Glioblastoma (GBM) is the most common primary brain cancer in adults. Despite the remarkable advancements in recent years in the realm of cancer diagnosis and therapy, regrettably, GBM remains the most lethal form of brain cancer. In this view, the fascinating area of nanotechnology has emerged as an innovative strategy for developing novel nanomaterials for cancer nanomedicine, such as artificial enzymes, termed nanozymes, with intrinsic enzyme-like activities. Therefore, this study reports for the first time the design, synthesis, and extensive characterization of innovative colloidal nanostructures made of cobalt-doped iron oxide nanoparticles chemically stabilized by a carboxymethylcellulose capping ligand (i.e., Co-MION), creating a peroxidase-like (POD) nanozyme for biocatalytically killing GBM cancer cells. These nanoconjugates were produced using a strictly green aqueous process under mild conditions to create non-toxic bioengineered nanotherapeutics against GBM cells. The nanozyme (Co-MION) showed a magnetite inorganic crystalline core with a uniform spherical morphology (diameter, 2R = 6-7 nm) stabilized by the CMC biopolymer, producing a hydrodynamic diameter (HD) of 41-52 nm and a negatively charged surface (ZP~-50 mV). Thus, we created supramolecular water-dispersible colloidal nanostructures composed of an inorganic core (Cox-MION) and a surrounding biopolymer shell (CMC). The nanozymes confirmed the cytotoxicity evaluated by an MTT bioassay using a 2D culture in vitro of U87 brain cancer cells, which was concentration-dependent and boosted by increasing the cobalt-doping content in the nanosystems. Additionally, the results confirmed that the lethality of U87 brain cancer cells was predominantly caused by the production of toxic cell-damaging reactive oxygen species (ROS) through the in situ generation of hydroxyl radicals (·OH) by the peroxidase-like activity displayed by nanozymes. Thus, the nanozymes induced apoptosis (i.e., programmed cell death) and ferroptosis (i.e., lipid peroxidation) pathways by intracellular biocatalytic enzyme-like activity. More importantly, based on the 3D spheroids model, these nanozymes inhibited tumor growth and remarkably reduced the malignant tumor volume after the nanotherapeutic treatment (ΔV~40%). The kinetics of the anticancer activity of these novel nanotherapeutic agents decreased with the time of incubation of the GBM 3D models, indicating a similar trend commonly observed in tumor microenvironments (TMEs). Furthermore, the results demonstrated that the 2D in vitro model overestimated the relative efficiency of the anticancer agents (i.e., nanozymes and the DOX drug) compared to the 3D spheroid models. These findings are notable as they evidenced that the 3D spheroid model resembles more precisely the TME of "real" brain cancer tumors in patients than 2D cell cultures. Thus, based on our groundwork, 3D tumor spheroid models might be able to offer transitional systems between conventional 2D cell cultures and complex biological in vivo models for evaluating anticancer agents more precisely. These nanotherapeutics offer a wide avenue of opportunities to develop innovative nanomedicines for fighting against cancerous tumors and reducing the frequency of severe side effects in conventionally applied chemotherapy-based treatments.