Gene. 120 (1992) 143-149 0 1992 Elsevier Science Publishers B.V. All rights reserved. 0378-1119/92/$0.5.00

143

GENE 06711

Acanthamoeba

castellanii

RNA

polymerase II transcription accurate initiation at the adenovirus major late promoter

in vitro:

(Recombinant DNA; amoeba; protozoa; HeLa extracts; cr-amanitin; conservation of transcription factors)

Feng Liu and Erik Bateman Department of Microbiology and Molecular Genetics, University of Vermont, Burlington, VT 05405-0068.

Received by ST. Case: 13 April 1992; Revised/Accepted:

USA

11 June/l2 June 1992; Received at publishers: 3 July 1992

SUMMARY

We have developed and characterized an efficient in vitro system from the protozoan, Acanthamoeba castellanii, that accurately initiates transcription from the adenovirus-2 major late promoter (Aw,C,P). Transcription by A. castellanii RNA polymerase II (pol II) is initiated at the same nucleotide (nt) that is used by HeLa extracts and is dependent upon adenovirus sequences located between nt -51 and the region around the tr~sc~ption start point (rsp). The results suggest that the A. castellanii transcription factors for pol II which determine the tsp and the promoter elements that they recognize have been functionally conserved during evolution.

In eukaryotes, several proteins are required for accurate and regulated transcription initiation by pol II. A partial list of factors required for basal transcription initiation includes TFIID, TFIIB, TFIIE, TFIIF and TFIIA (Saltzman and Weinmann, 1989; Sawadogo and Sentenac, 1990). These fractions are able to fiction on a core promoter as typified by the AdMLP, and under appropriate conditions are responsive to the positive effects of various upstream activator factors (Maniatis et al., 1987; Mitchell and Tjian, 1989). Several of the basal TFII have recently been cloned and expressed (Hoffmann et al., 1990; Lewin, 1990; Ha Co~e~~ondence to: Dr. E. Bateman, Dep~m~t of ~i~robiolo~ and Molecular Genetics, University of Vermont, Burlington, VT 05405-0068, USA. Tel. (802) 6568608; Fax (802) 6568749.

Abbreviations: A., Acanthamoeba; AWL?, adenovirus-2 major late promoter; bp, base pair(s); kb, kilobase or 1000 bp; nt, nucleotide(s); PMSF, phenylmethylsulfonyl fluoride; pol II, RNA polymerase II; TAF, TBP-associated factor(s); TBP, TATA box-binding protein; TFII, transcription factor(s) for pol Ii: TPCK, ~-tosyl-L-ph~ylalanine chloromethy1 ketone; tsp, ~~sc~ption start point(s).

et al., 1991; Peterson et al., 1991; Finkelstein et al., 1992), but it has not yet been possible to completely reconstitute transcription from cloned components. One of the factors, TFIID, probably occurs in the cell as a multi-protein complex consisting of TFIID and one or more proteins dubbed TATA box-binding protein (TBP)-associated factors (TAFs) (Pugh and Tjian, 1991). TFIID may additionally be complexed with another group of proteins termed adaptors or coactivators (Lewin, 1990), and has been shown to interact directly with TFIIA (Usuda et al., 1991), which may be a member of a class of proteins that regulate TFIID activities. Purified pol II is incapable of promoter site selection or initiation on intact DNA templates, these functions being provided by the TFII (Sawadogo and Sentenac, 1990). Pol II itself is a complex enzyme; for example, A. custezianiipol II contains twelve subunits (D’Alessio et al., 1979). The general TFII and pol II interact by forming a pre-initiation complex that is dependent on initial TFIID binding to the TATAAA region (Saltzman and Weinmann, 1989; Sawadogo and Sentenac, 1990). Addition of each protein to the initiation complex and subsequent initiation is likely to be an ordered process and involves an energy-dependent step

