(Received for publication, March 13, 1997, and in revised form, April 11, 1997)
From the Department of Pharmacology and the
§ Howard Hughes Medical Institute, University of Texas
Southwestern Medical Center, Dallas, Texas, 75235-9050
Substantial guanylyl cyclase activity was detected in membrane fractions prepared from Caenorhabditis elegans (100 pmol cGMP/min/mg at 20 °C or 500 pmol cGMP/min/mg at 37 °C), suggesting the potential existence of orphan cyclase receptors in the nematode. Using degenerate primers, a cDNA clone encoding a putative membrane form of the enzyme (GCY-X1) was obtained. The apparent cyclase was most closely related to the mammalian natriuretic peptide receptor family, and retained cysteine residues conserved within the extracellular domain of the mammalian receptors. Expression of the cDNA in COS-7 cells resulted in low, but detectable guanylyl cyclase activity (about 2-fold above vector alone). The extracellular and protein kinase homology domain of the mammalian receptor (GC-B) for C-type natriuretic peptide (CNP) was fused to the catalytic domain of GCY-X1 and expressed in COS-7 cells to determine whether ligand-dependent regulation would now be obtained. The resulting chimeric protein (GC-BX1) was active, and CNP elevated cGMP in a concentration-dependent manner. Subsequently, a search of the genome data base demonstrated the existence of at least 29 different genes from C. elegans that align closely with the catalytic domain of GCY-X1, and thus an equally large number of different regulatory ligands may exist.
Guanylyl cyclases appear to fall into two general families. (i) The membrane forms contain an extracellular ligand binding domain, a single transmembrane segment, and an intracellular protein kinase-like and cyclase catalytic domain; and (ii) the soluble forms contain two heterologous subunits, both of which contain a consensus cyclase catalytic domain (1, 2). These general features of the soluble and membrane forms appear to be conserved across many species, although expression of various invertebrate cDNAs encoding putative membrane guanylyl cyclases in mammalian cultured cells has not resulted in detectable enzyme activity (3-5).
Caenorhabditis elegans has proven a highly effective animal model for unraveling the functions of specific signaling pathways, the power resting particularly in the well defined lineage of each cell and the use of genetics. While cGMP has been measured (6) and a cGMP-dependent protein kinase has been purified from nematodes (7), C. elegans has been largely ignored as a means to define the functions of the various guanylyl cyclase/cGMP signaling pathways even though recent work on disruption of the cyclic nucleotide-gated ion channel has resulted in defective sensory neuron function (8, 9).
In preliminary studies on crude membrane fractions from C. elegans, guanylyl cyclase activity was easily measured, and PCR1 subsequently resulted in the identification of a cDNA clone encoding a putative guanylyl cyclase that was most closely related to the mammalian natriuretic peptide receptor family. The conservation of cysteine residues within the extracellular domain when compared with the mammalian natriuretic peptide receptors also suggested that the putative cyclase represented an orphan receptor. When GCY-X1 was expressed in COS-7 cells, although activity was low, it was reproducibly detected. To determine whether the nematode cyclase was capable of ligand-dependent regulation, we produced a chimeric protein consisting of the extracellular domain, transmembrane segment, and protein kinase homology domain of the mammalian guanylyl cyclase-B receptor (GC-B) and the putative catalytic domain of the C. elegans cyclase. This chimeric protein was active when transiently expressed in mammalian cells. Furthermore, the ligand (CNP) that binds to the mammalian cyclase, stimulated the activity of the nematode cyclase in a concentration-dependent manner. The fact that the cyclase catalytic domain of C. elegans can be regulated by the mammalian peptide supports the theory that GCY-X1 is an orphan receptor and suggests that a large family of apparent C. elegans guanylyl cyclases are also orphan receptors.
