Tre JouRNAL OF BIoLoGicaL CHEMISTRY
Vol. 243, No. 5, Issue of March 10, pp. 913-£22, 1968
Printed in U.S.A.

The Processive Degradation of Individual

Polyribonucleotide Chains

I. ESCHERICHIA COLI RIBONUCLEASE II*

Nancy G. Nossal AND MAXINE I, SINGER

(Received for publication, September 11, 1967)

From the National Institute of Arthritis and Metabolic Diseases, National Institutes of Health, Bethesda, M ary-

land 20014

SUMMARY

The experiments reported here suggest that Escherichia
coli ribonuclease II, an exonuclease, hydrolyzes a given poly-
ribonucleotide chain to completion before releasing a small,
resistant oligonucleotide and initiating hydrolysis of another
chain. Investigation of the products of hydrolysis of oligo-
and polyribonucleotides indicates that enzymatic degradation
starts at that end of the chain bearing a 3’-hydroxyl group.
The enzyme shows a strong preference for long polymers as
substrates, although oligonucleotides are hydrolyzed if pres-
ent at sufficiently high concentration.

 

Escherichia coli ribonuclease I] catalyzes the degradation of
single stranded polyribonucleotides to nucleoside 5’-monophos-
phates (1-3). The enzyme requires magnesium and in addition
is activated by potassium or ammonium ions. Helical forms of
RNA are not hydrolyzed. The enzyme has been characterized
as an exonuclease, since mononucleotides are the principal prod-
ucts during all stages of digestion, and large oligonucleotides do
not accumulate (8). The evidence presented in this paper in-
dicates that the enzyme catalyzes the hydrolysis of a polyribo-
nucleotide chain beginning at the 3’-hydroxy] end, and that it
sequentially liberates mononucleotides from the chain until a
small oligonucleotide (n = 2 to 4), which is itself comparatively
resistant to hydrolysis, is formed. Furthermore, the results
suggest that a given enzyme molecule tends to hydrolyze a single
RNA chain repeatedly until its 5’-hydroxvl chain end is released
as a small oligonucleotide, and only then begins the hydrolysis of
anew RNA chain. Shis mode of attack will be termed pro-
cessive degradation!

* This paper is dedicated to Professor Heinrich Brinkmann of
Swarthmore College on the occasion of his 70th birthday.

1 A preliminary report of these findings has been made (Nossa,
N. G., Toupert, G. P. C., anp Sincer, M. F., Fed. Proc., 36, 612
(1966)).

EXPERIMENTAL PROCEDURE

4C-UDP, “C-ADP, “C-GDP, *H-CDP, and unlabeled mono-
nucleotides were purehased.from Schwarz BioResearch. pAp
was the gift of Dr. H. G. Khorana. Crystalline BSA,? obtained
from Pentex, was dialyzed for several days against 1 mu EDTA
and then against H.0. Radioactive phenylalanine and serine
were purchased from New England Nuclear, and “C-valine was
obtained from Calbiochem. Sephadex G-100 was obtained from
Pharmacia. Polyacrylamide gels (Bio-Gel) were products of
Bio-Rad Laboratories.

Paper Chromatography and Paper Electrophoresis—Paper
chromatography on Whatman No. 1 or 3MM paper was carried
out with 1-propanol-concentrated NH,OH-H:0 (65:10:35),
Solvent 1 (4); 1-propanol-concentrated NH,OH:H,0 (35:10:35),
Solvent 2 (5); 95% ethanol-1 m ammonium acetate (40:60),
Solvent 3 (6); saturated (NH,):SO,-2-propanol-1 m sodium
acctate (80:2:18), Solvent 4 (7); and 2-propanol-H.O (70:30)
with NH; in the vapor phase, Solvent 5 (8). All proportions are
by volume. DEAE-cellulose paper chromatography (What-
man DE-81) was carried out for 6 to 9 hours with 1 M or 0.5 M
NH,HCOs (freshly prepared). The latter molarity is preferred
for the separation of small oligonucleotides (n = 2 to 5). Paper
electrophoresis was carried out in 0.5 M potassium phosphate
buffer, pH 7.0, at 13.4 volts per em, or in 0.05 « ammonium
formate buffer, pH 3.5, at 17.8 volts percm. All the above pro-
cedures were carried out at room temperature.

Analytical Procedures—For total alkaline hydrolysis, poly-
nucleotides were incubated for 18 hours in 0.3 or 0.4 1 KOH at
37°. Solutions then either were neutralized with 0.3 11 HCl and
subjected to paper electrophoresis or were neutralized with 1 m
HCIO, and chromatographed with Solvent 5.

Aquequs samples were counted in a liquid scintillation counter

2 The abbreviations used are: BSA, bovine serum albumin; poly
A, polyadenylic acid; poly U, polyuridylic acid; poly AU, copoly-
mer of adenylate and uridylate; poly AG, copolymer of adenylate
and guanylate; 1 Aeso unit is that amount of material giving an
absorbance of 1 at 260 mg in 1.0 ml of solution in a l-cm light path.
Abbreviations for specifically labeled polyribonucleotides are
described in Table I.

913
O14

in 10 ml of Bray’s solution (9) containing 0.1 ml of 1 ~ NH,OH.
Squares cut from Whatman No. 1 or J)-81 paper chromato-
grams were counted in Bray’s solution or in a solution containing
2 ¢ of 2,5-diphenyloxazole (PPO) and 0.1 g of 1,4-bis[2-(5-
phenyloxazolyl)]benzene (POPOP) per 473 ml of toluene.

Protein was determined by the method of Lowry et al. (10).

Enzymes— RNase H1 was a hydroxylapatite fraction (Fraction
VII) (8) purified from £. coli MRE 600 (11), a mutant lacking
RNase ] (12). The molecular weight of RNase II was estimated
to be approximately 65,000 by using the sucrose density gradient
method of Martin and Ames (13) with bovine serum albumin as
the reference marker (14). A unit of enzyme produces 1 umole
of AMP per hour at 37° in @ reaction mixture containing 2 mm
poly A,? 0.1 m Tris-HCl! (pH 7.5), 0.1 aw KCI and 1.5 mm MgCh.
The axsay was carried out as previously described (3). Oc-
casionally, 0.1 M ammonium bicarbonate buffer, pH 7.5, was used
to replace both the Tris-HCI] and the 0.1 m KC] to facilitate the
removal of salt before chromatography. NH,* ions can sub-
stitute for Kt ions with this enzyme (1).

Polynucleotide phosphorylasg was prepared from. .\icrococcus
lysodetkticus (15) or from EF. coli MRE 600 through the DEAE-
Sephadex step described by Williams and Grunberg-Manago (16),
Two different. preparations of enzyme from Jf. lysodetkticus were
used: one required a primer in order to polymerize nucleoside
diphosphates (15); the other, less purified fraction required no
primer (17). The £. coli fractions did not require primer. A
unit of enzyme is equivalent to the incorporation of 1 umole of
2P; into ADP in 15 min at 37°, with poly A as a substrate (17).

The .(zofobacter agilis endonuclease, which degrades poly A
to oligonucleotides containing phosphate monoesterified to the
terminal 5’-hydroxyl group, was the gift of Dr. Audrey Stevens
(18). Pancreatic RNase A was purchased from Sigma. F. coli
alkaline phosphatase was donated by Dr. Wallace Brockman
(ammonium sulfate fraction (see Reference 19)); this enzyme
(electrophoretically purified) was also purchased from Worthing-
ton. A unit of enzyme produces 1 pmole of P; from AMP per
hour at 37° when assayed in 0.1 x1 Tris-HIC), pH 9.0, at a concen-
tration of 0.012 m 5’-AMP.

Tl RNase was prepared by the procedure of Takahashi (20) or
was purchased from Calbiochem. A unit of enzyme produces a
change of 1 absorbance unit per ml in the following assay. Ycast
RNA (0.15 mg), 0.05 m Tris-HCl (pH 7.5), 0.002 m EDTA, and
enzyme in a volume of 0.2 ml were incubated for 15 min at 37°.
An equal volume of 0.259 uranium acetate in 497, perchloric acid
was added, and 0.1 ml of the clear supernatant fluid was diluted
to 1.5 ml prior to determining the absorbance at 260 mz.

