Macromolecular Crowding Induces a Molten Globule State in the C-Terminal Domain of Histone H1

Abstract

We studied the secondary structure of the C-terminal domains of the histone H1 subtypes H1° (C-H1°) and H1t (C-H1t) in the presence of macromolecular crowding agents (Ficoll 70 and PEG 6000) by IR spectroscopy. The carboxyl-terminal domain has little structure in aqueous solution but became extensively folded in the presence of crowding agents. In 30% PEG, C-H1° contained 19% α-helix, 28% β-sheet, 16% turns, and 31% open loops. Similar proportions were observed in 30% Ficoll 70 and for C-H1t in both crowding agents. The proportions of secondary structure motifs were comparable to those of the DNA-bound domain. Kratky plots of the small-angle x-ray scattering showed that in crowding agents the C-terminus had the compaction of a globular state. Progressive dissipation of the secondary structure and a linear increase in partial heat capacity with temperature together with increased binding of ANS indicated that the C-terminus is not cooperatively folded in crowded conditions. Native-like secondary structure and compactness in absence of folding cooperativity indicate that the C-terminus in crowding agents is in a molten globule state. Folding of the C-terminus in crowded conditions may increase the rate of the transition toward the DNA-bound state and facilitate H1 diffusion inside cell nuclei.

INTRODUCTION

H1 linker histones are involved in chromatin structure and gene regulation. It is currently accepted that histone H1 could have a regulatory role in transcription through the modulation of chromatin condensation. H1 may mediate transcription on a more specific level, participating in complexes that either activate or repress specific genes (1–7). Binding to scaffold-associated regions (8) and participation in nucleosome positioning (9) are other mechanisms by which H1 could contribute to transcriptional regulation.

H1 contains three distinct domains: a short amino-terminal domain (20–35 amino acids), a central globular domain (∼80 amino acids) consisting of a three-helix bundle and a β-hairpin, and a long carboxy-terminal domain (∼100 amino acids) (10). The amino acid composition of the C-terminus is dominated by Lys (∼40%), Ala (∼17%), and Pro (∼12%). The C-terminal domain is the primary determinant of H1 binding to chromatin in vivo (11,12). Several studies indicate that the ability of linker histones to stabilize chromatin folding resides in the C-terminal domain of the molecule (13,14). Preferential binding of histone H1 to scaffold-associated regions and activation of apoptotic nuclease also appear to be determined by the C-terminal domain (15,16).

In aqueous solution, the C-terminal domain is dominated by the random coil and turn-like conformations in rapid equilibrium with the unfolded state, but on interaction with DNA, it folds cooperatively (17). The DNA-bound structure is extremely stable and includes α-helix, β-sheet, turns, and open loops. The H1 carboxyl-terminal domain thus appears to belong to the so-called intrinsically disordered proteins undergoing coupled binding and folding (17–22). Like most natively disordered proteins, the C-terminus is of low hydrophobicity, large net charge, and low sequence complexity (23–25).

Proteins are usually studied in dilute solution; however, in the cellular environment, macromolecules and small-molecule solutes are present at high concentrations so that a significant fraction of the intracellular space is not available to other macromolecules. The excluded volume effects are predicted to favor the adoption of compact as opposed to expanded macromolecular conformations, resulting in a reduction of the total excluded volume. However, experimental evidence has shown that although crowding agents in general favor refolding (26–29), in some cases they appear to be ineffective (30,31). A role of molecular crowding in accelerating α-synuclein fibrillation has been described (32,33).

In this study we have used IR spectroscopy to estimate the proportions of secondary structure motifs of the C-terminus of H1° (C-H1°) and H1t (C-H1t) in the presence of Ficoll 70 and polyethyleneglycol (PEG) 6000 as macromolecular crowding agents. Our results show that crowding is highly effective in promoting folding of the C-terminal domain. In the presence of 30% PEG or Ficoll, the proportions of secondary structure motifs appear to be similar to those of the DNA-bound domain. In these conditions, Kratky plots of the x-ray scattering indicate that the compactness of the C-terminus is that of a globular state. However, binding of 1-anilinonaphthalene-8-sulfonate (ANS), thermal melting, and differential scanning calorimetry (DSC) show that in a crowded environment the C-terminus is not cooperatively folded. The properties of the C-terminus in crowding agents are thus those of a molten globule state.

