Proc. Nat. Acad. Sci. USA
Vol. 71, No. 11, pp. 4425-4428, November 1974

A Novel Form of RNA Polymerase from Escherichia coli

(M13/¢X174/rifampicin)

WILLIAM WICKNER AND ARTHUR KORNBERG

Department of Biochemistry, Stanford University School of Medicine, Stanford, California 94305

Contributed by Arthur Kornberg, September 6, 1974

ABSTRACT A new form of RNA polymerase, termed
RNA polymerase II, has been recognized as a large frac-
tion of the rifampicin-sensitive enzyme in E. coli. It is
physically separable from RNA polymerase (holoenzyme,
RNA polymerase I) by gel filtration and is distinguished
by its capacity to discriminate between M13 and ¢X174
viral DNA templates in priming DNA synthesis. This tem-
plate specificity is manifested only with saturating levels
of DNA unwinding protein and characterizes the priming
of DNA synthesis on viral single strands in cell-free ex-
tracts and in vivo. RNA polymerase IIE has less than 5%
of the specific activity of RNA polymerase I in transcribing
duplex DNA of phages \ and T4, salmon sperm DNA, and
the copolymer poly[d(A-T)}. Rifampicin inactivation of
RNA polymerase II releases a factor, presumably a small
subunit, which can be isolated and used to confer on
RNA polymerase I the properties of III, namely, dis-
crimination between M13 and ¢X174 templates in priming
DNA synthesis, and a relative inability to transcribe duplex
DNA.

The observation that rifampicin, a specific inhibitor of the
B-subunit of bacterial RNA polymerase, blocks conversion of
M13 single-stranded circular viral DNA (SS) to the duplex
replicative form (RF) in vivo first led us to postulate a role for
RNA priming in DNA synthesis (1). Silverstein and Billen
(2) found that conversion of ¢X174 SS to RF was not sensi-
tive to rifampicin under similar conditions. Soluble extracts
of gently lysed Escherichia coli, like the intact cells, were
rifampicin-sensitive in the conversion of SS to RF of M13,
but not of ¢X174 (3). Further studies (4, 5, 20) have shown
that replication of ¢X174 SS is primed by a multienzyme
RNA synthetic system using many of the same factors re-
quired for host chromosome replication. Recently, reconstitu-
tion of M13 SS to RF has been achieved (6) with purified
DNA polymerase III holoenzyme (7), DNA unwinding pro-
tein (8, 9), and RNA polymerase I (holoenzyme). These
enzymes resemble cells and crude extracts in producing M13
RF with a full-length linear complementary strand and a
unique gap (3, 6), but they have the nonphysiological prop-
erty of catalyzing rifampicin-sensitive conversion of ¢X SS
to RF.

A search for the enzymatic basis of M13 DNA-specific
priming has led us to purify a new form of RNA polymerase,
termed RNA polymerase III, the subject of this communica-
tion. This enzyme is assayed by its ability, in the presence
of DNA unwinding protein, to prime DNA polymerase III
holoenzyme replication of M13, but not ¢X, DNA. RNA poly-
merase III, purified approximately 100-fold by ammonium
sulfate fractionation and gel filtration, is estimated to be
50% pure by sodium dodecyl] sulfate—acrylamide gel analysis.

 

Abbreviations: SS, single-stranded circular viral DNA; RF,
double-stranded, circular viral DNA; ¢X, bacteriophage ¢X174.

4425

It is separated from RNA polymerase I during gel filtration.
In addition to its selectivity in priming DNA synthesis, it is
distinguished from RNA polymerases I and IT by the remark-
able property of being inactive in transcription of duplex
DNA under RNA polymerase I assay conditions. Upon ex-
posure to rifampicin, a small factor is released from RNA poly-
merase III which renders RNA polymerase I (holoenzyme)
both template-specific in priming viral DNA synthesis and in-
active in transcribing duplex DNA. These studies raise
questions about the physiological role of RNA polymerase IIT,
its structure and interconversion with RNA polymerase I,
and the role of each in the RNA synthetic events of the cell.

MATERIALS AND METHODS

Materials were from previously described sources (7). E.
coli B, harvested */, through the logarithmic phase of growth
on rich medium, was purchased from Grain Processing Corp.,
Muscatine, Iowa.

