A quantitative assay for Xenopus 5S RNA gene transcription in vitro

Cell, Vol. 24, 809-817.

June

1981,

Copyright

0 1981

by MIT

A Quantitative Assay for Xenopus 5s RNA Gene Transcription in Vitro W. Michael Wormington, Daniel F. Bogenhagen, Eddie Jordan and Donald D. Brown Department of Embryology Carnegie Institution of Washington 115 W. University Parkway Baltimore, Maryland 21210

Summary The in vitro transcription of Xenopus 5s RNA genes and of deletion mutants of these genes has been quantitated by assays that measure the efficiency of transcription and the ability to compete with the transcription of a “wild-type” 5S RNA gene. The difference between the competition strength of one repeating unit of X. borealis somatic 5S DNA and its plasmid vector is fifteenfold. tRNA genes and the adenovirus VA RNA genes are weak competitors of 5S RNA synthesis (and vice versa). Deletion of the 5’ flanking region reduces the competition strength of somatic but not oocyte 5S DNA. Except for the influence of the flanking sequence, the regions within the 5S RNA gene that determine competition strength are those that have been shown to interact with a specific transcription factor that is required for accurate initiation of 5S RNA transcription. The major oocyte (Xlo) and trace oocyte (Xlt) 5S RNA genes from X. laevis are transcribed as efficiently as somatic 5S DNAs but compete only one fourth as well. This fourfold difference in the competition strength is due to oocyte-specific base changes within the intragenic control region. Introduction Previous studies from our laboratory have described the construction of systematic 5’ and 3’ deletions of one repeating unit of somatic 5s DNA from X. borealis. Assaying the ability of these deletion mutants to direct accurate initiation of 5S RNA synthesis in vitro has shown that a region between residues 50-55 and 8083 of the 5s RNA gene is necessary and sufficient to direct specific initiation (Sakonju et al., 1980; Bogenhagen et al., 1980). These results also indicated that the sequences adjacent to the initiation site, including at least 15 nucleotides of the 5’ flanking sequence, influence the exact start site of transcription. These initial analyses of the transcription of 5s DNA deletion mutants were qualitative in nature. They detected the presence or absence of accurate transcription-initiation events, but were not designed to measure quantitative changes in transcription. To quantitate the contribution of the intragenic control region as well as any influence of the flanking sequences on the transcription of Xenopus 5s RNA genes, we have designed the following two assays. First, a transcrip-

tion-efficiency assay measures the number of 5s RNA molecules transcribed per gene per hour. Second, a transcription-competition assay measures the ability of increasing amounts of one gene to inhibit transcription of a second wild-type 5s RNA gene. These assays are carried out in a complex oocyte nuclear extract (Birkenmeier et al., 1978), which limits our understanding of the molecular interactions that are responsible for the differences we have detected. However, two components required for 5s DNA transcription have been identified: RNA polymerase Ill, and a 40 kd protein that acts as a positive transcription factor and binds to the internal control region of both oocyte and somatic 5S RNA genes (Engelke et al., 1980). The oocyte transcription factor specifically stimulates 5s DNA transcription upon its addition to the oocyte nuclear extract and may therefore be the limiting component in the competition assay (Pelham and Brown, 1980). The results of these competition experiments support this hypothesis. Results A Competition Assay for 5S DNA Transcription To measure the contribution of different regions of the 5s RNA gene to efficient transcription, altered 5s DNAs were tested for their ability to compete with transcription of a control 5S RNA gene, pXbs1. In a standard 15 ~1 reaction containing 10 ~1 oocyte nuclear extract at a concentration of 5 pg/ml of pXbs1, more than 85% of the total RNA synthesized is 5s RNA. DNA is limiting at this concentration (Birkenmeier et al., 1978) and varying amounts of competitor DNA can be added in increasing amounts to 5 pg/ml of pXbs1 in a fixed reaction volume. The amount of competitor DNA needed to reduce pXbs1 transcription by 50% is used as the index of competition strength. To distinguish the RNA products of the competitor and the standard templates in the same reaction, it was necessary that one of them be modified to produce an RNA with a different electrophoretic mobility. The +20 5s RNA maxigene (Sakonju et al., 1980) contains two decameric synthetic Barn HI linkers inserted between residues 40 and 41 of the gene. This maxigene retains 48 residues of the normal 5’ flanking sequence and supports synthesis of a “5S-like” transcript that initiates approximately 20 bp within the gene. This “5s” RNA transcript migrates more slowly than normal 5s RNA in partially denaturing 7 M urea/ polyacrylamide gels, presumably due to its altered secondary structure. The result of a competition assay between a fixed concentration of pXbs1 and increasing amounts of the +20 maxigene is shown in Figure 1A; the reverse competition is shown in Figure 1 B. The assays were quantitated by cutting out and counting the radioactive bands from the gel, in order to plot a graph of the dependence of transcription on the amount of com-

Cell 810

Figure 1. Competitions +20 Maxigene

between

pXbsf

and

(A) pXbs1 (5 pg/ml) competed by +20 maxigene. (Lane 1) Opg/ml; (lane 2) 5 fig/ml; (lane (lane 5) 30 3) 10 fig/ml; (lane 4) 20 &ml; Irg/ml; (lane 6) 40 pg/ml: (lane 7) 50 pg/ml; (lane 8) 75 &ml. (6) +20 maxigene (5 pg/ml) competed by pXbs1. (Lane 9) 0 pg/ml; (lane 10) 5 pg/ml; (lane 11) 10 pg/ml; (lane 12) 20 rg/ml; (lane 13) 30 pg/ml; (lane 14) 40 pg/ml; (lane 15) 50 W/ml; (lane 16) 75 pg/ml.