MATERIALS AND METHODS

Cloning, expression, and purification of the C-terminal domains of histone H1 subtypes

The C-terminal domains of histones H1° and H1t were obtained from recombinant Escherichia coli (M15) as described previously (17).

Circular dichroism spectroscopy

The C-terminal domains were at 0.3 mg/ml in 10 mM phosphate buffer, pH 7.0, plus 140 mM NaCl and in the presence of PEG 6000 or Ficoll 70. Spectra were obtained on a Jasco J-717 spectrometer in 1-mm cells at 20°C. The results were analyzed with Standard analysis software (JACSO) and expressed as mean residue molar ellipticity [θ]. The helical content was estimated from the ellipticity value at 222 nm (θ₂₂₂), according to the empirical equation of Chen et al. (34): percentage helical content = 100[θ₂₂₂/−39,500 × (1 − 2.57/n)], where n is the number of peptide bonds.

Infrared spectroscopy measurements

The C-terminal domains were at 5 mg/ml in 10 mM HEPES, pH 7.0, plus 140 mM NaCl and in the presence of PEG 6000 or Ficoll 70 at 300 g/L. Data were collected on an FT600 Bio-Rad spectrometer equipped with an MCT detector. Typically, 1000 scans for each background and sample were collected, and the spectra were obtained with a nominal resolution of 2 cm⁻¹ at 20°C. Secondary structure content was determined by curve fitting to the original spectrum using the component band positions identified by Fourier self-deconvolution as previously described (35,36). The baseline contributed by the solvent (10 mM HEPES, 140 mM NaCl, pH 7.0 in H₂O or D₂O and with or without PEG 6000 or Ficoll 70) was removed before fitting. Spectra were recorded both in H₂O and D₂O to distinguish between the overlapping contributions in D₂O of random coil and open loops. The vibrations of loops and the α-helix are found at similar positions in H₂O and D₂O, the α-helix at ∼1652 cm⁻¹ and loops at ∼1643 cm⁻¹. In contrast, deuteration has a major effect on the position of the random coil so that in H₂O it overlaps with the α-helix, whereas in D₂O it overlaps with loops. The α-helix was, therefore, estimated directly in D₂O, whereas loops were estimated directly in H₂O. When both α-helix and loops are present in the protein structure, the percentage of random coil can be obtained either from the difference of the components around 1552 cm⁻¹ in H₂O and D₂O or from the difference of the components around 1643 cm⁻¹ in D₂O and H₂O, as previously described (17).

Small-angle x-ray scattering

Measurements were performed in an MBraun instrument equipped with a Siemens Kristalloflex (Graz, Austria) 760 (K-760) generator, producing radiation with a wavelength of 1.54 Å (CuK_α), operating at 50 kV and 40 mA, and a Kratky Hecus camera. The collimator was a slit window, and the scattering was detected with a linear position sensitive detector OED-50M. The sample-to-detector distance was 268 mm. Samples were measured at 25°C. All samples were placed in glass capillaries of 1 mm diameter and 10 μm wall thickness. The background was subtracted by measuring blanks without protein. Both the samples and blanks were measured for 4 h unless otherwise stated. The concentration of protein was 4–10 mg/ml.

The scattering intensity I(Q) is represented as a function of the momentum transfer Q; Q = 4πsinθ/λ, where λ is the wavelength of x-rays and 2θ the scattering angle. The inner part of the scattering profile can be described by the Guinier approximation, (37). The scattering at zero angle, I₀, is proportional to the molecular mass of the particle and to the square of the contrast between the particle and the solvent. The radius of gyration, R_g, which is the root mean-square of the distances of all volume elements to the center of mass of the electronic volume of the particle, was obtained from the slope of a Guinier plot of lnI(Q) versus Q².

The R_g can also be measured with the Debye equation (38), where using the approximation given by Calmettes et al. (39): [P_D(x)]⁻¹ = 1 + 0.359 x^1.103. The R_g is given by R_g = (a/0.359b)^0.453, where a is the slope and b the intercept of the straight line.

