A 15-kDa interferon-induced protein is derived by COOH-terminal processing of a 17-kDa precursor.

An interferon-induced 15-kDa protein is synthesized from a precursor of higher molecular weight; the precursor contains 165 amino acids (17 kDa), whereas the stable product (15 kDa) contains 156 amino acids. The stable 15-kDa form is derived from the precursor 17-kDa form by the removal of eight amino acids from the COOH terminus and the methionine from the NH2 terminus. The existence of the precursor 17-kDa protein can be demonstrated after brief periods of in vivo labeling with [35S]methionine and by translation of mRNA in vitro.

* This paper is Contribution 4572-B from E. I. du Pont de Nemours and Co., Inc. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
D. Fahey and E. Knight, Jr., unpublished observations. well known post-translational removal of the Met-1 had occurred (3). However, more extensive NH2-terminal processing seemed unlikely, because no NHz-terminal signal peptide sequence is predicted by either the cDNA (1) or genomic DNA (2). Hence, we chose to investigate the possibility of posttranslational COOH-terminal processing. We report here that the 15-kDa protein isolated from IFN-treated cells is synthesized as a 17-kDa precursor and is rapidly converted to the 15-kDa form by removal of eight amino acids from the COOH terminus and the NHz-terminal methionine.

EXPERIMENTAL PROCEDURES
Materials-All reagents and materials were of the highest quality available and were purchased from standard suppliers. IFN-fl, specific activity 2 X 10' units/mg, was prepared as described (4).
Cyanogen Bromide Cleavage of the 15-kDa Protein-A solution of 1.7 nmol of reversed-phase HPLC-purified 15-kDa protein (5) in 0.10 ml of 70% formic acid was treated with 5 mg of cyanogen bromide for 24 h in the dark at 25 "C. The solution was evaporated to dryness twice in vacuo; the residue was dissolved in 10% formic acid, and a portion was fractionated by reversed-phase HPLC on a 4.5 X 50-mm Vydak C4 column equilibrated with 0.1% trifluoroacetic acid. Peptides were eluted with a gradient of acetonitrile, 0-60% (60 min), 0.5 ml/min, 25 "C, and were detected by monitoring at 220 nm.
Amino Acid Sequencing and FAB Mass Spectroscopy-Amino acid sequencing was performed with an Applied Biosystems gas phase sequenator, model 470. FAB mass spectroscopy was performed using a JEOL HXlOOHF double focusing magnetic sector mass spectrometer scanning the mass range 100-1500.
Cell Culture-Daudi cells were grown as previously described (5). Human osteosarcoma, human WISH, and human diploid fibroblast cells were grown in minimal essential medium supplemented with 10% fetal calf serum.
In Vivo and I n Vitro Labeling of Proteins-For long-term labeling were incubated at 37 "C for designated time periods, and the labeling was terminated by adding each 200 pl of cells to 10 ml of ice-cold phosphate-buffered saline. For a chase, 800 pl of cells were labeled as above for 5 min, then 10 pl of 10 mM methionine was added, and the 800 p1 of cells were diluted to 15 ml with RPMI 1640 growth medium containing 15% fetal calf serum and incubated at 37 "C for the designated chase times. To terminate the chase, 5 ml of the labeled cells were added to 10 ml of ice-cold phosphate-buffered saline. Cytoplasmic extracts were prepared as described above and used immediately for immunoprecipitation or stored at -70 "C. Proteins were labeled in vitro by translating poly(A+) RNA from IFN-treated cells in rabbit reticulocyte and wheat germ lysate systems. RNA was prepared by the guanidinium isothiocyanate-cesium chloride method (6), and poly(A+) RNA was prepared from total RNA by chromatography on oligo(dT) (7). Each reaction mixture in 50-pl total volume contained either rabbit reticulocyte or wheat germ lysate, 5 pg of 4520 poly(A*) RNA from IFN-8-treated Daudi cells, and 50 pCi of [36S] methionine. Chases were initiated by adding 5 pl of 10 mM methionine. Protein synthesis was terminated by adding SDS to 0.1% and placing the sample on ice.
[3JS]Methionine-labeled proteins were then analyzed by immunoprecipitation and SDS-PAGE. Immunoprecipitation-Apolyclonal antiserum to homogeneous 15-kDa protein was raised in a New Zealand White rabbit. Four pg of purified 15-kDa protein was initially injected intradermally in complete Freund's adjuvant. Subsequently, 4 pg was injected a t day 21 and at day 36; 20 pg was injected at day 50 and at day 61, all in incomplete Freund's adjuvant. For immunoprecipitations [=S]methionine-labeled cytoplasm from 1 X 10' cells was added to 200 pl of antibody buffer (0.5% Triton X-100, 0.15 M NaCI, 0.005 M EDTA, 0.002% NaN3, 0.05 M TriseHCI, pH 7.4, 1 mg/ml bovine serum albumin) followed by 20 pl of Protein A-Sepharose. After 5 min at 4 "C the Protein A-Sepharose was removed by centrifugation and discarded. To the supernatant 0.1-1.0 pl of antiserum was added. After 30-60 min a t 4 "C, 20 pl of Protein A-Sepharose was added, and the mixture was left at 4 "C for 30 min with periodic mixing. The Protein A-Sepharose was washed three times with antibody buffer and once with 0.01 M Tris.HCI, pH 7.4. Proteins were eluted from the Protein A-Sepharose by boiling for 2 min in loading gel buffer containing 2% SDS and 2% &mercaptoethanol, analyzed by SDS-PAGE on slab gels using the buffer system of Laemmli ( S ) , and then visualized by autoradiography.

