Search

The general human RNA polymerase III transcription factor (TF) IIIC1 has hitherto been ill defined with respect to the polypeptides required for reconstitution of its activity. Here we identify Homo sapiens TFIIIB" (HsBdp1) as an essential component of hTFIIIC1 and hTFIIIC1-like activities. Several forms of HsBdp1 are described. The 250-kDa form of HsBdp1, also designated the "transcription factor-like nuclear regulator," strictly co-eluted with TFIIIC1 activity over multiple chromatographic purification steps as revealed by Western blot with anti-HsBdp1 antibodies and by MALDI-TOF analysis. In addition, TFIIIC1 activity could be depleted from partially purified fractions with anti-HsBdp1 antibodies but not with control antibodies. Moreover, highly purified recombinant HsBdp1 could replace TFIIIC1 activity in reconstituted transcription of the VAI gene in vitro. Furthermore, smaller proteins of approximately 90-150 kDa that were recognized by anti-HsBdp1 antibodies co-eluted with TFIIIC1-like activity. Finally, cytoplasmic extracts from differentiated mouse F9 fibroblast cells that lacked TFIIIC1 activity could be made competent for transcription of the VA1 gene by the addition of TFIIIC1, TFIIIC1-like, or recombinant HsBdp1. These results suggest that HsBdp1 proteins represent essential components of TFIIIC1 and TFIIIC1-like activities.

By employing purified transcription factors and RNA polymerase III (pol III), we generated active pol III transcription complexes on the human 5S rRNA gene. These large complexes were separated by size exclusion chromatography from non- incorporated proteins. In addition, we succeeded in isolating specific intermediate stages of complex formation. Such isolated partial complexes require complementation with the missing activities for full transcription activity. One central finding is that a 5S DNA-TFIIIA-TFIIIC2-TFIIIBbeta complex could be isolated which had been assembled in the absence of the general pol III transcription factor IIIC1. Thus TFIIIC1 is not an assembly factor for other transcription factors. Although pol III has the potential to bind unspecifically to DNA, such polymerase molecules cannot be rendered initiation competent by direct recruitment to a 5S DNA-TFIIIA-TFIIIC2- TFIIIBbeta complex, but this process strictly requires additional TFIIIC1 activity. This clearly demonstrates that in contrast to yeast cells, hTFIIIB(beta), although required, does not suffice for the functional recruitment of polymerase III. These data document that TFIIIC1 is the second transcription factor required for the recruitment of pol III in mammalian cells.

We attempted to devise a transcription system in which a particular DNA sequence of interest could be inducibly expressed under the control of a modified polymerase III (pol III) promoter. Its activation requires a mutated transcription factor not contained endogenously in human cells. We constructed such a promoter by fusing elements of the β-lactamase gene of Escherichia coli, containing a modified TATA-box and a pol III terminator, to the initiation region of the human U6 gene. This construct functionally resembles a 5′-regulated pol III gene and its transcribed segment can be exchanged for an arbitrary sequence. Its transcription in vitro by pol III requires the same factors as the U6 gene with the major exception that the modified TATA-box of this construct only interacts with a TATA-binding protein (TBP) mutant (TBP-DR2) but not with TBP wild-type (TBPwt). Its transcription therefore requires TBP-DR2 exclusively instead of TBPwt. In order to render the system inducible, we fused the gene coding for TBP-DR2 to a tetracycline control element and stably transfected this new construct into HeLa cells. Induction of such a stable and viable clone with tetracycline resulted in the expression of functional TBP-DR2. This system may conceptually be used in the future to inducibly express an arbitrary DNA sequence in vivo under the control of the above mentioned promoter.

It has been shown that under specific conditions, transcription of protein coding genes can be efficiently initiated by RNA polymerase (pol) III in vitro. We examined the formation and composition of such pol III transcription complexes on the duck histone H5 and alphaA-globin promoters and found that the essential step for the formation of pol III transcription complexes on these pol II promoters was the stable binding of transcription factor (TF) IIIB-beta. For this process, the intact TFIIIB-beta complex, consisting of TBP and associated factors (TAFs) was needed and the prior association of pol III assembly factors was not necessary. We demonstrate for the first time that hTFIIIB-beta alone is able to bind to pol II promoter DNA. This resulted in a very stable complex which was resistant to high concentrations of heparin. Although immunodepletion revealed that TBP is essentially required for complex formation, other components of hTFIIIB-beta must also be involved, since TBP itself is unable to form heparin-resistant complexes and does not mediate pol III commitment per se. pol III is recruited to these pol II promoters in a strictly TFIIIC1 dependent way. After binding of TFIIIB-beta, the addition of TFIIIC1 and pol III were sufficient to yield productive pol III transcription complexes, which utilized the correct pol II initiation site. From these findings, we postulate that TFIIIC1 is involved in the recruitment of pol III and may thus form a bridge between TFIIIB-beta and the enzyme. This finding provides the first evidence for functional contacts between TFIIIC1 and pol III, which could be of general importance for the assembly of pol III transcription complexes.

Human TFIIIC1, a basal transcription factor essentially required for expression of all pol III genes, exerts its function without primarily binding to DNA. We report here the purification of a termination site binding activity (TBA) which was initially described to be contained in fractions designated as TFIIIC0. TBA specifically and strongly binds to the termination region of pol III genes with internal promoters and can be completely separated from TFIIIC1and a TFIIIC1related activity (TFIIIC1-like), proving that DNA-binding of TBA is independent of these latter activites. Although TBA is not essentially required for, it strongly stimulates pol III transcription from intragenic promoters. This stimulation strictly depends on the presence of TFIIIC1and is not observed in conjunction with TFIIIC1-like. We further present the identification of a novel activity, TFIIIU, which is also contained in crude fractions of TFIIIC0. TFIIIU can be separated from TBA by further purification and is essentially involved in transcription of the mammalian U6 gene. TFIIIU cannot be substituted for by any of the established U6 transcription factors and thus represents a novel U6 transcription factor.

