in Chromatin*

The conformational state of histones in isolated chicken erythrocyte chromatin was studied using procedures developed for probing surface proteins on membranes. Under controlled conditions, only exposed tyrosyl residues react with iodide radicals, generated either by the oxidant, chloramine-T (paratoluenesul-fonyl chloramide), or the enzyme lactoperoxidase, giv- ing monoiodotyrosine. Using 125-iodine, this study compared the reactive tyrosines in free and bound histones H., and Hg. The relative extent of iodination of these histones within (H4) and outside (Hs) of the nu- cleosomes was measured after extraction and gel elec- trophoresis. Each of the histones was further analyzed for the extent of specific tyrosine iodination by sepa- rating the tryptic peptides by high voltage electropho- resis. The identity of the labeled peptide was deter- mined by dansylation of the amino acids present in each hydrolyzed peptide. The results show that there is a difference in the conformational arrangement of these histones on chromatin and in the free forms, since in chromatin not all tyrosine residues are as accessible for iodination

The conformational state of histones in isolated chicken erythrocyte chromatin was studied using procedures developed for probing surface proteins on membranes.
Under controlled conditions, only exposed tyrosyl residues react with iodide radicals, generated either by the oxidant, chloramine-T (paratoluenesulfonyl chloramide), or the enzyme lactoperoxidase, giving monoiodotyrosine.
Using 125-iodine, this study compared the reactive tyrosines in free and bound histones H., and Hg. The relative extent of iodination of these histones within (H4) and outside (Hs) of the nucleosomes was measured after extraction and gel electrophoresis.
Each of the histones was further analyzed for the extent of specific tyrosine iodination by separating the tryptic peptides by high voltage electrophoresis. The identity of the labeled peptide was determined by dansylation of the amino acids present in each hydrolyzed peptide.
The results show that there is a difference in the conformational arrangement of these histones on chromatin and in the free forms, since in chromatin not all tyrosine residues are as accessible for iodination as in the denatured state. Residue 53 of histone Hg for instance is more reactive than residues 28 and 58, indicating that the segments containing the latter residues are involved in either protein-DNA or protein-protein interactions.
In histone Hq, preferential labeling of 2 of the 4 tyrosines present was also observed. Sequence data from histones show that there is a statistically nonrandom distribution of basic amino acids in each of the histones HI, H2,, HZ,,, H:$, Hq, and H, (Elgin and Weintraub, 1975). It can be calculated that such a distribution would limit the possibilities of conformational change of these proteins. Indeed, only a few definable configurations of histones can be derived (Chou and Fasman, 1974;Fasman et al., 1977).
There is evidence suggesting that the configurations of histones in solution, and thus in the free state, and those in the bound state in chromatin interacting with other molecules such as DNA or other proteins, are different. The topography of these histones bound in chromatin has been studied extensively by several techniques. 125.Iodine, as a chemical probe, has been used to study histone Hq in Xenopus laevis chromatin (Biroc and Reeder, 1976). It was concluded that 2 of the 4 tyrosyl residues in H, (tyrosines 72 and 98) are protected from reaction when this protein is bound to chromatin in the * This work was supported in part by Grant CRBS 300 from The National Foundation March of Dimes, and Grants ES 00454 and 01596 from the National Institutes of Health. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "u&~erti.sement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
native form, but some of the residues may be preferentially exposed if the chromatin is exposed to media increasing in ionic strength. The reactivity of the tyrosyl residues corresponded to the degree of freedom afforded to Hq during a stepwise extraction process. In these studies, chloramine-T was used as the oxidant. Since chloramine-T is a strong oxidizing agent, it is uncertain if the native conformation of a protein, particularly when it is involved in protein-protein or protein-DNA interactions, would not have been perturbed. Therefore, it is conceivable that the use of a milder catalyst for iodination would be desirable, to prevent disruption of the native protein conformation (Morrison and Schonbaum, 1976).
