The fatty acid synthesis pathway is a checkpoint for lipoteichoic acid synthesis in Staphylococcus aureus

lipoteichoic acid (LTA), fatty acid synthesis (FASII), anti-FASII adaptation, wall teichoic acid 18 (WTA), lipidomics, glycerol-3-phosphate, cardiolipin, antimicrobial bitherapy, Staphylococcus aureus, 19 Streptococcus agalactiae. 20 21 Short title: Fatty acid synthesis activity controls lipoteichoic acid production 22 23 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted April 7, 2026. ; https://doi.org/10.64898/2026.04.06.715823doi: bioRxiv preprint 2

24 25 Bacterial membranes comprise diverse lipids whose proportions vary according to environmental 26 conditions. How cells direct lipid flux toward specific products remains unclear. We address this 27 question in the human pathogen Staphylococcus aureus, where multiple lipid products compete for a 28 common precursor, the major phospholipid, phosphatidylglycerol (PG). One product, lipoteichoic acid 29 (LTA), is essential for cell division, envelope homeostasis, and virulence. Lipids and metabolites were 30 quantified to identify factors that prioritize LTA synthesis over the other PG-derived products. We 31 identify upstream fatty acid synthesis (FASII) pathway as a key control point for LTA production. 32 Inhibition of FASII by antibiotics or gene inactivation causes LTA depletion . FASII inhibition similarly 33 affects LTA in Streptococcus agalactiae, suggesting conservation of this LTA control strategy. Changes 34 in membrane fatty acids do not account for LTA depletion. Instead, we show that FASII inhibition 35 causes a drop in intracellular glycerophosphate (GroP), a precursor for both PG and LTA. Under these 36 conditions of GroP limitation, PG flux favors production of a non-GroP lipid, cardiolipin. Moreover, 37 combined inhibition of FASII and WTA blocks S. aureus growth, confirming the lethality of depleting 38 LTA and WTA simultaneously. This study resolves how S. aureus manages phospholipid flux, by 39 prioritizing the synthesis of GroP-rich LTA or of non-GroP-containing lipids according to FASII-40 controlled GroP availability. 41 42 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted April 7, 2026. ; https://doi.org/10.64898/2026.04.06.715823doi: bioRxiv preprint 3

43 44 Phospholipids are primary components of most bacterial membranes, and are essential for cell 45 integrity. In Staphylococcus aureus and other Gram-positive pathogens, phospholipids are building 46 blocks for membrane-anchored structures. Among them, lipoteichoic acid (LTA) represents ~12 mole % 47 of total membrane outer leaflet lipids [1]. S. aureus LTA comprises a glycerophosphate (GroP) polymer 48 (~25-mer) usually anchored to the membrane via a di-glucosyl diacylglycerol (DG-DAG) lipid [2]. In S. 49 aureus, about three quarters of GroP polymers are decorated with D-alanines, which reduce the overall 50 negative charge of the polymer and contribute to antimicrobial peptide resistance [1-5] [6-8]. The 51 dedicated enzymes responsible for LTA synthesis, control of the GroP polymer tail length, and turnover 52 of the GroP donor lipid in S. aureus are well characterized [9-14]. 53 LTA is implicated in basic bacterial processes of cell division, autolysis, and antimicrobial 54 resistance, and also mediates bacteria-host interactions, which all relate to its physical properties [1, 55 15-17]. It is proposed to produce a stiff repulsive brush, which together with wall teichoic acid (WTA), 56 creates a barrier contributing to turgor pressure maintenance on the cell exterior. Bacterial septation 57 is suggested to be coordinated by LTA binding to autolysin, while the structure and charge of LTA 58 outside the membrane would offer protection against the pressure gradient inside and outside the cell 59 [16, 18, 19]. Both features of LTA could contribute to its roles in cell division. 60 While ltaS is essential for LTA synthesis, genes mediating polymer anchoring, i.e., ypfP (also 61 called ugtP) encoding diacylglycerol glucosyltransferase, and ltaA, encoding the DG-DAG anchor 62 flippase, are not [20]. Both ypfP and ltaA mutants produce longer GroP polymers, and the polymer is 63 suggested to use alternative anchors (i.e., phosphatidylglycerol (PG) or lysyl-PG [20, 21]; in one 64 exception, the ypfP mutant derived from SA113 was LTA-negative [22]). In addition, the CozEb protein 65 is involved in flipping the glycolipid anchor, and mutants also produce longer polymers [23]. Overall, 66 these mutants produce comparable amounts of LTA as the wild type (WT) [23, 24]. 67 Despite its essential roles, LTA is reportedly dispensable or deleterious in specific conditions or 68 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted April 7, 2026. ; https://doi.org/10.64898/2026.04.06.715823doi: bioRxiv preprint 4 mutant backgrounds. Growth of an ltaS mutant (defective for the LTA synthase LtaS) is restored in high 69 osmolarity or low temperature [16, 18, 25, 26]. Growth defects in clpX [27], pgl [28], and cshA [29] are 70 rescued by mutations in ltaS; LTA was absent or reduced in these mutants. Conversely, ltaS mutant 71 growth was rescued by various suppressor mutations ( cozEb, sgtB, mazEF, and clpX; [30]). 72 Interestingly, expression of LtaS, which is essential for LTA synthesis, decreases sharply in stationary 73 phase [31]. These findings predict that regulation of LTA expression may be multifactorial and 74 condition-dependent, and that LTA is not always required for growth. 75 Remarkably, LTA synthesis is only one of 4 possible pathways using PG as a substrate, 76 suggesting that these pathways are in competition, creating a ‘metabolic fork dilemma’ ( Fig. 1). The 77 factors that determine which end-product is favored are unknown. Solving this dilemma may be 78 complex, as LTA synthesis builds upon upstream pathways that provide PG precursors, namely for fatty 79 acid (FA) and GroP synthesis, to generate the LTA lipid anchor and GroP polymer. Moreover, the FA 80 synthesis pathway (FASII) in S. aureus and numerous Bacillota is dispensible and growth is fully 81 compensated by environmental FAs [32-35]. The shift from FASII to exogenous FA utilization (called 82 here ‘FASIIbypass’) is accompanied by marked changes in lipid metabolism and protein expression: these 83 include energy savings by not using FASII, and reverse directionality of the glycerol-3-phosphate 84 acyltransferase PlsX compared to FASII (Supplementary Fig. S1A and S1B) [35]. Here we established a 85 causal link between FASII activity and LTA synthesis in S. aureus, and give evidence for the generality 86 of this control in other species. Our findings suggest a simple mechanism for shunting PG towards LTA 87 synthesis PG, based on availability of GroP, a metabolite shared by both products and controlled by 88 FASII. LTA depletion during FASII bypass leads to an increase in WTA, opening perspectives for a 89 bitherapy treatment as shown in a proof of concept demonstration. 90 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted April 7, 2026. ; https://doi.org/10.64898/2026.04.06.715823doi: bioRxiv preprint 5

