Appearance of Rapidly Labeled, High Molecular Weight RNA in Nuclear Ribonucleoprotein RELEASE

Chromatin and nuclear ribonucleoprotein (nRNP) have been prepared from a human carcinoma cell line. Following a 1-hour (3H)uridine pulse, 60 to 70% of the nuclear radioactivity, after removal of nucleoli, was found in the chromatin, the balance in nRNP. This was true whether the chromatin and nRNP were separated by velocity centrifugation or by isopycnic centrifugation on Metrizamide gradients. Radioactivity in chromatin and nRNP was found in high molecular weight RNA, with mean sedimentation coefficients of 20 S and 15 S, respectively, as determined on sodium dodecyl sulfate-sucrose gradients. Experiments on the kinetics of appearance of radioactivity in the RNA of the two fractions suggest that some of the chromatin-associated RNA is precursor to nRNP-RNA. The proteins of nRNP are complex as revealed by sodium dodecyl sulfate gel electrophoresis. The contamination by chromatin protein was estimated to be 5%. Experiments involving short pulses of (3H)tryptophan, and pulse-chase, suggested that the rapidly turning over proteins of nRNP were not complexed with RNA while still associated with chromatin. However, it was also shown that the radioactivity in nRNP following short pulses of (3H)tryptophan did not correspond to the major bands seen on stained sodium dodecyl sulfate gels. It is therefore concluded that the protein of nRNP consists of two classes: species present in large amounts, possibly common to all RNA in nRNP, which are relatively stable and may be complexed to RNA still associated with chromatin; and a large number of rapidly turning over species, each present in small amounts and associated with nRNP only after its release from chromatin.

to RNA in nRNP was not pelleted under these conditions (see "Results").
In 10 experiments, the recovery in chromatin of DNA layered over 60% sucrose (after removal of nucleoli) ranged from 55% to 100%. However, nuclear DNA in the nRNP fraction ranged from 3% to 9% with a mean of 6.4% * 0.5. Therefore, DNA not recovered in chromatin does not necessarily contaminate nRNP. This point will be returned to under "Discussion." Isolation of RNA-RNA was isolated from chromatin and nRNP by the high salt DNase digestion procedure described by Soeiro and Darnell (14), and run on SDS-sucrose gradients according to Derman and Darnell (15 Gel Electrophoresis-Ten per cent SDS-polyacrylamide gels (10 cm) were prepared and run as described by Weber and Osborn (18). A 3% stacking gel of 0.5 cm was used. Electrophoresis was at 4 mA/gel for 30 min followed by 8 mA/gel for 7 hours. Staining with Coomassie blue was also as described (18). The gels were scanned at 550 nm in a Gilford spectrophotometer at a scan rate of 1 cm/min and a chart speed of 1 minIinch.
The chromatin and nRNP pellets were resuspended in 0.01 M sodium phosphate buffer, pH 7.0; 0.1% SDS; and 0.1% @-mercaptoethanol and then dialyzed overnight against this buffer. It is not necessary to remove nucleic acid prior to electrophoresis (12,19,20). The banding pattern of the stained gels was highly reproducible.  (22) and assayed by the orcinol method (23). DNA was assayed according to Burton (24).  (Fig. 1). In three separate experiments. the chromatin-RNA similarly labeled was somewhat larger, with a mean coefficient of about 20 S, and material running up to 30 S (Fig. 1).

Incorporation
A second method of fractionation, sedimentation of the post nucleolar supernatant on Metrizamide-buoyant density grsdients, gave consistent results. Cells were labeled for 4 days with ["C]thymidine and 1 hour (or 15 min, not shown) with [*H ]uridine. Fig. 2 shows that 60% of the 'H was found in the same region of the gradient as ["C ]thymidine-labeled chromatin. The remainder, found at higher densities, was presumably in free nRNP.
