Skip to content
Dr Ben Bax

Dr Ben Bax

Reader

School of Biosciences

Email:
baxb@cardiff.ac.uk
Telephone:
+44 (0)29 22 51 1070
Location:
Main Building, Park Place, Cardiff, CF10 3AT

I am a Reader in Structural Biology in the Medicines Discovery Institute at Cardiff University. The goal of the Institute is to translate understanding of disease mechanisms into novel therapeutic approaches for patients in need of improved treatment options. Structural information about how compounds bind to their target proteins can help chemists design better molecules and can inform strategies for making novel therapeutics.

I worked for GlaxoSmithKline for eighteen years (1998-2016); supporting project teams with structural data, on a range of neuroscience, anti-microbial and other targets, including AMP receptor positive modulators (Ward, Bax and Harries, 2010; DOI: 10.1111/j.1476-5381.2010.00726.x

I supported the team who developed the new antibiotic gepotidacin (in phase III) with structural data (Bax, et al., 2010 Nature, 466, pp. 935-940). Published crystal structures of nineteen DNA-complex stablizing compounds (see table 1 on Research tab - and publications) suggest that for DNA gyrase conformationally flexible small molecules often make better inhibitors of this conformationally flexible drug target (protein/DNA complex) than more rigid small molecules.

I have a BSc in Physics and Chemistry from Nottingham University and a PhD in Protein Crystallography from the department of crystallography in Birkbeck College, University of London.  I have a passion for using structure-guided drug design to make new medicines to improve human health; and considerable experience as an industrial structural biologist (working for GlaxoSmithKline (GSK) from 1998-2016).

I had three excellent supervisors for my PhD, Tom Blundell, Peter Lindley and Christine Slingsby, and the structure obtained, of bB2-crystallin, was the first ‘domain swapped’ structure (Bax et al., 1990 - see Publications tab for details).  Before moving to industry in 1998 I worked on structural studies on a number of proteins including: ceruloplasmin (Zaitseva et al., 1996), PI 3-kinase (Panyotou et al., 1992; Dhand et al., 1994), protein kinase C (Srinivassan et al. 1996), 7S NGF (Bax et al., 1997), the small G-protein ARF (Greasely et al., 1995) and a complex of phosducin with the bg subunits of the heterotrimeric G protein transducin (Loew et al., 1998).

I joined SmithKlineBeecham (later GlaxoSmithKline) in 1998 to work as a protein crystallographer in a newly formed structural biology group.  Protein kinase structures solved included GSK-3b (Bax et al., 2001; Christopher et al., 2009; Gentile et al., 2011, 2012; Henley et al., 2017).  Crystal structures of AMPA receptor positive modulators helped advance chemistry on this challenging neuroscience target (Ward et al., 2010 a, b; Ward et al., 2011).  A major area of study was new antibiotics (and anti-cancer drugs) targeting bacterial type IIA topoisomerases (Bax et al., 2010, Chan et al., 2017, 2015, 2014, Miles et al., 2016, 2013, Srikannathasan et al., 2015, Agrawal et al., 2013, Wohlkonig et al., 2010; Germe et al., 2018; Bax et al., 2019).  The determination of structures of NBTIs in complexes with DNA and DNA gyrase helped the team developing gepotidacin (Gibson et al., 2019); gepotidacin is the first member of the NBTI class of anitbiotic to enter a phase III clinical trial. I continue to be interested in topoisomerase inhibitors.

In GSK I co-chaired the structural biology software group and I was the industrial representative on the CCP4 executive committee (CCP4 is a consortium that develops crystallographic software).  A talk from the 2016 CCP4 study weekend resulted in a paper entitled: ‘Getting the chemistry right: protonation, tautomers and the importance of H atoms in biological chemistry’.

I joined the Medicines Discovery Institute in Cardiff in 2018.

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2004

2001

2000

1999

1998

1997

1996

1995

1994

1993

1992

1991

1990

1989

1988

1987

Research interests

I am a structural biologist/crystallographer. The main focus of my research is to try to understand how compounds (small molecules) interact with and moderate the activities of proteins. The aim of my research is to help support chemists by providing structures to assist in structure guided drug design (including identifying happy and unhappy waters). 

Current major areas of research interest include:

  1. Structure-guided drug design with a focus on diseases of the central nervous system
  2. Inhibitors of bacterial type IIA topoisomerases (fluoroquinolones, twofold axis pockets-NBTIs etc.)

1. Structure-guided drug design with a focus on diseases of the central nervous system

Interests include AMPA receptors (10.1021/jm100679e), NMDA receptors and other targets. 

