Polynucleotide Phosphorylase of Micrococcus luteus

Primer-dependent polynucleotide phosphorylase catalyzes the polymerization of ADP even in the absence of oligonucleotide primer. The reaction is very slow compared to the rate in the presence of primer, but the extent of reaction is directly proportional to time. Autocatalytic kinetics and long lags, observed earlier with cruder enzyme preparations, are absent with highly purified enzyme. The products of the reaction are very long polymer chains and do not serve as primers. No short chain intermediates are detectable and polymerization is therefore processive. Polymerization by primer-independent enzyme is also processive. In unprimed synthesis the apparent Km for ADP with dependent enzyme is several orders of magnitude higher than that found with independent enzJ-me. In the presence of oligonucleotide primer, dependent enzyme catalyzes the addition of mononucleotide residues to the primer. The reaction is not processive; no very long chain polymer is made and short chain intermediates are readily detected. On the other hand, primer-independent enzyme does not incorporate primer, if supplied, into long chain polymer; under these conditions polymer synthesis de novo continues in a processive fashion.

The reaction is very slow compared to the rate in the presence of primer, but the extent of reaction is directly proportional to time. Autocatalytic kinetics and long lags, observed earlier with cruder enzyme preparations, are absent with highly purified enzyme.
The products of the reaction are very long polymer chains and do not serve as primers. No short chain intermediates are detectable and polymerization is therefore processive.
Polymerization by primer-independent enzyme is also processive.
In unprimed synthesis the apparent Km for ADP with dependent enzyme is several orders of magnitude higher than that found with independent enzJ-me.
In the presence of oligonucleotide primer, dependent enzyme catalyzes the addition of mononucleotide residues to the primer.
The reaction is not processive; no very long chain polymer is made and short chain intermediates are readily detected.
On the other hand, primer-independent enzyme does not incorporate primer, if supplied, into long chain polymer; under these conditions polymer synthesis de novo continues in a processive fashion.
Polynucleotide phosphorylase catalyzes the reversible syrthesis of polyribonucleotides from ribonucleoside diphosphates, with the release of inorganic phosphate.
1 The Eighth Edition of the Catalogue of the American Type Culture Collect,ion (19G3) indicates that strain number 4698, formerly classified as Micrococcus lysodeiklicus, is now reclassified as Micrococcus luteus. The enzyme described in this paper was purified from cells obtained commercially (Miles Laboratories, Inc.). The Miles Laboratories' cataloglIe designation is M. lysodeikticus (ATCC 4698). Therefore, we now refer to these cells as M. hteus. with unesterified C-3'-hydroxyl groups (l-3). The added oligonucleotide served as a primer and was incorporated into the resulting polymer as the 5' terminus.
More recently, polynucleotide phosphorylase of 1V. luteus was also obtained in a primer-independent form. l,eferred to as Form-P (4-7).
This enzyme may be converted into a primerdependent form (Form-T) by limited proteolysis with trypsin (4, 6-10).
Form-T is reconverted to a new primer-independent form by reduction with scllfhydryl reagents (6). Form-I is easily distinguishable from Form-T on polyacrylamide disc clcctrophore'sis while reduced Form-T has the same mobility as Form-T itself (6). The electrophoretic mobility of Form-T is identical with that of the early preparations that were primer dependent, as isolated (4).
Neither the primer independence of Form-I nor the Ijrimer dependence of Form-T is absolute.
Polymerization of Al)l' by Form-I may be stimulated up to 2-fold by oligonucleotide (6), while Form-T does catalyze polymerization in t)he absence of primer, albeit with difficulty.
Earlier studies on the kinetics of polymerization by primerdependent il!f. luteus enzyme, in the absence of oligonucleotides, indicated that a lag period occurred during which little or no reaction could be detected (3,8,9). After the lag period polymerization followed one of two courses. In some cases the reaction proceeded in a linear fashion (3) and in others it appeared to proceed at a constantly accelerating rate (8). The significance of these variations is unknown, given the fact that most of the enzyme preparations were relatively crude. Nevertheless t,hc results suggested that primer-dependent enzyme could initiate polymer synthesis only with difficulty.
