- Email: [email protected]
prolactin-secreting pituitary tumors by single-strand conformation polymorphism (SSCP) analysis
Mokc~clur und Ceilulur Endocrinology, X7 (1992) 125- 129 0 1992 Elsevier Scientific Publishers Ireland, Ltd. 0303-7207/92/$05.00
MOLCEL
125
02814
rdentification of G protein cy subunit mutations in human growth hormone (GH)- and GH/prolactin-secreting pituitary tumors by single-strand conformation polymorphism (SSCP) analysis R.T. Drews “, R.A. Gravel
(Received
KW wwd.s: Growth hormonePolymerase chain
26 March
b and R. Collu a
1992; accepted
28 May 1992)
and growth hormone/prol~ctin-s~crcting (human reaction; Single-stand conf(~rm~ltion p(~lymorpbisn~
pituitary) analysis
tumors;
G,,,
protein;
C;SP oncogene;
Summary
We have applied the polymerase chain reaction (PCR) and single-strand conformation polymorphism analysis (SSCP) to detect activating mutations in the G,, subunit gene, ampii~ing genomic DNA extracted from growth hormone (GH)- and GH/prola~tin (PRL)-secreting human pituitary tumors. Of 15 tumors tested six contained mutations in the analyzed regions of the G,,,. SSCP analysis revealed band shift in exon 8 in four GH- and in one GH/PRL-secreting tumors, and in exon 9 in one GH/PRLsecreting tumor. Direct sequencing of PCR reaction products identified the mutations as RZt,-H, R,,,,-S and RX, -C in exon 8 and QZZ7-L in exon 9. These results show the efficacy of PCR/SSCP analysis in the detection of G protein mutations and extend the generalization that these sites are hot spots in tumor-inducing mutations.
Introduction G proteins are a class of guanine nucleotidebinding (G) proteins that play a central role in the transduction of receptor-mediated signals to endogenous second messengers. In endocrine tissues, G proteins take part in a cascade of cellular events mediating control of cell protiferation and hormone secretion. Recently, somatic mutations have been reported that inactivate the GTPase site in the LYsubunit of G, rendering it constitu-
Correspondence to: Dr. Robert Collu, Research Center, Ste-Justine Hospital, 3175 C6te Ste-Catherine, Montkal, Qukbec H3T lCS, Canada.
tively active (Landis et al., 1989; Lyons et al., 1990). Two distinct codons in G,, have been found to be the site of such activating mutations in a subset of human growth hormone (GH)secreting pituitary tumors and one thyroid tumor. One of the mutation sites is at amino acid residue 201 where arginine is replaced with either cysteine (R &3 or histidine (RZf,,-H). The other mutation is at residue 227, equivalent to codon 61 in ru.s oncogene, where glutamine is replaced with either arginine (Q2,,-R) or histidine (Q,,,H). The mutated crs is unable to hydrolyze GTP with normal efficiency and remains associated with adenyIate cyclase for a Ionger period of time. This sequence of events leads to the constitutive increase of CAMP levels which in turn has
a rnitogenic effect in somatotrophs and thyrocytes (Landis et al., 1989; Lyons et al., 1990). In this study, we examined 13 GH- and two GH/prolactin IPRLI-secreting pituita~ tumors to further document the prevalance of these mutations. We used the polymerase chain reaction (PCR) to amplify exons 8 and 9 and used analysis by single-strand conformational polymorphism (SSCP) to determine if such mutations would be found in our tumor samples. Materials and methods Tissue material and nucleic acid e.~tra~tio~ Human pituitary tissues were obtained at surgery and immediately frozen at -70°C until processed. Thirteen GH- and two GH/PRLsecreting adenomas were studied. DNA was prepared from homogenized tissues under standard conditions using the cell lysis and DNA isolation protocol described by Hoar et al. (1988). PCR amplification of pituitary tissue DNA The primers for PCR reactions were chosen from sequences flanking exons 8 and 9 of the G,, subunit gene, as follows: exon 8: and exon 9: and
5’ CTACTCCAGACCTTTGCTTTAG 3’ ACAGCTGGTTATTCCAGAGGGA: 5 ’ GMTCTTGACATTCACCCCAGT 3’ AGCGACC~GATCCCTAACAAC.
