Regulation of Organogenesis by the Caenorhabditis elegans FoxA Protein PHA-4 J. Gaudet and S. E. Mango

C. elegans strains KK822 (par-1(ts) IV) (11) and EU1 (skn-1(zu67)/DnT1 IV;V (12)) were grown in liquid culture with OP50 as a food source, synchronized, and harvested. For KK822, we grew synchronized homozygotes at the restrictive temperature (25°) and isolated embryos by hypochlorite treatment (J. Sulston and J. Hodgkin, in The Nematode Caenorhabditis. elegans W. B. Wood, Ed. (Cold Spring Harbor Laboratory Press, Plainview, 1988) pp. 587-606). For EU1, we grew synchronized animals and plated young adults. skn-1/DnT1 worms are uncoordinated (Unc), while skn-1 homozygotes are non-Unc, allowing us to enrich for skn-1 homozygotes using a plate crawling assay. We isolated embryos by hypochlorite treatment. To extract total RNA from isolated C. elegans embryos, frozen pellets of embryos were crushed and then resuspended in RNA extraction buffer (1% lauroyl sarcosine, 0.1 M Tris base, 0.1 M NaCl, 20 mM EDTA), followed by several rounds of phenol:chloroform extraction and ethanol precipitation. Once-selected poly-A+ RNA was purified using the PolyTract Isolation Kit (Promega). par-1 cDNAs were labeled with Cy3 and skn-1 cDNAs were labeled with Cy5. Labeled cDNA was prepared from 5 g of poly-A+ RNA by the Huntsman Cancer Institute Microarray Core Facility. Construction and probing of the microarrays is described by Reinke et al. (10).

Genes in the positive set had an average par-1/skn-1 ratio ( 2.0, and were expressed above background in at least two of the three experiments. We chose a cut-off of par-1/skn-1 ( 2.0 based on an empirical analysis of our microarray data. Using this threshold, our positive set included 15 of 18 known pharyngeal genes, and only 2 known non-pharyngeal genes. A less stringent threshold (par-1/skn-1 ( 1.7) only slightly improved our detection of known pharyngeal genes (16 of 18) but signifcantly increased the number of false positives (8 known non-pharyngeal genes). A more stringent approach was to include as positives those genes which were more abundantly expressed in par-1 vs. skn-1 at the 95% confidence level (calculated as in 10). Using this approach, we obtain only 172 positives, including only 13 of 18 known pharyngeal genes and 2 known non-pharyngeal genes. Therefore, the threshold of par-1/skn-1 ( 2.0 gave the best balance between selectivity and specificity.

Construction of GFP reporters

Because C. elegans promoters are generally small (for examples, see P. G. Okkema, S. W. Harrison, V. Plunger, A. Aryana, A. Fire, Genetics 135, 385 (1993); and J. D. McGhee, M. W. Krause, in C. elegans II D. L. Riddle, T. Blumenthal, B. J. Meyer, J. R. Priess, Eds. (Cold Spring Harbor Laboratory Press, Plainview, 1997) pp. 147-184), we used ~500-1000 base pairs of sequence upstream of the predicted ATG start codon to generate our reporter constructs (see Figure 1). To construct GFP reporters, promoter fragments were amplified from wildtype (N2) genomic DNA using gene specific oligonucleotides. Each oligonucleotide pair was engineered with unique restriction sites to facilitate cloning of the promoter fragments into the GFP::HIS2B reporter vector pAP.10 (which was built from a GFP::HIS2B reporter kindly provided by G. Seydoux). This cloning strategy results in a transcriptional fusion to the GFP::HIS2B reporter, except in the case of the intron-containing reporters, in which GFP was fused in-frame to coding sequence in the second exon (for avr-15 and peb-1), or fourth or fifth exon (for H30A04.1) of the gene. To generate site-specific mutations within the promoter fragments, we used a PCR-based mutagenesis approach, as described by S. N. Ho, H. D. Hunt, R. M. Horton, J. K. Pullen, L. R. Pease, Gene 77, 51-59 (1989). Site-specific mutations were verified by sequencing across the promoter regions of each reporter. C. elegans transgenic lines were created using standard microinjection techniques (C. C. Mello, J. M. Kramer, D. Stinchcomb, V. Ambros, EMBO J. 10, 3959-3970 (1991)). Microinjection experiments with GFP reporters were performed with PCR fragments that extended from the promoter to the end of the unc-54 3' cassette present in pAP.10. Deletions within promoters were generated by PCR amplification using oligonucleotides internal to the promoter. Except where noted, all injection mixes contained 10 ng/L pRF4 cut with EcoRI, 2 ng/L of a GFP reporter, and 88 ng/L sheared herring sperm DNA. Transgenic lines were established by following the dominant Roller phenotype conferred by pRF4 (J. M. Kramer, R. P. French, E.-C. Park, J. J. Johnson, Mol. & Cell. Biol. 10, 2081-2089 (1990)). For each reporter tested, a minimum of 3 independent transgenic lines were examined.

