SSGCID
Seattle Structural Genomics Center for Infectious Disease

FUN_08

FUN_08: Mycobacterium cytidylate kinases: drug target for tuberculosis and proof-of-concept for the mycobacteria ortholog approach

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

SSGCID Functional Project Proposal:  PI, Wes Vam Voorhis, UW. Co-investigators: Justin Craig, UW; Jim Sacchettini, TX A&M; David Sherman SCRI. 

 We have determined the structure of two cytidylate kinases (CK) from two non-tuberculosis (TB) Mycobacterium species, M. smegmatis and M. abscessus.  We were not able to obtain the structure of the ortholog for M. tuberculosis (Mtb) CK but Mtb CK is a promising drug target.   Our recent manuscript that is under review on the utility of non-TB Mycobacterium spp. ortholog structures for TB drug discovery, hypothesized that the structure for non-TB orthologs, that are >55% identical to the corresponding M. tuberculosis orthologs, could be used to predict inhibitors for the M. tuberculosis orthologs.  The manuscript predicts that the inhibitors found for the M. smegmatis and/or the M. abscessus CKs would be effective against Mtb CK.  This is particularly interesting because Mtb CK is a promising drug target.  Mtb CK has been genetically validated as essential in several genome-wide screens of Mtb (Tuberculist.org). The active site of Mycobacterium spp. CKs has several hydrophobic regions for drug targeting and is predicted to be druggable by pocket programs.   Furthermore, Mycobacterium spp. CKs have no direct human ortholog and the human enzyme that carries out the same biochemical function (ATP + dCMP ó ADP + dCDP) has distinct differences in its active site from Mycobacterium CK, thus specific targeting to Mtb CK without toxicity to human cells should be possible. Thus CK is a promising drug target for Mtb. We have shown that the M. smegmatis, M. abscessus, and Mtb CKs are all enzymatically active and have developed a HTS-ready screen with M. smegmatis CK (the most active enzyme) using creation of ATP (ADP + dCDP ó ATP + dCMP) and measurement of ATP using luciferase (Kinase-glotm).   

The aim of this functional study is to discover ligands (inhibitors) for the M. smegmatis CK through HTS, then test their efficacy against the Mtb CK to test the hypothesis that inhibitors found in the M. smegmatis CK screen will also effectively inhibit Mtb CK.  We will look at a small diversity library at FHCRC HTS lab.  We will also export our assay to collaborators, such as Jim Sacchettini at Texas A & M, who have collections of mycobacterial active compounds, to see if any of them are inhibitors of CK, thus exposing their potential mechanism of action.  At the same time, we will co-crystalize inhibitors with M. smegmatis CK to find ways to improve the potency of the inhibitors.  Dr. Sherman’s lab would screen these CK inhibitors for activity against Mtb in culture. 

 

Specific Aim #1:  Screen diversity library at FHCRC, 25,000 compounds, for inhibition of M. smegmatis CK.
Work to be performed in the lab of HTS core lab at FHCRC, carried out during Q(s) Q1-Q2.

Use M. smegmatis CK HTS to screen diversity library at FHCRC HTS center. The entire 50K diversity library cannot be screened due to cost constraints so that we will focus on compounds predicted to penetrate the Mtb cell.  Follow up hits by dose response at FHCRC then purchase and confirm positive hits.

Specific Aim #2:  Screen Mtb cell active compounds from Texas A & M collection, for inhibition of M. smegmatis CK.
Work to be performed in the lab of Jim Sacchettini at Texas A & M, carried out during Q(s) Q1-Q2.

Screen 2K cell-active compound library.  If positives identified, rescreen with dose response and reconfirm with purchased or resynthesized compounds.

Specific Aim $3:  Test for inhibition of M. smegmatis CK hits against Mtb CK, co-crystallize with M. smegmatis CK, and test against Mtb cell growth.
Work to be performed in the labs of Wes Van Voorhis (Mtb CK enzyme assays) at UW and David Sherman (Mtb cell growth inhibiton) at Seattle BioMed, and co-crystallography work at Emerald Bio, carried out during Q(s) Q3-Q4. 

