During the excision of Tc1 transposons, flanking DNA is also often deleted (Plasterk, 1991). So although the available Tc1 insertion alleles apparently fail to disrupt kin-15 or kin-16 gene function, they do provide an efficient substrate for the isolation of small deletions. Zwaal et al. (1993) have demonstrated that in a mut-2 background Tc1-mediated deletions occur with sufficient frequency (10-3 per allele per generation ) that one can reasonably conduct a screen for spontaneous deletion alleles using PCR.
Because kin-15 and kin-16 may be redundant (see above), my first priority will be a screen for Tc1-mediated deletions that disrupt both genes. (Furthermore, in vitro mutagenized transgenes can be introduced into this double mutant strain to examine the phenotypic effects of individual gene mutations -- see below.) The PCR screen for this "double mutant" deletion, designated here as Df|kin-15+16, will be conducted as described previously (Zwaal et al., 1993; Williams et al., 1992, Bürglin et al., pers. comm) using the kin-16 (ms8::Tc1) allele, in which the Tc1 transposon is conveniently positioned between kin-15 and kin-16.
To conduct the PCR screen, 100 cultures with approximately ten YK8 (mut-2(r459) I; kin-16(ms8::Tc1) II; dpy-19(n1347) III) hermaphrodites will be established (note that YK8 also contains the mutator gene mut-2 to insure frequent transposon excision in the germline). After reproducing for two or three generations (whereupon the E. coli food source will be exhausted), approximately half of the worms will be harvested for PCR analysis; the other half will remain on the original plate.
After a rapid purification procedure requiring no extractions or ethanol precipitations (Williams et al., 1992; Bürglin et al., pers. comm), a crude lysate from each culture will be screened for desired deletions by PCR as described previously (Zwaal et al., 1993). The primers L5#2 and L5#13, which evenly flank the Tc1 insertion site (see Figure 2), are optimally positioned for the detection of "double mutant" deletions: Because the primers are normally >3 kb apart, no PCR product will be detected (under the appropriate conditions) in YK8 animals, except in those rare cases where misexcision of the Tc1 transposon has removed 1-2 kb of flanking DNA.
Because Tc1-mediated deletions are detected with the above procedure at a frequency of 10-2 - 10-3 per allele per generation (Zwaal et al., 1993; Plasterk, pers. comm.), at least one of the initial 100 cultures (total of ~1000 parental worms) should have a desired deletion event in the first generation. Once a positive culture is identified, we will return to the sibling worms saved on the original plate to set up subcultures so that the deletion chromosome can be cloned by multiple rounds of PCR analysis and sib-selection (if no positives are found, the screen will be repeated). Assuming a frequency of >1 deletion allele / 1000 animals (the estimated detection limit of this PCR assay; Bürglin et al., pers. comm.), approximately 3000 animals from the original positive culture must be screened by PCR to be confident of picking a deletion-bearing animal (Table 1). One possible sib-selection scheme, within the limitations of a standard temperature cycler, is shown in Table 1.
TABLE 1. Possible sib-selection strategy for cloning a deletion chromosome.
|
Expected frequency of deletion allele1 |
|
|
|
--in 1° positive culture: 1/1000 worms |
|
--30 subcultures of 100 |
|
--in 2° positive culture: 1/100 worms |
|
--30 subcultures of 10 |
|
--in 3° positive culture: 1/10 worms |
|
--30 subcultures of 1 |
1 The deletion allele may be at a lower frequency if partially or completely lethal (but should be propagated in heterozygotes, since large deletions of the kin-15 / kin-16 region have been routinely isolated; see, for example, Sigurdson et al., 1984).
2 At 95% confidence level; calculated using N = ln(1-P)/ln(1-f), where N = number to screen, P = desired probability (95%), and f = frequency of deletion allele (Sambrook et al, 1989).
3 Fewer rounds can be conducted if more cultures and PCR reactions are done at each round. For example, if 100 cultures are established, only two rounds are needed (however, this requires twice as much reagents, 200 vs. 90 plates and PCRs, and the available temperature cycler has only 30 slots.).
