Regulation and in Vitro Translation of Messenger Ribonucleic Acid for Cellulase from Auxin-treated Pea Epicotyls*

SUMMARY Polysomal RNA was isolated from pea epicotyls treated with (2,4dichlorophenoxy)acetic acid, and fractionated on oligo(dT)-cellulose to yield poly(A)-containing RNA. This RNA fraction was translated in a wheat embryo cell-free system and found to have more than 90% of the messenger activity in total polysomal RNA. Immunoprecipitation of the translation product by monospecific antibodies to pea cellulases (@-1,4-glucan 4-glucanohydrolase, EC 3.2.1.4) indicated that cellulase was synthesized in this sytem. The immunoprecipitate co-migrated with the buffer-soluble cellulase component in sodium dodecyl sulfate-gel electrophoresis. Buffer-insoluble cellulase was not detected in the in vitro translation products. Fractionation of mRNA from membrane-bound and free polysomes and their subsequent translation indicated preferential synthesis of buffer-soluble cellulase on membrane-bound polysomes. With the above techniques for assaying buffer-soluble cellulase mRNA, a IO-fold increase in the level of this messenger per tissue segment was observed

system and found to have more than 90% of the messenger activity in total polysomal RNA. Immunoprecipitation of the translation product by monospecific antibodies to pea cellulases (@-1,4-glucan 4-glucanohydrolase, EC 3.2.1.4) indicated that cellulase was synthesized in this sytem. The immunoprecipitate co-migrated with the buffer-soluble cellulase component in sodium dodecyl sulfate-gel electrophoresis.
Buffer-insoluble cellulase was not detected in the in vitro translation products. Fractionation of mRNA from membrane-bound and free polysomes and their subsequent translation indicated preferential synthesis of buffer-soluble cellulase on membrane-bound polysomes.
With the above techniques for assaying buffer-soluble cellulase mRNA, a IO-fold increase in the level of this messenger per tissue segment was observed within 48 hours following (2,4-dichlorophenoxy)acetic acid treatment. There was no evidence for pre-existing untranslated message for cellulase in control tissues.
Since there was no delay in the appearance of mRNA for cellulase, compared to a 24-hour lag in the increase of cellulase activity, it is suggested that translational as well as transcriptional controls are exerted on the biosynthesis of cellulase in vivo.
Analysis of the rates of peptide chain initiation and elongation, both in vivo and in vitro, indicated that peptide chain elongation may be rate-limiting during the lag phase of cellulase development.
Treatment of pea epicotyls with auxiu-type growth regulators such as the hormone indoleacetic acid, or the analogue (2,4-dichlorophenoxy)acetic acid, results in a specific increase in cellu-lase activity (l-5). This enzyme activity has now been resolved into two separate components (6). Pure and crude BS cellulases were applied in separate wells and electrophoresis was performed in the above buffer at 10 volts per cm for 120 min at 2". Antiserum was applied on a 2-mm wide strip of filter paper placed between the two wells parallel to the electrophoretic migration and allowed to diffuse for 24 hours in the cold.
The slides were washed, dried, and stained with acid fuchsin (6

RESULTS
Generation of Cellulase after (2,4-Dichlorophenoxy)acetic Acid Treatment-Total cellulase activity increased several fold after (2,4-dichlorophenoxy)acetic acid treatment of intact pea epicotyls (Fig. 1). This increase was comparable to that observed following treatment of the decapitated epicotyl with the hormone indoleacetic acid (l-5).
The activity was distributed almost equally between 1% and UI cellulase (cf. 6) throughout the course of development.
Both enzyme activities showed a distinct lag of about 24 hours, after which they increased at a linear 1021 rate. The parallel kinetics, and identical responses to inhibitors of protein and nucleic acid synthesis (data not shown), imply a common mode of biosynthesis of the two enzymes. Crude microsomes isolated from hormone-treated peas have been shown (7) to support in vitro protein synthesis and generate cellulase activity during incubation.
This suggested the pressence of polysomes containing mRNA for cellulase in the microsomal fraction.
This was due to a marked increase in the net amount of ribosomes in auxin-treated tissue and not due to conversion of monosomes to polysomcs. In order to investigate whether a specific mRNA for cellulase is formed in response to auxin treatment, we have attempted to isolate and translate the messenger RNA in vitro in a hcterologous cell-free system from wheat embryos.