144 and one or more phosphorylations of pol II, which occur on the C-terminal repeat (Saltzman and Weinmann, 1989). However, the detailed architecture of the pol II initiation complex remains to be elucidated, partly due to the difI% culty in obtaining transcription factors in a form without cross-cont~ination of one or more other factors. A. castellanii is a small, free living amoeba that has been exceptionally useful for in vitro studies on transcription by RNA polymerases I and III (Paule, 1990; Zwick et al., 1991). A. casteflunii can be induced by starvation to differentiate into a dormant cyst that has morphological properties distinct from growing cells. During encystment, a number of genes encoding 5s rRNA, 39s pre-rRNA and a thioredoxin-like protein are down-regulated, while expression of others, such as cellulose synthetase become active (Potter and Weisman, 1972; Paule, 1990; E.B. and J.-M. Wong, unpublished observation). Little is known about A. castellanii promoters for genes transcribed by pol II, nor how they are regulated constitutively or during differentiation, but the simplicity of its differentiative response and its suitability for biochemical studies make A. castellanii an attractive system for the study of gene expression. In this report we describe the development of a new in vitro transcription system from A. castellunii. Several genes from A. castellanii have been cloned and sequenced, but in all cases the cloned genes are members of gene families, and un~biguous demonstrations of in vivo expression or tsp use are not available. We therefore used the well-studied AdMLP (Sawadogo and Roeder, 1985), which is noted to function particularly well in vitro, even in heterologous systems. We show that initiation by A. castellanii pol II occurs at the same nt within the AdMLP that is used by a HeLa cell extract, and that initiation is dependent on promoter sequences between nt -51 and the region around the tsp.

RESULTS

AND DISCUSSION

(a) Transcription from the AdMLP by Acanthamoeba custellunii nuclear extracts In order to develop an accurate tr~scription assay, we used the plasmid pML(C~AT)~~ which contains the AdMLP attached to a DNA fragment lacking guanosine (Sawadogo and Roeder, 1985). The tsp in this promoter has been well established, as have the positions of critical promoter elements (Concino et al., 1984; Yu and Manley, 1984; Sawadogo and Roeder, 1985). The criteria for determining whether accurate tr~sc~ption occurs from pML(C~AT)~~ are that tr~sc~ption initiation by pol II must occur at the same site used by mammalian pol II in vivo or in vitro in a promoter-dependent manner. If so, the promoter containing fragment from plasmid pML(C,AT),, digested with EcoRI and &a1 [pMLC,AT),s/EcoRISmaI] is predicted to produce a 388~nt runoff transcript, whereas that from EcoRI and ~~~dIII-digested template ~pMLC~AT)~~~~coRI-~~~dIII] should produce a 418~nt runoff transcript (Fig. 1A). As shown in Fig. 2, A. custeZlunii nuclear extracts produce transcripts close to the predicted length when assayed with the adenoviral templates, suggesting that transcription is specifically initiated at a site within the AdMLP and proceeds through the G-free region. Because transcripts of the correct size might arise by artefactual mechanisms or be due to RNA pol~erase I or III transcription, we next performed several control experiments to establish the size and authenticity of the runoff products. Because transcription runs through the G-free stretch, treatment of transcripts with RNase Tl will not affect correctly initiated transcripts. If initiation is mediated by A. castellanii pol II, it will be sensitive to low concentrations of a-~anitin (Detke and Paule, 1978)

+1 +

EC0RI t-,-404

Sma I

TATAAA

G- Froo

Ad MLP

Mind III Smo I

HP8 II -

TATAAA

-51

Ad YLP

G-Free

1 Sma

Ndol t

G-Fro0

I

i

PUC 13

-200

.

388

.

418 bp

bpc

primer + n9

Fig. 1. Diagram (Sawadogo

of DNA fragments

and Roeder,

used for in vitro transcription.

1985). (C)The promoter-less

p(C,AT)

(A and B) Promoter-containing

DNA fragment containing

from pUC13. The distances, in bp, between the zsp ( + 1) and the template reactions is indicated, its 5’ end within the G-Free casette is at nt + 119.

