Active Site investigations on the Fe-centred NHase from Comamonas testosteroni Ni1

Analyzing the catalytic role of active site residues in the Fe-type nitrile hydratase from Comamonas testosteroni Ni1

Salette Martinez, Rui Wu, Karoline Krzywda, Veronika Opalka, Hei Chan, Dali Liu , Richard C. Holz

JBIC Journal of Biological Inorganic Chemistry (2015), 20/5, 885-894

A strictly conserved active site arginine residue (αR157) and two histidine residues (αH80 and αH81) located near the active site of the Fe-type nitrile hydratase from Comamonas testosteroni Ni1 (CtNHase), were mutated. These mutant enzymes were examined for their ability to bind iron and hydrate acrylonitrile. For the αR157A mutant, the residual activity (k _cat = 10 ± 2 s⁻¹) accounts for less than 1 % of the wild-type activity (k _cat = 1100 ± 30 s⁻¹) while the K _m value is nearly unchanged at 205 ± 10 mM. On the other hand, mutation of the active site pocket αH80 and αH81 residues to alanine resulted in enzymes with k _cat values of 220 ± 40 and 77 ± 13 s⁻¹, respectively, and K _m values of 187 ± 11 and 179 ± 18 mM. The double mutant (αH80A/αH81A) was also prepared and provided an enzyme with a k _cat value of 132 ± 3 s⁻¹ and a K _m value of 213 ± 61 mM. These data indicate that all three residues are catalytically important, but not essential. X-ray crystal structures of the αH80A/αH81A, αH80W/αH81W, and αR157A mutant CtNHase enzymes were solved to 2.0, 2.8, and 2.5 Å resolutions, respectively. In each mutant enzyme, hydrogen-bonding interactions crucial for the catalytic function of the αCys¹⁰⁴-SOH ligand are disrupted. Disruption of these hydrogen bonding interactions likely alters the nucleophilicity of the sulfenic acid oxygen and the Lewis acidity of the active site Fe(III) ion.

3D content at http://proteopedia.org/w/Journal:JBIC:32

A proposed catalytic mechanism for NHase via a cyclic intermediate assessed by QM/MM

Catalytic Mechanism of Nitrile Hydratase Subsequent to Cyclic Intermediate Formation: A QM/MM Study

Megumi Kayanuma, Mitsuo Shoji, Masafumi Yohda, Masafumi Odaka, and Yasuteru Shigeta

J. Phys. Chem. B, Article ASAP

DOI: 10.1021/acs.jpcb.5b11363

The catalytic mechanism of an Fe-containing nitrile hydratase (NHase) subsequent to the formation of a cyclic intermediate was investigated using a hybrid quantum mechanics/molecular mechanics (QM/MM) method. We identified the following mechanism: (i) proton transfer from βTyr72 to the substrate via αSer113, and cleavage of the S–O bond of αCys114–SO^– and formation of a disulfide bond between αCys109 and αCys114; (ii) direct attack of a water molecule on the sulfur atom of αCys114, which resulted in the generation of both an imidic acid and a renewed sulfenic cysteine; and (iii) isomerization of the imidic acid to the amide. In addition, to clarify the role of βArg56K, which is one of the essential amino residues in the enzyme, we analyzed a βR56K mutant in which βArg56 was replaced by Lys. The results suggest that βArg56 is necessary for the formation of disulfide intermediate by stabilizing the cleavage of the S–O bond via a hydrogen bond with the oxygen atom of αCys114–SO^–.

Nitrilase and nitrile hydratase from Pseudomonas sp. UW4

Having just finished a project where we looked at a range of nitrilases and what their preferred substrates are, it is always interesting to ponder what the bacterium actually wanted the enzyme for (as compared to the host of xenobiotics you threw at it). We have often had the situation where we have an enzyme which we reckon ought to be active but doesn't seem interested in any of the forty or so compounds we have in our simple screen.
There is a recent paper in Applied and Environmental Microbiology by Duca, Rose and Glick which is concerned with investigating the biosynthesis of indoleacetic acid (IAA), which is a plant growth hormone. This compound comes from indoleacetonitrile (IAN) and there are two obvious pathways to get from there to IAA- via the NHase and via a nitrilase. These workers cloned both enzymes in to E. coli, and then looked at their level of interest in IAN. Interestingly the nitrilase had a habit of producing a proportion of amide as well as the usual acid. Also of interest is that the enzymes have different pH and temperature optima (Nase like 50 degrees C and pH6, the iron-centred NHase likes 4 degrees C and pH7.5), though I wonder if the lower temperature for the NHase is due to the fundamental lack of stability of iron NHases rather than an adaption. Additionally, the authors use some bioinformatics to confirm their experimental findings that this is an aromatic nitrile active system.

A nitrile hydratase for cyanopyridines... (but it prefers aliphatic nitriles)

There has been a recent paper in Journal of Molecular Catalysis B on a nitrilase that converted cyanopyridines. This enzyme came from a strain of Pseudomonas putida. There is also a recent paper in the same journal entitled “Discovery of a new Fe-type nitrile hydratase efficiently hydrating aliphatic and aromatic nitriles by genome mining” by Xiaolin Peia, Lirong Yang, Gang Xu, Qiuyan Wang and Jianping Wu.which describes a nitrile hydratase, this time, from a Pseudomonas putida strain (F1) which they have shown to be able to turn over 3-cyanopyridine in a 1L fed batch reactor. Their activity data suggests it actually prefers aliphatic nitriles: acrylonitrile rates as 941 U/mg and valeronitrile as 535 U/mg as compared to 3-cyanopyridine at 26 U/mg. They describe how it was cloned into E. coli, needing the inclusion of an activator protein to get activity. They also include a nice phylogenetic tree showing the spread of known iron type NHases... plenty of examples in the Rhodococcus but spreading outwards into Pseudomonas.

S-M bond lengths

I was wondering what a "normal" bond length might be for a thiol to iron or cobalt might be, and then what might happen to that bond length might do if that bond was oxidized once and then twice. The best way to do this is to look at the small molecule crystallographic data so I got my friend Ross to comb through the usual database to find examples of the various bonding motifs.

It turns out that there are no examples of 1b in the CCD, and precious few of 1c (only 4), 2b (only 5) and 2c (a more respectable 15). There are loads of examples of 1d (actually 286) and 2d (102).
Anyway, there isnt much of a difference in average bond lengths between these types: 1c- 2.221 angstroms, 1d- 2.276 angstroms, 2b- 2.225 angstroms, 2c- 2.208 angstroms and 2d- 2.252 angstroms.
Here are a few histograms showing the distributions that lie behind these means.

Looking for the iron in AJ270

 Using PyMol I have been looking to see what the tunnel which connects the iron centre to the outside world looks like. Here are some pictures. Iron is represented as a warmpink sphere, and the pdb structure used is 2QDY.

Or close up as...

Common features of cobalt and iron based NHases

With a couple of local colleagues, I have a paper in press with the International Journal of Data Mining and Bioinformatics. It is the eventual result of Joe, a mathematician asking if I had any high dimensional data sets that needed interrogating. In terms of this blog, it has a section looking at what are the main primary sequence features (using a set of one hundred NHases) which are important in distinguishing cobalt centred NHases from iron centred ones. Thanks to Marcus for all the 3D structure analysis, and spotting the hinge/loop differences first suggested as relevant to differential stability by Miyanaga in 2001

http://northumbria.academia.edu/JoeFaith/Papers/320396/Predicting_Functional_Residues_of_Protein_Sequence_Alignments_as_a_Feature_Selection_Task