Isolation of Pea mRNA-When total polysomal RNA was isolated from control and (2,4-dichlorophenoxy)acetic acidtreated pea epicotyls as described under "Experimental Procedure," and fractionated on an oligo(dT)-cellulose column, the bulk of the RiYA passed through the column (Fig. 3). Subsequent washing with lower salt concentrations resulted in elution of a small peak of RNA.
The presence of poly(A) in these fractions was tested by hybridization of the unlabeled RNA to [3H]poly(U).
As shown in Fig. 3, most of the poly(A)containing RKA eluted with buffer at low ionic strength.
This was the effective fraction in supporting in vitro protein synthesis in the wheat embryo cell-free system ( Table I).
Translation of Pea mRNA-Characteristics of the wheat embryo cell-free system used for in vitro protein synthesis are given in Table I. It was completely dependent on addition of exogenous mRPI;A, and the process of initiation (301, since the inhibitor of initiation, aurintricarboxylic acid, prevented translation (see also Refs. 24,27 (2,4-dichlorophcnoxy). acetic acid-treated tissue (48 hours) and RNA was extract,ed from polysomes as described under "Experimental Procedure." RNA was dissolved i.n 0.4 M N&l, 10 mnf Tris, pH 7.0, and 0.5(y0 sodium dodecyl sulfate, and applied at A to an oligo(dT)-cellulose column which was equilibrated in the same buffer system. At H, the column was washed with this buffer system minus sodium dodecyl sulfate until no further RNA was eluted.
The small amount of RNA which remained bound to the column was eluted starting at, C with low ionic strength buffer, (0.01 M Tris, pII 7.6). The presence of poly(A) in RNA fraction was detected by hybridization to [3H]poly(U).
The fractions which eluted from the column starting at C had most of the poly(A)-containing 11NA. It, was precipitated with ethanol and used for translation studies. Con trols with rRNA or tRNA did not bind any poly(U). elongation factors are soluble and provided in the supcrnatant in this system (31). The pea ribosomal wash did not stimulate incorporation (cf. 17). Incorporation did not require the addition of unlabeled amino acids indicating that sufficient endogenous lcvcls were present (see also Ref. 24). The poly(h)-cow taining RNA prepared from pea polysomcs rcpresentcd about 90 70 of total messenger activity.
There was no apparent difference iu ability of poly(A) RNA from control and (2,4-G chlorophcnoxy)acctic acid-treated peas to support protein synthesis in the wheat embryo cell-free system. The rate of incorporation was proportional to amount of poly(A) ISA up to approximately 6 pg and the system was capable of reinitiation and release of the complctctl peptidc chains.
Characterization of Translation Product-The prcsencc of cellulase-like material in the translation products directed by pea mItNA was tested with antisera prepared against purified 13s aud HI ccllulases as described uuder "Experimental l'rocetlure." Following translation, reaction mixtures were ceutrifuged to remove ribosomes aud nascent protein, and antisera wcrc added to the supcrnataut.
As shown in Table II, there was very little immunoprecipitation in the product formed in vitro in a system directed by control pea mRNA.
However, in similar prcparations with the use of mIINA from (2,4-dichlorophenoxy)acetic The complete amino acid incorporation system derived from wheat embryos contained 100 Mg of ribosomes plus soluble factors in 400.~1 volume as described under "l:xperimental Procedure." Total pea polysomal RNA was extracted by phenol-chloroform, and fractionated by oligo(dT)-cellulose chromatography as described in Fig. 3. The yield of polg(A)-containing RNA was 2 to 3% of total ILNA. Additions were made to the complete system plus poly(A)-containing pea ILNA. Pea ribosomal wash was prepared from control pea polysomes by extraction with 0.5 nf KC1 and desalting on Sephadex C-25. Pea IDJA was isolnt,cd from control and (L,4-dichlorophenoxy)acetic acid-treated tissue as described in Fig. 3 and poly(A)-containing I<NA was translated in standard wheat embryo cell-free systems as in Table I (with 200 ~1 as the total volume per reaction mixture).
Immunoprecipitation reactions were performed with ,mti-BS and anti-131 rclliilase sera as described in Fig. 4. Aliynots treated with control rabbit serum precipitated 50 to 150 cpm.
This immunoprccipitatc was analyzed by sodium dodecyl sulfate-gel elcctrophoresis along with pure 1% and HI cellulases which had been labeled with [3H]dimcthylsulfate in vitro (6). As showy in Fig. 4, the bulk of the immunoprecipit,ated radioact,ivity co-migrated with 15s ccllulase but no radioactivity coincided with 131 cellulase.
