Journal Club: Harper et al 2017

It may be self-serving, but I thought it also apt for the first article I review for Journal Club be my own. My paper, titled Multiple substitutions lead to increased loop flexibility and expanded specificity in Acinetobacter baumannii carbapenemase OXA-239, is the culmination of my undergraduate research. I was looking at a clinical variant of the beta-lactamase enzyme OXA-23 which has 3 amino acid substitutions.

Some Background

Beta-lactams are penicillin-like antibiotics that function by inhibiting cell wall synthesis in gram-negative bacteria. Beta-lactamase enzymes are the primary method certain bacteria use to gain resistance to beta-lactams. One bacterial strain you may have heard of that uses beta-lactamase enzymes is MRSA – methicillin resistant staphylococcus aureus.

These beta-lactamase enzymes work by hydrolyzing the 4-membered beta-lactam ring in the antibiotic molecule (see figure below) before the drug can function by inhibiting cell wall formation. Different forms of these beta-lactamase drugs have been developed since Alexander Fleming discovered penicillin in 1928 (Penicillins, Cephalosporins, Carbapenems, Monobactams) all of which use this 4-membered beta-lactam ring. The alterations occur in other parts of the molecule, which prevent it from fitting into the beta-lactamase enzyme active site while still being able to inhibit cell wall formation.

Created with GIMP

Mutations in beta-lactamase enzymes cause small shifts in its 3D structure, allowing it to bind to new drugs and break that 4-member beta-lactam ring. Bacteria that lack this new version of the enzyme will be unable to reproduce while those that do will continue growing and passing that version of the enzyme to its progeny. This results in a new clinical variant of the bacteria that is resistant to that particular antibiotic. When a new clinical variant of one of these enzymes is discovered in a patient, its structure needs to be characterized in order to understand how it bypassed the newer antibiotics so new forms can be developed to then bypass the new enzyme variant.

What I Found

The beta-lactamase variant called OXA-239 has 3 mutations compared to the OXA-23 wild type version. There is a serine to leucine substitution at position 109 (S109L), an aspartate to asparagine substitution at position 222 (D222N), and a proline to serine substitution at position 225 (P225S). The P225S substitution was seen as a single mutation for OXA-23, called OXA-225, studied in our laboratory (Mitchell et al, 2015). Using steady-state kinetics assays, they found that this mutation confers increased activity against newer versions of antibiotic. Using a single D222N variant (one of the mutations in my clinical triple mutant), which is only 3 amino acids away from the P225, also conferred increased activity against some antibiotics. There was an even greater increase when both were together. However, the D222N mutation also lowered activity against a certain antibiotic called Imipenem. Adding the S109L mutation brought the activity against Imipenem back to normal. Using circular dichroism, we also found that the S109L mutation alleviated some loss of stability that the D222N and P225S mutations caused.

X-ray crystallography structures are shown in the figure below: OXA-23 wild-type is blue, OXA-239 with imipenem bound is yellow, OXA-239 with doripenem bound is cyan, and OXA-239 with cefotaxime bound is green.

Created with GIMP

The β6β7 arrow points to the loop where the P225S and D222N mutations occur; the loop at the top left is where the S109L mutation occurs; the space in between is the active site. Below are two figures: the left shows how each of the antibiotics fit inside the active site and the right shows how the loop is altered with each drug bound. A is doripenem, B is imipenem, C is cefotaxime.



The way the enzyme accommodates the three antibiotics bound in my structures was mostly the same as with the P225S single mutant, by allowing greater flexibility the β6β7 loop in the molecule. With the two carbapenem drugs (doripenem or imipenem) bound, the β6β7 loop takes roughly the same conformation as the OXA-23 WT. However, with the cephalosporin class cefotaxime bound the loop conforms into a hairpin by essentially zipping it up all the way to the end (as seen in green in C on the right above). This new conformation accommodates the large cephalosporin molecule and stabilizes the active site. The S109L mutation seemed to help accommodate the imipenem better by offering a more hydrophobic environment within the active site – it comes within Van der Waal’s distance from the imipenem molecule.