RESULTS
Since the molecular weight of the IFN-induced protein predicted from the genomic and cDNAs (17,890) did not agree with that of the purified protein estimated by SDS-PAGE (about 15,000), we suspected that post-translational modification to reduce the mass from 17 to 15 kDa might be occurring. We excluded amino-terminal processing from consideration, since the cDNA and genomic DNA do not predict a signal sequence prior to the NH2 terminus of the 15-kDa protein (1,2). Furthermore, the NH2-terminal sequence predicted from the DNA is the same as that found for the protein with the exception that the protein does not contain the initial methionine (1). Another possible mechanism for reducing the molecular weight, although not as frequently studied as NH2-terminal processing, is COOH-terminal processing, i.e. removal of COOH-terminal amino acids post-translationally. It therefore became essential to determine if the COOH-terminal amino acid sequence of the protein is different from that predicted from the genomic and cDNAs. We used two techniques to determine the carboxyl terminus of the 15-kDa protein: (a) amino acid sequencing of the peptides generated by cyanogen bromide cleavage; and (b) fast atom bombardment mass spectrometry (FAB-MS) to determine the molecular mass of the cyanogen bromide-generated peptides. Fig. 1B shows the amino acid sequence of the 15-kDa protein deduced from the genomic and cDNAs and confirmed by amino acid sequencing of peptides constituting 80% of the protein (1). This revised sequence contains 20 more amino acids at the COOH terminus than deduced originally from the cDNA (1). Cyanogen bromide cleavage at the methionines should generate four peptides with the COOH-terminal peptide containing 15 amino acids if there is no processing. Cyanogen bromide cleavage was performed; the peptides were separated by reversed phase HPLC (Fig. 1A) and subjected to FAB-MS analysis. Peak fraction 1 showed a protonated molecular ion (M+H)' of 785 daltons, which is identical to that calculated for a COOH-terminal peptide beginning at Asn-151 and ending with Gly-157 (Fig. IC). Amino acid sequencing of this peptide gave the sequence: Asn-Leu-Arg-Leu-Arg-Gly-Gly. This confirms that the 15-kDa protein isolated from IFN-induced cells ends with glycine and contains 156 amino acids (157 minus Met-1).
Since the isolated 15-kDa protein contains nine amino acids fewer than the 165 predicted from the DNA, it became of interest to determine if a larger precursor could be detected. In order to detect the 15-kDa protein with maximum sensitivity, a polyclonal antiserum was prepared in a rabbit against homogeneous 15-kDa protein, and the data in Fig. 2 show its specificity by immunoprecipitation. The antiserum immunoprecipitates a 15-kDa protein from IFN-@-treated Daudi, diploid fibroblast, WISH, and osteosarcoma cells. Pulse-chase experiments were performed with IFN-8-treated Daudi cells using the antiserum to detect by immunoprecipitation "15-kDa-like" proteins of larger molecular weight (Fig. 3). After 2 min of labeling with [3sS]methionine, proteins of 15 and 17 kDa were observed. Furthermore, the amount of the two proteins increased after 5 min of labeling (Fig. 3, lane 4). A chase was initiated after 5 min of labeling by the addition of an excess of nonradioactive methionine. The 17-kDa protein disappeared after 5 min of the chase (Fig. 3A, lanes 6 4 , whereas the 15-kDa protein increased and then became constant after 5-10 min of the chase. These data indicate that the 15-kDa protein is first synthesized as a 17-kDa precursor that is rapidly converted to the stable 15-kDa form. The 17-kDa protein is, we believe, the 165-amino acid protein predicted from the DNA, and the stable 15-kDa protein is the result of the removal of eight amino acids from the COOH terminus and one methionine from the NH, terminus of the 17-kDa precursor. The relative timing of these two events has not been determined.