2. Inhibitors of bacterial type IIA topoisomerases

Type IIA topoisomerases are essential enzymes that regulate DNA topology by creating a temporary four base-pair staggered double stranded DNA break. Compounds which stabilize DNA-cleavage complexes with bacterial type IIA topoisomerases include the highly successful fluoroquinolone class of drugs as well as two novel compounds in late stage clinical development, zoliflodacin and gepotidacin (a new class of antibiotic which GlaxoSmithKline (GSK) have in Phase III clinical trials for uncomplicated urinary tract infection and urogenital gonorrhoea).

Structures I determined, while working for GSK, included the first quinolone structure showing the important 'water-metal-ion bridge' (Wohlkonig et al., 2010; DOI: 10.1038/nsmb.1892).  I also solved the first X-ray crystal structures of DNA complexes showing the binding modes of NBTIs  such as gepotidacin  (Bax et al.,2010; Gibson et al., 2019) and QPT-1 derivatives such as zoliflodacin (Chan et al., 2015). Because several S. aureus DNA gyrase complexes with DNA (Bax et al., 2019) have static disorder around the twofold axis of the ‘dimer’ – biological coordinates of ‘single complexes’ are available below – in table 1. See publications under reference tab for more details.

Note - these S.aureus DNA gyrase crystal structures include structures with clear views of the TOPRIM domain metal-ion binding sites – and suggest a single moving mechanism for DNA-cleavage. A 2.98Å yeast structure (pdb code: 3L4K) complicated by static disorder around a crystallographic twofold that was originally refined with two metals at each active site has been re-refined to be consistent with unambigous high resolution structures and coordinates for this yeast rerefined structure are available below in table 2.

TABLE 1 Coordinates of biological complexes of S.aureus DNA gyrase GyrBA fusion truncate with DNA and compounds.

Co-ordinates for biological complexes are available (click to upload) in the columns labelled ‘Coordinates for first (or second) complex in asymmetric unit’. Note the numbering scheme used is different from PDB numbering. If the complex has twofold disorder around the axis of the complex two complexes are available, representing the two orientations of the biological complex observed in the crystal structure. *Note most of the DNA complexes listed have one or two complexes in the asymmetric unit; but in the two apo structures (2xco and 2xcq, the GyrBA dimer sits on a crystallographic twofold and there is half a dimer in the asymmetric unit).

The S.aureus gyrase DNA complexes are all approx C2 symmetric and compounds have been observed in four distinct pockets: 1 (and 1'), 2D (on the twofold axis in the DNA), 2A (on the twofold axis between the two GyrA subunits), 3 (and 3').

no

PDB code +

resolution

Inhibitor             Crystal coords. (BA-x numb.), Space-group  [cell (a,b,c Å, and a,b,g °) ] Coordinates for first complex in asym. unit* Coordinates for second complex in asym. unit*
     

1

1’

2D

2A

3

3’

     

1

2xcq 2.98

none

-

-

-

 -

-

-

2xcq-BA-x.pdb

P6122, 90,90,416  90,90,120

2xcq-c1.pdb

 

2

2xco 3.1

none

-

-

-

-

-

2xco-BA-x.pdb

P6122, 90,90,411  90,90,120

2xco-c1.pdb

 
                       

3

6fqv 2.6

none

-

-

-

-

-

6fqv-BA-x.pdb

P21, 93,125,155  90,96,90

6fqv-c1.pdb

6fqv-c2.pdb

4

5cdr

2.65

none

-

-

-

-

-

5cdr-BA-x.pdb

P61, 93,93,411 90,90,120

5cdr-c1.pdb

 
                       

5

5iwi 1.98

‘237

-

-

X

X

-

-

5iwi-BA-x.pdb

P61, 93,93,411 90,90,120

5iwi-c1a.pdb

5iwi-c1b.pdb

 

6

2xcs

2.1Å

‘423

-

-

X

X

-

-

2xcs-BA-x.pdb,

P61, 93,93,413 90,90,120

2xcs-c1a.pdb

2xcs-c1b.pdb

 

7

6qtk

2.31Å

gepo'

-

-

X

X

-

-

6qtk-BA-x.pdb

P61, 93,93,409   90,90,120

6qtk-c1.pdb

6qtk-c2.pdb

 
8

6qtp

2.37Å
gepo' - - X X - -

6qtp-BA-x.pdb

P21, 86,124,94  90,117,90

6qtp-c1.pdb

6qtp-c2.pdb
 

9

5iwm

2.5Å

‘237

-

-

X

X

-

-

5iwm-BA-x.pdb

P61, 94,94,413 90,90,120

5iwm-c1a.pdb

5iwm-c1b.pdb

 

10

4bul

2.6Å

‘587

-

-

X

X

-

-

4bul-BA-x.pdb

P61, 94,94,416 90,90,120

4bul-c1a.pdb

4bul-c1b.pdb

 