Enzymes. DNA polymerase III holoenzyme was prepared
as described (7). DNA unwinding protein was prepared ac-
cording to Weiner et al. (9). We thank Dr. Michael Chamber-
lin for gifts of RNA polymerase I holoenzyme and core (10)
as well as for rifampicin-resistant holoenzyme. Unless other-
wise noted, RNA polymerase I was prepared by a minor
modification of the method of Babinet (11).

Templates. M13 and ¢X174 SS were prepared as described
previously (12). Salmon sperm DNA was purchased from
Sigma. Phage T4 and \ DNAs were the generous gifts of Dr.
J. Thorner and Dr. J. M. Syvanen of this department.

Buffers. Buffer I contained 20% glycerol, 0.05 M Tris: HCl
(pH 7.5), 1 mM dithiothreitol, 10 mM MgCh, 1 mM EDTA,
and 0.15 M ammonium acetate. Buffer II contained 30%
sucrose, 0.02 M Tris-HCl (pH 7.5), 1 mM dithiothreitol, 2
mM MgCh, 0.1 mM EDTA, and 0.25 M ammonium acetate.
Buffer III was the same as buffer II but with only 0.04 M
ammonium acetate.

Assays of RNA Polymerase. DNA synthesis on M13 or
@X SS was measured in a final volume of 25 ul containing:
5 ul of assay buffer (10% sucrose, 0.05 M Tris-HCl (pH 7.5),
20 mM dithiothreitol, 0.05 M NaCl, 0.2 mg/ml of bovine
serum albumin), 3 ul of deoxynucleoside triphosphate mix-
ture [150 uM [a-**P]dCTP (400 cpm/pmole); 400 uM each
of dATP, dGTP, and dTTP; 40 mM MgCh], 1 ul of ribo-
nucleoside triphosphates (25 mM ATP, 5 mM each of GTP,
CTP, and UTP), 2.5 ul of DNA (500 pmoles of nucleotide),
3.5 ul of water, and 5 yl of DNA unwinding protein (1.8 yg).
These components (20.0 ul) were mixed and incubated for 2
min at 30°. Water, RNA polymerase, and DNA polymerase
4426 Biochemistry: Wickner and Kornberg

TaBLe 1. Purification of RNA polymerase III

 

Specific

Total activity,

Total protein, units/mg

Fraction units mg of protein

I. Extract 90 11,000 0.008

II. Ammonium sulfate I 74 595 0,12
III. Ammonium sulfate II 30 90 0.33
IV. Gel filtration I 27 60 0.45
V. Gel filtration IT 12 17 0.71

 

ILI holoenzyme (0.05 unit) were then added to chilled tubes
to a final volume of 25 yl. After a 10-min incubation at 30°,
acid-insoluble nucleotide was determined (13). One unit of
RNA polymerase III primes 1 nmole of deoxynucleotide in-
corporation into M13 RF per min at 30°.

RNA synthesis was measured in a volume of 50 yl con-
taining: 0.05 M Tris-HCl (pH 7.5), 0.4 mg/ml of bovine
serum albumin, 4 mM 2-mercaptoethanol, 15 mM MgCh, 1
mM [?H]JATP (10? cpm/pmole), 1 mM each of GTP, CTP,
and UTP, and 25 ug/ml of DNA template. Acid-insoluble
nucleotide was measured after 10 min at 30°.

RESULTS

Purification of RNA Polymerase ITI. All operations were at
04°. A summary of the purification is in Table 1.

Preparation of Extract. Two hundred grams of frozen E.
colt B cell paste (see Materials and Methods) was broken into
small pieces and placed in a 1-liter Waring Blendor. Lysis
buffer [10% sucrose, 0.05 M Tris- HC! (pH 7.5), 0.2 M am-
monium sulfate, and 10 mM spermidine-HC!] was added to
the suspension to a volume of one liter and the cells were
suspended with three full-power bursts. Lysozyme (100 mg)
was added and the suspension was immediately transferred
to five 250-ml plastic bottles placed on ice. After 30 min the
bottles were transferred to a 37° bath for 5 min and centri-
fuged at 0° for 60 min at 12,000 rpm in the Sorvall GSA rotor.
The clear amber supernatant is Fraction I (755 ml).

