly at some fibrotic components. UltraView, visualized with DAB, and counterstained with hematoxylin. (A,B) scale bar = 100 µm; (C,D) scale bar = 200 µm.

Figure 8

A meningioma sample (MGM-4) was stained with H09 (1.63 µg/mL), with nonspecific stain at some fibrotic components (arrows; score 1). UltraView, visualized with DAB, and counterstained with hematoxylin. Scale bar = 50 µm.

Figure 9

Sanger sequence chromatograms for three representative tissue samples that do not have the R132H mutation (A) and five tissue samples confirmed to have the R132H mutation (B). Corresponding amino acid sequences are labeled as AA. The c.395G > A (p.R132H) mutation location is underlined with yellow bars and labeled respectively. The presence of double chromatogram peaks corresponding to both guanine and adenine at nucleotide position c.395 indicates the presence of the c.395G > A (p.R132H) mutation and is annotated in the nucleotide sequences by the IUPAC nucleotide code R, whereas non-mutant tissue samples are identified by the presence of a single guanine peak at nucleotide position c.395 annotated in the nucleotide sequences by the code G.