Venom phosphodiesterase was prepared by a modification of
the method of Koerner and Sinsheimer (21) or was purchased
from Worthington; in the latter case it was freed of contaminat-
ing 5’-nucleotidase by the procedure of Keller (22). A unit pro-
duces 1 pmole of AMP per hour at 37° when incubated with 2 mar
poly A, 1 mm MgCl, and 10 mm Tris-HCl, pH 8.7. The #. coli
evclic diesterase preparation was 4 cold water wash fraction from
shocked F. colt (MRE 600) cellx, and was the gift of Dr. Leon
Heppel (23). One unit produces 1 ymole of P; per hour from
2' ,3’-cyclic UMP under the standard assay conditions (23).

A fraction containing a mixture of aminoacyl-tRNA synthe-

3 Unless noted otherwise, the concentrations of polynucleotides

and oligonucleotides are given as concentration of constituent
mononucleotides (or phosphorus).

Processive Exonucleolytic Digestion of Polyribonucleotide Chains. I

Vol. 248, No. 5

tases was prepared from EF. coli B by following the procedure of
Muench and Berg (24) through the DEAE-cellulose column step.
This preparation still contained 34 units of RNase IT per ml when
measured by the standard assay for RNase II]. The amino-
acylation of tRNA was proportional to the concentration of
tRNA under the following conditions. Reaction mixtures (0.1
ml) all contained 0.1 m sodium cacodylate buffer (pH 6.9), 4 mm
reduced glutathione, 0.01 a MgCl, 0.01 a KCl, and 20 yg of the
synthetase fraction. In addition they contained 0.5 mu 4C-
valine (5.5 X 108 cpm per umole), 0.87 mm ATP, and up to 3
Algo units of (RNA; or 0.25 mu “C-serine (5.9 < 108 cpm per
umole), 0.44 mu ATP, and up to 6 clog units of tRNA; or 0.5
mm #H-phenylalanine (9.6 x 106cpm per umole), 1.3 mw ATP,
and up to 2 Ago units of tRNA, The reaction was initiated by
the addition of the amino acid, and, after 10 min at 37°, duplicate
40-u1] aliquots were absorbed onto Whatman No. 3MM _ paper
dises and subjected to a modification (25) of the procedure of
Mans and Novelli (26) in order to determine the incorporation
of amino acid into tRNA. The dises were counted in 10 ml of
Bray’s solution. The results were corrected for any radioactivity
detected in control mixtures that lacked enzyme.

Oligonucleotides—(Ap)sA, (Ap)sA, and (Up); were obtained
from Miles Chemical Company and were characterized by
alkaline hydrolysis. The abbreviations used to describe the
specifically labeled oligonucleotides and polyribonueleotides
synthesized for this work are given in Table I,

(Ap)sU and (Ap)sC were prepared by using (Ap),A as the
primer for the polymerization of UDP and CDP, respectively,
by ./. lysodeikticus polynucleotide phosphorylase in the presence
of pancreatic RNase, which preferentially hydrolyzes the
phosphodiester bond following pyrimidine residues. Reaction
mixtures (2.0 ml), containing 0.1 m Tris-HCl (pH 8.2), 5 mm
MyCl, 2.45 mu (Ap)sA, 0.2 mg of BSA, 1.2 mg of RNase A,
0.018 m UDP or CDP, and 1.8 (UDP) or 3.0 (CDP) units of
polynucleotide phosphorylase, were incubated at 37° for 1 (UDP)
or 2} (CDP) hours. The resulting phosphorylated hexanucleo-
tides ((Ap)sUp and (Ap)sCp) were isolated by chromatography
in Solvent 3 and were eluted with H.0; 80% yields, based on the
(Ap)sA input, were obtained. The hexanucleotides were de-
phosphorylated with F. colt alkaline phosphatase as follows.
Reaction mixtures, containing 0.2 m Tris-HCl (pH 8.1), 2.5 mu
MegCle, 52 Aveo units of (Ap);Up, and 21 units of enzyme in a,
volume of 6.0 ml or 57 Age units of (Ap)sCp and 32 units of en-
zyme in 7.8 ml, were incubated for 2 hours at 37°. The solutions
were deproteinized with phenol and extracted with ether, and the
products were then isolated by chromatography with Solvent 3.
The yields of dephosphorylated hexamer and the ratios of AMP
to nucleoside after alkaline hydrolysis were: for (Ap)sU, 24%
and 4.1; for (Ap);C, 22% and 4.4. The poor yield was probably
primarily due to the partial solubility of oligonucleotides in
phenol.

(Ap)sG*p was similarly prepared (by Dr. D. M. Logan) with
polynucleotide phosphorylase but in the presence of Tl RNase
which is specifie for bonds following guanosine residues, The
reaction mixture (3.7 ml) contained 0.1 m ‘Tris-HC] (pH 8.2), 2
mm MgCh, 14.9 mm 4C-GDP (2.8 * 108 epm per pmole), 1.1
mM (Ap),A, 0.87 mg of BSA, 11 units of phosphorylase, and 750
units of Tl RNase (Calbiochem) and was incubated for 3 hours
at 37°. The product, (Ap);G*p, was separated by chromatog-
raphy as described above. The yield was greater than 90% and,
Issue of March 10, 1968

N.G. Nossal and M. F. Singer

915

TaB.y I
Abbreviations for polyribonucleotides
The preparation of these polyribonucleotides is described under ‘Experimental Procedure.” An asterisk indicates a radioactive

 

 

 

group.
Abbreviation Structure |
(pA)s* pApApApApA |
(Ap) 6A? ApApApaApApApA .
(Ap): ApApApApApApAp |
(Ap)sU ApApApApApU
(Ap)sC ApApApApApC
(Ap)sG*p ApApApApApG*p
(pA*)a(pA)a pA*pA*pA*pA. . pA
(Ap), U*? ApAp. . .ApU*
(Ap)nC* ApAp. . .ApC*
(Ap), G* ApAp. . -ApG*

|

Radioactive label and position | Average chain length

 

 

None |
None
None i
None .
None
NC; 3’-terminal Gp
MC; 5’-terminal pApApA 108
MC; 3’-terminal U 80 <n < 105
3H; 3’-terminal C : 38
MC; 3’-terminal G 56

 

* Homologous compounds are similarly abbreviated.

‘ Thirty-four per cent of the chains in this preparation were not dephosphorylated. Thus they had the structure ApAp. . .ApU*p.

after KOH hydrolysis, AMP and GMP were recovered in a ratio
of 4.2:1. The solution of (Ap)sG*p had little or no detectable
polynucleotide phosphorylase or T1 RNase activity.

4C-Labeled oligonucleotides of the general structure (pA)n,
where n = 2 to 5, were prepared by the digestion of ¥C-poly A
(6.5 * 104c¢pm per pmole of adenosine) with the .1. agilis endo-
nuclease. The procedure of Stevens and Hilmoe (18) was
followed, except that after chromatography of the oligonucleo-
tides with Solvent 1 they were further purified by chromatog-
raphy with Solvent 3. (pA)2 accounted for 16.5% of the prod-
uct, and gave a ratio, after alkaline hydrolysis, of adenosine to
2',3’-AMP to pAp of 1.00:0.0:0.91; (pA)s, 37.0%, 1.00:0.88:
0.85; (pA) 31.4%, 1.00:2.08:0.99; and (paA)s, 15.462, 1.00:
3.12:1.01.

Polymers—Unless specifically noted, all the polymers used in
the studies were synthesized in this laboratory, with the use of
polynucleotide phosphorylase from Af. lysodetkticus, and were
dialyzed extensively against } mm EDTA and then distilled
water before use. Their structures and corresponding abbre-
viations are outlined in Table I.

(pA*)3(pA)n was prepared by use of "C-pApApaA as the primer,
with a primer-requiring fraction of polynucleotide phosphorylase
from Af. lysodeikticus. The polymer was purified by repeated
extraction with phenol followed by extensive dialysis, first
against 0.04 m KCl containing 0.005 m disodium EDTA, and
then against H.O. The average chain length, determined from
the ratio of radioactivity in the terminal “C-(pA)s to the ab-
sorbance at 257 mu, was 108. The polymer was characterized
by chromatography for 117 hours with Solvent 3. Five per cent
of the radioactivity migrated with (pA)s, less then 1 per cent
migrated faster than (pA)s, and the remainder did not leave the
origin. After alkaline hydrolysis in the presence of carrier pAp,
the ratio of radioactivity in 2’,3’-AMP to that in pAp was 2.4:1,
suggesting that the 5’-terminal phosphate had been removed from
about 12% of the chains.

