Procollagen mRNA Sequences in Chick Embryo Fibroblasts Infected with Rous Sarcoma Virus CORRELATIO’N WITH PROCOLLAGEN SYNTHESIS*

Chick cells infected with Rous sarcoma virus are characterized by a wide variety of changes known collectively as transformation. Among these are decreases in the level of procollagen biosynthesis and in the level of procollagen mRNA. In this communication, we examine the time course of the decrease in procollagen biosynthesis, as measured by sodium dodecyl sulfate-polyacrylamide gel electrophoresis and collagenase assay, and compare it with the decrease in procollagen mRNA sequences measured by hybridization to a complementary DNA. Procollagen biosynthesis and procollagen mRNA sequences decrease simultaneously after infection. Even the initial decrease in procollagen biosynthesis, therefore, is due to a decline in the level of procollagen mRNA.


Chick cells infected with
Rous sarcoma virus are characterized by a wide variety of changes known collectively as transformation. Among these are decreases in the level of procollagen biosynthesis and in the level of procollagen mRNA.
In this communication, we examine the time course of the decrease in procollagen biosynthesis, as measured by sodium dodecyl sulfatepolyacrylamide gel electrophoresis and collagenase assay, and compare it with the decrease in procollagen mRNA sequences measured by hybridization to a complementary DNA. Procollagen biosynthesis and procollagen mRNA sequences decrease simultaneously after infection.
Even the initial decrease in procollagen biosynthesis, therefore, is due to a decline in the level of procollagen mRNA.
The study of cellular changes induced by viral infection may elucidate cellular mechanisms in viral replication and in the normal regulation of host cell macromolecular synthesis. In eukaryotes, viral infection may or may not be followed by production of virus, and if production does ensue, it may be passive or lytic in nature. Thus, the extent to which host function is disrupted is dependent on the nature of the infection. For example, herpes simplex virus causes degradation of poly(A)-containing RNA within 4 h in lytically infected Friend leukemia cells (1). In the case of polio virus, host mRNA, although retained, is no longer translated because of changes in the host's translation machinery (2). On the other hand, Rous sarcoma virus induces a variety of changes in chick embryo cells, collectively known as transformation (3), and utilizes a relatively small percentage of host cell synthetic machinery for its replication (4). A number of specific host functions, as well as the distribution of various proteins and mRNA sequences, are altered following transformation by RSV' (5); these include sugar transport (6), globin (7), sulfated proteoglycan (8), adenosine deaminase (9), fibronectin (lo-12), and collagen (10, 13-18). However, relatively little is known about the mechanism or timing of these changes. In the case of procollagen, it has been * This work was supported by National Institutes of Health Grants

AM11248
and DE02600. shown that both synthesis of the protein and the procollagen mRNA level, when monitored by in vitro translation or hybridization to procollagen cDNA, are reduced in transformed cells (10,17,18).
In this study, we have measured the kinetics of the decrease in procollagen biosynthesis and in procollagen mRNA sequences during the fist 84 h after infection of chick embryo fibroblasts with RSV. Specifically, we were interested in whether this initial reduction in procollagen synthesis could be attributed to events which affect the concentration of the message or to factors altering the activity of this mRNA. Our results indicate that procollagen synthesis is decreasing by 24 h after infection and that the time course of the decline in procollagen mRNA sequences and procollagen synthesis is similar.  (25), collagenase digestion (21), and SDS-polyacrylamide gel electrophoresis (19), and half was taken up in SET buffer (1% SDS, 5 m EDTA, 10 mM Tris, pH 7.5) containing 50 pg/ml of Proteinase K in SDS and extracted in phenol/chloroform as described above. Cell layers dissolved in 0.5 M ammonium hydroxide were precipitated with trichloroacetic acid for collagenase assay as described by Peterkofsky and Diegelmann (21). The supernatants of this precipitation were counted to determine the radioactivity in proline pools in normal and transformed cells. Tricbloroacetic acid-soluble pools of proline, expressed per pg of protein, in transformed cells were 1.4 times those in normal cells by 64 h. This could reflect larger pools of proline or higher specific activity of proline pools in transformed cells, or both. If the specitic activity of the proline pools is not affected by transformation, then changes in collagenase-sensitive counts per min in proline, per pg of protein, would reflect absolute differences in collagen synthesized in normal and transformed cells. If, on the other hand, the specific activity of proline pools is greater in transformed cells, then measurement of collagenase-sensitive proline, expressed per pg of protein, may actually inflate the apparent level of collagen I 2 FIG. 1. SDS-polyacrylamide gel electrophoresis of translation products directed by procollagen mRNA in the reticulocyte lysate system. Poly(A)-enriched RNA was fractionated on sucrose gradients as described under "Experimental Procedures" and fractions containing mRNA activity for procuI(1) and proa chains were used to prime translations. Aliquots of the translation mixture, labeled with [JH]proline, were incubated minus (Slot I) and plus (Slot 2) bacterial collagenase for 45 min at 37°C. Samples were electrophoresed directly on 5% SDS-polyacrylamide gels. made by transformed cells, relative to normal cells. Data expressed as the percentage of total counts per min which is collagenase-sensitive are, of course, unaffected by differences between the specific activity of proline pools in normal and transformed cells.