Relationship between RNA of Chmmotin and nRNP-Cultures of cells were labeled continuously with ['Hluridine for periods of from 30 s to 60 min. Fig, 3 illustrates that with a 30-s pulse. 100% of the radioactivity in the nucleus (after removal of nucleoli) was in the chromatin fraction, and none in the nRNP. By 90 s, radioactivity appeared in the nRNP fraction, with a corresponding decrease in the percentage found in chromatin. At 2 min, only 10% of the radioactivity was in chromatin, the balance was found in the nRNP. The relative amounts remained essentially constant up to M) min. although the percentage found in nRNP may have slowly increased. Fig. 4 shows that the specific activity of the RNA in chmmatin rose very quickly. while the increase was slower in the nRNP-RNA. These results suggest that some of the chromatin-RNA is precursor to the RNA isolated in nRNP.

'bonucleoprotein and Chromatin
Presence of Protein in nRNP- Fig.  5 shows the results of C&I, gradients of formaldehyde-fixed nRNP. The buoyant density of the RNA in nRNP was 1.43 (Fig. 5A), similar to what has been previously reported (1,3,5), and corresponding roughly to an RNP consisting of 60% protein. The same results were obtained for 1 min uridine pulse-labeled RNA in the nRNP fraction (not shown). Fig. 5B illustrates that at least 85% of the tryptophan-labeled protein was found at the same buoyant density and was therefore complexed to RNA. The small amount of apparently free protein in this gradient (density < 1.36) could have been due to incomplete fixation. or reversal of fixation during long centrifugation (26). These results indicate that the tryptophan-labeled protein in the nRNP was firmly associated with the RNA, since it has been recently shown that, at least in chromatin, it would otherwise not be fixed by formaldehyde treatment (27).
Size Distribution of nRNP-Following a l-hour ["Hluridine pulse, radioactivity in the nRNP fraction (post chromatin supernatant) sedimented between 1 S and 200 S with a modal value of about 76 S (Fig. 6A). When the RNP was isolated by centrifugation of the post chromatin supernatant. resuspended, and run on an identical sucrose gradient to that in Fig.  6A, the size distribution was not as broad (Fig, 6B). The modal value is a sharper peak at someivhat less than 76 S, and there is far less material sedimenting at larger S values. Finally, Fig.  6C shows that some labeled RNA of small size was not   sedimented with the nRNP and remained in the supernatant. This was usually 10 to 15% of the total radioactivity in the nRNP fraction, and may consist of RNA fragments or small nRNP. Because it is clear that the nRNP decreased in size during preparation, it is not surprising that very large hnRNA greater than 45 S was not found in this fraction ( Fig. 1) (see "Discussion").
This preparation had been simultaneously labeled for 1 hour with ["Cltryptophan.
We can see that the post chromatin supernatant contained a great deal of tryptophan-labeled protein at the top of the sucrose gradient (Fig. 6A) which did not pellet with the nRNP (Fig. 6B), but was, rather, found in the corresponding supernatant (Fig. 6C). This has previously been identified as the soluble nuclear protein fraction (28). In several experiments with [3H]tryptophan incorporation, where the counts were considerably higher and proper quantitation could be made, we found that 60 to 65% of the [3H]tryptophanlabeled material in the post chromatin supernatant pelleted with the nRNP. As noted, the CsCl, gradient shown in Fig. 5B suggests that all of the protein pelleted with the nRNP is firmly associated with it (27).
Characterization of Protein of nRNP- Fig.  7, A and B are scans of stained SDS gels of nRNP and chromatin preparations, respectively. These patterns were highly reproducible. The complete absence of histones in the nRNP preparation should be noted. The small, low molecular weight peaks in nRNP, which appear to be histone contaminants, are not. This is shown in the gel photograph in Fig. 7C. The small molecular weight proteins in nRNP clearly ran slightly behind the two low molecular weight histone bands of chromatin in the two separate preparations shown. The major nRNP band ran at a molecular weight of about 34,000, in close agreement with the molecular weight of the major protein of nRNP in other systems (1,2,4,5). Although the resolving power of one-dimensional gels is limited, these results point to a marked heterogeneity of the proteins of nRNP.