91 92 FASII inhibition induces S. aureus cell envelope changes during adaptation. S. aureus adapts to FASII 93 inhibition (FASII bypass) by using exclusively environmental FAs to produce membrane phospholipids. 94 FASIIbypass is a non-mutational event, but may also occur upon mutation of FASII initiation genes acc 95 and/or fabD [33, 34, 36, 37]. Bacteria adapted to an anti-FASII proliferate exponentially after an initial 96 latency phase (6 to 10 h according to conditions) ( Supplementary Fig. S1B ), and undergo profound 97 protein expression changes [33, 35]. S. aureus WT USA300 JE2 strain (called JE2) was examined by 98 transmission electron microscopy during adaptation to the anti-FASII (FabI inhibitor AFN-1252, 0.5 99 µg/ml [38]) ( Fig. 2A ). At 6 h post-anti-FASII treatment, prior to full adaptation, bacteria show a 100 pronounced transient increase in envelope thickness (~28.6 nm in non-treated bacteria, to 42.6 nm). 101 Once adapted (at 10 h), envelope thickness is slightly greater than that of non-treated bacteria (~29.7 102 nm, p=0.04), and cell division septa appear normally positioned. Overall, membrane contours were 103 more irregular in the FASII bypass bacteria ( Fig. 2A and inset, Supplementary Fig. S2 ). Alterations in 104 membrane integrity were observed in FASII bypass, as retention of a membrane-permeable drug, 105 ethidium bromide (EtBr), was greater in FASII bypass than in non-treated bacteria, particularly in 106 stationary phase (Fig. 2B). 107 A recent proteomics analysis of S. aureus responses to the anti-FASII triclosan [33, 35] was 108 rescreened to detect alterations in envelope biosynthesis functions (Supplementary Fig. S3). Notably, 109 the diacylglycerol glucosyltransferase YpfP (SAUSA300_0918, also called UgtP), which synthesizes the 110 LTA glycolipid diglucose anchor [11], was detected in non-treated bacteria, but dropped below 111 detection levels upon anti-FASII treatment. ypfP is not essential for growth or LTA formation, but GroP 112 polymers may be longer in the mutant than in WT strains [10]. The above observations were suggestive 113 of envelope structural differences between the two bacterial states, and motivated us to evaluate LTA 114 production in anti-FASII-adapted S. aureus. 115 116 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted April 7, 2026. ; https://doi.org/10.64898/2026.04.06.715823doi: bioRxiv preprint 6 FASII arrest leads to LTA depletion. LTA abundance was compared in the S. aureus JE2 non-treated 117 and FASIIbypass cultures by immunoblotting with LTA-specific antibody (Clone 55; [39]). LTA was readily 118 detected in non-treated bacteria, and at 6 h post-AFN-1252 treatment. However, LTA levels diminished 119 at 10 h post treatment (Fig. 3A). Upon dilution of FASIIbypass cultures, regrowth occurred without a lag 120 phase and LTA remained depleted, indicating that anti-FASII adaptation exerted a long-term effect; FA 121 profiles remained exogenous in this condition (Supplementary Fig. S4). LTA levels were also depleted 122 in the WT strain treated by another FASII inhibitor, platensimycin, which targets FabF (3-oxoacyl-(acyl-123 carrier-protein) synthase II) [34, 40, 41], showing that the observed effect was not antibiotic-specific 124 (Fig. 3B). An FA-auxotroph fabD deletion mutant, which relies on FASIIbypass for growth, similarly led to 125 LTA depletion (Fig. 3A, right). Strain specificity of the LTA response to FASII arrest was ruled out, as LTA 126 loss upon FASII bypass was observed in a different S. aureus lineage, NCTC_8325, shown for strains 127 RN4220-R and HG1-R (tested with AFN-1252; Supplementary Fig. S5). Thus all tested situations of S. 128 aureus FASIIbypass and growth compensation by exogenous FAs cause LTA depletion. 129 The potential generality of connecting FASII activity to LTA production was examined using the 130 major pathogen Streptococcus agalactiae, which produces LTA with a GroP polymer similar to that of 131 S. aureus [42]. In this and other streptococci, FASII enzymes are feedback-inhibited via exogenous FA-132 mediated activation of FabT, which represses FASII by forming an acyl-ACP-FabT complex. The 133 consequence is that streptococcal membrane phospholipids exclusively comprise exogenous FAs 134 whenever bacteria grow in lipid-containing environments [32, 43]. LTA was depleted when S. 135 agalactiae NEM316 was grown in medium supplemented with a single FA, here C17:1cis (Fig. 3C). We 136 conclude that blocking FASII causes LTA depletion and can be generalized to different S. aureus 137 lineages, and to a streptococcus species. 138 139 LTA levels are not restored by LtaS expression. Of the three main LTA synthesis enzymes, YpfP, the 140 antiporter flippase LtaA (SAUSA300_0917, encoded adjacently to ypfP), and the LTA synthase LtaS, 141 which catalyzes GroP polymerization ( SAUSA300_0703), the first two are not essential and may be 142 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted April 7, 2026. ; https://doi.org/10.64898/2026.04.06.715823doi: bioRxiv preprint 7 substituted by other S. aureus functions, although LTA gel migration and LTA localization are affected 143 [39]. During FASII bypass, detected traces of LTA migrate like the non-treated samples (e.g., Fig. 3) and 144 cell morphology appears normal ( Supplementary Fig. S2 ), suggesting that decreased YpfP levels are 145 not the cause of LTA depletion. We focused on LtaS as potentially explaining LTA depletion, as it 146 catalyzes GroP polymerization, and is essential for LTA synthesis [44]. We first evaluated LtaS status in 147 non-treated and FASII bypass S. aureus by comparing sensitivity to Congo Red, an LtaS inhibitor [45]. 148 Compared to non-treated S. aureus, Congo Red resistance of anti-FASII-adapted bacteria was ~10-fold 149 increased (Fig. 4A), suggesting that LtaS has a lesser role in this condition. We also asked whether LtaS 150 overexpression would restore LTA synthesis. The IPTG-inducible ltaS (iltaS) [9] established in the LAC 151 strain (ANG2505; kindly provided by Dr. A. Gründling, Imperial College, UK) was grown overnight in 152 SerFA containing 1 mM IPTG, without or with anti-FASII (AFN-1252, 0.5 µg/ml). Overnight cultures 153 were then washed and restarted without or with both IPTG and AFN-1252; growth and LTA production 154 were compared in the 4 conditions (Fig. 4B). Without IPTG, growth of the non-treated and anti-FASII-155 adapted iltaS strain nearly stopped, indicating that LtaS remained essential in both states. IPTG 156 addition restored growth of non-treated and FASIIbypass cultures (Fig. 4B). However, IPTG-induced LtaS 157 expression did not restore LTA production during FASII-bypass (Fig. 4C). 