Human transcription factor IIIC2 interacts with the TFIIIA-5S DNA complex and forms a ternary TFIIIA/IIIC2-5S DNA complex. Formation of this complex does not preclude simultaneous binding of TFIIIC2to the B-box sequence of a second template. This suggests that the domain(s) or subunit(s) required for indirect recognition of the 5S promoter by TFIIIC2 are different from those necessary for direct binding of TFIIIC2 to B-box-containing pol III promoters. Whereas TFIIIC2 is only required for transcription of the 'classical' pol III genes, TFIIIC1 is generally required for transcription of all pol III genes, including that of the U6 gene. The activity of TFIIIC1 strongly enhances specific binding of basal pol III factors TFIIIA, TFIIIC2 and the PSE binding protein (PBP) to their cognate promoter elements and it acts independently of the corresponding termination regions. Moreover, we characterize an activity, TFIIIC0, purified from phosphocellulose fraction C, which shows strong DNase I protection of the termination region of several pol III genes and which is functionally and chromatographically distinct from TFIIIC1 and TFIIIC2.

The intron-containing gene encoding human ribosomal protein S3a (hRPS3a) was isolated by utilizing a PCR-based strategy to detect a gene-specific intron which was subsequently used as a probe for cloning of the entire gene. The hRPS3a gene is composed of six exons and five introns spanning 5013 bp. As described for other hRP-encoding genes, the promoter lacks a canonical TATA sequence and a defined CAAT box. Primer extension experiments, as well as cell-free transcription, revealed that a cytosine functions as the major transcription start point in a polypyrimidine region, but a guanosine at position -1 was also able to initiate transcription. Hybridization analysis of chromosomal DNA from a panel of human-rodent somatic cell hybrids revealed that hRPS3a is encoded by a single locus in the human genome, present on chromosome 4

Transcription factor TFIIA, defined by its role in transcription by RNA polymerase II, is also involved in RNA polymerase III transcription of mammalian U6 small nuclear RNA genes. This finding was substantiated by experimental evidence including (i) extensive copurification of an activity required for U6 transcription with TFIIA, (ii) the comparable molecular dimensions of this activity and TFIIA, (iii) the identical heat stability of both activities, and (iv) functional analyses revealing that TFIIA facilitates the interaction of TFIID with the TATA box of the U6 gene. As was shown previously for TFIID, TFIIA is the second basal transcription factor which could be demonstrated to be involved in gene expression by two different RNA polymerases.

The duck histone H5 gene contains a 12 base pair (bp) sequence motif within the coding region, which shows homology in 10 out of 12 bp with the consensus sequence of the USF binding site in the Ad2ML-promoter. The functional equivalent of transcription factor USF, partially purified from whole cell extracts of duck erythrocytes (EUSF), was shown to interact with this intragenic sequence. Electrophoretic mobility shift analyses revealed the selective formation of a complex between this protein fraction and the duck H5 gene. Footprint assays with DNase I delineated specific binding to the intragenic sequence outlined above. Moreover this protein fraction, containing EUSF, transactivates the expression of the duck H5 gene in vitro and elimination of the USF-consensus sequence leads to a loss of stimulation but retains the basic transcription of the gene. These results suggest an as yet unknown functional role of EUSF in the expression of the H5 gene of the duck.

Differentiation in vitro of mouse F9 embryonal carcinoma (EC) cells to the parietal endoderm (PE) mimics processes of development of the early mouse embryo. This differentiation is accompanied by a dramatic down-regulation of all genes transcribed by RNA polymerase III (pol III). Complementation of extracts from cells, differentiated for various time periods with purified pol III transcription factors show for the first time that TFIIIC1 can substantially restore this impaired transcription, particularly in the early stages of differentiation. At later stages (day 7) the TBP (TATA-binding protein )-TAF complex, TFIIIBbeta, may also become limiting, which can contribute to but cannot account for the reduced transcription of type 2 promoters in PE cells. Because TFIIIBbeta is not required for the expression of type 3 promoters, other components must necessarily be involved, and our results show that U6 transcription can significantly be reactivated by TFIIIC1. By employing a variant type 3 promoter construct, which essentially requires a mutant form of TBP (TBP-DR2), we show that TBP is not limiting in PE extracts. The partial purification of pol III transcription factors from PE and EC cells revealed that TFIIIC2 activity could be purified from both cell types, whereas TFIIIC1 activity was dramatically reduced in extracts from PE cells.

In eukaryotes, TFIIIB is required for proper initiation by RNA polymerase III. In the yeast Saccharomyces cerevisiae a single form of TFIIIB (gammaTFIIIB) is sufficient for transcription of all pol III genes, whereas in extracts derived from human cells two different hTFIIIB complexes exist which we have previously designated as hTFIIIB-alpha and hTFIIIB-beta. Human TFIIIB-alpha is a TBP-free entity and must be complemented by TBP for transcription of pol III genes driven by gene external promoters, whereas hTFIIIB-beta is a TBP-TAF complex which governs transcription from internal pol III promoters. We show that hTFIIIB-beta cannot be replaced by yeast TFIIIB for transcription of tRNA genes, but that the B" component of gammaTFIIIB can substitute for hTFIIIB-alpha activity in transcription of the human U6 gene. Moreover, hTFIIIB-alpha can be chromatographically divided into activities which are functionally related to gammaTFIIIE and recombinant yB"90, suggesting that hTFIIIB-alpha is a human homolog of yeast TFIIIB". In addition, we show that yeast TBP can only be exchanged against human TBP for in vitro transcription of the human and yeast U6 gene but virtually not for that of the yeast tRNA4Sup gene. This deficiency can be counteracted by a mutant of human TBP (R231K) which is able to replace yeast TBP for transcription of yeast tRNA genes in vitro.