Since our preparation of chromatin and nRNP seems to effect a separation between released nuclear RNA and that still associated with chromatin, we have investigated whether we first find nRNP protein in chromatin, presumably associated with the transcript during synthesis. Fig. 8 shows that with continuous incorporation of [3H]tryptophan, there was no lag in the appearance of radioactivity in the nRNP when compared to either whole nuclei or chromatin. Incorporation of [3H]tryptophan into total cell protein could be stopped within 5 min by a chase with unlabeled tryptophan (not shown). Fig. 9 illustrates that this chase very rapidly stopped the accumulation of radioactive protein in nRNP. Certainly, when compared to total cell protein or chromatin protein, there was no lag. These experiments therefore suggest that the nRNP proteins which incorporate [3H]tryptophan during these short pulses are associated, very soon after synthesis, with nRNP. They are not first located in chromatin, or any other fraction, and shifted to nRNP after some period.
The labeled protein from these pulse and pulse-chase experiments was run on SDS gels. It is important to note that the radioactivity did not correspond to the positions of the bands in stained gels (Fig. 7, A and B), but was more uniformly distributed throughout the gel. This was particularly apparent in the nRNP.
Although the high molecular weight species of chromatin showed a relatively higher level of incorporation at 5 min, this was not true of nRNP (Fig. 1OA). In nRNP, these proteins showed a higher level of incorporation only after 30 min (Fig.  lOC), while the distribution of label remained the same in chromatin from 5 to 30 min. However, Fig. 11, A to C illustrates that during a chase with u.ilabeled tryptophan, the distribution of radioactivity in nRNP (or chromatin), did not change. Hence, these high molecular weight labeled proteins of nRNP could, again, not have been associated first with chromatin and then shifted to nRNP.

DISCUSSION
Our data indicate that rapidly turning over high molecular weight RNA can be isolated associated with chromatin and that some of this RNA is precursor to RNA found in nRNP. Monahan and Hall (29) have reported that precursor to hnRNA is found in chromatin isolated on Metrizamide gradients. This is consistent with the recent observations that globin sequences can be found in the chromatin-associated RNA from mouse fetal liver cells (30) and that the RNA content of chromatin is greatest in material isolated from tissues with the highest capacities for RNA synthesis (31,32), although in the latter reports, the nature of the RNA is unknown.
In assessing the problem of RNA breakdown in our preparation, it should be noted that the rapidly labeled RNA of both our chromatin and nRNP is not as large as the RNA that can be obtained from whole nuclei (33). Indeed, the RNA that we extract from HT-29 nuclei is larger than either chromatin or nRNP RNA (>50 S). Clearly, therefore, there is RNA breakdown during preparation, but this must be limited to nicking, since we routinely recover all of the nuclear radioactivity   (27). It would he desirable to demon-strate more conclusively that all protein in nRNP is pelleted only because of its association with RNA. One way would be to show that ribonuclease digestion of RNA in the post chromatin supernatant eliminates any pelleting of protein upon subsequent isolation of nRNP by centrifugation (see "Methods"). However, we have found, as have others (1,42) that not all of the RNA in nRNP is susceptible to digestion by nuclease and therefore a macromolecular, protein-containing complex might still exist. In addition, we have observed, as have others, (1) that after RNase digestion of nRNP, the protein precipitates. Therefore, although the nRNP protein is not found in the 122,000 x g supernatant after ribonuclease digestion, it no longer forms a discrete pellet, but rather is recovered from the walls of the centrifuge tube.
A very important question is how much of the nRNP-protein is chromatin contamination.
