A Conformational Study of the Binding of a High Mobility Group Protein with Chromatin*

The nature of the binding of a high mobility group protein (HMG 17) to native and Hl-H5-depleted chicken erythrocyte chromatin was studied, as a func-tion of ionic strength, using circular dichroism and thermal denaturation techniques. The circular dichro- ism properties of the HMG 17-reconstituted whole chromatin and H1-H5-depleted chromatin demonstrated that a condensation of chromatin structure occurred upon HMG 17 binding at low ionic strength (1 n m Na phosphate, 0.25 nm EDTA, pH 7.0). Thermal denaturation profiles confirmed this change in the structure of chromatin induced by HMG 17. Thermal denaturation profiles were resolved into three-component transitions. In general, a shift in the temperature of maximum dh/dT for each transition (T,,,) was observed for all transitions upon HMG 17 binding. DNA melting in the first transition, originating from linker regions of whole chromatin, was nearly totally depleted and was distributed mainly into the highest melting transition. The same trends were also observed in H1-H5-depleted chromatin. These results indicate that the binding sites of HMG 17 are situated in the linker regions immedi- ately adjacent to the core. The nature of the interaction of HMG 17 at higher ionic strength (50 mM NaC1,l mM Na phosphate, 0.25 nm EDTA, pH 7.0) with whole chromatin and H1-H5-depleted chromatin was found to be different but a decrease in [O] values was found in both chromatins. These observations suggest that HMG 17 does not loosen chromatin structure but produces an

The nature of the binding of a high mobility group protein (HMG 17) to native and Hl-H5-depleted chicken erythrocyte chromatin was studied, as a function of ionic strength, using circular dichroism and thermal denaturation techniques. The circular dichroism properties of the HMG 17-reconstituted whole chromatin and H1-H5-depleted chromatin demonstrated that a condensation of chromatin structure occurred upon HMG 17 binding at low ionic strength (1 n m N a phosphate, 0.25 nm EDTA, pH 7.0). Thermal denaturation profiles confirmed this change in the structure of chromatin induced by HMG 17. Thermal denaturation profiles were resolved into three-component transitions. In general, a shift in the temperature of maximum dh/dT for each transition (T,,,) was observed for all transitions upon HMG 17 binding. DNA melting in the first transition, originating from linker regions of whole chromatin, was nearly totally depleted and was distributed mainly into the highest melting transition. The same trends were also observed in H1-H5-depleted chromatin. These results indicate that the binding sites of HMG 17 are situated in the linker regions immediately adjacent to the core. The nature of the interaction of HMG 17 at higher ionic strength (50 m M NaC1,l m M Na phosphate, 0.25 nm EDTA, pH 7.0) with whole chromatin and H1-H5-depleted chromatin was found to be different but a decrease in [O] values was found in both chromatins. These observations suggest that HMG 17 does not loosen chromatin structure but produces an overall stabilization and condensation of structure. The implications of these results to the currently accepted models of transcriptionally active chromatin are discussed.
In eucaryotic nuclei, DNA is packaged into regularly spaced nucleoprotein units called nucleosomes (1)(2)(3). Digestion with nucleases indicates that transcriptionally active regions are also organized into nucleosome-like structures, but that they have an altered conformation from inactive chromatin (4). It is now believed that DNase I sensitivity reflects the potential of a gene to be copied or having had a history of transcription, since the globin gene remains sensitive in mature avian erythrocytes that are no longer active in transcription (4)(5)(6). Mononucleosomes released from nuclei during brief micrococcal nuclease digestion are found to be enriched in transcribing sequences (7) and are also enriched in non-histone chromosomal proteins, particularly HMG' 14 and 17 (8)(9)(10)(11). Actively transcribing genes are sensitized to DNase I in the presence of HMG 14 and 17 (12). This sensitivity can be abolished by the removal of HMG 14/17 from the nuclei and could be re-established by the addition of HMG proteins (6,12). Dixon and co-workers (9)(10)(11)13) extensively studied the digestion of trout testis nuclei and showed that HMG proteins can be major structural components of transcribing genes. The exact location of HMG proteins within the chromatin subunit is unknown.
Binding of HMG 14/17 to mononucleosomes has been previously studied (14)(15)(16). HMG 14/17 do not bind independently of each other and are found to substitute for each other (15,16). Thermal denaturation studies have indicated that the binding sites of these proteins were in the region between 145-166 base pairs where the ends of two full turns of DNA emerge (15) after being wound around the histone core. This is consistent with the model put forward by Goodwin et al. (8).
To our knowledge, the binding of HMG 17 with high molecular weight chromatin has not been studied in detail. To probe the possible effects of HMG 17 on chromatin structure and possibly on higher ordered chromatin structure, z.e. internucleosomal interactions, studies were performed with high molecular weight chicken erythrocyte chromatin. Previously it had been shown that the DNase I sensitivity of the globin gene in chicken erythrocyte chromatin was abolished by removzl of HMG 14/17 but re-established by the addition of HMG 17 (6). We report here the results of thermal denaturation and CD studies of the binding of HMG 17 to both native and Hl-H5, non-histone chromosomal protein-depleted high molecular weight chicken erythrocyte chromatin (stripped chromatin). These methods are known to be sensitive indicators of histone and DNA conformations within chromatin (17). CD results indicated that HMG 17 produced a compaction of whole chromatin and stripped chromatin structure. Thermal denaturation studies indicated that HMG 17 bound to the linker regions in chromatin causing stabilization of the structure.