158 The SpsB protease cleaves LtaS and affects LTA length and localization [46-48]. The LtaS 159 cleavage product does not synthesize LTA when alone [48], but the cleaved product appears to be 160 required for efficient LTA synthesis [47]. We asked whether FASII bypass creates conditions that reduce 161 LtaS cleavage, which would explain LTA depletion. To test this, we used PK150 to stimulate SpsB [49, 162 50], and then assessed the effects on LTA synthesis. PK150 had no effect on LTA levels in either non-163 treated or FASII bypass conditions ( Fig 4D ). MspA reportedly sequesters SpsB and mspA inactivation 164 stimulates LTA production [51]. However, mspA mutants, without or with PK150, did not restore LTA 165 during FASIIbypass (Fig. 4E). The above results show that neither LtaS nor LtaS maturation is limiting for 166 LTA synthesis during FASII bypass. We therefore considered upstream pathways as possible causes for 167 LTA depletion. 168 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted April 7, 2026. ; https://doi.org/10.64898/2026.04.06.715823doi: bioRxiv preprint 8 169 FASII inactivation, but not membrane FA composition, dictates LTA depletion. FASII and/or FASIIbypass 170 supplies FAs for phospholipid synthesis, which in turn provides the substrates for LTA synthesis ( Fig. 171 1). We asked whether changes in FA composition during FASII bypass might be responsible for LTA 172 depletion. The S. aureus LTA lipid anchor comprises mainly two FAs, ai15 and C18:0, that also dominate 173 membrane phospholipid composition [1]. In our initial studies showing LTA loss, medium was 174 supplemented with a three-FA mixture (C14:0, C16:0, and C18:1), none of which are dominant in LTA. 175 Possibly, these FAs cannot produce lipid anchors appropriate for LTA synthesis, which would explain 176 decreased LTA production. We therefore performed the same experiment using a proportionately 177 balanced mixture of FAs corresponding to those synthesized by S. aureus (C14:0, ai15, C16:0, C18:0, 178 and C20:0; called ‘Natural FA mix’), so that the FA profiles of FASII bypass and non-treated WT cultures 179 were similar (Fig 5A upper). LTA levels were significantly decreased in both anti-FASII-adapted WT S. 180 aureus and in the fadD deletion strain, even in the presence of this natural complement of FAs ( Fig. 181 5A, lower). These results suggest that changes in FA composition are not a main factor leading to LTA 182 depletion in FASIIbypass conditions. 183 The above findings led us to hypothesize that FASII inactivation, and not the membrane FA 184 composition, generates a condition that halts LTA production. To test this, we devised a means to 185 obtain the same membrane FA composition in two conditions, one where FASII is inhibited (FASIIbypass), 186 and the other where FASII remains active. For the latter, we used a JE2-derived plsX mutant devoid 187 of PlsX [52, 53]. In plsX, phospholipids comprise exogenous FAs in position 1 of the glycerophosphate 188 backbone, and mainly FASII-synthesized ai15 in position 2. As ai15 is predominant in position 2, we 189 obtained homogenous membrane FAs in FASII bypass strains and in plsX by supplementing growth 190 media with solely ai15. Membrane FA profiles of Δ plsX, ΔfabD, and FASIIbypass cultures grown in ai15-191 supplemented media were nearly identical (Fig. 5B upper). In striking contrast, compared to both anti-192 FASII-adapted and ΔfabD cultures where LTA levels were depleted, LTA abundance in the ΔplsX mutant 193 was similar to that in the non-treated WT strain (Fig. 5B lower). Therefore, changes in FA composition 194 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted April 7, 2026. ; https://doi.org/10.64898/2026.04.06.715823doi: bioRxiv preprint 9 cannot account for LTA depletion. We conclude that inhibition of FASII activity, and not the membrane 195 FA composition, is mainly responsible for LTA depletion. 196 197 Pools of GroP, but not of ATP, are reduced during anti-FASII-adapted growth. The above results show 198 that FASII activity is needed for complete LTA production, but not to provide the phospholipid 199 precursors, which are assured by exogenous FAs. We asked whether the altered metabolite balance 200 during FASIIbypass compared to FASII could account for the shift away from LTA synthesis. Examination 201 of the major metabolites (Fig. 6A) involved in LTA production pointed to 2 candidates, ATP and GroP. 202 ATP depletion leads to reduced LTA synthesis as shown in pioneering studies of Fischer on the 203 biochemistry of LTA synthesis; synthesis of one LTA molecule costs ~150 ATPs [1, 15]. However, ATP 204 pools were higher in FASII bypass compared to non-treated extracts, ruling out this explanation for LTA 205 depletion ( Fig. 6B ). A neutral or positive ATP balance might be expected during FASII bypass, as 206 incorporation of exogenous FAs economizes ATP cost of FA synthesis (estimated at ~1 ATP per 2-207 carbon elongation [54]), e.g., 8 ATPs are used to synthesize one molecule of C16:0. This compares to 208 just 1 ATP consumed by the Fak kinase [55] for incorporation of any length exogenous FA. 209 GroP is an essential substrate for both PG and LTA synthesis in S. aureus. Despite the low molar 210 proportion of LTA compared to PG, about 50 % of total S. aureus GroP is sequestered within the LTA 211 polymer in non-treated conditions [56]. Synthesis of one LTA GroP polymer involves sequential GroP 212 transfer from ~25 PG molecules ( Fig. 6A; [1, 2]). As PG is the GroP donor, its synthesis necessarily 213 precedes that of LTA. The GroP pool was about 36 % lower in the FASII bypass condition than in non-214 treated bacteria (Fig. 6C). We conclude that the net drop in GroP, which is essential for PG synthesis 215 followed by LTA GroP polymer synthesis, is the probable cause for LTA depletion in FASIIbypass. 216 217 Cardiolipin (CL) levels are increased in anti-FASII-adapted S. aureus, but do not affect LTA production. 218 LTA is the only PG product to consume GroP ( Fig. 6A). We asked whether other PG lipid products 219 compensate LTA depletion in the GroP-limiting conditions imposed by FASIIbypass. Whole cell lipids were 220 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted April 7, 2026. ; https://doi.org/10.64898/2026.04.06.715823doi: bioRxiv preprint 10 extracted from non-treated, and FASIIbypass S. aureus cultures prepared in different FAs, i.e., ‘Natural 221 FA mix’, ai15+C18:0 (the predominant FA moieties of LTA [1]), or ai15, the main S. aureus FA. The plsX 222 strain, which produces LTA, was also examined in the ai15 growth condition. The proportion of DG-223 DAG, the major LTA lipid anchor, decreased in FASIIbypass conditions, in keeping with both LTA and YpfP 224 depletion (Fig. 3, Supplementary Fig. S3 ). In contrast, the proportions of CL increased to different 225 extents according to the added FAs in all FASIIbypass S. aureus cultures (Fig. 7A). Of note, the lipid profile 226 of plsX was similar to that of the non-treated WT strain, despite comprising exclusively ai15 in its 227 membrane (Fig. 7A right, violet bars); this indicates that FASIIbypass triggers these lipid changes. 