Conditions providing enhanced transfection efficiency in rat pheochromocytoma PC12 cells permit analysis of the activity of the far-upstream and proximal promoter of the brain creatine kinase gene.

While brain creatine kinase (CKB) is expressed at highest levels in the brain, where it functions in regenerating ATP, the gene elements and protein factors regulating CKB transcription in neuronal and glial cells have not been identified. To investigate the regulation of CKB in neuronal cells, we examined the expression of the promoter proximal and 5' far-upstream regions of the rat CKB gene transiently transfected into rat PC12 pheochromocytoma cells. Initially, these experiments were hampered by the extremely low transfection efficiency of PC12 cells. We increased efficiency by greater than 200-fold by employing CaPO4-precipitated DNA transfection into PC12 cells which were optimized for transient transfection by: (i) culturing cells in polylysine-coated dishes to insure attachment throughout transfection; (ii) exposing cells to transfected DNA for an optimal time and employing a glycerol shock; and, most importantly, (iii) dissociating the characteristic self-adhesive clumps of PC12 into mostly single cells. Use of the plasmid expressing green fluorescent protein allowed identification of the transfected cells that averaged 10-20% of the total. Analyses of CKB promoter-CAT gene constructs showed that in PC12 cells expression of the proximal (0.2 kb) CKB promoter was low while expression of the 1.4 kb promoter was three fold higher and the 2.9 kb promoter was ten fold higher, suggesting the presence of at least two upstream cis-acting, positive regulatory elements. In agreement, the steady-state CKB mRNA level was higher in PC12 than in other neuronal cell lines examined, possibly reflecting the effects of positive upstream factors. The results are discussed in relation to how this economical and straightforward transfection procedure may be useful in identify factors regulating the transcription of CKB and other genes expressed in neuronal cells.

Proximal promoter of the rat brain creatine kinase gene lacks a consensus CRE element but is essential for the cAMP-mediated increased transcription in glioblastoma cells.

Our previous studies have shown that transcription of brain creatine kinase (CKB) mRNA in U87-MG glioblastoma cells is stimulated by a forskolin-mediated increase in cyclic AMP (cAMP) via a pathway involving protein kinase A (PKA) and the activation of Galphas proteins. In this report, we have employed transient transfection to investigate the rat CKB gene elements essential for the cAMP-mediated induction of rat CKB transcription in human U87 cells and have mapped the transcription start site of the induced CKB transcripts. We found that the level of induced transcription from the transfected genomic rat CKB gene was the same whether transcription was driven by 2.9 kb of CKB promoter plus 5' flanking sequence or the 0.2 kb CKB promoter, suggesting that the proximal CKB promoter was essential. Also, the level of induced transcription of the chloramphenicol acetyl transferase (CAT) reporter gene driven by the 2.9 kb CKB promoter was the same as with the 0.2 kb CKB promoter. Analyses of a series of 5' deletions of the 0.2 kb proximal CKB promoter showed that the sequences between -80 bp and +1 bp were essential for the cAMP-mediated induction of CKB transcription, despite the absence of a consensus cAMP response element (CRE) sequence in that region. In agreement, gel mobility shift assays showed that nuclear extracts from U87 cells contained a protein(s) which bound specifically to a [32P]CKB DNA probe containing the -60 bp to +1 bp sequence. Mapping the 5' end of the CKB transcripts showed that the initiation of the cAMP-induced transcription occurred almost exclusively from the downstream transcription start site, apparently under the initiation direction of the nonconsensus (-28) TTAA element and not the consensus (-60) TATAAATA element. The results are discussed with regard to nuclear protein factors which may be involved, and the possible cAMP-mediated increase in CKB transcription during myelinogenesis, since the differentiation of oligodendrocytes has previously been shown to be accelerated by increased intracellular cAMP.

Expression of the brain creatine kinase gene is low in neuroblastoma cell lines.