Construction of the pha-4(ts) strain

The pha-4(ts) strain was engineered from a pha-4 nonsense allele, zu225, and the temperature-sensitive mutation smg-1(cc546ts) (S. Getz, S. Zu, A. Fire, pers. comm.) In C. elegans, the products of the smg genes are required for mRNA surveillance. mRNAs carrying a premature stop codon are recognized by this surveillance system and rapidly degraded. A mutation in any one of the smg genes results in the stabilization of mRNAs with premature stops. By placing the nonsense pha-4 allele in a smg-1(ts) background, we render pha-4 synthesis sensitive to temperature. The zu225 allele introduces a premature stop codon (Gln384(stop) downstream of the predicted DNA-binding domain of PHA-4 (M. Horner and S.M., unpub. obs.). Although pha-4(zu225) is predicted to generate a truncated protein, a temperature insensitive strain (constructed with the null mutation smg-1(r861)) is viable, indicating that the C-terminal region of PHA-4 is dispensible for activity (L. Kaltenbach and S.M.., unpub. obs.). Results using the pha-4(zu225) allele are presented, but we obtained similar results using a different pha-4 nonsense allele, q500.

Viability tests with pha-4(ts)

For viability tests, embryos were collected from pha-4(ts) adults for two hours and were allowed to develop at the permissive temperature for either an additional hour ("early embryos", in which the pharynx primordium has formed), five hours ("late embryos", in which the pharynx has attached to the buccal cavity and pharyngeal cells are differentiating), or ten hours ("L1s", in which the pharynx is fully formed), prior to the shift to the restrictive temperature. Embryos were counted following collection, and the number of adults arising from the embryos was determined 3-6 days later to allow sufficient time for development. The fraction of embryos that gave rise to fertile adults is the "% viable".

Expression of pharyngeal reporters in pha-4(ts)

Embryos were collected from pha-4(ts) adults for two hours and were allowed to develop at the permissive temperature until hatching (~10-12 hours). Hatched larvae were grown at the restrictive temperature for 5 days, at which time we compared pharyngeal expression of these shifted animals and animals grown continuously at the permissive temperature. To restore PHA-4 activity in downshifted animals, larvae grown at the restrictive temperature (15°) for 5 days were returned to the permissive temperature (24°) for one day and then scored for pharyngeal GFP expression.

We observed a four day lag between the time animals were shifted to restrictive temperature and the first observed decrease in GFP expression. This lag probably reflected the stability of protein and possibly mRNA for both pha-4 and GFP (termed 'perdurance'). To examine the perdurance of GFP, we created a transgenic strain whose GFP expression was temperature sensitive, but not dependent on pha-4. We used a body wall muscle reporter whose expression is smg-inducible (myo-3::gfp::let-858; S. Getz, S. Zu, A. Fire, pers. comm.), and introduced this reporter into the smg-1(cc546ts) strain to generate a temperature sensitive GFP reporter. These transgenic animals grown at the permissive temperature and then shifted to the restrictive temperature showed no significant decrease in GFP expression until three days after downshift (data not shown). This result suggests that the perdurance of GFP is a significant factor in the persistent expression of our reporter constructs in pha-4(ts) animals.