(Ap),U* was prepared by controlled digestion of poly AU
(labeled with “C in the uridylic acid residues) with pancreatic
RNase, followed by dephosphorylation with E. colt alkaline
phosphatase. Poly AU was synthesized with polynucleotide
phosphorylase trom #. coli. ‘The ratio of ADP to "C-UDP in
the substrate mixture was 30:1, and the ratio of mononucleotides

in the product was 58:1. The polymer was purified by phenol
extraction and dialysis. Poly AU (103.5 umoles) was incubated
for 16 hours at 37° with 150 ug of pancreatic RNase, 0.14 m KCl,
and 0.054 m Tris-HCl, pH 7.5, in a volume of 14 ml. These con-
ditions produced one break per 46 nucleotides as measured by
phosphate labile to alkaline phosphatase. Thus most of the
uridine residues and few of the adenosine residues would be ex-
pected to be in a terminal position. The digest was adjusted to
0.1 «1 HC] and incubated for 4 hours at room temperature to
cleave the cyclic phosphate esters. The pH was adjusted to 8
with KOH, and Tris-HCI, pH 8.1, was added to give a final con-
centration of 0.25 m. The polymer was then incubated for 1
hour at 37° with 325 units of alkaline phosphatase (Brockman),
a 7-fold excess over an amount which quantitatively dephos-
phorylated AMP at the same concentration as the polymer
terminal phosphate. The dephosphorylated polymer was ex-
tracted with phenol and ether to remove the RNase and alkaline
phosphatase. The polymer (23.8 umoles) was fractionated on a
column of Bio-Gel P-150 (1.3 x 100 em), covered with a Il-em
layer of Bio-Gel P-80, by elution with 1 m ammonium bicarbon-
ate. The average chain lengths of the fractions were determined
from the ratio of radioactivity (in the terminal uridine) to the
absorbance at 237 mu. Selected fractions were pooled, and the
ammonium bicarbonate was removed by repeated lyophilization.
When the pooled fractions, containing material of average chain
length between 50 and 105, were hydrolyzed with KOH, 66% of
the radioactivity was found in uridine and the remainder in
UMP, indicating incomplete dephosphorylation of the polymer.
After incubation with a 10-fold greater concentration of alkaline
phosphatase under the above conditions, more than 96% of the
radioactivity was in uridine after alkaline hydrolysis.

(Ap),aC* was similarly prepared with polynucleotide phos-
phorylase (Af. lysodeikticus) with 400 umoles of ADP and 20
umoles of 7H-CDP. After incubation for 6 hours at 37° with
4 mg of pancreatic RNase and phenol extraction, polymer was
isolated by chromatography on Bio-Gel P-4 in 1 mm Tris-HCI,
pH 7.5, and lyophilized. The poly (Ap),C*p (31 zmoles) was
incubated for 30 min at 37° with 500 umoles of Tris-HCl, pH
9.0, and 60 units of alkaline phosphatase (Brockman) in a volume
of 5.0 ml. The polymer was chromatographed overnight in
Solvent 2, and the material remaining at the origin was eluted,
916

TABLE II
Hydrolysis of pipApApsApA by RNase IT

The incubation mixtures contained (per ml) 0.1 M ammonium
bicarbonate buffer (pH 7.5), 1.5 mm MgCl, 0.15 mg of BSA, and
the indicated concentrations of 4C-(pA); and enzyme. The total
volumes were: Experiment A, 0.5 ml; B, 0.2 ml; C,0.I ml. After
30 min at 37°, the reaction mixtures were heated for 5 min at 100°,
lyophilized two times to remove salt, and chromatographed for 6
hours on DEAE-eellulose paper with 0.5 1 ammonium bicarbon-
ate. The paper strips were cut into suitable sections, with the
use of known oligonucleotide markers to aid in location of prod-
ucts, and the sections were counted in the toluene scintillation
solution. The radioactivity was corrected for that obtained in

control experiments without enzyme.

 

 

 

 

 

 

Conditions Products*
Exper ae l
t | Concen- Concen- | pydrolysi ‘ : :
ee |eration of tration off PYFPAYSS | oa, | Aye} (pA). (pads! AMP
(pA)s* - enzyme !
~~ moles AMP id eee
uM untts/ml ielensome aa uM uM was ua
unit 1
A 2.1 3 0.002 0.6 0.0 1.6 ' 0.0 3.6
B 22.0 3 0.02 5.3 0.61 16.1 0.5 34.0
2.0; 1 | 0.03 | 13.4] 1.6! 66 0.6 15.4
| .
C 220 0.08 149 15.4' 54.0 0.0 , 127
220) | 1 0.06 205 4.0 10.0 0.0 27

 

 

 

* Concentrations are expressed as oligonucleotide.

Tas_e IIT
Hydrolysis of oligonucleotides by RNase IT
Reaction mixtures (0.2 ml) containing radioactive oligonucleo-
tide, enzyme, 10 ng of BSA, 0.1 wm NH AHCOs buffer (pH 7.5), and
1.5 mM MgCl. were incubated for 7 hours at 37° and then treated
as described in Table I.

 

 

 

 

 

Oligonucleotide substrate | Substrate undegraded
Compound Concentration? 0.2 RNase unit 3 RNase units

"mat Ge a

(pA)2 1 99 98

(paA)s 0.67 98 62

(pA), 0.50 41 4

(pA)s 0.40 45 3

« Concentrations are expressed as oligonucleotide.

lyophilized, and rechromatographed for 48 hours in the same
solvent, The polymer eluted from the origin contained 10.1
umoles of nucleotide and 1.6 105 epm, giving an A:C ratio
of 38:1. After alkaline hydrolysis, 975¢ of the radioactivity
was in cytidine and the remainder in CMP.

(Ap),G* was prepared from poly AG which had been made
with 200 wmoles of ADP and 10 uwmoles of MC-GDP. The
polymer was extracted with phenol, chromatographed on Bio-
Gel P-4 to remove mononueleotides, concentrated by lvophiliza-
tion, and dialyzed. ‘Tl RNase (4500 units) prepared by the
procedure of Takahashi (20) was incubated with 52.4 ymoles of
polymer in 9.9 ml for 21 hours at 37°. Toluene was acded after
4hours. The (Ap),Gp was extracted with phenol and chroma-
tographed on Bio-Gel P-4. The polymer (16.9 ymoles) was

Processive Exonucleolytic Digestion of Polyribonucleotide Chains. I

Vol. 243, No. 5

dephosphorylated by incubation with 4.8 units of EZ. cold cyclic
diesterase for 45 min at 37° with 0.1 ma CaCh, and 0.2 m sodium
acetate buffer, pH 6.1, in a volume of 4.0 ml. The solution was
heated for 10 min at 70°, chromatographed on Bio-Gel P-4, and
concentrated. The average chain length, calculated from the
ratio of radigactivity to total phosphate, was 56. More than
99°; of the radioactivity was found in guanosine after alkaline
hydrolysis, and 18°, of the chains were found to terminate
with adenosine.

E. coli B tRNA was prepared and stripped of amino acid by
the method of Zubay (27), or was purchased from Schwarz
BioResearch,

RESULTS

Experiments with Oligonucleotides

Degradation of Oligonucleotides—In a previous publication
from this laboratory (8), the results of experiments utilizing a
series of oligonucleotides of general structure (pA), as substrates
for RNase J] were described, Those data were interpreted to
indicate that (pA), would be degraded only if n were 7 or greater.
This interpretation was consistent with the fact (1, 3) that
oligonucleotides accumulate as resistant end products of the
digestion of long chain polyribonucleotides. Recently, the
hydrolysis of oligonucleotides was reinvestigated asx a function
of oligonucleotide concentration, and it became clear that
oligonucleotides of chain length less than 7 are indeed hydrolyzed
if their concentration in the reaction mixture is sufficiently high.
Table II presents data pertaining to the hydrolysis of (pA)s.
Ata concentration of 0.22 mm (expressed as chains of (pA)s), the
rate of AMP production is 0.06 that of the rate of poly A hy-
drolysis when the latter is measured at saturating concentration
(1.2 uM, expressed as polymer chains; average n = 1000). For
poly A, the half-maximal rate of hydrolysis was obtained at a
concentration of 0.012 wat (as chains), while the data in Table IT
suggest that. the concentration of (pA)s required for half-max-
imal rate is at least three orders of magnitude greater. Table
III shows that (pA), is hydrolyzed at about the same rate as
(pA)s, but that (pA)3 and (pA): are more resistant to hydrolysis.
In the experiments in Table ITI, all the oligonucleotides were
tested at similar concentrations, and the concentration of (pA)s
was close to the optimal value. Under the conditions of these
experiments (pA), was hydrolyzed to a very small extent.