The cytoplasmic creatine kinase (CKB) enzyme has a central role in the regeneration of ATP in the brain. We have shown previously that CKB mRNA levels in cultured primary rat brain astrocytes and oligodendrocytes are much higher than in primary neurons. It has been suggested that high CKB expression is essential for the energy-demanding functions of glial cells. Conversely, CKB may be repressed in most neuronal cells; however, CKB protein has previously been detected by immunohistochemistry in several distinct groups of neurons in the adult rodent brain. Presently, little is known of the factors responsible for the high CKB expression in glia and possible repression in neurons. In this report, we investigated if low CKB mRNA was characteristic of some established neuronal cell lines. CKB mRNA was found to be extremely low in mouse C1300 neuroblastomas NS20Y and N1E-115 but 10-fold higher in NG108-15, a hybrid cell composed of a C1300 neuroblastoma and a rat C6 glioma. Since we showed NG108-15 contained only rat CKB mRNA transcribed from the C6 glioma CKB gene, expression of CKB mRNA may be a manifestation of a glial property in NG108-15 cells. However, CKB mRNA expression in NG108-15 appeared not to be fully activated since it was still 5-fold lower than in (parental) C6 glioma and 10-fold lower than in cellular RNA from either total rat brain or cultured primary astrocytes. When neuronal differentiation was increased in NS20Y and N1E-115 by treating cells with prostaglandin E1 and theophylline, the extremely low CKB mRNA level was not significantly changed. In a comparative study, the CKB mRNA levels in NS20Y, N1E-115 and neuronal RT4-B8 and RT4-E5 cells (from the rat RT4 peripheral neurotumor) were at least 50-fold lower than that in C6 glioma and 100-fold lower than in cultured primary astrocytes. These cell lines may provide a system for the identification of factors involved in the possible repression of CKB in many neuronal cells.

Expression of the brain creatine kinase gene in rat RT4 peripheral neurotumor cell lines and its modulation by cell confluence.

Creatine kinases (CK) catalyze the reversible transfer of a high energy phosphate group between creatine phosphate and ADP to regenerate ATP in cell types where the requirements for ATP are extensive and/or sudden. Previously, we have shown in primary rat brain cell cultures that brain CK (CKB) mRNA levels are highest in astrocytes and oligodendrocytes and much lower in neuronal cells. However, little is known of the factors which regulate CKB expression in the central nervous system and peripheral nervous system. To begin to investigate these factors, we asked in this report (1) if this pattern of CKB expression was also characteristic of some established glial and neuronal cell lines derived from the PNS; (2) whether CKB expression could be rapidly modulated by culture conditions, and (3) if CKB is expressed in cells with characteristics of glial cell progenitors. In subconfluent cells, CKB mRNA and enzyme activity were found to be high in both the rat RT4 peripheral neurotumor stem cell RT4-AC36A and its glial cell derivative RT4-D6. Conversely, CKB mRNA and activity were 5- and 8-fold lower, respectively, in the neuronal derivative RT4-E5 and, more dramatically, CKB was undetectable in neuronal RT4-B8 cells. Maintaining RT4-D6 glial cells at confluence rapidly increased CKB enzyme activity by 7-fold, such that D6 cells contained about 25% of the CKB level in lysates prepared from either whole adult rat brain or primary cultures of rat brain astrocytes. The levels of CKB mRNA and immunoreactive protein were also correspondingly increased in confluent D6 cells. These confluence-mediated increases in CKB appeared to be due to cell-cell contact and not the depletion of serum growth factors or an increase in intracellular cAMP. This study indicates that CKB expression is highest in cells displaying glial properties and can be rapidly modulated by appropriate culture conditions. The results are discussed in relation to the factors which may regulate CKB expression in vivo.

p53 binds to a novel recognition sequence in the proximal promoter of the rat muscle creatine kinase gene and activates its transcription.

The rat muscle creatine kinase (CKM) gene promoter is unusual since it is one of the few cellular promoters containing a p53 response element which is located proximally (bp -168 to -57) to the transcription start site. We have previously shown that p53wt transactivates transcription in vivo of rat CKM, in CV-1 monkey kidney cells, through this 112 bp promoter-proximal fragment which contains at least five degenerate p53-binding elements. In this report, we employed the gel-shift assay and demonstrated that recombinant, immunoaffinity-purified mouse p53wt binds to this 112 bp CKM sequence and activates the in vitro transcription of the proximal CKM promoter by nuclear extracts from CV-1 cells. Also, a competitor plasmid containing this 112 bp CKM fragment interferred with the in vivo transactivation of CKM by p53. This CKM fragment, when cloned upstream of the rat brain creatine kinase (CKB) promoter, mediated the p53 transactivation of CKB. Analyses of p53wt and a series of missense mutants (altered in conserved region II of p53) showed that binding of p53 to the CKM promoter was required but was not sufficient for transactivation. The results are discussed in relation to the possible role of p53wt in the expression of CKM in cell types which may not express the myogenic transcription factors.