Transcription factor IIIC (TFIIIC) binds in a sequence-specific manner to RNA-polymerase-III-transcribed genes (e.g. tRNA genes). It sequesters other transcription factors into the preformed complex, thereby activating transcription by RNA polymerase III. The Dictyostelium discoideum homologue of TFIIIC was highly purified by affinity chromatography based on its tDNA-binding activity. This TFIIIC homologue is a multicomponent factor (molecular mass 380 kDa), which binds to the B-box element of the internal tRNA gene promoter without significant A-box interaction. Partially purified D. discoideum TFIIIC is able to functionally complement a human RNA polymerase III in vitro transcription system depleted of human TFIIIC. We provide evidence that partially purified D. discoideum TFIIIC interacts in vitro with gene-external B-box elements present down-stream of many D. discoideum tRNA genes.

Protein fractions containing TFIIA, a transcription factor known to be involved in transcription initiation by RNA polymerase II and 5'-regulated polymerase III genes (e.g. U6), were tested for their role in in vitro transcription of classical pol III genes. These fractions were shown to stimulate a basal transcription system, reconstituted from highly purified fractions hTFIIIB and hTFIIIC. We demonstrate that this stimulating activity isolated from HeLa cells coelutes over at least six chromatographic steps with hTFIIA. Moreover the native molecular mass and the stability of this activity against heat treatment are comparable to those of hTFIIA. Finally we show that recombinant TFIIA from Saccharomyces cerevisiae can substitute for the human factor in pol III transcription in vitro which proves that TFIIA is also involved in the efficient expression of classical pol III genes.

We analysed the formation of transcription complexes on the H5 gene of the duck which is efficiently transcribed in HeLa cell extracts in vitro. Upon deletion of its TATA-box, the fidelity of transcription of the H5 gene is maintained, although the efficiency of this process is significantly reduced. Selective inactivation of TFIID in whole cell extracts and reconstitution experiments either with human recombinant TFIID or a protein fraction from duck erythrocytes enriched in TFIID show that transcription of the TATA-less H5 promoter nevertheless requires the protein TFIID. Screening of promoter elements which could indirectly mediate the interaction of TFIID with a TATA-less H5 promoter led to the identification of a sequence element located about 40 base-pairs downstream from the H5 initiation site that shows partial homology to the USF consensus sequence. In electrophoretic mobility shift and footprinting studies we demonstrated a specific interaction of the erythroid factor USF (eUSF) with this downstream element. By isolating active transcription complexes we found that all components required for correct initiation remain stably associated with the H5 promoter irrespective of the presence or absence of the TATA box. Moreover, the reconstitution of eUSF and TFIID-depleted transcription complexes with purified protein fractions demonstrate that not only TFIID but also eUSF essentially participates in complex formation even on H5 promoter mutations lacking the TATA-box. Mutual interactions between eUSF and TFIID appear to stabilize the binding of TFIID in the presence or absence of its proper binding site.

The Xenopus tropicalis U6 gene is very poorly transcribed both when introduced into human cells by transfection, and in human cell-free extracts. By analysis of hybrid promoters constructed from human and Xenopus sequences in various combinations, we show that species specificity is mediated by the proximal sequence elements (PSEs) of the promoters. We demonstrate the PSE-dependence of U6 transcription in a fractionated extract of HeLa cells. One of the fractions required for transcription contains an activity designated PSE-binding protein (PBP), previously shown to bind to the PSE of the mouse U6 gene. Binding of PBP to various wild-type and hybrid U6 PSE sequences correlates with their activity in transcription in HeLa cell extracts. This provides strong evidence that PBP is the PSE-binding factor involved in U6 transcription. In addition, it suggests that the differential affinities of the promoters for PBP is responsible for the observed species specificity. The divergence between U snRNA promoters in different species contrasts with the relatively strong conservation of other families of RNA polymerase II and III transcribed gene promoters. Possible mechanisms by which this diversity could be generated are discussed.

Transcription factors, required for the basal expression of the mouse U6 gene were identified in extracts from HeLa cells. This gene is transcribed at least four times more efficiently than its human counterpart in extracts from mouse or HeLa cells and hence provides an excellent in vitro system for the identification of transcription factors involved in the basal expression of mammalian U6 genes. At least four separate protein components were found to be required in addition to RNA polymerase III for correct synthesis of U6 RNA in vitro. These correspond to: (i) TFIIIB; (ii) a heat labile activity contained in a protein fraction enriched in TFIID; (iii) an, as yet, uncharacterized component contained in the flow-through upon rechromatography on phosphocellulose, and finally; (iv) a protein specifically binding to the mouse U6 gene promoter and transactivating its expression. Transcription factors IIIA and IIIC are not involved in mammalian U6 transcription in vitro. The U6-specific transcription factor has a molecular mass of approximately 90 +/- 10 kDa. It specifically binds to the U6 gene from bp -42 to -78 on the coding and from bp -37 to -79 on the non-coding strand thereby centrally encompassing the PSE motif of the mouse U6 promoter. The binding activity of this protein is correlated with the efficiency with which the U6 gene is transcribed in vitro, thereby indicating a crucial role of the PSE-binding protein for U6 transcription.

Transcription factor IIIC from human cells (hTFIIIC) contains a 55 kDa polypeptide which specifically binds to the promoter of the VAI and 5S gene. This interaction can be abolished by depleting divalent metal cations from the free protein through chelation with EDTA. Prior association of the protein with its DNA-binding sequence renders the complex refractory to chelation by EDTA. Specific binding of hTFIIIC to its cognate promoter sequences--shown by electrophoretic mobility shift and DNase I protection assays--can be restored by the addition of zinc ions. In contrast to the binding of hTFIIIA to the 5S gene, which was monitored in parallel and which exclusively requires Zn2+, the binding of hTFIIIC to the VAI and 5S gene can also be reconstituted--albeit with a lower efficiency--by the transition metals Co2+, Fe2+ and Mn2+ but not by Ni2+ or Cu2+. These results show that hTFIIIC binds to its promoter sequences in a metal coordinated fashion which differs from that observed for the binding of hTFIIIA to the 5S gene.

