Yasumi Ohshima, for a friendly, productive and (very) long-distance collaboration on the let-23 cloning project. I thank the members of the Sternberg lab, who were all part of the creative and often crazy. These five let-23 functions can be independently mutated, indicating that the let-23 protein encodes tissue-specific functions.
The let-23 protein possesses putative ligand-binding, transmembrane, and tyrosine kinase domains, as well as cysteine-rich regions, all with the hallmarks of the EGFR family. Furthermore, genetic data indicate that let-23 acts upstream of the let-60 ras gene, supporting mammalian studies suggesting a link between EGFR and ras.
Summary
The publication included as Chapter II (Aroian and Sternberg, 1991) describes the first aspect of my research: genetic characterization of the let-23 locus. I adopted from Paul two different mutagenesis screens that relied on let-23 vulval phenotypes to isolate new let-23 alleles. Gregg Jongeward in the lab has since used two of these alleles to isolate let-23 suppressors and expand our understanding of vulval pathways.
Briefly, the results of this analysis are: (1) complete loss of let-23 function results in larval death; (2) let-23 mutations affect at least five tissues; (3) loss of let-23 function in the vulva leads to vulva absence. The let-23 gene required for Caenorhabditis elegans vulval induction encodes a tyrosine kinase of the EGF receptor subfamily.
Chapter I
Aroian and Paul W. Sternberg
To determine whether distance from the anchor cell was important, all but one VPC was ablated in the presence of the anchor cell ( Sternberg and Horvitz, 1986 ). This result suggests that the operation of the anchor cell signal does not require contact between the anchor cell and the VPCs. Thus, the anchor cell is required both for the induction of VPCs and for the proper movement of their great-granddaughters.
For example, in the case of let-23 RTK, mapping of the gene between an RFLP and a deficiency. However, in the wild-type hermaphrodite, the simplest model would be that the anchor cell signal dominates the effect of the Muv genes in P5.p-P7.p (Figure 15C; Herman and Hedgecock, 1990). In the presence of the anchor cell, the VPC closer to the inducing signal always selects the 1°.
Therefore, the anchor cell signal and high activation of the intrinsic 1° pathway can override the choice of the 2° fate.
ON OFF
The orientation is the same as in Figure 1. C) The elimination of one of the somatic gonad cells, the anchor cell, gives the same result, indicating that the anchor cell is necessary for. The anchor cell in the gonad releases a diffusible, graded signal that is transduced by the six vulval precursor cells, or VPCs. This induction of vulval fate occurs even if the anchor cell is ablated, so presumably Muv.
Unlike the Muv phenotype, the VPCs in a Hin mutant still behave as if they are responding to the anchor cell signal: (1) there is a gradation of VPC fate around the anchor cell; and (2) VPCs in a Hin hermaphrodite will select 3°. It is hypothesized that the influence of the anchor cell on vulvar fate selection is mediated through lin-3, let-23 RTK and let-60 ras, as the elimination of one of these genes mimics anchor cell ablation. As shown in the last panel, the signal from the anchor cell probably does not pass through the Muv genes (see text and Figure 16).
However, if the anchor cell is removed in these mutant animals, then the fates of all VPCs are variable. Thus, the anchor cell can influence VPCs even though lin-I5 is absent, suggesting that the anchor cell signal does not pass through lin-I5. The vulva detennation pathway (1°/2'vs.:J' insert) as defined ~cJoulje mutantphen~ The anchor cell signal is received ~let-Z:JKlK which allows let-23 KlKto overcome ~lin-15 inhibition.
NolEs: lin-lis shown to be consistent with the anchor cell signal sidenitis completely epistatic for all V ul genes shown. In the absence of the anchor cell signal, lin-15 and neighboring VPCs, a VPC will choose the 1° fate. The anchor cell is missing because lin-12(dJ mutations also result in no anchor cell being produced.
If anchor cell signaling is turned on, the inhibitory effects of lin-15 can be overcome. If anchor cell signal is close, an isolated VPC will choose 1° fate via LP.
First, lin-12 is active in the same cell, although not sufficient to inhibit the strmglyactive intrinsic pathway. In Pl.p, a medial afferent cell signal is received, moderately activating the intrinsic pathway which in turn activates lin-12. As a result, lin-12 is activated at a high enough level and the internal pathway to a low level of coughing will be successfully completed by lin-12 and Z' mm. p emits a side signal, which tends to be weak, since the inner path is turned fromL.
Net strain: I.P. In P8.p, only a small amount of anchor cell signal is received, weakly activating the intrinsic pathway and lin-12. As a result, the intrinsic pathway is blocked and the lin-12 is newer at a level high enough to promote Z'mm. The essential features of this model are: l) The level of the anchor signal is important for how strong the I.P. 2) 'The intrinsic pathway turns on Z' mm and the lateral signal.
