Purification and Characterization of a 19-Kilodalton Intracellular Protein AN ACTIVATION-REGULATED PUTATIVE PROTEIN

Activation of protein kinase C in T cells results in rapid phosphorylation of a 19-kDa intracellular pro- tein termed 19K. We report the purification of 19K from human peripheral T cells and an internal 20- amino acid sequence determined from this protein. It is shown that 19K is a novel cytoplasmatic protein which is phosphorylated in vitro by partially purified protein kinase C. 19K-specific antibodies, raised by immunizing rabbits with purified protein, were used to show that the 19K is expressed, and phosphorylated in response to protein kinase C activation, in several cellular systems. These antibodies were also used to precipitate 19K from both [3SS]methionine and 32Pi- labeled T cells. The data showed that 15 min of phorbol ester treatment has no effect on the rate of 19K synthesis but results in induction of 19K phosphorylation. However, we demonstrate, by Western blot analysis, that expression of 19K in primary peripheral T cells increased at least lo-fold over a period of 4 days after activation. The increase in 19K expression


Activation
of protein kinase C in T cells results in rapid phosphorylation of a 19-kDa intracellular protein termed 19K. We report the purification of 19K from human peripheral T cells and an internal 20amino acid sequence determined from this protein.
It is shown that 19K is a novel cytoplasmatic protein which is phosphorylated in vitro by partially purified protein kinase C. 19K-specific antibodies, raised by immunizing rabbits with purified protein, were used to show that the 19K is expressed, and phosphorylated in response to protein kinase C activation, in several cellular systems. These antibodies were also used to precipitate 19K from both [3SS]methionine and 32Pilabeled T cells. The data showed that 15 min of phorbol ester treatment has no effect on the rate of 19K synthesis but results in induction of 19K phosphorylation. However, we demonstrate, by Western blot analysis, that expression of 19K in primary peripheral T cells increased at least lo-fold over a period of 4 days after activation.
The increase in 19K expression correlates with initiation of DNA synthesis, and in proliferating T cells 19K comprises approximately 0.2% of total cytoplasmatic protein.
Thus, 19K is a novel putative protein kinase C substrate which is subject to activation associated up-regulation in human T cells.
Activation of protein kinase C has been implicated in the regulation of many biological processes including the regulation of T lymphocyte activation (for review, see Refs. 1 and 2). One means of studying regulation of this kinase is to analyze specific phosphorylation events of its cellular substrates. We have recently described activation of protein kinase C in human T cells after stimulation with synthetic diacylglycerol or via antibodies directed against either the CD2 antigen or CD3 complex (3)(4)(5). In these reports, protein kinase C activation was evaluated by the analysis of both an 80-kDa (termed 80K) and a 19-kDa (termed 19K) intracellular substrate. The 80K protein in T cells shares all of a number of biochemical features with a 80-kDa protein kinase C substrate originally described in fibroblasts, which suggests that these proteins are analogous (5)(6)(7)(8)(9)(10)(11) tion suggested that 19K may also be a substrate for protein kinase C (3,4). We now report the purification of 19K from human peripheral T cells and 20 residues of the internal amino acid sequence of this protein. The data suggest that 19K is a novel cytoplasmatic protein that can be used as a protein kinase C substrate in uitro. Immunoprecipitation with specific antibodies revealed that the 19K is expressed in several cellular systems, and is phosphorylated in response to protein kinase C activation. Most interestingly, expression of 19K increased at least lo-fold after activation of primary peripheral T cells. In the resulting proliferating T cell population, the 19K substrate constitutes almost 0.2% of the cellular proteins.