DNA (Figure 2). Since the maxigene is indistinpetitor guishable in its competition strength from “wild-type” pXbs1, we have used them interchangeably as the standard DNA in subsequent competition assays. In these experiments, 10 pg/ml of pXbs1 reduced the transcription of 5 pg/ml of maxigene by 50% (and vice versa). This DNA concentration is assigned a competition strength of 1 .O. Therefore, if 20 pg/ml of a cloned 5s DNA were to be required to reduce transcription of pXbs1 or maxigene by 50%, that DNA would be assigned a competition strength of 0.5. Competition due to the addition of the plasmid vector pBR322 has a strength of approximately 0.07, which is fifteenfold weaker than the strongest competitor, pXbs1 (see Figure 3). 5s DNA Competition Is Specific The specificity of this competition assay was determined by comparing the ability of other genes transcribed by RNA polymerase Ill to compete with transcription of pXbs1. As shown in Figure 3, pAdl23, containing the adenovirus VA1 and VA2 RNA genes, and pXltDNA”? , containing a Xenopus methionine tRNA gene (Telford et al., 1979), compete weakly for transcription of pXbs1. A 50% reduction in transcription of pXbs1 is observed with addition of 70 pg/ml of pAd123 or of 100 pg/ml of pXltDNA”? . Competition by a Drosophila arginine tRNA gene (Silverman et al., 1979) is identical to that seen with the Xenopus tRNA gene (data not shown). The reciprocal experiment, adding increasing amounts of pXbs1 DNA to a fixed concentration of pXltDNA”? , shows that 5s DNA is a weak competitor of tDNA transcription (Figure 4). The difference between the transcription-efficiency and transcription-competition assays is illustrated by

Competitor

Figure igene

2. Equivalence

of Competition

DNA Ipq/mll

between

pXbs1

and +20

Max-

Competitions and processing of reactions were performed as described in Experimental Procedures. Results are expressed as the fraction of radioactivity incorporated into 5s RNA or +20 maxigene RNA in the control reaction. The data are derived from the experiment shown in Figure 1 and additional experiments. (W) +20 maxigene competed by pXbs1. (0- - -0) pXbs1 competed by +20 maxigene.

the comparison of 5s RNA and tRNA transcription. Whereas the VA1 RNA gene is transcribed poorly in this extract, synthesizing only one transcript per gene per hour, both 5s RNA and tRNA genes are transcribed efficiently, supporting the synthesis of 15 and 13 transcripts per gene per hour, respectively. This disparity between the transcription efficiency and competition strength suggests that different factors limit the transcription of 5s RNA and tRNA genes in

55 RNA Gene Transcription 811

Compehtor

Figure 3. Competition Nuclear Extract. (C--O) +20 maxigene. (A- - -A) pBR322.

Figure

4.

Competition

(M) pXltDNA”“1 peted by pXltDNAm”l

of pXbs1

by Genes

CO--O)

pAd123.

between competed

pXltDNA”“‘1 by pXbs.1.

Transcribed (C---C0

in Oocyte pXltDNAm”l

and pXbs1 (W)

pXbs1

com-

this system. We presume that the extent to which tDNAs and the adenovirus VA1 RNA gene exceed the plasmid pBR322 in competition strength against a 5S RNA gene reflects their competition for some common factor required for transcription of all these genes. One such factor could be RNA polymerase III. Competition by 5’ Deletion Mutants of pXbs1 Sakonju et al. (1980) presented the analysis of a systematic set of deletion mutants approaching the Xbs 5s RNA gene from the 5’ side. These qualitative transcription studies showed that the entire 5’ flanking sequence and the first 50 nucleotides of the gene can be removed without loss of synthesis of 5S RNA or of a 5S-size RNA. We have screened these 5’ deletion

Figure pXbs1

5. Competition

of +20

DNA

Maxigene

(p9/ml)

by 5’ Deletion

Mutants

of

5’ deletions were constructed as described by Sakonju et al. (1980). (M) pXbs1. (o--o) -26. (X--X) -1. (A-A) +47. (A-A) +61.0--0 +66. (B--W) +74.

mutants for their ability to compete with transcription of the +20 maxigene, in order to quantitate the effects of the deletions on transcription. Deletion of the 5’ flanking sequence to within 26 residues of the initiation site has no effect on competition strength (Figure 5). Removal of the remainder of the 5’ flanking sequence reduces the competition strength to 0.4. This assay therefore detects a quantitative contribution of the 5’ flanking sequence to transcription of somatic 5S DNA, in addition to its influence on the initiation site previously reported. Sakonju et al. (1980) found that intragenic 5’ deletion mutants removing up to 50 residues of the 5S RNA gene synthesize a 5S-size transcript in which deleted sequences are replaced by plasmid residues. Figure 5 also shows the analysis of these intragenic deletion mutants in the competition assay. Deletion of the first 27 residues of the 5S RNA gene causes no further decrease in competition strength, as compared to 5’ spacer deletions (data not shown). Deletion to residue 47 reduces the competition strength to 0.25. As 5’ deletions enter the control region (beyond residue +50 of the 5S RNA gene), the competition strength decline8 gradually. The deletions pXbsA5’+55 and +66 still compete more efficiently than pBR322, but they do not support accurate initiation. Sakonju et al. (1981) have shown that these 5’ deletions beyond the border of the control region can bind the transcription factor, even though they do not support initiation, thus accounting for their activity in the competition assay. The reductions in competition strength upon deletion of Xenopus sequences are not due to the juxtaposition of deleterious plasmid sequences, since deletions to the same endpoint give the same competition