Once animals bearing the "double mutant" deletion have been cloned, the mutator allele (mut-2) and any background mutations (including dpy-19) will be removed by repeated outcrossing to a marked strain (Table 2). If Df|kin-15+16 is a recessive lethal mutation, not all F1 progeny will have the deletion (Table 2). The presence of the Df|kin-15+16 deletion can be determined by crossing individual F1 males with unc-4 hermaphrodites and checking a fraction of each F2 population by PCR; if any F2 fractions are negative for the deletion allele, the deletion-bearing parental strain must have been heterozygous. Once identified by PCR and their Unc-4 male progeny (Table 2), Df|kin-15+16 / unc-4 males will be outcrossed ten times to unc-4 hermaphrodites to insure (with >95% confidence) that all unlinked mutations have been removed. When finally outcrossed, the deletion allele will be maintained over the mnC1 crossover suppressor chromosome after crossing to unc-4 / mnC1 dpy-10 unc-52 males.
It is possible that the Df|kin-15+16-bearing F1 males (see Table 2) will not mate efficiently (perhaps due to some background mutation); in that case, individual F1 hermaphrodites will be repeatedly outcrossed to unc-4 / mnC1 males. In the F1 generation, Df|kin-15+16 / unc-4 animals will again be identified by PCR and their Unc-4 male progeny; in subsequent generations, this genotype will be identified as non-Unc animals that fail to produce Dpy Unc (mnC1 / mnC1) progeny when crossed to unc-4 / mnC1 males.
TABLE 2. Expected F1 and F2 outcross
progeny of Df|kin-15+16 hermaphrodites
(P0: Df|kin-15+16 /
Df|kin-15+16 OR Df|kin-15+16 / + x
unc-4 / mnC1 dpy-10 unc-52 males)1
|
|
|
|
|
|
|
Df|kin-15+16 / unc-4 |
x unc-4 / unc-4 -> |
Df|kin-15+16 / unc-4 & unc-4 / unc-4 |
|
|
|
Df|kin-15+16 / mnC1 |
x unc-4 / unc-4 -> |
Df|kin-15+16 / unc-4 & unc-4 / mnC1 |
|
|
|
+ / unc-4 |
x unc-4 / unc-4 -> |
+ / unc-4 & unc-4 / unc-4 |
|
|
|
+ / mnC1 |
x unc-4 / unc-4 -> |
+ / unc-4 & unc-4 / mnC1 |
|
|
1 Df|kin-15+16 indicates the "double mutant" deletion; the complete genotype is mut-2(r459) I; Df|kin-15+16 / Df|kin-15+16 or + II; dpy-19(n1347) III; unc-4 / mnC1 dpy-10 unc-52 males must be used because unc-4 / unc-4 males do not mate efficiently (Hodgkin et al., 1991).
2 If Df|kin-15+16 parent was homozygous, expect first two F1 classes only; if heterozygous, expect all four.
3 The presence of the Df|kin-15+16 and unc-4 chromosomes will be verified in the F2 by PCR analysis of sibling DNA and visual inspection of males (to distinguish from F1 self-progeny), respectively.
Rather than repeating the laborious PCR screen for deletions that disrupt kin-15 or kin-16 individually, the phenotype of the single mutants will be examined after complementation of the "double mutant" with plasmids that restore the function of each gene separately. This approach should be quicker (the necessary plasmids are already available or easily constructed) and avoids a potential pitfall in the analysis of a kin-16 deletion allele: Because kin-15 and kin-16 are transcribed as a polycistronic pre-mRNA (Spieth et al., 1993; Morgan and Greenwald, 1993), a deletion in the upstream kin-16 coding sequences may also disrupt kin-15 gene function.
kin-15 and kin-16 single mutant lines can be quickly constructed by crossing the "double mutant" strain to GS484 and GS485, which harbor kin-16::lacZ, kin-15(+) and kin-16(+), kin-15::lacZ transgenes, respectively (Morgan and Greenwald, 1993). However, because insertion of lacZ sequences into the upstream kin-16 transgene may disrupt kin-15 gene function as well (see above), in addition we will create a transgenic line harboring a kin-16 point mutation. This kin-16(0), kin-15(+) plasmid can be quickly constructed from pSKKX1, which contains a unique XhoI restriction site near the 5' end of the kin-16 coding sequences (at nucleotide 295; Morgan and Greenwald, 1993). Digestion of pSKKX1 with XhoI, fill-in of the resulting overhangs with the Klenow fragment of E. coli DNA polymerase I, and blunt end ligation will be performed to create a four base pair insertion and produce a frame-shift mutation.