Human transcription factor IIIA (htFIIIA), specifically required for transcription of the gene for 5S ribosomal RNA has been characterized with respect to some of its physical, immunological and functional properties. TFIIIA from HeLa cells, which selectively binds 5S RNA, is a monomer of approximately 35 kDa with a Stokes' radius of approximately 2.65 nm and a sedimentation coefficient of approximately 2.8 S. These values indicate that the human protein is of rather globular shape and hence diverges not only in molecular mass but also in most of the molecular properties from its highly asymmetric counterpart in Xenopus laevis oocytes. By raising specific polyclonal antibodies against hTFIIIA it was shown in Western immunoblots that there was no cross-reaction between anti-hTFIIIA antibodies and the amphibian protein. Conversely, monoclonal antibodies against three domains of X. laevis TFIIIA antibodies and the amphibian protein. Conversely, monoclonal antibodies against three domains of X. laevis TFIIIA did not cross-react with the human transcription factor. The polyclonal antisera raised against hTFIIIA specifically neutralized binding of the human transcription factor to 5S DNA and abolished in vitro transcription of 5S RNA but these antibodies were unable to inhibit 5S RNA synthesis in cellular extracts from Xenopus, Drosophila or yeast cells. Finally, the species variation of TFIIIA could be substantiated by electrophoretic mobility shift assays revealing preferential binding of hTFIIIA to the homologous 5S RNA gene.

Human transcription factor IIIC contains a 55 kDa polypeptide which specifically interacts with the Adenovirus 2 VAI gene promoter and which mimics most of the DNA binding properties of the entire factor. The specificity and affinity of this protein:DNA interaction was demonstrated by: (i) Separation of purified fractions of hTFIIIC by SDS PAGE, electrotransfer to nitrocellulose, renaturation of proteins and their subsequent binding to the VAI gene, (ii) recovery and renaturation of proteins from SDS gels and identification of a fraction of hTFIIIC with a molecular mass less than 68 kDa, which specifically binds to VAI DNA, (iii) correlating the differential binding activity of the renatured 55 kDa component of hTFIIIC to mutated Pol III promoters with the ability of the entire factor to form functional transcription complexes thereon, and finally by (iv) specific crosslinking of the 55 kDa DNA binding component of hTFIIIC to the photoaffinity labeled B-box promoter sequence of the VAI gene.

RNA polymerase III presents a versatile experimental system for the analysis of eukaryotic transcription. Transcription complexes can be generated in vitro from a limited number of transcription factors and they can subsequently be isolated in a complete and fully functional form. The analysis of the essential DNA-protein and protein-protein interactions underlying the formation of such complexes requires the purification of the transcription factors involved. We succeeded to purify TFIIIA, TFIIIB and TFIIIC from HeLa cells and describe here some of their properties. TFIIIA from human cells is structurally and immunologically different from its functionally equivalent counterpart in Xenopus laevis oocytes. This raises interesting questions with regard to the structure-function relationship of these two proteins. TFIIIC, purified from HeLa cells, was shown for the first time to primarily and specifically interact with the gene for ribosomal 5S RNA. These results could shed light onto the hitherto unknown but essential role of TFIIIC during the synthesis of ribosomal 5S RNA. The long-term goal of our investigations is to clone the genes coding for the polIII transcription factors and to study their expression during different stages of growth and differentiation.

DNA-directed RNA polymerase was solubilized from total HeLa cells. Three distinct classes of the enzyme could be clearly differentiated by their sensitivity toward α-amanitin. While form A is completely resistant to high concentrations (133 μg/ml) of this toxin, enzyme B is highly sensitive and is completely inhibited by concentrations of 0.1 μg/ml. In contrast, RNA polymerase C shows an intermediate behaviour (50% inhibition at 30% μg/ml). Separation of the three individual enzymes was achieved by chromatography on DEAE-cellulose (to separate enzyme B from A and C) and DEAE-Sephadex (to separate polymerase A from C). All three RNA polymerases were subsequently purified by phosphocellulose chromatography followed by sedimentation through glycerol gradients. Analysis of the purified enzymes by gel electrophoresis under denaturating conditions showed that the A enzyme consists of five subunits with molecular weights of 185, 128, 65, 41 and 32 × 10 3 . In contrast, polymerase B is composed of seven subunits in variable stoichiometry with molecular weights of 215, 175, 145, 123, 68, 43 and 31 × 10 3 respectively. The subunit structure of enzyme C is not entirely clear at present and remains to be established. In addition, RNA polymerase activities were solubilized from mitotic and middle-S phase cells in comparison to controls. With respect to amounts and/or activities of all three RNA polymerases A,B and C no significant differences were detectable between logarithmically growing, mitotic and middle-S phase cells.

Transcription factor hTFIIIC was purified from cytoplasmic extracts of HeLa cells using four different chromatographic steps. This procedure yields a protein fraction which actively supports transcription in reconstitution assays and contains five major polypeptide chains with a molecular mass ranging from 25 to 250 kDa as estimated by SDS-PAGE and silver staining. In this fraction a polypeptide with a molecular mass of approximately 110 kDa could be identified as a specific DNA-binding component of hTFIIIC. By electrophoretic mobility shift and footprinting analyses it could be demonstrated that purified hTFIIIC binds specifically to the 5S gene. The protected region encompasses the A-Box promoter element and flanking sequences extending toward the 5'-proximal end of the gene. By addition of hTFIIIC to preformed TFIIIA/5S DNA complexes, we observe an additive effect of both factors on the footprint boundaries.

The gene coding for the major human ribosomal 5S RNA was chemically synthesized and cloned into a pUC13 vector. This approach was taken, because attempts to isolate the human 5S gene have thus far yielded either pseudogenes or variant 5S genes of unknown function. The synthetic human gene was transcribed by RNA polymerase III either in a crude HeLa cell extract or in a system reconstituted from partially purified transcription factors. Comparative studies with the Xenopus laevis somatic 5S gene show that the human gene is transcribed with similar fidelity and an efficiency of about 80% under optimal conditions. The time-course of transcription and optimal concentrations of template and transcription factors were found to be similar for both genes studied. The synthetic gene described may prove useful to study its interaction with human transcription factors in a homologous system.

