Eur. J. Biochem. 77, 471 -477 (1977)
Conformational Studies on Histone H3 and Its CNBr Peptides Gregory MORRIS and Peter N. LEWIS Department of Biochemistry, University of Toronto (Received February 8, 1977)
Histone H3 and H3 peptides 1-120, 1-90, 91-135, 91-120 and 121-135 have been prepared and examined for salt-induced conformational changes by circular dichroism measurements. It was found that reduced histone H3 and the reduced peptides 1 - 120,91- 135 and 91 - 120 exhibit biphasic changes with the formation of a-helix and p structures. H3 peptide 1-90, on increasing the ionic strength to moderately high levels, monophasically formed appreciable quantities of a-helix and p structures, while peptide 121- 135 remained unfolded under all ionic strengths examined. All the above peptides except 121-135 also aggregate when the ionic strength is raised. The salt-induced near-ultraviolet circular dichroic spectra of histone H3 and peptide 1-90 were found to be very similar, suggesting that the conformational changes induced in the peptide 1 - 90 are essentially the same as those observed for the intact histone. These results support the contention that the polypeptide segments of this histone interact initially by parallel self-association followed by the formation of even larger aggregates on a longer time scale. Direct evidence [ l - 121 for the existence of a tetrameric histone H3-H4 complex in free solution of Z = 0.1 M and indirect evidence [13- 151 for the presence of a similar complex in chromatin has been obtained by a variety of experimental approaches. While the three-dimensional arrangement of the histones in this complex is still unknown, Lewis [12] has recently determined some of the features of the structure in free solution. In that study it was proposed that the initial structural changes that histone H4 undergoes by itself in solutions of moderate ionic strength [16] are probably involved in the process of complexation. We report here the results of a parallel study on the structure of histone H3 and its fragments that gives added support to the proposal [7,8,12] that tetramer formation involves the interaction of the C-terminal portions of a histone H3 dimer and a histone H4 dimer. MATERIALS AND METHODS
Preparation of Histone H3 and H3 CNBr Peptides Histone H3 was prepared from frozen calf thymus tissue by the method of Hooper and Smith [17] and was further purified by gel filtration on a column (3.2 x 150 cm) of Bio-Gel P-10 (Bio-Rad) eluted with 10 mM HCl. The histone H3 obtained was free from Abbreviations. CD, circular dichroism; CNBr, cyanogen bromide.
contamination ( < 5 %) by other histones as judged by electrophoresis on urea/acetic acid gels [181. Cleavage of histone H3 at methionine residues 90 and 120 was carried out as described by Brandt and von Holt [19]. The resulting lyophilized reaction mixture (from 100mg of histone H3) was dissolved in 4 ml of 8 M urea, 20 mM sodium phosphate pH 7 containing 2 % 2-mercaptoethanol and incubated for 2 h at 37 "C. This solution was then acidified to pH 2 with 1 M HCl and loaded onto a Sephadex G-50 column (2.7 x 120 cm) eluted with 10 mM HCl at a flow rate of 50 ml/h. The appropriate fractions were then pooled and lyophilized. The larger peptides eluted from the above column in a single peak with the void volume and were rechromatographed on a Bio-Gel P-10 column (2.7 x 120 cm) eluted with 10 mM HC1. In this way, histone H3 fragments 1-90, 91-135, 121-135 and 91-120 were obtained in quite pure form. H3 peptide 1- 120 was found to be contaminated with unreacted histone H3 and was further purified by preparative electrophoresis as follows. A large gel (2.2 x 10 cm) containing 15 % acrylamide, 2.6 M urea and 0.9 M acetic acid was prepared and 0.5 mg of crude fragment 1 - 120 (2 mg/ml in 8 M urea) was applied to the gel surface. The gel was electrophoresed for 160 min at 150 V with a reservoir buffer of 0.9 M acetic acid. The gel was removed from its tube and a thin longitudinal slice was cut with a razor blade and stained with Coomassie blue G [20] to visualize the protein bands. The remaining large gel piece was then sliced transversely to obtain fragment 1- 120
412
Conformational Studies on Histone H3 and Its CNBr Peptides
free from histone H3. This slice was placed into a dialysis bag and the peptide was eluted electrophoretically into the bag. Recovery by acetone precipitation from the solution in the bag was about 65 % of the theoretical value. Peptide purity was determined by amino acid analysis N-terminal assays and electrophoresis on the 30 polyacrylamide ureajacetic acid gels described elsewhere [ 161.
Assay for the Extent of Peptide Aggregation The tendency of histone H3 and its CNBr peptides to aggregate at pH 7, I = 0.1 M was measured by the phosphate gel electrophoresis method of Lewis [ll]. In certain cases, sedimentation velocity measurements were made at 20 "C on a Beckman model E ultracentrifuge equipped with Schlieren optics. Circular Dichroism CD spectra were measured at ambient temperatures on a Durrum-Jasco ORD/CD-15spectropolarimeter equipped with a SS-20 CD modification. Stock solutions of the peptides were prepared by dissolving 2 mg of peptide in 0.3 ml of 8 M urea, 20 mM sodium phosphate, 2 2-mercaptoethanol pH 7 and incubating for 2 h at 37 "C. The solution was then acidified to pH 2 with 1 M HC1 and desalted by means of a column of Sephadex G-25 (0.9 x 15 cm) eluted with Nz-saturated 10 mM HCI, 1 mM dithiothreitol. The peptide concentrations of the stock solutions were determined by a modification of the microbiuret assay of Itzhaki and Gill [21] using histone H4 as a calibrant. CD samples for the far-ultraviolet (200 250 nm) spectra run in 1-mm quartz cells were prepared by dilution of the stock solutions to 0.1 mg/ml. An aliquot of 10 mM NaOH equal to the volume of
