Structure and mechanism of the reversible photoswitch of a fluorescent protein

Andresen et al. 10.1073/pnas.0502772102.

Supporting Information

Files in this Data Supplement:

Supporting Table 1
Supporting Table 2
Supporting Methods
Supporting Table 3
Supporting Figure 6
Supporting Figure 7
Supporting Figure 8
Supporting Figure 9
Supporting Movie 1
Supporting Figure 10




Supporting Figure 6

Fig. 6. Structures of the MYG chromophores. (a) Orthogonal views of the MYG chromophores from the four asFP595 crystal structures [asFP595 (not irradiated), asFP595-A143S (not irradiated), asFP595-A143Sh*n, asFP595-S158V (not irradiated)] after least-squares superpositioning of all Ca atoms of the protein molecules. Atoms are color coded by atom type (carbon wt, cyan; carbon A143S, salmon; carbon S158V, beige; oxygen, red; nitrogen, blue). (b) Difference Fourier maps generated with phases calculated from the final model of the asFP595 (not irradiated) structure and with structure factor differences F(A143S) – F(wt) (Left, contoured at 2.5s), F(A143Sh*n) –F(wt) (Middle, contoured at 3s) and F(A143Sh*n) – F(A143S) (Right, contoured at 3s). The depicted chromophores of the cis and trans conformantion are intended to aid orientation. The maps show positive differences for features, which are present at a higher level in A143S compared with asFP595 (not irradiated) (Left), in A143Sh*n compared with asFP595 (not irradiated), and in A143Sh*n compared with A143S (Right). These data show the increasing fractions of the chromophores, which are in the cis conformation from asFP595 (not irradiated) over A143S to A143Sh*n. Color coding is as in a.





Supporting Figure 7

Fig. 7. Influence of A143S and S158V on the chromophor of asFP595. Both mutations, A143S and S158V, support the cis conformation of the chromophor. The mutation A143S stabilizes the cis conformation of the chromophor, whereas the mutation S158V blocks the transconformation. Displayed are the chromophores of asFP595 ("off" state, cyan carbon), asFP595-S158V ("on" state, salmon carbon) and asFP595-A143S (on state, magenta carbon). In asFP595-A143S, the serine stabilizes the cis chromophore through an additional H bond to the terminal oxygen of the hydroxyphenyl ring (left arrow). The valine introduced by the mutation S158V blocks the trans conformation of the chromophore in asFP595-S158V. This finding is shown by the overlapping electron densities of V158 and the wt trans chromophore (right arrow).





Supporting Figure 8

Fig. 8. Rearrangement and fluctuations of residues near the backbone break after isomerization. Smoothed rms deviation (rmsd) of MYG (all atoms) and neighboring residues 61–62 and 66–67 (backbone atoms) during equilibration after spontaneous transcis isomerization. The residues C62 and S66 directly adjacent to MYG exhibit a larger rms deviation (0.12 and 0.09 nm, respectively, averaged over the last 650 ps) than second neighbors S61 and K67 (0.10 and 0.05 nm, respectively). Residues S61 and C62 N-terminal of the chromophore show significantly larger rms deviation values than their respective C-terminal counterparts K67 and S66.





Supporting Figure 9

Fig. 9. Interactions with residues affect the isomerization process. For all four simulated pathways, the same residues close to the chromophore are crucial for chromophore positioning and/or isomerization. The sum of Coulomb and vdW MYG-protein interactions are shown, averaged for 10 trajectories. (a) During HTbot, all favorable MYG–protein interactions stay intact, and no significant energy barrier is found. (b) In contrast, along the HTtop pathway, T60 exhibits a significant energy barrier. (c and d) Both R pathways lead to unfavorable hydrophobic interactions. In particular, Rtop (c) is mainly inhibited by H197. However, this residue does not affect Rbot (d), which is largely hampered by T60 instead. Irrespective of the isomerization mechanism, the final cis state is stabilized with respect to trans by H197, A143, and M160. The favorable interaction to A143 found in the simulation corresponds to the hydrogen bond to S143 in the on state of the A143S mutant. The described features are not only prominent in the shown averages but also in all individual trajectories (error of 1-3 kJ/mol).





Supporting Figure 10

Fig. 10. Dihedral potentials used for the free MD simulations. Solid line, t; dashed line, j. The potential energy barriers of both dihedrals were varied between 0 (dashed) and 7 kcal/mol (dotted line) in test simulations.





Supporting Movie 1

Movie 1. Representative trajectory of spontaneous trans-cis isomerization through HTbot as observed in MD simulations starting from the excited asFP595 off state. To alleviate the observation of the trans-cis isomerization only selected residues are shown, whereas the underlying simulations accounted for the whole protein. Colors: cyan, MYG; asFP595 chain A, yellow; chain B, orange.