For example, along the lines of cfintracellularinhilitim (21), IDF receptor is supposed to activate phospholipaseG-y, which, among other things, results in the activation of protBn-kinase C, which in turn inactivates receptor (reviewed in Ullrich and& :hlessinger, 19D). Thus, one pathcflet-Z3RTK stirnulates vulval induction via.let-8Jras lutanother path oflet-Z3 RTK mightinhilitvulval induction via a negative feedlBcklcq>. In mct, such a signal is cmsis1Entwith the lateral signal previously discussed, since one aspect of the lateral signal might inhilit the I.Pin a neighboring cell.
This loss of only the hilitory pathway could occur because activation of the inhil:itory pathway (eg via. Pl.0y or lateral signal) is more sensitive loan initial reduction in let-Z3 RTKiOOs than activation of the stirn ulatory pathway ( via.let-8Jras ). The semi-vulval part of the genus coincides with the location of the anchor cell relative to the hybrid genus, so that for P5.p the vulval part faces posteriorly, for P7.p the vulval part faces anteriorly, and for P6.p the vulval part faces either anteriorly or posteriorly. These data are interpreted to mean that some factor necessary for the execution of vulval lineages is being asymmetrically distributed in each VPC (shaded portion of the cells) based on the direction of the anchor cell signal.
00 IOO~ H"
The let·21 gene encodes a tyrosine kinase of the EGF receptor subfamily (AROIAN et al. 1990) and acts as vialet-60 ras (HAN. The structure of the let-21 gene suggests that it could be a receptor for an inducible signal (AROIAN). et al. 1990). The penetrance of the P 12 transformation was measured in Unc progeny sy97 unc-4 and SJ97 unc-Jf/", nCI mothen.
This greater possibility is intriguing given that let-23 pl'Otein is a member of the ECF receptor tyrosine kinaSf. For example, an inhibitory pathway could involve intracellular downregulation of an inductive signal receptor that could be id. The whole operation of ltl·2} gen(' is therefore complicated. Id-2) genr- appears!.
Let-23 has been shown to encode a putative tyrosine kinase of the epidermal growth factor receptor subfamily. Because let-23 acts upstream of let-60 ras in the vulval determination pathway, the identification of the let-23 product provides support for a link in vivo. Here we report that Ir,·}3 codes for a protein from the epidermal growth factor (EGF) receptor tyrosine kinase subfamily.
With previous data indicating that Irl·}] acu upstream of the Irl·60 ras gene in the vulval determi·. One of the two hydrophobic stretches follows the initiator methionine and a basic residue (Arl) and appears to be a leader sequence for insense in the membraneJ•• The other hydro. If it occurs from the lene products, we suggest that kt-23 encodes the receptor for the vulval inductive liln&1.
Chapter IV
I decided to tie together my genetic and molecular studies of the let-23 gene by sequencing the mutations associated with various let-23 alleles. Most of the protein was tested with the results that approximately half (11/20) alleles were definitively localized (see Discussion). The resulting late-23 protein will lack 85% of the wild-type protein, including the entire cytoplasmic domain.
Of the non-wild-type splices, all but one result in a stop codon near the end of exon 16 (the splice that removes four bases from exon 17 encodes a long, nonsense protein starting at the end of exon 16). I am encouraged by the fact that mutations have been found all over the part of the gene. In addition, I am encouraged by the fact that some of the sequenced mutations are new to EGFR.
In support of B, n1045 may cause unusual receptor activation by deletion. An interesting possibility is that this inhibitory pathway involves the lin-12 pathway (see Chapter 1) of the five let-23 decisions. The diversity and sequence of 3' splice acceptor sites, including the altered wild-type site (from AG to AA) used as a result of the n1045 mutation, is unprecedented in the field of splicing.
One common feature of these unusual nl045 sites is the presence of either the penultimate A or the ultimate G of the wild-type AG site. The sequencing of the sy97 mutation provides strong evidence that tissue specificity of receptor tyrosine kinase may be present in the C-terminus. The molecular basis of familial apoE deficiency: an acceptor splice site mutation in the third intron of the deficient apoE gene.
The location and changes associated with three lethal mutations are indicated along with their amino acid. Location of the sy 11 and sy 12 mutations relative to the conserved cysteine spacing in the EGFR family (adapted from Aroian et al., 1990, Figure 6b). The numbers represent the number of amino acid residues between cysteines for let-23, DER and human EGFR.
The cysteine mutated in the sy12 allele is underlined and in bold; the change to tyrosine is shown. The stretch of five amino acids between two cysteines harboring the syll mutation is indicated by an underlined bold face. Wild-type genomic sequence from the end of exon 16 to the beginning of exon 17 (including intron 16) is shown.
A faint fourth band at 741 bases is the size expected for genomic DNA or an unspliced message. The wild-type size band of nl045 RNA is a mixture of at least six different types of RNA. The location of the sy97 mutation at the end of intron 17 is shown along with the location of the analogous change in intron 16 associated with nl045.