Thermal melting

For the analysis of the thermal stability of C-H1t in 30% Ficoll 70, the temperature of the sample was adjusted using a cover jack connected to a circulating thermostatic bath and monitored with a fitted external probe. Thermal analyses were performed by heating from 20 to 85°C at a rate of 1°C/min. For each degree of temperature interval, 400 interferograms were averaged, Fourier-transformed, and ratioed against background. Thermal analyses were performed in D₂O. The Ficoll contribution at each temperature was subtracted using a Ficoll sample at the same temperature and concentration.

ANS fluorescence assays

1-Anilinonaphthalene-8-sulfonate (ANS) binding assays were performed in a Cary Eclipse spectrofluorometer. The spectra were measured in 140 mM NaCl, 10 mM phosphate buffer, pH 7.0, at 20°C. The concentration of ANS was 0.3 mM, and the concentration of protein was 10 mM. The ANS emission was scanned between 400 and 650 nm with an excitation wavelength of 380 nm.

Differential scanning calorimetry

DSC measurements were carried out on a Microcal MC-2 calorimeter supplied with an Origin software package for data analysis and curve fitting. Protein and background scans were performed at a scan rate of 1.5°C/min between 20 and 100°C. We used RNase A as a control at a concentration of 150 μM. C-H1t was analyzed at a concentration of 100 μM in buffer and in buffer plus 30% Ficoll 70. The buffer was 10 mM phosphate, pH 7.0, plus 140 mM NaCl.

RESULTS

CD and IR spectroscopy analysis of the C-terminal domain of histone H1 in the presence of crowding agents

We used CD to explore the effects of different concentrations of Ficoll 70 and PEG 6000 (10, 20, and 30%) on the secondary structure of the C-terminal domains of the histone H1 subtypes H1° and H1t in physiological salt (0.14 M NaCl) (Fig. 1). We found that in 30% of both crowding agents C-H1° and C-H1t became significantly folded. In 30% PEG, the estimated values of α-helix were 16% for C-H1° and 13% for C-H1t. In 30% Ficoll, the estimated helical content was 16% for both C-terminal domains.

FIGURE 1.

Conformational transition of the C-terminal domain of histone H1 measured by CD. C-H1° in Ficoll (A) and PEG (C). C-H1t in Ficoll (B) and PEG (D). The buffer was 10 mM phosphate, pH 7.0, plus 140 mM NaCl. (Solid line) C-terminal domains in buffer; (dotted line) buffer plus 10% PEG or Ficoll; (dashed-dotted line) buffer plus 20% PEG or Ficoll; and (dashed line) buffer plus 30% PEG or Ficoll.

IR spectroscopy was used to further characterize the secondary structure of the C-terminal domains in crowding agents. In the absence of crowding agents, the amide I (D₂O) of both C-H1° and C-H1t was dominated by the unordered structure (at 1642 cm⁻¹ in D₂O) with ∼50% of the total intensity. The other main component (at 1663 cm⁻¹), with 28–31% of the total area in both C-H1° and C-H1t, was assigned to turns in rapid equilibrium with the unfolded state. Well-defined structure was represented by 9–12% of β-sheet (at 1626–1630 cm⁻¹) (Fig. 2 and Table 1).

FIGURE 2.

Amide I decomposition of the spectra of the C-terminal domain of histone H1. C-H1°, carboxy-terminal domain of histone H1°. C-H1t, carboxy-terminal domain of histone H1t. The spectra were recorded in D₂O at 20°C. The buffer was 10 mM HEPES plus 140 mM NaCl, pH 7.0. The protein concentration was 5 mg/ml.

TABLE 1.

Percentages (%) of secondary structure of the carboxy-terminal domain of histones H1° (C-H1°) and H1t (C-H1t) in presence of crowding agents

