Plants adapt to the availability of light throughout their lives because it regulates so many aspects of their growth and reproduction. To detect the level of light, plant cells use proteins called phytochromes, which are also found in some bacteria and fungi. Phytochrome proteins change shape when they are exposed to red light, and this change alters the behaviour of the cell. The red light is absorbed by a molecule known as chromophore, which is connected to a region of the phytochrome called the PHY-tongue. This region undergoes one of the key structural changes that occur when the phytochrome protein absorbs light, turning from a flat sheet into a helix.

Claesson, Wahlgren, Takala et al. studied the structure of a bacterial phytochrome protein almost immediately after shining a very brief flash of red light using a laser. The experiments revealed that the structure of the protein begins to change within a trillionth of a second: specifically, the chromophore twists, which disrupts its attachment to the protein, freeing the protein to change shape. Claesson, Wahlgren, Takala et al. note that this structure is likely a very short-lived intermediate state, which however triggers more changes in the overall shape change of the protein.

One feature of the rearrangement is the disappearance of a particular water molecule. This molecule can be found at the core of many different phytochrome structures and interacts with several parts of the chromophore and the phytochrome protein. It is unclear why the water molecule is lost, but given how quickly this happens after the red light is applied it is likely that this disappearance is an integral part of the reshaping process.

Together these events disrupt the interactions between the chromophore and the PHY-tongue, enabling the PHY-tongue to change shape and alter the structure of the phytochrome protein. Understanding and controlling this process could allow scientists to alter growth patterns in plants, such as crops or weeds.

Phytochrome photosensor proteins are crucial for the optimal development of all vegetation on Earth (Butler et al., 1959; Gan et al., 2014; Quail et al., 1995). Prototypical phytochromes can exist in two photochemical states with differential cellular signaling activity, called red light-absorbing (Pr) and far-red light-absorbing (Pfr) state (Figure 1—figure supplement 1). As a result, phytochromes can distinguish two colors of light, providing plants, fungi, and bacteria with primitive two-color vision. Light is detected by a bilin chromophore, which is covalently linked to the photosensory core of the protein (Wagner et al., 2005), comprising of PAS (Per/Arndt/Sim), GAF (cGMP phosphodiesterase/adenyl cyclase/FhlA) and PHY (phytochrome-specific) domains. Two propionate side chains additionally anchor the chromophore non-covalently to the protein (Figure 1b). The signaling sites of the phytochrome are found in its C- and N-terminal output domains, which vary between species. Important for the signaling is a stretch of amino acids in the PHY domain, called the PHY-tongue, which changes from a β-sheet in Pr into an α-helix in Pfr state (Essen et al., 2008; Yang et al., 2008; Takala et al., 2014; Sanchez et al., 2019) The chromophore connects to the PHY-tongue via a strictly conserved aspartatic acid, which is expected to play a crucial role in signal transduction.

Figure 1 with 5 supplements see all

Download asset Open asset

Key to phytochrome function is the primary photoresponse on picosecond time scales. Here, light signals are translated into conformational changes. The changes arise in the electronically excited bilin, but must then be transduced to the surrounding protein residues. This prepares the protein for a formation of the first intermediate (Lumi-R for prototypical phytochromes), in which isomerization of the D-ring has likely occurred (Rüdiger et al., 1983; Dasgupta et al., 2009; Yang et al., 2012; Rockwell et al., 2009; Ihalainen et al., 2018). The mechanism that leads to the first intermediate is currently not well understood, because crystallographic observations of phytochromes directly after photoexcitation have not been available.

To address this gap of knowledge, we recorded time-resolved serial femtosecond X-ray crystallographic (SFX) data of the PAS-GAF domains of the phytochrome from Deinococcus radiodurans (DrBphP_CBD) at 1 ps and 10 ps after femtosecond optical excitation. The experiments were performed in Japan, using the SPring-8 Angstrom Compact Free Electron Laser (SACLA) tuned to 7 KeV (Tono et al., 2015). For homogeneous excitation of the crystals, we photoexcited micrometer-sized crystals in a grease jet with a photon density of 1.7 mJ/mm^₂ ($1/e^2$ measure, see Materials and methods) into the flank of the absorption peak at 640 nm (Figure 1—figure supplement 1). Taking into account the significant light scattering in the grease-buffer mixture (Figure 1—figure supplement 1), we estimate that the average number of photons per chromophore is 0.5–1 (see Materials and methods). We recorded the SFX data at 1 ps for several excitation fluences (Figure 2). Lowering the excitation density tenfold from 1.7 mJ/mm^₂ photons to 0.2 mJ/mm^₂ resulted in a joint reduction of all difference signals. Critical signals, like the twist of the D-ring and the photodissociation of the pyrrole water from the chromophore sustained when lowering the excitation densities, indicating that the signal arises predominately from one-photon excitation. The refined structure in dark (DrBphP_dark), 2.07 Å resolution (Table 1), was very similar to our previous dark structure solved by SFX (5K5B, RMSD 0.646 Å and 0.610 Å for monomers A and B) (Edlund et al., 2016), but the present crystals contained two monomers in the asymmetric unit (Figure 1a, Figure 1—figure supplement 2, Figure 1—figure supplement 3). The refined 1 ps structure was solved to 2.21 Å (Table 1).

Figure 2

Download asset Open asset

From the time-resolved data, we calculated Fourier difference electron density maps (${|F_o|}^{light}-{|F_o|}^{dark}$), which report on the change of structure due to optical excitation (see Materials and methods). Briefly, the diffraction data for light and dark were scaled to each other and subtracted, assuming preservation of the phases (see Materials and methods for details). The map at 1 ps indicates many significant changes in difference electron density (Figure 1a) above the background level of 3.0 standard deviations (σ) (Figure 1—figure supplement 4). The changes cluster around the chromophore, with the strongest negative densities for the pyrrole water (monomer A: −8.2σ, B: −9.4σ, Table 2). The map at 10 ps contains similar significant features, but at weaker overall intensity (pyrrole water A: −5.0σ, B:−6.5σ) (Figure 1—figure supplement 4 and Figure 1—figure supplement 5). We ascribe this to a lower population of the activated state at 10 ps compared to 1 ps. Monomer A has a lower signal strength than monomer B, but provided a clearer difference map around the chromophore. We refined a structural model (DrBphP_1ps) using extrapolated structure factors (Figure 3—figure supplement 1; Pande et al., 2016). The refinement of the structure against the 1 ps data was successful using a photoexcitation density of 8%. However, we aborted our attempts to refine a structural model against the 10 ps data, as the model would have become unreliable due to an even lower photoactivation yield. We focus our discussion on monomer A and the 1 ps time point, although all conclusions are supported by monomer B and the features observed in the difference maps at 10 ps (Table 2, Figure 1—figure supplement 5).

First, we inspect the D-ring region at 1 ps (Figure 3). We observe strong negative difference density features on the atoms of the D-ring (marked I, II, III), correlating with density gains at both faces of the ring (IV, V, VI). These features strongly indicate that the D-ring twists. The positive feature IV homes the N-H and C = O groups, whereas V and VI indicate densities for the methyl and vinyl groups in the twisted ring (Figure 3c). Excellent agreement was obtained between the observed difference map and the difference map calculated from DrBphP_1ps ($F_c^{1ps}-F_c^{dark}$), when the D-ring twists (C14-C15-C16-ND) from around 20° in the dark to 60° monomer A) and 90° monomer B) at 1 ps (Figure 3c and e). Although the twisting movement is clearly indicated by the difference map, we judge the precision of the angle to be low and approximately ±25°.

Figure 3 with 4 supplements see all

Download asset Open asset

Concomitant with the twist of the D-ring, the C-ring translates by approximately 0.69 Å as indicated by the correlated negative (VII) and positive (VIII) electron density (Figure 3a). Furthermore, the C-ring propionate chain detaches from its conserved anchoring residues Ser272 and Ser274 (IX and X, Figure 3a). The strictly conserved His260 retracts from its position (XI and XII) and Tyr263 moves upward at 1 ps (XIII and XIV, Figure 3b). The water network connecting the C-ring propionate, the D-ring C = O, and His290 rearranges accordingly (Figure 3a). The excellent agreement between calculated and observed difference maps confirms these observations (Figure 3d and e, Figure 3—figure supplement 2). We conclude that the twist of the D-ring causes detachment of the C-ring propionate from the protein scaffold by dislocation of the C-ring, facilitated by the associated hydrogen bonding network.