RESULTS
Purification of Procollagen mRNA and Synthesis of Procollagen cDNA-Translation of sucrose gradient-fractionated chick tendon poly(A)-containing RNA revealed a broad region of procollagen mRNA activity spanning the marker 28 S rRNA peak (data not shown). Electrophoresis of translation products on SDS-polyacrylamide gels (Fig. 1, Slot I) showed that the mRNA activity in this region of the gradient stimulated the synthesis of proteins migrating in the region of chick procrl(1) and procu2 markers. The existence of doublets in each of these positions was variable and could have been due to partial proteolysis of the procollagens by the lysate itself. Fig.  1 demonstrates that these bands were sensitive to bacterial collagenase; the products obtained by translation of oviduct mRNA were not (data not shown). Interestingly, some mRNA template activity for procu2 migrated perceptibly earlier in the sucrose gradient than did procul(1) activity (data not shown). Procollagen cDNA was synthesized from pooled fractions of the gradient containing mRNA activity for both proa and procul(1) chains. and was still declining 84 h after infection. In contrast, procollagen synthesis in cells not infected with RSV actually increased somewhat over this time course, perhaps as a response to rapid proliferation upon plating. If the data are expressed as collagenase-sensitive counts per min per mg of cellular protein, instead of as a percentage of total counts per min, a similar decrease in infected cells is observed. This suggests that the decrease is an absolute one and is not only a change in the percentage of total protein synthesis devoted to collagen.

Kinetics of the Decrease in Procollagen Synthesis
Examination of procollagen mRNA levels by hybridization to procollagen cDNA revealed a decline in procollagen sequences following RSV infection. A decrease was apparent by 24 h after infection and the decline continued in parallel with the change in procollagen synthesis up to 84 h after infection. In contrast, in uninfected cells, the level of procollagen mRNA levels increased significantly until 36 h, then reached a plateau and returned to initial levels by 84 h. It thus appears that the decline upon infection in procollagen biosynthesis can be accounted for by a corresponding decrease in procollagen mRNA levels.
These results are corroborated by SDS-polyacrylamide gel electrophoresis analysis, as shown in Fig. 3. In this experiment, equal counts of proline-labeled cell layers were electrophoresed on 5% polyacrylamide gels. The decrease in the density of the bands co-migrating with procul(1) and prool2 chains over the time course is apparent. Several new proteins, which may be transformation-related, appear late in the time course of infection. Densitometric tracing of the fluorogram revealed that, at 48 h after infection, radioactivity in the two procollagen bands of the transformed cells (Fig. 3, Slot 5) was 43% of that in the control cells (Fig. 3, Slot I). By 84 h, transformed cell procollagen bands (Fig. 3, Slot 8) had decreased to 14% of the control value. These data are in good agreement with those obtained by collagenase digestion. DISCUSSION Clear definition of the transition from normal to transformed states in cells infected by RSV will help to characterize the points at which transformational control is exerted. Our work has demonstrated that the reduction in procollagen synthesis observed in RSV-transformed cells is mediated by a decrease in the level of procollagen mRNA sequences soon after infection. Other workers have shown that induction and maintenance of transformed characteristics in chick cells, infected by RSV, depends on the src gene product (26), a protein kinase (27). There is also evidence from experiments on cells infected with RSV, which was ts in the transformation function, that the src protein acts at both nuclear and cytoplasmic sites (28).
Since fibronectin, like procollagen, is a cell surface-associated and extracellular connective tissue protein which is reduced in RSV-transformed chick cells (10-G!), it is informative to briefly compare what is known about the mechanism of its reduction with the results reported here. In fully transformed cells, Olden and Yamada (11) have shown that changes in both biosynthesis and extracellular turnover contribute to the decreased level of fibronectin.
At least part of the decrease in biosynthesis is attributable to a decrease in the level of translatable fibronectin mRNA, as assayed in vitro (10). Experiments with chick cells transformed with RSV ts for transformation have also elucidated changes in post-translational regulation of this protein. Hynes and Wyke (12) found that ts RSV-transformed chick cells had increased cell-surface fibronectin within 4 h of shifting from permissive to restrictive temperatures and that this effect was insensitive to an inhibitor of protein synthesis, cycloheximide.
Beug et al. (28) have obtained compatible results with chick fibroblasts transformed by RSV ts for transformation.
In the latter work, cellsurface fibronectin measured by immunofluorescence was absent at the permissive temperature in both nucleated and enucleated cells, but reappeared upon shifting to the restrictive temperature.
It therefore seems that the transformationrelated reduction of fibronectin may involve an initial modulation of fibronectin levels which is independent of the change in the level of fibronectin mRNA. The results presented in this communication are of interest since this is the fist time that the reduced level of a cellular protein after infection by RSV could be correlated with the initial decrease in mRNA for that protein. Our experiments suggest that, in the case of procollagen, the regulation of procollagen mRNA is at least partially responsible for the initial decrease observed in procollagen biosynthesis.
Since the experiments presented here have measured cell layer procollagen labeled in a 30-min pulse period, these measurements do not necessarily reflect changes in turnover of cell surface or extracellular protein, or both. It will be interesting to compare the results of additional studies relating to the regulation of fibronectin and procollagen in transformed cells. The decrease in procollagen mRNA levels shown here could result from decreased synthesis, increased degradation, or a combination of these factors. Experiments are underway to distinguish among these possibilities.
Acknowledgments-We are grateful to Drs. D. Rowe and D. Lee for assistance with in vitro translation and immunoprecipitation techniques, respectively, and to Drs. D. Lee and B. Gallis for valuable discussions and critical reviews of the manuscript.