Isolated HeLa cell nuclei were employed to catalyze the synthesis of RNA in vitro. In the presence of low concentrations of alpha-amanitin (1 mug/ml), used to suppress the formation heterogeneous nRNA, these nuclei synthesize RNA very efficiently for extended periods of time (at least 60 min) at an elongation rate of about seven nucleotides per second. The product, analyzed on sucrose density gradients and polyacrylamide gels was found to exist of two predominant size classes. Synthesis of the 45-S ribosomal precursor was completely resistant even to high concentrations of alpha-amanitin (150 mug/ml) and hence was catalyzed by enzyme A (or I). A limited degree of processing of the 45-S precursor occurred in vitro. In addition, a second RNA class of low molecular weight (4-8 S) was synthesized by HeLa cell nuclei in the presence of 1 mug/ml alpha-amanitin in vitro. Analysis on 8% polyacrylamide gels resolved the RNA into four distinct components. Their synthesis was resistant to low (1 mug/ml) but clearly sensitive to high (150 mug/ml) concentrations of alpha-amanitin. Consequently the synthesis of all these small-molecular-weight RNA species is catalyzed by RNA polymerase C (or III). For the assessment of the initiation frequency of the individual classes of RNA, a new technique was developed independent of labelling the 5' end of the RNA molecule with the gamma-phosphate of the initiating nucleotide. It employs the double labelling of an RNA molecule with two different isotopes added sequentially at different stages of completion of the chain. From the incorporation ratio of the two isotopes into a particular class of RNA, conclusions can be drawn concerning their initiation frequency. The results obtained have shown a high reinitiation frequency for the small-molecular-weight RNA species at all stages of the incubation reaction. In contrast, reinitiation of the 45-S precursor RNA occurs only to a limited extent in isolated HeLa cell nuclei in vitro.

DNA-dependent RNA-polymerase C, initially shown to exist in rat liver tissue and characterized by its sensitivity toward high concentrations of α-amanitin, has been shown to exist in HeLa cells and to be present in variable amounts in the nucleus of the rat liver cell. Enzyme C has been shown to co-chromatograph with nucleolar RNA polymerase A (or I) on DEAE-cellulose and can clearly be separated from the latter by chromatography on DEAE-Sephadex. It is possible that RNA polymerase C may occur in multiple forms although the significance of this finding is unclear at present. From its similar chromatographic properties and sensitivity toward α-amanitin it is possible that RNA polymerase C corresponds to polymerase III.

The fact that DNA is transcribed into RNA has long been known and has become part of the “central dogma” of molecular genetics. The mechanisms responsible for controlling this process at the individual genes of higher cells (eukaryotes), however, are still not completely understood. RNA polymerases apparently require a number of auxiliary factors (transcription factors) for gene recognition. These factors combine with the enzyme at the gene to form a transcription complex. The structures of these complexes are starting to become clearer; most is known about the control of RNA polymerase III, the enzyme responsible for the synthesis of certain small RNA molecules. The model character of the results obtained with this system is clearly underlined by a series of recent publications. TF III A is an especially intensively studied protein, which is a positive regulator for the expression of ribosomal 5S RNA and possesses structural properties that were previously unknown in DNA-binding proteins. It is becoming increasingly evident that the “architecture” of TF III A is not an exotic curiosity but probably exemplifies a general structural plan.

The gene for tRNAMet1 from Xenopus oocytes was transcribed in a cell free system with components isolated from a HeLa cell-free extract. It was found that, apart from the established assembly of transcription factors IIIB and IIIC on tRNA genes into stable transcription complexes, these factors can also associate with the enzyme in the absence of DNA to form a functional polymerase III complex. These complexes can be isolated in a highly active form from the bulk of other cellular proteins by mild methods such as gel filtration or density gradient centrifugation. When associated with RNA polymerase III into a functional complex, the transcription factors IIIB and IIIC can clearly be differentiated from free transcription factors, which individually display a much lower relative molecular mass. The polymerase complexes are stable against 1 M KCl, rendering unlikely that they represent fortuitous aggregates including RNA polymerase III and transcription factors IIIB and IIIC. These complexes are sensitive to dilution and, whereas transcription factor IIIC binds to the enzyme more tightly, factor IIIB tends to leak from the complex upon dilution of the protein concentration. From these results it is clear that in addition to their function as DNA-binding protein(s), transcription factors IIIB and IIIC can directly interact with RNA polymerase III without prior binding to the promoter region of the gene to be transcribed.

Das DNA in RNA transkribiert wird, ist eine alte Erkenntnis und als solche Teil des „Zentralen Dogmas” der molekularen Genetik. Wie dieser Prozes jedoch auf die einzelnen Gene hingelenkt wird, ist bei hoheren Zellen (Eukaryonten) immer noch nicht vollig verstanden. Die RNA-Polymerasen bedurfen offensichtlich zur Gen-Erkennung einer ganzen Reihe helfender Faktoren (Transkriptionsfaktoren). Diese bilden am Gen zusammen mit dem Enzym einen Transkriptionskomplex. Der Aufbau dieser Komplexe beginnt klar zu werden, und am grosten ist das Wissen uber die Steuerung der RNA-Polymerase III, die fur die Synthese bestimmter kleiner RNA-Molekule zustandig ist. Wie sehr die Erkenntnisse, die an diesem System gewonnen wurden, Modellcharakter haben, erweist sich besonders deutlich in einer Reihe neuester Arbeiten. Ein besonders gut untersuchtes Protein (TF III A), das ein positiver Regulator fur die Expression ribosomaler 5S-RNA ist, hat Struktureigenschaften, die bisher bei DNA-bindenden Proteinen unbekannt waren. Es zeigt sich immer mehr, das die „Architektur” von TF III A nicht die einer exotischen Kuriositat ist, sondern wahrscheinlich als Beispiel fur einen generellen Bauplan gesehen werden mus.