It was generally believed that the lag represented a period during which the formation of polynucleotides occurred slowly and that the resulting chains could then act as primers, thereby leading to an increased reaction rate. The sometimes noted autocatalytic kinetics were consistent with such an interpretation.
These findings were in contrast to those obtained with rnzyme from other bacterial sources, which, in general, did not require primer except in special circumstances.
For example, highly purified Escherichia coli polynucleotide phosphorylase was not Issue of RIay 10, 1970 R. E. dloses and M. F. Xinger 241.5 dependent, on primer (1 I) but a lag phase. that coirld be climirrated by oligouucleotides, was induced by a variety of treatments (11. 12). In addition, some preparations of enzyme from Azotobncter agilis which could not later be reproduced, were greatly stimulated by oligonuclcotide primer (13). These data hare been reviewed in detail (14).
Using highly purified Form-T from JI. luteus, we love now re-examined the kinetics of polymerization by primer-dependent Form-T enzyme, and the results are reported here.
EXPl<:RIMENTAL PROCEDURE Nucleoti~es-ApApApA was obtained from Miles Laboratories. 311-pA1pApA was prepared by degrading 15 rrg of 3H-poly A with sheep kidney endonuclease according to the met,hod of . The digest, which contains primarily (PA), with n from 2 to 6, was deproteinized repeatedly wit,hchloroform and isoamyl alcohol (16). Deproteinization with phenol was avoided because of the solubility of oligonucleotides in phenol.
In the chloroform-isoamyl alcohol procedure less than l!$ of the oligomrcleotide is soluble in the organic phase. The solution of oligonucleotides was applied to a column (2.5 x 38 cm) of microgranular diethylaminoethyl cellulose (DE-52, Whatman) prepared as described by Tener (17) and equilibrated with 0.05 M NaCl containing 0.01 &r Tris-HCl, pH 7.5. The elution of oligonucleotides was carried out with a linear gradient (2 liters total) from 0.05 to 0.3 in NaCl in the presence of 7 M urea and 0.01 M Tris-I-ICI, pH 7.5 (17). Appropriate fractions were pooled and desalted by a modification of the method of Rushizky and Sober (18) in which ammonium bicarbonate was replaced by triethylarnrnonium bicarbonate, pH 7.9. The purity and chain length of the oligonucleotides was determined by paper chromat,ography with known markers.
Although the peaks on the column were symmetrical and well separated, significant contamination with oligonucleotides of both lower and higher chain length was detected.
Gel electrophoresis materials were from Canal Industrial Corporation, Bethesda, Maryland.
Enzyme-Polynucleotide phosphorylase was prepared from M. luteus through the hydroxylapatite step, as previously described (5, 7). The primer-independent enzyme (Form-I) had a specific activity of 53 phosphorolysis units per mg of protein (1 phosphorolysis unit is the amount of enzyme catalyzing the formation of 1 pmole of ADP from poly A in 15 min) and 193 polymerization units per mg (1 polymerization unit catalyzes the incorporation of 1 pmole of ADP into poly A in 15 min).
Protein was calculat,ed from the absorbance at 280 nip, with the extinction coefficient reported by Klee (7). Two preparations of primerdependent enzyme (Form-T) were used and both gave identical results. One (Preparation A) was prepared, by limited proteolysis with trypsin, from the Form-I enzyme just described (7). The second preparation of Form-T (Preparation B) was prepared by limited proteolysis of a crude enzyme preparation with trypsin (8). Although the latter material was then purified through the hydroxylapatite step, it did not attain the purity usually found at that stage. The specific activity was 12.3 phosphorolysis units per mg of protein.
Form-T enzyme was alkylatcd with N-ethylmaleimide as described previously (7).
Sheep kidney nuclcase was prepared according to the method of . Enzyme ilasa2/sPol~nierizat~ioi~ was routinely assayed by measuring the release of l'i from hDP, or other nucleoside diphosphates.