DNA (0.2 mg> amplification reactions were performed in a 50 ~1 volume containing 0.2 mM each of dATP, dCTP, dGTP, dTTP, 100 pm01 of each oligonucleotide primer, and 2.5 units of Taq polymerase (AmpIitaq, Perkin-Elmer Cetus, Norwalk, CT, USA), in 10 mM Tris-HCl (pH 8.31, 50 mM KCI, 1.5 mM MgCl, and 0.01% gelatin. Thirty cycles of amplification were used, consisting of 30 s of denaturation at 94°C 30 s of annealing at 6o”C, and 90 s of extension at 72°C in a Perkin-Elmer Cetus DNA Thermal Cycler. For amplification of DNA for SSCP analysis, the concentration of dATP in the PCR reaction tubes was reduced to 0.1 mM and 10 PCi of [(u“P]dATP (New England Nuclear, 3000 Ci/mmol) was added as radioactive tracer.
SSCP analysis of amplified material To prepare radioactive PCR samples for SSCP DNA analysis, a 2 pi of a lo- ’ dilution (in water) of the labetled PCR product was mixed with an equal volume of 95% formamide. 10 mM EDTA, 0.05% bromophenol blue, 0.05% xylene-cyanol FF. The sample was denatured by heating at 95°C for 5 min and immediately loaded on a 0.4 mm x 31.0 cm x 38.5 cm 6% polyac~lamide/lO~~ glycerol gel (Orita et al., 1989). Elcctrophoresis was done at room temperature overnight (16-20 h) at 15 mA (constant current). A non-denatured sample was always run beside a denatured sample. The gel was dried and autoradiographed on X-ray film (Kodak, Rochester, NY, USA). Sequencing SSCP positive and control samples were sequenced in two directions directly from PCR amplified material (Winship, 1989). Brieff y, 40-80 ng of double-stranded PCR products were purified on 5% non-denaturing polyacrylamide gel, and dried in the presence of one of the primers (+ 140 ng) used for the PCR reaction. After resuspension of the samples in the annealing buffer, sequencing was performed using [N‘“S]dATP and the dideoxy chain termination reaction (Sequenase 2.0, U.S. Biochemicals, Clevcland, OH, USA). Samples from the sequencing reactions were electrophoresed on a 6% polyacrylamide, 7 M urea gel. After removing the urea, the gel was dried and auto~diographed on X-ray film (Kodak, Rochester, NY, USA). Results The DNA of pituitary tumors (n = 15) was amplified by PCR using primers flanking exons 8 and 9 of the G,, subunit gene. The length of the amplification products was verified by agarose gel electrophoresis and corresponded to the predicted values (225 bp for exon 8 and 187 bp for exon 9; data not shown). Radioactive, amplified DNA samples were denatured and subjected to SSCP analysis in non-denaturing polyacryiamide gels containing 10% glycerol. Comparison of the mobility of the variously folded single-strand conformers of tumor and normal DNA revealed differences in five samples from exon 8 and in one
127
sample from exon 9. In Figs. 1 and 2, the arrowheads indicate the broadening of bands (Fig. 1, lane 5) or the generation of new bands (Fig. 1, lane 3; Fig. 2, lane 3) due to the altered migration of single strands bearing the mutations. PCR products from exons 8 and 9 from normal DNA and those showing differences from normal were sequenced directly using one or the other of the primers used for DNA amplification. In four GH-secreting tumors and in one GH/PRLsecreting tumor we found the simultaneous presence at codon 201 in exon 8 of both normal CGT (arginine) and mutant alleles. The latter were
Fig. 2. Identification of a mutation in exon 9 of the G,, gene by SSCP analysis. The autoradiogram of DNA samples PCR amplified and resolved on 6% polyacrylamide/lO% glycerol gel is shown. Lanes 1 and 2, control, lanes 2 and 3, GH/PRL-secreting and lanes 5 and 6, GH-secreting pituitary tumors. Lanes 2, 4 and 6 show the non-denatured (doublestranded) form of the product. The arrowheads mark those bands which exhibit differences in migration due to the presence of a mutation.
