The human folate receptors (hFRs) are important in the cellular accumulation of folates and antifolates. We described the structure of the human KB cell FR (hFR-KB) gene and identified two discrete promoter regions (P1 and P4) upstream from exons 1 and 4, respectively (Elwood et al., 1993). To further understand the molecular basis of hFR expression, we have now analyzed the basal transcription of the P4 promoter localized upstream of a major transcription start site. The sequence upstream from exon 4 contains several potential transcriptional factor-binding sites and a consensus initiator region sequence at the transcription start site but does not contain canonical TATA or CAAT boxes. While deletion of a 5' flanking sequence from nt -1023 to nt -605 of P4 promoter region decreases the luciferase reporter gene expression in KB cells to 54-70% of control construct, the removal of the sequence between nt -292 and nt -46 markedly decreases the activity to 3%. DNase I footprints and competitive mobility shift and supershift mobility assays indicate that Sp1 or Sp1-related nuclear protein(s) bind to three clustered GC-rich regions within the sequence between nt -292 and nt -46 of the hFR-KB P4 promoter. Both in vitro and in vivo analyses of the expression of promoter constructs containing site-specific mutation(s) of these three Sp1-binding sites and initiator sequence demonstrate that each of three Sp1 sites and the initiator sequence are required for optimum promoter activity and that they interact cooperatively in this P4 promoter of the hFR-KB gene.

We report the use of a fusion to the green fluorescent protein to visualize the assembly of the morphogenetic protein SpoIVA around the developing forespore during the process of sporulation in the bacteriumBacillus subtilis. Using a deconvolution algorithm to process digitally-collected optical sections, we show that SpoIVA, which is synthesized in the mother cell chamber of the sporangium, assembled into a spherical shell around the outer surface of the forespore. Time-lapse fluorescence microscopy showed that this assembly process commenced at the time of polar division and seemed to continue after engulfment of the forespore was complete. SpoIVA remained present throughout the late stages of morphogenesis and was present as a component of the fully mature spore. Evidence indicates that assembly of SpoIVA depended on the extreme C-terminal region of the protein and an additional region that directly or indirectly facilitated interaction among SpoIVA molecules. The N- and C-terminal regions of SpoIVA, including the extreme C terminus, are highly similar to the corresponding regions of the homologous protein from the distantly related endospore-forming bacterium Clostridium acetobutylicum, attesting to their importance in the function of the protein. Finally, we show that proper localization of SpoIVA required the expression of one or more genes which, likespoIVA, are under the control of the mother cell transcription factor ςE. One such gene wasspoVM, whose product was required for efficient targeting of SpoIVA to the outer surface of the forespore.

Bacillus subtilis gene ypfP, which is located at 196°on the genetic map, shows similarity to both the monogalactosyldiacylglycerol synthase gene of Cucumis sativus, which encodes a galactosyltransferase, and the murG genes of B. subtilis, Escherichia coli, Haemophilus influenzae, and Synechocystis sp. strain PCC6803, which encode N-acetylglucosaminyltransferases involved in peptidoglycan biosynthesis. Cells containing a null mutation of ypfP are shorter and rounder than wild-type cells during growth in Luria-Bertani medium and glucose minimal medium. In addition, the mutant cells preferentially undergo lysis when grown on solid Luria-Bertani medium.The chloroplast membranes of higher plants and eukaryotic algae contain two major galactolipids, monogalactodiacylglycerol (MGDG) and digalactosyldiacylglycerol. The final step in the synthesis of MGDG is catalyzed by MGDG synthase, a UDPgalactose:1,2-diacylglycerol 3-␤-D-galactosyltransferase, which transfers galactose from UDPgalactose to 1,2-diacylglycerol (9). A sequence of an MGDG synthase from cucumber (Cucumis sativus) is known and was shown to be similar to those of the MurG peptidoglycan biosynthesis enzymes of Escherichia coli and Bacillus subtilis (17). The rigid peptidoglycan layer of the bacterial cell wall is composed of long glycan chains that are cross-linked through peptide side chains. The glycan chains are alternating copolymers of N-acetylglucosamine (GlcNAc) and N-acetylmuramic acid (MurNAc) that are held together by ␤134 glycosidic linkages. The chains are formed by polymerization of disaccharide units that are assembled on a lipid carrier molecule at the inner surface of the cytoplasmic membrane (11,12,18). Synthesis of the disaccharide is catalyzed by MurG, a UDP-N-acetylglucosamine:Nacetylmuramyl-(pentapeptide) pyrophosphoryl-undecaprenol N-acetylglucosamine transferase, which transfers GlcNAc from the nucleotide sugar UDP-GlcNAc to undecaprenyl-pyrophosphoryl-MurNAc-pentapeptide to form undecaprenylpyrophosphoryl-MurNAc-(pentapeptide)GlcNAc (2, 14). Here we report that B. subtilis contains a gene, ypfP, that influences cell shape and viability, and whose inferred product resembles MGDG synthase of C. sativus and shows similarity to some conserved regions of MurG proteins.ypfP, which is 382 codons in length, was previously identified by the B. subtilis sequencing project and lies between metB and cspD at 196°on the chromosome (3). Our searches against available sequence databases have revealed that the inferred amino acid sequence of YpfP is most similar to that of MGDG synthase of C. sativus. An alignment of these proteins shows they exhibit 24% identity overall and 37% identity in four regions involving 183 YpfP residues (Fig. 1). In confirmation and extension of the results obtained by Shimojima et al. (17), we found that YpfP, like MGDG synthase, contains regions of great similarity to some conserved regions of the MurG proteins of B. subtilis (1,7,15), Haemophilus influenzae (5), E. coli (8,13,19), and Synechocystis sp. strain PCC6803 (10...

The primary structures of the human KB cell (FR-KB1) folate receptor (FR) and of a human placental (FR-P2) FR, proteins important in cellular accumulation of folates, have been deduced from cDNA sequences. Herein, we report a novel human FR cDNA (FR-P3) isolated from a placental library and the chromosomal organization of the human FR-P3 gene. Compared to the FR-P2 cDNA, the composite 1084 base-pair (bp) FR-P3 cDNA is homologous, but contains a unique 5' terminus and sequence differences within the open reading frame (ORF) and at the exon I-II junction. Polymerase chain reaction and RNase protection assays demonstrate that the FR-P3 cDNA represents the major transcript, and suggest that the FR-P2 cDNA is encoded by an independent FR gene. The nucleotide sequences of two non-overlapping human genomic clones contain the FR-P3 gene, which spans 5148 bp, is composed of five exons, and is polymorphic relative to 5' restriction sites. The transcript size (1084 bp) predicted from structural analysis of the FR-P3 gene correlates with the size (1100 bp) determined by Northern blots. Based on RNase protection assays, both FR-P3 and FR-KB1 transcripts are expressed in human fetal and adult tissues, and the abundance of each transcript varies among the tissues studied. These results indicate that the FR transcripts are products of independent, conserved genes; that neither FR gene is preferentially expressed during fetal development; and that specific FR transcripts are differentially expressed in human tissues, suggesting that transcription of each FR gene is regulated independently. The isolation of the FR-P3 gene will permit functional analysis of the cis and trans regulatory elements of the FR-P3 gene and the mechanisms involved in tissue-specific FR gene expression.