Cytoplasmic extracts from HeLa cells, capable of transcribing the cloned genes for ribosomal 5-S RNA, were employed to study the factors involved in this process. Two factors can be isolated, by gel filtration through Sephadex G-100, which are devoid of RNA polymerase activity. They significantly enhance the extent and specificity of the transcription of 5-S rRNA. Both proteins can jointly be purified by affinity chromatography on immobilized DNA containing the genes for ribosomal 5-S RNA from Xenopus borealis. Besides a protein of approximately 45 kDa, possibly corresponding to TF IIIA isolated from Xenopus oocytes, a second protein with a molecular mass of 22 ± 1 kDa stimulates the formation of 5-S RNA. This protein is contained in the breakthrough of DEAE-cellulose; it binds to and is eluted from phosphocellulose with 0.6 M KCl. In addition, it was found that the exclusion volume obtained after gel filtration on Sephadex G-100 contains functional complexes, which are capable of transcribing the cloned 5-S genes and hence contain all the required factors. Direct evidence is presented that the protein of 22 kDa described above is contained in and can be isolated from such complexes. It is postulated from indirect evidence that an additional factor with a molecular mass in excess of 100 kDa is required which can be removed from functional polymerase complexes by gel filtration through Bio-Gel A5m.

A homologous in vitro transcription system was developed in which the alpha A-globin gene of the duck was faithfully transcribed. Whole-cell extracts from duck erythrocytes were separated into fractions A, B, C and D by consecutive elution from phosphocellulose columns and were individually reconstituted in run-off transcription assays. Fractions A, C and D were required to achieve faithful initiation on the alpha A-globin gene. The latter fractions were mutually interchangeable with comparable fractions from HeLa cells. A fourth fraction, B, was not required but enhanced basal transcription when reconstituted with fractions A, C and D or a very low amount of HeLa whole-cell extract which by itself did not yield a detectable signal. Fraction B from duck erythrocytes was further purified by chromatography on DEAE-Sephadex and was shown to contain two trans-acting factors. One of these differentially acts on the alpha A-globin gene of the duck. The other component from duck erythrocytes surprisingly resembles the upstream stimulatory factor, previously isolated from HeLa cells. This latter protein binds to and trans-activates the adenovirus 2 major late promoter, but is not involved in the transcription of the alpha A-globin gene.

Chelation of divalent cations by 5 mM EDTA and subsequent removal by dialysis from a cytoplasmic HeLa cell extract leads to a complete loss of 5S rRNA transcription without affecting tRNA synthesis. Transcription complexes for 5S RNA can no longer be assembled in such a zinc-depleted extract and this ability can be fully restored only by the re-addition of 5 microM zinc. Reconstitution experiments with isolated protein fractions show that transcription factor A from HeLa-cells requires zinc to exert its specific function. Pre-formation of transcription complexes partially protects the metal ion against removal by chelation even in the presence of 1.8 M KCl. These results indicate that the zinc ions are bound to mammalian transcription factor IIIA which, in a transcription complex, binds very strongly to the 5S RNA gene. Cation depletion with 75 mM EDTA also suppresses tRNA transcription; an effect which is reversible by zinc addition. We conclude that beside for the binding of TF IIIA, zinc is also bound with a different affinity to a transcription component common to 5S and tRNA synthesis, most likely polymerase III itself.

The synthesis of ribosomal 5-S RNA in vitro in isolated chromatin from HeLa cells was studied by hybridization of the product to a plasmid DNA probe which contained the 5-S RNA genes from Xenopus mulleri. It was found that endogenous RNA polymerase C, contained in the chromatin, actively synthesizes RNA which hybridizes to the DNA for 5-S RNA. The molecular weight of this RNA, judged by gel filtration and electrophoresis in formamide-containing polyacrylamide gels, is of distinct 5-S size. Hybridisation of the product synthesized in vitro to the DNA for 5-S RNA can be competed quantitatively by the presence of unlabelled cytoplasmic 5-S RNA. Presaturation of the DNA-containing filters with cytoplasmic 5-S RNA prevents hybridisation, indicating that antisense transcripts are not synthesized to a measurable extent in this system. The addition of Escherichia coli RNA polymerase to chromatin stimulates the synthesis of bulk RNA, and beyond certain concentrations also leads to enhanced formation of hybridizable sequences. These are, however, predominantly contained in molecules of high molecular weight, indicating abberent initiations and/or terminations of the bacterial polymerase at random sites. These data re-emphasize the necessity of concurrent hybridization and size analyses for the assay of specific transcription. Homologous RNA polymerase C, added to chromatin either as a crude preparation in the presence of polymerase A or as purified enzyme, resulted in increased production of hybridizable 5-S RNA sequences, the formation of which was suppressible by high concentrations of α-aminitin. The RNA products synthesized under these conditions were predominantly asymmetric and hybridized in a discrete peak at the 5-S position. This is in contrast to results found for the bacterial enzyme which synthesized a high proportion of antisense transcripts on isolated chromatin. These findings indicate strongly that a particular RNA polymerase species of homologous source should be used when analyzing the transcription of specific genes in vitro.

In view of the widespread use of in vitro incubations of corpus luteum slices for a variety of purposes, experiments were conducted to establish optimum incubation conditions and to determine some of the sources of error variation observed in this system. A 2-hr incubation period was sufficient to detect an LH (luteinizing hormone) induced stimulation of progesterone synthesis; longer incubation times yielded higher levels of synthesis but did not increase the precision of the assay. The pH optimum for the incubation system was between 7.15 and 7.55. Addition of 1 μmole NADPH2 resulted in a consistent stimulation of progesterone synthesis, even in those groups already maximally stimulated with LH, suggesting that 2 different mechanisms are operative for the 2 stimulating agents. NADPH2 did not change the regression coefficient or the precision of the assay. Addition of nicotinamide (30 HIM) resulted in higher rates of progesterone synthesis, presumably by maintaining the integrity of the pyridine nucleoti...

The fact that DNA is transcribed into RNA has long been known and has become part of the “central dogma” of molecular genetics. The mechanisms responsible for controlling this process at the individual genes of higher cells (eukaryotes), however, are still not completely understood. RNA polymerases apparently require a number of auxiliary factors (transcription factors) for gene recognition. These factors combine with the enzyme at the gene to form a transcription complex. The structures of these complexes are starting to become clearer; most is known about the control of RNA polymerase III, the enzyme responsible for the synthesis of certain small RNA molecules. The model character of the results obtained with this system is clearly underlined by a series of recent publications. TF III A is an especially intensively studied protein, which is a positive regulator for the expression of ribosomal 5S RNA and possesses structural properties that were previously unknown in DNA-binding proteins. It is becoming increasingly evident that the “architecture” of TF III A is not an exotic curiosity but probably exemplifies a general structural plan.