Since the antiserum to 1% cellulase was monospecific, i.e. only one precipitation band was observed in crude and purified 1% preparations (Fig. 5), the material which co-migrated in sodium dodccyl sulfate gels with purified 1% cellulase was considered to be the 1% cellulase component. Poly(A)-containing RNA was isolated from these two sets of polysomes (Fig. 3) and used for in vitro translation in t.he wheat embryo cell-free system. The translation products were allowed to react with antiserum prepared against US cellu- (P) and crude buffer-soluble extract (C) were subjected to electrophoresis on an agarose slide as described under "Experimental Procedure," and antiserum to BS cellulase was applied as a strip between the two samples (boundary indicated). Reaction was allowed to proceed at 4" for 24 hours and slides were dried and stained (6).  Table  I. Immunoprecipitation with anti-BS serum was carried out as in Fig. 4. The lower incorporation per unit mRNA in this  table compared to data in Table  I  Anti-BS cellulase precipitate lase as above. Table III shows that total incorporation supported by the membrane-bound mRNA was comparable to that by mRNA from free polysomes, but there was a much higher percentage of radioactivity (4-fold) precipitated by anti-ES cellulase serum in the product made by the former.
These results indicate that 193 cellulase is preferentially synthesized on membrane-bound polysomes.

Increase in Translatable
Cellulase mRNA-The rapid rise in cellulase levels following (2,4-dichlorophenoxy)acetic acid treatment in viva (Fig. 1) could be a result of de novo synthesis of translatable mRNA for cellulase (induction), differential turnover, or recruitment of pre-existing mRNA into polysomes. The above methods provide a procedure for assaying the amount of functional messenger for cellulase in vitro. Accordingly, total poly(A)-containing RNA was isolated from whole tissue (including polysomal and non-polysomal RNA) at intervals following (2,4-dichlorophenoxy)acetic acid treatment, translated in the wheat embryo cell-free system, and the product was immuno- FIG. 6. Increase of mRNA for cellulase and appearance of cellulase activity following (2,4-dichlorophenoxy)acetic acid treatment. BS cellulase was assayed as in Fig. 1, and total cellular RNA was isolated from whole tissue (without fractionation of polysomes) as described under "Experimental Procedure." The poly(A)-containing fraction was isolated using oligo(dT)-cellulose chromatography and translated in the standard wheat embryo cell-free system. The product synthesized in vitro was allowed to react with antiserum to BS cellulase as in Fig. 4 in order to assay the level of mltNA for this enzyme. The percentage of mRNA for cellulase was calculated by comparing anti-BS-cellulase precipitable counts with the total translation product, assuming equal efficiency for translation of all messages. precipitated with anti-US cellulase serum. The levels of mRNA for US cellulase in relation to IlS cellulase activity are shown in Fig. 6. Following (2,4-dichlorophenoxy)acetic acid treatment, there is a linear increase in amount of translatable mRNA for 1% cellulase during the first 48 hours, whereas cellulase activity develops linearly only after a lag of about 24 hours. '1 he percentage of the total mRh'A which codes for cellulase increases markedly during this first 24.hour period, but it does not result in a net increase of the enzyme at that time.
There is no indication of pre-existing nonfunctional mRNA for ccllulase in zero time control tissue. The fact that a several-fold increase in cellulase mRNA precedes the increase in enzyme may mean that this RNA is also regulated at the level of translation in vivo. Translational Regulation of Protein Synthesis following (2,acetic Acid Treatment-Preliminary data on the rate of protein synthesis in (2,4-dichlorophenoxy)acetic acidtreated pea epicotyls indicated that the total amount of polysomes increased about SO(r, within 24 hours (Fig. 2), whereas the absolute rate of protein symhesis per segment did not in crease proportionately during this period. In order to investigate modulation of protein synthesis at the translational level, the relative rates of peptide chain initiation and elongation were determined in control and (2,4-dichlorophenoxy)acetic acidtreated tissue.
There is little likelihood that initiation of protein synthesis is rate limiting in the intact epicotyl since the percentage of total ribosomes present as polysomes remained very high with or Equal numbers of control (zero time) and (2,4-dichlorophenoxy)acetic acid-treated epicotyls were labeled with [ldC]leucine and [3H]leucine, respectively, as described under "Experimental Procedure." The segments were combined and lH:r4C ratios in supernatant (released protein) and polysome (nascent protein) fractions were measured.