Prostaglandin E1, E2, and cholera toxin increase transcription of the brain creatine kinase gene in human U87 glioblastoma cells.

The creatine kinase isoenzymes play an important role in maintaining ATP levels in some cell types during times of high energy demand. We have previously shown in primary cell cultures from rat brain that glial cells express much higher levels of brain creatine kinase (CKB) mRNA than neurons. In a separate earlier study we observed that transcription of CKB mRNA in glial cells can be stimulated by a forskolin-mediated increase in cAMP via a pathway involving protein kinase A (PKA). In this report, we show that the level of CKB mRNA in human U87 glioblastoma cells can be increased by either prostaglandin E1 (PGE1), prostaglandin E2 (PGE2), or cholera toxin (an activator of G alpha s proteins). The induction of CKB mRNA occurs rapidly (with maximal induction after 6 h), is at the level of transcription, and is mediated specifically through PKA. In addition, the results indicate that both PGE1 and PGE2 use the same or related signal transduction pathways to increase CKB transcription. These results suggest that in glial cells CKB mRNA can be regulated by extracellular signals acting through G-protein-coupled receptors. This study may contribute to an understanding of the mechanisms underlying the previously-reported, early postnatal increase in CKB enzyme activity in rat brain. The results are also discussed with regard to the potential involvement of the expression of prostaglandins and CKB during hypoxia and ischemia.

Mouse p53 represses the rat brain creatine kinase gene but activates the rat muscle creatine kinase gene.

The creatine kinases (CK) regenerate ATP for cellular reactions with a high energy expenditure. While muscle CK (CKM) is expressed almost exclusively in adult skeletal and cardiac muscle, brain CK (CKB) expression is more widespread and is highest in brain glial cells. CKB expression is also high in human lung tumor cells, many of which contain mutations in p53 alleles. We have recently detected high levels of CKB mRNA in HeLa cells and, in this study, have tested whether this may be due to the extremely low amounts of p53 protein present in HeLa cells. Transient transfection experiments showed that wild-type mouse p53 severely repressed the rat CKB promoter in HeLa but not CV-1 monkey kidney cells, suggesting that, in HeLa but not CV-1 cells, p53 either associates with a required corepressor or undergoes a posttranslational modification necessary for CKB repression. Conversely, mouse wild-type p53 strongly activated the rat CKM promoter in CV-1 cells but not in HeLa cells, suggesting that, in CV-1 cells, p53 may associate with a required coactivator or is modified in a manner necessary for CKM activation. The DNA sequences required for p53-mediated modulations were found to be within bp -195 to +5 of the CKB promoter and within bp -168 to -97 of the CKM promoter. Moreover, a 112-bp fragment from the proximal rat CKM promoter (bp -168 to -57), which contained five degenerate p53-binding elements, was capable of conferring p53-mediated activation on a heterologous promoter in CV-1 cells. Also, this novel p53 sequence, when situated in the native 168-bp rat CKM promoter, conferred p53-mediated activation equal to or greater than that of the originally characterized far-upstream (bp -3160) mouse CKM p53 element. Therefore, CKB and CKM may be among the few cellular genes which could be targets of p53 in vivo. In addition, we analyzed a series of missense mutants with alterations in conserved region II of p53. Mutations affected p53 transrepression and transactivation activities differently, indicating that these activities in p53 are separable. The ability of p53 mutants to transactivate correlated well with their ability to inhibit transformation of rat embryonic fibroblasts by adenovirus E1a and activated Ras.

Transcription of the brain creatine kinase gene in glial cells is modulated by cyclic AMP-dependent protein kinase.

The brain creatine kinase (CKB) gene is expressed in a variety of tissues with highest expression seen in the brain. We have previously shown in primary rat brain cell cultures that CKB mRNA levels are high in oligodendrocytes and astrocytes and low in neurons (Molloy et al.: J Neurochem 59:1925-1932, 1992). In this report we show that treatment of human U87 glioblastoma cells with forskolin and IBMX, to elevate intracellular cAMP, induces expression of CKB mRNA from the transiently transfected rat CKB gene by 14-fold and also increases expression from the endogenous human CKB gene. This induction of CKB mRNA i) is due to increased transcription; ii) occurs rapidly (with maximal induction after 6 hr; iii) requires the activity of protein kinase A (PKA), but iv) does not require de novo protein synthesis and, in fact, is superinduced in the presence of cycloheximide. Given the role of oligodendrocytes in the energy-demanding process of myelination and of astrocytes in ion transport, these results have physiological significance, since they suggest that changes in cellular energy requirements in the brain during events, such as glial cell differentiation and increased neuronal activity, may in part be met by a cAMP-mediated modulation of CKB gene expression. Of particular importance is the possible modulation of CKB gene expression during myelinogenesis, since oligodendrocyte differentiation has been shown previously to be stimulated by increases in cAMP.

