To improve the preferential GalR binding we synthesized M wh
To improve the preferential GalR2 binding, we synthesized M1152 where the N-terminal Gly residue was deleted, in analogy with previously designed M871 (Sollenberg et al., 2006). In addition to the poor affinity of M871 for GalR1, a recent study showed that it also hardly recognized by GalR3 (Sollenberg et al., 2010). This modification did improve the selectivity, although it also resulted in two-fold lower affinity to GalR2. The design of M1153 was based on the previously published M1145 (Runesson et al., 2009b). As with M1145, the N-terminal part of M1153 was elongated with four amino (R)-baclofen residues (RGRG) from the galanin-like peptide (GALP) sequence (Ohtaki et al., 1999, Runesson et al., 2009b). Furthermore, the N-terminal Gly was substituted with an Asn in order to disturb binding to GalR1 and GalR3 (Runesson et al., 2009b, Sollenberg et al., 2006). The presented ligand M1153 exhibited a noticeably higher selectivity than previously published ligands for GalR2 (Bulaj et al., 2008, Liu et al., 2001, Runesson et al., 2009b, Sollenberg et al., 2006), and it still retains high binding affinity, comparable with the affinity of galanin, for GalR2. We speculate that the improved binding affinity of M1153 could be explained by the possibility that Glu in the side chain of Lys can form additional hydrogen bonds, which may be important for the interactions with GalR2. The observed lower binding affinity of M1151, M1152 and M1153 towards GalR3 could be due to the orthogonally coupled Glu, since GalR3 has been proposed to have a relatively narrow binding pocket compared to GalR2 (). In addition to the remarkable binding properties of M1153, we also demonstrate its agonistic effects for GalR2 by measuring inositol phosphate production in CHO cells. M1153 showed a distinct additive effect in increasing IP production of 0.01μM galanin from concentration of 1μM. The effect increased when 10μM concentration of the M1153 was used, but no significant effect was seen at lower concentrations. It has been previously reported that M871 displays similar properties in vitro (Sollenberg et al., 2006) and in vivo (Kuteeva et al., 2008), but unfortunately it has not yet been verified that M1153 has agonistic effects in vivo similar to the in vitro results. As an implication of the fact that no side effects were seen during the in vivo experiments it would be useful to test M1153 on depression (Kuteeva et al., 2007, Porsolt et al., 1977, Steru et al., 1985) or on status epilepticus models (Bulaj et al., 2008) to confirm its agonistic effects. This study provides the first side-by-side comparison of highly galanin receptor-selective agonists on feeding behavior. Two of the evaluated subtype selective agonists M1153 and M1145 were shown to have no effect on the acute consumption of two highly palatable food substances reported to be stimulated by centrally administered galanin (Kyrkouli et al., 1990, Leibowitz, 2005, Tempel et al., 1988). In contrast, the GalR1 selective agonist M617 notably stimulated acute consumption of high-fat milk, this latter finding replicating and extending a previous study also showing M617 stimulation of cookie mash (Lundström et al., 2005). A concomitant lack of effect of both M617 and M1153 on activity suggested that neither peptide produced side effects that would influence feeding behavior. A previous study employed a similar approach, reporting no feeding stimulation by galanin peptide fragments, Gal(1–16) and Gal(2–29) (Wang et al., 1998). However, Gal(1–16) shows relatively poor affinity for GalR1, with at least a 25-fold lower binding affinity for GalR1 than M617, and it does not differentiate between GalR1 and GalR2. The lack of stimulation by Gal(2–29) reported in that study is consistent with those of the present study. Previous studies on feeding with the antagonistic ligands M40 and C7 reveal that these compounds can block feeding induced by galanin (Bartfai et al., 1993, Crawley et al., 1993). An in vivo study with the non-peptide galanin receptor ligand galnon showed that it strongly reduced food intake (Abramov et al., 2004). Other previous insights about GalR involvement in feeding behavior were reliant upon transgenic and knockout animals. Galanin-overexpressing mice demonstrate weight gain likely due to metabolic actions (Poritsanos et al., 2009) and the specific increased consumption of fat (Adams et al., 2008). More specifically, GalR1 knockout mice showed abnormal adaptation responses to dietary challenges of high fat and high glucose conditions, though normal regulation on low fat, chow diets (Holmes et al., 2003, Zorrilla et al., 2007, Wrenn et al., 2004). GalR2 knockout mice showed no differences in feeding behaviors or body weights (Gottsch et al., 2005). While this dissociation is generally consistent with our present observations, the profound methodological differences, including species, time frame, and genetic versus pharmacological manipulations, make comparisons at best indirect.