a

Sir3/N:

+
U

++
B

U

B

WT

αSir3

H4K16Q

αSir3

Supplementary Figure 1. Sir3 specifically binds to WT over H4-H16Q arrays in phosphate buffer
containing ~40 mM Na+. (a) Nucleosomal array capture assay as in Fig. 1a-b showing a titration of Sir3 on
WT and H4-K16Q arrays in 20 mM phosphate buffer.

Sedimentation coefficient (S)

Frictional ratio (f/f0)
Frictional ratio (f/f0)
Molecular weight (Da)

Sedimentation coefficient (S)

d

Sedimentation coefficient (S)

Frictional ratio (f/f0)

Partial Concentration

Frictional ratio (f/f0)

Saturated nucleosomal array

Partial Concentration

Boundary fraction (%)

Sedimentation coefficient (S)

Partial Concentration

Frictional ratio (f/f0)

Unsaturated nucleosomal array

Sedimentation coefficient (S)

c

Molecular weight (Da)

Sedimentation coefficient (S)

Partial Concentration

Boundary fraction (%)

b

Partial Concentration

Frictional ratio (f/f0)

Free DNA template

Partial Concentration

Boundary fraction (%)

a

Molecular weight (Da)

Mol. Weight (MDa)

Sed. Coefficient (s -13)

Frictional Ratio

DNA template

1.335 [1.313] (1.333, 1.347)

10.654 (10.648, 10.668)

7.499 (7.486, 7.540)

Unsaturated array

1.167 [1.964] (1.140, 1.195)

23.341 (23.324, 23.358)

2.587 (2.546, 2.627)

Saturated array

2.296 [2.614] (2.272, 2.321)

37.594 (37.561, 37.627)

2.269 (2.252, 2.285)

Supplementary Figure 2. 2DSA/GA-MC modeling using predicted partial specific volumes does not
accurately determine the molecular weights of complex chromatin macromolecules. (a-c) vHW and
2DSA/GA-MC plots of free 601-177-12 template DNA, an approximately half-saturated 12mer nucleosomal
array, and a saturated 12mer nucleosomal array, respectively. (d) The molecular weight, sedimentation
coefficient, and frictional ratio of samples in (a-c) as determined by 2DSA/GA-MC. Numbers in parentheses
are 95% confidence intervals. Numbers in brackets represent theoretical molecular weights calculated by a
sequence-based algorithm implemented in UltraScan3.

a

b

Saturated Array

Saturated Array + Sir3

Supplementary Figure 3. 2DSA fitting is appropriate for chromatin samples. (a-b) 2DSA experimental
vs. model scans (top, model is in red) and residuals (bottom) demonstrate good fits and random residuals for a
WT array unbound (a) and bound by Sir3 (b).

Boundary fraction (%)

b

601-177-1 DNA

Sedimentation coefficient (S)

Boundary fraction (%)

c

601-177-12 DNA

Sedimentation coefficient (S)

Sedimentation coefficient (s)

Sedimentation coefficient (s)

Sedimentation coefficient (s)

Lysozyme

Sedimentation coefficient (s)

Boundary fraction (%)

a

R 2 = 0.991
v = 0.726 mL/g
[0.724]

Density (g/mL)

R 2 = 0.996

v = 0.548 mL/g
[0.55]

Density (g/mL)

R 2 = 0.978

v = 0.568 mL/g
[0.55]

Density (g/mL)

Supplementary Figure 4. The partial specific volume of molecules can be determined via
sedimentation in solvents of known density. (a-c) vHW plots showing the sedimentation of molecules in
0% (light gray), 30% (dark gray), and 60% H2O18 (black) and plots of sedimentation coefficient vs. density for
(a) lysozyme, (b) 601-177-1 template DNA, and (c) 601-177-12 template DNA. The v is calculated by dividing
the slope of the fit line by the y-intercept. Numbers in brackets represent the v of the respective molecule as
predicted by UltraScan3.

