br Materials and methods br Results br

Materials and methods
Results
Discussion We have demonstrated the efficient generation of high numbers of bfCNs from hPSCs via an expandable basal forebrain NSC intermediate. This is of particular interest as bfCNs are one of the primary populations affected in AD, and our protocol provides a robust resource for their in vitro study (Holtzman et al., 2011; Everitt and Robbins, 1997; Davies and Maloney, 1976). This work is highly relevant in light of the recent advent of iPS technology, which has opened up the possibility of modeling diseases from a patient\'s own tissue (Han et al., 2011). In the amyloid the basal forebrain develops from the most ventral part of the embryonic telencephalon, and this ventral identity is induced by hedgehog signaling, via the SHH ligand (Fig. 8) (Ericson et al., 1995a; Shimamura et al., 1995; Pera and Kessel, 1997; Gunhaga et al., 2000). Regardless of the original cell line we could reproducibly and efficiently generate NSCs with a basal forebrain phenotype using EB-based, non-adherent differentiation or NAdD (Fig. 8). Specifically we have shown that NAdD generates higher levels of secreted SHH compared to differentiation as a monolayer, and this is true regardless of media composition. Furthermore, hedgehog signaling is initiated at the stage of early germ layer differentiation in the EB, the first stage of the NAdD protocol, and prior to the addition of any exogenous factors or morphogens, which could influence dorso-ventral cell fate. Timing of this ventral patterning is in agreement with events in the developing embryo where the hedgehog signaling is induced relatively early in development as the germ layers develop adjacent to the ventral neural tube (Fig. 8) (Briscoe and Ericson, 1999). Therefore we conclude that it is the unique 3D environment of the EB, which is sufficient to recapitulate these developmental events in vitro. It has been reported by several studies that a naive neural differentiation system will produce derivatives with telencephalic identity, and this seems somewhat the ‘default’ neural fate of hPSCs (Watanabe et al., 2005; Gaspard et al., 2008; Li et al., 2009). However dorso-ventral fate is highly variable between protocols, and this is likely due to differentiation conditions generating their own intrinsic dorso-ventral induction (Bissonnette et al., 2011; Watanabe et al., 2005; Gaspard et al., 2008; Li et al., 2009; Danjo et al., 2011; Wataya et al., 2008). In terms of ventral forebrain, very high numbers of neural cells expressing markers of the ventral/basal forebrain, have only previously been achieved by the addition of specific doses of extrinsic SHH or other factors thought to influence this fate (Bissonnette et al., 2011; Watanabe et al., 2005; Gaspard et al., 2008; Li et al., 2009; Danjo et al., 2011; Nat et al., 2012; Ma et al., 2012). Interestingly, however several studies support our evidence for the early induction of hedgehog signaling in the EB, however this has never been utilized for bfCN differentiation. It has previously been demonstrated in two studies by Iacovitti et al., where exogenous ventralizing factors were absent from an EB-based protocol, which efficiently produced midbrain neurons with a ventral phenotype (Iacovitti et al., 2007; Cai et al., 2009). Furthermore, other studies generating non-neural derivatives also establish that EB formation does generate high levels of hedgehog signaling (Mfopou et al., 2007; Maye et al., 2000; Wu et al., 2010). Therefore, although the utilization of endogenous signaling pathways is not often reported for use in hPSC differentiation, it is clear that hPSCs have the ability to generate their own signals, and this does occur in the absence of over-riding high concentration signaling factors that are present in many protocols. Expression of NKX2.1 and LHX8 in the NAdD generated NSCs was followed by expression of the bfCN transcription factor ISL1 upon terminal differentiation. Taken together this indicates that we were able to recapitulate embryonic events. Recently, two groups published findings indicating the generation of bfCNs from hES cells (Nilbratt et al., 2010; Bissonnette et al., 2011). In both cases bfCN differentiation was dependent on exposing hES cells to known factors required in the developing basal forebrain either by ectopic application in the differentiation media or by over-expression (Nilbratt et al., 2010; Bissonnette et al., 2011). An accurate developmental model of any neuronal population affected by neurodegeneration is key to therapeutic progress, as in terms of neuroprotection, regeneration or replacement, it is crucial that we understand the factors required for cell differentiation and survival (Nat and Dechant, 2011; Liu, 2011). We therefore conclude that our NAdD protocol, in the absence of exogenous factors, better recapitulates developmental events in the basal forebrain, and thus serves as an accurate model for studying the development of bfCNs.