Fig. I. Identification of the mutations in exon 8 of the G,, gene by SSCP analysis. The autoradiogram of DNA samples PCR amplified and resolved on 6% polyactylamide/lO% glycerol gel is shown. Lanes 1 and 2, control, lanes 3 and 4, GH-secreting and lanes 5 and 6, GH/PRL-secreting pituitary tumors. Lanes 2, 4 and 6 show the non-denatured (doublestranded) form of the product. The arrowheads mark those bands which exhibit differences in migration due to the presence of a mutation.
CAT (histidine) in a GH/PRL tumor, TGT (cysteine) in three GH tumors and AGT (serine) in one GH tumor (Fig. 3). As shown on Fig. 4, the second GH/PRL-secreting tumor also contained a normal and a mutant allele in exon 9 where, at codon 227, were simultaneously present the normal CAG (glutamine) and a mutant CTG (leucine) allele. Discussion A survey of 13 GH- and two GH/PRL-secreting pituitary adenomas revealed the presence
of
12X
six mutations in the (Y subunit of the G, protein. Five of them were found at exon 8, codon 201 (ADP ribosylation site), where arginine was replaced with either histidine, cysteine or serine (Rzcl,-H, Rzo,-C, R,,,,-S). One mutation was found
G
A
T
G
A
T
C
C
Fig. 3. Direct sequencing of PCR-amplified region of codon 227. U: Genomic, control DNA shows the presence of wild-type codon CAG (Gln). h: Genomic DNA from SSCP positive tumor shows the simultaneous presence of wild-type codon CAG (Gin) and mutated codon CTG (Leu). An arrow denotes the presence of the additional nucleotide.
Fig. 3. Direct sequencing of PCR-amplified region of codon 201, (I: Genomic, control DNA shows the presence of wild-type DNA from SSCP codon CGT (ARg). h, c. d: Genomic positive tumors shows the simultaneous presence of wild-type codon CGT (Arg) and mutated codons CAT (His), TGT (Cys) or AGT (Ser). The arrows denote the presence of the additional nucleotides.
in exon 9 where, in codon 227 (equivalent to codon 61 in the ras oncogene), glutamine was replaced with leucine (Q,,,-L). Mutations in these codons have been reported by Lyons et al. (1990) and our results reinforce their conclusion that these two sites are ‘hot spots’ for tumor-inducing mutations. The substitution of glutamine with leucine at position 227 is different from those reported by Lyons et al. (1990) who found glutamine to be replaced by arginine or histidine, but corresponds to the substitution originally found in the equivalent codon 61 of activated p21’“” oncogene. As already established by Zachry et al. (1990), expression of the G,, containing the Q12,-L mutation in Swiss 3T3 fibroblasts markedly decreases the GTPase activity and constitutively increases the level of CAMP. The elevated level of CAMP associated with the Qz2,-L substitution is similar to the outcome found originally for GH pituitary tumors containing G,, mutations Qz2,-R or Qm- H (Landis et al., 1989).