the stock solution used was added just before spectral measurement to bring the pH to 3.5 f 0.2. The ionic strength of the CD samples was adjusted to the required value by the addition of solid NaC1. The spectra were recorded at 2 nm/min, an average run taking about 30 min from the salt addition. Spectra were also recorded on the same samples after 24 h. The near-ultraviolet (250 - 300 nm) CD spectra of reduced histone H3 and peptide 1- 90 were measured at 4 mgjml in a 1-cm quartz cell. RESULTS Preparation and Characterization of Histone H3 CNBr Peptides The amino acid sequence [22] of histone H3 shown in Fig. 1 contains two methionine residues, one at position 90 and the other at 120. The elution profile on Sephadex G-50 resulting from partial cleavage at these two methionine residues by CNBr is shown in Fig. 2. Peaks 1, 2 and 3 shown in Fig. 2 were combined and rechromatographed on Bio-Gel P-10. The resulting elution profile is shown in Fig. 3. The peptides in peaks 4, 5 and 6 from the Sephadex separation and in peaks 1, 2 and 3 from the Bio-Gel separation were recovered and examined by amino acid analysis, N-terminal analysis and gel electrophoresis. A minor contaminant in peak 2 of the BioGel separation was uncleaved histone H3. This was removed from the desired product, peptide 1 - 120, by preparative electrophoresis. The results of amino acid analyses on the peptides recovered from Sephadex peaks 4, 5, 6 and the Bio-Gel peaks 1, 2 and 3 are shown in Table 1, together with the expected compositions. N-terminal determinations yielded the amino acids alanine, alanine, proline, alanine, alanine, ala10
Ala-Arg-Thr-Lys-Gln-Thr-Ala-Arg-Lys-Ser-Thr-Gly-Gly-Lys-A~a20 30 Pro-Arg-Lys-Gln-Leu-Ala-Thr-Lys-Ala-Ala-Arg-Lys-Ser-Ala-Pro40
Ala-Thr-Gly-Gly-Val-Lys-Lys-Pro-His-Arg-Tyr-Arg-Pro-Gly-Thr50 60 Val-Ala-Leu-Arg-Glu-Ile-Arg-Arq-Tyr-Gln-Lys-Ser-Thr-Glu-Leu70
Leu-Ile-Arg-Lys-Leu-Pro-Phe-Gln-Arg-Leu-Val-Arg-~lu-Ile-Ala-
DO
90
Gln-Asp-Phe-Lys-Thr-Asp-Leu-Arg-Phe-Gln-Ser-Ser-Ala-Val-Met-
100 Ala-Leu-Gln-Glu-Ala-Cys-Glu-Ala-Tyr-Leu-Val-Gly-Leu-Phe-Glu110
120
Asp-Thr-Asn-Leu-Cys-Ala-Ile-His-Ala-Lys-Arg-Val-Thr-Ile-Met-
130 Pro-Lys-Asp-Ile-Gln-Leu-Ala-Ar~-Arg-Ile-Arg-Gly-Glu-Arg-Ala.
Fig. 1. Amino acid sequence of calf thymus histone H3 [22]
G. Morris and P. N. Lewis
200
473
400
300
-400
300
500
Eluate ( m [ )
500
Eluate ( m l )
Fig.2. Elution profile from a Sephadex G-50 column of u CNBr digest of histone H3. The bars indicate the fractions pooled
Fig.3. Elution profile from a Bio-Gel P - I 0 column of the lurge CNBr peptides from CNBr-cleaved histone H3. The bars indicate the fractions pooled
Table 1. Amino acid composition of histone H3 fragnients n.d. = not determined Amino acid
( 1 - 120)
H3
obsd
cdkd
obsd
(91 - 135)
(91 120)
(121 - 135)
calcd
obsd
calcd
obsd
obsd
calcd
11 1 13 2 8 5 9 5 5 11 4 0 0 3 7 2 3
2.1 0.6 4.2 2.6 2.1 0.7 6.6 1.1 2.3 7.5 2.3 0.4 n.d. 2.7 4.5 0.8 1.0
2 1 5 3 2 0 6 1 2 7 2 1 2 4 5 1 1
1.9 0.9 1.4 2.6 2.1 0.9 4.5 0.3 1.4 4.8 2.0 0.2 n.d. 1.7 4.1 1.0 1.0
1.7 0.1 4.2 0.8 0.2 0.3 2.7 0.6 1.3 2.1 1.1 0.4 n.d. 1.2 1.0 0.1 0.3
1 0 4 1
(1-90)
calcd
obsd
~
calccd
mol/mol Lysine Histidine Arginine Aspartic acid Threonine Serine Glutamic acid Proline Glycine Alanine Valine Methionine Cysteine Isoleucine Leucine Tyrosine Phenylalanine
13.9 2.7 18.7 5.1 9.4 5.1 14.6 6.0 7.4 17.3 5.1 2.0 n.d. 6.5 12.5 2.9 4.1
13 2 18 5 10 5 15 6 7 18 6 2 2 7 12 3 4
11.2 2.2 13.3 4.5 9.7 5.5 13.3 5.6 6.8 16.5 5.9 0.4 n.d. 5.0 9.9 3.2 4.4
12 2 14 4 10 5 13 5 6 16 6 1 2 5 11 3 4
9.1 1.1 12.3 2.4 8.0 5.2 9.8 4.8 5.8 11.0 4.1 0 n.d. 3.0 8.0 2.1 3.3
nine expected for H3 peptides 91-135, 91-120, 121-135, intact histone H3, 1-120 and 1-90 respectively [22]. The results of electrophoresing the purified peptides and the total CNBr digest on urea/ acetic acid gels are shown in Fig.4. Densitometer measurements on these and duplicate gels showed H3 fragments 121-135, 1-320 and 1-90 to be more than 90 % homogeneous while peptides 91 - 120 and 91 - 135 were about 80 % homogeneous. The adjacent multiple bands in the 91-125 gel may be due to oxidized forms of the uncleaved methionine residue at position 120, resistant to CNBr cleavage.
1 1 1 2 2 0 4 0 1 5 2 0 2 2 4 1 1
0 0 2 1 1 2 0 0 0 2 1 0 0
Aggregation of Histone H3 and H3 Fragments
The tendencies of histone H3 and H3 peptides 1-120, 1-90, 91-135, 121-135 and 91-120 to associate in solutions of moderate ionic strength were examined by electrophoresis on neutral nondenaturing gels [ll]. The densitometer traces of the stained gels are shown in Fig. 5, along with the pattern obtained for the lyophilized digest. Histone H3 and peptide 1 - 120 (traces 2 and 3) form large aggregates and barely enter the gel. Peptides 1-90 and 121- 135 (traces 4 and 6) migrate as single bands with no protein
414
Conformational Studies on Histone H3 and Its CNBr Peptides
Fig. 4. Ureajacetic acidpolyacrylamide gels (30 %) of’ the histone H3 CNBr digest and purified peptides. Gel 1 H3, peptides 1 - 120, 1-90, 91 - 135, 91 120 and 121- 135 respectively
=
digest, gels 2 - 7
=
histone
~
Table 2. Aggregation properties oJ H3 fragments on non-denaturing phosphate gels Amounts in each band were determined by a dye elution method [I 11 Peptide
Aggregate band
Intermediate bands
Monomer band
% H3 1- 120 1-90 91 135 91 - 120 121 - 135 ~
-++ Fig.5. Electrophoresis of histone H3 and H3 CNBr peptides on neutralphosphate gels. Densitometer traces of stained gels (1) CNBr digest mixture, (2) histone H3, (3-7) peptides, 1-120, 1-90, 91 - 135,121 - 135 and 91 - 120 respectively. Migration was toward the cathode for all peptides except 91 - 120
98 97 5 18 5 0
0
0 0 79 93 0
2 3 95 3 2 100
remaining at the origin, while peptides 91 - 135 and 91-120 (traces 5 and 7) displayed more than one band. Addition of urea to the non-denaturing gels [12] reduced these multiple bands to one band in each case, indicating multimer formation in peptides 91-135 and 91-120. A dye elution method [ l l ] was used to quantify the distribution of peptide along the gel length in terms of aggregate, intermediate and monomer bands. The results of the analysis are shown in Table 2. Under the conditions of the experiment (0.05 M phosphate pH 7), we interpret the results to mean that histone H3 and peptide 1 - 120 aggregate extensively, peptides 91 - 120 and 91 - 135 form small multimers and peptides 1-90 and 121-135 do not self-interact. Interestingly, the total reaction mixture containing all the peptides and some uncleaved histone H3 appears to exhibit a gel pattern which is simply a linear combination of the gel patterns of the constituent peptides (see Fig. 5) suggesting that these peptides do not interact extensively with one another.