Cdl

612

strength, regardless (data not shown).

of their orientation

within pBR322

Competition by 3’ Deletion Mutants of pXbs1 A similar analysis of the competition strengths of the 3’ deletion mutants described by Bogenhagen et al. (1980) is shown in Figure 6. The 3’ deletions that do not interfere with transcription termination do not reduce the competition strength (data not shown). However, deletion mutants that enter the gene from its 3’ end and abolish termination overcompete in this assay (data not shown). This enhanced competition strength is presumed to be due to the elongation of transcription, leading to prolonged binding of elongation factors, perhaps limiting their availability for reinitiation. To overcome this difficulty, these deletion fragments were joined to a terminal fragment of the Xbs gene to provide a strong termination site without adding back any sequence known to be involved in directing initiation (Bogenhagen and Brown, 1981). The termination fragment contains the final 15 nucleotides of the Xbs 5s RNA gene and a portion of the 3’ flanking sequence. These recombined 5s RNA genes produce discrete transcripts that vary between 109 and 140 nucleotides in length, and that are more suitable templates to be assayed by the competition assay. A deletion mutant retaining the first 97 residues of the gene competes as well as pXbs1. However, a deletion to residue +87 or +83 reduces competition strength to 0.3 or 0.1, respectively. The 3’ deletion to residue +83 is the furthest deletion that still supports accurate transcription initiation (Bogenhagen et al., 1980). Deletions from the 3’ side to +80 and beyond result in gene fragments that cannot be distinguished from pBR322 in their competitive strength and do not support accurate transcription initiation. In contrast to the

5’ deletion series, the loss of competition by 3’ deletion mutants exactly coincides with their loss of ability to support accurate initiation of transcription. The 3’ end of the control region therefore must be intact in order to observe the influence of the 5’ flanking sequence on transcription. In addition, the region between residues +83 and +97 influences transcription quantitatively. This region has been shown to be protected by interaction with the transcription factor (Sakonju et al., 1981). Correlation of Competition Strength with Transcription Efficiency of 5s DNAs The number of 5s RNA molecules synthesized per gene per hour in the oocyte nuclear extract by altered 5s DNAs was determined in order to compare transcription efficiencies with competition strengths. Table 1 shows that the transcription efficiency of the deletion mutants closely parallels the results obtained by the competition-strength assay. Deletions to -26 on the 5’ side and to +97 on the 3’ side are transcribed with wild-type efficiency. There is a gradual decline in transcription rate with progressive deletions into the gene from the 5’ side and a sharp loss of efficiency between residues +97 and +83 in the 3’ series. As previously reported (Sakonju et al., 1980; Bogenhagen et al., 1980) deletions into the control region do not support specific transcription initiation. Competition Strength of Oocyte and Somatic 5s DNAs Both oocyte and somatic 5s DNAs are transcribed in the oocyte nuclear extract. To determine whether quantitative differences in the transcription of these genes exist, single repeating units of the major oocytetype (Xlo) and the trace oocyte (Xlt) genes from X. laevis were assayed for their efficiency of transcription and their ability to compete with transcription of the Table

__ Compe+,tor

Figure

6. Competition

3’ deletions were Brown (1981). f-1

(A-A)

of pXbs1

DNA

(pg/ml

by 3’ Deletion

constructed as described + maxigene. c-1

+83.G--DpBR322.

1

Mutants by Bogenhagen +97. (A-A)

and +87.

1. Transcription

Efficiencies

of 5’ and 3’ Deletions

of pXbs1

5s DNA

5s RNA Molecules/Gene/t-h

pXbs1

15

-26A5'

15

-1lA5’

9

+26A!i'

8

+55A5'

0.05

+66A5'

0.03

+ 124A3’

15

+97A3'

15

+07A3'

10

+63A3'

0.1

Incubations (15 Al) contained 10 pl germinal vesicle extract, 5 Ag/ml DNA, 500 pM each CTP, GTP and ATP. 100 PM w3’P-UTP (40 Ci/ mmole). Reactions were processed as described in Experimental Procedures. Results are the means of three separate experiments.

5S RNA Gene Transcription 813

Figure 7. Competitions between +20 gene and pXlt400 and a Somatic/Cocyte brid 5S RNA Gene

maxiHy-

(A) +20 maxigene (5 Ag/ml) competed by PXlt400. (Lane 1) 0 pg/ml; (lane 2) 5 pg/ml; (lane 3) 10 pg/ml; (lane 4) 20 pg/ml; (lane 5) 30 pg/ml; (lane 6) 40 Ag/ml; (lane 7) 50 pg/ ml: (lane 8) 75 Ag/ml. (B) +20 maxigene (5 Ag/ml) competed by somatic/oocyte hybrid 55 RNA gene, (Lane 1) 0 pg/ml: (lane 2) 5 pg/ml; (lane 3) 10 pg/ ml: (lane 4) 20 pg/ml; (lane 5) 30 Ag/ml; (lane 6) 40 pQ/mt; (lane 7) 50 pg/ml; (lane 8) 75 pghl.