When primer was used it was A1p,4pBpA unless otherwise noted.
Incubations were all at 37". At the end of the incubation period, the reaction was stopped by the addition of 20 volumes of ice-cold 2.57" perchloric acid. After 10 min in ice, the samples were centrifuged and the supernatant fluid used to assay for Pi by the method of Fiske and SubbaRow (19). When the incorporation of labeled substrates into polymer was measured, the acid-precipit'able material was collected on Whatman GF/C' glass filters, 2.4 cm, and washed three times with 3 ml of ice water, dried, and counted in toluene scintillator.
The general procedure for a time course was t'o assemble the reactants and begin the reactions with addition of enzyme. At the various time points, 0.05-ml samples were withdrawn and added to 0.95 ml of 2.5% perchloric acid. The zero time aliquot, which was removed immediately after the addition of enzyme, served as the blank.
Disc Gel ElectrophoresisDisc gel electrophoresis was used to check the purity of polynucleotide phosphorylase and to distinguish Form-I from Form-T. The procedures have been described (4, 7).
For analysis of poly A product, disc gel electrophoresis was carried out with 5% acrylamide gels according to the methods of Mcl'hie,Hounsell,and Gratzer (20) and Gould (21) except that samples were applied in a sample gel at pH X.9. For a period of 35 min, 5 ma per tube were applied at room temperature. The current was stopped when the bromphenol blue marker was still about 1 cm from the bottom of the tube. After completion of the run the gels were sliced into about 40 segments with an "egg slicer" (22) and the slices were placed into glass scintillation vials. Thirty percent hydrogen peroxide (75 ~1) was added, the vials were tightly capped, and then heated for 8 to 12 hours at 55" to dissolve the gel (23). At the end of that period, the vials were cooled, and 1.0 ml of water was added to each. After shaking, 10 ml of Triton-toluene scintillator (24) was added to each vial and the samples were counted.
As expected from the data of McPhie et al. (20) preliminary experiments established that polymers of chain length 700 or greater, migrate 1 cm or less into the separating gel under the conditions used. Polymers of chain length 730 and 1200, kindly provided by Dr. Gary Felsenfeld (25), were used to calibrate the gels. Some of the polymer remains in the stacking gel. ADP and oligonucleotides (n 5 6) migrate in front of the dye marker while tRNA is found just behind the dye marker.
Free enzyme (Form-I) migrates about 27% of the length of the gel; in a gel yielding 40 slices this corresponds to approximately Fraction 11. Slicing of the gels into segments did not afford reliably uniform slices, thus accounting for the variation in the number of slices obtained from each gel as vvell as for some of the irregularities in the pattern of radioactivity.
In the figures, the numbering of slices starts from the top of the gel and the first 3 to 5 slices, in general, represent stacking gel. The large peak of radioactivity at the bottom of the gel is mainly ADP.
Paper Chromatography-Certain reaction mixtures were  Similar kinetics were obtained when the course of the reaction was followed by measuring the formation of acid-insoluble poly A product.
In the absence of primer the amount of poly A  concentration. The reaction mixtures and procedure were described under "Experimental Procedure" except that the concentration of nucleoside diphosphate was varied as indicated, and the ratio of nucleoside diphosphate to Mg2+ was maintained at 4:l (1). The concentration of (Ap)zA was 0.27 mM. The concentration of Form-I enzyme was 15 pg per ml. With ADP the time of incubation was 20 min; with UDP it was 15 min. In all cases the extent of reaction was less than 25% and was in a range where extent of reaction is linear with time. Each point represents the average of duplicate determinations.
The velocity is expressed as nanomoles of Pi released per min per ~1 of stock enzyme solution (6.2 mg of protein per ml).
formed equals the amount of Pi released (Fig. 3). Thus, Form-T catalyzes polymerization in the absence of primer; the release of Pi is not caused by a phosphatase reaction.
In the presence of primer the amount of poly A formed equals the Pi released at early time points. At later times acid-precipitable poly ,4 is less than the Pi released.