Purification of 19K
itates were centrifuged (20 min, 12,000 rpm), dried under vacuum, and boiled for 5-10 min in sample buffer (2.5% SDS, 125 mM Tris-HCI, pH 6.8,8% glycerol, and 10% 2-mercaptoethanol). Samples were analyzed by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) or by two-dimensional SDS-PAGE as previously described (12). Slab gels were fixed and stained overnight (50% methanol, 10% acetic acid, and 0.1% Coomassie Brilliant Blue R-250), dried down onto paper, and subjected to autoradiography. Alkali-stable phosphoproteins were revealed by rehydration of dried gels followed by treatment of the gels with 0.9 M KOH at 56 "C for 90 min (6,13 Phosphoproteins-Peripheral human T cells, quiescent after 17 days of IL-2-dependent growth (3,17), were labeled with 32Pi for 6 h. After 15 min exposure to PDBu or solvent alone, cells were harvested. The heat-resistant Triton X-100-soluble fraction, prepared as described under "Experimental Procedures," was collected and equal quan-tities of acid-precipitable counts were analyzed on two-dimensional SDS-PAGE.
It is shown in Fig. 1 (left panels) that three distinct spots, termed a-c, are induced in the presence of PDBu. These spots have been partially characterized in a previous report, in which we present evidence that spots a-c represent various serine-phosphorylated forms of a 19-kDa protein termed 19K (3). This serine phosphorylated protein shows a remarkable resistance to alkali-treatment ( Fig. 1, right panels), and the phosphorylated 19K band can be clearly resolved and quantified after treatment of SDS-PAGE with 1 M KOH (3,4). Given that alkali resistance of phosphoproteins is normally due to phosphorylation of threonine or tyrosine residues, this result may appear unexpected, but several other reports have demonstrated alkali-resistant serine phosphorylation (13,18,19).
In parallel with 19K phosphorylation we also observe concomitant appearance of an alkali-stable 80-kDa phosphoprotein (p1 4.8) after PDBu treatment (termed 80K, Fig. 1). Previous reports have suggested a quantitative correlation between alkali-stable 19K and 80K phosphorylation after stimulation of T cells with phorbol esters or mitogenic T cell receptor-stimulating antibodies (3)(4)(5). Purification Strategy for the I9K Substrate--""P-Labeled proteins corresponding to spots a-c were excised from a twodimensional SDS-PAGE and eluted. Partial digestion with protease V8 of the respective protein reveals 3 distinct proteolytic fragments in both spots a and b (17.5, 15.0, and 13.0 kDa, Fig. 2 Lowerpanel, autoradiograph of protease V8 fragments of "'I-labeled spots a and b. Partially purified 19K, prepared as described under "Experimental Procedures," was radiolabeled with Na""I.
The labeled material was separated bv two-dimensional SDS-PAGE and proteins were eluted and analyzed as above.
also obtained after digestion of spot c, but the data was less clear since the yield from this spot was very low (data not shown). Since each phosphate group added onto a small polypeptide, such as 19K, significantly decreases the p1 (20), it seems likely that spots a-c represent the same protein in differently phosphorylated forms. The first step in our purification strategy was to produce an '""I-labeled 19K protein which could be used as a tracer preparation to identify 19K containing fractions during the subsequent large scale purification.
This was nessecary for two reasons: (i) we did not have an assay to detect the 19K protein and (ii) the abundance of the protein was too low to allow identification of 19K by staining of SDS-PAGE. To prepare a '""I-labeled 19K tracer, "'P-labeled proteins from PDBu-treated T-blasts were fractionated as described under "Experimental Procedures," and 19K-enriched material was '? labeled. This material was thereafter separated by twodimensional SDS-PAGE and several spots at relevant positions were excised after autoradiography.