Different methods of iodination are not simply different ways of preparing the same electrophile (I+); there are instead fundamental differences among the mechanisms of iodination. The rate of deiodination of radioiodinated proteins varies with the method of iodination (Krohn et al., 1977). Iodine monochloride, chloramine-T, and the lactoperoxidase method yield largely 3-iodotyrosine with small amounts of other iodinated amino acids; however, the chloramine-T product spectrum varies with the chloramine-T:protein ratio. In terms of radioiodine incorporation, the efficiency of the different methods, from greatest to least, is found to be: electrolytic > chloramine-T > lactoperoxidase > iodine monochloride. Lactoperoxidase has also been used in the probing of available tyrosines in proteins on cell membranes. It has been used successfully in the study of membrane proteins in intact cells (Phillips and Morrison, 1971;Hubbard and Cohn, 1972;Graham et al., 1975;Morrison and Schonbaum, 1976). Chloramine-T and lactoperoxidase iodination techniques have also been used to probe the accessibility of envelope proteins of vesicular stomatitis virions (Moore et al., 1974). The glycoprotein and, to a lesser extent, the matrix membrane protein of intact vesicular stomatitis virions have been specifically iodinated by oxidation with chloramine-T and the lactoperoxidase methods. The soluble or solid-state lactoperoxidase method of iodination has been used to probe the molecular morphology of Escherichia coli ribosomes (Michalski and Sells, 1975). From this study, it was concluded that the 30 S subunit undergoes a conformational change during its association with the 50 S subunit to form the 70 S monosome. The high molecular weight of these enzyme molecules prevents their entrance into the inner portion of the cell membrane; thus, only exposed tyrosyl residues are reactive. If chromatin also has a compacted subunit structure, then the structure of the histones should be amenable to an analogous study.
In this study, we have focused on two histones: Hs, the histone unique to nucleated erythrocyte chromatin, located outside of the nucleosome structure, and the corresponding H4 histone, whose presence within the chromatin subunits is ubiquitous.
Histone H, has been implicated as a major con-of Histones H4 and H;, densing protein responsible for the compact state, thus lowering transcriptional and replicative activities of the avian erythrocyte chromatin (Huang et al., 1977). Evidence supporting this notion includes the observation that histone H, increases in its proportion to other histones with the advances of erythrocyte maturation, signified by a gradual loss of DNA and RNA synthetic activity. By immunofluorescence assay, the arrangement of Hs within each nucleus has been observed t,o undergo definable changes with erythroid age. The dim and diffuse fluorescent patterns in the early embryonic erythroid cells change to a bright halo distributed peripherally to the inner nuclear membrane of adult erythrocytes (Mura et al., 1978). The changing Hs staining pattern within the nucleus may well reflect the supramolecular structural organization of the chromatin during various stages of gene activity. Histone Hs has been partially sequenced (Sautibre et al., 1976) and shown to have only 3 tyrosine residues at positions 28,53, and 58.
The sequence of Hq from chick erythrocytes is yet to be determined.
However, Hq from different species examined so far shows a high degree of sequence homology with that from calf thymus which contains 4 residues of tyrosine at positions 51, 72, 88, and 98 (Elgin and Weintraub, 1975). The results to be presented in this communication will demonstrate that the availability of these residues to iodination is equal when the histones are free in solution, but different in their native bound state. It is inferred from this study that histones interact in the chromatin with other components with sequence specificity.

Chicken erythrocytes
were obtained from Dover Poultry Products, Inc., Baltimore.
Pooled blood was collected from the production line slaughter of chickens within 1 min of decapitation. Gels were photographed while still wet, then dried down onto a piece of filter paper.
15% Polyacrylamide-Acetic Acid-Urea Slab Gel System-Total histones were also separated using a 15% polyacrylamide-acetic acidurea slab gel system. The acid-urea gels (15 x 11 cm) were prepared by the method of Panyim and Chalkley (1969) with the modification that slab gels were used instead of tubular gels. Urea was purchased from Sigma.
Samples containing 10 to 80 pg of protein in a volume of 20 ~1 or less were mixed with 15 ~1 of 0.1% Pyronin Y in 0.9 N acetic acid, 35% glycerol.