Ammonium Sulfate I. Solid ammonium sulfate (0.3 g/ml
of Fraction I) was added with rapid stirring. After 20 min, the
suspension was centrifuged for 10 min at 12,000 rpm in the
GSA rotor. Pellets were successively suspended in one-fifth
the volume of Fraction I in buffer I plus: (a) 0.30 g (d) 0.24 g
(c) 0.22 g (d) 0.20 g (e) 0.18 g and (f) 0.16 g of ammonium sul-
fate per ml. Suspensions were performed with glass~Tefion
homogenizers and were followed at once by centrifugation for
10 min at 12,000 rpm in the GSA rotor. The supernatant
from each centrifugation was mixed with an equal volume
of saturated, neutralized ammonium sulfate and, after 10
min, centrifuged as before. Each precipitate was dissolved
in 10 mi of buffer II and fractions of high specific activity
(usually d-f) were pooled: Fraction IT (44 ml).

Ammonium Sulfate 1. Fraction II was mixed with an equal
volume of saturated ammonium sulfate and centrifuged after
10 min. The centrifugation here and subsequently was for 5
min at 16,000 rpm in the Sorvall SS34 rotor. Pellets were
successively suspended in one-fifth the volume of Fraction I
in buffer II plus: (a) 0.24 g (0) 0.22 g (c) 0.20 g (d) 0.18 g and
(e) 0.16 g of ammonium sulfate per ml. After each centrifuga-

Proc. Nat. Acad. Sci. USA 71 (1974)

 

Priming of
MI3 ONA synthesis

  
  

200,

   
 
   

cRNA synthesis on
ig 74 DNA
“48> Priming of gx
\\ DNA synthesis
\

  

Acid-insoluble nucleotide (pmoles)

 

 

Fraction

Fic. 1. Gel filtration. Fraction III enzyme was applied to a
Bio-Gel A-5m column (240 ml, see Results) and 6-ml fractions were
collected. Fractions 25 to 33 were used for further purification of
RNA polymerase ITI.

tion, the supernatants were precipitated with saturated am-
monium sulfate as before. Each precipitate was dissolved in
5 ml of buffer III; fractions of high specific activity (usually
cand d) were pooled: Fraction III (12 ml).

Gel Filtration I. Fraction III was applied to a Bio-Gel A-5m
column (100-200 mesh, 240 ml) equilibrated with buffer IT.
Fractions that primed DNA synthesis on M13, but not ¢X-
174, SS (Fig. 1) were pooled: Fraction IV (60 ml).

Gel Filtration II. Fraction IV was concentrated by addition
of 0.3 g of solid ammonium sulfate per ml. After 30 min, the
suspension was centrifuged and the precipitate was dissolved
with 1.5 ml of buffer III (1.5 ml) and applied to a Bio-Gel
A-5m column (100-200 mesh, 60 ml) equilibrated with buffer
III. Fractions of peak RNA polymerase III specific activity
were pooled: Fraction V (14 ml). This fraction was stable for
at least 2 weeks at 0°.

Requirements for Specific Priming of M13 DNA Replicatwn.
Discrimination by RNA polymerase between M13 and ¢X
DNA for priming DNA synthesis requires two factors: (1)
the novel form of RNA polymerase, designated RNA poly-
merase III to distinguish it from the classic holoenzyme,
called RNA polymerase I, and (ii) DNA unwinding protein
in an amount sufficient to coat the single-stranded DNA
(Table 2). The unwinding protein markedly stimulates the
RNA polymerase I-primed reaction (6) but does so for both
M13 and ¢X DNA replication; the effect on the RNA poly-
merase III-primed reaction is to suppress specifically the
o@X DNA replication.