The lack of agreement after extensive reaction may reflect the redistribution of chain sizes as the reaction approaches equilibrium (26), as well as the relatively short chain product produced under these conditions (see below).
It is necessary to consider the possibility that the polymerization by Form-T enzyme in the absence of primer might actually reflect contamination of the enzyme with either Form-I or reduced Form-T rather than the catalytic activity of Form-T itself. However, several observations indicate that Form-T is indeed able to cat)alyze polymerization iu the absence of primer at a linear, albeit slow, rate.
To eliminate the problem of contamination by reduced Form-T, a sample of Form-T enzyme was first treated with fl-mercaptoethanol and then with N-ethylmaleimide as previously described (6, 7). The reduction converts Form-T to a primerindependent form; reaction with N-ethylmaleimide alkylates the reduced enzyme and reconverts it irreversibly to a primerdependent state. The treatment results in some loss of over-all activity (6, 7). The alkylated enzyme does, however, catalyze the polymerization of ADP; the reaction is linear with time and the rate is stimulated about lo-fold by addition of ApApApA (Fig. 4). Thus, the linear rate of polymerization by Form-T cannot be ascribed to contamination with reduced Form-T. The possibility that the reaction results from contamination of Form-T by Form-I also appears unlikely.
Electrophoresis of large amounts of Form-T on polyacrylamide gels indicated slight activity in the region characteristic of Form-I. However, the activity of Form-T is clearly discernible when gels are incubated with ADP and the resnlting polymer stained with acridine orange (4). In this situation the two enzyme forms are widely separated on the gel. In addition, when Form-T enzyme was passed over a column of superfine Sephadex G-200, which separates any contaminating Form-I from the Form-T (8), the activity in the peak was stimulated about 20-fold by ApApApA and polymerization of ADP in the absence of primer was again linear with time.
The experiments shown in Figs. 5 and 6 support the conclusion that Form-T has inherent polymerizing activity, and in addition they reveal an important difference between primerindependent and primer-dependent enzyme. The polymerization of ,ADP 'by Form-I (Fig. 5) and by Form-T (Fig. 6) was studied as a function of ADP concentration, and the data expressed on double reciprocal plots. The apparent K, values (obtained from the intersection of the lines on the horizontal ais) for' Form-I and Form-T, in the absence of primer, differ by several orders of magnitude.
In the presence of primer, the apparent K, for Form-T decreases as a function of primer concentration and approaches that found for Form-I in either the presence or absence of primer.
In addition, the apparent K, for reduced Form-T (Fig. 6)

Substrate
Specificity-In view of reports (9, 10) that trypsintreated polynucleotide phosphorylase demonstrates marked substrate specificity in polymerization, the polymerization UDP and CDl' was tested with Form-T.
Both substrates yielded linear t,ime courses in the absence of primer.
Under the conditions used, the relative activities toward ADP, UDP, and CDP were 1.1, 0.9, and 0.25 nmoles of Pi released per min per ~1 of enzyme solution, respectively.
The rates of polymerization of UDP and CDP were stimulated IO-to 20-fold by the addition of primer.5 The apparent K,n for IJDP with Form-I is shown in Fig. 5

Products of Polymerization Reaction
The experiments just described show that in contrast to less highly purified enzyme, the recent preparations of dependent enzyme show neither a lag nor any sort of autocatalytic kinetics in the polymerization reaction.
The earlier demonstrations of lags, or autocatalytic curves, were interpreted to indicate that some product of the reaction, presumably an oligonucleotide, was serving to stimulate the rate of polymerization.
We t'herefore studied the nature of the polymer products with t,he various enzyme forms.

Effect of Reaction Products on Rate of Polymerization by Form-T
Enzyme-Using the higlily purified Form-T, we investigated directly the possibility that stimulatory material is formed during the polymerization reaction.
A standard reaction mixture was set up containing Form-T enzyme and no primer. Samples were withdrawn at intervals and heated at 75" for 10 min to inactivate the enzyme.