The relative preference of RNase IL for long polymers, sug-
gested by the experiments just described, is shown directly by
the data in Table LV. The effect of mixing approximately equal
concentrations (as numbers of chains, not nucleotide residues)
of poly A and 4C-(p.\)5 on the independent rate of hydrolysis of
each compound was investigated. Poly A was present at
saturating concentration; the oligonucleotide was not. The
molar concentration of enzyme was much lower than the coneen-
tration of either substrate. The pentanuclcotide had little or
no effect on the hydrolysis of poly A, but the presence of poly A
inhibited the digestion of (pA). over 90S¢.

Tables IL and IV also contain data concerning the products of
degradation of (pA)s. Free (pA), does not accumulate as an
intermediate during the hydrolysis of (pA)s. In fact, at all
levels of digestion studied, the yield of (pA)s is higher than that
of (pA)s. It is of interest that, during the digestion of (Ap)«A
and (Ap); (Table V), more (Ap)sA than (Ap),A is produced,
These findings will be considered under “Discussion.”
Issue of March 10, 1968

Direction of Hydrolysis of Oligonucleotides by RNase I[—
Investigation of the products of digestion of oligonucleotides can
establish the direction of hydrolysis by an exonuclease (28-30).
The method is diagrammed in Fig. 1, with (Ap)«A as an example.
Since the products of RNase II digestion are 5’-mononucleotides,
adenosine will appear as a product only if hydrolysis starts at
the 5’-hydroxyl end. The products found upon digestion of
(Ap)sA are shown in Table V. No adenosine was detected; the
only products were 5‘-AMP and oligonucleotides of chain length
less than 6. Thus, the direction of hydrolysis ix 3’ to 5’. Ina
similar experiment, (Ap)3\ was degraded to 5’-AMP and resist-
ant oligonucleotide; no adenosine was detected.

The degradation of oligonucleotides of the general structure
(Ap)sX, where X is cytidine or uridine (Table VI), confirms the
conclusion that hydrolysis is 3’ to 5’, since 5’-XMP is always a
product, The data in Table VI are also relevant to the possible
preference of the enzyme for specific bases. Thus, as X is
varied, no very large differences in rate of hydrolysis of -ApX
phosphodiester bonds appear. Similar results have also been
obtained with (Ap)sG.

It is of interest that (Ap);, which is phosphorylated at the
3’-terminus, is hydrolyzed by RNase IL at the same rate as the
dephosphorylated heptamer. This observation is unexpected,
for snake venom phogsphodiesterase and polynucleotide phos-
phorylase act very slowly on oligonucleotides bearing a terminal
3’-phosphate group (31). The 3’,5’-nucleoside diphosphate was
identified as a product of the hydrolysis of oligonucleotide
phosphorylated at the 3’ chain end with the phosphorylated
oligonucleotide (Ap);G*p. The oligonucleotide (0.174 mole as
oligonucleotide) was incubated with 0.75 unit of enzyme for 5
hours at 37°. When the digest was chromatographed with
Solvent 5, all of the radioactivity was either at the origin or in
the area just ahead of the origin. The latter area was eluted and
subjected to electrophoresis in 0.05 M ammonium formate buffer,
pH 3.5. The major radioactive spot, migrating just in front of
GDP, was eluted and concentrated. The spectrum of the
nucleotide agreed with that of 5’-GMP. Total phosphate
determination gave 1.9 phosphates per guanosine residue.

Tanne IV
Hydrolysis of (pA)s in presence of poly A

Duplicate reaction mixtures, containing 2.15 ma poly A or
0.135 mm “C-(pA)s (or both), 0.1 m NH.HCOs; (pH 7.5), 1.5 mm
MgCl:, 0.1 mg of BSA, and 0.4 unit of enzyme in a volume of 0.2
ml, were incubated for 45 min at 37°. The reaction was started
by addition of enzyme. Hydrolysis of poly A was determined
by the absorption at 259 mp» of the acid-soluble product. Hy-
drolysis of (pA)s was determined by chromatography of the en-
tire reaction mixture on DEAE-cellulose paper with 0.56 m
NH,HCOs, and counting of the appropriate areas of the paper in
the toluene scintillation solution. With the pentamer alone, 2.7
myumoles of (pA)s (as oligonucleotide) were hydrolyzed, yielding
(in millimicromeles of oligonucleotide) 0.5 (pA)a, 2.3 (pA)s, 0.0
(pA)2z, and 5.4 AMP.

 

 

| AMP from
Substrate bow
| Poly A | (pA)s
mymoles
Poly A... 0... ccc eee eee 56
MC-(PA)s. oe 5.4
0.6

Poly A + MC-(pA)s. 0000... 53

N.G. Nossal and M. F. Sinaer

917
From 3'—OH chein end

ApApApApApApA —» pA + Oligonucleotide

From 5'—OH chain end

ApApApApApApA —= A+ pA + Oligonucleotide

Fru. 1. Produets expected from the hydrolysis of (Ap)«A

TABLE V
Products of hydrolysis of (Ap)eA and (Ap)z by RNase IT

In Experiment 1, (Ap)sA or (Ap)7 (0.052 zmole of oligonucleo-
tide) was incubated at 37° in a volume of 0.1 ml with 0.1 mM am-
monium bicarbonate buffer (pH 7.5), 1.5 mm MgCl, 5 ug of BSA,
and 1 unit of enzyme for 135 min. Each reaction was sct up in
duplicate. The reactions were stopped by heating for 5 min at
100°, and the NH,ITCO; was removed by repeated lyophilization.
The relative quantities of the mono- and oligonucleotides were
determined by chromatography of one of each pair on DEAE
cellulose paper with 0.5 m NH,HCO; for 5 hours. The spots were
eluted with 0.4 m KOH for 18 hours at 37°, and the absorbance at
259 mp was determined after correction for an appropriate paper
blank. The other duplicate was chromatographed with Solvent
2, and the area corresponding to marker adenosine was eluted
with 0.01 m TIC]. In Experiment 2, the reaction mixtures were
the same as described for Experiment 1 except that they con-
tained 0.6 unit of enzyme and were incubated for 30 min at 37°.

 

 

 

 

 

 

 

 

Products*
Substrate r .
(Ap}sA, | .
Cap | Cand | ania | Caos | Xtee | Sine
_ | % total absorbance
Experiment ] | \ {
(Ap) 6A ' 18 § 3 28 | (Il 44 0
(Ap); | 20 | 1 | 24 | 13 | 42
Experiment 2
(Ap) «A 54: 4 13 5 25
(Ap): | 58 | 1 | 12 5 | 25

 

« Expressed as ratio of absorbance eluted from spot to total
eluted absorbance.

These oligonucleotides were not separated by this chro-
matography.

TanLe VI

Digestion of (Ap)sX
Reaction mixtures (0.1 ml), containing 3.8 Aso units of oligo-
nucleotide (580 um oligonucleotide), 5 mm Tris buffer (pH 7.8),
1.5mm MgCh, 0.1m KCI, 5yg of BSA and 0.6 unit of enzyme, were
incubated for 60 min at 37°, heated for 2 min at 100°, and chro-
matographed in Solvent 4. The products were quantitatively
cluted with 0.01 m HCI, and the concentrations were determined
from the absorption at the wave length of maximum absorbance,

corrected for an appropriate paper blank.