	Buffer*		Ficoll†				PEG†
	D2O		D2O		H2O		D2O		H2O
	Band (cm−1)	%	Band (cm−1)	%	Band (cm−1)	%	Band (cm−1)	%	Band (cm−1)	%
Assignment C-H1°
Turns	1676	3	1685	1	1685	3	1680	1	1680	7
Turns	1661	31	1667	26	1668	21	1668	15	1668	10
α-Helix			1655	15			1656	19
α-Helix/random coil					1655	15			1656	19
Open loops					1643	23			1642	31
Random coil/open loops	1642	52	1664	23			1645	31
Random coil‡				0				0
Random coil¶						0				0
β-Sheet	1630	9	1635	14	1634	14	1635	18	1633	12
β-Sheet			1626	13	1625	13	1626	10	1623	11
Low-frequency β-sheet	1617	5	1616	8	1616	11	1616	6	1615	10
C-H1t
Turns	1685	2	1684	1	1681	8	1681	1	1681	8
Turns	1674	4	1668	29	1668	14	1666	23	1666	16
Turns	1663	28
α-Helix			1656	15			1655	14
α-Helix/random coil					1656	15			1656	14
Open loops					1645	23			1643	27
Random coil/open loops	1643	53	1645	23			1645	27
Random coil‡				0				0
Random coil¶						0				0
β-Sheet			1637	16	1635	15	1634	17	1634	13
β-Sheet	1626	12	1628	11	1625	16	1625	13	1624	11
Low-frequency β-sheet	1616	1	1616	5	1615	9	1615	5	1615	11

^*Buffer: 10 mM HEPES, 140 mM NaCl, pH 7.0.

^†The concentrations of Ficoll 70 and PEG 6000 were 30%.

^‡The value corresponds to the difference between the components at ∼1643 in D₂O and H₂O.

^¶The value corresponds to the difference between the components at ∼1655 in H₂O and D₂O.

Addition of either PEG or Ficoll forced the C-terminus to fold. In 30% PEG 6000, deconvolution of the amide I of C-H1° gave 19% α-helix (1656 cm⁻¹), 28% β-sheet (1635, 1626 cm⁻¹), 16% turns (1680, 1668 cm⁻¹), and 31% open loops (1642 cm⁻¹) (Fig. 3 and Table 1). Spectra were recorded both in H₂O and D₂O to distinguish between the overlapping contributions in D₂O of random coil and open loops as described under Materials and Methods. The proportions of the different secondary structure motifs in C-H1t in the presence of PEG were similar to those found in C-H1°, although slight differences were present: a lower proportion of α-helix (14%) and open loops (27%) and a higher proportion of turns (24%); the proportion of β-sheet was very similar (30%) (Figs. 3 and 4). The amount of random coil was <1% in both C-H1° and C-H1t. The results in Ficoll 70 were similar to those found in PEG (Figs. 3 and 4, Table 1).

FIGURE 3.

Amide I decomposition of the spectra of C-H1° in the presence of macromolecular crowding agents. The left column shows the spectra measured in D₂O, and the right column shows those measured in H₂O. The buffer was 10 mM HEPES plus 140 mM NaCl, pH 7.0, at 20°C. The protein concentration was 5 mg/ml. The concentration of Ficoll 70 and PEG 6000 was 30%.

FIGURE 4.

Amide I decomposition of the spectra of C-H1t in the presence of macromolecular crowding agents. The left column shows the spectra measured in D₂O, and the right column shows those measured in H₂O. The buffer was 10 mM HEPES plus 140 mM NaCl, pH 7.0, at 20°C. The protein concentration was 5 mg/ml. The concentration of Ficoll 70 and PEG 6000 was 30%.

Small-angle x-ray scattering

The conformation of the C-terminus was examined by small-angle x-ray scattering (SAXS). The changes in compaction can be qualitatively analyzed by means of Kratky plots (Q²I(Q) versus Q). The scattering curve in Kratky plots has a characteristic bell shape when the protein is in the native state or in the molten globule state. Such a maximum is absent in completely unfolded proteins or in premolten globule states (20,40,41). In 30% Ficoll, the C-terminal domain of H1t showed the characteristic maximum of a well-developed globular structure (Fig. 5 A). The position of the maximum (0.1 Å⁻¹) is that expected for the molecular mass of the C-terminus (M_r = 11,082) (41). The Guinier plot gave an R_g of 14.8 ± 0.7 Å (Fig. 5 B). In the absence of crowding agents, the Kratky plot lacked the maximum of globular conformations and exhibited the monotonic increase indicative of an expanded coil-like conformation (Fig. 5 A). The R_g in buffer solution was estimated with the Debye equation, which is more appropriate than the Guinier approximation in the case of unfolded proteins (39,42). In these conditions, the R_g was 25.0 ± 0.2 Å (Fig. 5 C). Although this value is considerably higher than the R_g in crowded conditions, it is smaller than the value of ∼31.5 Å that can be estimated for random coils with the expression: where N is the number of amino acid residues (24,43). This indicates that the C-terminus in crowded conditions is more compact than an unperturbed random coil polypeptide chain. This is consistent with the IR results, showing the presence of a residual proportion of turns and β-sheet in buffer solution. Turns may not contribute to a tightly folded core but may reduce the dimensions of the conformational ensemble.