(1 're lminary experiments demonstrated 1' that this treatment completely destroys polynucleotide phosphorylase activity but does not affect the activity of oligonucleotide primer.) The samples were then cooled rapidly, an amount of fresh enzyme equal to that originally present was added, and incubation at 37" was resumed.
The time course of the reaction was followed by Pi release in the usual manner both in the original mixture (Fig. 7, main time course) and in the treated saml,les (Fig. 7, it, B, and C). One portion of the original incubation mixture was heated at zero time as a control (Fig. 7/l).
The samples in Fig. 7, B  is produced during the polymerization reaction, a finding that is confirmed by experiments described below.
Thus, for example, when 67 nmoles of ADP are polymerized per 50 ~1 of reaction mixture, the concentration of polymer, in A%W residues, is 1.34 pmoles per ml. If this were composed entirely of tetranucleotide, it would represent a concentration of 0.34 ITIM. Fig. 1B shows that at such a concentration, polymerization should be stimulated approximately 20-fold. The autocatalytic curves obtained in work with crude enzymes could have resulted from the secondary breakdown of the poly A product into oligonucleotides by a contaminating endonuclease. As shown in Fig. 8, addition of endonuclease to reaction mixtures containing ADI' and Form-T enzyme does indeed lead to autocatalytic curves. The sheep kidney endonuclease used is known to generate oligonucleotides with unesterified terminal C-3'hydroxyl groups, and phosphomonoesterified terminal C-5'hydroxyl groups (i.e. (PA)~ (15)).
Size of Poly A Products-Direct analysis of the chain length of t,he polymer product was carried out by means of electrophoresis on polyacrylamide gels and paper chromatography.
The former procedure was used to detect poly A of chain length greater than 50 residues and the latter to study the formation of oligonucleotides. n 5 9. In the electrophoresis experiments, which will be The experiments were carried out as described for Fig. 9. The enzyme concentrations (Form-T, Preparation B) were 64 rg per ml in the absence of (Ap) SA and 32 rg per ml in the presence of (Ap) aA.
presented first, '*C-ADP was used as substrate and duplicate samples of the reaction mixtures were removed at various times. One sample was used to measure the I'i formed in the standard way. The other sample was added to EDTA to stop the reaction and the mixture was applied to the top of a 5% polyacrylamide gel. After electrophoresis the gels were sliced and the radioactivity in each segment was determined as described under "Experimental Procedure." The main purpose of these experiments was to study the products accumulated with Form-T enzyme, but control experiments with Form-I were also carried out. Fig. 9 shows the results obtained upon gel electrophoresia of the products of polymerization with Form-I in the absence and in t'he presence of (AP)~A.
In both cases, very high molecular weight poly A (top of the gel) is detectable even after very little total reaction. The first times shown on Fig. 9 represent 3% utilization of the available ADP in the experiment with no primer, and 7.5% when ApApApA was present. As the reaction proceeds, more and more large poly A accumulates.
Calibration of the gels with poly A of known size (see "Experimental Procedure") indicates that the product has a chain length of 700 or greater. It is also notable that relatively few counts were detected between the peak of ADP substrat,e (bottom of the gel) and the peak of large poly A. The radioactivity in this region appears to increase some with time and may result from a redistribution of chain length as the reaction proceeds (see discussion of Fig. 3 above) or a minor nuclease contamination.
Results of a similar experiment with Form-T enzyme are shown in Fig. 10. In the absence of (Ap),A, high molecular weight poly A is apparent at the earliest time studied, and as with Form-I (Fig. 9) the amount of polymer increases with increasing release of Pi. Again, only small amounts of labeled polymer are apparent between the ADP and the material at the top of the gel. The result with Form-T in the presence of (Ap),A is strikingly different (Fig. 10). At none of the points studied was any poly ,4 detected at the top of the gel. With time, the amount of poly A with mobility less than that of tRNA increases somewhat, and is apparent as a low level of radioactivity along the whole length of the gel. When Form-T enzyme is first reduced with mercaptoethanol to convert it to a primer-independent form (6), the appearance of high molecular weight poly A (top of the gel) is again detected early in the reaction (Fig. 11).
