Dear all,
I have less experience in this field. My Protein (200 aminoacid residues) is
stable in high salt buffer(1M KCl); however after extensive optimization by
salting-in (in 4-5 weeks time), am
getting very thin and improved plate-like single crystals with a size of 50 X
50 X (don't ask me about thickness). 25% PEG8k as cryo works well. Anyways, the
best crystal diffracted synchrotron X-rays upto 2.0 Å. Detected P21 by
pointless with unit cell a= 30.41 b=46.98 C=58.43 β=100.76 α=γ=90. Matthews
says 33% solvent/1 molecule in au; but there's a weak pseudo-translation of
12.54% (SFCHECK); no twinning (in another native dataset 30.44 47.01 57.64 90
100.2 90, there's no pseudo-translation).
Performed MR by PHASER with several (using FFAS03 server, possible distant)
homologs with 17-22% sequence identity.
1. Best initial search model(~60 aa residues)gave a PHASER output - RFZ=6.6
TFZ=4.0 PAK=0 LLG=30 & REFMAC rigid body R/Rfree : 52%/53%
2. Improved model(BUCCANEER fragmented/manually modelled/~110
residues/Origin-shifted) gave a PHASER output - RFZ=14.5 TFZ=11.0 PAK=0 LLG=266
LLG=266 & REFMAC
rigid body R/Rfree of 43%/44%; got stuck here!
Also in parallel did some SAD/MAD/SIR trials. Most of my heavy atom soaked
(based on native gel-shift assays) crystals cracked as they are really
thin-skinned even at lower concentrations (0.5mM -1mM)of the compounds.
Native: P21 30.41 46.98 58.43 90.00 100.76 90.00
(λ=1.4Å, SCALA-ed upto 2Å; overall - Rmerge=8.6% , Mul=2.7,
Mean(I/sd(I))=5.9, comp=97.8%, pseudo-translation vector 0.000 0.114 0.000)
Der 1: P21 30.36 46.84 58.38 90.00 100.92 90.00 (IODINE)
(λ=1.4Å, SCALA-ed upto 3.5Å; overall - Rmerge=6.2% , Mul=3.5,
Mean(I/sd(I))=10.8, comp=99.8%, Anom Mul=1.9, Anom comp=90.9%,
pseudo-translation vector 0.313 0.000 0.311)
2 more datasets from iodine soaks are non-isomorphic with < 3Å changes in the
unit cell with a very weak Anomalous signal. SHELXD detects 11 sites (ori/inv)
with decreasing occupancy in ascending order. Unable to detect iodine (bound,
if any) sites from SAD/SIR(AS) in autoSHARP; but however the density modified
map appears to be showing domain boundaries in which I am unable to interpret
secondary structural feautures.
Der 2: upon HgCl2 soaking one of the crystals turned out to be non-isomorphic
C2221,
confirmed with Pointless. Eventhough anomalous signal is very very
weak, still collected 3W MAD datasets on that crystal.
(seperately SCALA-ed the sets upto 3.5Å)
peak : C2221 61.15 171.69 47.12 90.00 90.00 90.00
(λ=1.006Å, Overall - Rmerge=12.3%, Mul=12.8, comp=100%, Anom
Mul=7.1, Anom comp=100%)
rem : C2221 61.34 172.11 47.26 90.00 90.00 90.00
(λ=0.989Å, Overall - Rmerge=12.7%, Mul=6.3, comp=100%, Anom
Mul=3.4, Anom comp=100%)
inf : C2221 61.11 171.52 47.07 90.00 90.00 90.00
(λ=1.007Å, Overall - Rmerge=17.4%, Mul=11.6, comp=100%, Anom
Mul=6.8, Anom comp=100%)
Matthews says 56% Solvent/1 molecule in au; yet there's again
pseudo-translation of 47.4% (0.166 0.166 0.000 - SFCHECK);no twinning. SHELXD/E
detects 6 sites (ori/inv) with decreasing occupancy in
ascending order. Tried autoSHARP, no Hg (bound, if any) sites were detected.
My questions are:
1) Is it possible that P21 transitioned to C2221, or is it a different crystal
form?
2) In both P21 and C2221, there's Pseudo translation. What are the chances for
more than 1 molecule in the au? (down below also added MOLREP RF peaks)
3) I think my MR approach is not the best, but is it a right path?
4) Is there any specific way to screen fragile crystals for heavy atom phasing?
5) Is it possible that the SHELXD detected sites are for weak anomalous
scatterers like sulphur by any chance (as my protein has 4 Cys & 3 Met)?
6) If Sulphur, is it fruitful to combine 2 to 3 P21 Datasets (collected at same
λ; but composite crystals with small changes in the unit cell) or C2221
datasets (single crystal) before
scaling in order to increase the redundancy for Sulphur SAD phasing?
7) Would also like to incorporate MR model (to try) with sulphur-SAD, is it a
tangible approach?
I hope am not asking too much. Your criticism, suggestions, comments,
references and answers will greatly help me.
Thank you,
Raspudin
P21 (native dataset) RF peaks,
theta phi chi alpha beta gamma Rf Rf/sigma
Sol_RF 1 0.00 0.00 0.00 0.00 0.00 0.00 9.196 10.78
Sol_RF 2 108.16 0.00 180.00 180.00 143.68 0.00 7.395 8.67
Sol_RF 3 106.86 11.92 179.99 11.92 -146.28 168.08 3.221 3.77
Sol_RF 4 176.50 0.00 180.00 0.00 -7.00 -180.00 2.816 3.30
Sol_RF 5 79.28 -128.59 179.72 50.66 -158.56 127.83 2.778 3.25
Sol_RF 6 0.00 -152.81 180.00 0.00 0.00 180.00 2.724 3.19
Sol_RF 7 105.47 32.14 179.65 32.79 -149.05 148.51 2.719 3.19
Sol_RF 8 104.31 41.01 179.98 41.01 -151.37 138.99 2.644 3.10
Sol_RF 9 98.58 -119.51 179.99 60.49 -162.85 119.51 2.598 3.04
Sol_RF 10 90.00 90.00 25.02 0.00 25.02 0.00 2.190 2.57
C2221 (peak λ dataset) RF peaks,
theta phi chi alpha beta gamma Rf
Rf/sigma
Sol_RF 1 0.00 0.00 0.00 0.00 0.00 0.00 1.422 12.23
Sol_RF 2 78.98 -90.00 180.00 90.00 157.96 90.00 0.4585 3.95
Sol_RF 3 50.55 -90.00 180.00 90.00 101.10 90.00 0.3412 2.94
Sol_RF 4 90.00 -180.00 90.00 90.00 90.00 -90.00 0.3387 2.91
Sol_RF 5 90.00 0.00 59.74 90.00 -59.74 -90.00 0.2710 2.33
Sol_RF 6 0.00 180.00 29.90 0.00 0.00 29.90 0.2655 2.29
Sol_RF 7 157.03 -90.00 179.99 90.00 45.95 90.00 0.2503 2.15
Sol_RF 8 90.00 -117.92 179.99 0.00 180.00 55.85 0.2502 2.15
Sol_RF 9 90.00 144.20 179.99 0.00 180.00 -108.40 0.1796 1.55
Sol_RF 10 97.47 63.66 179.89 64.09 -165.05 116.77 0.1632 1.40
