Properties of human heterochromatin protein 1 (HP1) paralogs. (a) Multiple sequence alignment of human HP1α, HP1β, and HP1γ. Red boxes with white letters show identical amino acids, while white boxes with red letters indicate amino acids with similar properties. The orange box highlights the unique block of four serine residues in HP1α NTE that can be constitutively phosphorylated in vivo. (b) Structures of the HP1α chromo (CD – PDB code 3fdt) and chromoshadow (CSD – PDB code 3i3c) domains. Positively and negatively charged residues are shown in blue and red licorice representations, respectively. The green dashed box marks the additional helix in the CSD compared to the CD. (c) HP1α CSD-CSD homodimer can dimerize via α-helix and β-sheet binding interfaces. (d) Amino acid composition of HP1 paralogs. Positive (Arg, Lys), Negative (Asp, Glu), Aromatic (His, Phe, Tyr, Trp), Aliphatic (Ala, Ile, Leu, Met, Val), Polar (Asn, Gln, Ser, Thr), and Other (Cys, Gly, Pro). (e, f) Differences in the net charge and charged amino acids content of each domain of the HP1paralogs. (g, h, i) Charge distribution in the sequences of the HP1 paralogs (blue = positively charged residues, red = negatively charged residues).

Sequence differences result in differences in conformation and phase separation of HP1 paralogs. (a) Rg distributions of HP1 paralogs in CG single homodimer simulations using the HPS-Urry model. (b) Average intramolecular contacts within one chain and average intermolecular contacts between two chains on each domain of the three HP1 paralogs. Two residues were considered to be in contact if the distance between them was less than 1.5 of their vdW arithmetic mean. (c, d, e) Pairwise 2D contact maps between two HP1 homodimers. The intramolecular interactions within one monomer and the intermolecular interactions between two monomers are shown in the two small triangles and the off-diagonal quadrant, respectively. The contact propensity of HP1β and HP1γ is normalized to the highest contact propensity of HP1α. The magenta boxes highlight strong electrostatic attractions between the hinge and CTE regions in HP1α. (f) Density profiles of HP1 paralogs in CG coexistence simulations. The inset shows the respective saturation concentrations. (g) Representative snapshots of each system in the CG coexistence simulations. (h) Phase diagrams of HP1 paralogs at different temperatures calculated in CG simulations where a dilute phase of free monomers co-exists with a dense phase. The critical temperatures of HP1α, HP1β, and HP1γ are 353.3 K, 276.1K, and 350.5 K, respectively. (i) Rg distributions of the HP1 paralogs as calculated in the CG phase coexistence simulations. (j) Average intermolecular contacts made by each domain of the HP1 paralogs in CG phase coexistence simulations. The error bars represent the standard deviation from triplicate simulation sets. (k, l, m) Intermolecular contacts within the condensed phase of HP1 paralogs along the sequence of each homodimer. The panels below each map show the average contacts per chain as a function of residue number. The contact propensity of HP1β and HP1γ is normalized to the highest contact propensity of HP1α. The CG coexistence simulations were conducted using the HPS-Urry model at 320K.

The interplay of IDRs and folded domains in mediating phase separation of HP1 paralogs. (a) Cartoons representing the three human HP1 paralogs. (b, c) Density profiles, saturation concentrations (inset), and average contacts made by each domain of HP1α with IDRs replaced with those from HP1β or HP1γ. (d, e, f) Density profiles, saturation concentrations, and average contacts made by each domain of HP1α chimeras whose hinge was swapped with the hinge from either HP1β or HP1γ (HP1α-βHinge and HP1α-γHinge, respectively), and HP1β and HP1γ chimeras whose hinge regions were replaced with the hinge of HP1α (HP1β-αHinge or HP1γ-αHinge, respectively). (g, h, i) Density profiles, saturation concentrations, and average contacts made by each domain of HP1α chimeras with folded domains replaced with those from HP1β or HP1γ (HP1α-βCDβCSD and HP1α-γCDγCSD, respectively). (j, k, l) Density profiles, saturation concentrations, and average contacts made by each domain of HP1α chimeras with either CD or CSD replaced with the corresponding domain from HP1β (HP1α-βCD and HP1α-βCSD, respectively) or HP1γ (HP1α-γCD and HP1α-γCSD, respectively). The cyan and orange dashed lines show the simulated saturation concentrations of wild-type HP1α and pHP1α, respectively. The error bars represent the standard deviation from triplicate simulation sets. The CG coexistence simulations were conducted using the HPS-Urry model at 320K.

Phase separation of HP1 heterodimers. (a, b, c) Density profiles, snapshots of the condensates, and saturation concentrations in the CG coexistence simulations performed with heterodimers HP1αβ, HP1αγ, and HP1βγ. (d) Phase diagram of HP1 heterodimer phase separation conducted at different temperatures. (e) Average intermolecular contacts for each domain of the HP1 heterodimers in CG co-existence simulations. The error bars represent the standard deviation from triplicate simulation sets. The CG coexistence simulations were conducted using the HPS-Urry model at 320K.

Phase separation of HP1 homo- and heterodimers with DNA. (a, b) Density profiles, saturation concentrations (inset), and snapshots of HP1 paralogs with one chain of 147 bp double-stranded DNA in the CG coexistence simulations. (c, d, e) Intermolecular contacts between HP1 homodimers and DNA. (f, g) Density profiles, saturation concentrations (inset), and snapshots of HP1 heterodimers with one chain of 147 bp double-stranded DNA in the CG coexistence simulations. (h, i, j) Intermolecular contacts between HP1 heterodimers and DNA. Preferential interactions between HP1 protein and DNA are shown in red. The error bars represent the standard deviation from triplicate simulation sets. The CG coexistence simulations were conducted using the HPS-Urry model at 320K.

DNA regulates the condensation of HP1 paralog mixtures. (a) Schematic of the ratios of HP1α and HP1β (or HP1γ) used to generate multicomponent phase diagrams. The total concentration of the mixture was fixed in all cases. (b, c) Multicomponent phase diagrams of HP1α mixing with HP1β (or HP1γ). The color code corresponds to the ratios shown in the schematic in (a). The circles and the squares show the protein concentrations in the dilute and the dense phases, respectively. The tie line (colored dashed line) connects the concentrations in the dense and dilute phases for each ratio. Phase existence lines (black dotted lines) define the two arms of the phase diagram at low and high concentrations. (d) Snapshots of the condensates in the CG coexistence simulations of HP1α+HP1β and HP1α+HP1γ mixtures at an equimolar concentration in the absence and presence of dsDNA. (e, f) The density profiles the mixtures at an equimolar concentration in the presence of dsDNA. (g, h) Multicomponent phase diagrams of HP1α mixing with HP1β or HP1γ in the presence of DNA. The 1:1 ratio data points for HP1α and HP1β mixing with DNA were excluded in (g) due to the instability of the condensate. The diamonds and the hexagons show the protein concentrations in the dilute and the dense phases, respectively. The error bars represent the standard deviation from triplicate simulation sets. The CG coexistence simulations were conducted using the HPS-Urry model at 320K.

Distinct interaction patterns of HP1 paralogs influence the localization patterns of HP1α, HP1β, and HP1γ in the absence and presence of DNA.