Thus, the experiments in Figs. 9 through 11 indicate that long chain poly A (n > 700) is synthesized very early during the polymerization reaction, in all cases except with Form-T enzyme in the presence of oligonucleotide primer. The observation that Form-I, even in the presence of primer, synthesizes long chain material was of interest and led us to investigate the nature of the short chain products of polymerization.

The incubations
used were similar to those just described, except that either the ADP or the oligonucleotide primer was labeled. Samples of the reaction mixtures were removed at various times and used to determine the release of Pi or the nature of the accumulated oligonucleotides (by chromatography on DEAE paper).
With Form-I enzyme, 3H-pApApA disappears aud the radioactivity is converted to longer oligonucleotides (Fig. 12). At the first time point, the main product is tetranucleotide, but as the reaction proceeds, longer labeled oligonucleotides are apparent. Very little, if any, of the oligonucleotide ever appears at the origin where long chain poly A is found. Therefore, while long polymer is being synthesized by Form-I in the presence of oligonucleotide (see Fig. 9), little if any of the polymer forms by extension of oligonucleotide primer. An experiment (not shown) identical with that shown in Fig. 12  origin, with a suggestion of peaks corresponding to the expected labeled tetra-and pentanucleotides.
In the absence of oligonucleotide, no counts at all appeared on the chromatogram between the ADP and the origin. In this case, too, the trinucleotide is utilized and longer oligonucleotides appear; however, the label is incorporated into the material at the origin, indicating that the polymer is synthesized by extension of the oligonucleotide chains. Furthermore, the rate of Q~ApApA utilization relative to Pi release is much higher for Form-T than for Form-I. When the same experiment (not shown) was repeated with unlabeled trinucleotide and radioactive ADP, the oligonucleotides between pApApA and the origin were labeled and the pattern of label was consistent with that shown for Form-T in Fig. 12 Form-T enzyme, the rate of incorporation of aI-I-pApAp-n-as essentially constant and there was approximately 1 pmole of 3H-pApi\pA incorporated per 40 pmoles of Pi released. With Form-I enzyme, an apparent incorporation of 3H-pApApA into acid-insoluble material was also noted; 1 pmolr of 3H-pApApA per 140 pmoles of l'i released.
However, a c~ontrol experiment showed that under these conditions, there is coprecipitation of 31-I-p-ZpX~)A and poly A; the extent of this coprecipitation depends on t,hc amount of poly A and accounts in large measure for the observed incorporation with FormI.
With the value of 1 ~mole of 3H-p-4pLYpA per 140 pmolcs of Pi released as the value for coprecipitation, the results \vith Form-T can be sllitably corrected.
This calculation indicates that 1 pmole of 3H-pApApA is incorl)oratcd per 56 pmoles of l'i released, a chain length consistent with the electrophoretic and c~llrorllatographic findings.

DISCUSSION
The evidence presented here shows that primer-dependent polynucleotide phosphorylase from ill. luteus (Form-T) does indeed catalyze polymerization in the absence of oligonucleot~ide primer albeit at a very slow rate. In distinction to earlier results (3, 8, 0, 13) the rate of the reaction appears to be linear with time and is directly proportional to the enzyme concentration. The autocatalytic kinetics and apparent long lag periods observed earlier were probably artifacts arising from contamination of the enzyme preparations with nucleases.
Polyribonucleotides which can prime the 1)olymerization of ADP by FormT are apparently not produced in significant amount during unl>rimed synthesis.
This conclusion is supported by several observations in addition to the linear polymerization rate. (a) In the experiment in which fresh enzyme was added to partly reacted and heated reaction mixtures, only a 50yC increase over the expected rate was observed.
Significant accumulation of l)rimer would have resulted in substantially greater stimulation.
The small increase may result from an undetectable amount of oligonucleotide present in the reaction mixtures or produced by degradation of polymer during heating. It may also represent protection of the enzyme against heat inactivation by the components of a partly reacted mixture. (b) Direct investigation of the size of the polymer product during unprimed synthesis by Form-T demonstrated that no short chain length material is produced to any significant extent.