Expression of the rat brain creatine kinase gene in C6 glioma cells.

We have recently shown that while brain creatine kinase (CKB) mRNA was detectable in RNA from cultured primary rat brain neurons, CKB mRNA was about 15-fold higher in primary astrocytes and 17-fold higher in oligodendrocytes (Molloy et al., J Neurochem 59:1925-1932, 1992). To begin to understand the molecular mechanisms responsible for brain glial cells containing the highest levels of CKB mRNA in the body, we have examined the expression of rat CKB mRNA in established C6 glioma cells. RNase-protection analysis showed the endogenous CKB mRNA levels in exponentially growing C6 were high and measured 50% of that in total RNA from rat brain lysate and 60% of that in cultured primary astrocytes and oligodendrocytes. The 5' and 3' ends of CKB mRNA in C6 were mapped to the same nucleotides as CKB mRNA from rat brain, indicating that the sites of in vivo transcription initiation and termination/polyadenylation of CKB mRNA in C6 are the same as in total rat brain RNA. The level of CKB enzyme activity in C6 whole cell lysates was among the highest of the glial cell lines which we measured. All creatine kinase enzyme activity present in C6 was found in the dimeric CKB isoform (BB), which is characteristic of CKB expression in the brain. A 2.9 kb gene fragment containing the basal CKB promoter and far-upstream 5' sequences was cloned upstream of the chloramphenicol acetyltransferase (CAT) gene and transfected into C6 cells. CAT activity was readily detectable in C6 and mapping of the 5' end of the CAT mRNA showed that transcription was directed from the correct initiation site. Since we found C6 cells were difficult to transfect, conditions were established which both maximized transfection efficiency and maintained normal C6 cell morphology. These results should permit the future identification of the nuclear trans-acting factors and the cognate cis-acting regulatory elements responsible for high CKB mRNA expression in brain glial cells.

Rat brain creatine kinase messenger RNA levels are high in primary cultures of brain astrocytes and oligodendrocytes and low in neurons.

Rat brain creatine kinase (CKB) gene expression is highest in the brain but is also detectable at lower levels in some other tissues. In the brain, the CKB enzyme is thought to be involved in the regeneration of ATP necessary for transport of ions and neurotransmitters. To understand the molecular events that lead to high CKB expression in the brain, we have determined the steady-state levels of CKB mRNA in homogeneous cultures of primary rat brain astrocytes, oligodendrocytes, and neurons. Northern blot analysis showed that whereas the 1.4-kb CKB mRNA was detectable in neurons, the level was about 17-fold higher in oligodendrocytes and 15-fold higher in astrocytes. The blots were hybridized with a CKB-specific 32P-antisense RNA probe, complementary to the 3' untranslated sequence of CKB, which hybridizes to CKB mRNA but not CKM mRNA. Also, the 5' and 3' ends of CKB mRNA from the glial cells were mapped, using exon-specific antisense probes in the RNase-protection assay, and were found to be the same in astrocytes and oligodendrocytes. This indicated that (a) the site of in vivo transcription initiation in astrocytes and oligodendrocytes was directed exclusively by the downstream, nonconcensus TTAA sequence at -25 bp in the CKB promoter that is also utilized by all other cell types that express CKB and (b) the 3' end of mature CKB mRNA was the same in astrocytes and oligodendrocytes. In addition, there was no detectable alternate splicing in exon 1, 2, or 8 of CKB mRNA in rat astrocytes and oligodendrocytes.(ABSTRACT TRUNCATED AT 250 WORDS)

Identification of cis-acting regulatory elements in the promoter region of the rat brain creatine kinase gene.