Electrophoretic mobility shift assays

The electrophoretic mobility shift assays were performed essentially as described by (8), except that 100 ng of herring sperm DNA was included to eliminate non-specific binding to radiolabeled DNA. Each competition contained 1/10 volume of a PHA-4 transcription/translation reaction (TNT T7 Quick System, Promega, using as template a PCR fragment containing the T7 promoter in front of the pha-4 cDNA) and 0.25 fmol of labeled duplex in a 20 L volume. Cold competitor was added from 100- to 3000-fold molar excess over labeled duplex. DNA was quantified spectrometrically as well as visually. We used an excess of protein in the binding reactions to obtain observable competition at 300-fold excess, allowing us to observe subtle affinity shifts in either direction. For comparison of different DNA sequences, we used a single protein preparation, and performed binding assays for each DNA in triplicate. The myo2-WT duplex was made by annealing two oligonucleotides, C183-1 (5' TGTCTCGTTGTTTGCCGTCGGATGTCTGCC3') and C183-2 (5' TGGGCAGACATCC GACGGCAAACAACGAGA 3'), and radiolabeled with the Klenow fragment of DNA PolI, [-³²P]-dCTP, and other dNTPs. Labeled duplex was purified over a Sephadex G50 spin column. Unlabeled competitor duplexes were prepared in the same manner, with cold dNTPs. PhosphorImager data were quantified using ImageQuant software, and the relative affinities of each annealed duplex was determined by measuring the ratio of shifted to total labeled duplex (DP/D_tot) in competition reactions (comp.) versus control reactions (cont.) lacking specific competitor. The data were plotted as "relative amount shifted" ([DP/D_tot]_comp/[DP/D_tot]_cont) versus concentration of competitor, allowing us to estimate the concentration of competitor oligonucleotide required to compete 50% of the shifted complex (IC₅₀). The IC₅₀value is inversely proportional to the binding affinity of the protein-DNA interaction (35).

Supplemental Table 1. Known pharyngeal genes detected by the microarrays. 'Average Log Ratio' is the average log2(par-1/skn-1) for each gene. Genes included in the positive set (top 15 lines) had an average log ratio (1.0. The three genes not detected as positives in our microarray experiments (myo-1. kel-1, and klp-3) are all first expressed in very late embryos. However, we still detect many genes first expressed in very late embryos (e.g. eat-20, F44A2.5, F53B3.3 [data not shown]), suggesting late genes have not been lost systemically. *Additional references for expression data are: tnc-2: H. Terami and H. Kagawa, pers. comm.; T11B7.4: C. McKeown and M. Beckerle, pers. comm.; ajm-1: W.A. Mohler, J.S. Simske, E.M. Williams-Masson, J.D. Hardin, and J.G. White, Curr Biol 8, 1087-1090 (1998); ceh-32: C. Dozier and T. Burglin, pers. comm.; exp-2: M.W. Davis, R. Fleishhauer, J.A. Dent, R.H. Joho, and L. Avery, Science 286, 2501-2504 (1999); lit-1: Y. Shostak and K. Yamamoto, pers. comm.; nid-1: S.H. Kang and J.M. Kramer, Mol Biol Cell 11, 3911-3923 (2000); kel-1: M. Ohmachi, A. Sugimoto, Y. Iino, M. Yamamoto, Genes to Cells, 4, 325-337 (1999); klp-3: M.L.A. Khan, et al., J Mol Biol, 270, 627-639 (1997). I. Hope's data comes from his Expression Pattern Database at http://129.11.204.86:591/default.htm (I. Hope, pers. comm.).
Gene	Average Log Ratio	Ref.
myo-2	2.86	(32,33)
tnc-2	2.37	*
T11B7.4	2.40	*
pha-4	2.18	(6, 8)
jam-1	1.98	*
ceh-22	1.83	(14)
F44A2.5	1.72	I. Hope
F53B3.3	1.51	I. Hope
F53H4.5	1.53	I. Hope
ceh-32	1.49	*
exp-2	1.32	*
lit-1	1.25	*
eat-20	1.10	(18)
nid-1	1.06	*
peb-1	1.05	(19)
myo-1	0.74	(32,33)
kel-1 (C47D12.7)	0.38	*
T09A5.2 (klp-3)	0.35	*

Supplemental Figure 1. Pharyngeal expression requires predicted PHA-4 binding sites. Cartoons illustrating the promoter fragments used to drive expression of a GFP reporter. Triangles represent positions of TRTTKRY sequences on the top or bottom strand, black boxes represent predicted coding sequence. 'Pharynx expression' is the relative strength of pharyngeal GFP at its peak in embryogenesis as assessed by visual examination. i) to iii) Introns containing (2 predicted PHA-4 binding sites are required for pharyngeal expression.