Once available, the kin-15 and kin-16 single and double mutants will first be examined under the dissecting microscope for any discernible phenotype, including lethality (readily discernible in synchronized populations as dying eggs or larvae). Subsequently, the cellular nature of any defects, including the cell lineage pattern, will be discerned using a light microscope equipped with Nomarski DIC optics.
Based on the expression pattern of the kin-16::lacZ fusion gene, my initial attention will obviously be focused on the development of the large hypodermal syncytium (hyp7). In particular, mutant animals will be examined by immunofluorescence microscopy after staining with the monoclonal antibody MH27 (specific for adherens junction), previously used by Podbilewicz and White (1994) to examine the pattern of hypodermal cell fusions during larval development.
The possible models for the regulation of KIN-15 and KIN-16 kinase activity (see Background and Significance) make specific predictions about which protein domains will be essential for the regulation of kinase activity. To identify the probable functional domains of KIN-15 and KIN-16, the evolutionarily conserved regions will be identified in related nematode species.
To start, kin-15 and kin-16 homologues will be cloned by screening the genomic libraries of a closely related species, Caenorhabditis briggsae (library gift of D. Riddle), and a slightly more distant nematode, Panagrellus redividus (library gift of C. Link). To minimize the effects of individual gene divergence, the probe will be a mixture of kin-15 and kin-16 cDNA sequences. Plating of each nematode genomic library on the appropriate lambda host, transfer of recombinant phage DNA to nitrocellulose membranes, and random oligonucleotide labelling of the probe will be performed as previously (Morgan and Greenwald, 1993; Sambrook et al., 1989). To account for the expected sequence divergence between the nematode homologues, probe hybridization and subsequent washes will be performed under reduced-stringency conditions as described in Way et al. (1991). Once positive plaques have been purified to homogeneity by repeated rounds of screening, the most strongly conserved restriction fragments of each insert will be distinguished by Southern (DNA) blot hybridization using the kin-15 + kin-16 cDNA probe under reduced-stringency conditions.
After subcloning these fragments into pBluescript vectors, phage-rescued single stranded DNA will be subject to DNA sequence analysis using the Sequenase enzyme (U.S. Biochemical), as performed previously (Morgan and Greenwald, 1993). The resulting sequence data will be analyzed with MacVector (IBI) software to identify any open reading frames and corresponding amino acid sequences with homology to C. elegans kin-15 and kin-16 gene sequences.
For positive phage inserts, extended sequence analysis will be conducted to identify the complete coding sequences and potential regulatory regions. After compiling raw DNA sequence data with AssemblyLign (IBI) software, the completed DNA sequence will be analyzed with MacVector , BLAST (NCBI) and other software available on the Internet (Henikoff, 1993) to identify potential protein-coding regions based on the presence of extended ORFs, observed regions of codon bias, and conservation of encoded amino acid sequences with KIN-15/16 and other PTKs (Devereaux et al., 1984).
Once the potential protein coding regions have been identified, conserved, potential cis regulatory DNA elements will be distinguished by comparing DNA sequences between the gene homologues. More importantly, amino acid sequence comparison will identify the conserved, presumably functional protein domains. The kinase domain, critical for enzyme activity, should be well conserved. The pattern of amino acid homology should allow us to distinguish between different models for the regulation of PTK activity. For example, if KIN-15 and KIN-16 interact with a cell surface protein, the amino terminal sequences should be more highly conserved than other regions of the protein. Alternatively, if kinase activity is solely regulated by intracellular factors, only sequences in the juxtamembrane, kinase insert, or C-terminal region (or combinations of these three) should be more highly conserved.
[This section deleted from Internet version.]