 

 

 

 

| tn
+ Total A . erminal
Substrate ron, cred AMP UMP or CMP ees
myumoles %
(Ap)sA 3.7 4.9 |
(Ap)sU 3.8 32.6 i 31.0 53
(Ap):C , 3.9 28.2 | 21.6 37

 
 

918
100; (Ap)n yite .
a +
porno
i
|
sol} .
i
' a
t
$
t
i
lo. A.
t
¢

 

j
OOF (Apia CoH
---~ Venom
—— PNAcse T
-
/
i
5OF /
i
f

 

PERCENT RADIOACTIVITY SOLUBILIZED

 

 

 

100+ (Ap)n G MC
_-_—“
eee eee
oo a
t
4
i
?
i
a Lo
! °
'
t
J
t 4
i .
i 4 c
f
0 T
0 50 {00

PERCENT ABSORBANCY SOLUBILIZED

Vig. 2. Hydrolysis of (Ap),U*, (Ap),C*, and (Ap),G*. A,
(Ap),U*. Reaction mixtures (0.1 ml), containing (per ml) 0.1 m
NILHCOs buffer (pH 7.5), 1.5 mm MgCh, 0.05 mg of BSA, and
either 1.02 mm polymer with 1.0 unit of RNase I] (@——@) or
0.76 mM polymer and 3 units of enzyme (O O), were ineubated
for periods from 5 to 120 min at 37°. The reaction was stopped by
adding perchloric acid. Aliquots of the supernatant solutions
were diluted with 0.05 m potassium phosphate buffer, pH 7.0, and
read at 257 my or counted in Bray's solution, Venom phospho-
diesterase reaction mixtures (0.1 ml), containing (per ml) 0.135 m
NI,WCO; buffer (pli 9.0), 1.02 mm polymer, and 0.45 unit of
enzyme, were incubated for 5 to 180 min and treated in the same
manner. 3B, (Ap)aC*. Reaction mixtures (0.1 ml) for RNase II
contained (per ml) 2.96 mm polymer, 0.1 M Tris-HCl (pli 7.5), 0.1
KCI, 1.5 mm MgCh, and 4 units of enzyme and were treated as in
A. Venom phosphodiesterase reaction mixtures had (per ml)
1.01 mm polymer, 12 mm Tris-HCl (pH 7.5), 1 mm MgCle, and 0.19
unit of enzyme. Aliquots (0.1-ml) were removed at various times ;
perchloric acid was added, and the supernatant was analyzed as
in A. C, (Ap),aG*. Reaction mixtures for RNase II contained
(per ml) 2.6 ma polymer, 0.1 m Tris-TIC! (pH 7.5), 0.1 m KCl, 1.5
mm MgCl, 0.6 mw EDTA, and 1.2 (O——O) or 1.6 (@—-®)
units of enzyme. Venom phosphodiesterase mixtures were as
described for B, except that they contained 2.6 mm polymer and
0.23 unit of enzyme. Aliquots (0.1-ml) were taken at various
times and analyzed as in A.

 

Processive Exonucleolytic Digestion of Polyribonucleotide Chains. I

Vol. 248, No. 5

Neither this nucleotide nor pAp was dephosphorylated when 2.9
mymoles were incubated with enough snake venom 5/-nucleo-
tidase (32) to dephosphorylate 280 mymoles of 5’-AMP. The
nucleotide (2.9 mumoles) was quantitatively converted to
guanosine when incubated with enough #. coli alkaline phos-
phatase to remove 500 mumoles of P; from pAp.

Experiments with Polyribonucleotides

Polymers Labeled at 3'-Hydrozyl End—By using polynucleo-
tides labeled specifically at one end of the chain, the extent of
hydrolysis of the labeled end can be compared with the extent of
hydrolysis of the molecule as a whole at any time during diges-
tion. Thus, with polymers of the general structure (Ap),X,
where X is radioactive, the degradation of the total chain can be
estimated by measuring the conversion of ultraviolet-absorbing
material to an acid-soluble form, while the hydrolysis of the 3’-
hydroxyl end is estimated by measuring the release of radio-
activity into an acid-soluble form. The results expected for an
enzyme which degrades in the 3’ to 5’ direction are exemplified by
those obtained with snake venom phosphodiesterase (Fig. 2)
(33); most of the radioactivity is released before a large per-
centage of the ultraviolet-absorbing material is solubilized.
Similar experiments with RNase IT yield markedly different
results (Fig. 2). ‘Thus, when the polymer (Ap),U* (containing
34°% of (Ap),U*p) is hydrolyzed by RNase IJ, the percentage of
ultraviolet-absorbing material and the percentage of radioac-
tivity released at all times, up to 100°, digestion, are identical.
Similar results were obtained with completely dephosphorylated
(Ap) al*, (Ap),.C*, and (Ap),G* (Fig. 2). Results such as those
shown in Fig. 2 might be expected if the enzyme hydrolyzed
bonds of the type -ApX more slowly than bonds of the type
-ApA. However, the experiments with oligonucleotides of the
type (Ap)sX (‘Table VI) show that the rate of hydrolysis of the
3’-terminal group does not depend markedly on the base in that
position. In order to eliminate possible endonuclease contamina-
tion as an explanation for these findings, the products of digestion
of (Ap),G* were studied as a function of time by chromatog-
raphy in Solvent 5 for 18 hours. At all times, radioactivity
from the 3’-terminal guanylic acid residue was found only in
mononucleotide or in polymeric material too large to migrate
from the origin. No small radioactive oligonucleotides were
detected. Thus this experiment, as well as others which are
summarized under “Discussion,” indicate that the results
described in Fig. 2 cannot be attributed to endonuclease. An
alternative interpretation of the data in Fig. 2 is that RNase I
begins hydrolysis of a polyribonucleotide at the 3’-hydroxyl end
and continues hydrolyzing that same chain until it is essentially
completely degraded (processive degradation). In contrast,
venom phosphodiesterase also begins hydrolysis by removing the
3’-hydroxyl-terminal moiety of a chain, but then releases that
chain and initiates hydrolysis of some other molecule.

Polymers Labeled at 5'-Hydroxyl End~The products of the
hydrolysis of (pA*)3(pA), (average chain length, 108) by RNase
I] were investigated as a function of time (Fig. 3). As expected
for an exonuclease that degrades in the 3’ to 5’ direction and
leaves a resistant oligonucleotide, no “C-AMP was detected until
the terminal stages of the digestion. Indeed, the entire substrate
was degraded completely to acid-soluble products before “C-
AMP was found. The data in Fig. 3 also show that, when only
30% of the total polymer was degraded, significant. amounts of
Issue of March 10, 1968

4 (236.) had accumulated in an oligonucleotide fraction con-
taining mainly di- and trinucleotides. No radioactivity was
detected between the origin and that fraction. This indicates
that those chains that were degraded at all were degraded
completely, When the radioactive oligonucleotide peaks from
the last two time points were pooled, lyophilized repeatedly to

 

 

CT T T T T T
50 tina 0 minutes ~)500°
\ gt 0 % Soluble
20- + 2000
1OL + 1000
Out { 0
30 Minutes
10 30 % Soluble 1000

 

 

 
   
 

   
     
 

é

5 OY 0 8
St0- ia 50 Minutes 41000 “
> 48 % Soluble oO
2 Z
ao 0

S A 80 Minutes

B l0r i} 75% Soluble 1000

9 Shee / \ PSs

1 on JN

+ OL ‘ 0
i 240 Minutes 7

°

100 % Soluble

i
s
3
°

425 Minutes
100 % Soluble

 

 

0 10 20 30 40
DISTANCE FROM ORIGIN (cm)

Fig. 3. Products of the hydrolysis of (pA*)3(pA),. by RNase II.
For each time point, duplicate reaction mixtures, containing 2.75
mM (pA*)3(pA)n, 0.1 m NH HCO; buffer (pH 7.5), 1.5 mu MgCh,
0.05 mg of BSA, and 3 units of enzyme in a total volume of 1 ml,
were incubated for the indicated times. One reaction mixture
was used to measure the distribution of radioactivity, and the
other for the distribution of ultraviolet absorbance. Aliquots
(50-41) were removed to determine the percentage of material
absorbing at 259 mu which was acid-soluble; the remainder was
lyophilized and chromatographed on DEAE-cellulose paper for
6} hours with 1 m NH,HCO;. Under these conditions (pA)s just
leaves the origin. The position of AMP is shown in the figure.
Each chromatogram was divided into 1.8-cm squares. One of
each duplicate series from a given time point and a blank series
for each paper were counted in the toluene scintillation solution.
The second series from each time point and a series of blanks from
each paper were eluted to determine the distribution of material
absorbing at 259 mu. The first four squares from the origin were
eluted for 36 hours at 37° with 0.4 m KOH, and the other squares
were eluted with 2m NH HCO; for 36 hours at 37°. In control
experiments, 90 and 105% of the radioactivity and 105 and 104%
of the absorbance were thus recovered from (pA*)3(pA)n and
4C-(pA)s, respectively. No further material was eluted by 0.4
KOH] from the squares which had been eluted with 2“ NH,HCOs.
The recovery of absorbance from successive time points was 40.4,
38.1, 41.7, 41.8, 48.5, und 43.2 Aess units, and recovery of radioac-
tivity was 4.3, 4.1, 4.1, 4.0, 4.4, and 4.6 X 103 epm, respectively.