In the end, we will write peer-reviewed paper regarding these CK inhibitors, their promise in TB drug discovery, and testing the hypothesis that Mycobacterium spp. orthologs can aid in Mtb drug discovery. We will seek additional funding for continuing medicinal chemistry, biological testing, and PK/ADME/Tox of inhibitors so this study will have lasting impact.

The funds for this functional project will be to support the work at expert laboratories, outside of SSGCID, including the FHCRC HTS lab, Dr. Sacchettini’s laboratory at Texas A&M, at Emerald Bio (outside of SSGCID effort), Dr. Van Voorhis’ lab (outside of SSGCID effort), and at Dr. Sherman’s laboratory at Seattle BioMed. No expenditure of hours from SSGCID staff is anticipated.  These efforts are clearly distinct from SSGCID contractual work in that they; are beyond the initial solution of structures,  involve primarily HTS to find inhibitors,  involve assay development, involve co-crystallography with ligands, and involve tuberculosis cell culture to test the effects of compounds.  These studies are clearly beyond the funded scope of SSGCID and could not be done without Functional Study funding.

FINAL REPORT

Functional Study 8: Mycobacterium cytidylate kinases.

Project lead: Justin Craig,  UW PPG
Project collaborators: Jim Sacchettini, TX A & M, College Station, TX, David Sherman, SCRI, Seattle, WA. 

Status
: Completed
Time-line:

ORIGINAL

TIMELINE

MILESTONE

ACHIEVED

Quarter 1

09.2014-11.2014

  • M1: Begin screening both cell-active and diversity libraries. (supports Aims 1,2)

X

Quarter 2

12.2014-02.2015

  • M2: Complete screening at FHCRC/Texas A&M and reconfirm on M. smegmatis CK and show inhibition of Mtb CK  (supports Aims 1,2)

X

Quarter 3

03.2015-05.2015

  • M3: Repurchase/resynthesize hits and confirm activity against Mtb and M. smegmatis CK (supports Aim 3)

  • M4:Test hits against Mtb in vitro (supports Aim 3)

  • M5:Begin to set up hits for co-crystallography against M. smegmatis CK (supports Aim 3)

 

Quarter 4

06.2014-08.2015

  • M6: Write up findings for publication and grant application or for a priority project for the Structure-based Drug

  • Discovery Consortium (SDDC) where chemistry efforts can be funded

  • M7: Finish setting up hits for co-crystallography against M. smegmatis CK (supports Aim 3)

  • M8:Finish screening on Mtb cells (supports Aim 3)

i.p.

i.p. in progress, to continue after project end, using Option Quantity funds



Summary: After the screening of more than 20,000 compounds failed to identify any inhibitors for the M.smegmatis CK enzyme, a computational approach was adopted. Interaction matrix data from the MTB proteome generated by the computational analysis of novel drug opportunities (CANDO) was parsed for compounds likely to interact with the CK enzyme. 40 compounds were identified and ranked accordingly. The top 20 ranked compounds were ordered and screened for inhibition in the activity assay. No compounds showed inhibitory characteristics even at high concentrations in the millimolar range. Clearly, while MTB CK is an attractive target for drug development this work illustrates that for this particular target an alternative approach to HT screening of molecular drug like compounds may be required to produce potential inhibitors. These findings are being written up for publication.

Specific Aim 1:  Screen diversity library at FHCRC, 25,000 compounds, for inhibition of M. smegmatis CK.