fragments

derived

a region lacking G nt (G-Free)

ends are shown. The position

from the plasmid with 5’-flanking

pML(&AT),,

sequences

derived

of the 20 nt primer used for primer extension

145 1

2

Comparison of the A. castellanii runoff transcripts to those from HeLa extracts (Fig. 3, lanes 1, 4, 9 and 10) shows that they are similar or identical in size. The HeLa transcripts serve as accurate RNA size markers, since they have been previously characterized (Sawadogo and Roeder, 1985). Surprisingly, the A. castellanii extracts are considerably more active on this promoter than extracts from HeLa cells, since we typically assay 5 ~1 of the A. castellanii extract and 20 ~1 of the HeLa extract prepared according to Dignam et al. (1983). Transcription is directed by pol II, since addition of 1 pg a-amanitin/ml to the transcription reaction completely inhibits synthesis of the runoff products, as well as most of the background transcription arising from nonspecific initiation (Fig. 3, lanes 1, 2, 4 and 5). The a-arnanitin sensitivity of A. castellanii RNA polymerases have been previously characterized (Detke and Paule, 1978) and as expected, transcription of the A. castellanii 5s rRNA-encoding gene by RNA polymerase III is unaffected by this concentration of a-amanitin (Fig. 3, lanes 7 and 8). As predicted, transcripts produced from pML(C,AT),,/ EcoRI-SmaI are unaffected by RNase Tl (Fig. 3, lane 3), demonstrating that the correct strand is transcribed. There is, however, some shortening by RNase Tl of the transcript made from pML(C,AT),dEcoRI-Hind111 (Fig. 3, lane 6). This is because the Hind111 site is in the multiple cloning site of the vector and the transcript contains several G residues at its 3’ end.

3

-418

‘,

388-

CCA tRNA

(b) Initiation occurs at the same site as in the mammalian extract Fig. 2. In vitro transcription SmaI;

lane 2, runoff

Hind

produced

transcript

III; lane 3, transcription

pML(C,AT),,/EcoRI-SmaI.

produced

using pML(C,AT),dEcoRIusing pML(C,AT),@oRI-

by the post-nuclear Numbers

pH

7.5/10

EDTA/2% of ATP,

mM

UTP,

mmol)/20-30 performed solution

Each 50 ~1 reaction MgCl,/l

glycerol/O.1

mM

DTT/75

PM CTP/0.03

(0.6 M Na.acetate

acrylamide

pH 5.2/0.2% and analyzed

gel under denaturing

where

(Bateman

and Paule,

nuclear

extracts

is essentially

mM

K.acetate/O.OZ

by addition SDS/200

(3000 Ci/ were

of 50 ~1 of stop

pg yeast tRNA/ml).

essentially

1988). The procedure identical

mM

mM each

DNA. Reactions

electrophoretically

conditions

by the

25 mM HEPES,

PM [a-32P]CTP

ng template

at 25°C for 60 mm and stopped

Samples were processed

tRNA end-labeling

block II (Stratagene)/0.4

pg of nuclear extract/50

using

show the predicted

mixture contained

units RNase

and GTPjl

supernatant

in the margin

size (in nt) of the runoffs; CCA tRNA indicates extract. Methods:

by A. castellanii nuclear

from the AdMLP

Lane 1, runoff transcript

extracts.

on a 6% polyas described

for preparation

to the methods

of Ohlson

Edlund (1986) and Zwick et al. (1991), with the critical exception

elseof and

that all

buffers contained 0.2 mM PMSF/l mM benzamidine/O.l mM TPCK/2 pg leupeptin per ml/2 pg pepstatin per ml. In addition, nuclei were washed once (Buffer A; Ohlsson ammonium SmaI and

and Edlund,

1986) prior to lysis with 0.4 M

sulfate and final extract preparation. pML(CzAT),s/EcoRIpML(C,AT),g/EcoRI-HindIII denote agarose gel-purified

DNA fragments

as shown in Fig. 1.

The tsp used by A. castellanii pol II on the AdMLP was more accurately mapped by primer extension analysis of the in vitro RNA transcripts (Fig. 4A). The single primer extension product is close to the predicted size of 119 nt in length as determined by comparison to size markers, and maps to the adenosine at nt + 1 (underlined) within the AMLP promoter sequence: 5’-8 TCGTCCTCACTCTC + 6; 3’ as determined by comparison to a nt sequence ladder derived from the template. The same primer extension product was obtained regardless of whether pML(C,AT)JEcoRI-SmaI or pML(C,AT),$ EcoRI-Hind111 was used as template, ruling out that primer extension products terminated at template ends (Fig. 4B). Products from the A. castellanii extract were also directly compared to those from a HeLa extract, again using primer extension (Fig. 4B), which shows clearly that they are identical in size. These results demonstrate unambiguously that A. castellanii pol II initiates at the adenosine nt previously identified as the mammalian tsp.