Values for relative rate of peptide chain elongation were calculated according to the method of Palmiter (26) from comparison of the ratios of 3H:11C in released protein to 3H:r4C in nascent protein. without (2,4-dichlorophenoxy)acetic acid-treatment (Fig. 2). No more than 90% ribosomes could be recruited into polysomes in this tissue even when the rate of pcptide chain elongation was deliberately retarded (by treatment with cycloheximide) in an effort to increase the number of ribosomes per messenger.
If the rate of initiation was limiting, then treatment with cycloheximide should cause a marked change in polysome size distribution (cf. 26, 32). In the intact pea tissue, cycloheximide (used at a level which partially inhibits protein synthesis) did not alter the polysome profile, indicating that peptide chain initiation was not rate limiting.
The relative rate of peptide chain elongation in vivo was mcasured by a double labeling ratio method as described by I'almiter (26). Differentially labeled tissue segments from control and (2,4-dichlorophenoxy)acetic acid-treated peas were combined and the amounts of labeled protein in polysomes (nascent) and soluble fractions (released) were compared (see "Experimental l'roccdure").
This method assumes that differences in labeling in the two tissues are not accompanied by major changes in mean molecular weight of the proteins synthesized.
The fact that polysome profiles from control and (2,4-dichlorophcnoxy)acetic acid-treated tissues showed similar size distributions of polysomes (Fig. 2, A and B), implies that the mRKA populations were comparable with respect to length.
Moreover, the mean molecular weight of proteins synthesized by control and (2,4-dichlorophenoxy)acetic acid-treated polysomes in vitro mere very close as measured by sodium dodecyl sulfate gel electrophoresis.* The results (Table 1V) show that (2,4-dichlorophenoxy)acctic acid-treatment causes a transient suppression of about 3070 in the relative rate at which polysomes complete their peptide chains in vivo.
In order to measure translational capacity of pea polysomes in vitro, assays were conducted using wheat supernatant in which concentrations of initiation and elongation factors were nonlimiting (24). Under these conditions, polysomcs primarily complete nascent chains and there is little reinitiation (24). As shown in Table V   Polysomes were isolated from control and (2,4-dichlorophenoxy)acetic acid-treated intact pea epicotyls (as in Fig. 2) and 100 pg were used in a standard amino acid-incorporating system (as in Table I), with supernatant enzymes derived either from wheat or pea as indicated.
The wheat supernatant was used at saturating levels (11,30). The pea supernatant was partially purified as described under "Experimental Procedure" and 100 pg of protein were used per reaction mixture.
This fraction was free of any inhibitor of protein synthesis as found in crude pea supernatant (28) Time after (2,4-dichlorophenoxy) (2,4-dichlorophenoxy)acetic acid. However, when the polysomes were provided with pea supernatant from (2,4-dichlorophenoxy)acetic acid-treated tissue, the ability of this supcrnatant to support pcptide chain elongation showed a transient decline followed by recovery.
This decrease was not due to the presence of inhibitors or to any reduction in ability of supernatant fraction for charging tRKA.
Since initiation was not affected in (2,4-dichlorophenoxy)acetic acid-treated tissue, it is suggested that soluble elongation factors may become rate limiting in peas during the first 24 hours after (2,4-dichlorophenoxy)acetic acid treatment.

Isolation
and in Vitro Translation of Cellulase mRNA--The observation that most putative mRNA from eukaryotic cells (s-14), including plants (14-16), contains poly(A) sequences was used to detect and purify a messenger fraction from auxiritreated peas. The fraction isolated by oligo(dT)-cellulose chromatography was found to contain most of the messenger activity when translated in a wheat embryo cell-free system (Table I). This system contains very little endogenous mRNA activity, and several exogenous messages have been translated in it with complete fidelity (16,33,34). Using pea mRKA, translation in the wheat embryo system proceeded independently of any tissue-specific factor(s), and it was capable of reinitiation and release of completed polypeptide chains (Table I).
Evidence for the synthesis of a specific cellulase protein in this heterologous system was obtained by identification of the translation product with antiserum prepared against purified cellulase. Since the total cellulase (RI + 1%) represented only a small fraction of total protein, several criteria were used to ensure homogeneity of the purified enzymes before preparation of antisera to them (6). Although the translation product was allowed to react with both 1% and RI antisera, only 15s cellulase was detected.
In earlier experiments (T), the cellulase activity which increased in microsomal fractions during protein synthesis in vitro was also buffer-soluble.
This indicates that mRNA for 1% cellulase is present in the polysomes of auxin-treated peas. This is the first mRNii for a plant enzyme to be isolated and translated in vitro. Lack of detection of 131 cellulasc in the in vitro translation products does not necessarily indicate that mRKA for this enzyme was not present or not translated in the system. Kewly formed l%I cellulase may have been insoluble in the reaction medium (which contained only approximately 0.05 hl total salts) and thus removed from the bulk of the soluble translation products along with ribosomes and nascent peptides.