The Bristol N2 strain of
C. elegans was grown in liquid culture and frozen at
80 °C (10). An aliquot of frozen worms (0.25 ml) was ground with
mortar and pestle under liquid nitrogen and resuspended in 1 ml of
ice-cold homogenization buffer (10% glycerol, 50 mM HEPES,
pH 7.4, 100 mM NaCl, 1 mM EDTA, 2 mM dithiothreitol, 10 µg/ml leupeptin, 1 µg/ml
pepstatin A, 10 µg/ml aprotinin). The worm suspension was centrifuged
at 2,000 × g for 10 min at 4 °C. The pellet was
discarded, and the supernatatant fluid was centrifuged for 20 min at
100,000 × g and 4 °C to generate a membrane fraction (pellet) and a cytosolic fraction (supernatant fluid). Membranes were washed twice with homogenization buffer and resuspended by sonication.
The nematode membrane fractions
were assayed for cyclase activity after resuspension. For expression in
COS-7 cells, cells were transiently transfected with eukaryotic
expression vectors using the DEAE-dextran method (11). Cells were
harvested 48 h after transfection in the presence of
homogenization buffer and resuspended by sonication. Particulate
fractions were isolated by centrifugation for 20 min at 100,000 × g. Membranes were washed twice with homogenization buffer
and resuspended by sonication. The assay for guanylyl cyclase activity
was performed in the presence of 5 mM MnCl2 and
0.1% Triton X-100 (12). Reaction mixtures containing 25 mM
HEPES, pH 7.4, 50 mM NaCl, 0.25 mM
3-isobutyl-1-methylxanthine (IBMX), 0.1% bovine serum albumin, 10 mM NaN3, 1 mM dithiothreitol, 100 µM GTP and 0.5 µCi [-32P]GTP (3000 Ci/mmol, DuPont NEN) in a total volume of 100 µl were incubated for
10 min at 37 °C unless indicated otherwise.
[
-32P]cGMP was determined as described
previously (12).
Degenerate primers previously used to amplify guanylyl cyclase sequences from rat (13) were used to obtain guanylyl cyclase sequences from a mixed stage C. elegans cDNA library (Stratagene). The sense primer (TGTGGACIGCICCIGA(A/G)C(A/T)(A/G/C/T)(C/T)) is based on a conserved region within the protein kinase-like domain of membrane guanylyl cyclases and the antisense primer (TAIGC(A/G)TCICCIA(C/T)IGT(C/T)TCITT(A/G)TA) is based on a conserved region within the catalytic domain. The PCR reaction was performed as described (13) and yielded a single 940-base pair band. The PCR product was digested with XbaI and EcoRI and ligated into EcoRI- and XbaI-digested pGEM3Z (Promega). Thirty subclones were sequenced using the Prism kit (Applied Biosystems) and were shown to be identical. The cDNA fragment was radiolabeled and used to screen a mixed stage C. elegans cDNA library (Stratagene). Six positive phage clones were isolated, and the Bluescript plasmid containing the insert was excised according to the manufacturer instructions. The clone with the longest insert (PBS.16A, 4.8 kilobases) was sequenced completely in both directions. The putative guanylyl cyclase cDNA was designated GCY-X1 (the gene has not yet received a nomenclature within the C. elegans genome project).
Isolation of the 5For RNA isolation, 1 ml of packed frozen C. elegans were thawed on ice, homogenized in RNA STAT-60 (Tel-Test) with a Brinkman Polytron homogenizer and passed twice through a 22-gauge needle. After addition of 2 ml of chloroform, the mixture was vortexed and centrifuged at 8,000 × g for 15 min. The upper phase was precipitated with isopropyl alcohol, and the precipitate was collected by centrifugation, washed with 70% ethanol, and dissolved in diethylpyrocarbonate-treated sterile water. Polyadenylated RNA was purified using the Fast Track kit (Invitrogen). 500 ng of polyadenylated RNA were reverse transcribed using the Superscript system (Life Technologies). The first strand cDNA was subjected to PCR using a sense primer corresponding to the SL-1 (14) or SL-2 RNA sequences (15) and a GCY-X1 antisense primer (GGGGGTACCGATGCACATAGGATAATCACA). The resulting PCR product was subcloned into pGEM3Z to give pSL1.700. The PCR product was sequenced in both directions.
