Patients with chronic ulcerative colitis are at increased risk of developing colorectal cancer. Although current hypotheses suggest that sporadic colorectal cancer is due to inability to control cancer stem cells, the cancer stem cell hypothesis has not yet been validated in colitis-associated cancer. Furthermore, the identification of the colitis to cancer transition is challenging. We recently showed that epithelial cells with the increased expression of aldehyde dehydrogenase in sporadic colon cancer correlate closely with tumor-initiating ability. We sought to determine whether ALDH can be used as a marker to isolate tumor-initiating populations from patients with chronic ulcerative colitis. We used fluorescence-activated cell sorting to identify precursor colon cancer stem cells from colitis patients and report both their transition to cancerous stem cells in xenografting studies as well as their ability to generate spheres in vitro. Similar to sporadic colon cancer, these colitis-derived tumors were capable of propagation as sphere cultures. However, unlike the origins of sporadic colon cancer, the primary colitic tissues did not express any histologic evidence of dysplasia. To elucidate a potential mechanism for our findings, we compared the stroma of these different environments and determined that at least one paracrine factor is up-regulated in the inflammatory and malignant stroma compared with resting, normal stroma. These data link colitis and cancer identifying potential tumor-initiating cells from colitic patients, suggesting that sphere and/or xenograft formation will be useful to survey colitic patients at risk of developing cancer. [Cancer Res 2009;69(20):8208-15]

Embryonic stem (ES) cells can generate neural progenitors and neurons in vitro and incorporate into the adult central nervous system (CNS) following transplantation, suggesting their therapeutic potential for treating neurological disorders. However, our understanding of the conditions that direct ES-derived neural progenitor (ESNP) migration and differentiation within different regions of the adult CNS is incomplete. Rodents treated with the chemoconvulsant kainic acid (KA) experience seizures and display hippocampal sclerosis, as well as enhanced hippocampal neurogenesis, similar to pathological findings in patients with temporal lobe epilepsy (TLE). To examine the potential for ESNPs to incorporate into the adult hippocampus and differentiate into hippocampal neurons or glia following seizure-induced damage, we compared the fates of ESNPs after they were transplanted into the CA3 region or fimbria 1 week following KA-induced seizures. After 4-8 weeks, ESNPs grafted into the CA3 region had migrated to the dentate gyrus (DG), where a small subset adopted neural stem cell fates and continued to proliferate, based on bromodeoxyuridine uptake. Others differentiated into neuroblasts or dentate granule neurons. In contrast, most ESNPs transplanted into the fimbria migrated extensively along existing fiber tracts and differentiated into oligodendrocytes or astrocytes. Hippocampal grafts in mice not subjected to seizures displayed a marked tendency to form tumors, and this effect was more pronounced in the DG than in the fimbria. Taken together, these data suggest that seizures induce molecular changes in the CA3 region and DG that promote region-specific neural differentiation and suppress tumor formation.

Mouse embryonic stem cells can differentiate in vitro into cells of the nervous system, neurons and glia. This differentiation mimics stages observed in vivo, including the generation of primitive ectoderm and neurectoderm in embryoid body culture. We demonstrate here that embryonic stem cell lines mutant for components of the Hedgehog signaling cascade are deficient at generating neurectoderm-containing embryoid bodies. The embryoid bodies derived from mutant cells are also unable to respond to retinoic acid treatment by producing nestin-positive neural stem cells, a response observed in cultures of heterozygous cells, and contain cores apparently arrested at the primitive ectoderm stage. The mutant cultures are also deficient in their capacity to differentiate into mature neurons and glia. These data are consistent with a role for Hedgehog signaling in generating neurectoderm capable of producing the appropriate neuronal and glial progenitors in ES cell culture.

Stem cell therapies for neurodegenerative disorders require accurate delivery of the transplanted cells to the sites of damage. Numerous studies have established that fluid injections to the hippocampus can induce lesions in the dentate gyrus (DG) that lead to cell death within the upper blade. Using a mouse model of temporal lobe epilepsy, we previously observed that embryonic stem cell-derived neural progenitors (ESNPs) survive and differentiate within the granule cell layer after stereotaxic delivery to the DG, replacing the endogenous cells of the upper blade. To investigate the mechanisms for ESNP migration and repair in the DG, we examined the role of the chemokine CXCL12 in mice subjected to kainic acid-induced seizures. We now show that ESNPs transplanted into the DG show extensive migration through the upper blade, along the septotemporal axis of the hippocampus. Seizures upregulate CXCL12 and infusion of the CXCR4 antagonist AMD3100 by osmotic minipump attenuated ESNP migration. We also demonstrate that seizures promote the differentiation of transplanted ESNPs toward neuronal rather than astrocyte fates. These findings suggest that ESNPs transplanted into the adult rodent hippocampus migrate in response to cytokine-mediated signals.

Summary:Complex partial seizures arising from mesial temporal lobe structures are a defining feature of mesial temporal lobe epilepsy (TLE). For many TLE patients, there is an initial traumatic head injury that is the precipitating cause of epilepsy. Severe TLE can be associated with neuropathological changes, including hippocampal sclerosis, neurodegeneration in the dentate gyrus, and extensive reorganization of hippocampal circuits. Learning disabilities and psychiatric conditions may also occur in patients with severe TLE for whom conventional anti-epileptic drugs are ineffective. Novel treatments are needed to limit or repair neuronal damage, particularly to hippocampus and related limbic regions in severe TLE and to suppress temporal lobe seizures. A promising therapeutic strategy may be to restore inhibition of dentate gyrus granule neurons by means of cell grafts of embryonic stem cell-derived GABAergic neuron precursors. "Proof-of-concept" studies show that human and mouse embryonic stem cell-derived neural precursors can survive, migrate, and integrate into the brains of rodents in different experimental models of TLE. In addition, studies have shown that hippocampal grafts of cell lines engineered to release GABA or other anticonvulsant molecules can suppress seizures. Furthermore, transplants of fetal GABAergic progenitors from the mouse or human brain have also been shown to suppress the development of seizures. Here, we review these relevant studies and highlight areas of future research directed toward producing embryonic stem cell-derived GABAergic interneurons for cellbased therapies for treating TLE.