Table 1. Summary of the structural data and core details for the HMBS crystal structures in the PDB as of mid-2021 (also included in this table is the entry in AlphaFoldDB; Jumper et al., 2021 ▸).
| PDB code (publication reference) | Biological source | Crystal parameters† | Title of PDB entry | Colour of crystal | The highest difference Fourier (F o − F c) electron-density peak (viewed in Coot) and any specific comments therefrom‡ | PDB Validation Report assessment (clashscore; my own specific comments of interest based on the PDB report)‡ |
|---|---|---|---|---|---|---|
| 1pda (Louie et al., 1992 ▸, 1996 ▸) | Escherichia coli | 1.76 Å; P21212; 88, 75.9, 50.5 Å; Z = 1 | Structure of porphobilinogen deaminase reveals a flexible multidomain polymerase with a single catalytic site | Yellow§, i.e. inactive form of the enzyme | 10.6σ. The functionally interesting peak is peak 6 at −5.5σ on the second cofactor ring. It is stated that this ring is at 90% occupancy. The cofactor is in the inactive oxidized state. There are ten peaks above ±5.0σ, the Coot default threshold. | Clashscore 6. Significant negative electron-density difference density (3σ) on the DPM314A second ring is shown in the report. |
| 1ah5 (Hädener et al., 1999 ▸) | Escherichia coli | 2.40 Å; P21212; 88.47, 76.15, 50.79 Å; Z = 1 | Reduced form selenomethionine-labelled hydroxymethylbilane synthase determined by MAD | Colourless, i.e. active form of the enzyme | 7.0σ. There are seven peaks above ±5.0σ. The 7.0σ peak is close to the second ring of the cofactor, with no obvious explanation. Peaks 2 and 5 are in the mobile loop region, i.e. Leu61 and Pro40, respectively. | Clashscore 8 |
| 1ypn (Helliwell et al., 1998 ▸) | Escherichia coli | 2.40 Å; P21212; 88.47, 76.15, 50.79 Ŷ; Z = 1 | Reduced form hydroxymethylbilane synthase (K59Q mutant) crystal structure after 2 h in a flow cell determined by time-resolved Laue diffraction | Turned red during the time-resolved diffraction sequence | No F o − F c peaks above 5.0σ. | Clashscore 8 |
| 2ypn (Nieh et al., 1999 ▸) | Escherichia coli | 2.3 Å; P21212; 88.06, 75.73, 50.35 Å; Z = 1 | Hydroxymethylbilane synthase | Colourless, i.e. the active form of the enzyme | 5.4σ. This is the only peak and is located above the bridge between the C1 and C2 cofactor rings. | Clashscore 4 |
| 1gtk (Helliwell et al., 2003 ▸) | Escherichia coli | 1.66 Å; P21212; 87.5, 75.9, 50.1 Å; Z = 1 | Time-resolved and static ensemble structural chemistry of hydroxymethylbilane synthase | Colourless, i.e. the active form of the enzyme | 11.1σ. There are 21 peaks above 5.0σ. These are unmodelled split-occupancy peaks, bound waters and some signs of radiation damage. | Clashscore 11. Significant negative electron-density difference density (3σ) on the DPM315A second ring and positive on the bridge between the two rings. |
| 3ecr (Song et al., 2009 ▸) | Human | 2.18 Å; P212121; 71.61, 81.061, 109.211 Å; Z = 2 | Structure of human porphobilinogen deaminase | −9.6σ at Glu76B (radiation damage?). There are 14 peaks above ±5.0σ. These are misoriented side chains (e.g. Trp201A) and unmodelled waters. | Clashscore 13 | |
| 3eq1 (Gill et al., 2009 ▸) | Human | 2.8 Å; P21212; 81.083, 104.435, 109.732 ņ†; Z = 2 | The crystal structure of human porphobilinogen deaminase at 2.8 A resolution | 7.1σ. 13 peaks above ±5.0σ. These are misoriented side chains (e.g. Gln29B, Trp198B) and unmodelled waters. There is a 5.2σ peak at the side chain of Arg195B and O3 and O4 of DPM401B. There is another at 5.1σ near to Thr145B and DPM401B and likewise at 5.1σ for DPM400A. | Clashscore 10 | |
| 4htg (Roberts et al., 2013 ▸) | Arabidopsis thaliana | 1.45 Å; C2; 141.573, 37.271, 55.069 ņ†; Z = 1 | Porphobilinogen deaminase from Arabidopsis thaliana | 6.1σ. There are six peaks above 5.0σ. These are mismodelled side chains or unplaced waters, but two amino acids, Asp78A (−5.9σ) and Glu32A (−5.5σ), show radiation damage. | Clashscore 6 | |