+20 5s RNA maxigene. As shown in Figure 7A, a single Xlt gene with an intact 5’ flanking sequence (pXlt400) has a competition strength only one-fourth that of an intact somatic 5s RNA gene. This result does not reflect a species-specific difference, since a single repeating unit of X. laevis somatic (pXlsl1) 5s DNA has the same competition strength as does pXbs1 (Table 2). Figure 8 shows that pXlt400 is overcompeted by addition of a very low concentration of + 20 maxigene. This is the expected result of competition between two genes with very different affinities for the same factor. Deletion of the 5’ flanking sequence of Xlt 5s DNA to a naturally occurring Hind Ill site 10 residues upstream from the initiation site (pXltl91) does not affect competition strength (Table 2). Both Xlt 5s DNA derivatives, pXlt400 and pXltl91, are transcribed nearly as efficiently as pXbs1, directing the synthesis of 14 transcripts per gene per hour. Whereas deletion of the somatic 5s DNA 5’ flanking sequence reduces transcription, deletion of most of the Xlt 5s DNA 5’ flanking region has only a minor effect on the efficiency of pXlt transcription. Thus the competition strength and rate of transcription of Xlt 5s DNA relative to somatic 5s DNA are independent of the presence of the 5’ flanking sequence of this oocyte-specific 5s RNA gene. However, the 5’ flanking sequences of the oocyte-specific genes are not irrelevant to transcription. Deletion of the Xlt spacer to residue -10 and the Xlo spacer to residue -11 causes inaccurate transcription initiation (data not shown), as previously shown for analogous 5’ flanking deletions of the Xbs 5s RNA gene (Sakonju et al., 1980). Separately cloned fragments of the major oocyte-type gene and pseudogene from X. laevis have been tested in these two assays (Table 2). These genes are also transcribed efficiently, but compete poorly with pXbs1. This pronounced difference in competition strength

Table 2. Competition Oocyte and Somatic

Strengths 5S DNAs

and Transcription

Efficiencies

5S DNA

Competition Strength

5S RNA Molecules/ Gene/Hr

pXbs.1

1 .o

15

1 .o

15

1 .o

15

pXlt

0.25

15

pXltA5’-10

0.22

13

0.25

14

0.25

14

0.25

14

pXls11 +20

maxigene

pXl0 pXloAB’-1

1

pXlo “pseudogene”

of

represents the first functional difference between oocyte and somatic 5s RNA genes that we have been able to detect. Two sorts of sequence differences between oocyte and somatic 5s RNA genes could contribute to the weakened competition strength of oocyte-type genes. Except for short blocks of homology, the 5’ flanking spacer regions of the five sequenced examples of Xenopus 5s DNA families are poorly conserved (Korn and Brown, 1978; Brown et al., 1979). These extensive sequence differences in the oocyte 5s DNA flanking region could be responsible for the difference in competition strength just as deletion of the Xbs 5s DNA flanking sequence impairs competition. Alternatively, the oocyte-type 5s RNA genes could compete poorly due to oocyte-specific base changes within the intragenic control region. Although Xenopus 5s RNA gene sequences are highly conserved in general, the oocyte-type genes differ from somatic-type genes at four positions: residues 30, 53, 55 and 79. Three of these sequence differences are located within the intragenic control region. To test the importance of these intragenic oocyte-

Cell 814

IO

IO

Xbs

70

GAAAGTGCCCW\TATCGTCTGATCTCGGAAGCCAAGCAGGGTCGGGCCTGG~AGTAC~

Hybrid

: TC---C

x1t

---C-T--CPTC---C--

Xl0

-T--C-+TA--

x1o’Ygene

--T--C--TG-TM-------c-

Figure 9. DNA Sequences and Oocyte 5s DNAs

of the Internal

Control

Regions

of Somatic

Noncoding-strand DNA sequences are shown. Oocyte-specific base Changes differing from the somatic nucleotide sequence are indicated. DNA sequences were determined by the method of Maxam and Gilbert (1977). Xbsl has an A residue at position 34, where the majority of Xbs repeating units have a C.

Discussion Competitor

Figure

8. Competition

(-1 pXbs1 by pXbs1.

between

competed

DNA lrg/mll

pXbs1

by pXlt400.

and pXlt400 (-1

pXlt400

competed

specific base changes, we constructed a hybrid Xbs 5s RNA gene containing Xlt 5s DNA-specific base changes within the control region. This was accomplished by hybridizing an intragenic restriction fragment of Xlt 5s DNA to a single-stranded copy of the Xbs 5s RNA gene cloned in the vector Ml 3 mp5. This primer was extended in vitro with E. coli DNA polymerase I (Klenow fragment) and the double-stranded regions were joined with T4 DNA ligase according to the method of Hutchison et al. (1978). The resulting hybrid gene was selected by the presence of a Hpa II restriction site at gene residue 62, introduced into the Xbs 5s RNA gene by priming with the intragenic Xlt 5s RNA gene fragment. The sequence of this somatic/oocyte hybrid is compared with the “parental” genes in Figure 9. The hybrid gene sequence is identical to the Xbs 5s RNA gene, except for four Xltspecific base changes at residues 53, 55, 56 and 62. The base changes at residues 56 and 62 are base transitions characteristic of the Xlt 5s RNA gene but not of the Xlo 5s RNA gene. The introduction of these four Xlt-specific base changes into the Xbs 5s RNA gene caused the hybrid gene to compete as poorly as the Xlt parent, as shown in Figure 7B. Since only the two base changes at residues 53 and 55 are shared among weakly competing oocyte-type genes, it is likely that only these two of the four base-pair changes produced are responsible for the reduction in competition strength. This is further supported by the observation that the Xlo pseudogene competes as poorly as the oocyte-type gene, but has additional base changes within the control region that appear to have no apparent influence on its ability to compete with +20 maxigene transcription.