146 RNass

-

tl

-

+

-

-

+

-+--+-

ainanlt in Ml

*

2

345

6

”

-

_

-

-

+

_

-

7

8

9

10

F

am

(c

e418 -388

Fig. 3. Control experiments to establish the authenticity of transcripts. The additions shown at the top of each lane were made to transcription reactions containing the following templates; lanes 1-3, pML(C*AT),~~coRI-~~~1; lanes 4-6, pML(C~AT)~~~EcoRI-NindIII; lanes 7 and 8, supercoiled pAcSS.3. Lanes 9 and 10 show transcripts produced by a HeLa extract using pML(C,AT),dEcoRK-SmaI and pML(C,AT),dEcoRI-ltrind III, respectively. For RNase Tl digestion of transcripts, samples were processed as above after transc~ption, except that tRNA was omitted. The mixture was dissolved in 50 ~1 water, heated to 90°C for 5 min and chilled on ice to destroy possible RNA secondary structures. 50 units of RNase Tl were added and incubated at 37°C for 15 min. The reaction was then processed and analyzed by electrophoresis as for standard tr~s~ription reactions as described in the legend to Fig. 2. Numbers and arrows in the margins show the product lengths (in nt) and their position. 5s indicates the position of the 5s rRNA transcript in lanes 7 and 8.

A

B

12

34M -

.

147

MPW4 5

7.5

K+(mM)

12.5 15

10

40

80

60

100

120

b

0-l 4

I

I

I

1

,

,

6

6

10

12

14

16

WI++

Fig. 5. Optimization as shown. in arbitrary

of transcription

A densitometer

reaction

conditions.

was used to determine

units is shown graphically.

(c) Extract characteristics

The MgCl,

were otherwise

or K.acetate

transcript

as described

and cation optima

with a nt sequencing

ladder

60

concentrations

of transcription

analysis of A. custellanii and HeLa in vitro transcripts.

Fig. 4. Primer extension

40

derived from pML(C,AT),,.

140

80 K+

A. castellanii nuclear extracts were prepared using a modification of previously published procedures (Ohlsson and Edlund, 1986; Zwick et al., 1991), in which crude nuclei are lysed in high salt and the resulting extract concentrated by precipitation with ammonium sulfate. Successful preparation of A. castellanii extracts capable of accurate initiation by pol II was dependent on including several protease inhibitors (PMSF, benzamidine, TPCK, leupeptin and pepstatin) during nuclei isolation. In contrast, extracts active for transcription by RNA polymerase III can be prepared without these protease inhibitors, suggesting that the protease-sensitive component is peculiar to pol II gene transcription. We do not know which extract component is protease-sensitive, but one possibility is pol II, since the C-terminal repeat of the largest subunit is protease

the AdMLP

20

ww

the relative amount

All methods

‘oM4_ ImY)

used for in vitro transcription

at each concentration,

reactions

and the relative absorbance

were varied of the bands,

in Fig. 2.

sensitive and can influence the efficiency of transcription in vitro (see Zehring et al., 1988; Saltzman and Weinmann, 1989; Sawadogo and Sentenac, 1990; Zehring and Greenleaf, 1990; Usheva et al., 1992; Koleske et al., 1992). In accordance with this idea, we have noticed distinct chromatographic behaviour of pol II prepared with or without leupeptin, pepstatin and TPCK (E.B. and F.L., unpublished). The post-nuclear supernatant contains no polymerase activity assayed specifically (Fig. 2, lane 3) or nonspecifitally (not shown),and we found that prior to lysis, the nuclei can be washed in low salt buffer with quantitative retention of pol II activity assayed specifically or nonspecifically, and this step removes about 60% of the total protein from the crude nuclei and results in an improvement in transcription activity. Pol II is by far the most (Panel A) Comparison The sequence

of the A. custellunii primer extension

of the transcribed

strand

is shown.

