Cell-free Synthesis of Rat Parotid Preamylase*

Poly(A)-containing RNA from rat parotid gland directs the cell-free synthesis of several products in the reticulocyte lysate translation system including a very prominent 58,000-dalton polypeptide which is immunoreactive with anti-alpha-amylase. Purified alpha-amylase has a molecular weight estimated as 56,000 daltons. The 58,000-dalton, cell-free product and alpha-amylase share common peptides as determined by analysis of their limited proteolysis digests. The cross-reactivity and peptide homology suggest that the cell-free product may be a precursor of mature alpha-amylase. While the NH2 terminus of alpha-amylase is blocked, that of the 58,000-dalton product evidently is not, and automated sequence analysis has yielded its partial sequence as: Met-X-Phe-Phe-Leu-Leu-Leu-X-Leu-Ile-X-Leu-X-X-X-X-X-X-X-X-X-Phe-X-X-X-X-X-Ile-X-X-Leu-Phe. The highly hydrophobic nature of the NH2 terminus of the 58,000-dalton, cell-free product suggests that, like other secreted polypeptides, the extra piece may play a role in the transport and secretion of the mature alpha-amylase.

Poly(A)-containing RNA from rat parotid gland directs the cell-free synthesis of several products in the reticulocyte lysate translation system including a very prominent 58,000-dalton polypeptide which is immunoreactive with anti-a-amylase. Purified a-amylase has a molecular weight estimated as 56,000 daltons. The 58,000-dalton, cell-free product and a-amylase share common peptides as determined by analysis of their limited proteolysis digests.
The cross-reactivity and peptide homology suggest that the cell-free product may be a precursor of mature a-amylase. While the NH2 terminus of a-amylase is blocked, that of the 58,000-dalton product evidently is not, and automated sequence analysis has yielded its partial sequence as: Met-X-Phe-Phe-Leu-Leu-Leu-X-Leu-Ile-X-Leu-X-X-X-X-X-X-X-X-X-Phe-X-X-X-X-X-Ile-X-X-Leu-Phe.
The highly hydrophobic nature of the NH2 terminus of the 58,000-dalton, cell-free product suggests that, like other secreted polypeptides, the extra piece may play a role in the transport and secretion of the mature a-amylase.
Rat parotid gland, highly specialized for the synthesis of a few proteins (l), provides a system amenable to study the secretion of enzymes. The major protein component in the gland is u-amylase which accounts for about 30% of the total proteinaceous content of its secretory granules. This protein, consisting of a single polypeptide chain with a molecular weight of 56,000 (2) was shown to be synthesized on membrane-bound polyribosomes of rough endoplasmic reticulum (3). The newly synthesized protein is transported from the rough microsomes to the smooth endoplasmic reticulum, then to the Golgi apparatus, and is finally sequestered in secretory vesicles (4,5). Several other secretory proteins have been shown to be synthesized and secreted by analogous mechanisms (6,7). The rate of secretion of Lu-amylase depends upon hormonal stimulation of the /I-adrenergic receptors and is triggered by catecholamines (8,9). Recent reports indicate that most secretory proteins are synthesized as precursors with a hydrophobic extension peptide at the NH2 terminus.
We have investigated the synthesis of cY-amylase in the rat parotid gland. This paper describes the isolation of mRNA from rat parotid coding for a polypeptide sharing common sequences with cu-amylase as judged by immunoprecipitation with antiamylase and peptide patterns. The primary translation product has a molecular weight larger than authentic (Yamylase and the partial sequence of the isolated polypeptide reveals a cluster of hydrophobic amino acid residues at the NH2 terminus. This indicates that the main cell-free translation product of rat parotid mRNA represents a precursor to cu-amylase similar in its NHz-terminal sequence to precursor of other secretory proteins previously described.

MATERIALS AND METHODS
[""SlMethionine (500 to 1000 Ci/mmol) and [4,5-  , and centrifuged for 3 min at 500 x g. The sample layer was removed by aspiration, the tube was cut above the pellets, and the pellets were washed three times with PBS and finally dissolved either in 2% NaDodSOI, 1% mercaptoethanol, and 10% glycerol, when counted or subjected to gel electrophoresis, or in 40% acetic acid for samples destined for sequence analysis.
Products Analysis-The postmitochondrial parotid extract containing [""Slmethionine-labeled amylase and the putative precursor synthesized in reticulocyte lysate and also labeled with ["'S]methionine were electrophoresed on NaDodSOl/polyacrylamide gels (10%). The gels were dried and autoradiographed, and the corresponding bands were excised and re-electrophoresed overnight into the dialysis bag at 100 V. The bag content was dialyzed against water, the precipitate was removed by centrifugation, and the supernatant was lyophilized.