Cytoplasmic extracts from HeLa cells, containing transcription factors and intact RNA polymerase C, were used to transcribe the cloned genes for ribosomal 5-S RNA from Xenopus borealis. It was found that the endogenous enzyme in these extracts synthesizes discrete products of 5-S size which hybridize predominantly to the 5-S RNA gene region and reveal an oligonucleotide pattern characteristic of authentic 5-S ribosomal RNA. The overall transcription of the plasmic DNA is greatly repressed in the presence of the cytoplasmic extract. The specific production of 5-S RNA by the cytoplasmic extract depends on an optimal DNA concentration, beyond which the excess DNA is transcribed unspecifically. The crude extracts contain a factor which can be enriched by chromatography on phosphocellulose and which binds to DNA. In the presence of the cytoplasmic extract, this factor significantly stimulates 5-S RNA synthesis at high DNA concentration. No stimulation is observed at limiting DNA concentrations, already saturated by endogenously contained factor. The synthesis of 5-S RNA in this system depends on the appropriate template since it is not observed with the cloned genes for 5-S from Drosophila. The cytoplasmic extracts endogenously contain a limited capacity to transcribe the genes for ribosomal 5-S RNA. This process can be stimulated significantly by the addition of optimal concentrations of purified RNA polymerase C, which by itself transcribes these genes in an entirely random fashion. Reconstitution of plasmid DNA with the DNA-binding protein and purified polymerase does not lead to the production of 5-S RNA. From these data we conclude that the cytoplasmic extracts contain additional factor(s) which are deficient in the purified polymerase molecule and which are required together with the DNA-binding protein for the specific synthesis of ribosomal 5-S RNA. The observed stimulation of 5-S RNA synthesis is not seen with RNA poiymerase C from yeast cells, indicating an evolutionary divergence of functional significance in the polymerase molecule.

Aflatoxin B1 inhibits rat liver nucleoplasmic RNA polymerase B (40–50%) and cytoplasmic RNA polymerase C activities (25–35%) if applied in vivo. Nucleolar RNA polymerase A activity is not inhibited under the same conditions. Aflatoxin B1 has no effect on the activities of purified RNA polymerase enzymes A, B and C or on [3H]UTP incorporation of isolated rat liver nuclei in vitro. Aflatoxin B1, upon preincubation with a rat liver microsomal fraction, however, is apparently converted to a compound which then inhibits the activities of purified nucleoplasmic RNA polymerase (20–40%), cytoplasmic RNA polymerase (10–20%) and the incorporation of [3H]-UTP into isolated nuclei (38%). Nucleolar RNA polymerase activity is not affected under these conditions. The microsomal fraction, which most effectively converts aflatoxin B1 to an inhibitor of RNA polymerase is also the most effective cellular fraction catalysing the metabolism of aflatoxin B1.

The polyanion heparin has been employed to study the interaction of rat liver DNA-dependent RNA polymerase A and its template under various conditions. Heparin very efficiently inhibits polymerase molecules, which are not bound to DNA or are associated with the template in a loose, i.e., non-specific fashion. Purified nucleoli, isolated from rat liver nuclei, contain RNA polymerase A in abundant quantities of which only a portion is bound in a transcriptional complex. Excess enzyme, which is contained in the nucleolus in a quasi free form, can be transferred to an exogenously added template and can be completely inhibited by the prior addition of heparin. The enzyme contained in a transcriptional complex, however, initiated in vivo and completing these RNA chains in vitro, is fully resistant to heparin. In contrast to these results it has been found that RNA polymerase A extracted from nuclei and purified by various chromatographic steps does not form heparin-resistant complexes, even after the enzyme has been bound to the DNA template. Moreover it has been found that purified RNA polymerase A transcribes truly native DNA extremely poorly, indicating that the enzyme is highly deficient in the act of initiation on duplex DNA. It is therefore questionable whether the interaction of the purified enzyme and isolated DNA represents binding to true initiation complexes as is observed in the intact nucleolus.

Isolated HeLa cell nuclei synthesize ribosomal 5S RNA of very nearly correct sequence and size. The in vitro product was resolved according to size on formamide-containing polyacrylamide gels and the fractions were subsequently hybridized to recombinant DNA containing the 5S genes from Xenopus mulleri. It could be shown that the 5S RNA synthesized in vitro differed only very slightly in size from the mature species labeled in vivo and contained a few extra nucleotides in some of the molecules. Analysis of the 3'-terminal base of molecules synthesized independently with four different nucleotides showed that the chains were almost exclusively terminated with uridine. Digestion of the in vitro product with ribonuclease T1 and analysis of the oligonucleotides on DEAE-Sephadex A-25 in the presence of 7 M urea revealed that the molar yield of the internal fragments agreed well with the expected theoretical values. The 3'-terminal fragments, however, were found to be present in three different species with the sequences CUUOH, CUUUOH, CUUUUOH which occurred with a frequency of about 60%, 20%, and 20%, respectively. From these data we conclude that 5S RNA synthesis in isolated HeLa cell nuclei was correctly initiated but that termination occurred with a slight ambiguity, adding either one or two uridine residues to some of the chains.

Transcription factor IIIB (TFIIIB), which by itself does not bind stably or specifically to DNA, was purified from cytoplasmic extracts of HeLa cells using five different chromatographic steps. This procedure yields one predominant polypeptide which represents 90% of the most highly purified preparation and shows a relative molecular mass of 60,000, when analyzed on sodium dodecyl sulfate-polyacrylamide gels. A similar value was obtained for the native protein by rate zonal centrifugation on glycerol gradients. From these data we conclude that TFIIIB from HeLa cells has a Mr of 60,000 +/- 5,000 and that it functions as a single polypeptide. Highly purified TFIIIB was required and sufficient for the specific transcription of the Xenopus laevis and human tRNA and 5 S RNA genes as well as those for VA RNA when reconstituted with RNA polymerase III and the other appropriate transcription factors.