Turning our attention to the B-ring, we find that the B-ring propionate breaks its salt bridge to Arg254 (Figure 3—figure supplement 3). However, this is not caused by movements of the chromophore backbone, as we observe little change on the B-ring itself. Instead, we find that a water bridge between the B- and C-ring propionates is broken as indicated by negative difference electron densities on the waters (Figure 3—figure supplement 3). Additionally, the highly conserved helix from Ser257 to Val269, moves away from the chromophore by an average of 0.36 Å in monomer A and 0.62 Å in monomer B (distances relative to the pyrrole water, Figure 3—figure supplement 4). The changes of the D-ring are transduced to Ser257 via the side chains of His260 and Tyr263, and as a result, the hydrogen bond of Ser257 to the B-ring propionate group breaks. The amino acids in the stretch are over 50% conserved (Figure 3—figure supplement 4), suggesting that it has evolved to transfer an ultrafast signal. We conclude that relaxation of the protein is necessary for the detachment of the B-ring propionate from the protein scaffold.

Next to the changes around the D-ring, the maps reveal strong difference electron density on the A-ring (XVIII and XIX), Asp207 (XX to XXIII) (Figure 4a) and the pyrrole water (XV) (Figure 4b). When interpreted and modelled as downward movement of the A-ring and Asp207 and photodissociation of the pyrrole water from the chromophore, excellent agreement between calculated and observed difference electron density is obtained (Figure 4—figure supplement 1). The A-ring is covalently attached to the protein backbone in phytochromes (Song et al., 2014), which renders complete isomerization impossible, but is sufficiently flexible to accommodate the proposed changes. The pyrrole water may either move to feature (XVI), or occupy an anisotropic, worm-shaped feature which extends from the A-ring to the D-ring (XVII) (Figure 4b).

Figure 4 with 1 supplement see all

Download asset Open asset

Furthermore, correlated negative and positive electron density features are observed on backbone atoms of the highly conserved stretch from Pro201 to His209. These difference electron density features indicate that the residues move away from the centre of the chromophore by an average of 0.54 Å and 0.57 Å in monomers A and B, respectively (Figure 3—figure supplement 4b). The stretch includes Asp207 and it is located between the A-ring and the PHY domain, which makes it plausible that the changes in the chromophore cause this protein rearrangement. The changes are complemented by significant rearrangements of a stretch of waters and a conserved Tyr176 (Figure 4c).

The structure of DrBphP_CBD 1 ps after photoexcition reveals changes of the biliverdin chromophore and the surrounding residues. We find a twist of the D-ring, displacement of the C-ring, and associated changes of the water network which connects the D-ring, the C-ring propionate, and His290. Further, we identify a disruption of the salt bridge between the B-ring propionate and the Arg254, and significant changes around the A-ring, Asp207 and the pyrrole water. The changes are retained at 10 ps, even though at a lower population (Figure 1—figure supplement 5). The extensive and coordinated structural changes in the binding pocket (Figure 3) manifest a liberation of the chromophore from the protein scaffold, which we propose to be necessary for the conformational rearrangements to occur in the downstream photoconversion to Pfr.

Infrared spectral data indicate significant reorganization of the chromophore and several amino acids including the PHY-tongue region as early as in Lumi-R state, which is the first known ground state intermediate in the photoconversion from Pr to Pfr (Ihalainen et al., 2018; van Thor et al., 2007). However, the structure of the bilin and the binding pocket in Lumi-R is not known, because structural information is missing. Since the quantum yield of reaching the Lumi-R state is low (on the order of 10%), spectroscopic investigation of the mechanism is difficult, and it is currently not fully established how the Lumi-R state is reached. Crystallographic data does not report on whether the chromophore is electronically excited or not, hence we cannot determine whether the structure that we observe is in a relaxed excited state or in a ground state. The time delay (1 ps) supports that our structure presents an intermediate enroute to the Lumi-R state.

The D-ring of the bilin chromophore isomerizes around the C15-C16 bond from (Z) in Pr to (E) in Pfr (Rüdiger et al., 1983; Burgie et al., 2016; Takala et al., 2014; Yang et al., 2008; Essen et al., 2008). Circular dichroism spectroscopy and solid-state NMR spectroscopy have indicated that the position of the D-ring inverts from an ’$\alpha$’-facial (Pr) to a ’$\beta$’-facial (Pfr) position in cyanobacterial and plant phytochromes, whereas it stays ’$\alpha$’-facial in bacterial phytochromes (Rockwell et al., 2009; Song et al., 2011; Song et al., 2018). Based on anticipated steric clashes with the C-ring methyl group, it has been proposed that the D-ring rotates counter-clockwise in plant and cyanobacterial phytochromes, but clockwise in bacterial phytochromes (Rockwell et al., 2009). Moreover, spectroscopy has shown that the D-ring of the bilin chromophore is already isomerized in the Lumi-R state (van Thor et al., 2007; Heyne et al., 2002; Yang et al., 2012).

Seemingly contradictory, we now observe that the D-ring is rotated counter-clockwise by tens of degrees for the bacterial DrBphP at 1 ps time delay (Figure 3c and e). The conformation is strongly supported by the difference map. It contains two positive peaks (V and VI in Figure 3c), which indicate the new position of the vinyl and methyl group of the D-ring. We tested models in which the D-ring was rotated in a clockwise direction, but the agreement with the experimental difference map decreased. Thus, it may be that the D-ring indeed rotates counter-clockwise in bacterial phytochromes, similar to plant and cyanobacterial phytochromes. For complete isomerization, this would mean that the C-ring moves out of the way during the rotation. We observe significant movements of the C-ring, which may be an indication for that such a mechanism is possible. Raising a note of caution, we cannot fully exclude that the truncation of our phytochrome construct or the crystal packing influences the direction of rotation. NMR studies have reported conformational heterogeneity in the chromophore binding pocket of phytochromes in solution (Song et al., 2011; Lim et al., 2018; Song et al., 2018; Gustavsson et al., 2020). Crystallization could select one of the conformations, which may have a preferred rotation in the counter-clockwise direction. More experiments are needed to clarify this question.

It is interesting to compare the structural changes at 1 ps time delay to the changes observed in the conversion between Pr and Pfr (Burgie et al., 2016; Takala et al., 2014; Stojković et al., 2014). Major changes include a flipped D-ring, changes in conserved residues of the chromophore-binding pocket, for example Tyr176, His201 and Phe203, and refolding of the PHY tongue. The PHY tongue is not included in our construct, but Tyr176 and Phe203 are associated with difference electron density features in our maps (Table 2). However, the movements are much smaller at 1 ps compared to the Pr-to-Pfr transition. This is not unexpected, given the short time delays, but it shows that the residues are tightly coupled to the chromophore. Interestingly, the Pr and Pfr structures also reveal a sliding movement of the entire chromophore (Yang et al., 2011; Burgie et al., 2016; Takala et al., 2014). This requires that the propionic groups have to break their bonds to the protein scaffold. Our data indicate that this is part of the primary photoresponse.

The photodissociation of the pyrrole water from the chromophore is a surprising finding. The pyrrole water is ubiquitously found in phytochrome structures (Essen et al., 2008; Yang et al., 2008; Otero et al., 2016; Burgie et al., 2016; Wagner et al., 2005; Burgie et al., 2014; Schmidt et al., 2018; Yang et al., 2011). Our fluence dependent SFX data show that the negative density on the pyrrole water is the last signal to disappear when lowering the photon excitation densities 10-fold (Figure 2). This makes us confident that the photodissociation reaction is not caused by multi-photon effects. The removal of the water requires significant energy, because the hydrogen bonds to the A-, B-, and C-rings of the chromophore and the backbone C = O group of Asp207 have to be broken. We do not think that the twist of the D-ring causes this through direct steric interactions, because there is no contact between the pyrrole water and the D-ring. Rather, it may be triggered by an excited state charge redistribution between the pyrrole water and the chromophore, for example by ultrafast proton or electron transfer (Toh et al., 2010). Such charge re-distributions are typically facilitated by changes in geometry (Nosenko et al., 2008) and may therefore be caused indirectly by structural changes of the A-, C-, or D-rings, but this requires further investigation.

Conformational changes of the A-ring, Asp207 and the pyrrole water have not been considered to occur on picosecond time scales. The strictly conserved Asp207 is a key residue for signal transduction because it connects the chromophore to the PHY-tongue in Pr and Pfr (Essen et al., 2008; Yang et al., 2008; Takala et al., 2014). Its displacement suggests, together with the relocation of the residue stretch surrounding it, that disruption of the GAF-PHY interface may occur as early as 1 ps after photoexcitation (Figure 3—figure supplement 4b). With a hydrogen bond to the pyrrole water and in tight steric contact with the A-ring, Asp207 thereby acts as an extended arm of the chromophore. We propose that the photodissociation of the pyrrole water from the bilin and the change of the A-ring are integral parts of ultrafast phytochrome signaling toward the PHY domain.

We demonstrate that within 1 ps, the D-ring twists, that the chromophore is liberated from the protein (Figure 5a) and that movements of the pyrrole water, the A-ring and Asp207 lead to signaling directed toward the PHY-tongue (Figure 5b). When mapped on the structure of the complete photosensory core module (Takala et al., 2014), both changes work together to destabilize the Arg466:Asp207 salt bridge. Tyr263 moves up, caused by the twist of the D-ring, and Asp207 moves down, caused by changes of the A-ring, retracting both residues from the salt bridge.