The Competition Assay This assay mea8ures quantitative differences in the relative abilities of various genes transcribed by RNA polymerase Ill to compete with each other for factors limiting their transcription in an extract from oocyte nuclei. Using systematically constructed deletion mutants of 5s DNA, we have been able to identify nucleotide sequences that influence the competition strength of 5s RNA genes. Similar competition assays recently reported by Sprague et al. (1980) have revealed quantitative differences in the influence of 5’ flanking sequences of altered silkworm alanine tRNA genes transcribed in Xenopus and Bombyx extracts. Additional findings have been described for the flanking sequence of the adenovirus VA RNA genes (Fowlkes and Shenk, 1980). We have observed that the competition strength and transcription efficiency of genes may differ. The oocyte and somatic 5s RNA genes, as well as tRNA genes, exhibit similar transcription efficiencies in the Xenopus oocyte nuclear extract, yet these genes have decidedly different competition strengths. In contrast, the VA RNA genes compete about as strongly as tRNA genes for transcription of a 5s RNA gene, even though the VA RNA genes are not efficiently transcribed in the oocyte nuclear extract. These transcription reactions are carried out in a crude extract containing several components required for the accurate transcription of 5s RNA genes. Two of these factors have been identified: RNA polymerase Ill and a specific transcription factor. The latter is a 40 kd protein that is required for 5s DNA transcription in oocyte extracts (Engelke et al., 1980; Pelham and Brown, 1980). The following evidence implicates the interaction of this factor with 5s DNA as the primary explanation for the differences we have observed in competition strength and transcription efficiency among various 5s RNA genes. First, the transcription factor is the limiting component in the oocyte nuclear extract. The rate of tran-

5s RNA Gene 815

Transcription

scription of 5s RNA genes can be augmented by its addition to the extract (Pelham and Brown, 19801, whereas addition of purified RNA polymerase III has no stimulatory effect (Kern et al., 1979). Second, tRNA genes and the adenovirus VA RNA genes compete poorly in this assay. The reciprocal competition is equally poor, indicating that the competition between these genes is for relatively abundant common factors in the extract (such as RNA polymerase Ill). The 5s DNA transcription factor does not alter transcription of tRNA or the VA RNA genes (Pelham and Brown, 19801, nor does it bind to these genes in any specific manner in vitro (Sakonju et al., 1981). Third, some, but not all, of the competition data from the deletion series correlate with the qualitative borders of the internal control region derived from previous transcription assays (Sakonju et al., 1980; Bogenhagen et al., 1980) and from purified factorbinding experiments (Sakonju et al., 1981). As summarized in Figure 10, the 3’ deletion series shows that sequences downstream from residue 97 have no influence on competition strength. These deletions also do not affect the site of initiation of transcription (Bogenhagen et al., 1980) or binding of the factor (Sakonju et al., 1981). Deletion to residue 87 or 83 allows specific initiation of transcription but at a reduced rate. These deletion mutants are progressively weaker competitors. The 3’ border of the region protected by the transcription factor is residue 95/97 (Engelke et al., 1980; Sakonju et al., 1981). Mutants deleted to residue 83 still bind the factor, albeit more weakly than the control gene (Sakonju et al., 1981). Mutants deleted to residue 80 and beyond no longer initiate

l.Of @ ft .65 .‘i; .6 -

+97 +11s,124

-26

-II I -I

E e .4.-: 5 .*;

+2* +67

5'

3' 0

-40

-20

+I

Figure 10. Summary Mutants of pXbs1

l

20

+40

of Competition

+60

+ 60

Strengths

+100

transcription, do not bind to the transcription factor and fail to compete in this assay. The 5’ deletion series reveals a more complicated pattern of competition strength. Deletions extending beyond residue 50 of the gene support little or no 5s RNA synthesis (Sakonju et al., 1980). The transcription factor protects the wild-type gene to approximately nucleotide 40/46 (Engelke et al., 1980; Sakonju et al., 1981). Yet deletions of the 5’ flanking sequence, where there is no detectable binding of the transcription factor, influence transcription of Xbs 5s DNA by all assays that have been utilized. They reduce the competition strength and transcription efficiency of somatic 5s DNA, and they alter the start site of transcription. We do not see the first two quantitative effects with deletions of the flanking sequences of oocyte 5s RNA genes, but the start site of transcription of oocyte 5s DNA is altered. Deletions within the 5’ flanking sequence that influence transcription initiation have lost part or all of a conserved sequence located - 17 to - 12 in the spacers of all five Xenopus 5s RNA genes (Kern and Brown, 1978). The influence of the 5’ flanking sequence of somatic 5s RNA genes as measured in these assays depends on an intact control region. For example, a complete 5’ flanking sequence and the first 80 nucleotides of the gene is no stronger than the plasmid pBR322 in its ability to compete for somatic 5s DNA transcription. From this we conclude that the binding of the transcription factor to the control region is required but not sufficient for maximal transcription of somatic 5s DNA in this assay system. When the transcription factor is bound to the control region it influences RNA polymerase III, perhaps in conjunction with one or more as yet unidentified factors, to start transcription at the correct nucleotide of the gene. The sequence at and around the start site determines the efficiency and accuracy of this second interaction. The results of the 3’ deletion series can be viewed as a direct measure of the binding of the transcription factor to the control region. Deletions from the 5’ side into the control region that can no longer support accurate initiation still compete in the assay system, albeit weakly. Recent experiments by Sakonju et al. (1981) give a satisfactory explanation for this apparent paradox, since these extensive 5’ deletions can still bind the transcription factor to the remainder of the control region.