We have investigated the transcription in a HeLa whole-cell extract of two evolutionary widely separated structural genes coding for duck alpha A globin and yeast actin. Transcription of isolated DNA fragments of the duck alpha A globin gene increases linearly up to relatively high concentrations of DNA. Size analyses and S1 mapping of the transcripts synthesized in vitro on either linear DNA fragments or supercoiled templates reveal that the alpha A globin RNA is initiated at the in vivo cap site and remains unspliced. The same assay conditions were used to transcribe the yeast actin gene. In contrast to the duck gene, size analyses and S1 mapping of the RNA products synthesized on both linear DNA fragments and the supercoiled template containing the actin gene show that the transcripts found in vitro do not stem from the in vivo cap site. The promoter of the yeast actin gene is not recognized in this system in vitro.

A protein factor which stimulates the transcription of native templates catalyzed by enzyme B from rat liver has been isolated from the cytoplasm of this tissue. It has been purified extensively by chromatography through DEAE-cellulose and CM-cellulose, hydroxylapatite, gel-filtration, rechromatography on DEAE-cellulose and isoelectrofocussing. The protein is of low molecular weight (±30000) with an isoelectric point of 9.6. This factor predominantly stimulates enzyme B (±500%) without markedly effecting other RNA polymerase species from this tissue. Experiments have been adressed to the question concerning the mechanism through which this phenomenon operates. The stimulation manifests itself without noticeable time-lag, immaterial of whether the factor is added before or after initiation of RNA synthesis and approximate estimation of the initiation complexes formed with or without factor, measured by direct filtration on millipore filters, lends no proof to the assumption that initiation is elevated. The degree of stimulation, measured on a template blocked by actinomycin D, thus allowing initiation but limiting propagation, is markedly diminished. This observation, taken together with the pronounced sensitivity of the factor-induced RNA synthesis toward pyrophosphate, an inhibitor of phosphodiester bond formation, suggests that the observed effect operates through the chain-elongation step. This assumption is supported by the fact that Ks, determinations for CTP are not altered by the factor in the face of a greatly enhanced value for V, suggesting that the rate of chain propagation is raised. Experiments in which the 5′ termini of the RNA were labelled with purine nucleoside [γ-32P] triphosphates show that the average chain length of the product is increased through the action of the stimulatory protein. The factor-induced stimulation is most pronounced on native templates of high molecular weight and it is possible that the protein studied aids in the unwinding process of such a DNA or enhances the rate of chain propagation in some other way.

DNA-dependent RNA polymerase (E.C. 2.7.7.6.) has in recent years been isolated from a number of different eukaryotic organisms and tissues and in all examined cases this enzyme has been found to occur in multiple forms [1]. With the exception of mitochondrial RNA polymerase [2–5], these enzymes are located in the cell nucleus although recent reports describe the existence of a cytoplasmically located enzyme [6, 7]. The biological significance and possible biosynthetic interelationships between the cytoplasmic and nuclear enzymes as well as the question concerning the role which these enzymes may play in controlling rate and specificity of RNA synthesis are incompletely understood at present.

Transcription factor IIIA (TFIIIA) was purified from cytoplasmic extracts of HeLa cells by developing a simple and efficient procedure employing phosphocellulose under widely differing ionic conditions followed by affinity chromatography on immobilized human 5 S genes. This procedure yielded a fraction containing human TFIIIA activity and a protein of 35 kDa as its major component. Moreover, we succeeded in renaturing the activity of human transcription factor IIIA (hTFIIIA) isolated after preparative sodium dodecyl sulfate-polyacrylamide gel electrophoresis and in identifying a polypeptide of 35 kDa with the transcription activity. This value differs from that reported for Xenopus TFIIIA. It could be demonstrated by footprinting analyses that hTFIIIA specifically binds to the internal control region of the human 5 S rRNA gene. The limits of protection slightly differ at the 3' border of the internal control region from those imprinted by Xenopus TFIIIA on the same gene. Comparative footprint analyses of hTFIIIA on the human and frog somatic 5 S rRNA gene, measured in titration, competition, and salt-stability experiments, demonstrated a higher affinity of the human factor to the homologous gene. These results, together with the difference in molecular mass of these functionally analogous proteins, reemphasize the importance of homologous systems for the analysis of mechanisms involved in gene regulation.

The DNA binding properties of human transcription factor PBP, which specifically binds to the proximal sequence element of mammalian U6 genes and which plays a pivotal role during their transcription, were analyzed both qualitatively and quantitatively. As a prerequisite, we analyzed the optimal conditions for DNA binding of the PBP by assaying the stability of the interaction against increasing concentrations of salt, dithiothreitol, and heparin. The protein, which does not induce DNA bending, has a characteristic sensitivity against elevated temperatures and precipitously loses activity between 41 and 43 degrees C, a property which can be used for selective inactivation of the protein. Subjection of the PBP to limited proteinase K treatment showed that the protein consists of at least two functional domains, one of which is required for DNA binding. The PBP binds to the PSE with a much higher specific equilibrium constant (Ks = 1.33 x 10(11) M-1) than to nonspecific DNA (Kn = 1.18 x 10(5) M-1). The association and dissociation rates of PBP.PSE interactions were quantitatively determined by kinetic analyses. The pronounced lag phase during the initiation reaction of mammalian U6 transcription in vitro is probably correlated with the slow binding of the PBP to its target sequence. Once formed, however, the PBP.PSE complex is very stable and has a much lower dissociation (kd = 1.84 x 10(-5) s-1) than association rate constant (ka = 0.18 x 10(6) M-1 s-1). Collectively, the results demonstrate that the PSE binding protein stably associates with a high affinity to its cognate promoter sequence, and this process represents one of the primary events in the formation of the preinitiation complex on the U6 gene. Finally, we analyzed the effect of individual base pair mutations within mammalian U6 PSE sequences on the binding of the PBP.