Protease V8 digestion of 'e"I-labeled material eluted from 2 distinct spots yielded fragment sizes that were similar to those obtained with "'P-labeled 19K (Fig. 2, compare upper and lowerpanels). The data shows that the 17.5-and 13.0-kDa protease V8 fragments can be detected in '""I-labeled materials, while a 15.8-kDa band was detected instead of the 15.0-kDa band in '"P-labeled 19K. However, alkaline phosphatase treatment of '"'I-labeled spot a and b, followed by isoelectric focusing, strongly indicated that spots a and b are differently phosphorylated forms of the 19K protein. Thus, the p1 of both spots became identical after dephosphorylation and was increased to around pH 7 (data not shown). The dephosphorylation experiment taken together with the data shown in Fig. Purification of 19K 2, suggested that lZ51-labeled spots a and b were likely to correspond to the 32P-labeled spots a and b. Large scale purification was initiated by mixing dephosphorylated ?-labeled spots a and b with Triton X-lOO-solubilized T cell blasts. Purification was thereafter obtained by employing the strategy summarized in Table I and described under "Experimental Procedures." The analysis of various fractions was initially based on the assumption that 19K would co-purify with the 'Y-labeled 19K tracer, which could be simply tracked by following the radioactivity. The final purified product was a 19-kDa protein which was subsequently used to immunize a rabbit. The resulting immune serum, that is described in the next section, was thereafter used to quantitate the yield of the 19-kDa protein during purification by quantitative Western blot analysis. It is shown in Table I that the 19-kDa protein was purified 530-fold and the final product appeared homogeneous by both SDS-PAGE and HPLC reverse phase analysis. Thus, this 19-kDa protein constitutes almost 0.2% of the Triton X-lOO-soluble protein in a population of actively proliferating T cell blasts. Moreover, the 19-kDa protein was quantitatively recovered in the cytoplasmic fraction after hypotonic lysis of cells, and we have no evidence so far for nuclear or membrane association (data not shown).
The Purified 19-kDa Protein Is Identical to 19K- Fig. 3A shows the 19-kDa protein after either FPLC Mono-S chromatography (lane 1) or after the final purification step on a HPLC reverse phase column (lane 2). This silver-stained gel shows that only one band at 19-kDa is resolved after the final purification step. After lZ51-labeling of this polypeptide it could be demonstrated that the purified product yields the same protease V&specific fragments as the 'Y-labeled 19K tracer (17.5, 15.8, and 13.5 kDa, data not shown).
Since 19K is phosphorylated in uivo after activation of protein kinase C, we performed in uitro phosphorylation of the purified 19-kDa protein in the presence of partially purified protein kinase C. Fig. 3B shows that the 19-kDa protein could be phosphorylated in vitro in the presence of this protein kinase C preparation and the appropriate cofactors (compare lanes 1 and 2). As a control the protein kinase C preparation was also incubated without the 19K protein which did not result in detectable phosphorylation (lanes 3 and 4). In other experiments we compared the phosphorylation levels of the 19-kDa protein with histone III-S and the result indicated that the 19-kDa protein was almost as good a protein kinase C substrate in vitro as histones (data not shown).
To provide further evidence that the purified 19-kDa protein is identical to the 19K protein kinase C substrate, an antiserum raised against the 19-kDa protein was employed. Fig. 4A shows that this antiserum only recognized a 19-kDa band in Western blot analysis of the purified material. Moreover, immunoprecipitation from [3"S]methionine-labeled cells shows that only one band is specifically precipitated (Fig. 4B). It is also shown in the same figure that 15 min of PDBu treatment does not influence [35S]methionine incorporation into this 19-kDa protein. Immunoprecipitation from 32Pilabeled cells demonstrated that the same 19-kDa protein is phosphorylated in response to PDBu (Fig. 4C). Most importantly, this phosphorylation event is resistant to alkali-treatment of gels (Fig. 40). Thus, these experiments confirm that the purified 19-kDa protein is identical to 19K in addition to demonstrating the specificity of the antiserum.