+120

of 5’ and 3’ Deletion

The Xbs 5s DNA repeating unit is shown below. Solid vertical bars: competition strengths of 5’ deletions. Open vertical bars: competition strengths of 3’ deletion mutants. Horizontal rectangle: the 5s RNA gene. Cross-hatched region (from residues +50 to +80): region required for initiation of 5s RNA synthesis. Solid horizontal bar: region (from +40/46 to +95/97) that is protected by binding to the transcription factor.

A Difference between Oocyte and Somatic 5s RNA Genes Oocyte-type 5s DNAs are transcribed as efficiently as somatic 5s RNA genes but they compete only onefourth as well. The major oocyte (Xlo) and trace oocyte (Xlt) 5s DNAs of X. laevis have indistinguishable competition strengths and rates of transcription. The weak competition strength of oocyte 5s DNA

Cell 816

compared to somatic 5s DNA has been shown to be the result of base changes within the control region. A hybrid 5s RNA gene with the base changes within the control region that are characteristic of the trace oocyte gene flanked by the somatic 5s DNA sequences competes as weakly as the oocyte-type genes. The Xlo pseudogene has a competition strength equivalent to that of the Xlo gene. As shown in Figure 9, the pseudogene, in addition to the specific base changes previously described, contains additional base changes not found within other oocyte-type control regions. These changes appear to have no influence on competition, since the competition strength of the pseudogene is equivalent to that of the Xlo gene. These results indicate that the failure to detect pseudogene transcripts in vivo is not due to the inactivation of a “promoter” sequence as proposed by Jacq et al. (1977). Preliminary results obtained with oocyte injection of the recloned pseudogene indicate that, although it is transcribed, its transcripts are unstable and are subsequently degraded (M. Wormington, unpublished observations). Recently, it has been shown that transcription of 5s RNA genes in somatic cells requires a transcription factor that is related but not identical to the oocyte factor (Pelham et al., 1981). Extracts prepared from somatic cells transcribe oocyte and somatic 5s DNAs with comparable efficiencies and somatic 5s RNA genes have a fourfold greater competition strength in these extracts than do oocyte 5s DNAs, just as we have described for oocyte extracts. This difference in competition strength cannot by itself account for the developmental inactivation of the oocyte-type 5s RNA genes in somatic cells (Ford and Southern, 1973). However, if two or more molecules of the transcription factor interact in a cooperative fashion upon binding to the control region, a small difference in the affinity of the protein for the two types of 5s RNA genes could become accentuated at low concentrations of the protein. The importance of a cooperative effect has been postulated to explain the differential interaction of repressor molecules, with lambda operator regions having different affinities for the protein (Ptashne et al., 1980). A similar mechanism could contribute to the differential expression of 5s RNA genes in vivo. Experimental

5S DNA with an intact 5’ flanking sequence (pXlt400) and a 5’ flanking sequence deletion mutant with its deletion endpoint 10 residues from the initiation site (pXltl91) have been previously described (Peterson et al., 1980). The construction and nomenclature of 5’ and 3’ deletion mutants of pXbs1 have also been previously described (Sakonju et al., 1980; Bogenhagen et al., 1980). pAd123 containing the adenovirus VA1 and VA2 RNA genes has been described (Pelham and Brown, 1980). pXltDNA”Y , containing a single X. laevis methionine tRNA gene (Engelke et al., 19801, was a gift from R. Roeder. The Ml3 mp5 cloning vector and host strain JM103 were obtained from J. Messing. Recloning of the Xlo Gene and Pseudogene The recombinant plasmid pXlo31 (Fedoroff and Brown, 1978), containing a 660 bp repeating unit of X. laevis major oocyte 5S DNA, was digested with Hind Ill and the 5s DNA insert was purified by agarose gel electrophoresis. The purified fragment was then selfligated with T4 DNA ligase and subsequently digested with Fnu 4H. This resulted in the production of a 193 bp fragment containing the Xlo gene flanked by 11 and 69 residues on its 5’ and 3’ sides, respectively, and a 467 bp fragment containing the pseudogene preceded by 11 bp of 5’ flanking sequence followed by the spacer region on its 3’ side beyond the naturally occurring Hind Ill site. This provides a termination site for the pseudogene transcript in a cluster of T residues in the spacer approximately 120 residues downstream from the initiation site. and results in the production of a discrete RNA transcript (Bogenhagen and Brown, 1981). The gel-isolated fragments were filled in with T4 DNA polymerase and blunt-end-ligated to previously phosphorylated Barn HI linkers (Sakonju et al., 1980). Following digestion with Barn HI, gene- and pseudogene-containing fragments were gel-purified and ligated to pBR322 and were used to transform E. coli strain HBlOl as described by Bogenhagen et al. (1980). The recombinant plasmids were screened, amplified and purified as described by Sakonju et al. (1980). All procedures were in accordance with the NIH guidelines (P2/EKl). The Xlo genecontaining plasmid was designated as pXlo316 and the pseudogene clone as pXlo31$1. Competition Assays and Analysis of Product The oocyte nuclear extract was prepared as described by Birkenmeier et al. (1978). The transcription reactions contained, in a total volume of 15 81. 10 mM HEPES (pH 7.4). 70 mM NH,CI, 0.1 mM EDTA, 2.5 mM dithiothreitol, 6% glycerol, 200 pM each of GTP. CTP and ATP. 20 pM u-~*P-UTP (-10 Ci/mmole). 5 pg/ml of pXbs1 or +20 maxigene and 0.5 to 200 pg/ml of competitor DNA. Reactions were started by addition of 10 ~1 of oocyte nuclear extract (equivalent to 10 oocyte nuclei) and were incubated for 90 min at 20°C. The assays were terminated by addition of 185 81 of 0.3 M sodium acetate in 1 x SET containing 0.5% SDS (SET is 0.15 M NaCI, 5 mM EDTA. 50 mM Tris-HCI [pH 7.41). The solution was extracted with 100 pl of SET-saturated phenol and precipitated with 3 volumes of ethanol at -20°C. The RNA was resuspended in 99% formamide containing 1 mM EDTA. 0.05% each of bromophenol blue and xylene cyanol and was subjected to electrophoresis in thin 7 M urea, 8% polyacrylamide gels as described (Sakonju et al., 1980). The autoradiogram was used to guide excision of radioactive bands for quanhtation by Cerenkov counting.