| 4mlq (Azim et al., 2014 ▸) | Bacillus megaterium | 1.6 Å; P212121; 53.01, 65.12, 96.78 Å; Z = 1 | Crystal structure of Bacillus megaterium porphobilinogen deaminase | 7.4σ. There are just four peaks, but the top three are at DPM401A. See Supplementary Fig. S1. The interpretation of what these mean is unclear. Azim and coworkers model two conformations of the cofactor at 50% and their paper carries an extensive discussion of the cofactor states and colourations (pink as well as yellow and colourless). | Clashscore 2 | |
| 4mlv (Azim et al., 2014 ▸) | Bacillus megaterium | 1.46 Å; P212121; 53.32, 65.78, 97.21 Å; Z = 1 | Crystal structure of Bacillus megaterium porphobilinogen deaminase | 9.0σ. There are 23 peaks above 5σ. These are mainly unmodelled split-occupancy side chains and possible bound waters. Peak 8 next to Val61A is an extended feature (blob); peaks 18 and 23 are similar. | Clashscore 3. The PDB Validation Report ligands and difference density show quite a number of peaks over both DPMs. | |
| 5ov5 (Guo et al., 2017 ▸) | Bacillus megaterium | 1.81 Å; P212121; 49.16, 62.7, 91.83 Å; Z = 1 | Bacillus megaterium porphobilinogen deaminase D82E mutant | 8.2σ. There are 13 peaks above 5.0σ. These are mainly unmodelled bound waters. There is evidence of radiation damage at Glu307A, Glu290A, Met175A and Cys241A. | Clashscore 2 | |
| 5ov6 (Guo et al., 2017 ▸) | Bacillus megaterium | 1.81 Å; P212121; 48.99, 62.69, 91.26 Å; Z = 1 | Bacillus megaterium porphobilinogen deaminase D82N mutant | 9.2σ. Four peaks. All four are unmodelled bound waters. | Clashscore 2 | |
| 5ov4 (Guo et al., 2017 ▸) | Bacillus megaterium | 2.8 Å; P212121; 49.092, 62.487, 91.391 ņ†; Z = 1 | Bacillus megaterium porphobilinogen deaminase D82A mutant | 5.1σ. Just two peaks, which look like side-chain disorders. | Clashscore 3 | |
| 5h6o (T. Funamizu, M. Chen, Y. Tanaka, K. Ishimori & T. Uchida, unpublished work) | Vibrio cholerae | 2.7 Å; I4122; 94.23, 165.34 Å; Z = 1 | Porphobilinogen deaminase from Vibrio cholerae | 6.3σ. Four peaks; three are perhaps bound waters, but being rather close to the protein may be series-termination errors. One is an unmodelled ‘blob’. | Clashscore 7 | |
| 5m6r (Pluta et al., 2018 ▸) | Human | 2.7 Å; P21; 68.91, 81.2, 79.69 Å, 93.01°; Z = 2 | Human porphobilinogen deaminase in complex with reaction intermediate | Colourless crystal. See Supplementary Fig. S5 of Pluta et al. (2018 ▸). | 6.7σ. 15 peaks. These are unmodelled bound waters or indistinct blobs. They may instead be series-termination errors. | Clashscore 1. The two tetrapyrroles are high-quality (2F o − F c) fits. |
| 5m7f (Pluta et al., 2018 ▸) | Human | 2.78 Å; P21; 70.42, 80.95, 76.22 Å, 95.7; Z = 2 | Human porphobilinogen deaminase in complex with DPM cofactor | Colourless crystal. See Supplementary Fig. S4 of Pluta et al. (2018 ▸). | 7.0σ. Seven peaks. The top peak is a fairly extensive blob (Supplementary Fig. S2), as are peaks 3, 4 and 7. The latter two are possibly glimpses of the mobile loop 58A–74A. | Clashscore 2 |
| 7aaj (Bustad et al., 2021 ▸) | Human | 1.8 Å; P212121; 81.163, 84.608, 108.913 ņ†; Z = 2 | Human porphobilinogen deaminase in complex with cofactor | 6.4σ. 17 peaks. The top two peaks form a fairly extended blob, as do peaks 5 and 6. Perhaps glycerol? The remainder are side chains that could be modelled slightly better or possible bound waters. | Clashscore 4. The PDB Validation Report mentions that ‘The analyses of the Patterson function reveals a significant off-origin peak that is 23.61% of the origin peak, indicating pseudo-translational symmetry (tNCS)’. | |