product

son of extension products of transcripts produced by A. custellunii extracts (lanes 1 and 3) or HeLa extracts (lanes 2 and 4). pML(C,AT),s/SmuI used as transcription template in lanes 1 and 2; pML(CaAT),s/HindIII in lanes 3 and 4. A 20-nt primer (5’-ATGATGATAGATTTGGGAAA), plementary

to the G-Free

cassette

(Jones et al., 1985). Samples spond to the sequence

beginning

were analyzed

of the -non-transcribed

at nt + 119 (relative to the tsp) was used in primer extension electrophoretically strand.

as in Fig. 2. The sequence

shown,

reactions

and the 5’ terminal

essentially

from

(Panel B) Compari-

as described

was com-

elsewhere

nt, T + 1, within the figure corre-

148 abundant

polymerase

present

in A. castellanii nuclear

ex-

tracts, accounting for approx. 80% of nonspecific transcription. Extracts additionally contain an activity that labels tRNA, presumably by CCA addition, significant amounts of GTP, and perhaps other nucleotides. We optimized the activity of nuclear extracts by varying the K.acetate and MgCl, concentrations in the transcription reactions. Transcription showed an optimum K*acetate concentration of about 80 mM and the optimum MgCl, concentration was 10 mM (Fig. 5). Chloride at a concentration of 75 mM was inhibitory on transcription from the AdMLP when compared to acetate (not shown), and consequently, all extracts are routinely assayed in acetate. The optimum amount of DNA was approx. 50 ng/50 ~1 reaction, but we found this to vary from extract to extract. (d) Acunthamoeba castelfunii in vitro transcription is dependent on the AdMLP Transcription from the AdMLP by mammalian pol II is dependent on the TATA box, and can be stimulated by additional upstream-binding factors such as USF (Sawadogo and Roeder, 1985). We tested which sequences were required for initiation by A. castellanii pol II by comparing the level of transcription from templates containing AdMLP 5’-sequences downstream from nt -404 or nt -51 and from a promoterless G-free casette (pC,AT) in which AdMLP sequences are replaced by vector sequences (Fig. 1). The template containing only AdMLP sequences downstream from nt -51 is transcribed as efficiently as a template containing promoter sequences downstream from nt -404 (Fig. 6, lanes 1 and 2). However, no product is made from the template in which promoter sequences have been replaced by vector sequences (Fig. 6, lane 3) indicating that transcription is dependent on AdMLP sequences between nt -51 and nt +lO. These results indirectly suggest that general transcription factors from A. castellanii extracts are functional on the AdMLP, whereas there is apparently no counterpart to human stimulatory factors such as USF, since the USFbinding site is deleted in the -5 1 mutant. It is probable that the TATAAA box and initiator region are the major elements utilized by our extracts, since the 5’ deletion to nt -51 is fully active and because previous deletion analyses have not shown a requirement for other promoter elements within this region (Concino et al., 1984; Yu and Manley, 1984). We are currently testing this idea using homologous A. castellanii promoters as well as the AdMLP. The A. castellanii transcription components that direct initiation of transcription by pol II appear to have been conserved during evolution, since A. castellanii pol II initiates transcription at the same site within the AdMLP as mammalian pol II, suggesting a similar manner of promoter

388

Fig. 6. Transcription tained S’deletions nt -51.

from AdA4LP deletions. up to the following

Lane 3 shows

Fig. 1). Transcription in Fig. 2. The position

transcription

reactions

Transcribed

positions;

templates

lane 1, nt -404;

from p(C,AT),$NdeI-SmaI

were performed

and analyzed

of the 38%nt runoff transcript

con-

lane 2, (see

as described

is shown.

recognition by the respective factors and similar interactions among them. In support of this notion, we have recently cloned A. castellanii TFIID which shows approx. 85% aa sequence identity to human TFIID in its conserved C-terminal domain, and is functional in TFIID-depleted HeLa extracts (Wong et al., 1992). However, it has been suggested that a component other than TFIID determines the distance from the TATAAA box to the tsp, and that this component (possibly TFIIB) differs between yeast and humans (Buratowski et al., 1988). Our results suggest that the equivalent A. castellanii factor closely resembles the human factor since A. casteflanii extracts choose the same tsp as used in human extracts. A. castellanii extracts differ from yeast extracts in that yeast transcription initiation from the AdMLP does not occur at nt + 1, but begins at sites within the G-free cassette about 63-69 nt downstream from the TATA element (Lue et al., 1989). (e) Conclusions (I) We have developed a new transcription system from A. castezlaniithat accurately initiates transcription from the AdMLP.

