Until a project led by my lab's first postdoc, Alex Jaramillo, & stellar RA Kyra Taylor, my thoughts on the inoculum effect (IE) - the more bacteria there are, the greater the conc of antibiotic required to kill them - were fuzzy.
In a new preprint, imo we clarified the carbapenem IE: carbapenemases act as shared goods for bacterial communities. All carbapenemase-producing strains had a strong IE; those resistant due to porin deficiency had no IE at all.
This has key consequences…
🧵 1/n
(2/n)
…because carbapenems are a really important class of antibiotics (the broadest spectrum abx we use in clinical practice), and this is a huge effect: MICs (the amt of antibiotic that prevents bacterial growth) can vary by up to 1000x in some strains as cell density increases, and not change at all in others.
We found that this difference is a predictable consequence of mechanism of resistance.
Preprint here: https://www.biorxiv.org/content/10.1101/2023.05.23.541813v2
(3/n)
Some background: there are two main mechanisms of resistance to carbapenems, which (like penicillin) inhibit bacterial cell wall synthesis (focusing here on carbapenem resistant Enterobacterales, aka CRE):
1. carbapenemase production ("CP-CRE") - break the drug down
2. porin deficiency - don't let the drug into the periplasm where it can act
Because carbapenems act outside the cytoplasm (in the periplasm, where the cell wall is made), carbapenemases need to act outside the cell…
4/n
…which means these carbapenemases end up shared with neighbors. (In theory they could be constrained to the periplasm, but at least above a certain cell density, this doesn't seem to be the case: all CP-CRE have a strong IE.
By contrast, porin deficiency, which prevents carbapenems from getting into the periplasm, is a cell-autonomous resistance mechanism that has no effect on neighbors, hence no IE.
We verified this in >100 clinical CRE: 63 CP-CRE all had an IE; none of 47 non-CP-CRE did.
5/n
So why is this important? We thought of diagnostic, clinical, and mechanistic implications.
1. Clinical diagnostics: we measure resistance at a certain cell density to guide antibiotic use, so having MICs that vary widely is a problem. The IE is known, so guidance is given on acceptable inoculum range ("CLSI range" in figs) - but even in this range, MICs can vary widely. In fact, 1/3 of CP-CRE changed meropenem resistance class in the allowed range, and 25% tested S at an approved inoculum.
9/n
That's most of the story. A couple other reflections (see preprint for more):
2. Clinical implications: Porin-deficient strains take up carbapenems less well, but also nutrients, making them less fit. Yet studies haven't shown a consistent difference in outcomes in CP-CRE vs non-CP-CRE, despite this fitness difference. Most bacteremias have v low cell density (<10 cfu/mL), where CP-CRE MICs are << non-CP-CRE. Maybe greater resistance at low cell density offsets fitness cost in non-CP-CRE?
11/n
Anyway, the clearest way I can summarize what I learned from this project is: the carbapenem inoculum effect depends on the mechanism of resistance of the isolate, not the mechanism of action of the antibiotic. (So we should perhaps call it the carbapenemase inoculum effect?)
In some ways I feel like I should have known that. Maybe you already did? Happy to hear anyone's input - this is still under peer review, and I'm sure there's more to learn from it.
Roby Bhattacharyya