Work completed Q2. A high through-put ready activity assay for M. smegmatis CK based on the creation of ATP (ADP + CDP ó ATP + dCMP) and measurement of ATP using luciferase (Kinase-glotm) was developed for screening of the compound libraries available at FHCRC. Since no inhibitors of the CK enzyme were found in the literature or the BRENDA enzyme database, the assay relied on the subtraction of an essential substrate (CDP) as a control for reduced activity. For screening the required number of compounds, approximately 200 mgs of M.smegmatis CK protein was purified and validated as active. The assay was modified and validated for high through-put 384 well format to use with FHCRC equipment. Screening was performed in 2 batches. The 1st batch (Spectrum compound library of 2300 compounds) yielded reliable statistics (z prime scores above 0.8) and 3 putative positive hits, however they were ruled out as false positives due to their known inhibition of the luciferase enzyme in Kinase-glotm. The 2nd batch consisted of 15000 compounds from the Chem Div Targeted Diversity Library (TDL). These compounds represent a subset of the full 50,000 compound TDL. This library is designed to provide a diverse compound set against eclectic biological targets with limited structure/ligand information. The results also yielded reliable statistics and 18 putative positive hits. None of these compounds appear in the literature as known inhibitors of luciferase. Follow up screening at the FHCRC of the putative positive hits re-confirmed inhibition in the CMK assay. However, for all but one of these compounds the inhibition was found to be due to the compound inhibiting the luciferase enzyme in Kinase-glotm and not the cytidylate kinase itself. This was determined by measuring ATP alone by Kinase-glotm with and without compound. The single remaining putative positive compound was ordered from Chem Div and tested again for CMK inhibition as well as luciferase inhibition at the UW PPG. While CMK inhibition was again confirmed in dose response experiments, it was also confirmed that the compound inhibited luciferase, contradicting earlier results. As a further test to determine if the compound interacts with CMK, a thermal shift assay was performed with CMK and the compound. The assay showed no melting temperature shift of the protein in the presence of the compound.  Therefore this compound was also abandoned as a putative positive. In completing Specific Aim 1, a total of 17500 compounds have been screened in accordance with funding. This work completes Milestone 1 and contributes to Milestones 2 and 3.

Specific Aim 2:  Screen Mtb cell active compounds from Texas A & M collection, for inhibition of M. smegmatis CK.

Completed Q2. Purified M. smegmatis CK protein from the SSGCID pipeline was shipped to James Sachettini’s lab at Texas A&M University. The cytidylate Kinase-glotm based enzyme assay was validated in high through-put format for Texas A&M equipment and then implemented to screen. Compounds chosen for screening were derived from two libraries, one from SRI and the other from Texas A&M University, tested in earlier whole cell assays against Mycobacterium tuberculosis Mc2-7000. Small molecule compound plates containing approximately 2600 compounds were screened in the assay as singlets at a concentration of approximately 20 µM. 9 compounds were observed as having some inhibitory effect on cytidylate kinase and were chosen for retest. Serial dilution series of the compounds from 50 µM to 10 nM were prepared and tested in the assay. None of the compounds which were re-tested demonstrated similar inhibitory activity as that observed initially. Although some apparent inhibitory activity was observed in screening whole cell inhibitory compounds in the cytidylate kinase enzyme assay, no confirmed active compounds were identified. In completing specific aim 2, 2600 compounds were screened. Bringing a total of 20,100 compounds screened. This work completes Milestone 2 and contributes to Milestones 1 and 3.

Specific Aim 3:  Test for inhibition of M. smegmatis CK hits against Mtb CK, co-crystallize with M. smegmatis CK, and test against Mtb cell growth.

Not started. Work for Specific Aim 3 could not be started due to the lack of inhibitors found in our screening efforts.

Work performed in Q4. In order to address the lack of inhibitors and after input from the contract’s Advisory Boards (TSB and SWG) on February 18, the scientific leadership team (SLT) decided to screen any hits generated by the computational analysis of novel drug opportunities (CANDO) performed by Ram Samuldara. Interaction matrix data for the entire MTB proteome (that was either in the PDB or could be modelled) against 3000+ compounds was parsed for the CMK enzyme. 40 compounds were identified and ranked according to potential interaction. The top 20 ranked compounds where screened in the M.smeg CK activity assay at high (milli to micro-molar range) and low (micro-molar range) concentrations. None of the compounds showed any inhibition of the CMK enzyme, therefore we ceased screening of the lower ranked compounds. Additionally, we tested the compound strausproine which was not a product of CANDO but is a well known and potent protein kinase inhibitor to assess the potential of screening against protein kinase inhibitor libraries. Since none of  the CANDO compounds nor staurosporine exhibited inhibition we are unable to begin work on Specific aim 3. Since our efforts represent a large amount of screening and computational analysis and since this enzyme is a well validated potential drug target we are currently drafting a paper of our findings for publication.