G. Morris and P. N. Lewis
475 0
:::!/f -75
-
-
4 1-120 L
I
I
I
X (nm)
91 -135 210
225 X(nm)
240
-21$q Fl
-5.0
121 -135
91 -120
-7.5
210
225
240
7.5
210
225
240
1(nm) (nm) Fig.6. Far-ultraviolet CD spectra ofhistone H3 and its fragments in solutions of various ionic strengths. Histone H3 in (1) water, (2) 0.01 M
NaC1, (3) 0.025 M NaC1, (4) 0.10 M NaCl measured directly, (5) 0.1 M NaCl after 24 h ; peptide 1-120 in (1) water, (2) 0.05 M NaCI, (3) 0.1 M NaC1, (4) 0.2 M NaCl measured directly, (5) 0.2 M after 24 h ; peptide 1-90 in (1) water, (2) 0.5 M NaC1, (3) 1.0 M NaCl directly and after 24 h ; peptide 91 - 135 in ( 1 ) water, (2) 0.5 M NaCl measured directly, (3) 0.5 M NaCl after 24 h; peptide 91 - 120 in (1) water, (2) 0.25 M NaCl measured directly, (3) 0.25 M NaCl after 24 h; peptide 121- 135 in 0- 1 M NaCl measured directly and after 24 h
Peptide 91 - 120 does not contribute to this pattern as it is negatively charged at pH 7 and therefore does not migrate into the phosphate gel, .as run. Circular dichroic studies on H3 peptide 1-90 reported in the next section indicate that secondary structures are induced at high salt concentrations ( 2 0.5 M NaC1). In order to determine whether or not these changes are accompanied by aggregation, a sedimentation velocity run of H3 peptide 1- 90 in 0.5 M NaCl was made. All of this peptide was observed to sediment as a single symmetrical peak of 17 S supporting the suggestion [16] that structural changes
in the histones (except H1) are usually accompanied by intermolecular association.
Salt-Induced CD Spectral Changes in the Far Ultraviolet (200 - 250 nm) Histone H3 and peptides 1 -90, 1 - 120, 91 - 120 and 91 - 135 undergo conformational changes on increasing the ionic strength of the solvent as the farultraviolet CD spectra in Fig.6 show. H3 peptide 121- 135 did not undergo any detectable changes even in 1 M NaC1, The changes observed for histone H3
416
Conformational Studies on Histone H3 and Its CNBr Peptides
Table 3. Secondary structures in histone H3 and H3 peptides from CD measurements The salt concentration is the lowest at pH 3.5 which gave the maximum change. The structure values given are in residues, the total change observed was the sum of the fast and slow changes Peptide
NaCl
I
I
I
I
I
I
I
I
fi structure
or-Helix fast
slow
fast
slow
15 19 15 1 0 0
4 1
9 10 15 0 0 0
10 3 0 0 2 0
M H3 1-120 1- 90 91-135 91-120 121-135
0 10
0.20 1.00 0.50 0.25 1.00
0
2 3 0
.
If
I I 250
H3
260 I
and peptides 1- 120, 91 - 120 and 91 - 135 occurred on two distinct time scales. Isenberg and co-workers [23 - 251 have studied this phenomenon for intact histone H3 as well as for the other histones at pH 7. It was found by these researchers that the magnitude of both the fast and changes depended upon the salt concentration, the protein concentration and the ambient temperature. Our results for H3 and the 1-120, 91-120 and 91-135 at pH 3.5 these dependencies. Measurements were also made at pH 7 {not shown). The only difference observed was that the amount of salt required to effect a given change was slightly less than at pH 3.5, this effect being also observed for histone H4 [16]. The spectra shown in Fig. 6 were analyzed for their secondary structure content by a constrained leastsquares analysis using the a-helix and p-structure curves from Chen et al. [26] and the 10mM HC1 randomcoil spectrum of H3 peptide 1-90 as reference spectra in the range 208 - 244 nm [27]. The results of this analysis for the maximal fast (< 30 min) and slow (> 24 h) salt-induced changes in terms of the number of residues involved in a-helixes and p structures are given in Table 3. Similar results were obtained by using the vector projection method of Baker and Isenberg [27]. The onset of a-helix formation in histone H3 and peptide 1- 120 involves both the fast and slow phases with the major portion ( w 15 residues) in the fast phase. Peptide 1-90, which does not undergo a slow change, attains the helix content of the fast step for the intact molecule, albeit at a much higher salt concentration. The increased number of helical residues in peptide 1- 120 is probably due to the fact (unpublished observation) that the slow change for this peptide is more rapid than that for histone H3 itself, thereby complicating the separation of the two changes. These results suggest that in the intact molecule the a-helix is located before residue 90 on the
250
260
270 280 Wavelength, k ( n m ) I I
270 280 Wavelength, X ( n m )
I
1
290
30(
I
290
300
Fig. 7 Near-ultraviolet CD spectra of histone H3 and peptide 1-90. Histone H3 in water and in 0 15 M NaCl measured directly and after 24 h, peptide 1- 90 in water and in 0.5 M NaCl measured directly and after 24 h
polypeptide chain and that the slow changes involve residues in the region of 91 - 120 as peptide 121- 135 undergoes no salt-induced changes. Only a very small amount of helix is induced in peptides 91 - 120 and 91-135 and this is associated mainly with a slow change. fl structures are associated with both the fast and slow step changes as can be seen from Table 3. The large amount of this conformation induced in peptide 1-90 during the fast step suggests that a major fi structure locus in histone H3 occurs before residue 90. Salt-Induced CD Spectral Changes in the Near- Ultraviolet (250 - 300 nm) Region
The aromatic CD spectra of histone H3 and peptide 1- 90 are shown in Fig. 7. The results shown are typical of those obtained from four separate experiments. The fast-step change spectrum for histone H3 and the spectrum for peptide 1- 90 have virtually the same shape except that the amplitude of the H3 spectrum is exactly two thirds that of the 1-90
G. Morris and P. N. Lewis
spectrum. As there are three tyrosine residues in histone H3 at positions 41, 54 and 99, the fast-step changes in histone H3 almost certainly involve residues 41 and 54. The spectrum resulting from the slow change, presumably due to tyrosine-99, is not greatly changed from that due to the fast change. While these results do not allow us to deduce the secondary structures in which the tyrosines residues are involved, the spectra in Fig.7 suggest that the fast change occurring in histone H3 and the change induced in peptide 1-90 at high salt concentrations involve the very same structures at least around chain sites 41 and 54.