MATERIALS AND METHODS
Preparation of HMG 17-HMG 17 was prepared by a modified method of Walker and Johns (18). The procedure adapted is briefly discussed below. All procedures were carried out at 4 "C and all ' The abbreviations used are: HMG, high mobility group proteins; T,, temperature of maximum dh/dTfor each transition; %HT, percent of total hyperchromicity found in a given transition; SDS, sodium dodecyl sulfate; Na phosphate, pH 7.0, prepared from NaHzPOs and adjusted to pH 7.0. buffers contained 0.5 m~ phenylmethylsulfonyl fluoride (Sigma, stock solution is 0.1 M in dry dioxane) as a proteolytic inhibitor. The pellet nuclei (obtained as described by Walker and Johns (18)) from 2 liters of blood from freshly killed chickens were extracted twice with 400 ml of 5% perchloric acid (v/v) and centrifuged at 20,000 X g for 30 min after each extraction. The combined extract was filtered through a sintered glass filter (medium porosity) and acidified to 0.3 M HC1 and precipitated by addition of 6 volumes of acetone. Approximately 2.3 g of dry protein was obtained.
The protein obtained was &solved in approximately 60 ml of 75 mM boric acid, 0.1 M NaC1, pH 8.8. The pH of the protein solution was adjusted to 8.8 by addition of NaOH, and the solution was dialyzed against the same buffer overnight. This solution was concentrated, by immersing the dialysis tubing containing protein in dry Sephadex powder, to 30-ml volume. This solution was applied to a CM-Sephadex C-25 column (2.6 X 60 cm, Pharmacia), and the proteins were fractionated using a linear salt gradient