RNA polymerase III has feeble transcriptional activity. RNA
polymerase III was 15- to 70-fold less active than RNA poly-
merase I in transcription of standard templates, such as 4,
T4, and salmon sperm DNA, and poly[d(A-T)] (Table 3).
When the two polymerases were present in the same incuba-
tion, the RNA polymerase I remained active, thus indicating
the absence of a diffusible inhibitor to account for the in-
activity of RNA polymerase III.
Proc. Nat. Acad. Sci. USA 71 (1974)

TABLE 2. Requirements for specific priming of M13

Template for

 

DNA synthesis
st
Form of __(pmoles)
RNA polymerase Unwinding protein M13 ox
I - 13 15
iit = 225 107
I + 232 122
Ill + 188 6

 

DNA synthesis on M13 or ¢X SS was assayed with 0.01 unit
of RNA polymerase III or RNA polymerase I (Materials and
Methods). DNA unwinding protein was added, where indicated.

The properties of RNA polymerase I observed with the
preparation obtained by the Babinet procedure were also
observed with another preparation prepared by M. Cham-
berlin. RNA polymerase I holoenzyme primed M13 and ¢X
replication equally well, while the core enzyme was totally
inactive (Table 4), in contrast to the behavior of RNA poly-
merase III, which primed M13, but not $X, replication and
was inactive in transcription of poly {d(A-T) ]. Rifampicin at 5
ug/ml of completely inhibited RNA polymerase III, as it is
known to do to RNA polymerase I.

Rifampicin Releases a Diffusible Faclor from RNA Poly-
merase III. RNA polymerase III treated with rifampicin and
then precipitated with ammonium sulfate to separate pro-
teins from free, unbound rifampicin was inactive in priming
DNA, but when mixed with RNA polymerase I prevented it
from priming ¢X SS while still permitting it to prime M13.
This result suggested that rifampicin released a factor from
RNA polymerase III that could confer the M13-specificity
property on RNA polymerase I. This diffusible factor, re-
leased by rifampicin, was isolated by Sephadex G-150 gel
filtration (Fig. 2), using as an assay its inhibition of the prim-
ing of ¢X DNA synthesis by RNA polymerase I. Little or
none of this factor activity was seen on gel filtration in a con-
trol experiment with RNA polymerase II] that had not been
exposed to rifampicin. As judged by gel filtration, the fac-
tor is smaller than hemoglobin, but its size and other prop-
erties remain to be determined.

The rifampicin-induced release of a diffusible factor was
also studied with the aid of a rifampicin-resistant RNA poly-
merase I (I-rif®) (Table 5). When I-rif® was mixed with RNA
polymerase III in the absence of rifampicin (Exps. 1, 3, and
5), RNA-primed DNA synthesis and transcription showed ap-

TaBLe 3. Transcription of duplex DNA

 

 

RNA synthesis (pmol) on: DNA synthesis
Form of RNA Salmon poly- __{pmol) on:
polymerase sperm T4 a {d(A-T)] M13 oxX
I 62 84 146 132 56 41
Tl 4 3 2 8 49 3
I+HI 63 50 138

 

As described in Materials and Methods, 0.005 unit (defined for
priming of DNA synthesis) of RNA polymerase I (1 ug) or RNA
polymerase ITI (2 ug) was added. In some assays, 0.005 unit of
each polymerase was mixed and added together to the assay.

RNA Polymerase IIT 4427

Taste 4. Comparison of forms of RNA polymerase I and LIT
in priming of DNA synthesis and in transcription

 

 

DNA
Gocl) en: RNA synthesis
Form of — (pmol) on
RNA polymerase M13 oX poly [d(A-T)]
T-holoenzyme (Chamberlin*) 12 10 13
I-holoenzyme [Babinet (11)] 38 36 83
I-core (Chamberlin*) 0 0 10
It 66 5 2
It + rifampicin 0 0 0

 

Holoenzyme and core RNA polymerase I, RNA polymerase I
prepared by the method of Babinet (11), and Fraction IV RNA
polymerase IIT (0.007 unit) were assayed as described in Ma-
tertals and Methods. Where indicated, the enzyme was incubated
for 5 min at 0° with 5 ug/ml of rifampicin before addition to the
assay.