The electrophoresis buffer consisted of 0.9 N acetic acid. Gel running time was approximately 6 h at 150 V (continuous voltage). Quantitation of Histones in Gels-Quantitation of histones within gels was based on a linear relationship between protein and Coomassie G-250 staining intensity.
Under the staining and destaining conditions described for sodium dodecyl sulfate-and acid-urea polyacrylamide slab gel electrophoresis, histone Hs exhibits a linear relationship up to 25 pg of total histone protein, and histone H, exhibits a similar relationship up to 35 pg of total histone protein.
When specific bands were cut from the gel and counted for radioactivity, the specific activity of the histone protein was determined. El&ion of Protein Bands from SDS-and Acid-Urea-Polyacrylamide Slab Gels-Histones are eluted from acetic acid-urea-polyacrylamide gels by extraction in 66% acetic acid. Individual histone bands were cut from the slab gels and homogenized using the Tri-R-Stir-R homogenizer. allowed to proceed for an exposure time varying from 1 h to 2 weeks. For gels or paper sheets with low radioactivity, DuPont Cronex Hi-Plus intensifying screens (8 x 10 inches) were used to decrease the exposure time. 125-Iodine was counted in a Packard Auto-Gamma scintillation spectrometer and a Beckman Gamma 4000 counter. The counting efficiency of both units was 70 to 75%. For high voltage electrophoresis paper with exceptionally low radioactivity associated with it, strips (1 x 3 cm) were cut out and counted individually. Chloramine-2' lo&nation-The procedure used for iodination by chloramine-T (paratoluenesulfonyl chloramide) was that of Biroc and Reeder (1976), except that the order of addition was altered to maximize iodine incorporation. Proteins were iodinated in 6-cm screwtop reaction bottles. The reaction mix was constantly swirled using magnetic fleas. The reaction mix, with appropriate order of addition of reagents, was as follows: 1) up to 100 pg of histone protein in 100 ~1 of water; 2) 100 ~1 of 10 mM Tris, pH 8.0, 2 mM EDTA; 3) 2 ~1 of ""I-sodium in dilute NaOH, 100 @i/PI; 4) 5 ~1 of 0.12 mM KI (0.6 nmoles); 5) 5 al of 1 mM L-tyrosine (5 nmol, Mann); 6) 10 ~1 of 7 mM chloramine-T (70 nmol, Eastman  (Biroc and Reeder, 1976). The column, consisting of a Pasteur pipette with 1.5 ml of bed volume, was equilibrated with 30% acetic acid and, to avoid loss when desalting small amounts of protein, 1.0 ml of 0.25% Triton X-100 was passed through the column fist. The elution profile of the Bio-Gel column is such that histones elute at the void volume and free '""I at the included volume. The entire digest sample was streaked at an origin 11 cm from the end of Whatman 3MM chromatographic paper (234 x 57 cm). Aliquots of 10 ~1 were streaked and dried using a hot air dryer.
The paper was moistened with pH 3.5 buffer (5%, acetic acid, 0.5% pyridine, Eastman), blotted to remove excess buffer, and electrophoresed at 3000 V for 1% h for histone H, and 1% h for L------l. were again lyophilized and counted, and 10 ~1 of 0.2 M sodium bicarbonate was added to each. The sample was thoroughly mixed and lyophilized, and 10 ~1 of distilled water plus 10 ~1 of dansylchloride (2.5 mg/ml acetone, Pierce) were added to the reaction mix. The sample tube was then evacuated under a stream of nitrogen and sealed. The reaction was allowed to proceed for 1 h at room temperature.

Peptide Identification by Thin Layer
Chromatography-Samples were lyophilized over sodium hydroxide pellets and dissolved in 10 ~1 of acetone:acetic acid (3:2, v/v). Spotting was 1 cm in on both sides of Polyamide P6 (Macherey-Nagel, Inc.) or Micropolyamide F1700 (Schleicher & Schuell) thin layer plates (7.5 x 7.5 cm). The first dimension solvent was formic acid:water (3:200, v/v), and the second dimension solvent was benzene:acetic acid (9~1, v/v). The plates were air-dried and heated to 100°C for I min to enhance fluorescence under ultraviolet light. Individual amino acids comprising the peptides were treated in the same manner, as standards.

