The effect of serum in downregulating FASII in S. agalactiae observed by Brinster et al.1 suggests redundancy in fatty-acid synthesis upon exposure to a fatty-acid-rich environment. We confirmed that serum leads to a downregulation of S. agalactiae fab genes, but demonstrated unaltered fab expression in S. aureus (Fig. 1a). This might be due to differences in exogenous fatty-acid uptake, for example, it may be driven by cell-membrane hydrophobicity. S. aureus adapts its membrane hydrophobicity to escape innate-immune molecules, such as antimicrobial peptides and amphipathic fatty acids2,3, reiterating the specificity of host–pathogen interactions4,5.

Figure 1: S. aureus FASII inhibition is not rescued by exogenous fatty acids.
figure 1

a, Transcription levels of fab genes in mid-logarithmic S. agalactiae (left panel) in Todd Hewitt medium and S. aureus (right panel) in Mueller Hinton (MH) medium, with or without 50% human serum. The fold change was calculated using the standard curve method with gyrA as control (average ± s.d.; n = 4). b, Growth of wild-type (empty pAJ96 vector) and fabI antisense S. aureus strains in liquid MH with and without 150 ng ml-1 tetracycline (left panel) and MH + 50% human serum with and without 250 ng ml-1 tetracycline (right panel) (n = 2). The right panel also shows growth of fabI antisense S. aureus (induced with 250 ng ml-1 tetracycline) with 12-hour serum pre-adaptation (dotted blue line) (n = 2). Values are represented as averages ± s.d. c, Growth of wild-type (empty pAJ96 vector) and fabI antisense strains on solid medium with 20% human serum or 0.1% Tween-80 and incubated for 24 hours at 37 °C. FabI antisense is induced by adding 300 ng ml-1 tetracycline (MH + Tween-80) and 500 ng ml-1 tetracycline (MH + 20% human serum). d, In vivo activity of CG400549 against S. aureus in mouse peritonitis model. Percentage survival is shown after single oral dose (positive control is linezolid).

PowerPoint slide

Validating the effect of exogenous fatty acids for S. aureus requires inactivation of the drug target. Knockdown of enoyl-acyl-carrier-protein (ACP) reductase (FabI), a key enzyme in S. aureus FASII, was achieved by expression of tetracycline-induced fabI antisense RNA, which has previously been shown to have target selectivity6. In contrast to S. agalactiae1, we observed that fabI downregulation severely compromised S. aureus growth in both the absence and presence of serum (Fig. 1b and c), Tween-80 and oleic and stearic fatty acids (data not shown). To show whether S. aureus can adapt to serum fatty acid, we pre-incubated the fabI antisense strain for 12 hours in 50% serum. Even after this long fatty-acid exposure, the fabI antisense strain was growth-compromised in 50% serum (Fig. 1b). These differences can be explained by mechanistic and physiological diversity among bacterial FASII enzymes7. Whereas streptococci have FabK as the ACP reductase, staphylococci possess FabI as the fatty-acid elongation counterpart7. Additionally, only streptococci possess FabM, a cis–trans isomerase, suggesting that they have a different mechanism for formation of fatty acids8,9.

Brinster et al.1 reported highly increased minimum inhibitory concentrations (MICs) in the presence of serum for hydrophobic compounds like cerulenin (which targets FabB/FabH) and triclosan (targeting FabI) for several Gram-positive pathogens10,11. We tested the inhibitory activities of these compounds, and also platensimycin12 (targeting FabF/FabB) and CG400549 (targeting FabI; ref. 13). All inhibitors lost activity completely against S. agalactiae when these bacteria were grown in the presence of serum or Tween-80 (Table 1), suggesting exogenous serum compensatory mechanisms. However, for S. aureus serum or Tween-80 had little impact on the MICs of FASII inhibitors. The slightly increased MIC values for triclosan are due to its significant plasma protein binding (about 99.5%) (Table 1). Furthermore, pre-incubation with exogenous fatty acids did not rescue the growth inhibition of S. aureus by CG400549 (data not shown). To determine whether serum inhibits cellular fatty-acid synthesis, we measured the incorporation of [1-14C]-acetate into membrane fatty acids. Serum had no significant effect on [1-14C]-acetate incorporation in S. aureus, suggesting that these bacteria continue fatty-acid synthesis even in the presence of serum. In contrast, for S. agalactiae we observed a significant decrease in [1-14C]- acetate incorporation in the presence of serum (data not shown). These results suggest functional differences in FASII pathways among different Gram-positive pathogens.

Table 1 Susceptibilities of S. aureus and S. agalactiae to FASII pathway inhibitors

Final validation of drug targets is accomplished by in vivo pharmacodynamics. We tested CG400549 in a mouse peritonitis model and observed a dose-dependent survival of mice (Fig. 1d). Brinster et al.1 pointed out that bacterial adaptation to serum fatty acid is a gradual process. We observed that dosing of CG400549 in neutropenic mice, 6 hours after systemic infection (which allows host adaptation), killed about log10[colony-forming units] = 1.7 ± 0.3 in infected spleens (in untreated controls no bacteria were killed). These results and earlier reports on the efficacy of FabI inhibitors in rodent models14, including the clinical efficacy of Isoniazid, a FASII inhibitor in Mycobacterium tuberculosis15, argue against the broad claims by Brinster et al.1 that FASII is not a pharmaceutically attractive target.


S. aureus ATCC29213 and S. agalactiae NEM316 strains were used. Expression levels of FASII genes and the control gyrA were determined using quantitative real-time polymerase chain reaction (PCR). FabI antisense6 was cloned into the tetracycline-inducible pAJ96 vector (A. Neill, University of Leeds, UK) and electroporated into RN4220(pSTE2) (T. Hauschild, University of Bialystok, Poland). Swiss mice were intraperitoneally infected with S. aureus (108 colony-forming units) and immediately treated orally with CG400549 (percentage survival). Neutropenic mice were infected with S. aureus and dosed orally 6 hours post-infection with CG400549 (three times a day at 100 mg kg-1) for two days and killed to determine spleen colony-forming unit counts. References were purchased from Fasgen (triclosan), Sequoia Products (linezolid/cerulenin), Interchim (CG400549) and BioAustralis (platensimycin).