129
Five of the six mutations detected by us were in codon 201. This prevalence fits with that found by Lyons et al. (1990) who reported that 16 of 18 mutations in G,, in their GH tumors were located in codon 201. Furthermore, as also supported by our results, they demonstrated the bias toward arginine to cysteine substitutions (14 R to C) at this site. The arginine to serine substitution presently identified had not previously been reported in GH tumors but, recently, Suarez et al. (1991) documented this mutation in a thyroid papillary tumor. Of special note is our finding of G,, subunit mutations in both GH/PRL pituitary adenomas. These mutations previously not described in GH/PRL-secreting tumors are associated in GH tumors with constitutively increased cellular levels of CAMP (Lyons et al., 1990). Although not enough tissue was available to assay CAMP levels, it is therefore possible that they were also constitutively increased in these two GH/PRL-secreting adenomas. We have applied an approach combining analysis of PCR amplified regions of the G,, gene by SSCP with direct sequencing of candidate PCR products to the identification of point mutations. Our results confirm the efficacy of this analysis for the detection of single nucleotide substitutions, although we note, as in Fig. 1, that the user must be alert to subtle differences in band patterns that can be generated due to mutation. Application of PCR-SSCP analysis should pro-
vide an effective way to identify unknown genetic lesions in the functional domains of the (Y subunits genes and also of other proteins involved in the process of cellular signal transduction and oncogenesis. Acknowledgments Supported by a grant from the Canadian Network of Centers of Excellence to R.A.G., by a grant from the Medical Research Council of Canada to R.C. and by a postdoctoral fellowship from the Interservice Club Council (Telethon of Stars) to R.T.D. We thank Beverly Akerman for introduction to SSCP analysis. References Hoar, D.I., Haslam, D.B. and Starozik, D.M. (1988) Prenatal. Diagn. 4, 241-247. Landis, CA., Masters, S.B., Spada, A., Pace, A.M., Bourne, H.R. and Vallar, L. (1989) Nature 340, 692-696. Lyons, J., Landis, CA., Harsh, H., Vallar, L., Grunewald, K., Feichtinger, H., Duh, Q.Y., Clark, O.H., Kawasaki, E., Bourne, H.R. and McCormick, F. (1990) Science 249, 655-659. Orita, M., Suzuki, Y., Sekiya, T. and Hayashi, K. (1989) Genomics 5, 874-879. Suarez, H.G., du Villard, J.A., Caillou, B., Schlumberger, M., Parmentier, C. and Monier, R. (1991) Oncogene 6, 6777 680. Winship, P.R. (1989) Nucleic Acid Res. 17, 1266. Zachry, I., Masters, S.B. and Bourne, H.R. (1990) Biochem. Biophys. Res. Commun. 168, 1184-l 193.
MOLCEL
125
02814
rdentification of G protein cy subunit mutations in human growth hormone (GH)- and GH/prolactin-secreting pituitary tumors by single-strand conformation polymorphism (SSCP) analysis R.T. Drews “, R.A. Gravel
(Received
KW wwd.s: Growth hormonePolymerase chain
26 March
b and R. Collu a
1992; accepted
28 May 1992)
and growth hormone/prol~ctin-s~crcting (human reaction; Single-stand conf(~rm~ltion p(~lymorpbisn~
pituitary) analysis
tumors;
G,,,
protein;
C;SP oncogene;
Summary
We have applied the polymerase chain reaction (PCR) and single-strand conformation polymorphism analysis (SSCP) to detect activating mutations in the G,, subunit gene, ampii~ing genomic DNA extracted from growth hormone (GH)- and GH/prola~tin (PRL)-secreting human pituitary tumors. Of 15 tumors tested six contained mutations in the analyzed regions of the G,,,. SSCP analysis revealed band shift in exon 8 in four GH- and in one GH/PRL-secreting tumors, and in exon 9 in one GH/PRLsecreting tumor. Direct sequencing of PCR reaction products identified the mutations as RZt,-H, R,,,,-S and RX, -C in exon 8 and QZZ7-L in exon 9. These results show the efficacy of PCR/SSCP analysis in the detection of G protein mutations and extend the generalization that these sites are hot spots in tumor-inducing mutations.