* Holoenzyme RNA polymerase I has been purified by an un-
published procedure employing phosphocellulose chromatography
(Prof. M. Chamberlin and Dr. N. Gonzalez, University of Cali-
fornia, Berkeley).

proximately additive results. In contrast, when rifampicin-
treated RNA polymerase III was mixed with RNA poly-
merase I (Exps. 2, 4, and 6), the priming of ¢X replication
was depressed whereas that of M13 was stimulated; RNA

 

 

 

 

Bo} void
volume
| ; SAMPLE A:
hemoglobin rifampicin-
treated
60+
c
2
3
2
x °F
20+ SAMPLE B:
control
P~-&
7 ‘
7 N.
7 .
ae ‘\
- ae") 1 4 >
6 8 10 2 14 16 18 20

Fraction

Fic. 2. Isolation of a factor released from RNA polymerase
III by rifampicin. Samples of RNA polymerase III (100 ug in
0.1 ml buffer IIT) were incubated for 5 min at 0° (A) with rifam-
picin (2 wg) or (B) without the drug. Each sample was then
mixed with 100 ul of saturated ammonium sulfate. After 10 min
at 0°, the precipitates were collected by centrifugation (0°, 5
min, 16,000 rpm in the Sorvall SE12 rotor). Each precipitate
was suspended twice in 3 ml of a mixture of equal volumes of
buffer I and saturated ammonium sulfate and collected by
centrifugation (as above). Rifampicin-treated (sample A) and
untreated RNA polymerase III (sample B), dissolved in 100 yl
buffer I were each filtered over a Sephadex G-150 column (0.6 X
8 em, equilibrated with buffer I at 4°). Aliquots (3 ul) of the
100-yl fractions were assayed for their ability to inhibit RNA
polymerase I-primed ¢X SS -+ RF. Incorporation of 20 pmoles

- of deoxynucleotide into the acid-insoluble fraction was observed

in the uninhibited reaction.
4428 Biochemistry: Wickner and Kornberg

TaBLe 5. Rifampicin release of a diffusible factor from RNA

 

 

polymerase ITT
DNA
synthesis RNA

Form of (pmol) on: synthesis

Experi- RNA eS (pmol) on
ment polymerase Rifarnpicin M13 4X T4
1 iil _ 55 5 0
2 WI + 0 0 0
3 I-rif® - ll 8 ll
4 I-rif® + 8 11 10
5 I-rif® + IIT - 96 17 9
6 Frif® + HI + 36 2 2

 

Rifampicin-resistant RNA polymerase I (I-rif®, 1 ug) and
RNA polymerase III (2 yg, rifampicin-sensitive) were assayed
as described in Materials and Methods. Where indicated, the
enzymes were incubated for 5 min at 0° with 2 ug of rifampicin
per ml.

synthesis on a T4 duplex DNA template was likewise de-
pressed upon mixing of the two rifampicin-treated enzymes.
In sum, rifampicin, known to act on the § subunit, appears
to release a small factor from RNA polymerase ITI, which can
cause RNA polymerase I to discriminate between M13 and
¢X SS and to become relatively inert in transcription of
duplex DNA.

DISCUSSION

The capacity of cell extracts to discriminate in the replication
of M13 and $X single-stranded DNA in a rifampicin-sensitive
reaction was not maintained in a purified RNA polymerase-
multienzyme system. Our efforts to restore discriminatory
ability to the purified system have led us to a new form of
RNA polymerase, termed RNA polymerase III. This enzyme
is physically separable from RNA polymerase I (Fig. 1) and
is further recognized by (1) its selective ability to prime the
replication of M13, but not ¢X DNA, and (#2) its inability to
transcribe standard duplex DNA. The close relation of RNA
polymerase III to RNA polymerase I is shown by its sensi-
tivity to rifampicin. Upon exposure to rifampicin, RNA poly-
merase III releases a factor, presumably a low-molecular-
weight subunit, that inhibits RNA polymerase I transcription
and confers the ability to prime M13 but not ¢X replication.
This factor may be the major difference between RNA poly-
merases I and III.