FIGURE 5.

Small-angle x-ray scattering of the C-terminal domain of histone H1t. (A) Kratky plots (Q²I(Q) as a function of Q) in 30% Ficoll (1) and in buffer solution (2). (B) Guinier plot (ln I(Q) as a function of Q²) of the protein in 30% Ficoll. (C) Representation of the x-ray scattering of the protein in buffer solution using the approximation to the Debye function (I(Q)⁻¹ as a function of Q^2.206); Q is in Å⁻¹. The Q ranges used for the fits were 0.035 < Q < 0.084 for the protein in Ficoll and 0.0074 < Q < 0.0385 for the protein in buffer solution.

ANS binding

The C-terminal domains were tested for 1-anilinonaphthalene-8-sulfonate (ANS) binding in either the absence or presence of PEG and Ficoll (Fig. 6). This hydrophobic dye has been used as a probe for the formation of collapsed, partially folded states such as the premolten and molten globules (44,45). In the absence of crowding agents, the emission of ANS was not affected by the presence of the C-terminus. In contrast, in Ficoll, the λ_max of fluorescence of protein-bound ANS shifted from 525 to 485 nm, and the fluorescence intensity increased ∼11-fold. In PEG, a shift from 525 to 501 nm and an ∼7-fold increase in fluorescence intensity were observed. (Fig. 6, A and B).

FIGURE 6.

Effects of macromolecular crowding on the C-terminal domain of histone H1. (A) Fluorescence emission spectra of 0.3 mM ANS in the presence of C-H1° without crowding agent (solid line) and with 30% PEG 6000 (dashed line) or 30% Ficoll 70 (dotted line). (B) Fluorescence emission spectra of 0.3 mM ANS in the presence of C-H1t without crowding agent (continuous line) and with 30% PEG 6000 (dashed line) or 30% Ficoll 70 (dotted line). (C) Thermal profiles of the secondary structure motifs obtained from amide I (D₂O) decomposition in 140 mM NaCl, 30% Ficoll. (D) Temperature dependence of the partial heat capacity (C_p) in the absence (dashed line) and in the presence of 30% Ficoll 70 (continuous line). The DSC curve of ribonuclease A is shown for comparison (dotted line). Binding, thermal denaturation, and DSC experiments were performed in 10 mM phosphate buffer, pH 7.0, plus 140 mM NaCl.

Thermal melting and DSC

To better understand the features of the folded states of the C-terminus in crowding agents, we analyzed the thermal melting by IR spectroscopy and the change in partial heat capacity (C_p) with temperature by DSC of C-H1t in 30% Ficoll. The changes in the amide I (D₂O) components were monitored in the 20-85°C interval. The α-helix and the β-sheet were lost fairly linearly as temperature was increased; all defined structure was lost by 85°C. Concomitantly with the decrease in α-helix and β-sheet, the random coil component increased from 23% to 56% (Fig. 6 C).

In parallel with the dissipation of secondary structure, the C_p increased almost linearly between 25°C and 85°C, where it leveled off coinciding with the complete dissipation of the α-helix and β-sheet (Fig. 6 D). Cooperatively folded proteins have a significant enthalpy difference between their native and unfolded states, which leads to large heat capacity peaks corresponding to heat absorption on denaturation (46). Thermal melting and C_p measurements indicate that, despite globular compaction and native-like secondary structure, the C-terminus is not cooperatively folded and, instead, has features characteristic of the molten globule state (47–50), which possesses an extensive secondary structure but does not have a fixed tertiary structure.