228 If CL were required during FASII bypass as a compensatory mechanism, a cardiolipin synthesis 229 mutant would fail to adapt to anti-FASII. We assessed FASIIbypass in a cls1cls2 mutant, which lacks both 230 cardiolipin synthases [57]. Growth and LTA depletion in FASII bypass conditions were similar in WT and 231 cls1cls2 strains (Fig. 7B and 7C respectively). Thus, while proportions of CL increase in response to anti-232 FASII, CL is not required for adaptation. This might be expected, as CL is enriched in the inner 233 membrane leaflet (Fig. 6A; [58]), while LTA is exposed at the bacterial surface, ruling out a functional 234 compensation. We therefore considered that non-lipid factors might support S. aureus growth during 235 FASIIbypass. 236 237 WTA levels are increased in anti-FASII-adapted S. aureus , and required for adaptive growth and 238 survival. Wall teichoic acid (WTA) and LTA are proposed to cooperate to assure bacterial envelope 239 integrity, and WTA is required when LTA is absent [10, 16, 59]. Our previous proteomics heat map 240 showed increases in 4 out of 5 detected WTA biosynthetic proteins in response to anti-FASII treatment 241 (Supplementary Fig. S3). We assessed production of WTA as a possible factor compensating LTA loss 242 during FASII bypass growth. S. aureus WTA and LTA yields were compared in non-treated and FASIIbypass 243 conditions (Fig. 8A and 8B, respectively). S. aureus FASIIbypass exhibited 2-fold greater WTA than the 244 non-treated culture (p=0.05), while LTA amounts diminished by greater than 10-fold (p=5x10 -5). 245 Increased WTA production might contribute to S. aureus survival during anti-FASII treatment by 246 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted April 7, 2026. ; https://doi.org/10.64898/2026.04.06.715823doi: bioRxiv preprint 11 compensating for reduced levels of LTA. 247 As FASII-bypass reduces LTA and raises WTA levels, we asked whether FASII and WTA synthesis 248 inhibitors would act synergistically. To test this, bitherapy assays were performed on S. aureus JE2, 249 using AFN-1252 as anti-FASII, and targocil [60] as anti-WTA. Inhibition of growth and a >4-log reduction 250 in 15 h survival indicate a synergistic biostatic effect of the combined treatment ( Fig. 8C ). These 251 promising results suggest a bitherapy strategy to enhance the effects of two drugs that reach their 252 targets, where single treatment is ineffective. 253 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted April 7, 2026. ; https://doi.org/10.64898/2026.04.06.715823doi: bioRxiv preprint 12

254 255 The FA building blocks of all membrane phospholipids are synthesized by FASII and/or obtained from 256 environmental sources via FASII bypass. We showed that while both pathways generate FAs and 257 phospholipids, FASII, but not FASII bypass, promotes LTA production. The FASII activity requirement for 258 LTA production elucidated here in S. aureus also applies to a streptococcus ( S. agalactiae), where 259 external FAs repress FASII [32], suggesting the generality of FASII-controlled LTA synthesis. 260 Two explanations for the differential effects of FASII and FASII bypass on LTA synthesis were 261 excluded. First, a regulatory effect of FASII on the essential LTA synthesis enzyme LtaS was ruled out, 262 as LtaS overproduction did not restore LTA during FASIIbypass (Fig. 4). Second, a change in membrane FA 263 composition during FASII bypass did not cause LTA depletion, as the same FA composition obtained by 264 manipulating two different pathways, one being FASII bypass, and the other conserving FASII but 265 inactivating PlsX, resulted in different LTA outcomes ( Fig. 5B ). These exclusions led us to focus on 266 metabolites potentially underlying LTA synthesis that might be dictated by FASII and FASII bypass 267 pathways. 268 PG, the major S. aureus phospholipid, is the substrate for four possible products (Fig. 1). Among 269 them, LTA synthesis is doubly costly: First, LTA hordes an estimated half of bacterial GroP in polymer 270 form, which it receives from ~25 PGs. Second, by transferring their GroP moieties, the PGs must be 271 recharged with GroP in an ATP-dependent process [1, 56, 61]. LTA is the only PG product that calls for 272 this costly tradeoff, a case of “deshabiller Pierre pour habiller Paul”, where numerous PGs are 273 dismantled to produce one LTA. Our findings showing GroP depletion by FASII bypass lead to a simple 274 model in which the ‘choice’ of synthesizing LTA depends on the amount of available GroP, and on 275 whether PG production is slowed due to insufficient GroP ( Fig. 6A and 6C). Reduced LTA synthesis 276 could favor production of competing PG products that do not use GroP, i.e., CL, lysyl-PG, and/or 277 lipoprotein. Higher proportions of CL detected during FASII bypass are consistent with this proposal. 278 These results connect the bacterial metabolic state to expression levels of a multifunctional surface 279 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted April 7, 2026. ; https://doi.org/10.64898/2026.04.06.715823doi: bioRxiv preprint 13 structure, LTA, which contributes to cell integrity and division, and virulence. They tie in with our 280 previous observations showing slowed virulence kinetics of FASIIbypass adapted bacteria [35]. 281 The reasons for the reduction in GroP pools upon FASII bypass are not yet known. Unlike FASII, 282 FASIIbypass relies on reverse PlsX activity (Fig. S1), which alters enzyme-product homeostasis, and causes 283 accumulation of phospholipid intermediates [33, 34]. Previous work in Bacillus subtilis showed that 284 accumulated lysophosphatidylglycerol (LPA), a phospholipid intermediate, is unstable, generating 285 dephosphorylated monoacylglycerol; the GroP moiety is lost in this process [62]. We speculate that 286 FASIIbypass and consequent PlsX reversal desynchronizes phospholipid synthesis enzymes, similarly 287 leading to LPA accumulation and dephosphorylation of the GroP moiety. Futile GroP turnover from 288 phosphatidic acid (PA) intermediates would explain its depletion during FASII bypass (schematized in 289 Supplementary Fig. S6, salmon color inset). 290 Rapid growth of S. aureus and S. agalactiae by FASII bypass indicates that LTA depletion is 291 compensated by other factors, as also suggested from the numerous suppressor mutants and 292 conditions that alleviate the LTA requirement [25, 27-31]. CL is increased, but not required for 293 FASIIbypass; this might be expected, due to its enriched location in the inner membrane leaflet [58], 294 rather than the outer leaflet where LTA is located. The roles of other outer leaflet lipids, i.e., 295 lipoproteins, in compensating LTA depletion remain to be explored. The overall increase in WTA yields 296 and accrued sensitivity to a WTA synthesis inhibitor point to a more dominant role for WTA during 297 FASIIbypass, likely by compensating LTA depletion ( Fig. 8). Simultaneous inactivation of LTA and WTA is 298 lethal, possibly due to a collapse of the protective stiff cell wall structure that requires at least one of 299 these components [16, 18, 59]. Our findings showing the collateral effects of anti-FASII in altering S. 300 aureus virulence factor expression and reducing LTA can be exploited to design synergistic 301 antimicrobial bitherapy, combining an anti-FASII and an anti-WTA, to eliminate S. aureus. 302 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted April 7, 2026. ; https://doi.org/10.64898/2026.04.06.715823doi: bioRxiv preprint 14