Procedures

Materials Restriction endonucleases Barn HI, Fnu 4H. Hha I, Hind Ill and Rsa I were purchased from Bethesda Research Labs. E. coli DNA polymerase I (Klenow) and nuclease Sl were from Boehringer Mannheim. Barn HI linkers were obtained from Collaborative Research. r3*PATP (3000 Ci/mmole), a-32P-UTP. w~‘P-GTP and a-32-dATP (-400 Ci/mmole) were from Amersham. Sources of Plasmid DNAs Recombinant plasm& containing single repeating units of X. borealis (pXbs1) and X. laevis (pXlsl1 I somatic 5S DNA. X. laevis trace oocyte

Construction of a Somatic/Oocyte Hybrid 5s RNA Gene The construction of the somatic/oocyte hybrid was performed by in vitro repair following the method of Hutchison et al. (1978). A derivative of pXbs1 was cloned in the Hind Ill site of Ml 3 mp5 vector to provide a supply of single-stranded template. The single-stranded DNA was purified from polyethylene glycol-precipitated phage by phenol extraction as described by Heidecker et al. (1980). and was further purified by chromatography on hydroxyapatite. A primer approximately 50 nucleotides long was prepared from a combined Hha I and Rsa I digestion of purified insert from pXlt400 (Peterson et al., 1980). This Xlt 5s DNA fragment spans gene residues 28 to 77. but was treated with Sl nuclease (Bogenhagen et al., 1980) to degrade

zS7RNA

Gene

Transcription

the cohesive end generated by Hha I digestion. This was necessary since the Hha I site is not contained within the Xbs 5S DNA sequence and the primer would otherwise have been mispaired with the Xbs 5s DNA template strand at the 3’ OH end of the primer. Approximately 4 pmole of duplex primer were annealed with 1 pmole of M 13-cloned Xbs DNA and in vitro repair was accomplished by incubation for 4 hr at 14% with 2 U Klenow fragment of E. coli DNA polymerase I in a 50 pl reaction containing 200 AM deoxynucleoside triphosphates. 6 PCi a-32P-dATP. 400 AM ATP. 50 mM Tris-HCI (pH 7.5). 10 mM MgCIP, 2 mM dithiothreitol and 1 U T4 DNA ligase. The reaction was terminated by heating at 68% for 10 min followed by chromatography on Sephadex G-100 equilibrated with 10 mM Tris-HCI (pH 7.5), 1 mM EDTA. The excluded peak was collected and an aliquot was treated with 200 U/ml Sl nuclease in Sl buffer (Bogenhagen et al., 1980) for 10 min at 0°C. The reaction was extracted with phenol followed by ether and the DNA was collected by ethanol precipitation. DNA was resuspended in TEN buffer (Cohen at al., 1972) and was used to transform CaCb-treated JM103 host cells. Since the hybrid confers no selectable properties, resulting Ml3 clones were screened by restriction analysis of crude isolates of RF DNA for inserts containing a new Xlt-specific Hpa II restriction site at gene residue 62. Following identification of the proper hybrid clone in Ml 3. the insert was excised and recloned in pBR322 for competition-transcription analysis. The DNA sequence changes were confirmed by the chemical DNA-sequence technique of Maxam and Gilbert (1977). and the 5S RNA synthesized in vitro was characterized by Tl RNAase digestion and fingerprint analysis (Sakonju et al., 1980).

RNA in kidney 241, 7-12.

Acknowledgments

Peterson, R. C.. Doering. J. L. and Brown, D. D. (1980). Characterization of two Xenopus somatic 5S DNAs and one minor oocytespecific 55 DNA. Cell 20. 131-l 41.