11

2xcr

3.5Å

‘423

-

-

X

X

-

-

2xcr-BA-x.pdb

P212121 113,165,308 90,90,90

2xcr-c1a.pdb

2xcr-c1b.pdb

2xcr-c2a.pdb

2xcr-c2b.pdb

                       

12

5npp 2.22Å

‘237 + Thp2

-

-

X

X

X

X

5npp-BA-x.pdb

P61, 93,93,410 90,90,120

5npp-c1a.pdb

5npp-c1b.pdb

 
                       

13

5npk 1.98Å

Thp1

-

-

-

 -

X

X

5npk-BA-x.pdb

P21, 89,121,169 90,90.1,90

5npk-c1.pdb

5npk-c2.pdb

14

6qx1

2.65Å

Benzois’3

-

-

-

-

X

X

6qx1-BA-x.pdb

P61, 93,93,409   90,90,120

6qx1-c1.pdb

 
15

6qx2

3.4
Benzois’3 - - - - X X

6qx2-BA-x.pdb

P21, 188, 410,94  90,120.2,90

Six complexes

in asym. unit.

Poor resolution

 
                       

16

5cdp 2.45Å

Etop.

X

-

-

 -

-

-

5cdp-BA-x.pdb

P61, 93,93,411 90,90,120

5cdp-c1.pdb

 

17

5cdm 2.5Å

QPT-1

X

X

-

 -

-

-

5cdm-BA-x.pdb

P61, 94,94,412 90,90,120

5cdm-c1.pdb

 

18

5cdn 2.8Å

Etop.

X

X

-

-

-

-

5cdn-BA-x.pdb

P21, 90, 170, 125,  90, 102, 90

5cdn-c1.pdb

5cdn-c2.pdb

19

5cdq 2.95Å

Moxi.

X

X

-

 -

-

-

5cdq-BA-x.pdb

P21, 88, 171,126,  90, 103, 90

5cdq-c1.pdb

5cdq-c2.pdb

20

6fqm 3.06Å

IPY-t1

X

X

-

 -

-

-

6fqm-BA-x.pdb

P21 88, 172, 125,  90, 103, 90

6fqm-c1.pdb

6fqm-c2.pdb

21

6fqS 3.11Å

IPY-t3

X

X

-

 -

-

-

6fqs-BA-x.pdb

P61, 94,94,420 90,90,120

6fqs-c1a.pdb

6fqs-c1b.pdb

 

22

5cdo 3.15Å

QPT-1

X

X

-

 -

-

-

5cdo-BA-x.pdb

P21, 91,170, 125,  90, 103, 90

5cdo-c1.pdb

5cdo-c2.pdb

23

2xct 3.35

Cipro.

X

X

-

 -

-

-

2xct-v2-BA-x.pdb

P21, 89,123,170 90,90.3,90  90

2xct2-v2-c1.pdb

2xct2-v2-c2.pdb

Footnote: ‘237 = GSK945237; ‘423 = GSK299423; gepo = geoptidacin; ‘587 = GSK966587; Thp2 = thiophene 2; Thp1 = thiophene 1; Benzois’3 = benzoisoxazole3; Etop. = etoposide; QPT-1 = QPT-1; moxi. = moxifloxacin; IPY-t1 = imidazopyrazinone-tricyclic 1; ; IPY-t3 = imidazopyrazinone-tricyclic 3; cipro = ciprofloxacin.

Table 2 Coordinates of biological complexes for the deposited and re-refined crystal structures of 3L4K

Because 3L4K sits on a crystallographic twofold axis, the observed 2.98Å electron density is effectively a convolution of two structures superposed, related by the crystallographic twofold axis. This makes refinement and interpretation of the electron density more challenging, and more ambiguous than would be the case for a 2.98Å X-ray crystal structure not suffering from such static disorder. Below are presented coordinates from the two interpretations of the data: 3lk4.pdb and the derived complexes, 3l4k-c1a.pdb and 3l4k-c1b.pdb are the originally published interpretation (Schmidt et al., 2010), while RR-3l4k.pdb and RR-3l4k-c1a.pdb and RR-3l4k-c1b.pdb are from the re-refinement coordinates (see Bax et al., 2019 for details).

PDB  file Active site 1 Active site 2    
 

Metal site occupancies

WHD Tyr 782

Metal site occupancies

WHD Tyr 782'

Crystallographic coordinates

Coordinates for biological complex

 

A

B

 

A

B

     

Original 3L4K

1.0

1.0

Tyr

1.0

1.0

Tyr

3l4k.pdb

3l4k-c1a.pdb

3l4k-c1b.pdb

Re-refined RR-3L4K

0.5

0.5

Tyr

0.5

0.5

Tyr

RR-3l4k.pdb

RR-3l4k-c1a.pdb

RR-3l4k-c1b.pdb