Boundary fraction (%)

d

Boundary fraction (%)

e

Boundary fraction (%)

f

Sed. coefficient (S)

R 2 = 0.963
v = 0.654 mL/g

Density (g/mL)

Sed. coefficient (S)

16.7 S

Sedimentation coefficient (S)

R 2 = 1.000

v = 0.658 mL/g

Density (g/mL)

Sed. coefficient (S)

22.4 S

Sedimentation coefficient (S)

R 2 = 0.956

v = 0.669 mL/g

Density (g/mL)

22.8 S
Sed. coefficient (S)

Boundary fraction (%)

c

Sedimentation coefficient (S)

Sedimentation coefficient (S)

R 2 = 0.967

v = 0.677 mL/g

Density (g/mL)

26.7 S
Sed. coefficient (S)

Boundary fraction (%)

b

13.5 S

Sedimentation coefficient (S)

R 2 = 1.000

v = 0.671 mL/g

Density (g/mL)

35.0 S
Sed. coefficient (S)

Boundary fraction (%)

a

Sedimentation coefficient (S)

R 2 = 0.974

v = 0.695 mL/g

Density (g/mL)

Supplementary Figure 5. The partial specific volume of nucleosomal arrays increases with histone
octamer saturation and S. (a-f) Determination of the v of arrays in Fig. 2 as in Supplementary Fig. 4.

Boundary fraction (%)

v = 0.669 mL/g

Density (g/mL)

Boundary fraction (%)

Sed. coefficient (S)

R 2 = 0.956

Density corrected

Sedimentation coefficient (S)

Sed. coefficient (S)

Viscosity corrected

v = 0.667 mL/g

Density (g/mL)

Sed. coefficient (S)

22.8 S

Sedimentation coefficient (S)

R 2 = 0.967

v = 0.677 mL/g

Density (g/mL)

Boundary fraction (%)

Sedimentation coefficient (S)

R 2 = 0.935

Density corrected

Sedimentation coefficient (S)

Sed. coefficient (S)

Viscosity corrected

Sedimentation coefficient (S)

R 2 = 0.997

v = 0.677 mL/g

Density (g/mL)

35.0 S

Sedimentation coefficient (S)

R 2 = 0.974

v = 0.695 mL/g

Density (g/mL)

Boundary fraction (%)

Boundary fraction (%)

c

Sedimentation coefficient (S)

Sed. coefficient (S)

Boundary fraction (%)

Boundary fraction (%)

b

22.4 S

Density corrected

Sedimentation coefficient (S)

Viscosity corrected
Sed. coefficient (S)

Boundary fraction (%)

Boundary fraction (%)

a

Sedimentation coefficient (S)

R 2 = 0.954

v = 0.692 mL/g

Density (g/mL)

Supplementary Figure 6. The partial specific volume of nucleosomal arrays is independent of
viscosity, and the sedimentation distribution of chromatin samples is highly reproducible. (a-c) The v
determination of array samples in Supplementary Fig. 5c,d,f shown as used in Fig. 2 (top panels) and
corrected for viscosity (bottom panels). The vHW distributions corrected for density are in the top right panels.

H4K16Q + Sir3

Sedimentation coefficient (S)
Boundary fraction (%)

c

Tris

Sedimentation coefficient (S)

Boundary fraction (%)

d

Tris + MgCl 2

Sedimentation coefficient (S)

Boundary fraction (%)

e

Phosphate + MgCl 2

Sedimentation coefficient (S)

Sedimentation coefficient (s)
Sedimentation coefficient (s)

Boundary fraction (%)

b

Sedimentation coefficient (s)

Sedimentation coefficient (S)

Sedimentation coefficient (s)

WT + Sir3

Sedimentation coefficient (s)

Boundary fraction (%)

a

R 2 = 0.997
v = 0.726 mL/g

Density (g/mL)

R 2 = 0.999

v = 0.686 mL/g

Density (g/mL)

R 2 = 0.986

v = 0.763 mL/g

Density (g/mL)

R 2 = 0.961

v = 0.574 mL/g

Density (g/mL)

R 2 = 0.935

v = 0.595 mL/g

Density (g/mL)

Supplementary Figure 7. The partial specific volume of arrays decreases during Mg++-induced folding
but increases upon Sir3 binding. (a-b) Example v determinations of WT and H4K16Q arrays with Sir3.
Average v’s from three experiments were used for 2DSA/GA-MC in Fig. 3. (c-d) Example v determinations of
extended and folded arrays in Tris. Average v’s from three experiments were used for 2DSA/GA-MC in Fig. 4.
(e) Example v determinations of folded arrays in phosphate buffer.