N. G. Nossal and M. F. Singer

 

 

919
SEPHADEX G-100
oa RNAose I
‘ o 0%
a
03k (40"7
0.2
ale
2
€
o
S Jy A A
& 0s
z , Venom
2 0%
oO | 10%
2 04 a 2%
<
0.3
02
Ol

 

 

 

 

 

100 '20

Fic. 4. Sephadex G-100 chromatography of the products of the
degradation of poly A by RNase II and venom phosphodiesterase.
The reaction mixtures described below were layered on a Sephadex
G-100 column (2 X 37 em) at 4°, and the column was developed
with 0.1 m NH,HCO; (pH 8). The flow rate was approximately 6
ml per hour, and 1-ml fractions were collected. RNase II incuba-
tions contained 0.75 mm poly A (Fraction III (34)), 0.1m Tris-HCl
(pH 7.5), 0.1 m KCl, 1.5 mm MgCl, 0.02 mg of BSA, and enzyme in
a total volume of 0.4 ml. The reaction mixtures were incubated
in the chamber of a Beckman spectrophotometer (approximately
28°), and the degradation was followed by the hyperchromic shift
at 259 my in a cell with a 1-mm path length. The amount of en-
zyme and times of incubation were: O, no enzyme, 0 min; @, 1
unit, 60 min; A, 2 units, 52 min. Venom phosphodiesterase reac-
tion mixtures were incubated in a spectrophotometer cuvette as
described above and contained 0.75 mm poly A (Fraction IIT (34)),
0.063 m Tris-HCI] buffer (pH 9.0), and enzyme in a volume of 0.4
ml. The amount of enzyme and times were: O, no enzyme, 0
min; @, 0.45 unit, 17 min; A, 0.45 unit, 80 min. The position of
the peak of (Ap); is shown for reference.

remove NH,HCO;, and chromatographed on Whatman DE-81
paper with 0.5 m NH JICOs,, 4.0% of the radioactivity migrated
with 5’-AMP, 23.99% with (pA)2, 66% with (pA)s, 5.8% with
(pA), and 0.38% with (pA)s. No radioactivity was detected in
compounds migrating more slowly than (p<A)s.

Size of Polymer Substrate during Hydrolysis—The average
size of the residual polymer substrate, during the course of
digestion by RNase II or venom phosphodiesterase, was in-
vestigated by gel filtration on Sephadex G-100. Regardless of
the extent of degradation, RNase II digests of poly A (Fig. 4)
contained only poly A of the same size as the original substrate,
920

 

SEPHADEX 6-100
UMP

ABSORBANCY 262 mz

 

 

 

120

mi.

Fig. 5. Sephadex G-100 chromatography of RNase II digests of
poly U. The incubations contained 2.9 mm poly U, 0.04 « Tris
(pH 7.5), 0.04 m KCI, 0.5 mm MgCl, 0.02 mg of BSA, and enzyme
in a volume of 0.1 ml. Reaction mixtures were incubated at 37°.
Cold HO (0.3 ml) was added, and the mixture was layered on a
Sephadex G-100 column as described in Fig. 4. O, no enzyme,
0 min; @, 2 units, 15 min; A, 2 units, 50 min; A, 2 units, 75 min.
The elution profile of (Up); is given for reference.

TABLE VII
Degradation of tRNA

In Experiment 1, the reaction mixture, containing 47 A2zeo
units of tRNA (Schwarz), 0.05 m Tris-HCI! (pli 7.5), 0.1 m KCI,
1 ma MgCh, and 36 units of RNase IJ in a volume of 0.6 ml, was
incubated at 37°. Controls without enzyme or with neither
enzyme nor tRNA were run simultaneously. At 0 and 240 min,
duplicate 50-u] aliquots were removed, and the extent of hydroly-
sis was determined by the uranium acetate-perchlorie acid
method. No degradation occurred in the absence of enzyme.
At the same times, 40-u] and 50-1 aliquots were taken to measure
valine and serine acceptor activity, respectively (see ‘(Experi-
mental Procedure’). Results shown are the average of duplicate
determinations. In Experiment 2, the conditions were the same
as in Experiment 1 except that the RNase IJ digestion mixture
contained 2 mm MgCle, and 20-ul aliquots were used to determine
acceptor activity. The conditions for Experiment 3 were like
those for Experiment 1 except that 63 units of tRNA prepared
by the method of Zubay (27) were used.

 

 

 

 

Experiment {tRNA degraded | Acceptor amino acid Initial accentor rast
%  %
1 13.9 Serine 0.52 i 2
1 13.9 Valine 1.40 | 23
2 9.3 Phenylalanine 1.26 2
3 19 Valine 0.76 | 27

 

* Nanomoles of amino acid accepted per pmole of tRNA.

AMP, and very short oligonucleotides in the shoulder of the
mononucleotide peak. Similarly, digests of poly U (Fig. 5) had
only the original substrate, UMP, and small amounts of very
short oligonucleotides. These would be the expected products of
processive degradation; no significant concentration of inter-
mediate length polymers would accumulate. In contrast,
venom phosphodiesterase digests of poly A do contain poly-
nucleotides of intermediate length, as shown by the significant
absorbance in the fractions between the original substrate and

Processive Exonucleolytic Digestion of Polyribonucleotide Chains. I

Vol. 243, No. 5

AMP. <A comparison of Figs. 4 and 5 shows that the fractiona-
tion on the gel under these conditions is not based solely on size,
since UMP is eluted before AMP. However, within a homol-
ogous series, the volume required for elution decreases logarith-
mically with increasing chain length (35).

Degradation of tRNA—tRNA is a comparatively poor substrate
for RNase II (2, 3), but it is degraded slowly by high levels of
enzyme. It was of some interest to study the effect of this
digestion on the ability of tRNA to accept amino acids, since
degradation by two other exonucleolytic cnzymes, namely,
snake venom phosphodiesterase and polynucleotide phospho-
rylase, both of which degrade in the 3’ to 5’ direction, have
yielded markedly different results. Thus, degradation with the
venom enzyme quantitatively destroys amino acid acceptor
activity when only 5% of the RNA has been degraded to soluble
fragments (36). On the other hand, after degradation of 20% or
more of a sample of tRNA by polynucleotide phosphorylase, the
efficiency of the residual tRNA in accepting amino acids is the
same as that of the starting material (37, 38). Table VIT shows
that digestion of tRNA by RNase II yields results similar to
those obtained with polynucleotide phosphorylase. Even
after more than 10% of the molecule has been degraded to acid-
soluble material, acceptor activity remains, Furthermore, the
data suggest that some species of tRNA are more susceptible to
degradation by RNase II than others. For example, the same
solution of tRNA lost 235% of its total valine acceptor activity,
but only 2% of its serine acceptor activity, when 13.9% of the
tRNA was digested.

DISCUSSION

Interpretation of these studies on the mechanism of action of
RNase IL is impossible unless contaminating endonuclease
activity can be ruled out. An earlier report from this laboratory
concluded that the action of highly purified RNase II on syn-
thetic polyribonucleotides is primarily, if not exclusively,
exonucleolytic (3). On the other hand, Spahr (1), using a much
less purified enzyme preparation, reported significant endo-
nucleolytic degradation of high molecular weight RNA from
bacteriophage R-17. ‘The experience in this laboratory since our
earlier work has all tended to support our original conclusion,
The additional evidence concerning the absence of significant
endonuclease activity in our RNase II preparations will be sum-
marized here.