Introduction G proteins are a class of guanine nucleotidebinding (G) proteins that play a central role in the transduction of receptor-mediated signals to endogenous second messengers. In endocrine tissues, G proteins take part in a cascade of cellular events mediating control of cell protiferation and hormone secretion. Recently, somatic mutations have been reported that inactivate the GTPase site in the LYsubunit of G, rendering it constitu-
Correspondence to: Dr. Robert Collu, Research Center, Ste-Justine Hospital, 3175 C6te Ste-Catherine, Montkal, Qukbec H3T lCS, Canada.
tively active (Landis et al., 1989; Lyons et al., 1990). Two distinct codons in G,, have been found to be the site of such activating mutations in a subset of human growth hormone (GH)secreting pituitary tumors and one thyroid tumor. One of the mutation sites is at amino acid residue 201 where arginine is replaced with either cysteine (R &3 or histidine (RZf,,-H). The other mutation is at residue 227, equivalent to codon 61 in ru.s oncogene, where glutamine is replaced with either arginine (Q2,,-R) or histidine (Q,,,H). The mutated crs is unable to hydrolyze GTP with normal efficiency and remains associated with adenyIate cyclase for a Ionger period of time. This sequence of events leads to the constitutive increase of CAMP levels which in turn has
a rnitogenic effect in somatotrophs and thyrocytes (Landis et al., 1989; Lyons et al., 1990). In this study, we examined 13 GH- and two GH/prolactin IPRLI-secreting pituita~ tumors to further document the prevalance of these mutations. We used the polymerase chain reaction (PCR) to amplify exons 8 and 9 and used analysis by single-strand conformational polymorphism (SSCP) to determine if such mutations would be found in our tumor samples. Materials and methods Tissue material and nucleic acid e.~tra~tio~ Human pituitary tissues were obtained at surgery and immediately frozen at -70°C until processed. Thirteen GH- and two GH/PRLsecreting adenomas were studied. DNA was prepared from homogenized tissues under standard conditions using the cell lysis and DNA isolation protocol described by Hoar et al. (1988). PCR amplification of pituitary tissue DNA The primers for PCR reactions were chosen from sequences flanking exons 8 and 9 of the G,, subunit gene, as follows: exon 8: and exon 9: and
5’ CTACTCCAGACCTTTGCTTTAG 3’ ACAGCTGGTTATTCCAGAGGGA: 5 ’ GMTCTTGACATTCACCCCAGT 3’ AGCGACC~GATCCCTAACAAC.
DNA (0.2 mg> amplification reactions were performed in a 50 ~1 volume containing 0.2 mM each of dATP, dCTP, dGTP, dTTP, 100 pm01 of each oligonucleotide primer, and 2.5 units of Taq polymerase (AmpIitaq, Perkin-Elmer Cetus, Norwalk, CT, USA), in 10 mM Tris-HCl (pH 8.31, 50 mM KCI, 1.5 mM MgCl, and 0.01% gelatin. Thirty cycles of amplification were used, consisting of 30 s of denaturation at 94°C 30 s of annealing at 6o”C, and 90 s of extension at 72°C in a Perkin-Elmer Cetus DNA Thermal Cycler. For amplification of DNA for SSCP analysis, the concentration of dATP in the PCR reaction tubes was reduced to 0.1 mM and 10 PCi of [(u“P]dATP (New England Nuclear, 3000 Ci/mmol) was added as radioactive tracer.