Figure 5

Download asset Open asset

Our data reveal a highly collective primary photoresponse for phytochromes. This is consistent with the fact that most point mutations of conserved residues alter, but do not inhibit, photoconversion (Wagner et al., 2008). The ultrafast structural changes are more extensive than in bacteriorhodopsin, photoactive yellow proteins, and in a fluorescent protein (Pande et al., 2016; Nogly et al., 2018; Coquelle et al., 2018). While previously observed ultrafast backbone movements have been interpreted as ’protein quakes’ for myoglobin and bacteriorhodopsin (Barends et al., 2015; Nogly et al., 2018; Toh et al., 2010), the present backbone motion in the phytochrome binding pocket are much more directed (Figure 3—figure supplement 4). The changes occur in highly conserved regions of the protein and are part of the collective signaling response of the entire binding pocket.

Phytochromes have to be able to stabilize the bilin and to direct its photoisomerization from two photochemical ground states, Pr and Pfr. These differ both structurally and electronically, which precludes a single reaction trajectory for isomerization in the two directions. With this in mind, the observed primary photoresponse is reasonable. The structural signal is highly delocalized already at 1 ps, causing near-simultaneous liberation of the chromophore and initial signal transduction. We propose that these reaction trajectories stabilize each other, navigating the protein into a productive reaction path. The multidimensional reaction trajectory is consistent with the low quantum yields for photoconversion (Lamparter et al., 1997), which are characteristic for the phytochrome superfamily. Whereas the twisting motion of the D-ring has been the working model for phytochrome activation and is now confirmed, the photodissociation of the pyrrole water is highly surprising. We propose that both chemical events work together and enable phytochrome proteins to translate light information into structural signals, guiding the growth and development of plants, fungi, and bacteria on Earth.

Crystallography data have been submitted to protein data bank (PDB) dark:ID: D_1292104678 and PDB ID: 6T3L 1ps:ID: D_1292104679 and PDB ID: 6T3U. Raw diffraction images are deposited under ID 121 at CXIDB.

1. 1. Claesson E
   2. Wahlgren WY
   3. Takala H
   4. Pandey S
   5. Castillon L
   6. Kuznetsova V
   7. Henry L
   8. Panman M
   9. Carrillo M
   10. Kübel J
   11. Nanekar R
   12. Isaksson L
   13. Nimmrich A
   14. Cellini A
   15. Morozov D
   16. Maj M
   17. Kurttila M
   18. Bosman R
   19. Nango E
   20. Tanaka R
   21. Tanaka T
   22. Fangjia L
   23. Iwata S
   24. Owada S
   25. Moffat K
   26. Groenhof G
   27. Stojković EA
   28. Ihalainen JA
   29. Schmidt M
   30. Westenhoff S

   (2020) CXIDB

   The primary structural photoresponse of phytochrome proteins captured by a femtosecond X-ray laser.

   https://doi.org/10.11577/1607859
2. 1. Claesson E
   2. Takala H
   3. Wahlgren W
   4. Pandey S
   5. Schmidt M
   6. Westenhoff S

   (2020) RCSB Protein Data Bank

   ID 6T3L. PAS-GAF fragment from Deinococcus radiodurans phytochrome in dark state.

   http://www.rcsb.org/structure/6T3L
3. 1. Claesson E
   2. Takala H
   3. Wahlgren W
   4. Pandey S
   5. Schmidt M
   6. Westenhoff S

   (2020) RCSB Protein Data Bank

   ID 6T3U. PAS-GAF fragment from Deinococcus radiodurans phytochrome 1ps after photoexcitation.

   http://www.rcsb.org/structure/6T3U

1. 1. Adams PD
   2. Afonine PV
   3. Bunkóczi G
   4. Chen VB
   5. Davis IW
   6. Echols N
   7. Headd JJ
   8. Hung LW
   9. Kapral GJ
   10. Grosse-Kunstleve RW
   11. McCoy AJ
   12. Moriarty NW
   13. Oeffner R
   14. Read RJ
   15. Richardson DC
   16. Richardson JS
   17. Terwilliger TC
   18. Zwart PH

   (2010) PHENIX: a comprehensive Python-based system for macromolecular structure solution

   Acta Crystallographica Section D Biological Crystallography 66:213–221.

   https://doi.org/10.1107/S0907444909052925

   • PubMed
   • Google Scholar
2. 1. Barends TR
   2. Foucar L
   3. Ardevol A
   4. Nass K
   5. Aquila A
   6. Botha S
   7. Doak RB
   8. Falahati K
   9. Hartmann E
   10. Hilpert M
   11. Heinz M
   12. Hoffmann MC
   13. Köfinger J
   14. Koglin JE
   15. Kovacsova G
   16. Liang M
   17. Milathianaki D
   18. Lemke HT
   19. Reinstein J
   20. Roome CM
   21. Shoeman RL
   22. Williams GJ
   23. Burghardt I
   24. Hummer G
   25. Boutet S
   26. Schlichting I

   (2015) Direct observation of ultrafast collective motions in CO myoglobin upon ligand dissociation

   Science 350:445–450.

   https://doi.org/10.1126/science.aac5492

   • PubMed
   • Google Scholar
3. 1. Barty A
   2. Kirian RA
   3. Maia FR
   4. Hantke M
   5. Yoon CH
   6. White TA
   7. Chapman H

   (2014) Cheetah: software for high-throughput reduction and analysis of serial femtosecond X-ray diffraction data

   Journal of Applied Crystallography 47:1118–1131.

   https://doi.org/10.1107/S1600576714007626

   • PubMed
   • Google Scholar
4. 1. Battye TG
   2. Kontogiannis L
   3. Johnson O
   4. Powell HR
   5. Leslie AG

   (2011) iMOSFLM: a new graphical interface for diffraction-image processing with MOSFLM

   Acta Crystallographica. Section D, Biological Crystallography 67:271–281.

   https://doi.org/10.1107/S0907444910048675

   • PubMed
   • Google Scholar
5. 1. Burgie ES
   2. Wang T
   3. Bussell AN
   4. Walker JM
   5. Li H
   6. Vierstra RD

   (2014) Crystallographic and electron microscopic analyses of a bacterial phytochrome reveal local and global rearrangements during photoconversion

   Journal of Biological Chemistry 289:24573–24587.

   https://doi.org/10.1074/jbc.M114.571661

   • PubMed
   • Google Scholar
6. 1. Burgie ES
   2. Zhang J
   3. Vierstra RD

   (2016) Crystal structure of Deinococcus phytochrome in the photoactivated state reveals a cascade of structural rearrangements during photoconversion

   Structure 24:448–457.

   https://doi.org/10.1016/j.str.2016.01.001

   • PubMed
   • Google Scholar
7. 1. Butler WL
   2. Norris KH
   3. Siegelman HW
   4. Hendricks SB

   (1959) Detection, assay, and preliminary purification of the pigment controlling photoresponsive development of plants

   PNAS 45:1703–1708.

   https://doi.org/10.1073/pnas.45.12.1703

   • Google Scholar
8. 1. Coquelle N
   2. Sliwa M
   3. Woodhouse J
   4. Schirò G
   5. Adam V
   6. Aquila A
   7. Barends TRM
   8. Boutet S
   9. Byrdin M
   10. Carbajo S
   11. De la Mora E
   12. Doak RB
   13. Feliks M
   14. Fieschi F
   15. Foucar L
   16. Guillon V
   17. Hilpert M
   18. Hunter MS
   19. Jakobs S
   20. Koglin JE
   21. Kovacsova G
   22. Lane TJ
   23. Lévy B
   24. Liang M
   25. Nass K
   26. Ridard J
   27. Robinson JS
   28. Roome CM
   29. Ruckebusch C
   30. Seaberg M
   31. Thepaut M
   32. Cammarata M
   33. Demachy I
   34. Field M
   35. Shoeman RL
   36. Bourgeois D
   37. Colletier JP
   38. Schlichting I
   39. Weik M

   (2018) Chromophore twisting in the excited state of a photoswitchable fluorescent protein captured by time-resolved serial femtosecond crystallography

   Nature Chemistry 10:31–37.

   https://doi.org/10.1038/nchem.2853

   • PubMed
   • Google Scholar
9. 1. Dasgupta J
   2. Frontiera RR
   3. Taylor KC
   4. Lagarias JC
   5. Mathies RA

   (2009) Ultrafast excited-state isomerization in phytochrome revealed by femtosecond stimulated raman spectroscopy

   PNAS 106:1784–1789.

   https://doi.org/10.1073/pnas.0812056106

   • PubMed
   • Google Scholar
10. 1. Duisenberg AJM

    (1992) Indexing in single-crystal diffractometry with an obstinate list of reflections

    Journal of Applied Crystallography 25:92–96.