| 7aak (Bustad et al., 2021 ▸)‡‡ | Human | 1.7 Å; P212121; 81.241, 86.127, 107.39 ņ†; Z = 2 | Human porphobilinogen deaminase R173W mutant crystallized in the ES2 intermediate state | 6.5σ. 18 peaks. These are side chains that could be modelled slightly better, or possibly bound waters or split-occupancy glycerol. There is evidence of radiation damage at Cys247A. | Clashscore 3. As with 7aaj, mention of tNCS. The two tetrapyrroles are high-quality (2F o − F c) fits. | |
| 7cd0 (Sato et al., 2021 ▸)‡‡ | Human | 2.31 Å; P212121; 81.416, 81.366, 108.85 ņ†; Z = 2 | Crystal structure of the 2-iodoporphobilinogen-bound ES2 intermediate form of human hydroxymethylbilane synthase | 7.1σ. Six peaks above 5.0σ. One peak is a side chain that could be remodelled. The others are perhaps series-termination peaks; anyway, it is not obvious how to model them. | Clashscore 7. Also mention of tNCS. The two tetrapyrroles are high-quality (2F o − F c) fits. The 2-iodoporphobilinogen (2F o − F c) is a bit broken up on the two, non-iodo, side chains (even though contoured at 0.7 r.m.s.d.) but is OK. | |
| 7ccy (Sato et al., 2021 ▸)§§ | Human | 2.40 Å; P212121; 74.011, 81.237, 109.097 ņ†; Z = 2 | Crystal structure of the 2-iodoporphobilinogen-bound holo form of human hydroxymethylbilane synthase | There are no peaks above 5.0σ. | Clashscore 7. Also mention of tNCS. The 2-iodoporphobilinogen (2F o − F c) here is completely well formed (compared with 7cd0). | |
| 7ccz (Sato et al., 2021 ▸) | Human | 1.79 Å; P212121; 81.454, 79.542, 108.678 ņ†; Z = 2 | Crystal structure of the ES2 intermediate form of human hydroxymethylbilane synthase | 5.9σ. Nine peaks above 5.0σ. Arg22A could be remodelled (peak 4). The other peaks may be bound waters, but most are too close to the protein atoms and if placed would lead to clashes. | Clashscore 4. No tNCS indicated, although the space group and unit cell are very similar to those of 7cd0. The two tetrapyrroles are high-quality (2F o − F c) fits. | |
| 7ccx (Sato et al., 2021 ▸) | Human | 1.84 Å; P212121; 70.411, 80.815, 109.194 Å ††; Z = 2 | Crystal structure of the holo form of human hydroxymethylbilane synthase | 6.55σ. There are six peaks above 5.0σ. These are split-occupancy or unmodelled waters in nonfunctional locations. | Clashscore 4 | |
| AlphaFoldDB (jointly with the EMBL–EBI) (Jumper et al., 2021 ▸; Tunyasuvunakool et al., 2021 ▸) | Escherichia coli | E. coli hydroxymethylbilane synthase |
†
Resolution; space group; unit-cell parameters; Z.
‡
The PDB Validation Report concerns the derived model and not the details that have not been modelled. The F o − F c map is easily inspected in the molecular-graphics visualization system Coot (Emsley et al., 2010 ▸), which was used to describe the unmodelled peaks in this table.
§
A nice picture of a yellow HMBS crystal is shown in Supplementary Fig. 3 of Azim et al. (2013 ▸).
¶
These were set to the unit-cell parameters of PDB entry 1ah5. At 12 h the unit-cell parameters were 87.52, 75.92, 50.12 Å. This was a monochromatic data set to 2.0 Å resolution measured on BM14 at ESRF.
††
These are likely not at a precision of three decimal places.
‡‡
I made calculational checks of PDB entries 7aak and 7cd0, where tNCS was indicated in the PDB Validation Report. I checked their space groups with Zanuda (Lebedev & Isupov, 2014 ▸) and confirmed their correctness.
§§
Raw diffraction-image data for PDB entry 7ccy were held at in the XRDa at Protein Data Bank Japan, which meant that I was able to process with them with iMosflm to a slightly better resolution of 2.2 Å instead of 2.4 Å. I confirmed that the electron-density maps were the same. There were also 23 Gbytes of raw diffraction-image data for PDB entry 7cd0. At the present time, due to the block structure of the files (32 × 800 Mbytes), I was unable to process these diffraction images myself.