The products were redissolved in water and 5 pl (5000 cpm) were mixed with 15 ~1 of buffer containing 0.1 M Tris-HCl (pH 6.8), 0.5% NaDodSOd, 10% glycerol, and 0.0001% bromphenol blue. The samples were heated to 100°C for 2 min and digested for 20 min at 37"C, with either chymotrypsin or papain, at the enzyme concentration as indicated in the legend to Fig. 3. The digestion products were then analyzed by NaDodSOd-polyacrylamide gel (15%) electrophoresis according to Cleveland et al. (28). Partial Sequence Analysis-The total cell-free products that were labeled with one radioactive amino acid at a time were immunoprecipitated, as described above, and subjected to automated Edman

RESULTS
Isolation of mRNA-Analytical immunoprecipitation of rat parotid polysomes with antiamylase" had indicated that the level of mRNA specific for cu-amylase is augmented approximately lo-fold when fasted rats were stimulated by injection of catecholamines, as compared to normal or fasted rats. Preliminary attempts to extract the RNA from the isolated parotid polysomes were unsuccessful. Rat parotid gland contains a considerable level of RNase (25, 30) which poses a serious obstacle to the isolation of an intact RNA. In the intact cell, however, the compartmentalization of the cytoplasm possibly limits the damage wrought by RNase. It was found that the method of Kirby (23) was very efficient in overcoming the effect of RNase. Accordingly, the intact, frozen parotid glands from rats injected with isoproterenol were thawed directly into the denaturative medium and total RNA was extracted by the phenol/cresol method. The enrichment for poly(A)-containing RNA was achieved by two subsequent cycles of oligo(dT)-cellulose chromatography. Twice bound poly(A)-RNA represented approximately 1.7% of the total RNA. Translation of mRNA and Nature of Cell-free Products-Addition of total rat parotid poly(A)-RNA to mRNAdependent rabbit reticulocyte lysate cell-free systems resulted in 3-to 25-fold stimulation of the radioactive amino acids incorporation into proteins, depending upon the labeled amino acids used. The ["'Slmethionine-labeled cell-free products were fractionated on NaDodS04-polyacrylamide slab gels and analyzed by autoradiography.
One of the most prominent polypeptides synthesized in the system was that with a molecular weight of approximately 58,000 (Fig. lb). Only this polypeptide was specifically immunoprecipitated from the total cell-free products by sequential treatment with goat antiserum prepared against purified cu-amylase, followed by addition of purified rabbit anti-goat antibodies (Fig. Id). Authentic Lu-amylase migrates somewhat faster in this system (Fig. lc), giving an expected molecular weight of 56,000 (2) (Fig. 2). However, it should be remembered that molecular weight determination based on NaDodSOJpolyacrylamide gel electrophoresis are prone to error, especially when one is dealing with glycoproteins, which show a misleadingly high apparent molecular weight. Therefore, the difference in molecular weights between the cell-free corresponding product and a-amylase, which is glycoprotein (2), may be even greater. (39,ooO cpm); C, ["Hlisoleucine (22,000 cpm). Cycle zero represents a blank cycle (without phenyl isocyanate) which was used to wash out potential radioactive contaminants.
Peptide Analysis-The relationship between authentic aamylase and cell-free product immunoprecipitable with antiamylase was further investigated by partial proteolytic digestion. The [3"S]methionine-labeled a-amylase and 58,000dalton cell-free product were eluted from NaDodS04/ polyacrylamide gels (Fig. 3, a and f) and after limited digestion with chymotrypsin or papain, the digests were resolved on NaDodS0,/15% polyacrylamide gels. Autoradiograms show that with papain the in vivo (Fig. 3, d and e) and in vitro (Fig.  3, i and 1) peptide patterns are identical. Limited digestion with chymotrypsin revealed two identical peptides and an additional one which migrates faster in a-amylase (Fig. 3, b,   FIG. 5. Radioactivity recovered at each sequenator cycle from the cell-free products immunoprecipitated with antiamylase as described under "Materials and Methods." A, products labeled with [:'H]leucine (250,000 cpm); B, semilog plot of data corrected for background and out-of-step degradation. c and g, h). This peptide probably represents the NHP-terminal peptide of a-amylase with the extension of an "extra piece." Based on the above data, it was concluded that the 58,000dalton polypeptide in vitro product was identical to a-amylase but contained an additional polypeptide chain of approximately 2,000 daltons and may be alluded to as the preamylase.