    https://doi.org/10.1107/S0021889891010634

    • Google Scholar
11. 1. Edlund P
    2. Takala H
    3. Claesson E
    4. Henry L
    5. Dods R
    6. Lehtivuori H
    7. Panman M
    8. Pande K
    9. White T
    10. Nakane T
    11. Berntsson O
    12. Gustavsson E
    13. Båth P
    14. Modi V
    15. Roy-Chowdhury S
    16. Zook J
    17. Berntsen P
    18. Pandey S
    19. Poudyal I
    20. Tenboer J
    21. Kupitz C
    22. Barty A
    23. Fromme P
    24. Koralek JD
    25. Tanaka T
    26. Spence J
    27. Liang M
    28. Hunter MS
    29. Boutet S
    30. Nango E
    31. Moffat K
    32. Groenhof G
    33. Ihalainen J
    34. Stojković EA
    35. Schmidt M
    36. Westenhoff S

    (2016) The room temperature crystal structure of a bacterial phytochrome determined by serial femtosecond crystallography

    Scientific Reports 6:35279.

    https://doi.org/10.1038/srep35279

    • PubMed
    • Google Scholar
12. 1. Emsley P
    2. Lohkamp B
    3. Scott WG
    4. Cowtan K

    (2010) Features and development of coot

    Acta Crystallographica. Section D, Biological Crystallography 66:486–501.

    https://doi.org/10.1107/S0907444910007493

    • PubMed
    • Google Scholar
13. 1. Essen L-O
    2. Mailliet J
    3. Hughes J

    (2008) The structure of a complete phytochrome sensory module in the pr ground state

    PNAS 105:14709–14714.

    https://doi.org/10.1073/pnas.0806477105

    • Google Scholar
14. 1. Gan F
    2. Zhang S
    3. Rockwell NC
    4. Martin SS
    5. Lagarias JC
    6. Bryant DA

    (2014) Extensive remodeling of a cyanobacterial photosynthetic apparatus in far-red light

    Science 345:1312–1317.

    https://doi.org/10.1126/science.1256963

    • PubMed
    • Google Scholar
15. 1. Gustavsson E
    2. Isaksson L
    3. Persson C
    4. Mayzel M
    5. Brath U
    6. Vrhovac L
    7. Ihalainen JA
    8. Karlsson BG
    9. Orekhov V
    10. Westenhoff S

    (2020) Modulation of structural heterogeneity controls phytochrome photoswitching

    Biophysical Journal 118:415–421.

    https://doi.org/10.1016/j.bpj.2019.11.025

    • Google Scholar
16. 1. Henderson R
    2. Moffat JK

    (1971) The difference fourier technique in protein crystallography: errors and their treatment

    Acta Crystallographica Section B Structural Crystallography and Crystal Chemistry 27:1414–1420.

    https://doi.org/10.1107/S0567740871004060

    • Google Scholar
17. 1. Heyne K
    2. Herbst J
    3. Stehlik D
    4. Esteban B
    5. Lamparter T
    6. Hughes J
    7. Diller R

    (2002) Ultrafast dynamics of phytochrome from the Cyanobacterium synechocystis, reconstituted with phycocyanobilin and phycoerythrobilin

    Biophysical Journal 82:1004–1016.

    https://doi.org/10.1016/S0006-3495(02)75460-X

    • PubMed
    • Google Scholar
18. 1. Ihalainen JA
    2. Gustavsson E
    3. Schroeder L
    4. Donnini S
    5. Lehtivuori H
    6. Isaksson L
    7. Thöing C
    8. Modi V
    9. Berntsson O
    10. Stucki-Buchli B
    11. Liukkonen A
    12. Häkkänen H
    13. Kalenius E
    14. Westenhoff S
    15. Kottke T

    (2018) Chromophore-Protein interplay during the phytochrome photocycle revealed by Step-Scan FTIR spectroscopy

    Journal of the American Chemical Society 140:12396–12404.

    https://doi.org/10.1021/jacs.8b04659

    • PubMed
    • Google Scholar
19. 1. Lamparter T
    2. Mittmann F
    3. Gärtner W
    4. Börner T
    5. Hartmann E
    6. Hughes J

    (1997) Characterization of recombinant phytochrome from the Cyanobacterium synechocystis

    PNAS 94:11792–11797.

    https://doi.org/10.1073/pnas.94.22.11792

    • PubMed
    • Google Scholar
20. 1. Lehtivuori H
    2. Rissanen I
    3. Takala H
    4. Bamford J
    5. Tkachenko NV
    6. Ihalainen JA

    (2013) Fluorescence properties of the chromophore-binding domain of bacteriophytochrome from Deinococcus radiodurans

    The Journal of Physical Chemistry B 117:11049–11057.

    https://doi.org/10.1021/jp312061b

    • PubMed
    • Google Scholar
21. 1. Lim S
    2. Yu Q
    3. Gottlieb SM
    4. Chang CW
    5. Rockwell NC
    6. Martin SS
    7. Madsen D
    8. Lagarias JC
    9. Larsen DS
    10. Ames JB

    (2018) Correlating structural and photochemical heterogeneity in cyanobacteriochrome NpR6012g4

    PNAS 115:4387–4392.

    https://doi.org/10.1073/pnas.1720682115

    • PubMed
    • Google Scholar
22. 1. McCoy AJ
    2. Grosse-Kunstleve RW
    3. Adams PD
    4. Winn MD
    5. Storoni LC
    6. Read RJ

    (2007) Phaser crystallographic software

    Journal of Applied Crystallography 40:658–674.

    https://doi.org/10.1107/S0021889807021206

    • PubMed
    • Google Scholar
23. 1. Murshudov GN
    2. Skubák P
    3. Lebedev AA
    4. Pannu NS
    5. Steiner RA
    6. Nicholls RA
    7. Winn MD
    8. Long F
    9. Vagin AA

    (2011) REFMAC5 for the refinement of macromolecular crystal structures

    Acta Crystallographica Section D Biological Crystallography 67:355–367.

    https://doi.org/10.1107/S0907444911001314

    • PubMed
    • Google Scholar
24. 1. Nakane T
    2. Joti Y
    3. Tono K
    4. Yabashi M
    5. Nango E
    6. Iwata S
    7. Ishitani R
    8. Nureki O

    (2016) Data processing pipeline for serial femtosecond crystallography at SACLA

    Journal of Applied Crystallography 49:1035–1041.

    https://doi.org/10.1107/S1600576716005720

    • PubMed
    • Google Scholar
25. 1. Nogly P
    2. Weinert T
    3. James D
    4. Carbajo S
    5. Ozerov D
    6. Furrer A
    7. Gashi D
    8. Borin V
    9. Skopintsev P
    10. Jaeger K
    11. Nass K
    12. Båth P
    13. Bosman R
    14. Koglin J
    15. Seaberg M
    16. Lane T
    17. Kekilli D
    18. Brünle S
    19. Tanaka T
    20. Wu W
    21. Milne C
    22. White T
    23. Barty A
    24. Weierstall U
    25. Panneels V
    26. Nango E
    27. Iwata S
    28. Hunter M
    29. Schapiro I
    30. Schertler G
    31. Neutze R
    32. Standfuss J

    (2018) “Retinal isomerization in bacteriorhodopsin captured by a femtosecond x-ray laser,”

    Science 361:eaat0094.

    https://doi.org/10.1126/science.aat0094

    • PubMed
    • Google Scholar
26. 1. Nosenko Y
    2. Wiosna-Sałyga G
    3. Kunitski M
    4. Petkova I
    5. Singh A
    6. Buma WJ
    7. Thummel RP
    8. Brutschy B
    9. Waluk J

    (2008) Proton transfer with a twist? femtosecond dynamics of 7-(2-pyridyl)indole in condensed phase and in supersonic jets

    Angewandte Chemie International Edition 47:6037–6040.

    https://doi.org/10.1002/anie.200801350

    • PubMed
    • Google Scholar
27. 1. Otero LH
    2. Klinke S
    3. Rinaldi J
    4. Velázquez-Escobar F
    5. Mroginski MA
    6. Fernández López M
    7. Malamud F
    8. Vojnov AA
    9. Hildebrandt P
    10. Goldbaum FA
    11. Bonomi HR

    (2016) Structure of the Full-Length bacteriophytochrome from the plant pathogen Xanthomonas campestris provides clues to its Long-Range signaling mechanism

    Journal of Molecular Biology 428:3702–3720.