Construction of Eukaryotic Expression Vectors for GCY-X1To introduce a mammalian translation
initiation sequence (16) just 5 to the predicted initiator methionine,
PCR was performed on pSL1.700 with the primers
(GGGCTGCAGGCCACCATGCTCAGATGGCTAACTCTACTT) and
(CCCTCTAGAGTCGACTGTAACGAAATTTCTCGCAGCCAC). The product was digested with XbaI and PstI and ligated to
pGEM3Z. The resulting construct was digested with SalI and
ligated with the SalI/XhoI fragment of pBS.16A to
create pKZ.GC. The full-length GCY-X1 cDNA was isolated
by XbaI/PstI digestion of pKZ.GC and was ligated to XbaI/PstI-restricted pCMV5 (17) to give
pCMV5.GCY-X1.
Primers (GGGGGTACCAGCATATTGGATAATTTACTCAAAAGA) and (CCCGTCGACCTATGCAAATTCTTCGCCAAA) were used in the PCR with pBS.16A as template to amplify a GCY-X1 fragment containing an Acc65I site. The PCR product was digested with Acc65I and SalI and ligated into pBluescript (KS). The resulting construct was sequenced and digested with Acc65I and SalI. The insert was cloned into Acc65I- and SalI-digested pCMV5.GC-B to create pCMV5.GC-BX1.
Intact Cell cGMP MeasurementsCells transiently transfected with guanylyl cyclase expression vectors were washed with Dulbecco's modified Eagle's medium (DMEM) and incubated with DMEM containing 0.25 mM IBMX for 10 min at 37 °C. The medium was then switched to DMEM/0.25 mM IBMX and various concentrations of CNP (Peninsula Laboratories), and cells were incubated for 10 min at 37 °C. Perchloric acid (0.5 N) was added to stop the reaction, and the amount of cGMP formed was measured by radioimmunoassay (12).
Production of AntiseraA 15-mer peptide (GYKDVEIPDFGEEFA) corresponding to the predicted carboxyl terminus of GCY-X1 was coupled to the purified protein derivative (Serumstaateninstitut, Copenhagen, Denmark), and antiserum was produced (18). The peptide conjugate was injected into 8-pound female New Zealand White rabbits.
Western Blot AnalysisMembranes prepared from transiently transfected cells were subjected to SDS-PAGE. The proteins in the gel were electrophoretically transferred to nitrocellulose, which was incubated with primary antisera (1:1,000) and subsequently with a goat anti-rabbit antibody conjugated to horseradish peroxidase (Biosource International). The immunoreactive bands were visualized with the ECL reagent (Amersham Life Science).
Gene LocalizationA polytene blot of 958 yeast artificial chromosomes (YAC) (gift of Alan Coulson, MRC laboratory for Molecular Biology) was probed with radiolabeled fragments corresponding to GCY-X1 (Arg50-Ser195 or Ser195-Thr497). The blot was prehybridized at 65 °C in 30 mM sodium phosphate, pH 6.2, 100 mM NaCl, 0.075% (w/v) sarkosyl, 17% (w/v) dextran sulfate and washed at 60 °C in 15 mM sodium phosphate, pH 6.2, 100 mM NaCl, 0.5 mM EDTA. The filter was subjected to autoradiography to visualize the positively hybridizing YACs.
Guanylyl cyclases in
most cells exist in both a soluble and particulate fraction. To
initially characterize the localization as well as the relative
activities of guanylyl cyclases of C. elegans, worms were
homogenized and fractionated by high speed centrifugation.
Surprisingly, given the lack of any report on cyclase activity in the
nematode, cGMP was formed at a rate of 50-600 pmol/min/mg membrane
protein, dependent on the assay temperature (Fig. 1).
The activity of rat lung, the most active mammalian tissue for soluble
enzyme activity, is about 100 pmol/min/mg protein (19), whereas
invertebrate spermatozoa possess the highest activities of guanylyl
cyclase found in particulate fractions, about 10 nmol/min/mg protein
(20). The guanylyl cyclase activity in C. elegans membranes is relatively higher at lower temperatures (20-30 °C) than at higher temperatures (35-40 °C) when compared with GC-B (Fig. 1). Since C. elegans prefers soil temperatures that are nearer
20 °C, these temperature-dependent differences could be
physiologically relevant.
