ISOLATION OF PROGESTERONE-DEPENDENT COMPLEMENTARY DEOXYRIBONUCLEIC-ACID FRAGMENTS FROM RHESUS-MONKEY ENDOMETRIUM BY SEQUENTIAL SUBTRACTIVE HYBRIDIZATION AND POLYMERASE CHAIN-REACTION AMPLIFICATION

The steroid sex hormone progesterone (P) induces the expression of a variety of genes through a signal transduction pathway mediated by the P receptor, a DNA-binding regulator of transcription. To identify genes and gene networks that are P dependent in the rhesus endometrium, we used a powerful technique employing subtractive hybridization coupled with polymerase chain reaction (PCR). Poly(A)(+) RNA was isolated from both P-dominant (days 21-23 of artificial menstrual cycles) and estrogen (E)-dominant (days 9-13) endometrium. The two classes of RNA were converted to cDNA, ligated to EcoRI adaptors, and amplified by PCR using an adaptor-complimentary primer. E-dominant cDNA was labeled with biotin, hybridized in excess to P-dominant cDNA (PcDNA), and complexed with streptavidin. Labeled cross-hybrid sequences common to both populations were subtracted by phenol-chloroform extraction. The remaining cDNA fragments were amplified by PCR. After four rounds of hybridization/amplification, the subtracted PcDNA was analyzed for P-dependent sequences by semiquantitive PCR. Initial analysis revealed that housekeeping genes were undetectable in subtracted cDNA, but a previously characterized P-dependent gene was retained. Three of five clones sequenced at random from the subtracted library exhibited P-inducibility/dependency by PCR analysis of E and PcDNA. One of these, an 835-basepair fragment designated H5, may represent a novel P-dependent gene, as no comparable homology could be found with existing sequences in GenBank and Swissprot databases. We estimate that the procedure described here resulted in highly significant enrichment of up-regulated cDNA fragments from P-dominant tissue.