    https://doi.org/10.1016/j.jmb.2016.04.012

    • PubMed
    • Google Scholar
28. 1. Pande K
    2. Hutchison CD
    3. Groenhof G
    4. Aquila A
    5. Robinson JS
    6. Tenboer J
    7. Basu S
    8. Boutet S
    9. DePonte DP
    10. Liang M
    11. White TA
    12. Zatsepin NA
    13. Yefanov O
    14. Morozov D
    15. Oberthuer D
    16. Gati C
    17. Subramanian G
    18. James D
    19. Zhao Y
    20. Koralek J
    21. Brayshaw J
    22. Kupitz C
    23. Conrad C
    24. Roy-Chowdhury S
    25. Coe JD
    26. Metz M
    27. Xavier PL
    28. Grant TD
    29. Koglin JE
    30. Ketawala G
    31. Fromme R
    32. Šrajer V
    33. Henning R
    34. Spence JC
    35. Ourmazd A
    36. Schwander P
    37. Weierstall U
    38. Frank M
    39. Fromme P
    40. Barty A
    41. Chapman HN
    42. Moffat K
    43. van Thor JJ
    44. Schmidt M

    (2016) Femtosecond structural dynamics drives the trans/cis isomerization in photoactive yellow protein

    Science 352:725–729.

    https://doi.org/10.1126/science.aad5081

    • PubMed
    • Google Scholar
29. 1. Pandey S
    2. Bean R
    3. Sato T
    4. Poudyal I
    5. Bielecki J
    6. Cruz Villarreal J
    7. Yefanov O
    8. Mariani V
    9. White TA
    10. Kupitz C
    11. Hunter M
    12. Abdellatif MH
    13. Bajt S
    14. Bondar V
    15. Echelmeier A
    16. Doppler D
    17. Emons M
    18. Frank M
    19. Fromme R
    20. Gevorkov Y
    21. Giovanetti G
    22. Jiang M
    23. Kim D
    24. Kim Y
    25. Kirkwood H
    26. Klimovskaia A
    27. Knoska J
    28. Koua FHM
    29. Letrun R
    30. Lisova S
    31. Maia L
    32. Mazalova V
    33. Meza D
    34. Michelat T
    35. Ourmazd A
    36. Palmer G
    37. Ramilli M
    38. Schubert R
    39. Schwander P
    40. Silenzi A
    41. Sztuk-Dambietz J
    42. Tolstikova A
    43. Chapman HN
    44. Ros A
    45. Barty A
    46. Fromme P
    47. Mancuso AP
    48. Schmidt M

    (2020) Time-resolved serial femtosecond crystallography at the european XFEL

    Nature Methods 17:73–78.

    https://doi.org/10.1038/s41592-019-0628-z

    • PubMed
    • Google Scholar
30. 1. Quail PH
    2. Boylan MT
    3. Parks BM
    4. Short TW
    5. Xu Y
    6. Wagner D

    (1995) Phytochromes: photosensory perception and signal transduction

    Science 268:675–680.

    https://doi.org/10.1126/science.7732376

    • PubMed
    • Google Scholar
31. 1. Ren Z
    2. Perman B
    3. Srajer V
    4. Teng TY
    5. Pradervand C
    6. Bourgeois D
    7. Schotte F
    8. Ursby T
    9. Kort R
    10. Wulff M
    11. Moffat K

    (2001) A molecular movie at 1.8 A resolution displays the photocycle of photoactive yellow protein, a eubacterial blue-light receptor, from nanoseconds to seconds

    Biochemistry 40:13788–13801.

    https://doi.org/10.1021/bi0107142

    • PubMed
    • Google Scholar
32. 1. Rockwell NC
    2. Shang L
    3. Martin SS
    4. Lagarias JC

    (2009) Distinct classes of red/far-red photochemistry within the phytochrome superfamily

    PNAS 106:6123–6127.

    https://doi.org/10.1073/pnas.0902370106

    • PubMed
    • Google Scholar
33. 1. Rüdiger W
    2. Thümmler F
    3. Cmiel E
    4. Schneider S

    (1983) Chromophore structure of the physiologically active form (P(fr)) of phytochrome

    PNAS 80:6244–6248.

    https://doi.org/10.1073/pnas.80.20.6244

    • PubMed
    • Google Scholar
34. 1. Sanchez JC
    2. Carrillo M
    3. Pandey S
    4. Noda M
    5. Aldama L
    6. Feliz D
    7. Claesson E
    8. Wahlgren WY
    9. Tracy G
    10. Duong P
    11. Nugent AC
    12. Field A
    13. Šrajer V
    14. Kupitz C
    15. Iwata S
    16. Nango E
    17. Tanaka R
    18. Tanaka T
    19. Fangjia L
    20. Tono K
    21. Owada S
    22. Westenhoff S
    23. Schmidt M
    24. Stojković EA

    (2019) High-resolution crystal structures of a myxobacterial phytochrome at cryo and room temperatures

    Structural Dynamics 6:054701.

    https://doi.org/10.1063/1.5120527

    • PubMed
    • Google Scholar
35. 1. Schmidt A
    2. Sauthof L
    3. Szczepek M
    4. Lopez MF
    5. Escobar FV
    6. Qureshi BM
    7. Michael N
    8. Buhrke D
    9. Stevens T
    10. Kwiatkowski D
    11. von Stetten D
    12. Mroginski MA
    13. Krauß N
    14. Lamparter T
    15. Hildebrandt P
    16. Scheerer P

    (2018) Structural snapshot of a bacterial phytochrome in its functional intermediate state

    Nature Communications 9:4912.

    https://doi.org/10.1038/s41467-018-07392-7

    • PubMed
    • Google Scholar
36. 1. Song C
    2. Psakis G
    3. Lang C
    4. Mailliet J
    5. Gartner W
    6. Hughes J
    7. Matysik J

    (2011) Two ground state isoforms and a chromophore D-ring photoflip triggering extensive intramolecular changes in a canonical phytochrome

    PNAS 108:3842–3847.

    https://doi.org/10.1073/pnas.1013377108

    • Google Scholar
37. 1. Song C
    2. Psakis G
    3. Kopycki J
    4. Lang C
    5. Matysik J
    6. Hughes J

    (2014) The D-ring, not the A-ring, rotates in Synechococcus OS-B′ Phytochrome

    Journal of Biological Chemistry 289:2552–2562.

    https://doi.org/10.1074/jbc.M113.520031

    • Google Scholar
38. 1. Song C
    2. Mroginski MA
    3. Lang C
    4. Kopycki J
    5. Gärtner W
    6. Matysik J
    7. Hughes J

    (2018) 3d structures of plant phytochrome A as pr and pfr from Solid-State NMR: implications for molecular function

    Frontiers in Plant Science 9:1–15.

    https://doi.org/10.3389/fpls.2018.00498

    • Google Scholar
39. 1. Stojković EA
    2. Toh KC
    3. Alexandre MT
    4. Baclayon M
    5. Moffat K
    6. Kennis JT

    (2014) FTIR spectroscopy revealing Light-Dependent refolding of the conserved tongue region of bacteriophytochrome

    The Journal of Physical Chemistry Letters 5:2512–2515.

    https://doi.org/10.1021/jz501189t

    • PubMed
    • Google Scholar
40. 1. Takala H
    2. Björling A
    3. Berntsson O
    4. Lehtivuori H
    5. Niebling S
    6. Hoernke M
    7. Kosheleva I
    8. Henning R
    9. Menzel A
    10. Ihalainen JA
    11. Westenhoff S

    (2014) Signal amplification and transduction in phytochrome photosensors

    Nature 509:245–248.

    https://doi.org/10.1038/nature13310

    • PubMed
    • Google Scholar
41. 1. Toh KC
    2. Stojkovic EA
    3. van Stokkum IH
    4. Moffat K
    5. Kennis JT

    (2010) Proton-transfer and hydrogen-bond interactions determine fluorescence quantum yield and photochemical efficiency of bacteriophytochrome

    PNAS 107:9170–9175.

    https://doi.org/10.1073/pnas.0911535107

    • PubMed
    • Google Scholar
42. 1. Tono K
    2. Nango E
    3. Sugahara M
    4. Song C
    5. Park J
    6. Tanaka T
    7. Tanaka R
    8. Joti Y
    9. Kameshima T
    10. Ono S
    11. Hatsui T
    12. Mizohata E
    13. Suzuki M
    14. Shimamura T
    15. Tanaka Y
    16. Iwata S
    17. Yabashi M

    (2015) Diverse application platform for hard X-ray diffraction in SACLA (DAPHNIS): application to serial protein crystallography using an X-ray free-electron laser

    Journal of Synchrotron Radiation 22:532–537.

    https://doi.org/10.1107/S1600577515004464

    • PubMed
    • Google Scholar
43. 1. van Thor JJ
    2. Ronayne KL
    3. Towrie M

    (2007) Formation of the early photoproduct lumi-R of cyanobacterial phytochrome cph1 observed by ultrafast mid-infrared spectroscopy

    Journal of the American Chemical Society 129:126–132.

    https://doi.org/10.1021/ja0660709

    • PubMed
    • Google Scholar
44. 1. Wagner JR
    2. Brunzelle JS
    3. Forest KT
    4. Vierstra RD

    (2005) A light-sensing knot revealed by the structure of the chromophore-binding domain of phytochrome

    Nature 438:325–331.

