Regulation of Cell Polarity and RNA Localization in Vertebrate Oocytes - International Review of Cell and Molecular Biology - page 146

whereas later transport of gdf1 during stages III/IV is dependent on micro-
tubule and microtubule motors ( Yisraeli et al., 1990 ). These motors include
plus-end-directed microtubule motors Kif5b (kinesin heavy chain/kinesin I)
and Kif3 (heterotrimeric kinesin/kinesin II) ( Betley et al., 2002; Messitt et al.,
2008; Yoon and Mowry, 2004 ). Evidence suggested these motors coordinate
transport nonredundantly. Kif5b and Kif3 are both found in gdf1 -RNP com-
plexes ( Messitt et al., 2008; Yoon and Mowry, 2004 ) and inhibition of either
using anitbodies or dominant-negative constructs reduced localization of gdf1
( Betley et al., 2002 ; Messitt et al., 2008 ). Heterologous rescue was possible in
overexpression experiments, suggesting that Kif5b and Kif3 perform similar
functions but are not fully redundant endogenously ( Messitt et al., 2008 ).
Interestingly, in these experiments, injected gdf1 transcripts collected in the
vegetal cytoplasm distal from the GV but did not fully translocate to the cor-
tex. These data suggest that localization through the proximal vegetal cyto-
plasm requires motors other than Kif5b/Kif3b, whereas these mediate
more cortical events. gdf1 and other late pathway mRNAs likely utilize the
subpopulation of plus-end cortical microtubules that exist in the vegetal hemi-
sphere ( Messittetal.,2008 ). The use of different motor proteins may be nec-
essary to coordinate dynamic interactions of the localizing RNP cargoes with
microtubules, ultimately resulting in net flux of RNAs to the vegetal cortex
over the course of oogenesis.
Deletion and site-directed mutagenesis have defined the gdf1 localization
element as a 340-nt long region of the 3 0 -UTR ( Mowry and Melton, 1992 ),
which contains several sets of repeated motifs ( Deshler et al., 1997; Gautreau
et al., 1997; Lewis et al., 2004 ). Short motifs with consensus sequences
UUUCU (VM1) and UUCAC (E2) are necessary for localization of injected
gdf1 transcript ( Deshler et al., 1997; Gautreau et al., 1997; Lewis et al., 2004 ).
Clustering of these motifs is suggested to be a key feature for LE function
( Betley et al., 2002; Kwon et al., 2002; Lewis et al., 2004 ). Several gdf1
RNA-binding proteins that interact with the gdf1 localization element have
been identified and characterized biochemically ( King et al., 2005 ).
The first of these identified was Insulin-like growth factor 2 mRNA-
binding protein 3 (Igf2bp3/Vg1RBP/Vera), a conserved KH-domain
RNA-binding protein that binds to UUCAC motifs in the gdf1 LE
( Deshler et al., 1997, 1998 ). Igf2bp3 associates with the ER and is thought
to mediate its interactions with gdf1 ( Deshler et al., 1997 ). Interestingly, the
presence of UUCAC motifs and Igf2bp3 binding are associated with local-
ized RNAs in many species of chordates ( Betley et al., 2002 ), indicated a
widely conserved mechanism.