April 2002 Research Report

Preparation of new qtable for sda and uhbd (and pka calculations since the first and last residues were previously uncalculated)

The qtable includes parameters for the amino acid residues, GTP and Mg²⁺ in CharmM 19 (protein) and CharmM 27 for the guanine and ATP to produce charges for GTP.

UHBD Results with new qtable.dat

The results of these calculations show that there are distinct positive and negative portions of both proteins, even with GTP/Mg²⁺ bound. There is a negative region over the GTP phosphates. Generally, the docked regions to the nitrogenase structure are complementary; this is shown on my Huenfeld Poster.

UHBD (E=electrostatic energy)

kffhgtp312: E (protein dielectric= 2, solvent dielectric=78.5): 0.179763E+05

E (protein and solvent dielectric=78.5): 0.556415E+03 kcal/mol

total charge on list of atoms: 0.000

kffhCOMgtp312.ins (-2.337310 to 2.302471)

kffhCOMgtp312u.ins (-1.1316584 to 1.586764)

ftsYgtp509: E (protein dielectric= 2, solvent dielectric=78.5): 0.177836+05

E (protein and solvent dielectric=78.5): 0.553785+03 kcal/mol

total charge on list of atoms: -6.000

kffhCOMgtp312.ins (-3.660003 to 2.570318)

kffhCOMgtp312u.ins (-2.509691 to 1.747120)

In light of the journal discussion today, should these electrostatics calculations be redone with protein dielectric= 1 as Simonson suggests?

Presentation at Huenfeld Meeting (Apr. 12-13):

I presented a poster wh9ich is posted on the web page. It contains research results through early April 2002. As I mentioned, Jeremy Smith was worried I was overlapping with his prior Research Proposal. Bert de Groot liked the displayed pictures of the electrostatics calculations.

Some notes to be thinking about in regards to the SRP:receptor complex

Perhaps for the SDA, we should use the non-GTP/Mg²⁺ bound forms of the SRP and receptor as the literature says that Ffh and FtsY act as GTP activating proteins for each other and both have little GTP activity by themselves.

Maybe Jeremy's group can use their Molecular Kinematics and Conjugate Peak Refinement methods once we have a vialbe complex to test.

Some SRP questions that need answered: How does the presence of the gamma phosphate in the SRP GTPase lead to a change in the functional site? What is the conformational flexibility of the nucleotide bound and unbound forms of the SRP and receptor? What are the different functional states in their mode of action? We can use MD to look at the apo-, GDP and GTP/Mg²⁺ bound forms. Are the changes in conformation due to thermal fluxuations or separate regions of conformational space? The I-box specific to SRP GTPases exhibit a change in the conformation of the beta-helices in regards to the nucleotide bound and free forms.

The GTP in Ras is hydrolyzed in an SN₂ mechanism, direct in-line transfer of the GTP gamma-phosphate to water with the inversion of configuration around the phosphate group. For Ras the question remains, which residue is the catalytic base that activates water for the nucleophilic attack? Do the SRP GTPases also act in an SN₂ manner, which residue is the catalyzing base? What is the nature of the transition state? Is is dissociative or associative? (FT-IR experiments have been used to try to answer this question.) We would like to visualize the dynamic coupling of the reaction steps and thw local conformation of the GTP binding site. Looking at the roles of other GAP for GTPases (such as Ras), what are the roles of the SRP:SR acting as GAPs for each other?

Upon GTP hydrolysis, the gamma phosphate depart destabilizes the GTP binding site leading to a conformational change in two regions: the Switch I and Switch II. In the Switch I in Ras, Tyr32 reorients and a modication of the structure around the Mg²⁺ is observed. In the switch II region there is a collapse of the ordered coil to helix transition at the N-terminus and a reorientation of the alpha 2 helix. GTP hydrolysis could be examined explicitly with QM/MM calculations for Ras and the SRP GTPases. Nucleophilic attack in Ras is due to Gln61, this residue corresponds to His85 in EF-Tu.