    https://doi.org/10.1038/nature04118

    • PubMed
    • Google Scholar
45. 1. Wagner JR
    2. Zhang J
    3. von Stetten D
    4. Günther M
    5. Murgida DH
    6. Mroginski MA
    7. Walker JM
    8. Forest KT
    9. Hildebrandt P
    10. Vierstra RD

    (2008) Mutational analysis of Deinococcus radiodurans bacteriophytochrome reveals key amino acids necessary for the photochromicity and proton exchange cycle of phytochromes

    Journal of Biological Chemistry 283:12212–12226.

    https://doi.org/10.1074/jbc.M709355200

    • PubMed
    • Google Scholar
46. 1. White TA
    2. Kirian RA
    3. Martin AV
    4. Aquila A
    5. Nass K
    6. Barty A
    7. Chapman HN

    (2012) CrystFEL : a software suite for snapshot serial crystallography

    Journal of Applied Crystallography 45:335–341.

    https://doi.org/10.1107/S0021889812002312

    • Google Scholar
47. 1. Winn MD
    2. Ballard CC
    3. Cowtan KD
    4. Dodson EJ
    5. Emsley P
    6. Evans PR
    7. Keegan RM
    8. Krissinel EB
    9. Leslie AG
    10. McCoy A
    11. McNicholas SJ
    12. Murshudov GN
    13. Pannu NS
    14. Potterton EA
    15. Powell HR
    16. Read RJ
    17. Vagin A
    18. Wilson KS

    (2011) Overview of the CCP4 suite and current developments

    Acta Crystallographica. Section D, Biological Crystallography 67:235–242.

    https://doi.org/10.1107/S0907444910045749

    • PubMed
    • Google Scholar
48. 1. Yang X
    2. Kuk J
    3. Moffat K

    (2008) Crystal structure of Pseudomonas aeruginosa bacteriophytochrome: photoconversion and signal transduction

    PNAS 105:14715–14720.

    https://doi.org/10.1073/pnas.0806718105

    • PubMed
    • Google Scholar
49. 1. Yang X
    2. Ren Z
    3. Kuk J
    4. Moffat K

    (2011) Temperature-scan cryocrystallography reveals reaction intermediates in bacteriophytochrome

    Nature 479:428–432.

    https://doi.org/10.1038/nature10506

    • PubMed
    • Google Scholar
50. 1. Yang Y
    2. Linke M
    3. von Haimberger T
    4. Hahn J
    5. Matute R
    6. González L
    7. Schmieder P
    8. Heyne K

    (2012) Real-time tracking of phytochrome's orientational changes during Pr photoisomerization

    Journal of the American Chemical Society 134:1408–1411.

    https://doi.org/10.1021/ja209413d

    • PubMed
    • Google Scholar

Acceptance summary:

The paper reports an excellent application of free electron laser crystallography to a biologically important problem. Only few such applications, all on different systems, have been published so far. The results reveal the detailed mechanism at picosecond and atomic resolution by which rearrangement of the chromophore in a phytochrome enable plants, bacteria and fungi to sense red light.

Decision letter after peer review:

Thank you for submitting your article "The primary structural photoresponse of phytochrome proteins captured by a femtosecond X-ray laser" for consideration by eLife. Your article has been reviewed by two peer reviewers, and the evaluation has been overseen by a Reviewing Editor and Cynthia Wolberger as the Senior Editor. The reviewers have opted to remain anonymous.

The reviewers have discussed the reviews with one another and the Reviewing Editor has drafted this decision to help you prepare a revised submission.

Summary:

In their manuscript Claesson et al. describe time-resolved crystallographic measurements to visualize picosecond structural changes in the ligand-binding domain of a bacterial phytochrome. Such photosensors allow plants, fungi and bacteria to detect light and adapt their physiological responses. The actual photochemical reaction happens in the ultrafast time regime at an bilin cofactor which later transmits the energy into conformational changes of the protein. It is quite remarkable how much the protein has already changed within a picosecond after a photon has been absorbed.

The only way to structurally resolve such ultrafast reactions is using one of the new X-ray laser facilities. These provide intense X-ray pulses with a few tens of fs in lengths, which in combination with femtosecond pump laser allows to resolve high-resolution structural snapshots at precise time delays after photoactivation. The method is very powerful but demanding and by far not routine, with only very few published examples.

This paper describes the first use of an XFEL in a pump-probe experiment to monitor the photoactivation of a phytochrome, in this case the DrBphP(PAS-GAF) bidomain. The authors show that the B- and C-ring propionates loose contact with the protein, that the pyrrole water disappears, the D-ring rotates anti-clockwise and that the A- and B-rings move downwards within 1 ps of photon absorption. Most of these observations are unexpected.

The phytochrome reaction described here is clearly different from previous XFEL studies and a very valuable addition to our knowledge on how organisms sense and use light. The primary photoresponse in a phytochrome is of considerable interest and the structural details described in the manuscript will find a wide readership.

The manuscript was a pleasure to read and the study is overall well done.

Essential revisions:

1) The authors provide PDB file coordinates etc. for the dark and 1 ps datasets but not for that collected at 10 ps. Why not? As the authors show the relevant difference maps, they must have a structural solution too. This additional information is important and would enhance the scientific value of the paper considerably.

2) The authors describe anticlockwise rotation of the D-ring at 1 ps. They should address the fact that this contradicts the time-resolved IR polarization data that indicated D-ring rotation at 30 ps and that a clockwise rotation had been expected.

3) The authors must comment on the differences between the 1 ps (and 10 ps!) structure/s and that of the final Pfr state (which is not reached in this PAS-GAF construct). For example, the D-ring rotation they show is only partial (in Pfr the D-ring rotates much further), there appear to be no changes in the aromatic side groups that attend ring D (there a major differences in Pfr) and, most importantly, not only are the A- and B-rings positioned similarly in Pr and Pfr, the pyrrole water in seen in both (implying that the differences described in this paper represent a push-button effect – which might of course be correct).

4) Rather sensibly, the authors say they are unsure if the 1 ps structure is still in S1 or if it is a ground state. However, in the supplementary information they say explicitly that they "demonstrate that the D-ring twists in the excited state". Whereas this is quite likely to be true (and has been suggested earlier), they don't demonstrate it.

5) The authors suggest a possible "ultrafast hydrogen or electron transfer" charge redistribution model to explain their observations. It would make more sense to say "proton" rather than "hydrogen".

6) There are additional issues surrounding the fluence rates used. Strong scattering might complicate things, but that is just a problem the authors have with their methods. The difference signals they report continue to increase between the two highest fluences used, implying that – even in microcrystals – there will be significant light gradients within each crystal. Is this a problem? On the other hand, the fluence rates are likely to be very high despite the scattering, whereby the classical problem of multi-photon events arises. This reviewer does not think that the statement that "lowering the excitation density resulted in a joint reduction of all difference signals" is entirely correct. Given that the authors plainly state that "The overall aim of the refinement was to maximise the agreement between the observed and calculated difference maps", they should discuss the distinct possibility of artefacts more thoroughly.

7) The laser fluence is a critical parameter not only for conducting but also for evaluating the results of ultrafast time-resolved crystallographic experiments. The authors are very aware of this and have attributed much of the supplementary information and some of the available beamtime to this issue. However, much of their discussion is based on calculations and assumption and the reviewer could not find the original values used during the experiments. Please add the used laser pulse energy, spot size and pulse duration to the Materials and methods section describing fluence calculations, so the interested reader can follow the argumentation better.

8) The missing parameters aside, the stated value of 0.65 photons per molecule is based on many assumptions including a correction from displacement of the laser (only mentioned in Figure 2) and a factor of 100 through scattering from the grease (Figure 1—figure supplement 1 states "at least a two orders of magnitude… depending on the mixing procedure"). In truth, we do not know how many photons per molecule reach the chromophore and this can only ever be an average in crystals anyway. The reviewer suggests that the authors replace the value of 0.65 photons per molecule by a range to better reflect this uncertainty. The authors should also think about replacing the statement "the excitation density was in the one photon regime" used in the main text and Materials and methods section with an expression that is less matter of fact. Importantly this reviewer does not think that this uncertainty does devaluate the biological interpretation of the data since critical features including photodissociation of the pyrrol water occur at much lower laser fluence. The higher photon densities just lead to better maps by increasing activation levels.

9) In a related issue, one reviewer finds it strange that the difference maps for the 10 ps data are much weaker compared to the ones at 1 ps despite better statistics in terms of multiplicity and to some extent also resolution based on the crystallographic Table 1. There is no reason given in the manuscript for the lower intensity of the difference map features. Is the initial photoactivated state just decaying that rapidly (possible but my guess is no) or were different pump-probe settings used for the 10 ps data? Please include any changes in the experimental settings between the two delays in the manuscript.

10) The 1 ps difference map is of high quality even though the authors state low quantum yields for the photoactivation of phytochromes in the order of 10%. Capturing low activation levels reliably is an important advantage of the technique and this study demonstrates this nicely. However, it would be interesting to have an estimate for the achieved levels of activation within the crystal to see how close they get to the stated quantum yields.

