SSGCID
Seattle Structural Genomics Center for Infectious Disease

FUN_13

Characterization and elaboration of fragment compounds against Ebola matrix proteins.

THE PROJECT PROPOSAL IS PRESENTED HERE AND THE FINAL REPORT IS PRESENTED BELOW.


SSGCID Functional Project Proposal:  Matt Clifton, Manager, UCB, Bainbridge Island, WA

The first known outbreaks of the Ebola virus occurred in Africa in 1976 with high mortality rates from severe and often fatal hemorrhagic fever. Although initially contained to small areas and limited to a few number of people, the 2013-2015 epidemic outbreak of the Ebola Zaire strain (EBOV) has affected over 26,000 people with over 11,000 deaths. Several additional strains are known (Reston, Tai Forrest), as well as Marburg and Cueva viruses from the same family. The 19 kb Ebolavirus RNA genome encodes seven proteins including a nucleoprotein (NP), a large protein (L) containing RNA-dependent RNA polymerase (RdRp) and methyl transferase domains, a glycoprotein (GP), and several accessory matrix proteins (VP24, VP30, VP35, VP40). SSGCID has previously published crystal structures of several of the matrix proteins in collaboration with Erica Ollmann Saphire at TSRI (Clifton, M.C. J. Infect. Dis. 2015; Clifton, M.C. Acta Cryst F 2014). Separate efforts are underway to understand the structural architecture of the L protein, especially the RdRp domain, which may be a good target for nucleotide analog inhibitors. During year 8 we have been performing fragment screening on the VP30 proteins from Zaire (SSGCID target ID: EbzaA.17250.a.EW11, VCID 11214; 1.75 Å resolution) and Marburg (SSGCID target ID: MalvA.17250.a.D11, VCID 6605 and 11260; 2.2 Å resolution) strains. Although we have not yet completed both screens, we have obtained proof of concept initial hits and expect analysis of the results to be completed by the end of Year 8.  We propose the following Function Study in Year 9 to follow-up on these initial results.

Specific Aim 1:  Initial Fragment Hit Characterization for VP30.
Work to be performed at UCB, carried out during Q(s) Q1-Q2.

We plan to use a variety of techniques to characterize the initial fragment hits identified by STD-NMR. These include differential scanning fluorimetry (DSF) and additional NMR experiments (WLG, HSQC, competition STD-NMR to determine similar or different site binders). These additional results will help prioritize fragments for co-crystallization efforts. Although an ideal lead molecule would target both strains, it may be beneficial to also study molecules which only bind in a strain-specific manner to understand potential resistance mechanisms that may arise down the line.

Specific Aim 2:  Co-crystal structure determination of VP30 with fragments.
Work to be performed at UCB, carried out during Q(s) Q2-Q3.

Our initial apo crystals structures of VP30 from the two different strains showed considerable structural differences, indicating the potential for different conformations available in solution. We will attempt soaking experiments, as well as co-crystallization with the various fragments that target one or both strains of VP30.

Specific Aim 3:  Chemical elaboration of initial hits.
Work to be performed at UCB, carried out during Q(s) Q2-Q4.

Given the compounds identified by initial STD-NMR, we will start to elaborate these hits using an “analog-by-catalog” approach. Chemical synthesis of select compounds may be necessary, although are not currently planned. Once available, these new compounds will be characterized by a panel of biophysical binding experiments (STD-NMR, DSF) and co-crystallization. At this point, we should have enough data on a few different chemical series for publication in a high-impact journal.

The results will be used to generate and qualify lead chemical matter targeting the Ebolavirus VP30 protein. The deliverables will be experimental results from DSF, competition STD-NMR and other NMR experiments; co-crystal structures deposited into the Protein Data Bank; and ultimately a manuscript detailing the experimental results and conclusions.

We project approximately 400 hours at UCBfor this research. Additional funds will be necessary for catalog SAR.

 

TIMELINE

MILESTONE

Quarter 1

09.2015-11.2015

  • Differential scanning fluorimetry (DSF) analysis of fragment hits identified (supports Aim 1)
  • Additional NMR experiments (supports Aim 1)

Quarter 2

12.2015-02.2016

  • Initiate Co-crystallization of initial hits (supports Aim 2)

 

Quarter 3

03.2016-05.2016

  • Analog by Catalog (supports Aim 3)
  • Characterization of catalog compounds (supports Aim 3)

Quarter 4

06.2016-08.2016

  • Submit manuscript detailing fragment screening, characterization and first round of chemical elaboration

 

FINAL REPORT

Functional Study 13: Fragment compounds against Ebola proteins.