RESULTS
The extent of iodination of histones varies considerably, depending on whether they are in the free state or bound with FIG. 5. Autoradiography.
other nucleoproteins in the chromatin matrix. Total histones extracted from chicken erythrocyte chromatin were subjected to iodination, and the extent of labeling compared with those on the chromatin at an equivalent range of concentrations. The results in Fig. 1 show that bound histones are much less accessible to iodination than the free ones. At an equivalent total histone concentration, only 15% of the iodinatable residues are available when bound in the chromatin. Homogenization and sonication of the chromatin, however, increase the extent of iodination.
The conditions used for shearing the chromatin were those generally applied to chromatin preparations for transcription studies and no systematic studies were made here.
Kinetics of Iodination-The extent of iodination as measured by radioactive iodine covalently bound to the free and bound histones was studied kinetically by two procedures using chloramine-T and lactoperoxidase. The results (Figs. 2,  A and B) show that a rapid incorporation is observed within 30 min following either procedure, after which the lactoperoxidase-catalyzed reaction diminishes to a plateau. In terms of specific activities of iodinated total histone, the chloramine-T method is approximately loo-fold more efficient than the lactoperoxidase method. While the lactoperoxidase-catalyzed reaction is less effective in the total extent of iodination, it is affected prominently by the enzyme/substrate ratio in the reaction. The kinetic experiments involving optimal iodination time, using lactoperoxidase coupled to Sepharose 4B, are consistent with the results of David and Reisfeld (1974).
Nature of the Iodination Product-In iodination, one or more iodide radicals may be added to tyrosine generally resulting in mono-and diiodotyrosine.
If an even distribution of these two forms exists on different histones and their residues, an erroneous estimation would be made in the interpretation of the results. Care was taken to check the nature of labeling under the reaction conditions specified. The results obtained by high voltage electrophoresis of the protease-R digests of labeled histone Hg indicated that monoiodotyrosine was indeed the major product. Histidine, another possible site of iodination, was not significantly reactive under the conditions used. An example of the paper electrophoretogram of an analysis is depicted by its corresponding autoradiograph in Fig. 3.
Extraction of Histones from Native Chromatin-Total histones were selectively extracted for further analysis to determine the extent of label of individual histones. These histones remain intact after iodination as demonstrated by gel electrophoresis and autoradiography (Figs. 4,A and B). Labeling Profile of Individual Histones when Bound to Chromatin-The relative iodination efficiency of various histones was of particular importance to this study. The proportion of histones recovered from these iodination procedures was measured after separation of the extracted histones by polyacrylamide gel electrophoresis. The bands were stained, photographed, and densitometric tracings were made on positive transparencies from the photographic negatives for quantitation of the proteins, Afterwards, the gels were dried and autoradiographed.
Autoradiography, representing the radioactivity of each band, was also quantitated by densitometric tracing. The autoradiograph densitometry was confirmed by cutting out the bands and counting directly for radioactivity. The results are summarized in Table I. It can be seen that not all histones on the chromatin are iodinated to the same specific activity as those extracted from chromatin prior to iodination.
In chromatin, internucleosomal histones HI and H5 appear to be preferentially iodinated, along with intranucleosomal histone Ha.
Iodinated histones Hq and Hr, were extracted from gels for of Histones H4 and H5 further analysis to determine the extent of labeling of their tyrosine residues at specific sites of their sequences. After 66% acetic acid extraction from gels at 4O"C, the histones remained intact, as demonstrated by rerunning extracted samples on SDS-gels, sectioning the gels, and checking radioactivity for co-migration with standards of Hq and H,.