Table 1. Crystallographic data and refinement

Crystal

wt asFP595

(off state)

asFP595-S158V

asFP595-A143S

asFP595-A143Sh*n, 1 min

asFP595-A143S h*n, 5 min

Data collection

Radiation source

BW6, DESY

Rot. anode

BW6, DESY

BW6, DESY

Rot. anode

Wavelength, Å

1.05

1.5418

1.05

1.05

1.5418

Space group

C2221

C2221

C2221

C2221

C2221

Unit cell (Å, °)

a = 76.2, b = 126.5, c = 94.1

a = 75.4, b = 125.3, c = 92.6

a = 76.0, b = 126.3, c = 93.8

a = 76.5, b = 125.9, c = 93.9

a = 75.5, b = 125.4, c = 93.1

Resolution, Å

30.0 – 1.30

30.0 – 1.70

30.0 – 1.45

30.0 – 1.45

30.0 – 1.86

Reflections

Unique

111,048

48,031

79,647

77,265

37,242

Redundancy

4.7

4.4

3.8

3.7

3.8

Completeness, %

99.7 (99.8)

99.0 (98.0)

99.6 (99.2)

96.1 (94.0)

99.4 (95.8)

I/s, I

27.2 (2.0)

16.9 (1.9)

22.6 (1.7)

25.7 (2.4)

10.5 (2.0)

Rsym*

0.057 (0.384)

0.075 (0.557)

0.051 (0.533)

0.058 (0.403)

0.096 (0.539)

Refinement

Resolution, Å

20.0 – 1.30

20.0 – 1.70

20.0 – 1.45

20.0 – 1.45

30.0 – 1.90

Reflections, no./%

105329/99.6

45596/99.1

75627/99.7

73359/96.1

33321/99.6

Test set, %

5

5

5

5

5

Rwork

15.3 (21.2)

18.1 (28.5)

17.3 (27.2)

17.6 (22.0)

18.6 (26.2)

Rfree

17.9 (25.5)

22.2 (31.0)

19.9 (31.7)

20.0 (24.6)

22.6 (34.3)

ESU, Å

0.031

0.083

0.046

0.046

0.122

Contents of a.u.

Protein atoms

3637

3658

3637

3637

3637

Water oxygens

659

463

577

574

566

Mean B-factors, Å2

Wilson

15.5

23.4

18.1

18.0

23.4

Protein

18.2

15.6

18.7

19.1

27.2

Water

33.7

41.6

34.7

34.2

40.6

Ramachandran

Preferred

91.2

91.6

91.3

92.0

92.3

Add. allowed

8.5

7.9

8.5

7.7

7.4

Gen. allowed

0.3

0.5

0.3

0.3

0

Disallowed

0

0

0

0.3

0.3

Rmsde geometry

Bond lengths, Å

0.012

0.009

0.012

0.012

0.012

Bond angles, °

1.60

1.46

1.59

1.71

1.30

Rmsd B-fac., Å2

Main chain bonds

2.5

1.7

1.7

1.7

1.37

Main chain angles

3.2

2.5

2.2

2.3

1.94

Side chain bonds

2.8

2.1

2.2

2.1

1.55

Side chain angles

3.7

3.2

3.1

3.1

2.24

PDB entry

2A50

2A52

2A53

2A54

2A56

a.u., asymmetric unit; rmsd, root-mean-square deviation. Data in parentheses are for the last 0.02Å (wt) or for the last 0.05Å (S158V and A143S).

*Rsym(I) = (ShklSi[½Ii(hkl) – <I(hkl)> ½] / ShklSi[Ii(hkl)]; Ii(hkl) – intensity of the ith measurement of hkl; <I(hkl)> – average value of hkl for all i measurements

Rwork = åhkl[½½Fobs½k½Fcalc½½] / åhkl[½Fobs½]; Rfree = åhklÌT[½½Fobs½k½Fcalc½½] / å hklÌT[½Fobs½]; hklÌT – test set

ESU, estimated overall coordinate error based on maximum likelihood.





Table 3. Ca root mean square deviations to asFP595 mol1

Molecule

Ca rmsd, Å

No. of fitting Ca

wt asFP595 mol 2

1.0

226

asFP595-A143S on mol 1

0.2

229

asFP595-A143S on mol 2

1.0

226

asFP595-A143S off mol 1

0.2

229

asFP595-A143S off mol 2

1.0

226

asFP595-S158V mol 1

0.3

229

asFP595-S158V mol 2

1.0

226

KFP comparison with wt mol 1

1.2

227

KFP comparison with wt mol 2

0.3

226

asFP595 tetramers contain two identical dimers. Each dimer, which is slightly asymmetric, contains two asFP595 molecules (molecule 1 and 2, respectively). Ca root mean square deviations are calculated for a comparison with asFP595 molecule 1. Numbers are rounded on the first position after the decimal point.

This Article

  1. PNAS September 13, 2005 vol. 102 no. 37 13070-13074
  1. AbstractFree
  2. Figures Only
  3. Full Text
  4. Full Text (PDF)
  5. » Supporting Information