No material of intermediate chain length was detected either on gel electrophorcsis (which would reveal rather high molecular weight intermediates) or on ion exchange paper chromatography (which would reveal small oligonucleotide intermediates). Poly A of chain lengthgreater than 700 is detectable aftera small percentage of t.he AD1 has reacted, and as polymerization proceeds, the amount of this material increases. Priming by this product is not expected since it is known that long chain poly A does not serve as a primer for primer-dependent J-l. Zuteus enzyme (12, 28, 29) .6 Thus, unprimcd synthesis by form-T enzyme appears to be processive (24) ; long chain polymers are synthesized one at a time without the accumulation of short chain intermediates. Polymerization by primer-independent 1)olynucleotide phosphorylaseis also proccssivc, as indicated previously in experiments with E. coli enzyme,' and demonstrated here for M. lute1L.s en-2422 Vol. 245,No. 9 The preferred reaction in this instance is the processive synthesis de nova of large polymer. Early experiments (32) also showed that oligonucleotide primer is extended in chain length with primer-independent enzyme. Although the point was not made explicitly at the time, the published data (32) show clearly that under such conditions the disappearance of pApApA primer in the presence of UDP can be largely accounted for by the production of (PA)~~U and (~A)~(pu)t.
Although ADP polymerization by Form-I in the presence of oligonucleotide occurs essentially without primer utilization, oligonucleotide may stimulate the rate of Pi release up to 2-fold (6). The mechanism of this stimulation would appear to be different from that operstin g with primer-dependent enzyme. It may be related to previous observations on the stimulatory effect of oligonucleotides of structure (Np), on primer-dependent A. agilis enzyme (32) ; because of the 3'-phosphate group such oligonucleotides cannot be incorporated into polymer and do not serve as primers for IV. Zuteus Form-T (1, 2, 8). The small stimulation of Form-I by oligonucleotides may also. be related to recent observations of Godefroy,Cohn,. Very early in the polymerization of nucleoside diphosphates by primer-independent E. coli enzyme a lag phase is observed which can be overcome by very low concentrations of oligonucleotides as well as long chain poly A. With both enzyme forms chain growth by addition of new nucleoside monophosphate residues to an added primer appears to be nonprocessive, in keepin g with the nonprocessive phosphorolysis of oligonucleotides (27). This is the preferred reaction with Form-T, while synthesis de nouo is preferred with Form-I, even when primer is supplied.
A practical consequence of this finding is that Form-T is more suitable for the synthesis of oligonucleotides (33) and block copolymers (26). The question as to why Form-I makes such inefficient use of added primer is of interest for future work.
The affinity of the two enzyme forms for oligonucleotides, when the latter are substrates for phosphorolpsis, is identical (27, 34). The apparent, contradiction might be resolved if the manner of oligonucleotide binding were different in the forward and back reaction as described in Scheme 1 of Reference 34. In the forward reaction the oligonucleotide would be bound so that the 3'-hydroxyl nucleoside can act as an acceptor of incoming nucleotide residues, that is, in the number 2 subsite (34). In phosphorolysis. however, the 3'-hydroxyl nucleoside would be bound to subsite number 1, the subsite occupied by ADP in polymerization according to the published scheme (34). The apparent contradiction may, however, have another less obvious explanation.
Thus, the affinity of an oligonucleotide in phosphorolysis is independent of the concentration of Pi, suggesting that the binding may be the same in both phosphorolysis and polymerization.
An interesting generalization from the data presented here is that polymer synthesis de nooo is always processive, while synthesis by additioll to primer is not. During synthesis de nova, the growing polymer must go through stages where n equals values equivalent to those observed for oligonucleotide intermediates in primed synthesis (i.e. n = 3 to 8). Such compounds are released free in primed synthesis but, are never detectable in synthesis de no2ro. Clearly there is an unknown mechanistic difference between initiation of a new chain and extension of added chains that accounts for this. This conclusion is supported by the fact, that synthesis by extension of primer never becomes processive and the resulting chains never reach the length of those made de novo.