SSCP analysis of amplified material To prepare radioactive PCR samples for SSCP DNA analysis, a 2 pi of a lo- ’ dilution (in water) of the labetled PCR product was mixed with an equal volume of 95% formamide. 10 mM EDTA, 0.05% bromophenol blue, 0.05% xylene-cyanol FF. The sample was denatured by heating at 95°C for 5 min and immediately loaded on a 0.4 mm x 31.0 cm x 38.5 cm 6% polyac~lamide/lO~~ glycerol gel (Orita et al., 1989). Elcctrophoresis was done at room temperature overnight (16-20 h) at 15 mA (constant current). A non-denatured sample was always run beside a denatured sample. The gel was dried and autoradiographed on X-ray film (Kodak, Rochester, NY, USA). Sequencing SSCP positive and control samples were sequenced in two directions directly from PCR amplified material (Winship, 1989). Brieff y, 40-80 ng of double-stranded PCR products were purified on 5% non-denaturing polyacrylamide gel, and dried in the presence of one of the primers (+ 140 ng) used for the PCR reaction. After resuspension of the samples in the annealing buffer, sequencing was performed using [N‘“S]dATP and the dideoxy chain termination reaction (Sequenase 2.0, U.S. Biochemicals, Clevcland, OH, USA). Samples from the sequencing reactions were electrophoresed on a 6% polyacrylamide, 7 M urea gel. After removing the urea, the gel was dried and auto~diographed on X-ray film (Kodak, Rochester, NY, USA). Results The DNA of pituitary tumors (n = 15) was amplified by PCR using primers flanking exons 8 and 9 of the G,, subunit gene. The length of the amplification products was verified by agarose gel electrophoresis and corresponded to the predicted values (225 bp for exon 8 and 187 bp for exon 9; data not shown). Radioactive, amplified DNA samples were denatured and subjected to SSCP analysis in non-denaturing polyacryiamide gels containing 10% glycerol. Comparison of the mobility of the variously folded single-strand conformers of tumor and normal DNA revealed differences in five samples from exon 8 and in one
127
sample from exon 9. In Figs. 1 and 2, the arrowheads indicate the broadening of bands (Fig. 1, lane 5) or the generation of new bands (Fig. 1, lane 3; Fig. 2, lane 3) due to the altered migration of single strands bearing the mutations. PCR products from exons 8 and 9 from normal DNA and those showing differences from normal were sequenced directly using one or the other of the primers used for DNA amplification. In four GH-secreting tumors and in one GH/PRLsecreting tumor we found the simultaneous presence at codon 201 in exon 8 of both normal CGT (arginine) and mutant alleles. The latter were
Fig. 2. Identification of a mutation in exon 9 of the G,, gene by SSCP analysis. The autoradiogram of DNA samples PCR amplified and resolved on 6% polyacrylamide/lO% glycerol gel is shown. Lanes 1 and 2, control, lanes 2 and 3, GH/PRL-secreting and lanes 5 and 6, GH-secreting pituitary tumors. Lanes 2, 4 and 6 show the non-denatured (doublestranded) form of the product. The arrowheads mark those bands which exhibit differences in migration due to the presence of a mutation.
Fig. I. Identification of the mutations in exon 8 of the G,, gene by SSCP analysis. The autoradiogram of DNA samples PCR amplified and resolved on 6% polyactylamide/lO% glycerol gel is shown. Lanes 1 and 2, control, lanes 3 and 4, GH-secreting and lanes 5 and 6, GH/PRL-secreting pituitary tumors. Lanes 2, 4 and 6 show the non-denatured (doublestranded) form of the product. The arrowheads mark those bands which exhibit differences in migration due to the presence of a mutation.
CAT (histidine) in a GH/PRL tumor, TGT (cysteine) in three GH tumors and AGT (serine) in one GH tumor (Fig. 3). As shown on Fig. 4, the second GH/PRL-secreting tumor also contained a normal and a mutant allele in exon 9 where, at codon 227, were simultaneously present the normal CAG (glutamine) and a mutant CTG (leucine) allele. Discussion A survey of 13 GH- and two GH/PRL-secreting pituitary adenomas revealed the presence
of
12X
six mutations in the (Y subunit of the G, protein. Five of them were found at exon 8, codon 201 (ADP ribosylation site), where arginine was replaced with either histidine, cysteine or serine (Rzcl,-H, Rzo,-C, R,,,,-S). One mutation was found
G
A
T
G
A
T
C
C
Fig. 3. Direct sequencing of PCR-amplified region of codon 227. U: Genomic, control DNA shows the presence of wild-type codon CAG (Gln). h: Genomic DNA from SSCP positive tumor shows the simultaneous presence of wild-type codon CAG (Gin) and mutated codon CTG (Leu). An arrow denotes the presence of the additional nucleotide.