DISCUSSION The data presented in the previous section indicate that the fast-step salt-induced conformational changes in histone H3 involve self-association, extensive helix and B structure formation located primarily before residue 90. The slow-step changes in this histone can be attributed to the section between 91 and 120 and are accompanied by aggregation and the formation of a small amount of helix and fi structures. The region beyond residue 120 in histone H3 appears to be in the random coil condition even under the highest of salt concentrations examined. Histone H3 has been previously studied by CD [23,28], nuclear magnetic resonance [29], electron paramagnetic resonance [30], difference spectroscopy [31] and chemical accessibility determinations [32]. The CD results obtained here for the intact histone are generally in agreement with those obtained by D’Anna and Isenberg [23] and Adler et al. [28]. However, unlike these researchers, we find no loss in helix content on standing. This may be due to the lower pH at which our measurements were made, thereby inhibiting sulfhydryl oxidation. The NMR studies by Bradbury and his colleagues [29] indicate that the interacting region in histone H3 is located between residues 42- 110 in good agreement with our peptide data. The difference spectroscopy results of Palau and Padros [31] support our data from peptide 1- 90 in that tyrosine-41 is involved in the structural changes. At the moment, we have no direct evidence for the exact location of the a-helix and B-structures in histone H3. However, theoretical studies predict that the most likely [33] location of a-helix in histone H3 is from residues 67-111. Since the tyrosine residues at 41 and 54 adopt a specific conformation as evinced by their CD spectra and are unlikely to be involved
477
in a helix [33], this region might be assigned to structures. REFERENCES
1 . Kornberg, R. D. & Thomas, J. 0 . (1974) Science (Wash. D.C.) 184, 865- 867. 2. Roark, D. E., Geoghegan, T. E. & Keller, G. (1974) Biochem. Biophys. Res. Commun. 59, 542- 547. 3. D’Anna, J. A. & Isenberg, I. (1974) Biochemistry, 13, 49934997. 4. D’Anna, J. A. & Isenberg, I. (1974) Biochem. Biophys. Res. Commun. 61,343 - 347. 5. Sperling, R. & Bustin, M. (1975) Biochemistry, 14, 3322-3331. 6. Rubin, R. L. & Moudrianakis, E. N. (1975) Biochemistry, 14, 1718-1726. 7. Weintraub, H., Palter, K. & Van Lente, F. (1975) Cell, 6, 85110. 8. Moss, T., Cary, P. D., Crane-Robinson, C. & Bradbury, E. M. (1976) Biochemistry, 15, 2261 -2267. 9. Sperling, R. & Bustin, M. (1976) Nucleic Acids Res. 3, 12631275. 10. Lewis, P. N. (1976) Biochem. Biophys. Res. Commun. 68, 329335. 11. Lewis, P. N. (1976) Can. J . Biochem. 54, 641 -649. 12. Lewis, P. N. (1976) Can. J . Biochem. 54, 963-970. 13. Bonner, W. M. & Pollard, H. B. (1975) Biochem. Biophys. Res. Commun. 64,282 - 288. 14. Burton, D. R., Hyde, J. E. & Walker, I. 0. (1975) FEBS Lett. 55,77 - 80. 15. Thomas, J. 0. & Kornberg, R. D. (1975) FEBS Lett. 58, 353358. 16. Lewis, P. N., Bradbury, E. M. & Crane-Robinson, C. (1975) Biochemistry, 14, 3391 - 3400. 17. Hooper, J. A. & Smith, E. L. (1973) J . Biol. Chem. 248, 32553260. 18. Panyim, S. & Chalkley, R. (1969) Arch. Biochem. Biophys. 130, 337- 346. 19. Brandt, W. F. &von Holt, C. (1974) Eur. J . Biochem. 46,407417. 20. Reisner, A. H., Nemes, P. & Bucholtz, C. (1975) Anal. Biochem. 64,509-516. 21. Itzhaki, R. & Gill, D. M. (1964) Anal. Biochem. 4,401 -410. 22. DeLange, R. J., Hooper, J. A. & Smith, E. L. (1972) Proc. Natl Acad. Sci. U.S.A. 69, 882- 884. 23. D’Anna, J. A. & Isenberg, I. (1974) Biochemistry, 13, 49874992. 24. Smerdon, M. J. & Isenberg, I. (1974) Biochemistry, 13, 40464049. 25. D’Anna, J. A. & Isenberg, I. (1974) Biochemistry, 13, 20932098. 26. Chen, Y. H., Yang, J. T. & Martinez, H. M. (1972) Biochemistry, 11,4120-4131. 27. Baker, C. C. & Isenberg, I. (1976) Biochemistry, 15, 629-634. 28. Adler, A. J., Moran, E. C. & Fasman, G. D. (1975) Biochemistry, 14,4179-4185. 29. Bradbury, E. M., Cary, P. D., Crane-Robinson, C. & Rattle, H. W. E. (1973) Ann. N.Y. Acad. Sci. 222, 266-289. 30. Palau, J. & Padros, E. (1972) FEBS Lett. 27, 157- 160. 31. Palau, J. & Padros, E. (1975) Eur. J . Biochem. 52, 555-560. 32. Palau, J. & Daban, J. R. (1974) Eur. J. Biochem. 49, 151- 156. 33. Lewis, P. N. & Bradbury, E. M. (1974) Biochim. Biophys. Acta, 336.153-164.