Isolation of a Guanylyl Cyclase cDNA
The guanylyl cyclase
activity in membrane fractions of C. elegans suggested a
potential role for cGMP signaling in the nematode. We therefore used
PCR to amplify receptor guanylyl cyclase sequences from a C. elegans cDNA library. Degenerate oligonucleotide primers directed against conserved regions of the mammalian receptor guanylyl cyclases were used for the amplification. From various PCR clones, only
a single sequence (GCY-X1) was obtained and used to screen a C. elegans cDNA library. Several cDNA clones were
isolated, but even the longest cDNA lacked the 5-end, as evidenced
by the lack of a splice leader RNA sequence and a predicted signal
peptide sequence. To isolate the 5
-end, PCR was performed on
first-strand cDNA using a sense primer corresponding to the
C. elegans trans-splice leader RNAs, SL-1 or SL-2, and an
antisense GCY-X1 primer. Only the product obtained with the
SL-1 primer was derived from GCY-X1 sequence. The first
predicted methionine occurs 30 nucleotides after the SL-1 sequence and
is in frame with the previously determined GCY-X1 sequence
(Fig. 2). Stop codons appear in the other two reading
frames preceding the first methionine. Following this Met is a
hydrophobic stretch of amino acids that would comprise a signal peptide
sequence, and the mature protein is predicted to begin at
Gly24 (21). The predicted GCY-X1 protein has an
Mr 145,000 and a single transmembrane segment
(22). There are eight consensus N-glycosylation sites within
the putative extracellular domain. GCY-X1 is predicted to
contain about 560 amino acids extracellularly, 23 amino acids within
the membrane-spanning region, and 600 amino acids intracellularly.
Between the juxtamembrane region and the Gly loop of the kinase-like
domain, a 150-amino acid stretch occurs that bears no identity to
sequences in the GenBankTM/EBI Data Bank. This region is
rich in basic amino acids (12 Lys), and contains eight consensus
protein kinase A phosphorylation sites as well as one consensus
protein-tyrosine kinase phosphorylation site (23). Both GC-A and GC-B
are known to be regulated by phosphorylation/dephosphorylation reactions (24-26), and therefore, this region of GCY-X1
may indicate an important regulatory region. The intracellular domain
is about 55% identical to GC-A, and phylogenetic analysis places the
C. elegans cyclase closest to the natriuretic peptide
receptors (GC-A and GC-B) (Fig. 3). Furthermore,
cysteine residues conserved in the extracellular domain of the
natriuretic peptide receptors are also conserved in GCY-X1
(data not shown).
Gene Localization
A polytene blot with 958 YAC clones
allowing 95% coverage of the C. elegans genome (gift of
Alan Coulson, MRC laboratory for Molecular Biology) was hybridized with
two different GCY-X1 cDNA probes. Both probes
hybridized to the same YAC clones, Y67H12 and Y39E12 (data not shown).
A YAC clone, Y62H3, that overlaps with the 5 portions of Y67H12 and
Y39E12 resulted in no signal, and thus the gene for GCY-X1
lies between unc-73 and unc-38 on chromosome I
(data not shown) (27).
Subsequently, a large number
(>29) of C. elegans genomic DNA sequences homologous to
guanylyl cyclases were found within the C. elegans genome
sequencing project. An alignment of a highly conserved region within
the catalytic domains of 17 of these putative guanylyl cyclases is
given in Fig. 4. Determination of whether or not these
sequences actually encode guanylyl cyclase activity is critical,
although there have been consistent failures to express invertebrate
putative membrane forms of guanylyl cyclases in mammalian cultured
cells (3-5).