The functional organization of the rat brain creatine kinase (ckb) promoter was analyzed by deletion, linker scanning, and substitution mutagenesis. Mutations were introduced into the ckb promoter of hybrid ckb/neo (neomycin resistance gene) genes, and the mutant genes were expressed transiently in HeLa cells. Expression was assayed by primer extension analysis of neo RNA, which allowed the transcription start sites and the amount of transcription to be determined. Transfections and primer extension reactions were internally controlled by simultaneous analysis of transcription from the adenovirus VA gene located on the same plasmid as the hybrid ckb/neo gene. We demonstrate that 195 bp of the ckb promoter is sufficient for efficient in vivo expression in HeLa cells. A nonconsensus TTAA element at -28 bp appears to provide the TATA box function for the ckb promoter in vivo. Two CCAAT elements, one at -84 bp and the other at -54 bp, and a TATAAA TA element (a consensus TATA box sequence) at -66 bp are required for efficient transcription from the TTAA element. In addition, we present evidence that the consensus beta-globin TATA box responds to the TATAAATA element in the same way as the ckb nonconsensus TTAA element.

Identification of a novel TA-rich DNA binding protein that recognizes a TATA sequence within the brain creatine kinase promoter.

The rat brain creatine kinase gene possesses a structurally complex promoter with multiple potential regulatory elements. Two CCAAT sequences, a TATAAATA sequence and a TTAA sequence are found within the first one hundred base pairs. We present evidence that favors the allocation of the downstream TTAA sequence as the potential TATA box. We show that the CCAAT sequences and the upstream TATAAATA sequence are binding sites for potential regulatory factors and that sequences in this region are capable of regulating expression from the downstream TTAA sequence. We suggest that the protein that binds to the upstream TATAAATA sequence is not a classical TFIID factor but rather may serve to block the binding of TFIID and/or to promote transcription from the downstream start site. We have been able to define conditions in vitro under which binding to this upstream TATAAATA sequence does not occur. Under these conditions we are able to detect transcription from both potential TATA sequences, a situation which we have been unable to detect in vivo. Our experiments suggest the existence in HeLa and brain nuclei of a protein that recognizes the concensus TATAAATA sequence, that is distinct from TFIID, and that may function in part to deny access of TFIID to this potential promoter element.

5,6-Dichloro-1-beta-D-ribofuranosylbenzimidazole inhibits transcription of the beta-hemoglobin gene in vivo at initiation.

The mechanism by which 5,6-dichloro-1-beta-D-ribofuranosylbenzimidazole (DRB) inhibits transcription of the beta-major Hb gene in vivo was investigated in differentiated Friend erythroleukemic cells (FELC). Steady-state nuclear RNA from untreated and DRB-treated FELC was analyzed by the S1 nuclease assay using [32P]DNA probes labeled at their 3' end close to the site of transcription initiation. DRB severely inhibited transcription of full-length beta-major Hb mRNA precursor molecules and did not increase the production of short, promoter-proximal transcripts. In addition, nascent beta-major RNA transcripts were labeled with [alpha-32P]UTP in nuclei isolated from DRB-treated FELC and untreated FELC and hybridized to separate restriction fragments spanning the entire beta-major Hb gene. DRB inhibited transcription of the promoter-proximal and promoter-distal beta-Hb DNA restriction fragments uniformly by 75-80% and did not detectably increase the amount of short transcripts. Moreover, a brief reversal of the DRB inhibition in vivo increased the number of short, nascent, promoter-proximal beta-Hb transcripts apparently as a result of reinitiation. These data indicate that DRB inhibited transcription of the beta-major Hb gene in vivo at initiation.

Improved preparation of poly(A) Sepharose and the isolation of oligo(U) containing heterogeneous nuclear RNA from Friend erythroleukemic cells.

When poly(A) sepharose (prepared according to previously published procedures) was stored in aqueous buffer at 4 degrees C for 5 days or longer, it bound nonspecifically a high percentage of the input RNA which could not be eluted with formamide. We have found that treatment with ethanolamine, followed by dehydration with ethanol yielded poly(A) sepharose which was stable for many months and possessed a low degree of nonspecific binding. Chromatography on poly(A) sepharose permitted the specific isolation of that fraction of Friend erythroleukemic cell heterogeneous nuclear RNA (hnRNA) which contained oligo(U) sequences. Approximately 10% of the hnRNA which contained a poly(A) sequence [poly(A+)] also contained an oligo(U) sequence. Interestingly, prior HCHO denaturation of the hnRNA enhanced binding of the poly(A+) oligo(U+) hnRNA to poly(A) sepharose by tenfold. This suggested that the oligo(U) sequence may be in a region with secondary structure, possibly an intramolecular duplex with the 3' poly(A). Friend cell oligo(U) sequences ranged from 20 to 50 nucleotides in length and, thus, were similar to the oligo(U) sequences which heretofore had only been shown to be present in HeLa cell hnRNA. These results established that rodent cell hnRNA contain oligo(U) sequences and demonstrate, for the first time, that hnRNA containing both a poly(A) and an oligo(U) sequence can be separated from other classes of hnRNA. In addition, conditions are presented for the removal of HCHO from nucleic acid.