G . Morris and P. N. Lewis *, Department of Biochemistry, University of Toronto School of Medicine, Medical Sciences Building, Taddle Creek Road, Toronto, Ontario, Canada, M5S 1A8
*
To whom correspondence should be addressed.
p
Conformational Studies on Histone H3 and Its CNBr Peptides Gregory MORRIS and Peter N. LEWIS Department of Biochemistry, University of Toronto (Received February 8, 1977)
Histone H3 and H3 peptides 1-120, 1-90, 91-135, 91-120 and 121-135 have been prepared and examined for salt-induced conformational changes by circular dichroism measurements. It was found that reduced histone H3 and the reduced peptides 1 - 120,91- 135 and 91 - 120 exhibit biphasic changes with the formation of a-helix and p structures. H3 peptide 1-90, on increasing the ionic strength to moderately high levels, monophasically formed appreciable quantities of a-helix and p structures, while peptide 121- 135 remained unfolded under all ionic strengths examined. All the above peptides except 121-135 also aggregate when the ionic strength is raised. The salt-induced near-ultraviolet circular dichroic spectra of histone H3 and peptide 1-90 were found to be very similar, suggesting that the conformational changes induced in the peptide 1 - 90 are essentially the same as those observed for the intact histone. These results support the contention that the polypeptide segments of this histone interact initially by parallel self-association followed by the formation of even larger aggregates on a longer time scale. Direct evidence [ l - 121 for the existence of a tetrameric histone H3-H4 complex in free solution of Z = 0.1 M and indirect evidence [13- 151 for the presence of a similar complex in chromatin has been obtained by a variety of experimental approaches. While the three-dimensional arrangement of the histones in this complex is still unknown, Lewis [12] has recently determined some of the features of the structure in free solution. In that study it was proposed that the initial structural changes that histone H4 undergoes by itself in solutions of moderate ionic strength [16] are probably involved in the process of complexation. We report here the results of a parallel study on the structure of histone H3 and its fragments that gives added support to the proposal [7,8,12] that tetramer formation involves the interaction of the C-terminal portions of a histone H3 dimer and a histone H4 dimer. MATERIALS AND METHODS
Preparation of Histone H3 and H3 CNBr Peptides Histone H3 was prepared from frozen calf thymus tissue by the method of Hooper and Smith [17] and was further purified by gel filtration on a column (3.2 x 150 cm) of Bio-Gel P-10 (Bio-Rad) eluted with 10 mM HCl. The histone H3 obtained was free from Abbreviations. CD, circular dichroism; CNBr, cyanogen bromide.
contamination ( < 5 %) by other histones as judged by electrophoresis on urea/acetic acid gels [181. Cleavage of histone H3 at methionine residues 90 and 120 was carried out as described by Brandt and von Holt [19]. The resulting lyophilized reaction mixture (from 100mg of histone H3) was dissolved in 4 ml of 8 M urea, 20 mM sodium phosphate pH 7 containing 2 % 2-mercaptoethanol and incubated for 2 h at 37 "C. This solution was then acidified to pH 2 with 1 M HCl and loaded onto a Sephadex G-50 column (2.7 x 120 cm) eluted with 10 mM HCl at a flow rate of 50 ml/h. The appropriate fractions were then pooled and lyophilized. The larger peptides eluted from the above column in a single peak with the void volume and were rechromatographed on a Bio-Gel P-10 column (2.7 x 120 cm) eluted with 10 mM HC1. In this way, histone H3 fragments 1-90, 91-135, 121-135 and 91-120 were obtained in quite pure form. H3 peptide 1- 120 was found to be contaminated with unreacted histone H3 and was further purified by preparative electrophoresis as follows. A large gel (2.2 x 10 cm) containing 15 % acrylamide, 2.6 M urea and 0.9 M acetic acid was prepared and 0.5 mg of crude fragment 1 - 120 (2 mg/ml in 8 M urea) was applied to the gel surface. The gel was electrophoresed for 160 min at 150 V with a reservoir buffer of 0.9 M acetic acid. The gel was removed from its tube and a thin longitudinal slice was cut with a razor blade and stained with Coomassie blue G [20] to visualize the protein bands. The remaining large gel piece was then sliced transversely to obtain fragment 1- 120
412
Conformational Studies on Histone H3 and Its CNBr Peptides
free from histone H3. This slice was placed into a dialysis bag and the peptide was eluted electrophoretically into the bag. Recovery by acetone precipitation from the solution in the bag was about 65 % of the theoretical value. Peptide purity was determined by amino acid analysis N-terminal assays and electrophoresis on the 30 polyacrylamide ureajacetic acid gels described elsewhere [ 161.
Assay for the Extent of Peptide Aggregation The tendency of histone H3 and its CNBr peptides to aggregate at pH 7, I = 0.1 M was measured by the phosphate gel electrophoresis method of Lewis [ll]. In certain cases, sedimentation velocity measurements were made at 20 "C on a Beckman model E ultracentrifuge equipped with Schlieren optics. Circular Dichroism CD spectra were measured at ambient temperatures on a Durrum-Jasco ORD/CD-15spectropolarimeter equipped with a SS-20 CD modification. Stock solutions of the peptides were prepared by dissolving 2 mg of peptide in 0.3 ml of 8 M urea, 20 mM sodium phosphate, 2 2-mercaptoethanol pH 7 and incubating for 2 h at 37 "C. The solution was then acidified to pH 2 with 1 M HC1 and desalted by means of a column of Sephadex G-25 (0.9 x 15 cm) eluted with Nz-saturated 10 mM HCI, 1 mM dithiothreitol. The peptide concentrations of the stock solutions were determined by a modification of the microbiuret assay of Itzhaki and Gill [21] using histone H4 as a calibrant. CD samples for the far-ultraviolet (200 250 nm) spectra run in 1-mm quartz cells were prepared by dilution of the stock solutions to 0.1 mg/ml. An aliquot of 10 mM NaOH equal to the volume of