Alternatively, the newly synthesized ISI peptides may have been different from the native molecule and not recognized by 131 antiserum. Concurrent increase of mRSA for 131 and 13s cellulases probably occurs since I# and RI cellulase activities develop with paralleled kinetics (Fig. 1) and there is no precursor-product relationship between them (6). Translation of the mRNA for I<1 cellulase, however, may depend on more complex cellular organization, e.g. a membrane site into which newly completed buffer-insoluble peptides are lodged.
Efforts to locate this messenger and translate it arc in progress.
With respect to the intracellular site for cellulasc biosynthesis, the mRT\;A isolated from membrane-bound polysomes contained most of the 13s cellulase message (Table III).
Like other secreted proteins (35), cellulase may be preferentially synthesized on membranes and transported to the cell wall through Golgi or endoplasmic reticulum vesicles (36). I'rcferential synthesis of another secreted protein (rat growth hormone) in vitro has also been demonstrated (37) with mRNA from membrane-bound polysomes.
Induction of Cellulase nzRiliA-Cellulase activity has been shown to be specifically generated in pea epicotyl tissue as a result of auxin treatment (l-5).
Evidence that this increase in cellulase is due to the appearance of specific mRKA was provided by assaying level of mRSX in a cell-free system for translation.
In order to rule out the possibility that a nonfunctional reserve of mRKA (not in polysomes) pre-existed in untreated tissue, mRNA was prepared from total cellular RSA, and poly(A)-containing RNA was tested for ability to support in vitro synthesis of cellulase. Fig. 6 demonstrates that at zero time there was relatively little mRTC'A for cellulase in untreated tissue, and that the messenger level increased almost linearly for at least 48 hours following (2,4-dichlorophenoxy)acetic acidtreatment.
Unlike some systems, e.g. ovalbumin induction (19) where the development of ovalbumin mRKA and the protein are coincident, a specific increase in cellulase mRNA occurs before the appearance of cellulase activity.
Mfects of inhibitors of Rn'A and protein synthesis in vivo (l-3) showed that both RSA and protein synthesis are essential for development of cellulase activity.
Thus present data is consistent with the contention that cellulase mRT\;A is synthesized de novo as a result of (2,4-dichlorophenoxy)acetic acid treatment.
Regulation of Cellulase Synthesis following Auxin Treatment-One possible explanation for an increase in mRNA for cellulasc before the increase in enzyme activity following (2,4-dichlorophenoxy)acetic acid-treatment ( Fig. 6) is inefficiency in the translation mechanism during the early period (0 to 24 hours). The initiation steps in translation did not appear to be rate limiting during that time since the polysome levels were very high (80 to 90y0) with or without auxin treatments (Fig. 2). However, analyses of the rate of peptide chain elongation both in vivo (Table IV) and in vitro (Table V) indicated that a transient reduction of 30 to 407, occurred during the lag period of cellulase development.
Elongation rates returned to normal (zero time) levels when cellulase activity began to rise linearly (24 to 48 hours).
Reasons for this temporary reduction in rate of translation are not fully understood.
The available evidence (Table V) suggests that polysomes themselves are equally active in control and (2,4-dichlorophenoxy)acetic-acid-treated tissue when supplemented with exogenous soluble factors from wheat supernatant, but similar factors from pea appear to become rate limiting.
Hormonal effects on translation have also been observed in other plant tissues, e.g. in cotton embryos, where the message for carboxypeptidase is available but not translated in the presence of the hormone abscisic acid (38,39).
Since the appearance of cellulase in auxin-treated peas was detected by assaying for activity only and not for enzyme protein, the interpretation of translational controls in this system should be treated as tentative.
Other mechanisms may be regulating cellulase biosynthesis in this system. For example, differential turnover rates of mRNA or enzyme, or both, could operate during development in such a way that net mRNA levels increase before enzyme.
Hormones have been shown to affect polysome levels (40,41) and changes in bound ribonuclease activity (42) during growth in the epicotyl.
The availability of ribosomes to attach to mRNA could also be a limiting factor if levels of mRNA increase before the number of ribosomes.
Finally, little is known about compartmentation of functional messenger in eucaryotic tissue.
The present study shows that a specific mRNA for ccllulase increases as a result of auxin treatment, and the process appears to be regulated at both transcriptional and translational levels.