Medium version | Full size version

Supplemental Figure 2. Electrophoretic mobility shift assays with PHA-4. (A-B) Different exposures of the same gel shown in Figure 3B of the main text. Here the full gel is shown to indicate equal addition of labeled duplex to each reaction and equal loading of the binding reactions on the gel. Lanes are (from left to right): F = free labeled duplex; M = labeled duplex and 1/10 volume of reticulocyte lysate; P = labeled duplex and 1/10 volume of in vitro transcribed/translated PHA-4. Numbered lanes are labeled duplex, 1/10 volume of in vitro transcribed/translated PHA-4, and cold competitor duplex. Lanes 1-4 contain myo2-WT cold competitor duplex at 100x, 300x, 1000x, and 3000x molar excess. Lanes 5-8 contain D2096-WT at 100x, 300x, 1000x, and 3000x molar excess. Lanes 9-12 contain D2096-mut at 100x, 300x, 1000x, and 3000x molar excess.

Medium version | Full size version

Supplemental Table 2. Expression of microarray positives, assayed with GFP::HIS2B reporters. Positives are ranked according to their average log2(par-1/skn-1) ratio; *F28F9.1 has a par-1/skn-1 ratio below the cut-off for inclusion in the positive set (i.e. <2.0). Expression: pm = pharynx muscle, mc = marginal cells, e = epithelial cells, vpi = pharynx-intestinal valve, bm = body muscle, epid = epidermis, gd = gonad, rect = rectal cells. Predicted PHA-4 sites are TRTTKRY sequences present in the promoters or introns of reporters. a reporter containing ~1.1 kb of H30A04.1 sequence is not expressed in the pharynx, while the addition of another ~400 bp of downstream sequence results in pharyngeal expression. H30A04.1 is eat-20 (18); this identification was not known when H30A04.1 was selected for analysis. The expected occurance of the sequence TRTTKRY is approximately once per 296 base pairs, or 3.378 times per 1000 kbp (assuming an average G+C content of 40%). Therefore, the probability of a given 1 kbp sequence containing 'n' TRTTKRY sites is determined by the Poisson distribution, such that P(n) = (3.378n)(e-3.378)/n! Therefore, P(0) = 0.034, P(1) = 0.115, P(2) = 0.195, P(3) = 0.219, P(4) = 0.185, P(5) = 0.125, P(6) = 0.070, P(7) = 0.034, P(8) = 0.014, P(9) = 0.005, P(10) = 0.002, etc.
Gene Reporter	Predicted Product	Average log2 (par-1/skn-1)	Rank (of 240)	Pharyngeal Expression	Non-pharyngeal Expression	# Predicted PHA-4 Sites
D2096.6	unknown	2.82	14	pm, mc, e	epid, gut, rect	3
M05B5.2	unknown	1.96	53	all	gut, rect, gd	5
C44H4.1	Leu-rich repeats	1.69	81	all	-	9
C27C12.6	DM domain	1.67	84	pm8, vpi	-	5
ZK816.4	unknown	1.46	110	pm	-	3
H30A04.1
(1.5 kb)	EGF repeats	1.10	197	pm3,4,6	neurons	4
F45G2.2	myosin HC	1.09	202	pm	bm	3
T05E11.3	chaperonin	1.03	231	all	gut, rect	6
F10B5.3	Zn-finger	1.81	68	-	epid	4
H30A04.1
(1.1 bp)	EGF repeats	1.10	197	-	neurons	0
K08A8.2	HMG Box	1.10	200	-	-	1
ZK685.2	LRR repeats	3.22	6	-	gut	0
F28F9.1	homeodomain	0.93	*	-	gut, neurons	0

This file is in Adobe Acrobat PDF format. If you have not installed and configured the Adobe Acrobat Reader on your system, please see Help with Printing for instructions.

Supplemental Table 3