Modification of RNA polymerase functions by addition or
substitution of subunits has been shown in phage T3 and T4
infections of E. coli (14, 15) and in SP82 infection of B. subtilis
(16). Chao and Speyer (17) found that EB. coli RNA poly-
merase is changed in chromatographic properties and tem-
plate specificity when cells enter a stationary phase. How-
ever, this form of RNA polymerase is not separated from
RNA polymerase I by gel filtration and, while inactive in
transcription of natural duplex DNA, it is still quite active
on poly [d(A-T)}. In contrast, RNA polymerase II] has been
obtained from cells growing exponentially, is separable from

Proc. Nat. Acad. Sct. USA 71 (1974)

RNA polymerase I by gel filtration, and is unable to tran-
scribe poly (d(A-T)]. Ishihama and coworkers (18) identified
yet another form of the enzyme, which they called RNA
polymerase II. It contains a different « subunit with a molecu-
lar weight of 56,000 rather than 90,000. Yet RNA poly-
merases I and II were not distinguishable in their capacities
to transcribe duplex DNA.

The ability of RNA polymerase III to maintain the cellular
discrimination between M13 and ¢X and its abundance in
extracts of gently lysed cells suggest that RNA polymerase
Til may constitute a significant fraction of the cellular RNA
polymerase. The complete inability of cell extracts to prime
¢X replication with rifampicin-sensitive enzymes points to a
predominance of RNA polymerase III over RNA polymerase
I or else to a specific repressor of the latter in priming the
replication of ¢X single-stranded DNA (19). These studies
raise questions that can be answered only by purification of
RNA polymerase III to homogeneity so that the factor
releasable by rifampicin can be characterized and its functions
in regulating transcriptional activity properly assessed.

This work was supported by grants from the National Insti-
tutes of Health and the National Science Foundation. W.W. is a
Fellow of the Mellon Foundation.

1. Brutlag, D., Schekman, R. & Kornberg, A. (1971) Proc.

Nat. Acad. Sci. USA 68, 2826-2829.

Silverstein, S. & Billen, D. (1971) Biochim. Biophys. Acta

247, 383-390.

Wickner, W., Brutlag, D., Schekman, R. & Kornberg, A.

(1972) Proc. Nat. Acad. Sci. USA 69, 965-969.

Schekman, R., Wickner, W., Westergaard, O., Brutlag, D.,

Geider, K., Bertsch, L. L. & Kornberg, A. (1972) Proc. Nat.

Acad. Sci, USA 69, 2691-2695.

Wickner, R., Wright, M., Wickner, 8. & Hurwitz, J. (1972)

Proc. Nat. Acad. Sci. USA 69, 3233-3237.

Geider, K. & Kornberg, A. (1974) J. Biol. Chem. 249, 3999-

4005. :

Wickner, W. & Kornberg, A. (1974) J. Biol. Chem., in press.

Sigal, N., Delius, H., Kornberg, T., Gefter, M. & Alberts, B.

(1972) Proc. Nat. Acad. Sci. USA 69, 3537-3541.

Weiner, J. A., Bertsch, L. L. & Kornberg, A. (1974) J. Biol.

Chem., in press.

10. Berg, D., Barrett, K. & Chamberlin, M. (1971) in Methods
in Enzymology, eds. Grossman, L. & Moldave, K. (Academic
Press, New York), Vol. 21, pp. 506-519.

11. Babinet, C. (1967) Biochem. Biophys. Res. Commun. 26,
639-644.

12. Wickner, W., Schekman, Rt., Geider, K. & Kornberg, A.
(1973) Proc. Nat. Acad. Sci. USA 70, 1764-1767.

13. Jovin, T. M., Englund, P. T. & Kornberg, A. (1969) J. Biol.
Chem. 244, 2996-3008.

14. Dharmgrongartama, B., Mahadik, S. P. & Srinivasan, P. R.
(1973) Proc. Nat. Acad. Sci. USA 70, 2845-2849.

15. Horvitz, H. R. (1973) Nature New Biol. 244, 137-140.

16. Spiegelman, G. B. & Whiteley, H. R. (1974) J. Biol. Chem.
249, 1476-1482.

17. Chao, L. & Speyer, J. F. (1973) Biochem. Biophys. Res.
Commun. 51, 399-405.

18. Fukuda, R., Iwakura, Y. & Ishihama, A. (1974) J. Mol.
Biol. 83, 353-367.

19. Hurwitz, J., Wickner, S. & Wright, M. (1973) Biochem.
Biophys. Res. Comm. 51, 257-267.

20. Schekman, R., Weiner, A. & Kornberg, A. (1974) Science,
in press.

pe Ww ON

an

or &

so