Tryptic Analysis of Iodinated Histones (E/S) Ratio Determination-Histones
Hq and Hg, free or bound to chromatin, were recovered after iodination for tryptic analysis. The optimal enzyme to substrate ratio for the hydrolysis was empirically determined by examining the digestion products by high voltage paper electrophoresis.
As shown in Fig. 5, an E/S ratio of 1:lO yields three peptides carrying the radiolabeled tyrosines as expected from the H:, sequence. At a high E/S (of 2: 1) or lower E/S (of 1:50) an anomalous peptide appeared. Within the range of E/S ratios tested, however, Hq yielded the same four major peptides and one additional minor one. For quantitation of the radioactivity in the peptides, an E/S ratio of 1:lO was thus chosen, but for the identification of the peptides an E/S ratio of 1:lOO was chosen to reduce possible contamination of histone peptides by trypsin peptides resulting from autodigestion.

-Tyr-Ile-Lys-
(2) 58 -Ser-His-Tyr-Lys- In order to assign the proper identity to the three bands shown on the high voltage paper electrophoretogram for Hs (Fig. 5), the bands were excised and the peptides were eluted and subjected to complete acid hydrolysis. The resulting amino acids were dansylated and analyzed by two-dimensional thin layer chromatography along with standards containing a mixture of the respective dansylated amino acids.
By this procedure, the slowest migrating band was assigned to the peptide containing tyrosine residue 28 (peptide digest 1, top), the middle one (peptide digest 2) containing tyrosine 53, and the fastest migrating one (peptide digest 3), tyrosine 58 (bottom). This is evidenced by zhe unique presence of the  (Biroc and Reeder, 1976).
Chicken erythrocyte Hq was purified by electrophoretic elution (Fig. 7) and its tryptic peptides identified along with those from calf thymus Hq. In this case, the dansylated NH2 termini were analyzed.
The  Fig. 5 and text for conditions.) The paper electrophoretogram was cut and counted after the separation. The results are shown in histograms. +--m-6WO DISTANCE FRCM CfKlN thymus (CT) tryptic peptides are Ile (peptide I), Thr (peptides 2 and 3), and Asp (peptide 4). The NH*-terminal dansyl amino acids for chicken erythrocyte (CE) Hq are shown to be Ile (peptide I) and Thr (peptides 2 and 3), and Asp for peptide 4 inferred. The results are shown in Fig. 8. With this information, the four major iodinated peptides for chicken erythrocyte histone H4 were assigned (migration towards cathode): 51 -Ile-Ser-Gly-Leu-Ile-Tyr-Glu-Glu-Thr-Arg-

72
(2) (3) -Asp-Ala-Val-Thr-Tyr-Thr-Glu-His-Ala-Lys- (4) However, the mobilities of the chicken erythrocyte Hq peptides containing tyrosine differ considerably from purified calf thymus H, digest peptides under identical electrophoresis conditions (Fig. 9). Since the nature of this difference is not known, it could be due to a genuine heterogeneity, or, more likely, a difference in modification.
The sequence assignment is thus tentative.
Quantitation of Monoiodotyrosine in Labeled Peptides-With the tryptic peptides from Ha and Hq identified and the iodination conditions set forth, the ratios between radioactivity in these peptides after various iodination reaction times were compared (Figs. 10 and 11) and the quantitative results of all comparisons are summarized in Tables II and III. Fig. 1OA shows that the Hg peptide containing Tyr 53 was preferentially iodinated when this histone was bound to native chromatin. This is in contrast to the same histone labeled in Tryptic peptides of samples iodinated for different reaction times were separated by high voltage paper electrophoresis, pH 3.5 (0.5% pyridine, 5% acetic acid). The reaction times for Hs bound to native chromatin were: 0 min; 15 min; 30 min; 60 min; and 120 min. The relative radioactivity of the three tyrosine-containing peptides is shown here in Part A, (left) for the 15, 30-, and 60-minute samples in histogram form. Peptides 1, 2, and 3 contain tyrosines 28, 53, and 58, respectively (see Fig. 6 (NC). Tryptic peptides of samples iodinated for different reaction times were separated by high voltage paper electrophoresis, pH 3.5 (pyridine/acetate). The reaction times were: 2, 15 min; 3, 30 min; 4, 60 min; 5, 120 min. The relative radioactivity is summarized in Table III. DISCUSSION In this study, the accessibility of tyrosine residues in histones from chicken erythrocytes was analyzed. The major finding is that differences in the accessibility of various tyrosine residues in different regions of a histone sequence on the chromatin may be assessed by either chloramine-T or lactoperoxidase iodination.
