Worm Breeder's Gazette 14(3): 30 (June 1, 1996)

These abstracts should not be cited in bibliographies. Material contained herein should be treated as personal communication and should be cited as such only with the consent of the author.

Protein Interactions Between the Receptor Tyrosine Kinase LET-23, LIN-7 and LIN-2

Susan M. Kaech, Stuart K. Kim

Stanford University School of Medicine, Stanford, CA 94305

The activity of LIN-2, LIN-7 and the receptor tyrosine kinase LET-23 are
required for the vulval precursor cell P6.p to receive and transduce the
anchor cell signal that induces the 1! vulval cell fate.  P6.p, like the
other Pn.p cells, is a polarized epithelial cell; its basal surface
faces the anchor cell and its apical surface faces the cuticle.  In
wild-type animals, a LET-23::GFP fusion protein is found in a lateral
ring that is co-localized with and basal to the vulval precursor cell
junctions.  LIN-2 and LIN-7 are necessary for subcellular localization
of LET-23 (the putative receptor for the inductive ligand), since
LET-23::GFP is mislocalized and distributed across the apical surface of
the Pn.p cells in lin-2 and lin-7 mutants (Simske et al., (1996) Cell
85, 195-204).  
        We are interested in the mechanism of LET-23 localization, and
are testing whether there might be direct protein interactions between
LET-23 and LIN-2 or LIN-7.  LIN-2 is a member of a family of proteins
called MAGUKs (for Membrane-Associated GUanylate Kinase) (Hoskins et
al., (1996) Development 122, 97-111).  MAGUKs have been found to
specifically localize to cellular junctions or to regions of cell-cell
contact.  MAGUKs contain at least one PDZ domain, which are domains
known to mediate protein-protein interactions.  LIN-7 contains a PDZ
domain, but does not contain other domains typically found in MAGUKs.
        To test if LIN-2, LIN-7 or LET-23 interact with each other, we
used the yeast two-hybrid system and in vitro binding experiments.  We
found that full-length LIN-7 interacts with both a 200 amino acid region
spanning the PDZ of LIN-2 and with the intracellular region of LET-23 in
the yeast two-hybrid system.  However, we did not detect an interaction
between LIN-2 and LET-23 in yeast. 
        in vitro binding experiments were used to confirm the yeast
two-hybrid results, and to delineate the regions of LIN-7 that bind to
LET-23 and LIN-2.  We tested the ability of different glutathione
S-transferase (GST)::LIN-7 fusion proteins, purified from bacteria and
immobilized on glutathione agarose beads, to specifically bind to LET-23
or LIN-2 from cell lysates.  First, we found that GST::LIN-7 fusion
proteins containing either full-length LIN-7 or only the PDZ of LIN-7
(amino acids 171-297) could bind full-length LET-23 in a cell lysate
from Sf9 cells expressing LET-23 from a baculovirus vector (kindly
provided by Andy Golden).  Second, we observed that GST::LIN-7 fusion
proteins containing either full-length LIN-7 or the N-terminus (amino
acids 1-125) of LIN-7 could bind LIN-2 expressed from transiently
transfected Cos7 cells.  We did not observe any binding between the PDZ
of LIN-7 and LIN-2.  Therefore it seems there are two distinct regions
of LIN-7 that mediate the interactions with LIN-2 and LET-23 (see
illustration).  The N-terminus of LIN-7 (amino acids 1-125) binds to the
PDZ domain of LIN-2, and the C-terminus of LIN-7 (amino acids 171-297)
binds to the intracellular region of LET-23.  
        The above data suggest the possibility that LIN-2, LIN-7 and
LET-23 form a ternary complex.  To test this idea, we used a yeast
"three" hybrid system.  As reported above, we did not see an interaction
between LIN-2 (fused to the GAL4 activation domain, ACT::LIN-2) and the
intracellular region of LET-23 (fused to the GAL4 DNA binding domain,
DBD::LET-23) in the yeast two-hybrid system.  Interestingly, when we
expressed LIN-7 tagged with an SV40 nuclear localization signal in the
presence of ACT::LIN-2 and DBD::LET-23, it appeared that the three
proteins were able to interact.  This ternary interaction was specific
to the combination of LIN-2, LIN-7, and LET-23 as we saw no protein
interactions with the combination of DBD::LET-23, LIN-7 and the yeast
protein ACT::SNF4.  These data suggest that LIN-2 binds LIN-7 and LIN-7
binds LET-23 to form a ternary complex.  Experiments are in progress to
test if these interactions are required for vulval development in vivo.