Just because it is not in the PDB does not mean that noone has ever
tried it. I think I can attest to seeing pretty much all the
Lanthanides tried at my beamline. Indeed, Hampton sells a Lanthanide
kit, and I recommend using it as every Lanthanide has a slightly
different ionic radius. For example, Holmium appears to be able to
substitute for Mg (Ku, Smith & Howell. 2007).
As for Sc, Ti, Zr, Nb and Th, all these have edges outside the
"normally" useful wavelength range between 2 A and ~0.8 A. I imagine
this probably reduces the frequency of their use, as does the
radioactivity of Th. Nevertheless, until today there was no place to
publish most of the heavy atom soaks I have seen. In fact, I have ~50
TeraBytes of diffraction data to upload to the new JFCE, but I have been
unable to find their submission site. Any help?
As for the other "missing" heavy atoms, I just found that a hafnium
cluster can be seen in 1o7t and a Zr cluster in 1xc1, so perhaps parsing
is to blame for missing those? Also, how can you tell if a particular
metal was used to solve a structure, but the authors only deposited
their native? For example, there was a Dy derivative used to solve
insulin (Blundell & Johnson, 1976), but apparently that atom never made
it into the PDB. Thulium was used to solve 1b79 (Fass et al. 1999), but
that is not apparent from the PDB entry either. Indeed, it is a very
common practice to neglect depositing derivative data and derivative
structures, and the original MAD/MIR data do tend to be lost to history
(which is bad news for the beamline used to collect them).
I will add that I have never seen a Germanium derivative, and I also
think that it would be cool to solve the first structure using
Pepto-Bismol (Bismuth subsalicylate) as the heavy ion reagent. Then
Proctor & Gamble can add "poor electron density" to their list of things
that Pepto can cure.
-James Holton
MAD Scientist
Thomas Womack wrote:
A perusal of the PDB reveals that the game of Periodic Table bingo still
has eleven rounds to run:
scandium, titanium, germanium, zirconium, niobium, neodymium,
dysprosium, thulium, hafnium, bismuth and thorium remain absent from PDB
entries.
OK, many of these are elements that would rather be refractory oxides or
jet-engine components than hexammines, and niobium chloride clusters
don't seem to be as water-stable as Ta6Br14, but why have neodymium,
dysprosium and thulium so consistently been left out there in the cold
rather than admitted to the warmish embrace of carboxyl groups? There
must somewhere be a protein with a site that cries out for ThCl2(2+), an
unexpectedly water-stable cation.
Tom