Removal of HI and H5 from High Molecular Weight Whole
Chromatin-Stripping of lysine-rich histones H1 and H5 was carried out using a modification' of the procedure described by Thoma and Koller (21). Whole chromatin (A260 = 5.0) was adjusted to 0.4 M NaCl, 0.05 M Na phosphate, pH 7.0, with an equal volume of 0.8 M NaC1,O.l M Na phosphate, pH 7.0. Dowex AG 50W-X2 resin, which had been converted to the Na' form and then equilibrated with 0.4 M NaCl, 0.05 M Na phosphate, pH 7.0, was slowly added with gentle stirring (1 d of resin/mg of DNA) at 0 "C. The stirring was continued for 1 h. The resin was allowed to settle at 0 "C and the supernatant containing stripped chromatin was removed. No lysine-rich histones were detected on 13% polyacrylamide-SDS gels (22)  Reconstitution of HMG I7 with Stripped Chromatin-Reconstitution of HMG 17 with stripped chromatin was done exactly as described for whole chromatin.
Thermal Denaturation-Chromatin and HMG 17-reconstituted samples were thermally denatured in 0.25 mM EDTA, 1.0 m~ Na phosphate, pH 7.0. The solvent composition was achieved by exhaustive dialysis at 4 "C. Denaturation profiles were recorded using a Gilford UV-VIS microprocessor-controlled spectrophotometer system 2600 with the thermal programmer 2527 accessory. The initial absorbance of the samples was between 0.9 and 1.2 (AzM) in I-cm cells at 25 "C. Quartz cells of path length 1 cm (0.3 m l ) with tight-fitting Teflon stoppers, supplied by Gilford, were used for thermal denaturation studies. Chromatin samples were degassed by gently bubbling helium gas through them before filling the cuvettes. Absorbance and air as blank. Cells were heated electrically from 35-97 "C at 0.25 "C/ temperature of the solutions were recorded at 1-min intervals using min. Plots of hyperchromicity (h) and its derivative, dh/dT, against temperature were obtained using a PDP 11 computer. Data acquisition and reduction were performed as described previously (20, 26). The derivative dh/dT, the denaturation envelope, was resolved into component thermal transitions by Guassian curve-fitting on a DuPont 310 curve resolver. The ratio of A~/ A~H ) was always less than 0.01. The A33Q value did not change throughout the melting transition.
Spectroscopic Analysis-Absorption spectra were recorded on a Cary 14 spectrophotometer at 23 "C. Circular dichroism measurements were made at 23 "C on a Cary 60 spectropolarimeter equipped with a 6001 CD accessory in a 1.0-cm path length cell for the DNA region as described previously (27), with the original photomultiplier tube replaced with an end-on Hamamatsu tube, No. R375. Concentrations of DNA nucleotide residue were determined by absorption at 258 nm using €258 = 6800 cm" mol nucleotide". The concentrations of the sample used were 1.0-2.0 X M nucleotide residue in solutions of 0.25 mM EDTA, 1.0 m~ Na phosphate, pH 7.0. The reported spectra represent averages of three independent preparations. CD spectra in the protein region were recorded using a 0.2-cm cell (Optical Cell Corp.). The molar protein concentration of core was assumed to be %w of the molar DNA concentration in nucleotides in stripped chromatin. The molecular weight of core was assumed to be 108,500 and the mean amino acid residue weight was taken as 111. For HMG 17, ellipticity calculations were made using a mean residue weight of 104 (23).
SDS-Polyacrylamide Gel Electrophoresis-Electrophoresis of histones was performed on sodium dodecyl sulfate-polyacrylamide gels by the modified Laemmli (28) method as described by Maize1 (22). A 25-cm separating gel of 13% acrylamide was employed with a 1.0-cm stacking gel of 3% acrylamide. Electrophoresis was performed as described previously (20).
Quantitation of HMG 17-HMG 17 concentrations were determined by the assay of Lowry et al. (30) and UV spectral measurements. Standard curves were generated using bovine serum albumin which had been carefully calibrated against H4 (correction factor = bovine serum albumin miUigrams/d X 0.755). HMG 17 concentrations were then calculated by using A: : , , , , , = 8.67 based upon H4 as standard in the method of Lowry et al. (30). The staining response of H4 and HMG 17 was found to be linear in the range of the reconstitution experiments and the average staining ratio of HMG 17/H4 was found to be 0.97 f 0.03. dialysis tubing (Spectrum Medical Industries Inc., Los Angeles, CA) Miscellaneous-Dialysis was carried out using Spectrapor No. 3 which had been prepared as described previously (20). Glass-distilled water and analytical grade reagents were used throughout.

Evaluation of HMG I7 Bound After Reconstitution
The amount of HMG 17 bound during reconstitution was determined by SDS-gel electrophoresis. The protein contents of various reconstitutes are listed in Table I. A typical gel scan (Fig. 1) shows the relative migration of HMG 17 and the histones in HMG-reconstituted whole chromatin. The relative staining intensities of HMG 17 were found to be 0.20 and 0.44 with respect to H4 for input molar ratios of 1  Hl-H5-stripped chromatin. This loss of material may be partly due to the adsorption of this protein on the dialysis tubing during gradient dialysis. It is known that 1 mol of HMG 17 is present for every 10 or 20 nucleosomes in whole nuclei (12). Thus, the amounts of HMG 17 bound after reconstitution are far above the ratio present in whole nuclei.