Essential revisions:

1) The authors provide PDB file coordinates etc. for the dark and 1 ps datasets but not for that collected at 10 ps. Why not? As the authors show the relevant difference maps, they must have a structural solution too. This additional information is important and would enhance the scientific value of the paper considerably.

Unfortunately, we were not successful in generating a satisfactory structural model against the data at 10ps. This structural model would not provide significant scientific insight beyond what we state in the paper, i.e. that the structural changes are similar at 1ps and 10ps. However, we demonstrate this statement more directly with Figure 1—figure supplement 5, which compares the difference densities at 1ps and 10 ps in detail.

The reason for the low reliability of a structural model at 10 ps is the following: The difference maps for 1ps and 10ps show similar features, but the signal strength is lower at 10ps, indicating that the population is lower at 10ps compared to 1ps (see also comment 9). A general restraint in time-resolved crystallography is that the measured amplitudes of the structure factor after laser excitation contain a mixture of molecules in the resting and activated state. The F_o-F_o difference maps presented in the paper are the most unbiased presentation of the structural changes between the resting and excited state. Refining a structure requires the reconstruction of the structure factor of the "pure" activated state, i.e. by removing the contribution of the resting state molecules to the data. This can be done, but introduces uncertainty because the photoactivation yield has to be estimated. We are confident that we got this right for the 1ps structures (8% yield corresponding to α = 25, see below and Materials and methods section), but with the 10ps structure the uncertainty would be higher, since the population of the photoactivated state is even lower. Given an expected low reliability of the model, refining a structural model would not provide a significant scientific insight beyond the finding that the structural changes are similar at 1ps and 10ps.

We have written a more elaborate justification in the revised manuscript, in the first paragraph of the Discussion, and in the Materials and methods section. We have also strengthened the statement that the structural changes at 1ps are retained at 10ps in the main text and added Figure 1—figure supplement 5.

2) The authors describe anticlockwise rotation of the D-ring at 1 ps. They should address the fact that this contradicts the time-resolved IR polarization data that indicated D-ring rotation at 30 ps and that a clockwise rotation had been expected.

We thank the reviewers for pointing this out! In the revised version, we now discuss the direction of the rotation more extensively, in particular with respect to Karsten Heyne’s IR work and the CD spectra of Rockwell and Lagarias. The direction of motion may be influenced by the truncated PAS-GAF construct that we use (PHY domain missing) and by crystal packing. To the best of our knowledge, the direction of rotation has not been observed directly but it has been inferred from steric arguments and from CD/IR measurements that indicate the end states. Unfortunately, a structure of a plant phytochrome or cyanobacterial phytochrome in the light state is not available. This is now discussed in the Discussion.

3) The authors must comment on the differences between the 1 ps (and 10 ps!) structure/s and that of the final Pfr state (which is not reached in this PAS-GAF construct). For example, the D-ring rotation they show is only partial (in Pfr the D-ring rotates much further), there appear to be no changes in the aromatic side groups that attend ring D (there a major differences in Pfr) and, most importantly, not only are the A- and B-rings positioned similarly in Pr and Pfr, the pyrrole water in seen in both (implying that the differences described in this paper represent a push-button effect – which might of course be correct).

We thank the reviewers for this comment. We have inserted a paragraph in the Discussion; it provides interesting new thoughts to the paper.

4) Rather sensibly, the authors say they are unsure if the 1 ps structure is still in S1 or if it is a ground state. However, in the supplementary information they say explicitly that they "demonstrate that the D-ring twists in the excited state". Whereas this is quite likely to be true (and has been suggested earlier), they don't demonstrate it.

We thank the reviewers for pointing out this inconsistency, and have corrected the sentence in the supplementary information accordingly.

5) The authors suggest a possible "ultrafast hydrogen or electron transfer" charge redistribution model to explain their observations. It would make more sense to say "proton" rather than "hydrogen".

We agree and have changed the word to 'proton'.

6) There are additional issues surrounding the fluence rates used. Strong scattering might complicate things, but that is just a problem the authors have with their methods. The difference signals they report continue to increase between the two highest fluences used, implying that – even in microcrystals – there will be significant light gradients within each crystal. Is this a problem? On the other hand, the fluence rates are likely to be very high despite the scattering, whereby the classical problem of multi-photon events arises. This reviewer does not think that the statement that "lowering the excitation density resulted in a joint reduction of all difference signals" is entirely correct. Given that the authors plainly state that "The overall aim of the refinement was to maximise the agreement between the observed and calculated difference maps", they should discuss the distinct possibility of artefacts more thoroughly.

We thank the reviewers for the comments on photon fluence. We address them below in response to reviewers’ comment 8.

7) The laser fluence is a critical parameter not only for conducting but also for evaluating the results of ultrafast time-resolved crystallographic experiments. The authors are very aware of this and have attributed much of the supplementary information and some of the available beamtime to this issue. However, much of their discussion is based on calculations and assumption and the reviewer could not find the original values used during the experiments. Please add the used laser pulse energy, spot size and pulse duration to the Materials and methods section describing fluence calculations, so the interested reader can follow the argumentation better.

We fully agree and have added the original parameters in the Materials and methods.

8) The missing parameters aside, the stated value of 0.65 photons per molecule is based on many assumptions including a correction from displacement of the laser (only mentioned in Figure 2) and a factor of 100 through scattering from the grease (Figure 1—figure supplement 1 states "at least a two orders of magnitude… depending on the mixing procedure"). In truth, we do not know how many photons per molecule reach the chromophore and this can only ever be an average in crystals anyway. The reviewer suggests that the authors replace the value of 0.65 photons per molecule by a range to better reflect this uncertainty. The authors should also think about replacing the statement "the excitation density was in the one photon regime" used in the main text and Materials and methods section with an expression that is less matter of fact. Importantly this reviewer does not think that this uncertainty does devaluate the biological interpretation of the data since critical features including photodissociation of the pyrrol water occur at much lower laser fluence. The higher photon densities just lead to better maps by increasing activation levels.

Indeed, there is likely significant inhomogeneity in the photoexcitation of the crystals. We have therefore changed the number to a sensible range: 0.5 – 1 photons/molecule. We have also added a more detailed explanation of how we arrive at the number/range, stating the original parameters and the factors that we believe should be considered when evaluating photon fluence in grease jets.

We agree to the reviewers’ view that the statement about the one-photon regime was too categorical. We have therefore changed it to “Critical signals, like the twist of the D-ring and the photodissociation of the pyrrole water sustained when lowering the excitation densities, indicating that the signal arises predominantly from one-photon excitation.”.

A related change in the manuscript is that we now state in the legend of Figure 1—figure supplement 1 that we tried to reproduce the sample preparation conditions at the XFEL as closely as possible when performing the absorption/scattering measurement of the grease/buffer/crystal mixtures.

9) In a related issue, one reviewer finds it strange that the difference maps for the 10 ps data are much weaker compared to the ones at 1 ps despite better statistics in terms of multiplicity and to some extent also resolution based on the crystallographic Table 1. There is no reason given in the manuscript for the lower intensity of the difference map features. Is the initial photoactivated state just decaying that rapidly (possible but my guess is no) or were different pump-probe settings used for the 10 ps data? Please include any changes in the experimental settings between the two delays in the manuscript.

As the reviewers correctly state, the data statistics are better for 10ps compared to 1ps, but the difference signals are weaker at 10ps. This indicates that the population of the activated state is lower. We make this statement more clearly in the document. The lower population could be due to (a) decay of the activated states in the crystals and/or (b) annihilation between neighbouring activated states. While it would be interesting to find out the reason, we cannot provide a definite answer based on our data.

We estimated the noise levels of the difference maps by reducing the contour levels of our difference maps and by identifying the level at which spurious, noisy difference features appeared throughout the structure. We defined this as the background levels in our difference maps. The levels for 1ps are σ=3.0 (0.0095e/ų) and for 10ps σ=3.4 (0.0094e/ų) (Figure 1—figure supplement 4). Thus, the absolute electron density background level is lower for 10ps compared to 1ps, in agreement with the better statistics for 10ps compared to 1ps.

There were not any changes in the experimental settings, but a drift of the pump laser can never be fully excluded. This would lead to a change in spatial overlap between X-ray and optical laser and it would cause a difference in photoactivation yield.

We state now clearly that the photoactivation yield was lower at 10ps. We also state that the settings for 1ps and 10ps were nominally the same.

10) The 1 ps difference map is of high quality even though the authors state low quantum yields for the photoactivation of phytochromes in the order of 10%. Capturing low activation levels reliably is an important advantage of the technique and this study demonstrates this nicely. However, it would be interesting to have an estimate for the achieved levels of activation within the crystal to see how close they get to the stated quantum yields.