Amino Acid Sequence Analysis of 19K-During initial attempts to determine a NH*-terminal amino acid sequence of 19K it was found that the NH2 termini was completely blocked. This was also the case with the 19K preparation obtained using a milder purification protocol, involving hypotonic lysis rather than Triton X-100 solubilization. We therefore cleaved 25 wg of 19K with cyanogen bromide and separated the resulting peptides by reverse phase chromatography. Three distinct 19K peptides could be isolated: a 13-kDa fragment and two small fragments of approximately l-3 kDa (accurate sizing of these two peptides was not obtained since they were poorly resolved by the IO-20% SDS-PAGE system employed). The NH, termini of the 13-kDa fragment was found to be completely blocked which suggests that this fragment is derived from the NH:! termini of 19K. One of the small peptides, presumably located at the COOH-terminal end, could, however, be sequenced with the expected yield (Fig. 5). A computer search with the resulting 20-amino acid sequence revealed no significant homology with any other protein contained in the EMBL or the Swiss gene data base. A remarkable feature of this peptide sequence is that 65% of the amino acids have either a positive or negative charge. There is also a striking abundance of lysine and glutamic acid (4 out of 20 in each case).

Phorbol Ester-induced
Phosphorylation of 19K in Transformed Cell Lines-The experiments presented above have all been performed on activated human T cells which have been maintained in IL-2. The 19K-specific antiserum described in the present report will make it possible to evaluate 19K expression and phosphorylation in essentially all types of cell lines. In the experiment shown in Fig. 6, 19K was immunoprecipitated from 3 different 32Pi-labeled cell lines: the leukemic T cell line Jurkat, the Burkitt lymphoma-derived B cell line Daudi, and the cervical adenocarcinoma cell line HeLa. The data demonstrate that phosphorylated 19K can be detected in all three cell lines and that the level of phosphorylation increases after PDBu treatment. The results also suggest that the phosphorylation level of 19K in the absence of PDBu is higher in these transformed cell lines than in peripheral T cells (compare Fig. 6 with Fig. 4). This elevation of the constitutive 19K phosphorylation level is consistently observed and it may reflect differences in basal protein kinase C activity in various cellular systems. The main conclusion that can be drawn from Fig. 6, however, is that the 19K substrate is expressed and phosphorylated in response to a protein kinase C activator, in several cellular systems. This conclusion is further supported by immunoprecipitation experiments employing diploid fibroblasts and the HepG2 hepatoma cell line (data not shown).

Activation-associated
Expression of 19K in Primary Peripheral T Cells-It is shown above that 19K constitutes about 0.2% of the detergent-soluble proteins in exponentially grow-  (Table I). To study the expression of 19K during the course of T cell activation, peripheral T cells were isolated and stimulated with an anti-CD3 antibody in combination with IL-2. Fig. 7 demonstrates that these primary cells have a characteristic two-day lag before S-phase progression can be detected by [3H]dThd uptake (17). At the indicated day after activation, cell lysates were prepared and analyzed for 19K content by Western blotting. The result shows that the expression of 19K increases lo-fold after 4 days of activation and that this increase parallels S-phase progression (Fig. 7). As a control the same blot was also probed with anti-triose-phosphate isomerase antibodies and the data shows that this "housekeeping" enzyme remains relatively constant during the course of activation. We conclude from this experiment that the level of 19K expression in primary T cells correlates with cellular activation.

DISCUSSION
The purification strategy in this report involved an initial analytical scale purification and preparation of an '251-labeled 19K tracer. Our assumption was that the 19K protein would co-purify with this radioactive tracer during large scale purification. Identity of purified 19K was established by raising specific antibodies which were used for immunoprecipitation of phosphorylated 19K from 32P,-labeled cells. Thus, specific precipitation of a 19-kDa alkali-stable phosphoprotein was obtained in PBDu-treated human T cells. In addition to demonstrating the specificity of the antibodies, immunoprecipitation experiments also demonstrated rapid PDBu-produced phosphorylation of 19K in cellular systems that are distinct from T cells (Fig. 6).