149 (2) Initiation occurs at the adenosine nt within the AdMLP that was previously identified as the tsp utilized by mammalian pol II (Sawadogo and Roeder, 1985; Weil et al., 1979). (3) Transcription is dependent on the AdA4LP region that includes the TATAAA box and sequences around the

Jones,

K.A., Yamamoto,

factors

A.J.,

Buratowski,

transcription merase Lewin,

S., Nonet,

II CTD and TFIID.

B.: Commitment

R.A.:

A novel

link between the RNA poly-

Cell 69 (1992) 883-894.

and activation

interactions.

at pal II promoters:

a tail of

Cell 61 (1990) 1161-1164.

P.M., Sugimoto,

by yeast RNA polymerase

are functionally conserved with respect to their human counterparts, suggesting that many of the interactions that occur during transcription have remained unchanged throughout evolution.

transcription

in vitro. Cell 42

M. and Young,

factor reveals a functional

Lue, N.F., Flanagan,

(4) The multiple components wihin A. castellanii extracts

kinase promoter

(1985) 559-572. Koleske,

protein-protein

tsp.

K.R. and Tjian, R.: Two distinct

bind to the HSV thymidine

K. and Komberg,

II at the adenoviral

R.D.: Initiation

major late promoter

in

vitro. Science 246 (1989) 661-664. Maniatis,

T., Goodboum,

and tissue-specific Mitchell,

S. and Fischer,

gene expression.

P.J. and Tjian,

J.A.: Regulation

Science

R.: Transcriptional

cells by sequence-specific

of inducible

236 (1987) 1237-1245.

regulation

DNA binding proteins.

in mammalian

Science 245 (1989)

371-378. Ohlsson,

ACKNOWLEDGEMENTS

H. and Edlund,

factors Paule,

We wish to thank Dr. R.G. Roeder for the gift of plasmids pML(C,AT),, and p(C,AT),,, Dr. Mike Zwick and Dr. M.R. Paule for the A. custellanii 5s rRNA gene, and Dr. Greg Gilmartin for HeLa extracts. Supported in part by grant No. EY 08706 from the National Eye Institute to E.B., NSF Vermont EPSCOR Grant No. RI18610679, and a grant from the Lucille B. Markey Charitable Trust to the Department of Microbiology and Molecular Genetics, University of Vermont.

M.R.:

In search

interactions

of nuclear

Cell 45 (1986) 35-44.

of the single transcription

factor.

Nature

344

O., Admon,

A.,

(1990) 819-820. Peterson,

M.G.,

Reinberg,

Inostroza,

J., Maxon,

M.E.,

D. and Tjian, R.: Structure

human general transcription Potter, J.L. and Weisman,

Flares,

and functional

factor HE. Nature

R.A.: Correlation

and in vitro during the encystment

properties

of

354 (1991) 369-373.

of cellulose synthesis

of Acanthamoeba.

in vivo

Develop.

Biol.

28 (1972) 472-479. Pugh,

B.F. and Tjian,

requires

R.: Transcription

a multisubunit

TFIID

from a TATA-less

complex.

Genes

promoter

Develop.

5 (1991)

1935-1945. Sahzman,

A.G. and Weinmann,

of RNA polymerase Sawadogo, REFERENCES

T.: Sequence-specific

with the insulin gene enhancer.

R.: Promoter

II transcription.

specificity and modulation

FASEB

M. and Roeder, R.G.: Interaction

tion factor with the adenovirus

J. 3 (1989) 1723-1733.

of a gene-specific

major late promoter

transcrip-

upstream

from the

TATA box region. Cell 43 (1985) 165-175. Bateman,

E. and Paule, M.R.: Promoter

transcription. Buratowski,

S., Hahn,

yeast TATA

S., Sharp,

Sawadogo,

RNA

protein

adenovirus

L.: Function

in a mammalian

of a

major late promoter

TATA box and initiation

in vitro.