Purified histones are more susceptible to iodination, whether using chloramine-T ( Fig. 1) or lactoperoxidase' as a catalyst, than histones bound to chromatin. Chromatin in the native form showed fewer reactive tyrosine residues than sheared chromatin (Fig. 1). These results suggest that some tyrosine residues are either buried or involved in interaction with other proteins and that shearing of chromatin changes their native configuration.
The extent of iodination for histones free in solution is not only greater than that for chromatin-bound histones, but it differs qualitatively as well (Table I). With use of the lactoperoxidase method, the tyrosine residues in HI, HR, and H:, are labeled more than those of He,, HZ),, and H, when bound to chromatin. If H, and Hs are located outside the nucleosomal core (i.e. in the spacer region), one would expect them to be more accessible to iodination than the histone core. However, since Hi1 is within the core, its high reactivity was surprising. It is possible that portions of HZr containing the reactive tyrosines extend outside of the nucleosomes and serve as reactive sites for iodination.
The size of the oblate nucleosome structure has previously been estimated at 100 x 50 A. For this size, it would seem a priori that a large molecule like lactoperoxidase (-60 8, diameter) would have difficulty penetrating the chromatin matrix. Lactoperoxidase-catalyzed iodination, while superior to chloramine-T in terms of mildness, must be carefully controlled as I2 generated from iodide may nevertheless penetrate a compact structure. We have used an excess substrate to enzyme:iodine ratio as recommended by Morrison and coworkers (1974) to avoid this problem. Whether lactoperoxidase will react with a nucleosome and induce change in its conformation is not known; however, no detectable changes have been observed with membrane proteins.
The use of 125-iodine to probe protein conformation has been used to study histone H., in Xenopus laevis chromatin (Biroc and Reeder, 1976) and the present results extend the applicability to both histones Hq and H, in another chromatin system. The data obtained in this laboratory with chicken erythrocyte Hq, although comparable, can not be directly compared with those of earlier works since a complete sequence of neither X. laevis nor chicken H.1 has been determined. It is significant that in both systems only 2 of the 4 tyrosine residues of H, on chromatin are relatively available for iodination. While many differences exist between the chromatin of Xenopus cells and chicken erythrocytes, notably the lack of transcriptional activity, longer nucleosomal DNA, and the presence of HR in nucleated chicken erythrocytes, the same basic nucleosomal structure in which a specific H,l conformation plays a role must be there. Our studies imply that those tyrosine residues not readily accessible for iodination in the chromatin are involved in interaction with DNA and with other chromosomal proteins. It can be pointed out that this is the first study showing a conformational difference between free and bound histone Hz by radioiodination.
Hn, being internucleosomal, is expected to be more available for iodination than the intranucleosomal histones such as H,. This is generally confirmed ( Table  I).
The meaning of this general quantitation, however, could be ' G. R. Griffiths and I'. C. Huang, unpublished observations.  Table  II). In contrast, the chloramine-T reaction, while favoring the labeling of Tyr 53, did not give the same results for the other residues (Method A, Table II).
The changes in profile of label with time are noted in Table  II. Fig. 11 shows a distinct change in labeling profile in the major tryptic peptides for H, bound to chromatin. This is in contrast to the relatively uniform labeling of all four peptides, through the respective tyrosines they carry, in free H., (Fig. 9 and Table  III). Peptide 2 specifically can be seen to increase in labeling with reaction time, although iodinated only to about one-tenth of peptides 1 and 4 (Table   III). Similarly, the tyrosine residue in peptide 3 was also iodinated to a lesser extent throughout the reaction.