the stock solution used was added just before spectral measurement to bring the pH to 3.5 f 0.2. The ionic strength of the CD samples was adjusted to the required value by the addition of solid NaC1. The spectra were recorded at 2 nm/min, an average run taking about 30 min from the salt addition. Spectra were also recorded on the same samples after 24 h. The near-ultraviolet (250 - 300 nm) CD spectra of reduced histone H3 and peptide 1- 90 were measured at 4 mgjml in a 1-cm quartz cell. RESULTS Preparation and Characterization of Histone H3 CNBr Peptides The amino acid sequence [22] of histone H3 shown in Fig. 1 contains two methionine residues, one at position 90 and the other at 120. The elution profile on Sephadex G-50 resulting from partial cleavage at these two methionine residues by CNBr is shown in Fig. 2. Peaks 1, 2 and 3 shown in Fig. 2 were combined and rechromatographed on Bio-Gel P-10. The resulting elution profile is shown in Fig. 3. The peptides in peaks 4, 5 and 6 from the Sephadex separation and in peaks 1, 2 and 3 from the Bio-Gel separation were recovered and examined by amino acid analysis, N-terminal analysis and gel electrophoresis. A minor contaminant in peak 2 of the BioGel separation was uncleaved histone H3. This was removed from the desired product, peptide 1 - 120, by preparative electrophoresis. The results of amino acid analyses on the peptides recovered from Sephadex peaks 4, 5, 6 and the Bio-Gel peaks 1, 2 and 3 are shown in Table 1, together with the expected compositions. N-terminal determinations yielded the amino acids alanine, alanine, proline, alanine, alanine, ala10
Ala-Arg-Thr-Lys-Gln-Thr-Ala-Arg-Lys-Ser-Thr-Gly-Gly-Lys-A~a20 30 Pro-Arg-Lys-Gln-Leu-Ala-Thr-Lys-Ala-Ala-Arg-Lys-Ser-Ala-Pro40
Ala-Thr-Gly-Gly-Val-Lys-Lys-Pro-His-Arg-Tyr-Arg-Pro-Gly-Thr50 60 Val-Ala-Leu-Arg-Glu-Ile-Arg-Arq-Tyr-Gln-Lys-Ser-Thr-Glu-Leu70
Leu-Ile-Arg-Lys-Leu-Pro-Phe-Gln-Arg-Leu-Val-Arg-~lu-Ile-Ala-
DO
90
Gln-Asp-Phe-Lys-Thr-Asp-Leu-Arg-Phe-Gln-Ser-Ser-Ala-Val-Met-
100 Ala-Leu-Gln-Glu-Ala-Cys-Glu-Ala-Tyr-Leu-Val-Gly-Leu-Phe-Glu110
120
Asp-Thr-Asn-Leu-Cys-Ala-Ile-His-Ala-Lys-Arg-Val-Thr-Ile-Met-
130 Pro-Lys-Asp-Ile-Gln-Leu-Ala-Ar~-Arg-Ile-Arg-Gly-Glu-Arg-Ala.
Fig. 1. Amino acid sequence of calf thymus histone H3 [22]
G. Morris and P. N. Lewis
200
473
400
300
-400
300
500
Eluate ( m [ )
500
Eluate ( m l )
Fig.2. Elution profile from a Sephadex G-50 column of u CNBr digest of histone H3. The bars indicate the fractions pooled
Fig.3. Elution profile from a Bio-Gel P - I 0 column of the lurge CNBr peptides from CNBr-cleaved histone H3. The bars indicate the fractions pooled
Table 1. Amino acid composition of histone H3 fragnients n.d. = not determined Amino acid
( 1 - 120)
H3
obsd
cdkd
obsd
(91 - 135)
(91 120)
(121 - 135)
calcd
obsd
calcd
obsd
obsd
calcd
11 1 13 2 8 5 9 5 5 11 4 0 0 3 7 2 3
2.1 0.6 4.2 2.6 2.1 0.7 6.6 1.1 2.3 7.5 2.3 0.4 n.d. 2.7 4.5 0.8 1.0
2 1 5 3 2 0 6 1 2 7 2 1 2 4 5 1 1
1.9 0.9 1.4 2.6 2.1 0.9 4.5 0.3 1.4 4.8 2.0 0.2 n.d. 1.7 4.1 1.0 1.0
1.7 0.1 4.2 0.8 0.2 0.3 2.7 0.6 1.3 2.1 1.1 0.4 n.d. 1.2 1.0 0.1 0.3
1 0 4 1
(1-90)
calcd
obsd
~
calccd
mol/mol Lysine Histidine Arginine Aspartic acid Threonine Serine Glutamic acid Proline Glycine Alanine Valine Methionine Cysteine Isoleucine Leucine Tyrosine Phenylalanine
13.9 2.7 18.7 5.1 9.4 5.1 14.6 6.0 7.4 17.3 5.1 2.0 n.d. 6.5 12.5 2.9 4.1
13 2 18 5 10 5 15 6 7 18 6 2 2 7 12 3 4
11.2 2.2 13.3 4.5 9.7 5.5 13.3 5.6 6.8 16.5 5.9 0.4 n.d. 5.0 9.9 3.2 4.4
12 2 14 4 10 5 13 5 6 16 6 1 2 5 11 3 4
9.1 1.1 12.3 2.4 8.0 5.2 9.8 4.8 5.8 11.0 4.1 0 n.d. 3.0 8.0 2.1 3.3
nine expected for H3 peptides 91-135, 91-120, 121-135, intact histone H3, 1-120 and 1-90 respectively [22]. The results of electrophoresing the purified peptides and the total CNBr digest on urea/ acetic acid gels are shown in Fig.4. Densitometer measurements on these and duplicate gels showed H3 fragments 121-135, 1-320 and 1-90 to be more than 90 % homogeneous while peptides 91 - 120 and 91 - 135 were about 80 % homogeneous. The adjacent multiple bands in the 91-125 gel may be due to oxidized forms of the uncleaved methionine residue at position 120, resistant to CNBr cleavage.
1 1 1 2 2 0 4 0 1 5 2 0 2 2 4 1 1
0 0 2 1 1 2 0 0 0 2 1 0 0
Aggregation of Histone H3 and H3 Fragments
The tendencies of histone H3 and H3 peptides 1-120, 1-90, 91-135, 121-135 and 91-120 to associate in solutions of moderate ionic strength were examined by electrophoresis on neutral nondenaturing gels [ll]. The densitometer traces of the stained gels are shown in Fig. 5, along with the pattern obtained for the lyophilized digest. Histone H3 and peptide 1 - 120 (traces 2 and 3) form large aggregates and barely enter the gel. Peptides 1-90 and 121- 135 (traces 4 and 6) migrate as single bands with no protein
414
Conformational Studies on Histone H3 and Its CNBr Peptides
Fig. 4. Ureajacetic acidpolyacrylamide gels (30 %) of’ the histone H3 CNBr digest and purified peptides. Gel 1 H3, peptides 1 - 120, 1-90, 91 - 135, 91 120 and 121- 135 respectively
=
digest, gels 2 - 7
=
histone
~
Table 2. Aggregation properties oJ H3 fragments on non-denaturing phosphate gels Amounts in each band were determined by a dye elution method [I 11 Peptide
Aggregate band
Intermediate bands
Monomer band
% H3 1- 120 1-90 91 135 91 - 120 121 - 135 ~
-++ Fig.5. Electrophoresis of histone H3 and H3 CNBr peptides on neutralphosphate gels. Densitometer traces of stained gels (1) CNBr digest mixture, (2) histone H3, (3-7) peptides, 1-120, 1-90, 91 - 135,121 - 135 and 91 - 120 respectively. Migration was toward the cathode for all peptides except 91 - 120
98 97 5 18 5 0
0
0 0 79 93 0
2 3 95 3 2 100
remaining at the origin, while peptides 91 - 135 and 91-120 (traces 5 and 7) displayed more than one band. Addition of urea to the non-denaturing gels [12] reduced these multiple bands to one band in each case, indicating multimer formation in peptides 91-135 and 91-120. A dye elution method [ l l ] was used to quantify the distribution of peptide along the gel length in terms of aggregate, intermediate and monomer bands. The results of the analysis are shown in Table 2. Under the conditions of the experiment (0.05 M phosphate pH 7), we interpret the results to mean that histone H3 and peptide 1 - 120 aggregate extensively, peptides 91 - 120 and 91 - 135 form small multimers and peptides 1-90 and 121-135 do not self-interact. Interestingly, the total reaction mixture containing all the peptides and some uncleaved histone H3 appears to exhibit a gel pattern which is simply a linear combination of the gel patterns of the constituent peptides (see Fig. 5) suggesting that these peptides do not interact extensively with one another.