Thermal Denaturation Studies
Whole Chromatin HMG I7 Reconstitute-Whole chromatin and HMG 17-reconstituted samples were thermally denatured in solutions containing 0.25 m EDTA, 1.0 mM Na phosphate, pH 7.00. The derivative hyperchromicity profiles of whole chromatin and HMG 17-reconstituted samples are presented in Fig. 2. The transition midpoint (T,) as well as the relative areas of the thermal transitions (%HT) of both control and the HMG-reconstituted whole chromatin are given in Table 11. Whole chromatin in the present study exhibited three well defined transitions with transition midpoints at 62, 71.6, and 83.1 "C. The (Table  11). We assume that -23 base pairs of DNA melting in the fist transition originate from the linker regions between the adjacent nucleosomes, The second transition, representing 64 base pairs, originates from the unstacking of weakly bound core DNA immediately adjacent to the linker region, and the third transition of 112 base pairs is from core DNA. When 0.88 mol of HMG 17/nucleosome was reconstituted with whole chromatin, the fist transition almost disappeared, leaving 2% of the total hyperchromicity, and the material was distributed to the second and third transitions. A rise of about 4 "C in the T,,, of the transition I, was observed, whereas the T,,, of transition 111, remained unaffected. This observation is a direct indication that HMG 17 binds to the iinker regions in whole chromatin. When HMG 17 binds to the linker region, most of the DNA present in that region melts in transition 111, , as the %HT lost in transition I, is found in transition 1 1 1 , .

Thermal Denaturation Studies with Stripped Chromatin-
Histones H1 and H5 have been assigned to be in the internucleosomal regions of whole chromatin (31). Our studies of the HMG interaction indicated the probable binding of HMG to the linker region in the presence of H1. To further investigate the role of the H1 in HMG binding, thermal denaturation studies were performed with stripped chromatin (Hl-H5 and non-histone chromosomal protein-depleted).
The melting of stripped chromatin was studied in 0.25 mM EDTA, 1.0 mM Na phosphate, pH 7.0. Stripped chromatin at this ionic strength showed three well defined transitions in the denaturation profde as shown in Fig. 2 and Table 11.
For Hl-H5-stripped chromatin, the fist transition represents 50% of the hyperchromicity which corresponds to 100 base pairs of DNA, originating from 60 base pairs of the linker and 40 base pairs of the core region. Since Hl-H5 are completely absent in core chromatin, linker region represents free naked DNA. The second transition, located at 70.8 "C, represents 15% of the total hyperchromicity or 30 base pairs of DNA associated with core adjacent to linker regions. The 35% of the total hyperchromicity melting in the third transition represents the remaining 70 base pairs of DNA which are strongly associated with core histones. When 0.46 mol of HMG 17/nucleosome bound to core chromatin (stripped), 8% of the hyperchromicity in the transition I, was lost and dis-  17 is different from that of Hl/H5 binding to core chromatin even though the CD values for the two samples are the same (Table 111).

CD Studies of HMG 17-reconstituted Whole Chromatin-
The circular dichroism spectrum of whole chromatin above 250 nm is entirely due to the DNA component of chromatin, and thus can be used as a sensitive probe of DNA conformational changes of whole chromatin (17). Whole chromatin and its HMG 17-reconstituted products were examined in solutions containing 0.25 m~ EDTA, 1.0 mM Na phosphate, pH 7.0. CD spectra in the DNA region were recorded from 350 to 250 nm as shown in Fig. 3 For HMG-reconstituted whole chromatin, a reduction in ellipticities was always found. For bound ratios of 0.40 and 0.88 mol of HMG 17/200 base pairs of DNA, the values of [8]282.5 = 3700 f 300 and 3300 f 300 deg cm*.dmol", respectively, were obtained. The ellipticities at the shoulder at 272.5 nm were also decreased. The values of the ellipticities of whole chromatin and HMG 17-reconstituted samples are given in Table 111.