Investigations of the products of the hydrolysis of oligonucleo-
tides have given no indication of endonucleolytic cleavage even
though rather high concentrations of enzyme were used. All of
the products are consistent with an exonucleolytic attack from
the 3’-hydroxyl end. The stoichiometry of the reaction products
of the digestion of (Ap)sA (Table V) and (pA)s (legend, Table
IV) show that one small oligonucleotide from the 5’-hydroxy]
chain end and a corresponding number of mononucleotide units
are formed from each molecule of heptamer or pentamer that is
degraded.

Endonuclease activity would also have been detected when
the products of polyribonucleotide digestion were studied by
chromatography on paper or by gel filtration on Sephadex G-100.
However, no oligonucleotides longer than tetramers moved from
the origin when the products of the degradation of (Ap),G* and
(pA*)3(pA),, were chromatographed in Solvent 5 and on DEAE-
cellulose paper, respectively. Under these conditions even
nonamers have a finite mobility and would have been detected.
Issue of March 10, 1968

Similarly, Sephadex filtration of the products of poly A and
poly U digestion showed no products migrating between the
substrate and the small oligonucleotide derived from the 5/-
hydroxyl chain end. Under these conditions, chains one-half (or
less) the length of the original substrate could have been casily
detected,

Recent. studies by Bishop, Koch, and Levintow (39) with
double stranded poliovirus RNA also suggest that highly
purified RNase [I is free of endonuclease. Thus, although the
infectivity of the double stranded RNA is rapidly lost upon
treatment with our Fraction VII of RNase II, the treated RNA
sediments exactly like the control in 5 to 20% sucrose density
gradient (Fig. 9 of Reference 39). A rapid lowering of the
average size of the RNA chains would have been expected in the
presence of endonuclease. It should be noted, however, that if
an endonuclease hydrolyzed only cne chain of the double helix,
hydrogen bonding might hold the chains together, making a few
such breaks per molecule undetectable by this method. The
enzyme concentration in these experiments was approximately
100 times in excess of that necessary to hydrolyze completely a
quantity of poly A equivalent to the viral RNA. It is of course
possible that for some unknown reason the R-17 phage RNA used
by Spahr is a better substrate for an endonuclease activity
asscciated with RNase II than are the substrates we have used;
on the other hand, the observed endonuclease activity in his
experiments may have reflected the purity of the enzyme tested.
In any case it is clear that the purified RNase If used in the
present studies does not significantly hydrolyze the internal
diester bonds of the synthetic polyribonucleotides used.

IT Nase I] appears to hydrolyze both oligonucleotides and
polyribonucleotides, beginning at the 3’-hydroxy] chain end,
This is shown by the liberation of XMP from hexamers of the
type (Ap)sX (where X is cytidine, guanosine, or uridine), by the
absence of adenusine in the products of the digestion of (Ap)sA,
and by the production of pAp and pGp upon hydrolysis of
(Ap); and (Ap)sGp, respectively. Similarly, XP is found at all
stages during the digestion of polymers of the type (Ap).X
(where X is uridine, cytidine, or guanosine), but “C-AMP from
a polymer labeled in the first 3 residues at the 5’ hydroxy] chain
end is found only during the terminal stages of the degradation.

While hydrolysis proceeds from the 3’-hydroxy] chain end,
studies with polymers labeled at the 3’-terminal end show that
the proportion of the terminal nucleotide and the proportion of
the molecule as a whele which are acid-soluble are similar’at all
stages of the digestion (Mig. 2). This result contrasts sharply
with that found for snake venom phosphodiesterase (Fig. 2)
(33) and various E. coli DNA exonucleases (33, 40), in which
most of the label from the 3’-terminal nucleotide residue is
solubilized when less than 10% of the total molecule has been
degraded. The results with RNase II are most easily interpreted
as reflecting a processive degradation in which the enzyme
repeatedly attacks a single polyribonucleotide chain, hydrolyzing
it completely to 5’-mononucleotides and a small resistant
oligonucleotide, before releasing that oligonucleotide and at-
taching to another chain. The products found during the
digestion of the 5’-terminally labeled (pA*)s(pA)» are also
consistent with the processive mechanism (Fig. 3). Thus the
proportion of the radioactivity from the 5’ chain end which is
found as oligonucleotide, and the proportion of the entire chain
hydrolyzed to 5’-AMP, are similar throughout the course of the
degradation. ‘There is, in fact, a small lag in the liberation of

N. G. Nossal and M. F. Singer

921

labeled oligonucleotide compared to AMP. This may reflect the
preference of the enzyme for the longer chains, which have a
higher ratio of 8C- to #C-AMP. This lag was decreased when
the substrate being hydrolyzed was more homogeneous with
respect. to size! It should be pointed out. that all experiments
presented here were carried out at enzyme concentrations far
below the substrate concentrations. Experiments with  stoi-
chiometric amounts of enzyme would be expected to vield results
similar to those obtained with venom phosphodiesterase.

Independent evidence for the processive mechanism of degrada-
tion was obtained from the study of the size of the residual
substrate during digestion, by gel filtration on Sephadex G-100
(Figs. 4 and 5). Even when 65°% of the poly A was converted to
AMP and resistant oligonucleotide, no polymers shorter than the
original substrate were apparent. A small change in length of
the portion of the poly A that was excluded by the gel would not
have been detected, but a shortening of the bulk of the polymer
would have been apparent under these conditions. The results
obtained with the venom phosphodiesterase under the same
conditions point out the ability of the Sephadex column to reveal
polymers that are shorter than the initial substrate. The
decrease in chain size observed, especially at the stage when only
10‘ of the polymer was degraded to AMP, is somewhat greater
than that expected for simple exonucleolytic attack, and suggests
that the venom preparation was probably contaminated with
endonuclease. The contamination must have been quite small,
however, since, when the same preparation was used for the
studies with (Ap),U*, most of the UMP-"C was liberated before
10% of the absorbance was solubilized (Fig. 2).

‘The effect of RNase IT degradation on the capacity of tRNA
to accept amino acids is also consistent. with a processive attack.
When 10% of the total tRNA was degraded, some of the mole-
cules must not have lost even a single nucleotide, since the
terminal -pCpCpa sequence must be intact for loading to vccur
(36). Consequently, other tRNA chains, which do lose their
incorporation ability, are probably degraded much more than
10%.

Studies with short oligonucleotides showed that even with
these comparatively poorly bound substrates the enzyme tends |
to attack in processive fashion. Thus, with compounds of the
type (Ap)sX, approximately equal quantities of AMP and XMP.
were produced at a time when up to 50° of the hexamer was still
undegraded. Random attack would have resulted in a more
rapid formation of XMP compared to AMP, Random attack
would also have led to the accumulation, during the early stages
of the digestion, of oligonucleotides 1 unit shorter than the
substrate, as has been observed with snake venom phospho-
diesterase (28, 41). In contrast, with RNase IT, the oligonucleo-
tide 1 unit shorter does not accumulate during any stage of the
digestion of (pA)s or (Ap)sA. Rather, the most abundant
oligonucleotide throughout the digestion ix (pA); from chains
phosphorylated at. the 5’-hydroxyl end, such ag (pA)s, and (Ap)sA
from dephosphorylated chains, such as (Ap)sA (Tables TV and
V).!

The experiments reported here do not elucidate the molecular
mechanism of processive degradation by RNase 1], Under con-
ditions in which (pA)s does not accumulate as a product of the
digestion of poly A, poly A strongly inhibits the hydrolysis of
free (pA)s (Fable IV). This suggests that the (pA)s unit formed

4 Unpublished observations.
922

transitorily during the hydrolysis of poly A is not equivalent to
free (pA)s. Thus, the polyribonucleotide behaves as if it were
bound to the enzyme until it is degraded to an oligonucleotide of
a size that no longer permits firm attachment. The interaction
of polymer and enzyme may involve multiple sites or a large
continuous site, thus assuring binding to the enzyme even after
the terminal unit is released. The substrate might then shift
position relative to the catalytic site of the enzyme, so that its
new terminal residue could be hydrolyzed. For example, recent
studies with lysozyme (42, 43) and various proteases (44-46) sug-
gest that the active sites of these endodepolymerases are large
enough to bind 6 or 7 monomer units simultaneously. Alterna-
tively, the observed results may reflect the fact that the binding
and diffusion constants of long polymers are such that there is a
strong probability that released polymer molecules will bind
again to the enzyme before they can move away. The reader is
referred to the “Discussion” in the accompanying paper (34) for
some further speculation on the mechanism of processive degrada-
tion.