The authors thank J. Messing for providing the Ml 3 cloning vector and host. The manuscript was greatly improved by the helpful criticisms of S. McKnight, H. Pelham and S. Sakonju. W. M. W. is an NIH postdoctoral fellow. D. F. B. is a fellow of the Helen Hay Whitney Foundation. This research was supported in part by a grant from the NIH. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. Received

January

29, 1981;

revised

March

12. 1981

Bogenhagen. in Xenopus 261-270.

D. F. and Brown, D. D. (1981). 5S DNA required for transcription

E. (1978). A nuclear transcribes 5S RNA Nucleotide sequences termination. Cell 24,

Bogenhagen, D. F., Sakonju. S. and Brown, D. D. (1980). A control region in the center of the 5S RNA gene directs specific initiation of transcription. II. The 3’ border of the region. Cell 19, 27-35. Brown, D. D.. Korn. L. J., Birkenmeier, E. H.. Peterson, R. and Sakonju. S. (1979). The in vitro transcription of Xenopus 5S DNA. In Eukaryotic Gene Regulation, R. Axel. T. Maniatis and C. F. Fox, eds. (New York: Academic Press). p. 54. Cohen, S. N.. Chang, A. C. Y. and Hsu. L. (1972). Nonchromosomal antibiotic resistance in bacteria: genetic transformation of Escherichia co/i by R-factor. Proc. Nat. Acad. Sci. USA 69, 211 O-21 14. Engelke. D. R.. Ng, S.-Y., Shastry. B. S. and Roeder, R. G. (1980). Specific interaction of a purified transcription factor with an internal control region of 5S RNA genes, Cell 7 9, 717-728. Fedoroff, N. V. and Brown, D. D. (1978). The nucleotide sequence of oocyte 5S DNA in Xenopus laevis. I. The AT-rich spacer. Cell 73. 701-716. Ford,

P. J. and Southern,

E. M. (1973).

Different

sequences

of Xenopus

laevis.

Fowlkes. D. M. and Shenk, T. (1980). Transcriptional of the adenovirus VA1 RNA gene. Cell 22, 405-413.

Nature control

New Biol. regions

Heidecker, G., Messing, J. and Gronenborn, B. (1980). A versatile primer for DNA sequencing in the Ml3 mp2 cloning system. Gene 10, 69-73. Hutchison, C. A., Ill, Phillips, S., Edgell, M. H.. Gillman. S., Jahnke, P. and Smith, M. (1978). Mutagenesis at a specific position in a DNA sequence. J. Biol. Chem. 253, 6551-8560. Jacq. C.. Miller, J. R. and Brownlee. G. G. (1977). A pseudogene structure in 5S DNA of Xenopus laevis. Cell 12, 109-l 20. Korn, L. J. and Brown, D. D. (1978). Nucleotide sequence of Xenopus borealis oocyte 5S DNA: comparison of sequences that flank several related eucaryotic genes. Cell 75, 1145-l 156. Korn. L. J.. Birkenmeier, E. H. and Brown, D. D. (1979). Transcription initiation of Xenopus ribosomal 5S RNA genes in vitro. Nucl. Acids Res. 7, 947-958. Maxam. A. M. and Gilbert, W. (1977). A new method DNA. Proc. Nat. Acad. Sci. USA 74, 560-564.

for sequencing

Pelham, H. R. B. and Brown, D. D. (1980). A specific transcription factor that can bind either the 5S RNA gene or 55 RNA. Proc. Nat. Acad. Sci. USA 77, 4170-4174. Pelham, H. R. B.. Wormington. W. M. and Brown, D. D. (1981). Similar 5S RNA transcription factors in Xenopus oocytes and somatic cells. Proc. Nat. Acad. Sci. USA, 78, 1760-1764.

Ptashne. M.. Jeffrey, A., Johnson, A. D., Maurer, R.. Meyer, B. J., Pabo. C. 0.. Roberts, T. M. and Sauer. R. T. (1980). How the A repressor and cro work. Cell 19, l-l 1. Sakonju, S.. Bogenhagen. D. F. and Brown, D. D. (1980). A control region in the center of the 5S RNA gene directs specific initiation of transcription. I. The 5’ border of the region. Cell 79. 13-25. Sakonju. S.. Brown, D. D.. Engelke, D., Ng. S.-Y., Shastry. B. S. and Roeder. R. G. (1981). The binding of a transcription factor to deletion mutants of a 5S ribosomal RNA gene. Cell 23, 665-669. Silverman, S.. Schmidt, 0.. Soll. D. and Hovemann, 8. (1979). The nucleotide sequence of a cloned Drosophila arginine tRNA gene and its in vitro transcription in Xenopus germinal vesicle extracts. J. Biol. Chem. 254, 10290-l 0294.

References Birkenmeier. E. H., Brown, D. D. and Jordan, extract of Xenopus laevis oocytes that accurately genes. Cell 15, 1077-1086.

cells and ovaries

for 55

Sprague, K. U.. Lawson, D. and Morton, D. (1980). 5’ flanking sequence signals are required for activity of silkworm alanine tRNA genes in homologous in vitro transcription systems. Cell 22, 171178. Telford, J. L.. Kressmann. A., Koski. R. A., Grosschedl, R., Muller. F., Clarkson. S. G. and Birnstiel, M. L. (1979). Delimitation of a promoter for RNA polymerase III by means of a functional test. Proc. Nat. Acad. Sci. USA 76. 2590-2594.