G. Morris and P. N. Lewis
475 0
:::!/f -75
-
-
4 1-120 L
I
I
I
X (nm)
91 -135 210
225 X(nm)
240
-21$q Fl
-5.0
121 -135
91 -120
-7.5
210
225
240
7.5
210
225
240
1(nm) (nm) Fig.6. Far-ultraviolet CD spectra ofhistone H3 and its fragments in solutions of various ionic strengths. Histone H3 in (1) water, (2) 0.01 M
NaC1, (3) 0.025 M NaC1, (4) 0.10 M NaCl measured directly, (5) 0.1 M NaCl after 24 h ; peptide 1-120 in (1) water, (2) 0.05 M NaCI, (3) 0.1 M NaC1, (4) 0.2 M NaCl measured directly, (5) 0.2 M after 24 h ; peptide 1-90 in (1) water, (2) 0.5 M NaC1, (3) 1.0 M NaCl directly and after 24 h ; peptide 91 - 135 in ( 1 ) water, (2) 0.5 M NaCl measured directly, (3) 0.5 M NaCl after 24 h; peptide 91 - 120 in (1) water, (2) 0.25 M NaCl measured directly, (3) 0.25 M NaCl after 24 h; peptide 121- 135 in 0- 1 M NaCl measured directly and after 24 h
Peptide 91 - 120 does not contribute to this pattern as it is negatively charged at pH 7 and therefore does not migrate into the phosphate gel, .as run. Circular dichroic studies on H3 peptide 1-90 reported in the next section indicate that secondary structures are induced at high salt concentrations ( 2 0.5 M NaC1). In order to determine whether or not these changes are accompanied by aggregation, a sedimentation velocity run of H3 peptide 1- 90 in 0.5 M NaCl was made. All of this peptide was observed to sediment as a single symmetrical peak of 17 S supporting the suggestion [16] that structural changes
in the histones (except H1) are usually accompanied by intermolecular association.
Salt-Induced CD Spectral Changes in the Far Ultraviolet (200 - 250 nm) Histone H3 and peptides 1 -90, 1 - 120, 91 - 120 and 91 - 135 undergo conformational changes on increasing the ionic strength of the solvent as the farultraviolet CD spectra in Fig.6 show. H3 peptide 121- 135 did not undergo any detectable changes even in 1 M NaC1, The changes observed for histone H3
416
Conformational Studies on Histone H3 and Its CNBr Peptides
Table 3. Secondary structures in histone H3 and H3 peptides from CD measurements The salt concentration is the lowest at pH 3.5 which gave the maximum change. The structure values given are in residues, the total change observed was the sum of the fast and slow changes Peptide
NaCl
I
I
I
I
I
I
I
I
fi structure
or-Helix fast
slow
fast
slow
15 19 15 1 0 0
4 1
9 10 15 0 0 0
10 3 0 0 2 0
M H3 1-120 1- 90 91-135 91-120 121-135
0 10
0.20 1.00 0.50 0.25 1.00
0
2 3 0
.
If
I I 250
H3
260 I
and peptides 1- 120, 91 - 120 and 91 - 135 occurred on two distinct time scales. Isenberg and co-workers [23 - 251 have studied this phenomenon for intact histone H3 as well as for the other histones at pH 7. It was found by these researchers that the magnitude of both the fast and changes depended upon the salt concentration, the protein concentration and the ambient temperature. Our results for H3 and the 1-120, 91-120 and 91-135 at pH 3.5 these dependencies. Measurements were also made at pH 7 {not shown). The only difference observed was that the amount of salt required to effect a given change was slightly less than at pH 3.5, this effect being also observed for histone H4 [16]. The spectra shown in Fig. 6 were analyzed for their secondary structure content by a constrained leastsquares analysis using the a-helix and p-structure curves from Chen et al. [26] and the 10mM HC1 randomcoil spectrum of H3 peptide 1-90 as reference spectra in the range 208 - 244 nm [27]. The results of this analysis for the maximal fast (< 30 min) and slow (> 24 h) salt-induced changes in terms of the number of residues involved in a-helixes and p structures are given in Table 3. Similar results were obtained by using the vector projection method of Baker and Isenberg [27]. The onset of a-helix formation in histone H3 and peptide 1- 120 involves both the fast and slow phases with the major portion ( w 15 residues) in the fast phase. Peptide 1-90, which does not undergo a slow change, attains the helix content of the fast step for the intact molecule, albeit at a much higher salt concentration. The increased number of helical residues in peptide 1- 120 is probably due to the fact (unpublished observation) that the slow change for this peptide is more rapid than that for histone H3 itself, thereby complicating the separation of the two changes. These results suggest that in the intact molecule the a-helix is located before residue 90 on the
250
260
270 280 Wavelength, k ( n m ) I I
270 280 Wavelength, X ( n m )
I
1
290
30(
I
290
300
Fig. 7 Near-ultraviolet CD spectra of histone H3 and peptide 1-90. Histone H3 in water and in 0 15 M NaCl measured directly and after 24 h, peptide 1- 90 in water and in 0.5 M NaCl measured directly and after 24 h
polypeptide chain and that the slow changes involve residues in the region of 91 - 120 as peptide 121- 135 undergoes no salt-induced changes. Only a very small amount of helix is induced in peptides 91 - 120 and 91-135 and this is associated mainly with a slow change. fl structures are associated with both the fast and slow step changes as can be seen from Table 3. The large amount of this conformation induced in peptide 1-90 during the fast step suggests that a major fi structure locus in histone H3 occurs before residue 90. Salt-Induced CD Spectral Changes in the Near- Ultraviolet (250 - 300 nm) Region
The aromatic CD spectra of histone H3 and peptide 1- 90 are shown in Fig. 7. The results shown are typical of those obtained from four separate experiments. The fast-step change spectrum for histone H3 and the spectrum for peptide 1- 90 have virtually the same shape except that the amplitude of the H3 spectrum is exactly two thirds that of the 1-90
G. Morris and P. N. Lewis
spectrum. As there are three tyrosine residues in histone H3 at positions 41, 54 and 99, the fast-step changes in histone H3 almost certainly involve residues 41 and 54. The spectrum resulting from the slow change, presumably due to tyrosine-99, is not greatly changed from that due to the fast change. While these results do not allow us to deduce the secondary structures in which the tyrosines residues are involved, the spectra in Fig.7 suggest that the fast change occurring in histone H3 and the change induced in peptide 1-90 at high salt concentrations involve the very same structures at least around chain sites 41 and 54.