The possibility that an exohydrolase might remove many units
from one polymer molecule before beginning the degradation of
another chain was previously considered by Bailey and French
(47). Using specifically labeled amylase chains, they showed
that sweet potato @-amylase removes between 3 and 4 maltose
units during each period of association with the substrate. Some
endopeptidases also hydrolyze a given protein molecule exten-
sively before initiating the hydrolysis of another. However, in
these latter cases, it appears that the hydrolysis of the first bond
converts the protein to a form which is a better substrate than
the original molecule (48, 49).

Studies of the products of the degradation of E. coli messenger
RNA in vivo have led to the conclusion that both RNase LI and
polynucleotide phosphorylase may be involved (50, 51). It: is
therefore interesting that polynucleotide phosphorylase also
appears to degrade by a processive attack on single polymer
chains (84, 52). Preliminary studies suggest that an RNA
exonuclease from Ehrlich ascites tumor cells may also function
in this way (58). The mechanism may be advantageous because
it allows the complete degradation of particular messenger RNA
molecules without the accumulation of partial breakdown prod-
ucts which might interfere with protein synthesis,

Acknowledgment—The participation of Mr. G. P.C. Tolbert in
the intial stages of this work is gratefully acknowledged.

REFERENCES

. Spaur, P.F., J. Biol. Chem., 239, 3716 (1964).

. SINGER, M. F., anp Touserr, G., Science, 146, 593 (1964).

. Sincer, M. F., anp Toipert, G., Biochemistry, 4, 1319 (1965).

. Hurst, R. O., Lirvue, J. A., ann Burver, G. C., J. Biol.
Chem., 188, 705 (1951).

. Jones, O. W., Townsenp, E. E., Soper, H. A., anpb Hepret,
L. A., Biochemistry, 3, 238 (1964).

. Tuacu, R. E., in G, L. Caxvons ano 1). R. Davies (Editors),
Procedures in nucleic acid research, Warper and Row, New
York, 1966, p. 520.

7. Markuay, R., anv Smiru, J. D., Blochem. J., 49, 401 (1951).

8. MarkuaM, R., ano Suirn, J. D., Biochem. J., 62, 552 (1952).

9. Bray, G. A., Anal. Biochem., 1, 279 (1960).

. Lowxy, O.H., Roseproucn, N.J., Farr, A. L., anp Ranpaun,
R. J., J. Biol. Chem., 193, 265 (1951).

. Sincer, M. F., in G. L. Canton: ano D. R. Davies (Editors),
Procedures in nuclete acid research, Harper and Row, New
York, 1966, p. 192.

Cammack, K. A., anpD Wave, H. F., Biochem. J, 96, 671 (1965).

Martin, R., anp AMES, B., J. Biol. Chem., 236, 1372 (1961).

howe

rn an

12.
13.

a

Processive Exonucleolytic Digestion of Polyribonucleotide Chains. I

14.

15.
16.
17.

18.
19.

20.
al.
22.
23.

24.

25.
26.

29.
30.
31.
32.
33.
34.
35.
36.
37,
38.
39.
40.

41.

42.
43.

44,
45.

46.

49.

51

52.

Vol. 243, No. 3

Epsaut, J. T., in W. Neunara ano K, Bartey (Editors), The
proteins, Vol. I, Part B, Academic Press, New York, 1953, p.
549.

Singer, M. F., axp O'Brien, B., J. Biol. Chem., 238, 328
(1963).

Winurams, F. R., axp Gruxperc-Manaco, M., Btochim.
Biophys. Acta, 89, 66 (1964).

Singgn, M. F., ann Guss, J. K., J. Biol. Chem., 287, 182 (1962).

Stevens, A., AND Hitmoe, R., J. Biol. Chem., 236, 3016 (1960).

Torniant, A., in G. L. Cantoni and LD. R. Davies (Editors),
Procedures in nucleic acid research, Harper and Row, New
York, 1966, p. 224.

Takanasul, K., J. Biochem. (Tokyo), 49, 1 (1961).

Koerner, J. F., anp SinsHEIMER, R.L., J. Biol. Chem., 228,
1049 (1957).

Keuier, E. B., Biochem. Biophys. Res. Commun., 17, 412
(1964).

Neu, H. C., ano [leppez, L. A., J. Biol. Chem., 240, 3685
(1965).

Mvencu, K. H., axp Bers, P., in G. L. Caxronr anp 1). RB.
Davies (Editors), Proredures in nucleic acid research, Harper
and Row, New York, 1966, p. 375.

Maxman, M. H., anp Canroni, G. L., Biochemistry, 4, 1434
(1965).

Mans, R. J., AND Nove, G. 1D., Arch. Biochem. Biuphys.,
94, 48 (1961).

27. ZuBay, G., J. Mol. Biol., 4, 347 (1962).
28.

RazzeuL, W. E., anp KHorana, H. G., J. Biol. Chem., 234,
2114 (1959).

Singer, M. F., Hitmos, R. J., anp Grunperc-Manaco, M.,
J. Biol. Chem. , 235, 2705 (1960).

ScHEUERBRANDT, G., DUFFIELD, A. M., anv Nusszaum, A.L.,
Biochem. Biophys. Res. Commun., 11, 152 (1963).

FE.ix, F., Porrer, J. L., axp Laskowsx1, M., J. Bial. Chem.,
235, 1150 (1960).

Leuman, I. R., Roussos, G. G., anp Pratt, LE. A., J. Biol.
Chem., 287, 819 (1962).

Leuman, I. R., J. Biol. Chem., 236, 1479 (1960).

Kvee,C.B., anp Singer, M.F., J. Biol. Chem., 243, 923 (1968).

Houn, T., anp ScHauuEr, H., Biochim. Biophys. Acta, 188,
466 (1967).

Preiss, J., Dieckmann, M., aNp Bere, P., J. Biol. Chem., 236,
1748 (1961).

Singer, M. F., Lusorsxy, S., Morrison, R. A., any Canront,
G.1L., Blochim. Biophys. Acta, 38, 568 (1960).

Monier, R., anp GrunBeRG-Maxaco, M., Acides ribonu-
cléiques et polyphosphates, structure, synthése et fonctions, .
Collog. Int. Centre Nat. Rech. Sct. (Parts), 163 (1962).

Bisnop, J., Kocn, G., axp Levinrow, L., in J. 8. Courer
and W. Paranxcuycu (Editors), The molecular biology of ,
viruses, Academic Press, New York, 1967, p. 335.

Leuman, I., Progr. Nucleic Acid Res., 2, 83 (1963).

Houiey, R. W., Mapison, J. T., ano Zamir, A., Biochem.
Biophys. Res. Commun., 17, 389 (1964).

Puiniivs, D. C., Proc. Nat. Acad. Sct. U. 8. A., 57, 484 (1967).

Ruruey, J. A., ano Gaves, V., Proc. Nat. Acad. Set. U.S. A.,
57, 496 (1967).

Scuecnter, J., ann Bercer, A., Biochemistry, 5, 3371 (1966).

ScHecuTer, I., anp Bencer, A., Biochem. Biophys. Res. Com-
mun., 27, 157 (1967).

BrepacuHer, L. J., anp Benner, M. L., Biochem. Biophys.
Res. Commun., 27, 176 (1967).

. Bartey, J. M., ano Frencn, D., J. Biol. Chem., 226, 1 (1957).
48.

Ginsuura, A., AND ScHacuman, IL. K., J. Biol. Chem., 285,
108 (1960).

LiInpERStTROgM-LaNG, K. U., Lane medical lectures, Stanford
University Press, Palo Alto, California, 1952, p. 64.

. Sincer, M. F., anv Leper, P., Annu. Rev. Biochem., 35, 195

(1966).

. Kiviry-Vocet, T., ano Exson, 1)., Biochim. Biophys. Acta,

138, 66 (1967).
TuHane, M. N., Guscuunaurr, W., Zacnav, H., anp Grun-
BERG-Manaco, M., J. Mol. Biol., 26, 403 (1967).

. Lazarcs, H. M., ano Sporn, M. P., Proc. Nat. Acad. Set.

U.S. A., 67, 1386 (1967).