Project lead: Silvia Delker, UCB

Status: Completed

Time-line: September 2015 - August 2016

ORIGINAL

TIMELINE

MILESTONE

ACHIEVED

Quarter 1

09.2015-11.2015

  • M1: Differential scanning fluorimetry (DSF) analysis of fragment hits identified (supports Aim 1)
  • M2: Additional NMR experiments (supports Aim 1)



X

Quarter 2

12.2015-02.2016

  • M3: Initiate Co-crystallization of initial hits (supports Aim 2)

X

Quarter 3

03.2016-05.2016

  • M4: Analog by Catalog (supports Aim 3)
  • M5: Characterization of catalog compounds (supports Aim 3)

X

Quarter 4

06.2016-08.2016

  • M6: Submit manuscript detailing fragment screening, characterization and first round of chemical elaboration

 

In progress

Summary: The primary goal of the functional study is to characterize and elaborate fragment compounds against Ebola accessory matrix protein VP30.  During the latter half of the functional study we achieved milestones M4 and investigated compound analogs for milestone M5 (in Q4).  In order to better understand VP30 in relationship to fragments, wild-type (apo - solved), “phosphorylated”, and “de-phosphorylated” EbzaA.17250.a protein is being explored as well as wild-type MalvA.17250.a.  Milestone M3 is still in progress, we continue to optimize co-crystal conditions, but soaking experiments may be more successful.  In Q4, weak density was observed in one structure and optimization is underway. To better characterize the protein, determining whether conformational changes occur within active and inactive protein, “phosphorylated” and “dephosphorylated” mutants were generated in EbzaA.17250.a  The relationship of active and inactive protein with fragments is being explored using NMR experiments and co-crystallization. The new constructs are predicted to crystallize similarly to wild-type, but anticipate conformational changes, possibly resulting in monomeric vs dimeric structures – exposing new surfaces exposed to small molecules.

Specific Aim 1Initial Fragment Hit Characterization for VP30. – completed Q1

Specific Aim 2Co-crystal structure determination of VP30 with fragments.

  • In progress. We completed milestone M3 initiating co-crystallization trials with VCID 11214 (EbzaA.17250.a.EW11). We setup crystallization trials; reproducing apo crystals (PDB ID: 5DVW) and exploring co-crystallization experiments with 6 compounds identified as mutual hits with both 11214 and 6605. 
  • We reproduced apo crystals (EbzaA.17250.a.EW11) and generated co-crystals with 2/6 shared hits with EbzaA.17250.a.EW11 protein.  The diffraction was ~2.0A, but contained no fragments.  Co-crystal and soak trials are underway with both apo and co-crystal conditions.  In Q4, weak density, possibly small molecule, was detected in one soak experiment and optimization is being pursued.
  • Based on conformational differences observed within same construct (EbzaA.17250), presumed related to crystal packing, in Q4 “phosphorylated” and “de-phosphorylated” mutants were designed and synthesized.  These constructs are undergoing crystallization trials now.

 Specific Aim 3:  Chemical elaboration of initial hits

  • In Q1 we completed preliminary analysis of 1D STD-NMR results, identifying unique and mutual hits as well as characterizing the (4) common motifs.  At least one representative of each motif (also a mutual hit) is currently in co-crystallization trials.
  • In Q3, based on the preliminary structural motifs of mutual hits, we completed milestone (M4) identifying analogs by catalog.  In Q4, we tested new compounds for 1D STD-NMR characterization with both VCID 6605 (MalvA.17250.a.D11) and VCID 11214 (EbzaA.17250.a.EW11).  Not all of the analogs were soluble in NMR conditions, but of the compounds that were screened, 4 ligands were identified as ≥10% STD, with 3 shared hits with both proteins from singleton experiments (M5).
  • In progress.  The active and inactive constructs are also be investigated with the mutual hit fragments to characterize whether the phosphorylated state contributes to binding, affinity, and stability of small molecules.