Sequence Analyses-To gain more information on the nature of putative preamylase, a partial amino acid sequence of the cell-free products labeled with radioactive amino acids and indirectly immunoprecipitated with antiamylase was determined by automatic Edman degradation. Figs. 3 and 4 show that the release of peaks of radioactivity is associated with certain cycles of Edman degradation of the immunoprecipitated cell-free products. Radioactive peak from the product labeled with [35S]methionine occurs only at Cycle 1 (Fig.  4A); with ["Hlphenylalanine-labeled product the peaks are at Cycles 3, 4, and 22 (Fig. 4B); with ["Hlisoleucine-labeled product the peaks appear at Cycles 10 and 28 (Fig. 4C); products labeled with ["Hlleucine show radioactive peaks at positions 5, 6, 7, 9, 12, 31, and 48 (Fig. 5A). Flat and low background of radioactivity was obtained from sequencer runs (30 cycles) of precursor labeled with [JH]valine, thus showing that this amino acid is not present in the preamylase (not shown). Discrete radioactive peaks were recovered from Stquencer runs of the labeled precursor. In the semilog plot (Fig. 5B), the peaks lay on a straight line, thus showing that they originated from one protein species. The results, sum-Met-X-Phe-Phe-Leu-Leu-Leu-X-Leu-IIe-X-Leu-X-X-X-X-X-X-X-X-X-Phe-X-X-X-X-X-Ile-X-X-Leu-Phe FIG. 6. Proposed structure and partial amino acid sequence of the NH, terminus of rat preamylase. marized in Fig. 6, illustrate the partial sequence of first 35 amino acids of putative precursor to a-amylase. DISCUSSION This report shows that partially purified rat parotid mRNA directs the in vitro synthesis of several distinct polypeptides in a reticulocyte lysate cell-free protein synthesizing system. A considerable fraction of the polypeptide material produced in this system was specifically immunoprecipitated with antiamylase. The amount of selective immunoprecipitate varied from 8 to 30% of the total products synthesized, depending on amino acid used for labeling. The immunoprecipitated material appeared to be homogeneous in size with a molecular weight of about 58,000 which is larger than purified rat parotid a-amylase by 2,000. It was shown by means of limited proteolysis, in NaDodS04-containing buffer, that the peptide patterns generated independently by two proteases with different specificities are almost identical for authentic amylase and the corresponding polypeptide synthesized in vitro. On the basis of this evidence, it is concluded that the immediate translation product of parotid amylase mRNA is a polypeptide somewhat larger than the in vivo secreted, mature Lu-amylase and will be referred to as preamylase.
The fidelity of translation of several mRNAs in the reticulocyte lysate system has been demonstrated.
In most cases the products have an NHz-terminal methionine residue (18,20,32,33), which has been identified as the initiator methionine (20,32,34). Sequence analysis of preamylase reveals a methionine residue at Cycle 1, suggesting that it is the initial translation product of mRNA.
Parotid amylase has a blocked NH2 terminus" and its primary structure has not been published.
At the moment, therefore, it is not possible to align the amino acid sequences of cr-amylase and preamylase and determine the exact size of the extra piece. The partial sequence of preamylase reveals a cluster of hydrophobic amino acid residues at the NH2 terminus; the region within positions 3 to 12 encompasses hydrophobic residues with the sequence Phe-Phe-Leu-Leu-Leu-X-Leu-Ile-X-Leu.
As emphasized by Schechter and Burstein (33), this structural feature is characteristic of the extra pieces found in vitro on the translation product of messenger RNAs coding for a variety of other secretory proteins. It is generally accepted that secretory proteins are synthesized on microsomes. According to signal hypothesis (22), the amino acid extension of 15 to 30 amino acid residues at the NH, terminus (signal peptide) is recognized by membrane receptors and directs the polysomes synthesizing these proteins to the endoplasmic reticulum. The growing nascent chains are vectorially discharged across the microsome membrane to the Golgi area and stored in secretory granules.
Transient signal peptides associated with preproteins have been postulated to be rapidly cleaved by microsomal proteases (10, 35, 36) during each cycle of protein synthesis, while the exported protein is maturated on the other side of the membrane. The processing of preproteins has recently been shown in different systems (10, 20, 34, 35, 36). However, preliminary experiments in our laboratory to demonstrate the conversion of preamylase to amylase using rat parotid microsomes added co-translationally, as the source of specific protease, were unsuccessful. This was probably due to the presence of ribonuclease in the preparation which hampered translation. As " Unpublished observation.
yet, only the protease from rough microsomes of dog pancreas, having a rather broad specificity towards preproteins, could be solubilized with aid of detergents (37). We are currently investigating the use of this system to accomplish preamylase maturation in vitro.
Amylase obtained from cleavage of preamylase in a cell-free system devoid of blocking activity, may have a free NH2 terminus and could then be sequenced providing additional evidences for the size of the extra piece. The in vitro isolation and translation of the mRNA coding for dog pancreatic preamylase has been recently reported (21). Pancreatic amylase, was previously shown to be completely unrelated to parotid enzyme by lack of immunological crossreactivity, different amino acid composition (2) and different peptides pattern (38), although both enzymes possess the same enzymatic activity. It was, therefore, suggested that the two enzymes are different unique molecules and thus the product of distinct genes. The primary sequence of pancreatic amylase has not been established. It will be of interest to compare these two unrelated amylases and to ascertain whether there is homology in their presequences.
Preamylase synthesized in uitro does not display any enzymatic activity for hydrolysis of starch. This activity, if it existed, would most likely be detectable, as corresponding amounts of amylase do show starch hydrolysis (data not shown). The absence of biological activity may well be due to the extra piece. The hydrophobic extension at the NH, terminus can prevent the proper folding of the preprotein, which is prerequisite for acquiring the full biological activity (39). The short lived extra piece is removed prior to maturation, liberating the biologically active molecule in the native conformation. Such a mechanism may play an important role in control of the secretion of enzymes.