Fig. 3. Direct sequencing of PCR-amplified region of codon 201, (I: Genomic, control DNA shows the presence of wild-type DNA from SSCP codon CGT (ARg). h, c. d: Genomic positive tumors shows the simultaneous presence of wild-type codon CGT (Arg) and mutated codons CAT (His), TGT (Cys) or AGT (Ser). The arrows denote the presence of the additional nucleotides.
in exon 9 where, in codon 227 (equivalent to codon 61 in the ras oncogene), glutamine was replaced with leucine (Q,,,-L). Mutations in these codons have been reported by Lyons et al. (1990) and our results reinforce their conclusion that these two sites are ‘hot spots’ for tumor-inducing mutations. The substitution of glutamine with leucine at position 227 is different from those reported by Lyons et al. (1990) who found glutamine to be replaced by arginine or histidine, but corresponds to the substitution originally found in the equivalent codon 61 of activated p21’“” oncogene. As already established by Zachry et al. (1990), expression of the G,, containing the Q12,-L mutation in Swiss 3T3 fibroblasts markedly decreases the GTPase activity and constitutively increases the level of CAMP. The elevated level of CAMP associated with the Qz2,-L substitution is similar to the outcome found originally for GH pituitary tumors containing G,, mutations Qz2,-R or Qm- H (Landis et al., 1989).
129
Five of the six mutations detected by us were in codon 201. This prevalence fits with that found by Lyons et al. (1990) who reported that 16 of 18 mutations in G,, in their GH tumors were located in codon 201. Furthermore, as also supported by our results, they demonstrated the bias toward arginine to cysteine substitutions (14 R to C) at this site. The arginine to serine substitution presently identified had not previously been reported in GH tumors but, recently, Suarez et al. (1991) documented this mutation in a thyroid papillary tumor. Of special note is our finding of G,, subunit mutations in both GH/PRL pituitary adenomas. These mutations previously not described in GH/PRL-secreting tumors are associated in GH tumors with constitutively increased cellular levels of CAMP (Lyons et al., 1990). Although not enough tissue was available to assay CAMP levels, it is therefore possible that they were also constitutively increased in these two GH/PRL-secreting adenomas. We have applied an approach combining analysis of PCR amplified regions of the G,, gene by SSCP with direct sequencing of candidate PCR products to the identification of point mutations. Our results confirm the efficacy of this analysis for the detection of single nucleotide substitutions, although we note, as in Fig. 1, that the user must be alert to subtle differences in band patterns that can be generated due to mutation. Application of PCR-SSCP analysis should pro-
vide an effective way to identify unknown genetic lesions in the functional domains of the (Y subunits genes and also of other proteins involved in the process of cellular signal transduction and oncogenesis. Acknowledgments Supported by a grant from the Canadian Network of Centers of Excellence to R.A.G., by a grant from the Medical Research Council of Canada to R.C. and by a postdoctoral fellowship from the Interservice Club Council (Telethon of Stars) to R.T.D. We thank Beverly Akerman for introduction to SSCP analysis. References Hoar, D.I., Haslam, D.B. and Starozik, D.M. (1988) Prenatal. Diagn. 4, 241-247. Landis, CA., Masters, S.B., Spada, A., Pace, A.M., Bourne, H.R. and Vallar, L. (1989) Nature 340, 692-696. Lyons, J., Landis, CA., Harsh, H., Vallar, L., Grunewald, K., Feichtinger, H., Duh, Q.Y., Clark, O.H., Kawasaki, E., Bourne, H.R. and McCormick, F. (1990) Science 249, 655-659. Orita, M., Suzuki, Y., Sekiya, T. and Hayashi, K. (1989) Genomics 5, 874-879. Suarez, H.G., du Villard, J.A., Caillou, B., Schlumberger, M., Parmentier, C. and Monier, R. (1991) Oncogene 6, 6777 680. Winship, P.R. (1989) Nucleic Acid Res. 17, 1266. Zachry, I., Masters, S.B. and Bourne, H.R. (1990) Biochem. Biophys. Res. Commun. 168, 1184-l 193.