CD Studies of HMG 17-Reconstituted Hl-H5-depleted
Chromatin (Core Chromatin)-The CD spectra of control stripped chromatin and its reconstitution products with HMG 17 in 0.25 mM EDTA, 1.0 m~ Na phosphate, pH 7.0, are seen in Fig. 4.
Stripped chromatin exhibits a larger maximum ellipticity in the DNA region than does the native whole chromatin due to the removal of Hl-H5 histones. Complete removal of Hl-H5 causes an increase of ellipticity, s. from 4100 to 5000 deg    Table I. Molar ellipticity values; error f300.
cm'.dmol" (Table 111). This increase in ellipticity may represent the release of tertiary structural constraints placed upon DNA by lysine-rich histones. Reconstitution of HMG 17 to Hl-H5-depleted chromatin is accompanied by a reduction in ellipticity. When 0.84 mol of HMG l7/nucleosome were bound with stripped chromatin a CD spectrum was obtained that was identical to that of whole chromatin. This indicates that HMG 17 can produce the same conformational changes that are induced by Hl-H5 in whole chromatin. Since H1 is assumed to bind outside the core, i.e. in the linker region, HMG 17 may also be binding in a similar manner.
The CD spectrum of chromatin above 250 nm has a lower ellipticity than native DNA. Cowman and Fasman (32) interpreted the chromatin CD spectrum in the nucleotide region as being composed of two independent components. One component is a conventional B-type DNA spectrum and the other, a *-type spectrum with a negative [8]275 contribution due to the DNA wrapped around the core histones (33). The protein-free linker DNA gives a simple B-type spectrum with a positive ellipticity -9000 deg cm'.dmol". Another explanation for the decrease in [8] values of DNA in chromatin is attributed to a change in the winding angle of DNA when it is wrapped around core histones in chromatin (34), i.e. a secondary structural change. Since the binding of HMG 17 is accompanied by a further reduction in [e] values in the nucleotide regions of both native whole chromatin and core chromatin, it is assumed that an additional \k band contribution or change in winding angle of additional DNA is produced in the HMG-bound complexes of chromatin. This would be in agreement with data from thermal denaturation which indicated a loss of DNA from the linker region upon binding of HMG 17. This implies that some unbound DNA becomes associated with the core upon HMG 17 binding, but other conformational changes resulting in a reduction of ellipticity cannot be excluded. The magnitude of the reduction in the [8]282.5 value was found to be essentially the same for native whole chromatin (from 4100 to 3400 deg cm2. dmol") and for stripped chromatin (from 5000 to 4100 deg cm2.dmol") on binding -1 mol of HMG 17/nucleosome. This means that the *-type band contribution or change in the winding angle produced by HMG 17 is essentially the same in whole and core chromatin, although there are additional conformational changes caused by H1 in whole chromatin.

Circular Dichroism Studies: High Ionic Strength
Recent reports have shown that nucleosomes and chromatin can assume different conformations depending upon the salt concentration (20, 26,35,36). Therefore, the binding of HMG 17 to whole chromatin and H1-H5-depleted chromatin has been studied in a medium containing 50 m NaC1, where differences in condensation have been shown to exist relative to low ionic strength (31). In addition, it has been reported that there are different binding sites for HMG 17 in chromatin effective at various ionic strengths (11,15).

CD Studies of HMG 17"reconstitutecE Whole Chromatin-
The CD spectra of the HMG 17 reassembled whole chromatin complex and control whole chromatin in a medium containing 50 m~ NaC1, 0.25 mM EDTA, 0.1 m~ Na phosphate, pH 7.0, were studied (Fig. 5). Above 50 m NaCl, chromatin slowly started precipitating, so that CD spectrum could not be taken.
Control whole chromatin had ellipticity peaks at [8]282. 5 and [8]272,5 with values of 2900 and 2300 k 300 deg cm2.dmol", respectively (Table 111). When approximately 1 molecule of HMG 17 was bound for every 2 nucleosomes (0.40 mol/nucleosome), the CD spectrum of the whole chromatin was unaffected, indicating that no condensation had occurred. When 1 molecule of HMG 17 was bound for every nucleosome (0.88 mol/nucleosome), the [81282.:, decreased to 2500 deg cm2dmol", thus indicating some compaction of structure. Thus, it would appear that for condensation to occur at higher ionic strength it is necessary to have HMG 17 associated with each nucleosome. This might imply an interaction between neighboring nucleosomes at higher ionic strengths. From the data presented in Table I11 it is not possible to determine whether the binding sites are all equivalent. Kuehl et al. (11) have shown that chromatin contains several classes of HMG 17-binding sites with different association constants. The data presented above may be viewed in the light of possibly two  different binding sites. CD Studies of HMG 17-reconstituted HI-HS-depleted Chromatin-The CD spectrum of Hl-H5-depleted chromatin in 50 mM NaCl was very similar to that of whole chromatin at low ionic strength (Fig. 6). When 0.46 mol of HMG 17 was bound/200 base pairs of DNA of stripped chromatin, a decrease in [@]z,z.z of about 500 deg cm2 .dmol" was observed (Table 111). When an additional amount of HMG-17 (0.84 mol of HMG 17) was bound/nucleosome no further decrease in ellipticity was observed. Thus, when =l mol of HMG 17 was bound12 nucleosomes in stripped chromatin, the HMG 17 caused compaction of the DNA; however, no further condensation occurred when the ratio of HMG 17 was increased to -1 mol/nucleosome. Thus, the mechanism of binding HMG 17 at 50 mM NaCl differs in whole chromatin and stripped chromatin. Therefore, H1 changes the mode and effect of binding of HMG 17 at higher ionic strength.
The [8]282,5 for whole chromatin was lower than that found for a reconstitute of 0.84 mol of HMG 17/nucleosome with stripped chromatin in 50 m~ NaC1. Thus, if HMG proteins displace H1 in actively transcribing chromatin, a decondensation could occur.
Circular Dichroism Spectral Studies in the Protein Region-To investigate whether HMG 17 produces conformational changes in the core histones upon binding, CD spectral measurements were carried out in the protein region from 250 to 200 nm. The region below 250 nm is mainly due to the optical activity of the histone peptide chromophore. The magnitude of ellipticities a t 208 and 220 nm is a measurement of the secondary structure of the core protein (17). Identical CD spectra were observed below 250 nm for both HMG 17reconstituted core chromatin and the control core chromatin under identical conditions. Thus, binding of HMG 17 does not alter the conformation of the core histones.