We have previously demonstrated a quantitative correlation between 19K phosphorylation and phosphorylation of the previously described 80K protein kinase C substrate (3)(4)(5). Moreover, phosphorylation of these substrates in T cells is only observed with ligands that are known to activate protein kinase C (i.e. synthetic diacylglycerol and specific receptor stimulation), while nutritional factors are without effects (3c. 32Pl-labellng , Panel A, the resulting material after the final purification step on a HPLC reverse phase column (0.2 pg) was separated on 15% SDS-PAGE and transferred to nitrocellulose. Western blotting was performed as described under "Experimental Procedures" with either preimmune sera (0.2%) or anti-19K immune sera (0.2%) as indicated. Panel B, peripheral human T cells, quiescent after 2 weeks of IL-2-dependent growth, were labeled with ["'Slmethionine for 30 min. PBDu (20 rig/ml) or solvent alone was added, as indicated,during the least 15 min before the cells were lysed in Triton X-100. Immunoprecitation was performed as described under "Experimental Procedures" with either preimmune sera or anti-19K immune sera as indicated. Panel C, peripheral human T cells, quiescent after 2 weeks of IL-2-dependent growth, were labeled with '*P, for 6 h, PBDu (20 rig/ml) or solvent alone was thereafter added, as indicated, 15 min before the cells were lysed in Triton X-100. Immunoprecitation was performed as described under "Experimental Procedures" with either preimmune sera or anti-19K immune sera as indicated. Panel D, the indicated part of the gel shown in panel C was rehydrated and alkali-treated and re-exposed for 96 h (exposure time in panel C was 24 h). Molecular masses are given in kilodaltons. 6). It has also been shown in previous reports that increased levels of intracellular calcium or CAMP failed to induce 19K and 80K phosphorylation in T cells (4,21). Thus, several lines of evidence suggest that 19K is one of the substrates that can be phosphorylated by protein kinase C in intact cells. Phosphorylation of 19K by protein kinase C is also suggested by the present report, since 19K serves as an in vitro substrate for a partially purified protein kinase C preparation (Fig. 3B). However, future identification of the phosphorylated serine residues, in both in uivo and in uitro phosphorylated 19K, is required before we can firmly conclude which kinase(s) is directly involved in phosphorylation.
A protein with similar characteristics to 19K has previously been described by Feuerstein and Cooper (19) in the myeloid cell line HL-60. Their studies of rapid phosphorylation events induced by phorbol esters revealed a 17-kDa protein (later more accurately sized as l&4-kDa) (22) with a similar appearance to 19K on two-dimensional SDS-PAGE. This protein (originally designated pp17) is phosphorylated on serine residues and shows the same remarkable resistance to alkali treatment as 19K (19 showed that pp17 is also expressed and phosphorylated in human T cells, thus suggesting that pp17 and 19K are the same protein. The present study shows that 15 min of PDBu treatment, which is sufficient to induce maximal 19K phosphorylation, has no detectable effect on the rate of 19K synthesis (Fig.  4B). However, we demonstrate by Western blot analysis that expression of 19K increased at least lo-fold after activation of primary T cells, and this increase was paralleled by the onset of DNA synthesis (Fig. 7). Interestingly, by studies on ["%]methionine-labeled T cells, and analysis of cell lysates by two-dimensional SDS-PAGE, Cooper et al. (23) have recently shown that the synthesis of pp17 is dramatically increased during the process of T cell activation. This result provides further evidence that pp17 and 19K may be the same protein. The observed activation associated up-regulation of this putative protein kinase C substrate warrents considerable further work on the regulation and function of this protein.
The amino acid sequence of a cyanogen bromide fragment suggests that 19K is a novel putative protein kinase C substrate. This amino acid sequence has been confirmed by a recent isolation of cDNA clones encoding the major part of the 19K polypeptide (to be published when a full length cDNA is obtained). The cloning strategy was based on immunoscreening of a X gtll cDNA library, and the identity of the 19K cDNA was demonstrated by affinity purification of 19Kspecific antibodies on the expressed recombinant protein. DNA sequence analysis of a cDNA clone, which lacks approximately 145 base pairs in the 5' end, reveals a 13.7-kDa open reading frame which contains the 20-amino acid sequence shown in Fig. 5A (33 amino acids from the COOH-terminal