DISCUSSION The data presented in the previous section indicate that the fast-step salt-induced conformational changes in histone H3 involve self-association, extensive helix and B structure formation located primarily before residue 90. The slow-step changes in this histone can be attributed to the section between 91 and 120 and are accompanied by aggregation and the formation of a small amount of helix and fi structures. The region beyond residue 120 in histone H3 appears to be in the random coil condition even under the highest of salt concentrations examined. Histone H3 has been previously studied by CD [23,28], nuclear magnetic resonance [29], electron paramagnetic resonance [30], difference spectroscopy [31] and chemical accessibility determinations [32]. The CD results obtained here for the intact histone are generally in agreement with those obtained by D’Anna and Isenberg [23] and Adler et al. [28]. However, unlike these researchers, we find no loss in helix content on standing. This may be due to the lower pH at which our measurements were made, thereby inhibiting sulfhydryl oxidation. The NMR studies by Bradbury and his colleagues [29] indicate that the interacting region in histone H3 is located between residues 42- 110 in good agreement with our peptide data. The difference spectroscopy results of Palau and Padros [31] support our data from peptide 1- 90 in that tyrosine-41 is involved in the structural changes. At the moment, we have no direct evidence for the exact location of the a-helix and B-structures in histone H3. However, theoretical studies predict that the most likely [33] location of a-helix in histone H3 is from residues 67-111. Since the tyrosine residues at 41 and 54 adopt a specific conformation as evinced by their CD spectra and are unlikely to be involved
477
in a helix [33], this region might be assigned to structures. REFERENCES
1 . Kornberg, R. D. & Thomas, J. 0 . (1974) Science (Wash. D.C.) 184, 865- 867. 2. Roark, D. E., Geoghegan, T. E. & Keller, G. (1974) Biochem. Biophys. Res. Commun. 59, 542- 547. 3. D’Anna, J. A. & Isenberg, I. (1974) Biochemistry, 13, 49934997. 4. D’Anna, J. A. & Isenberg, I. (1974) Biochem. Biophys. Res. Commun. 61,343 - 347. 5. Sperling, R. & Bustin, M. (1975) Biochemistry, 14, 3322-3331. 6. Rubin, R. L. & Moudrianakis, E. N. (1975) Biochemistry, 14, 1718-1726. 7. Weintraub, H., Palter, K. & Van Lente, F. (1975) Cell, 6, 85110. 8. Moss, T., Cary, P. D., Crane-Robinson, C. & Bradbury, E. M. (1976) Biochemistry, 15, 2261 -2267. 9. Sperling, R. & Bustin, M. (1976) Nucleic Acids Res. 3, 12631275. 10. Lewis, P. N. (1976) Biochem. Biophys. Res. Commun. 68, 329335. 11. Lewis, P. N. (1976) Can. J . Biochem. 54, 641 -649. 12. Lewis, P. N. (1976) Can. J . Biochem. 54, 963-970. 13. Bonner, W. M. & Pollard, H. B. (1975) Biochem. Biophys. Res. Commun. 64,282 - 288. 14. Burton, D. R., Hyde, J. E. & Walker, I. 0. (1975) FEBS Lett. 55,77 - 80. 15. Thomas, J. 0. & Kornberg, R. D. (1975) FEBS Lett. 58, 353358. 16. Lewis, P. N., Bradbury, E. M. & Crane-Robinson, C. (1975) Biochemistry, 14, 3391 - 3400. 17. Hooper, J. A. & Smith, E. L. (1973) J . Biol. Chem. 248, 32553260. 18. Panyim, S. & Chalkley, R. (1969) Arch. Biochem. Biophys. 130, 337- 346. 19. Brandt, W. F. &von Holt, C. (1974) Eur. J . Biochem. 46,407417. 20. Reisner, A. H., Nemes, P. & Bucholtz, C. (1975) Anal. Biochem. 64,509-516. 21. Itzhaki, R. & Gill, D. M. (1964) Anal. Biochem. 4,401 -410. 22. DeLange, R. J., Hooper, J. A. & Smith, E. L. (1972) Proc. Natl Acad. Sci. U.S.A. 69, 882- 884. 23. D’Anna, J. A. & Isenberg, I. (1974) Biochemistry, 13, 49874992. 24. Smerdon, M. J. & Isenberg, I. (1974) Biochemistry, 13, 40464049. 25. D’Anna, J. A. & Isenberg, I. (1974) Biochemistry, 13, 20932098. 26. Chen, Y. H., Yang, J. T. & Martinez, H. M. (1972) Biochemistry, 11,4120-4131. 27. Baker, C. C. & Isenberg, I. (1976) Biochemistry, 15, 629-634. 28. Adler, A. J., Moran, E. C. & Fasman, G. D. (1975) Biochemistry, 14,4179-4185. 29. Bradbury, E. M., Cary, P. D., Crane-Robinson, C. & Rattle, H. W. E. (1973) Ann. N.Y. Acad. Sci. 222, 266-289. 30. Palau, J. & Padros, E. (1972) FEBS Lett. 27, 157- 160. 31. Palau, J. & Padros, E. (1975) Eur. J . Biochem. 52, 555-560. 32. Palau, J. & Daban, J. R. (1974) Eur. J. Biochem. 49, 151- 156. 33. Lewis, P. N. & Bradbury, E. M. (1974) Biochim. Biophys. Acta, 336.153-164.
G . Morris and P. N. Lewis *, Department of Biochemistry, University of Toronto School of Medicine, Medical Sciences Building, Taddle Creek Road, Toronto, Ontario, Canada, M5S 1A8
*
To whom correspondence should be addressed.
p