303 304 Strains, media, and growth conditions. The following S. aureus strains were used: USA300_FPR3757 305 JE2 [63]); RN-R and HG1-R, which are derivatives of RN4220 and HG001 respectively, whose fakB1 306 alleles were repaired; note that the NCTC 8325 lineage carries a deletion in fakB1, resulting in a defect 307 in exogenous FA utilization [64]. A USA300 fabD mutant (removing the malonyl CoA-ACP 308 transacylase) was created as described [40]. The following strains were generously provided as follows: 309 strain ANG2505, IPTG-inducible expression of ltaS, described in RN4220 [9], and established in USA300 310 LAC, from Dr. Angelika Grundling (Imperial College London, UK); mspA transposon insertion mutants 311 corresponding to USA300_FPR3757 positions 2379899 or 2380097, from Dr. Paul Fey (University of 312 Nebraska, USA) [63]; USA300_FPR3757-derived LAC strain with a double deletion of cls1 and cls2, from 313 Dr. Andreas Peschel (University of Tubingen, Germany) [57]. ANG2505 and mspA mutants were 314 selected on 5 µg/ml erythromycin prior to experiments. Strains were cultured in BHI, or SerFA (BHI 315 comprising 10 % decomplemented calf serum with 250 µM FAs (Laradon, Sweden) prepared as an 316 equimolar mixture of C14:0, C16:0, and C18:1). Where specified, single FAs or a mixture simulating the 317 FAs endogenously produced by S. aureus (‘natural mix’ containing C14:0, 6.5%; ai15, 40.4%; C16:0, 318 6.3%; C18:0, 34.1%; C20:0, 12.6%) were prepared in BHI containing 10% delipidated bovine serum 319 (Eurobio, France) and used at the 250 µM final concentration. Cultures were started from independent 320 colonies from BHI solid medium, and inoculated overnight as SerFA pre-cultures grown aerobically at 321 37°C in SerFA. They were then diluted in SerFA to OD 600 = 0.1 without or with antibiotics, and growth 322 was followed for at least 10 h. FASII inhibitors were AFN-1252 (anti-FabI; Bioaustralis, Australia) or 323 platensimycin (anti-FabF, MedChemExpress, France), both used at 0.5 µg/ml. PK150 (Tebubio, France), 324 which stimulates SpsB activity [50], was used at 1-2 µM as indicated. S. aureus growth experiments 325 were performed with aeration at 37°C in tubes or using a plate reader (Tecan Spark, Tecan France). For 326 the latter, cultures were deposited in 96 well plates at initial OD 600 = 0.05 and growth was monitored 327 for 24h at 37°C. All growth experiments were performed in at least three biological replicates. S. 328 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted April 7, 2026. ; https://doi.org/10.64898/2026.04.06.715823doi: bioRxiv preprint 15 agalactiae NEM316 [32] was grown at 37°C without shaking in BHI containing 0.25% BSA without or 329 with 100 µM C17:1. This FA represses FASII without deterring growth [43, 65]. 330 331 Electron microscopy. Bacteria were fixed with 2% glutaraldehyde in 0.1 M Na cacodylate buffer (pH 332 7.2) for 3 hours at room temperature. Samples were then contrasted with 0.2% Oolong Tea Extract in 333 cacodylate buffer, postfixed with 1% osmium tetroxide containing 1.5% potassium cyanoferrate, and 334 gradually dehydrated in ethanol (30% to 100%). The samples were substituted in a mixture of ethanol 335 and Epon and embedded in Epon resin. Thin sections (70 nm) were collected onto 200 mesh copper 336 grids and counterstained with lead citrate. The grids were examined using a Hitachi HT7700 electron 337 microscope at 80 kV, and images were captured using a charge-coupled device camera. Envelope 338 thickness was assessed on 25-30 cells and at least 2 measurements were done per cell using ImageJ 339 FIJI software. and calculations of mean, standard deviation, and 2-tailed T-tests used for statistical 340 significance were performed using GraphPad Prism 9.5.1 and Excel software. 341 342 Ethidium bromide retention. S. aureus JE2 was precultured in SerFA and transferred to fresh medium 343 without or with AFN-1252, to obtain exponential (OD 600 = 1) and stationary (overnight) cultures. 0.5 344 OD600 units were collected for each sample and washed twice in PBS in the same volumes. 100 µl of 345 each washed sample was transferred to a black 96-well plate, and ethidium bromide (1 µg/ml) was 346 added. Ethidium bromide retention was monitored essentially as described [66, 67] on a Tecan plate 347 reader at 37°C for 60 min, using excitation wavelength 539 nm and emission wavelength 600 nm. The 348 20 min time point was used to assess differences between samples. Calculations of mean, standard 349 deviation, and 2-tailed T-tests used for statistical significance were performed using Excel software. 350 351 LTA Detection by immunoblotting LTA in S. aureus USA300 JE2 and LAC strains, and iltaS was assessed 352 in the specified OD 600 and conditions. Cultures were inoculated in SerFA and then subcultured 353 overnight in SerFA without or with anti-FASII. The following day, cultures were transferred to the same 354 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted April 7, 2026. ; https://doi.org/10.64898/2026.04.06.715823doi: bioRxiv preprint 16 fresh media and grown to OD600 between 1 and 3. Ten OD 600 units of each culture were collected and 355 centrifuged at 8,000 rpm for 5 min, washed once with phosphate-buffered saline (PBS) containing 356 0.02% Triton, and then washed twice with PBS. Pellets were stored at -80 °C prior to cell extractions. 