Induction of premature termination of transcription of the mouse beta-globin gene by 5,6-dichloro-1-beta-D-ribofuranosylbenzimidazole (DRB).

Hybridization of pulse-labeled RNA from DRB-treated Friend cells to mouse beta-globin cDNA revealed that the appearance of beta-globin mRNA in the cytoplasm was inhibited by greater than 87%. To examine the effect of DRB (125 microM) on HnRNA synthesis, nuclear RNA was electrophoresed in methyl mercuric hydroxide gels, transferred to nitrocellulose, and hybridized with beta-globin specific probes. Full-length nuclear transcripts, while present in untreated cells, were not detected in DRB-treated cells. Using restriction enzymes, the cloned beta-globin gene was divided into fragments proceeding from the 5' gene region to the 3' gene region. RNA labeled in vitro by transcription in nuclei isolated from DRB-treated cells hybridized only to the promoter proximal DNA fragment. Transcripts hybridizing to fragments from both the 5' and 3' regions of the gene were produced in nuclei from untreated cells. Together these results indicate that DRB causes premature termination of transcription within the beta-globin gene.

The isolation and characterization of HeLa cell messenger RNA-like molecules containing uridylic acid-rich oligonucleotide sequences.

HeLa cell mRNA, isolated from mRNPs released from polysomes by EDTA to minimize contamination from nuclear RNA, has been separated into four classes of molecules: 1) those containing poly(A) sequences [poly(A+)] as well as oligo(U) sequences [oligo(U+)] (4 to 13%); 2) poly(A+) oligo(U-) (approximately 52%); 3) poly(A-) oligo(U+) (approximately 2%); 4) poly(A-) oligo(U-) (approximately 33%). The oligo(u) segments are 89% UMP and range from 20 to 50 nucleotides in length. The poly(A+) oligo(U+) mRNAs appear to contain the oligo(U) in a region with secondary structure (possibly an intramolecular duplex with the 3'-poly(A) since they did not bind to poly(A) Sepharose without prior HCHO modification. HCHO modifies the exocyclic amino groups of CMP, GMP, and AMP and prevents hydrogen bonding but reacts only slightly with UMP. After removal of the excess HCHO, the oligo(U) of the mRNA was free to bind to the poly(A) Sepharose. Control experiments indicated that the binding of oligo(U+) mRNA to poly(A) Sepharose was specific and the HCHO did not cause cross-linking of the RNAs. The poly(A+) oligo(U+) RNAs appeared to contain one oligo(U)/molecule and could re-bind to poly(A) Sepharose with 31 to 75% efficiency. The poly(A+) oligo(U+) mRNA class was more heterogeneous in size and slightly larger (approximately 3 kb average length) than the other mRNA classes.

5,6-Dichloro-1-Beta-D-ribofuranosylbenzimidazole inhibits initiation of nuclear heterogeneous RNA chains in HeLa cells.

The nucleoside analog 5, 6-dichloro-1-beta-d-ribofuranosylbenzimidazole (DRB) at 75 to 150 micromolar concentrations inhibits the synthesis of nuclear heterogeneous RNA (hnRNA) in HeLa cells by 60 to 70 percent. The sedimentation profile of hnRNA labeled with (3H)uridine for 45 seconds after brief treatment (45, 90, or 180 seconds) with DRB showed a progressive decrease in the labeling of shorter hnRNA molecules relative to longer molecules. Prior exposure of the cells to actinomycin D, an inhibitor of RNA chain elongation, did not alter the sedimentation profile of hnRNA. These results suggest that DRB preferentially inhibits the initiation of hnRNA chains so that after exposure to DRB for a brief period the longer nascent chains still remain to be finished and thus incorporate a greater share of the pulse label. By progressively increasing the time of exposure to DRB, and measuring the rate of increase in the average size of the labeled, nascent RNA, it was estimated that the chains were growing at rates between 50 and 100 nucleotides per second.