Secreted protein acidic and rich in cysteine promotes glioma invasion and delays tumor growth in vivo

Secreted protein acidic and rich in cysteine (SPARC) is highly expressed in human astrocytomas, grades II-IV. We demonstrated previously that SPARC promotes invasion in vitro using the U87MG-derived clone U87T2 and U87T2-derived SPARC-transfected clones, A2b2, A2bi, and C2a4, in the spheroid confrontation assay. Additional in vitro studies demonstrated that SPARC delays growth, increases attachment, and modulates migration of tumor cells in extracellular matrix-specific and concentration-dependent manners. Therefore, we propose that SPARC functionally contributes to brain tumor invasion and delays tumor growth in vivo, and that the effects of SPARC are related to the level of SPARC secreted into the extracellular matrix. To test these hypotheses, we stereotactically injected these clones into nude rat brains (six animals were injected per clone). Animals were sacrificed on day 7 to assess growth and invasion for all clones at the same time in tumor development. To determine whether SPARC delayed but did not inhibit growth, rats were injected with U87T2 or clone A2b2, and the animals were sacrificed on days 9 (U87T2) and 20 (A2b2), when the animals demonstrated neurological deficit. Brains were removed, fixed, photographed, paraffin embedded, and sectioned. Sections were then serially stained with H&E for morphological assessment of invasion and to measure tumor volume, immunohistochemically stained to visualize SPARC, subjected to in situ hybridization with the human AluII DNA-binding probe to identify human cells, and immunohistochemically stained with MIB-1 to measure proliferation index. The results demonstrate that SPARC promotes invasion in vivo at day 7. Both the low (A2bi) and the high (A2b2) SPARC-secreting clones produced invasive tumors, invading with fingerlike projections and satellite masses into adjacent brain, as well as along the corpus collosum. The intermediate SPARC secreting clone (C2a4) primarily migrated as a bulk tumor along the corpus collosum. SPARC significantly decreased tumor growth at day 7, as measured both by adjusted MIB-1 proliferation indices (U87T2 = 95.3 +/- 1.4 versus A2bi = 73.4 +/- 4.0, A2b2 = 30.8 +/- 6.7 and C2a4 = 15.7 +/- 13.0) and tumor volumes (U87T2 = 13.4 +/- 0.6 mm(3) versus A2bi = 4.5 +/- 0.6 mm(3), A2b2 = 1.1 +/- 0.1 mm(3), and C2a4 = 0.4 +/- 0.1 mm(3)). Furthermore, SPARC delayed but did not inhibit tumor growth. The patterns of invasion and the extent of growth delay correlated with the level of SPARC expression. We propose that the ability of SPARC to promote invasion depends on the level of its secretion and the resultant modulation of the level of adherence and motility induced. This demonstration that SPARC functionally contributes to brain tumor invasion in vivo suggests that SPARC is a candidate therapeutic target for the design of therapies directed toward inhibition of the invasive phenotype.