CONCLUSIONS AND SUMMARY
The binding of HMG 14/17 to mononucleosomes has been shown to produce two major additional slow moving bands in nondenaturing polyacrylamide gels (14)(15)(16) indicating the binding of 1 and 2 molecules of HMG 14/17 per nucleosome. The reaction of HMG 14/17 with nucleosomes at a low ionic strength was found to be reversible and noncooperative. The mechanism of HMG 17 binding was dependent upon ionic strength of the medium. At a higher ionic strength, only one additional band of HMG 17 bound nucleosomes was found in the polyacrylamide gels (2 molecules of HMG 14/17 per nucleosome). Thus, the interaction of HMG 17 with nucleosomes at higher ionic strength was cooperative in nature (15). Differences in binding at low and high ionic strength were also obtained in our CD studies.
Goodwin et al. (8) postulated that in active chromatin, H1 is replaced by the binding of the NH2-terminal basic regions of HMG 14/17 to 10-20 base pairs of DNA immediately contiguous to the core region and the COOH-terminal region binds to the core particle itself, weakening the histone DNA interaction in that region. The DNase I digestion studies on poly(dA-dT)-reconstituted core histone complexes containing HMG 14 showed altered digestion rates at selected sites which is in agreement with the above suggestion (16). However, Sandeen et al. (15) demonstrated that HMG 14/17-stabilized mononucleosomes as detected by thermal denaturation studies and concluded that HMG 14/17 bound to 20-25 base pairs at the end of the core particle. The results obtained herein from the thermal denaturation studies indicate that the binding of HMG 17 occurs to both the linker DNA and to the DNA contiguous to the core region of chromatin and that it stabilizes the structure. HMG 14/17 can be cross-linked to both H1 and the core histones, even with a "zero length" condensing agent (38). This again suggests that the position of HMG 14/17 is contiguous with the nucleosome core and H1 in chromatin. Thus, binding of HMG 17 need not interfere in any way with H1 binding in chromatin. The CD and thermal denaturation studies herein indicated that a condensation and stabilization of chromatin structure was produced by HMG 17 binding, which is in agreement with the reduced DNase I sensitivity of bulk nucleosome-HMG 14/17 complexes (15).
From the above discussion, it is evident that the question remains to be answered as to how HMG 14/17 confers special DNase I sensitivity to active gene sequences in chromatin. It should be pointed out that an overall stabilization of chromatin does not necessarily mean the inhibition of RNA polymerase binding to local segments of chromatin. HMG 17 may be involved in the modifkation of the internucleosomal interactions. Weintraub (38) has proposed a model wherein the condensation of chromatin is necessary to align recognition sites for RNA polymerases. Another explanation may be that the postsynthetic modifications of histones such as acetylation, phosphorylation, and methylation (39), coupled with the binding of HMG 14/17 produce the necessary conformational changes for transcription. In addition, HMG 14/17 can undergo postsynthetic modification, such as ribosylation by poly(ADP ribose) (40), acetylation (41), methylation (42), and phosphorylation (43). Thus, the modification of HMG 17 may also have a role in their association with the active gene.
From the results obtained herein, there was no evidence to suggest a weakening of internucleosomal interactions in high molecular weight chromatin, which would have been undetectable in the previous monomer studies. It has been shown that HMG 17 causes an overall compaction and stabilization of native chromatin and H1-H5-depleted chromatin, and it is evident that the binding of HMG 17 to chromatin, as studied herein, would not be responsible for the increased DNase I sensitivity found in active genes.