We estimated the quantum yield of photoactivation in the crystals to be 8%, corresponding to a factor α = 25, which was used to compute the extrapolated structure factor for the structure refinement. This should be considered to be the ratio between activated and non-activated phytochrome molecules, averaged over the entire crystal. We have now included information on how we determined the photoexcitation quantum yield in the supplementary information. The 8% corresponds to values which are typical in time-resolved spectroscopy experiments on femtosecond time scales.

Author details

1. Elin Claesson

   Department of Chemistry and Molecular Biology, University of Gothenburg, Gothenburg, Sweden

   Contribution

   Planned experiments, Purified the protein and prepared the microcrystals, Acquired the data at SACLA, Computed difference electron density maps and refined the structural models, Wrote the paper with input from all authors

   Contributed equally with

   Weixiao Yuan Wahlgren and Heikki Takala

   Competing interests

   No competing interests declared

2. Weixiao Yuan Wahlgren

   Department of Chemistry and Molecular Biology, University of Gothenburg, Gothenburg, Sweden

   Contribution

   Purified the protein and prepared the microcrystals, Acquired the data at SACLA, Computed difference electron density maps and refined the structural models

   Contributed equally with

   Elin Claesson and Heikki Takala

   Competing interests

   No competing interests declared

   "This ORCID iD identifies the author of this article:" 0000-0003-0413-165X

3. Heikki Takala

   1. Department of Biological and Environmental Science, Nanoscience Center, University of Jyvaskyla, Jyvaskyla, Finland
   2. Department of Anatomy, Faculty of Medicine, University of Helsinki, Helsinki, Finland

   Contribution

   Purified the protein and prepared the microcrystals, Acquired the data at SACLA, Computed difference electron density maps and refined the structural models

   Contributed equally with

   Elin Claesson and Weixiao Yuan Wahlgren

   Competing interests

   No competing interests declared

   "This ORCID iD identifies the author of this article:" 0000-0003-2518-8583

4. Suraj Pandey

   Physics Department, University of Wisconsin-Milwaukee, Milwaukee, United States

   Contribution

   Acquired the data at SACLA and processed the crystallography data

   Competing interests

   No competing interests declared

5. Leticia Castillon

   Department of Chemistry and Molecular Biology, University of Gothenburg, Gothenburg, Sweden

   Contribution

   Purified the protein and prepared the microcrystals, Acquired the data at SACLA

   Competing interests

   No competing interests declared

6. Valentyna Kuznetsova

   Department of Biological and Environmental Science, Nanoscience Center, University of Jyvaskyla, Jyvaskyla, Finland

   Contribution

   Acquired the data at SACLA, Measured the UV-Vis spectroscopic data

   Competing interests

   No competing interests declared

7. Léocadie Henry

   Department of Chemistry and Molecular Biology, University of Gothenburg, Gothenburg, Sweden

   Contribution

   Acquired the data at SACLA

   Competing interests

   No competing interests declared

8. Matthijs Panman

   Department of Chemistry and Molecular Biology, University of Gothenburg, Gothenburg, Sweden

   Contribution

   Acquired the data at SACLA

   Competing interests

   No competing interests declared

   "This ORCID iD identifies the author of this article:" 0000-0003-3853-123X

9. Melissa Carrillo

   Department of Biology, Northeastern Illinois University, Chicago, United States

   Contribution

   Acquired the data at SACLA

   Competing interests

   No competing interests declared

10. Joachim Kübel

    Department of Chemistry and Molecular Biology, University of Gothenburg, Gothenburg, Sweden

    Contribution

    Acquired the data at SACLA

    Competing interests

    No competing interests declared

11. Rahul Nanekar

    Department of Biological and Environmental Science, Nanoscience Center, University of Jyvaskyla, Jyvaskyla, Finland

    Contribution

    Acquired the data at SACLA

    Competing interests

    No competing interests declared

12. Linnéa Isaksson

    Department of Chemistry and Molecular Biology, University of Gothenburg, Gothenburg, Sweden

    Contribution

    Acquired the data at SACLA

    Competing interests

    No competing interests declared

13. Amke Nimmrich

    Department of Chemistry and Molecular Biology, University of Gothenburg, Gothenburg, Sweden

    Contribution

    Acquired the data at SACLA

    Competing interests

    No competing interests declared

14. Andrea Cellini

    Department of Chemistry and Molecular Biology, University of Gothenburg, Gothenburg, Sweden

    Contribution

    Acquired the data at SACLA

    Competing interests

    No competing interests declared

15. Dmitry Morozov

    Department of Chemistry, Nanoscience Center, University of Jyvaskyla, Jyvaskyla, Finland

    Contribution

    Provided the geometry restrains biliverdin in the excited state

    Competing interests

    No competing interests declared

16. Michał Maj

    Department of Chemistry and Molecular Biology, University of Gothenburg, Gothenburg, Sweden

    Contribution

    Acquired the data at SACLA

    Competing interests

    No competing interests declared

17. Moona Kurttila

    Department of Biological and Environmental Science, Nanoscience Center, University of Jyvaskyla, Jyvaskyla, Finland

    Contribution

    Measured the UV-Vis spectroscopic data

    Competing interests

    No competing interests declared

18. Robert Bosman

    Department of Chemistry and Molecular Biology, University of Gothenburg, Gothenburg, Sweden

    Contribution

    Acquired the data at SACLA

    Competing interests

    No competing interests declared

19. Eriko Nango

    1. Department of Cell Biology, Graduate School of Medicine, Kyoto University, Kyoto, Japan
    2. RIKEN SPring-8 Center, Hyogo, Japan

    Contribution

    Acquired the data at SACLA

    Competing interests

    No competing interests declared

20. Rie Tanaka

    1. Department of Cell Biology, Graduate School of Medicine, Kyoto University, Kyoto, Japan
    2. RIKEN SPring-8 Center, Hyogo, Japan

    Contribution

    Acquired the data at SACLA

    Competing interests

    No competing interests declared

21. Tomoyuki Tanaka

    1. Department of Cell Biology, Graduate School of Medicine, Kyoto University, Kyoto, Japan
    2. RIKEN SPring-8 Center, Hyogo, Japan

    Contribution

    Acquired the data at SACLA

    Competing interests

    No competing interests declared

22. Luo Fangjia

    1. Department of Cell Biology, Graduate School of Medicine, Kyoto University, Kyoto, Japan
    2. RIKEN SPring-8 Center, Hyogo, Japan

    Contribution

    Acquired the data at SACLA

    Competing interests

    No competing interests declared

23. So Iwata

    1. Department of Cell Biology, Graduate School of Medicine, Kyoto University, Kyoto, Japan
    2. RIKEN SPring-8 Center, Hyogo, Japan

    Contribution

    Acquired the data at SACLA

    Competing interests

    No competing interests declared

24. Shigeki Owada

    1. RIKEN SPring-8 Center, Hyogo, Japan
    2. Japan Synchrotron Radiation Research Institute, Hyogo, Japan

    Contribution

    Acquired the data at SACLA

    Competing interests

    No competing interests declared

25. Keith Moffat

    Department of Biochemistry and Molecular Biology and Institute for Biophysical Dynamics, University of Chicago, Chicago, United States

    Contribution

    Conceived the experiments

    Competing interests

    No competing interests declared

26. Gerrit Groenhof

    Department of Chemistry, Nanoscience Center, University of Jyvaskyla, Jyvaskyla, Finland

    Contribution

    Provided the geometry restrains biliverdin in the excited state

    Competing interests

    No competing interests declared

27. Emina A Stojković

    Department of Biology, Northeastern Illinois University, Chicago, United States

    Contribution

    Acquired the data at SACLA

    Competing interests

    No competing interests declared

28. Janne A Ihalainen

    Department of Biological and Environmental Science, Nanoscience Center, University of Jyvaskyla, Jyvaskyla, Finland

    Contribution

    Conceived the experiments, Acquired the data at SACLA

    Competing interests

    No competing interests declared

    "This ORCID iD identifies the author of this article:" 0000-0002-8741-1587

29. Marius Schmidt

    Physics Department, University of Wisconsin-Milwaukee, Milwaukee, United States

    Contribution

    Conceived the experiments, Acquired the data at SACLA, Computed difference electron density maps and refined the structural models, Wrote the paper with input from all authors

    For correspondence

    smarius@uwm.edu

    Competing interests

    No competing interests declared

30. Sebastian Westenhoff

    Department of Chemistry and Molecular Biology, University of Gothenburg, Gothenburg, Sweden

    Contribution

    Conceived the experiments, Planned experiments, Acquired the data at SACLA, Computed difference electron density maps and refined the structural models, Wrote the paper with input from all authors

    For correspondence

    sebastian.westenhoff.2@gu.se

    Competing interests

    No competing interests declared

    "This ORCID iD identifies the author of this article:" 0000-0002-6961-8015

• 3,169

  Page views

• 393

  Downloads

• 56

  Citations

Article citation count generated by polling the highest count across the following sources: Crossref, Scopus, PubMed Central.