Nucleic

Acids

R.: The

site are both

Res.

12 (1984)

J.M., Spindler,

S.R. and Paule, M.R.: DNA-dependent

II from Acanthamoeba

RNA

castellanii. J. Biol. Chem. 254 (1979)

S. and Paule, M.R.: DNA-dependent

Acanthamoeba

castellanii. Biochim.

RNA polymerase

Biophys.

II from

Acta 520 (1978) 376-

J.D., Lebovitz,

initiation mammalian Finkelstein,

R.M. and Roeder, R.G.: Accurate

by RNA polymerase

S.M., Greenblatt, general initiation

CF.,

Li, J., Chavez,

J. and Burton,

Usuda,

Y., Kubota,

ing the general transcription

D.: Cloning initiation

689-695. Hoffmann, A., Sinn, E., Yamarnoto, and Roeder,

R.G.:

Highly

with presumptive

tor (TFIID).

Nature

encoding

RAP74,

by RNA polymerase of a human

factor IIB. Nature

factor

conserved

core domain

a

gene encod352 (1991)

and unique

motifs in a human

346 (1990) 387-390.

H.: Affinity purification

HA from HeLa cell nuclear

extracts.

EMBO

10.

on purified RNA polymerase

Cell 18 (1979) 469-484. Wang, J.M., Liu, F. and Bateman,

E.: Cloning

castellanii gene encoding

mutants

II and DNA.

and expression

transcription

and functional

of the adenovirus

factor

of the TFIID.

analyses

for base-

2 major late promoter.

Nucleic

Zehring, W.A., Lee, J.M., Weeks, J.R., Jokerst, The C-terminal

repeat domain

M. N

TATA fac-

R.S. and Greenleaf,

of RNA polymerase

in vivo but is not required

for accurate

A.L.:

II largest subunit transcription

ini-

tiation in vitro. Proc. Natl. Acad. Sci. USA 85 (1988) 3698-3702. Zehring, W.A. and Greenleaf, A.L.: The C-terminal repeat domain of RNA polymerase

regulatory

Cell 69

Acids Res. 24 (1984) 9309-9321.

II. Na-

T., Wang, J., Roy, A., Horikoshi,

protein.

Weil, P.A., Luse, D.S., Segah, J. and Roeder, R.G.: Selective and accurate initiation of transcription at the Ad2 major late promoter in a

is essential

ture 355 (1992) 464-467. Ha, I., Lane, W.S. and Reinberg,

terminus

A., Berk, A.J. and Handa,

of transcription

substitution

D.P., Wang, B.Q., Fang,

Z.F.: A cDNA

factor for transcription

C., Reinberg,

between the nonphosphorylated

II and the TATA-binding

Yu, Y.-T. and Manley, J.L.: Generation

from isolated

nuclei. Nucleic Acids Res. 11 (1983) 1475-1489.

A., Kostrub,

A., Lu, H.; Houbavi,

Gene 117 (1992) 91-97.

transcription

II in a soluble extract

B (II) and general

59 (1990) 711-754.

(1992) 871-881.

Acunrhamoeba

392. Dignam,

Rev. Biochem.

E., Goldring,

soluble system dependent

4085-4091. Detke,

Usheva, A., Maldonado,

J. 10 (1991) 2305-23

polymerase

A.: RNA polymerase

Annu.

form of RNA polymerase J.P. and Weinmann,

7423-7433. D’Alessio,

factors.

D. and Aloni, Y.: Specific interaction

transcription

334 (1988) 37-42.

for transcription

M. and Sentenac,

transcription

M.F., Lee, R.F., Merryweather,

necessary

during ribosomal

P.A. and Guarente,

element-binding

system. Nature Concino,

occlusion

Cell 54 (1988) 985-992.

II is not required

for transcription

factor

Spl to

function in vitro. J. Biol. Chem. 265 (1990) 8351-8353. Zwick, M.G., Imboden, M.A. and Paule, M.R.: Specific transcription an Acanfhnmoeba extracts.

castellanii 5S RNA

gene in homologous

Nucleic Acids Res. 19 (1991) 1681-1686.

of

nuclear