A 15-kDa Protein Is Derived by COOH-terminal Processing
Further evidence for a 17-kDa precursor was obtained from in uitro translation of mRNA from IFN-&treated Daudi cells. Fig. 3B shows that only the 17-kDa protein is synthesized in 5 min by the rabbit reticulocyte system. The 15-kDa protein appears after a 25-min chase, increases after a 55-min chase, and remains stable thereafter (Fig. 3B, lanes 8-12). The 17-kDa precursor, however, disappears after a 55-min chase. The same result was obtained when mRNA from IFN-/?-treated diploid fibroblast cells was translated by the rabbit reticulocyte system (Fig. 3B, lanes 13). Both results demonstrate that the reticulocyte system can synthesize and process the 17-kDa protein. Although the wheat germ system also synthesizes the 17-kDa protein, this system cannot process it (Fig. 3B, lanes 4-6).

DISCUSSION
The results of the in vivo pulse-chase and in vitro synthesis experiments, taken together with the data showing that the 15-kDa protein isolated from Daudi cells has eight fewer amino acids at its COOH terminus than predicted from the DNA, strongly suggest that the 15-kDa protein is derived by COOH-terminal processing of a 17-kDa precursor. COOHterminal processing of proteins has only recently been reported (9), and its biological role is still unclear. It has been suggested that COOH-terminal sequences may serve as a signal for the sorting of proteins within a cell, possibly within the lysosome (9,lO). Furthermore, a specific COOH-terminal sequence, Lys-Asp-Glu-Leu, has been shown to be involved in the accumulation of proteins in the lumen of the endoplasmic reticulum (11). The role of COOH-terminal processing of the 17-kDa protein is unknown, because the biological function of the 15-kDa protein remains unknown. Recently an interesting sequence homology has been found between the 15-kDa protein and ubiquitin (12). When the sequences of two ubiquitin molecules are aligned head to tail along the 156-amino acid sequence of the 15-kDa protein, several regions of similarity are evident. Furthermore, the same authors report isolation of an IFN-induced ubiquitin cross-reactive protein whose mass is 15 kDa (12). They suggest that this similarity and cross-reactivity indicate that the 15-kDa protein is a functionally distinct isoform of ubiquitin. Further studies will be needed to substantiate this interesting hypothesis.
Our results demonstrate that the 15-kDa protein is derived from a 17-kDa precursor by COOH-terminal processing and removal of the NH2-terminal methionine. This processing alone is probably responsible for the apparent mass change of 2000 estimated by SDS-PAGE. To investigate the biological role of the 15-kDa protein and the enzymes involved in the COOH-terminal processing, larger amounts of the 15-and 17-kDa proteins will be needed than are currently available from cell culture. To accomplish these goals, we have expressed the 17-kDa precursor in Escherichia coli and are modifying its cDNA so as to obtain expression of the 15-kDa protein. Both proteins will then be purified for use in biological studies.