The M-domain and signal sequence and RNA could be modelled into Ffh (GTP free form). There is a similar structure for SRP54. The protein is stabilized by the signal sequence in the nucleotide free form. The signal sequence binding affinity is higher in the ribosome bound form of the Ffh.

The eukaryotic SRbeta associates with the membrane in the GDP bound or empty states. FtsY associates with anionic phospholipids. There are two lipid binding sites on the NG domain and one on the A domain. Lipid association was found to increase its GTPase activity. It i strongly negatively charged, there are ~50 negative charges at physiological pH. It is electrostatically attracted to the anionic phospholipids, the R/K rich region may be the one to interact with the membrane. To interact with the lipid, a partial unfolding is needed.

The SRP and the signal peptide and the ribosome come to the membrane where the FtsY also diffuses to and associates to the SRP and the membrane. Subsequently, both bind GTP, the signal peptide is released and both hydrolyze their bound GTPs. The 4.5S RNA enhances the GTP activity when Ffh is complexed with FtsY and induces complex formation 400-fold when GTP is already bound. GTP hydrolysis of GTP-Fffh*GTP-FtsY is faster than dissociation of the complex. FtsY exists naturally in the cytosol and the inner membrane. GTP binding by FtsY is essential for compelx formation with Ffh. GTP-dependent SRP:SR complex formation leads to a conformational change i n the I-box of FtsY near Trp343. GTP binding of FtsY, but not hydrolysis, is required to release the SRP from the signal sequence. GTP binding is biphasic, indicating a 2-step binding mechanism. Fluorescence evidence indicates that the GTP binds to FtsY and the SRP and the complex subsequently forms. GTP hydrolysis leads to complex dissociation. The I-box stabilizes the nucleotide free FtsY.

We could look at the MutS structure and superimpose Ffhand FtsY on this for a model and interface (1ewq.pdb- Obmolova MutS, 1e3m.pdb- Lamers MutS). Irmi says that a complex forms when both the SRP and SR are bound GTP and the GTP binding domains are paired and the M and A domains are opposite (C2V relationship like that they proposed).

The SO₄ bound structures should not be considered as true apo forms as it sits in the active site. The P-loops should close in the true apo.

How do I know when I've got a correct model of the complex? The is not enough experimental evidence with SRP GTPases. I could look at all the ATP/GTP binding proteins in complexes (heterodimers and homodimers) that have structures listed in the Protein Databank. The SRP proteins could be superimposed onto these to obtain new complex possible conformations. Electrostatics could then be used to evaluate which is a viable model and this could be used for further calculations. There are 4 GEF/G-protein structures available. The interface of Tiam1 and Rac1 contacts are all in the SI and SII regions.

SDA results with new qtable.dat

I have SDA results for both GTP/Mg²⁺ bound and unbound forms of the SRP and its receptor. The ionic strength was set to 50 mM and the protein dielectric=4. This was used to model possible allowed conformations for the SRP:SR complex. The proteins are modeled into the large Boltzmann factor regions (the small ones are noise). There are many possible conformations, such as the alpha-helical bundle of one GTPase interacting with the GTP binding region of the other. This is highly unlikely considering the proteins act as GAPs for each other.

SDA results with distance constraint

The SRP is ~72 x 40 Angstroms in size. The SR is ~74 x 30 Angstroms in size. I have results for the GTP/Mg²⁺ bound forms of the SRP and its receptor. The first constraint is small, ~15 Angstroms, the default Razif used for Barnase and Barstar. The second constraint is larger, ~25 Angstroms, probably closer to what should be allowed for this large protein pair.

Displaying with VMD- but this does not do surface calculations

Future:

ASC surface calculation- hydrophobic, numerical area (total)

GRASP surface calculation- electrostatic potential or surface, can view results with Swiss pdb Viewer.

Quanta surface calculation- quantitative hydrophobics

view all SDA results with Molsurfer- adsi for electrostatics viewing

ADS surface

FTdock

Zdock

NACCESS- compute solvent accessible areas and the residue types overall

Presentation at Darmstaedt Meeting (May 7-8)