357 FastPrep (program 2; 6.5 m/s, 50 s x 3) was applied to all samples. Bradford protein assay was 358 performed on lysates to determine sample concentrations. Ten µg equivalents of protein were 359 deposited in wells of 15% polyacrylamide-SDS gels. Gel contents were transferred to a PVDF membrane 360 (BIO-RAD, France) using the Power Blotter System cassette (Thermo Scientific, France) with the 361 ‘medium’ molecular weight program for 7 minutes. PVDF membranes were incubated with primary 362 LTA-specific antibody (Clone 55; HyCult Biotechnology, Holland) and horseradish peroxidase-363 conjugated goat anti-mouse secondary IgG antibody (ThermoFisher, France), diluted to 1:2,500 and 364 1:10,000, respectively. Immunoreactive LTA species were detected using enhanced 365 chemiluminescence (ThermoScientific, France), analyzed with the Image Lab 5.0 software (ChemiDoc 366 MP Imaging System, BioRad, France), and quantified with ImageJ Fiji software [68]. LTA 367 immunodetection experiments were performed on independently prepared samples at least 10 times 368 in WT JE2 in both non-treated and FASII bypass conditions, and as indicated for other S. aureus and S. 369 agalactiae assays. 370 371 Congo Red resistance. Non-treated and anti-FASII-adapted S. aureus JE2 cultures were grown to OD600 372 ~2 and then adjusted to OD600 = 1 to perform spot test dilutions (undiluted=UD). For each dilution, 5 µl 373 aliquots were spotted onto SerFA solid medium without or with 2.5 mg/ml Congo Red. Plates were 374 incubated 24 h at 37°C, and photographed. 375 376 FA Extraction. The equivalent of one OD600 unit of bacterial culture was centrifuged at 8,000 rpm for 5 377 min, washed once in 0.9% NaCl containing 0.02% Triton X-100, then twice in 0.9% NaCl at the same 378 speed and time. Cell pellets were subjected to membrane lipid extraction as described [33, 69]. Gas 379 chromatography was performed in split-splitless injection mode on an AutoSystem XL Gas 380 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted April 7, 2026. ; https://doi.org/10.64898/2026.04.06.715823doi: bioRxiv preprint 17 Chromatograph (Perkin-Elmer) equipped with a ZB-Wax capillary column (30 m x 0.25 mm x 0.25 mm; 381 Phenomenex, France). Data were analyzed using the TotalChrom Workstation program (Perkin-Elmer). 382 S. aureus and S. agalactiae FA peaks were detected between 12 and 32 minutes of elution and 383 identified based on their retention times compared to purified esterified FA standards. 384 385 Lipid Extraction and Profiling. For lipid extraction, 100 OD 600 unit equivalents were collected from 386 bacteria grown to OD 600 ≈ 3. Bacteria were washed as above for FA extraction. Pellets were freeze-387 dried overnight and stored at -80°C. Lipid extractions were performed as described [70]; all steps were 388 performed in glass tubes. Briefly, pellet lipids were extracted with 4.75 ml of extraction buffer 389 (chloroform, methanol, and 0.3 % NaCl in a 1:2:0.8 ratio) incubated at 80°C for 15 min, followed by 1 390 h vortexing at room temperature. This procedure was repeated once, but with a 30 min vortexing step. 391 Then, 2.5 ml each of chloroform and 0.3 % NaCl were added consecutively, and phase separation was 392 achieved by 15 min centrifugation at 4,000 rpm. The lower phase was collected and evaporated under 393 nitrogen gas. Dried total lipid extract was weighed and stored at -20°C. Prior to lipid analysis, samples 394 were solubilized in chloroform to obtain a concentration of 70 mg/mL of dried extract. 395 Lipid analysis was performed using Normal Phase Liquid Chromatography (NPLC) as described 396 [71] with a Dionex Ultimate 3000 RSLC system (ThermoFisher Scientific, Germany) equipped with two 397 quaternary pumps, an autosampler, and a column oven. The RSLC system was coupled online to a 398 Corona Ultra charged aerosol detector (Corona-CAD) and to an LTQ Orbitrap Velos Pro mass 399 spectrometer equipped with a linear ion trap and an orbital trap analyzer (ThermoFisher Scientific, 400 Germany). Lipid classes were quantified by using a mixture of commercial standards (Avanti, Germany) 401 in chloroform, containing mono-glucosyl diacylglycerol (MG-DAG, 840523P), di-glucosyl diacylglycerol 402 (DG-DAG, 840524P), phosphatidylglycerol (PG, 841138P) and cardiolipins (CL, 840012P); the equimass 403 mixture was injected at concentrations ranging from 0.025 to 0.5 mg/mL for each component. Cyanur-404 phosphatidylethanolamine (Cyanure-PE(16:0-16:0), 870287P, Avanti, Fr) was used as an internal 405 standard. Five µL samples were injected and concentrations of each lipid class were determined with 406 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted April 7, 2026. ; https://doi.org/10.64898/2026.04.06.715823doi: bioRxiv preprint 18 the calibration curves obtained by corona-CAD detection. Lysyl-phosphatidylglycerol (Lysyl-PG) was 407 quantified using the PG calibration curve. Mass spectrometry was used to confirm the lipid species 408 detected and estimate its average molecular mass. Data were analyzed using Thermo Xcalibur Qual 409 Browser. The proportion of each lipid is presented as the mean ± SD of two biological replicates using 410 GraphPad Prism 9.5.1. 411 412 ATP measurements. Non-treated and anti-FASII-adapted cultures were grown in SerFA medium to 413 OD600 = 2. For each sample tested, 100 µl of culture was collected and placed in a 96 white opaque-414 multiwell plate (Nucleon, France). 100µl of Bactiter-Glo (Promega, France) was added to each sample. 415 The plate was incubated at 25°C with shaking (160 rpm) for 5 min before measuring luminescence 416 (Tecan plate reader). Calculations of relative expression, standard deviation, and 2-tailed paired T-tests 417 used for statistical significance were performed using Excel software. 418 419 GroP measurements. JE2 cultures were prepared in non-treated (NT) and anti-FASII-adapted (AD) 420 conditions in SerFA medium, supplemented with AFN-1252 (0.5 µg/ml) for AD strains. The following 421 day, cultures were diluted to OD₆₀₀ = 0.05 in fresh medium and grown to OD₆₀₀ = ~3. Twenty-five OD600 422 equivalent units were collected and centrifuged at 8000 rpm for 5 min at room temperature. Pellets 423 were flash-frozen in liquid nitrogen to halt enzymatic reactions. Cells were then treated with 1 mL 5% 424 perchloric acid (HClO₄) for 30 minutes on ice, followed by thorough resuspension. Lysates were 425 centrifuged at 12100 rpm (14000 g) for 30 minutes at 4°C, and supernatants were collected and stored 426 at -80°C until analysis. 427 GroP was quantified by liquid chromatography coupled with mass spectrometer detector (LC-428 MS). Samples were injected for liquid chromatography using a Supelco-F5 column (2.1 × 150 mm; 3 429 µm) with a mobile phase consisting of water with 0.1% formic acid (phase A) and acetonitrile with 0.1% 430 formic acid (phase B). Elution was performed using a gradient program as follows: the run starts at 5% 431 phase B for 2 minutes, then increases to 100% phase B over 13 minutes, maintaining these conditions 432 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted April 7, 2026. ; https://doi.org/10.64898/2026.04.06.715823doi: bioRxiv preprint 19 for 6 minutes before returning to 5% phase B in 1 minute, followed by re-equilibration for 10 minutes. 