The nRNP contains an average of 6% of the nuclear DNA. However, when the cells were labeled with ["Clthymidine, 47% of the radioactivity in nRNP pelleted, (density > 1.6) when the formaldehyde-fixed nRNP was spun on CsCl, (as compared to 2% and 7% when the label was [3H]uridine and [3H]tryptophan, respectively). Therefore, of the 6% of the total nuclear DNA in nRNP, almost half is essentially protein-free, and the protein-associated DNA (DNAp) in nRNP is only 3% of the total nuclear DNA. One can consider the amounts of RNA in nRNP and chromatin to be roughly equal (following a l-hour pulse of [3H]uridine, the specific activity of nRNP-RNA is about 60% that of chromatin-RNA, and the nRNP fraction contains about 60% of the radioactivity in chromatin). Therefore the ratio of nuclear DNA to nRNP-RNA is 20 (just as it is for chromatin-RNA). However, only 3% of the DNA is present in nRNP as DNAp, so the ratio of DNAp in nRNP to RNA in nRNP is 0.6. This chromatin present in nRNP has a protein to DNA ratio of 1 (no histone, Fig. 7). Because the protein to RNA ratio in nRNP is about 4, the contamination due to chromatin proteins in nRNP is '/4 x 0.6 x 100 = 15%.
This calculation can be done in another way. When the actual total protein to DNA ratio in nRNP is measured, it is found to be variable, as expected, since DNA is a contaminant. The values range from 5.1:1 to 17.7:1 (six determinations) with an average of 9.7:1. However, 50% of this DNA is protein-free, so the true ratio of chromatin protein to total DNA in nRNP is 1 (protein:DNA of histone-free chromatin) x 0.5 = 0.5. The average chromatin protein contamination is therefore the true ratio of chromatin protein to DNA in nRNP divided by the measured ratio of total nRNP protein to DNA: 0.5/9.7 x 100 = 5%, and, similarly, ranges from 3% to 10%. Some of these assumptions are approximations, but the contamination of nRNP with chromatin protein does not seem to be sufficiently high to change the interpretation of the data presented. Perhaps, the most straightforward indication of limited contamination is the completely different labeling patterns of the proteins in the two fractions (Figs. 10 and 11).
In considering our data from the pulse and pulse-chase experiments (Figs. 8 and 9), we do not know the rate of synthesis, cytoplasmic (or nuclear) pool sizes, or rate of transport of the protein in the nRNP and chromatin fractions. These results should therefore be approached cautiously. Nevertheless, in no case did we obtain results which would indicate that most of the rapidly turning over nRNP-protein is complexed to the transcript during synthesis or when the transcript is otherwise associated with chromatin. It should be noted however, that observations of Beermann (43), on puffs of the giant chromosomes of Chrionomus, suggested that protein did become associated with RNA to form RNP while still associated with chromatin. This observation has been repeated in the Balbiani rings and other puffs of Dipteran tissue (8,(44)(45)(46), the loops of lampbrush chromosomes (47)(48)(49), and most recently in regions of mammalian chromatin active in RNA synthesis (9). Furthermore, Scott and Sommerville (11) have shown that proteins isolated from RNP released into the nucleoplasm are antigenically similar to nonbasic proteins of lampbrush loops. We have recently obtained evidence which suggests that proteins which protect a portion of the RNA of both chromatin and nRNP from nuclease digestion are similar. 3 Our results, however, are not inconsistent with this previous work on the presence of nRNP in chromatin, since we have also shown that following short pulses of [3H]tryptophan, the radioactivity in nRNP does not correspond to the major bands on stained SDS gels. We suspect, as already stated by Georgiev (l), Stevens and Swift (8), and Berendes (40), that these stable proteins, present in large amounts and therefore possibly common to all hnRNA (5), bind to growing transcripts and are involved in packaging and transport (1,8,40), while the rapidly turning over proteins are added later, and we suggest, are involved in processing. Enzymes capable of cleaving hnRNA (51) and adding poly(A) to hnRNA have been identified in nRNP (52).
There is, in fact, evidence that protein added to RNA still associated with chromatin is stable. Pelling (53) has shown that RNA-synthesizing structures in Chironomous puffs are not labeled with radioactive amino acids up to 2 hours after label injection, and Clever (45) has shown the same to be true for puffs induced by ecdysone. Gall  inhibitors of protein synthesis do not affect the structure of lampbrush loops (54) or the initial induction of puffs by ecdysone (45,55), though subsequent puff formation, also dependent on RNA synthesis (55), is inhibited.
We have also shown that the major chromatin proteins do not turn over rapidly. This is consistent with several reports that much (non-histone) chromosomal protein exhibits great stability (56,57).