DISCUSSION

The C-terminal domain of histone H1 is mostly unstructured in dilute solution but folds extensively on interaction with DNA. The C-terminal domain of histone H1 can thus be considered an intrinsically disordered protein (18,19). Some intrinsically disordered proteins may adopt compact conformations in crowded environments (31). The C-terminal domain appears to belong to this category of intrinsically disordered proteins. In the presence of Ficoll 70 or PEG 6000, the C-terminal domain becomes extensively structured with 14–19% α-helix, 22–31% β-sheet, 16–26% turns, and 20–30% loops, depending on the subtype and crowding agent. Folding is associated with the differentiation of a limited hydrophobic core, as shown by the increased ANS binding by the C-terminus in crowded conditions. Interestingly, the high positive charge of the C-terminus (∼40% Lys) does not impair folding in crowded conditions.

The proportions of the secondary structure motifs of the C-terminus in crowded conditions are comparable to those of the DNA-bound domain (17), although the α-helix was proportionally more abundant in the DNA complexes, and the β-sheet more abundant in crowded conditions (Table 2).

TABLE 2.

Percentages of secondary structure of the C-terminal domain in crowded conditions and bound to DNA

	Complexes with DNA*		Ficoll		PEG
	C-H1°	C-H1t	C-H1°	C-H1t	C-H1°	C-H1t
Turns/310 helix	33–34	40–42	24–27	22–30	16–17	24
α-Helix	24	21	15	15	19	14
Open loops	18	15	23	23	31	27
β-Sheet	16	13	27	27–31	22–28	24–30
Random coil	<1	<1	<1	<1	<1	<1

^*These data were taken from Roque et al. (17).

Despite the low sequence identity between C-H1° and C-H1t (∼30%), the proportions of their secondary structure motifs appear to be similar, as also occurs in the DNA-bound domains (17). This suggests that selective constraints may act to preserve the secondary structure of the C-terminus in the different subtypes.

The compactness of the C-terminus in the presence and in the absence of crowding agents was examined using SAXS. Global shape changes were analyzed by means of Kratky plots (Q²I(Q) versus Q). The Kratky representation has a characteristic maximum in globular states (native and molten globule). This maximum is absent in nonglobular states (random coil, premolten globule) (20). In the presence of crowding agents, the Kratky plot of the x-ray scattering by the C-terminus of H1t in 30% Ficoll exhibited the bell-shaped appearance typical of globular states. In the absence of crowding agents, the Kratky plot increased monotonously as observed for many unfolded proteins. The decrease of the R_g from 25.0 Å in buffer solution to 14.8 Å in crowding agents confirms the condensation of the C-terminus in crowded conditions.

Analysis of the folded state of the C-terminus in crowding agents by thermal melting and DSC reveals the progressive dissipation of secondary structure in the temperature interval 20–85°C, which is paralleled by a linear increase in partial heat capacity, lacking the heat absorption peak typical of cooperatively folded proteins. These features indicate that the C-terminus in crowded conditions is not cooperatively folded despite native-like secondary structure and compaction and has, instead, the characteristics of the molten globule state (47,51). We have previously shown that on interaction with DNA, the secondary structure of the protein becomes extremely stable and melts cooperatively (17). The transition from the molten-globule state to the DNA-bound cooperatively folded state is probably driven by increased hydrophobicity after charge compensation of the positive charge of the abundant Lys residues of the C-terminus by the DNA phosphates.

The H1 globular domain is thought to be responsible for the location of H1 by interacting with a specific site in the nucleosome. The dependence of the tightening of the C-terminus molten globule structure on charge compensation may afford a mechanism to avoid the interference of the strong electrostatic binding of the C-terminus with binding site recognition by the globular domain because binding of the C-terminus to unspecific sites, such as nucleosomal DNA, already partially charge-compensated by core histones, may lead to partial folding and lower affinity.

The similarity of the proportions of secondary structure motifs of the domain in crowded conditions and bound to DNA suggests that in the presence of crowding agents the structure approximates that of the DNA-bound domain. An advantage related to chromatin dynamics of the condensation of the chain into a native-like structure could be an increased rate of the transition toward the native DNA-bound state (51). The folding of the C-terminus in crowded conditions might also facilitate diffusion of H1 inside cell nuclei. This effect would compensate for the slowing down of protein diffusion by crowding (52,53). Improved diffusion could have functional consequences because H1 molcules exchange in vivo between chromatin binding sites through a soluble intermediate (54).

In summary, the molten globule state of the C-terminus and the tightening of the structure on interaction with DNA could kinetically favor H1 binding and exchange.