Frictional ratio (f/f0)

Frictional ratio (f/f0)

b

Partial Concentration

Sedimentation coefficient (S)

Sedimentation coefficient (S)

Partial Concentration

Boundary fraction (%)

a

Molecular weight (Da)

Relative Percent

Sed. Coefficient (S)

Mol. Weight (kDa)

Frictional Ratio

Monomer

68.97 %

4.824 (4.815, 4.832)

113.43 (112.48, 114.38)

1.701 (1.693, 1.708)

Dimer

31.03 %

7.819 (7.806, 7.832)

260.53 (258.73, 262.33)

1.826 (1.819, 1.834)

Supplementary Figure 8. Sir3 exists as a mixture of monomers and dimers in solution. (a) Left panel,
vHW analysis of Sir3 at 171 nM (corresponding to the concentration used for 2 monomers of Sir3 per
nucleosome in Fig. 1d-e and 3) in phosphate buffer. Middle and right panels, GA-MC plots of S vs. molecular
weight and f/f0 vs. molecular weight. (b) 2DSA/GA-MC statistics show 69% of Sir3 in solution is a monomer
(113 kDa), and 31% exists as an oligomer with a molecular weight most closely corresponding to a dimer
(theoretical molecular weight is 226 kDa).

Boundary fraction (%)

a
40 mM Na+
150 mM Na+

Sedimentation coefficient (S)
WT

Height = 1.60 (1.56, 1.64)

H4K16Q

Height = 1.62 (1.58, 1.65)

Count

Count

b

0.2 μm

0.2 μm

Height (nm)

Height (nm)
WT + Sir3

H4K16Q + Sir3

Height = 2.09 (1.86, 2.31)

Count

Height = 3.29 (3.03, 3.55)

Count

c

0.2 μm

0.2 μm

Height (nm)

Height (nm)

Supplementary Figure 9. Sir3-array structure in 150 mM Na+ closely resembles Sir3-array structure in
40 mM Na+. (a) vHW analysis of 171 nM Sir3 in phosphate buffer containing ~40 mM Na+ or in phosphate
buffer brought to 150 mM Na+. (b) WT and H4-K16Q arrays in phosphate buffer brought to 150 mM Na+ are
equivalent in structure and height to arrays in phosphate buffer alone (compare to Fig. 5b). (c) WT and
H4-K16Q arrays in phosphate buffer brought to 150 mM Na+ are similar in structure and height to arrays in
phosphate buffer alone in the presence of 2 Sir3 monomers/nucleosome (compare to Fig. 5c).

Monomer

Sedimentation coefficient (S)

Frictional ratio (f/f0)

Frictional ratio (f/f0)

b

Partial Concentration

Sedimentation coefficient (S)

Partial Concentration

Boundary fraction (%)

a

Molecular weight (Da)

Sed. Coefficient (S)

Mol. Weight (kDa)

Frictional Ratio

2.224 (2.222, 2.226)

29.119 [26.336] (28.965, 29.274)

1.498 (1.492, 1.503)

Supplementary Figure 10. Sir3 BAH exists as a monomer in solution. (a) Left panel, vHW analysis Sir3
BAH at 1.71 μM in phosphate buffer. Middle and right panels, GA-MC plots of S vs. molecular weight and f/f0
vs. molecular weight. (b) 2DSA/GA-MC statistics show 100% of Sir3 BAH in solution is a monomer. Number
in brackets is the expected molecular weight.