433 The flow rate was set at 0.25 mL/min, and detection was carried out by mass spectrometry using an 434 Nx8060 triple quadrupole ion analyzer (Shimadzu, France) equipped with an electrospray ionization 435 (ESI) probe operating at 250°C. Detection was performed in multiple reaction monitoring (MRM) 436 mode, with the following transitions: 171.15>79.00; 171.15>127.10 and 171.15>103.05 (in negative 437 mode) and 173.10>99.05;173.10>81.10 and 173.10>63.05 (in positive mode). Calibration curves were 438 previously generated using commercial standards in the range of 0.2 to 50 pmol injected. Calculations 439 of relative expression, standard deviation, and 2-tailed paired T-tests used for statistical significance 440 were performed using Excel software. 441 442 WTA detection. Non-treated and anti-FASII-adapted S. aureus JE2 were prepared in SerFA without or 443 with AFN-1252 (0.5µg/ml) and cultured to OD600 = 2-3. Extractions were performed on 60 or 180 OD600 444 equivalent units of bacterial cultures as described [72]. Samples were migrated on 15% native 445 polyacrylamide gel electrophoresis (PAGE) and visualized using Alcian blue and silver nitrate staining 446 as described [72]. Gels were scanned on an Epson scanner and WTA peaks were quantified using 447 ImageJ Fiji software [68]. Standard deviation, and 2-tailed paired T-tests for statistical significance were 448 performed using Excel software. 449 450 Combination anti-FASII and anti-WTA treatment effects on S. aureus growth and survival. Growth of 451 JE2 was assessed in SerFA supplemented or not with AFN-1252 (0.5µg/ml) without or with the TarG 452 inhibitor targocil (anti-WTA, 15 µg/ml) [60]. Cultures were deposited in 96 well plates at initial OD600 = 453 0.05 and growth was monitored for 24 h at 37°C using a Tecan plate reader. Results correspond to 454 means of three biological replicates. Survival to single and combined treatment was performed using 455 the above growth conditions, except that targocil was added at 0, 5, 10 or 15 µg/ml. Bacterial survival 456 was assessed by serial dilution platings after 15h growth. All experiments were performed in triplicate. 457 458 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted April 7, 2026. ; https://doi.org/10.64898/2026.04.06.715823doi: bioRxiv preprint 20 459

460 461 We are grateful to Angelika Gründling (Imperial College London, UK) for insightful discussion of this 462 work. We thank Professors A. Gründling, Andreas Peschel (University of Tubingen, Germany), and Paul 463 Fey and Jennifer Endres (University of Nebraska) for their generous gifts of strains. We thank Marie-464 Françoise Noirot-Gros and Hasna Toukabri (Micalis), and team colleagues Jasmina Vidic, Philippe 465 Gaudu and David Halpern for valuable discussions and advice concerning this work. PW was awarded 466 a Franco-Thai PhD scholarship from Campus France. We gratefully acknowledge funding from 467 Fondation pour la Recherche Medicale (DBF20161136769; AG), the Agence Nationale de la Recherche 468 (ANR-16-CE15-0013; AG, PTC), and the ANR under the umbrella of the Joint Programming Initiative on 469 Antimicrobial Resistance (JPIAMR) ANR funding (ANR-22-AAMR-0007; AG, PTC). We thank the Région 470 Ile-de-France for financial support in the acquisition of instruments for the SAMM core facility (AS, BP). 471 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted April 7, 2026. ; https://doi.org/10.64898/2026.04.06.715823doi: bioRxiv preprint 21

472 473 1. Fischer W. Lipoteichoic acid and lipids in the membrane of Staphylococcus aureus . Med 474 Microbiol Immunol. 1994;183(2):61-76. Epub 1994/05/01. doi: 10.1007/BF00277157. PubMed PMID: 475 7935161. 476 2. Percy MG, Grundling A. Lipoteichoic acid synthesis and function in gram-positive bacteria. Annu 477 Rev Microbiol. 2014;68:81-100. Epub 2014/05/14. doi: 10.1146/annurev-micro-091213-112949. 478 PubMed PMID: 24819367. 479 3. Kiriukhin MY, Neuhaus FC. D-alanylation of lipoteichoic acid: role of the D-alanyl carrier protein 480 in acylation. J Bacteriol. 2001;183(6):2051-8. Epub 2001/02/27. doi: 10.1128/JB.183.6.2051-481 2058.2001. PubMed PMID: 11222605; PubMed Central PMCID: PMCPMC95102. 482 4. Neuhaus FC, Baddiley J. A continuum of anionic charge: structures and functions of D-alanyl-483 teichoic acids in gram-positive bacteria. Microbiol Mol Biol Rev. 2003;67(4):686-723. Epub 2003/12/11. 484 doi: 10.1128/MMBR.67.4.686-723.2003. PubMed PMID: 14665680; PubMed Central PMCID: 485 PMCPMC309049. 486 5. Schultz BJ, Snow ED, Walker S. Mechanism of D-alanine transfer to teichoic acids shows how 487 bacteria acylate cell envelope polymers. Nat Microbiol. 2023;8(7):1318-29. Epub 2023/06/13. doi: 488 10.1038/s41564-023-01411-0. PubMed PMID: 37308592; PubMed Central PMCID: 489 PMCPMC10664464. 490 6. Collins LV, Kristian SA, Weidenmaier C, Faigle M, Van Kessel KP, Van Strijp JA, et al. 491 Staphylococcus aureus strains lacking D-alanine modifications of teichoic acids are highly susceptible 492 to human neutrophil killing and are virulence attenuated in mice. J Infect Dis. 2002;186(2):214-9. Epub 493 2002/07/23. doi: 10.1086/341454. PubMed PMID: 12134257. 494 7. Peschel A, Otto M, Jack RW, Kalbacher H, Jung G, Gotz F. Inactivation of the dlt operon in 495 Staphylococcus aureus confers sensitivity to defensins, protegrins, and other antimicrobial peptides. J 496 Biol Chem. 1999;274(13):8405-10. Epub 1999/03/20. doi: 10.1074/jbc.274.13.8405. PubMed PMID: 497 10085071. 498 8. Weidenmaier C, Peschel A. Teichoic acids and related cell-wall glycopolymers in Gram-positive 499 physiology and host interactions. Nat Rev Microbiol. 2008;6(4):276-87. Epub 2008/03/11. doi: 500 10.1038/nrmicro1861. PubMed PMID: 18327271. 501 9. Grundling A, Schneewind O. Synthesis of glycerol phosphate lipoteichoic acid in Staphylococcus 502 aureus. Proc Natl Acad Sci U S A. 2007;104(20):8478-83. Epub 2007/05/08. doi: 503 10.1073/pnas.0701821104. PubMed PMID: 17483484; PubMed Central PMCID: PMCPMC1895975. 504 10. Hesser AR, Matano LM, Vickery CR, Wood BM, Santiago AG, Morris HG, et al. The length of 505 lipoteichoic acid polymers controls Staphylococcus aureus cell size and envelope integrity. J Bacteriol. 506 2020;202(16). Epub 2020/06/03. doi: 10.1128/JB.00149-20. PubMed PMID: 32482719; PubMed 507 Central PMCID: PMCPMC8404710. 508 11. Kiriukhin MY, Debabov DV, Shinabarger DL, Neuhaus FC. Biosynthesis of the glycolipid anchor in 509 lipoteichoic acid of Staphylococcus aureus RN4220: role of YpfP, the diglucosyldiacylglycerol synthase. 510 J Bacteriol. 2001;183(11):3506-14. Epub 2001/05/10. doi: 10.1128/JB.183.11.3506-3514.2001. 511 PubMed PMID: 11344159; PubMed Central PMCID: PMCPMC99649. 512 12. Matsuoka S, Chiba M, Tanimura Y, Hashimoto M, Hara H, Matsumoto K. Abnormal morphology 513 of Bacillus subtilis ugtP mutant cells lacking glucolipids. Genes Genet Syst. 2011;86(5):295-304. Epub 514 2012/03/01. doi: 10.1266/ggs.86.295. PubMed PMID: 22362028. 