GCY-X1 cDNA Encodes a Guanylyl Cyclase
To
formally demonstrate that GCY-X1 is a guanylyl cyclase, it
is necessary to show that the respective protein catalyzes the conversion of GTP to cGMP. Several groups attempting to express other
invertebrate drosophila (4, 5) or sea urchin membrane cyclases (3) have
failed. The cyclases most closely related to GCY-X1, GC-A
(28) or GC-B (29), however, express high enzymatic activity in various
mammalian cultured cells. When GCY-X1 was expressed in
COS-7 cells and enzyme activity of membranes estimated in the presence
of Mn2+/Triton X-100, cyclic GMP was formed at a rate of
about two times the rate of vector-transfected cells (Fig.
5B). Although substantially lower than that seen with
the mammalian cyclases, the elevation was consistently observed. The
relatively low activity could be due to low expression levels (see Fig.
5B, inset). The addition of micromolar
concentrations of ANP, BNP, or CNP to the isolated COS-7 cell membranes
did not stimulate GCY-X1 when assays were conducted in the
presence of Mg2+. Given that Koller et al. (30)
have succeeded in producing mammalian GC-A/GC-B chimeric enzymes
capable of responding to ligand, we designed experiments that would
test the hypothesis that GCY-X1 was an orphan receptor. A
GC-B/GCY-X1 chimeric expression vector was constructed. The
rat GC-B cDNA clone contains an Acc65I site preceding
the hinge region of the protein, an amphipathic helix proposed to
mediate dimerization of GC-A (31). This fortuitous restriction site
facilitated the swapping of a cDNA fragment encoding the presumed
dimerization motif and catalytic domain of GC-B with the respective
cDNA fragment of GCY-X1. The resulting chimera (GC-BX1), shown schematically in Fig. 5A,
comprises amino acids 1-798 of GC-B and amino acids 1111-1233 of
GCY-X1. GC-BX1 was expressed at levels much
higher than GCY-X1 (Fig. 5B, inset). cGMP elevations in COS-7 cells transfected with GC-B or
GC-BX1 were examined as a function of CNP concentration.
The mammalian/nematode chimeric enzyme responded to the added CNP at
concentrations equivalent to that of GC-B (Fig. 6) (32).
EC50 values for GC-B and GC-BX1 were identical
(3 × 108 M).
This is the first time an invertebrate membrane guanylyl cyclase activity has been successfully expressed in a mammalian cultured cell, confirming the hypothesis that GCY-X1 is a guanylyl cyclase. Although the ligand for GCY-X1 is not yet known, the extracellular domain most closely resembles the mammalian natriuretic peptide receptors, suggesting that if the C. elegans cyclase is a receptor, the ligand may fall into the natriuretic peptide family. Koller et al. (30) had previously demonstrated that the extracellular domains of GC-A and GC-B could be interchanged with retention of ligand sensitivity but that such chimeric constructs with GC-C (receptor for heat-stable enterotoxin and only about 10% identical at the amino acid level within the extracellular domain) were unsuccessful. Thus, it was postulated that if GCY-X1 is an orphan receptor, given it appears most closely related to the mammalian natriuretic peptide receptors, the cyclase catalytic domain of the nematode enzyme could be sensitive to regulation by a mammalian extracellular, ligand-binding domain. This was the case, and therefore by extension, it is suggested that many or all of the cyclases that are part of the large family (>29) of apparent cyclases in the nematode represent orphan receptors.
The large number of cyclases in the nematode also raises the question of whether a much larger family of cyclases exists in the mammal. Only six membrane forms of cyclase have been reported in the mammal, three of which remain orphans (2, 13), despite a mammalian genome about 30 times larger than that of C. elegans. PCR and low stringency hybridization have been used to search for other mammalian guanylyl cyclases, but the closest apparent relatives to GCY-X1 continue to be the natriuretic peptide receptors. The positive results with the chimeric receptor constructed here support the hypothesis that GCY-X1, in fact, is the C. elegans natriuretic peptide receptor homologue.
The nucleotide sequence(s) reported in this paper has been submitted to the GenBankTM/EMBL Data Bank with accession number(s) L80003[GenBank].
We are grateful to Lynda Doolittle for sequencing the cDNA and Debbie Miller for synthesizing the oligonucleotide primers. We also thank David Foster for carefully reading the manuscript.