PLoS ONE (Jan 2015)

Comprehensive Glycomics of a Multistep Human Brain Tumor Model Reveals Specific Glycosylation Patterns Related to Malignancy.

  • Jun-ichi Furukawa,
  • Masumi Tsuda,
  • Kazue Okada,
  • Taichi Kimura,
  • Jinhua Piao,
  • Shinya Tanaka,
  • Yasuro Shinohara

DOI
https://doi.org/10.1371/journal.pone.0128300
Journal volume & issue
Vol. 10, no. 7
p. e0128300

Abstract

Read online

Cancer cells frequently express glycans at different levels and/or with fundamentally different structures from those expressed by normal cells, and therefore elucidation and manipulation of these glycosylations may provide a beneficial approach to cancer therapy. However, the relationship between altered glycosylation and causal genetic alteration(s) is only partially understood. Here, we employed a unique approach that applies comprehensive glycomic analysis to a previously described multistep tumorigenesis model. Normal human astrocytes were transformed via the serial introduction of hTERT, SV40ER, H-RasV12, and myrAKT, thereby mimicking human brain tumor grades I-IV. More than 160 glycans derived from three major classes of cell surface glycoconjugates (N- and O-glycans on glycoproteins, and glycosphingolipids) were quantitatively explored, and specific glycosylation patterns related to malignancy were systematically identified. The sequential introduction of hTERT, SV40ER, H-RasV12, and myrAKT led to (i) temporal expression of pauci-mannose/mono-antennary type N-glycans and GD3 (hTERT); (ii) switching from ganglio- to globo-series glycosphingolipids and the appearance of Neu5Gc (hTERT and SV40ER); (iii) temporal expression of bisecting GlcNAc residues, α2,6-sialylation, and stage-specific embryonic antigen-4, accompanied by suppression of core 2 O-glycan biosynthesis (hTERT, SV40ER and Ras); and (iv) increased expression of (neo)lacto-series glycosphingolipids and fucosylated N-glycans (hTERT, SV40ER, Ras and AKT). These sequential and transient glycomic alterations may be useful for tumor grade diagnosis and tumor prognosis, and also for the prediction of treatment response.