515 13. Grundling A, Brogan AP, James MJ, Ramirez-Guadiana FH, Roney IJ, Bernhardt TG, et al. PgpP is 516 a broadly conserved phosphatase required for phosphatidylglycerol lipid synthesis. Proc Natl Acad Sci 517 U S A. 2025;122(5):e2418775122. Epub 2025/01/27. doi: 10.1073/pnas.2418775122. PubMed PMID: 518 39869797; PubMed Central PMCID: PMCPMC11804483. 519 14. Mychack A, Evans D, Gilles T, James MJ, Walker S. Staphylococcus aureus uses a GGDEF protein 520 to recruit diacylglycerol kinase to the membrane for lipid recycling. Proc Natl Acad Sci U S A. 521 2025;122(12):e2414696122. Epub 2025/03/18. doi: 10.1073/pnas.2414696122. PubMed PMID: 522 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted April 7, 2026. ; https://doi.org/10.64898/2026.04.06.715823doi: bioRxiv preprint 22 40100631; PubMed Central PMCID: PMCPMC11962490. 523 15. Koch HU, Haas R, Fischer W. The role of lipoteichoic acid biosynthesis in membrane lipid 524 metabolism of growing Staphylococcus aureus. Eur J Biochem. 1984;138(2):357-63. Epub 1984/01/16. 525 doi: 10.1111/j.1432-1033.1984.tb07923.x. PubMed PMID: 6697992. 526 16. Erickson HP. How Teichoic Acids Could Support a Periplasm in Gram-Positive Bacteria, and Let 527 Cell Division Cheat Turgor Pressure. Front Microbiol. 2021;12:664704. Epub 2021/05/28. doi: 528 10.3389/fmicb.2021.664704. PubMed PMID: 34040598; PubMed Central PMCID: PMCPMC8141598. 529 17. Holtje JV, Tomasz A. Lipoteichoic acid: a specific inhibitor of autolysin activity in Pneumococcus. 530 Proc Natl Acad Sci U S A. 1975;72(5):1690-4. Epub 1975/05/01. doi: 10.1073/pnas.72.5.1690. PubMed 531 PMID: 239401; PubMed Central PMCID: PMCPMC432610. 532 18. Erickson HP. How bacterial cell division might cheat turgor pressure – a unified mechanism of 533 septal division in Gram-positive and Gram-negative bacteria. BioEssays. 2017;39(8):1700045. doi: 534 10.1002/bies.201700045 535 19. Kanampalliwar A, Shah M, Park Y, Jeong B, Lawson PA, Bell M, et al. The role of lipoteichoic acid 536 in Staphylococcus aureus cell wall integrity. BioRxiv. 2025. doi: 537 https://doi.org/10.1101/2025.01.16.633316 538 20. Grundling A, Schneewind O. Genes required for glycolipid synthesis and lipoteichoic acid 539 anchoring in Staphylococcus aureus . J Bacteriol. 2007;189(6):2521-30. Epub 2007/01/09. doi: 540 10.1128/JB.01683-06. PubMed PMID: 17209021; PubMed Central PMCID: PMCPMC1899383. 541 21. Guyet A, Alofi A, Daniel RA. Insights into the Roles of Lipoteichoic Acids and MprF in Bacillus 542 subtilis. mBio. 2023;14(1):e0266722. Epub 2023/02/07. doi: 10.1128/mbio.02667-22. PubMed PMID: 543 36744964; PubMed Central PMCID: PMCPMC9973362. 544 22. Fedtke I, Mader D, Kohler T, Moll H, Nicholson G, Biswas R, et al. A Staphylococcus aureus ypfP 545 mutant with strongly reduced lipoteichoic acid (LTA) content: LTA governs bacterial surface properties 546 and autolysin activity. Mol Microbiol. 2007;65(4):1078-91. Epub 2007/07/21. doi: 10.1111/j.1365-547 2958.2007.05854.x. PubMed PMID: 17640274; PubMed Central PMCID: PMCPMC2169524. 548 23. Barbuti MD, Lambert E, Myrbraten IS, Ducret A, Stamsas GA, Wilhelm L, et al. The function of 549 CozE proteins is linked to lipoteichoic acid biosynthesis in Staphylococcus aureus . mBio. 550 2024;15(6):e0115724. Epub 2024/05/17. doi: 10.1128/mbio.01157-24. PubMed PMID: 38757970; 551 PubMed Central PMCID: PMCPMC11237490. 552 24. Burtchett TA, Shook JC, Hesse LE, Delekta PC, Brzozowski RS, Nouri A, et al. Crucial Role for 553 Lipoteichoic Acid Assembly in the Metabolic Versatility and Antibiotic Resistance of Staphylococcus 554 aureus. Infect Immun. 2023;91(7):e0055022. Epub 2023/06/22. doi: 10.1128/iai.00550-22. PubMed 555 PMID: 37347167; PubMed Central PMCID: PMCPMC10353449. 556 25. Oku Y, Kurokawa K, Matsuo M, Yamada S, Lee BL, Sekimizu K. Pleiotropic roles of 557 polyglycerolphosphate synthase of lipoteichoic acid in growth of Staphylococcus aureus cells. J 558 Bacteriol. 2009;191(1):141-51. Epub 2008/10/28. doi: 10.1128/JB.01221-08. PubMed PMID: 559 18952789; PubMed Central PMCID: PMCPMC2612411. 560 26. Corrigan RM, Abbott JC, Burhenne H, Kaever V, Grundling A. c-di-AMP is a new second 561 messenger in Staphylococcus aureus with a role in controlling cell size and envelope stress. PLoS 562 Pathog. 2011;7(9):e1002217. Epub 2011/09/13. doi: 10.1371/journal.ppat.1002217. PubMed PMID: 563 21909268; PubMed Central PMCID: PMCPMC3164647. 564 27. Baek KT, Bowman L, Millership C, Dupont Sogaard M, Kaever V, Siljamaki P, et al. The Cell Wall 565 Polymer Lipoteichoic Acid Becomes Nonessential in Staphylococcus aureus Cells Lacking the ClpX 566 Chaperone. mBio. 2016;7(4). Epub 2016/08/11. doi: 10.1128/mBio.01228-16. PubMed PMID: 567 27507828; PubMed Central PMCID: PMCPMC4981727. 568 28. Zeden MS, Gallagher LA, Bueno E, Nolan AC, Ahn J, Shinde D, et al. Metabolic reprogramming 569 and altered cell envelope characteristics in a pentose phosphate pathway mutant increases MRSA 570 resistance to beta-lactam antibiotics. PLoS Pathog. 2023;19(7):e1011536. Epub 2023/07/24. doi: 571 10.1371/journal.ppat.1011536. PubMed PMID: 37486930; PubMed Central PMCID: 572 PMCPMC10399904. 573 29. Khemici V, Prados J, Petrignani B, Di Nolfi B, Berge E, Manzano C, et al. The DEAD-box RNA 574 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted April 7, 2026. ; https://doi.org/10.64898/2026.04.06.715823doi: bioRxiv preprint 23 helicase CshA is required for fatty acid homeostasis in Staphylococcus aureus . PLoS Genet. 575 2020;16(7):e1008779. Epub 2020/07/31. doi: 10.1371/journal.pgen.1008779. PubMed PMID: 576 32730248; PubMed Central PMCID: PMCPMC7392221. 577 30. Karinou E, Schuster CF, Pazos M, Vollmer W, Grundling A. Inactivation of the Monofunctional 578 Peptidoglycan Glycosyltransferase SgtB Allows Staphylococcus aureus To Survive in the Absence of 579 Lipoteichoic Acid. J Bacteriol. 2019;201(1). Epub 2018/10/17. doi: 10.1128/JB.00574-18. PubMed 580 PMID: 30322854; PubMed Central PMCID: PMCPMC6287468. 581 31. Sharkey LKR, Guerillot R, Walsh CJ, Turner AM, Lee JYH, Neville SL, et al. The two-component 582 system WalKR provides an essential link between cell wall homeostasis and DNA replication in 583 Staphylococcus aureus. mBio. 2023;14(6):e0226223. Epub 2023/10/18. doi: 10.1128/mbio.02262-23. 584 PubMed PMID: 37850732; PubMed Central PMCID: PMCPMC10746227. 585 32. Brinster S, Lamberet G, Staels B, Trieu-Cuot P, Gruss A, Poyart C. Type II fatty acid synthesis is 586 not a suitable antibiotic target for Gram-positive pathogens. Nature. 2009;458(7234):83-6. doi: 587 10.1038/nature07772. PubMed PMID: 19262672. 588 33. Kenanian G, Morvan C, Weckel A, Pathania A, Anba-Mondoloni J, Halpern D, et al. Permissive 589 Fatty Acid Incorporation Promotes Staphylococcal Adaptation to FASII Antibiotics in Host 590 Environments. Cell Rep. 2019;29(12):3974-82 e4. Epub 2019/12/19. doi: 591 10.1016/j.celrep.2019.11.071. PubMed PMID: 31851927. 592 34. Morvan C, Halpern D, Kenanian G, Hays C, Anba-Mondoloni J, Brinster S, et al. Environmental 593 fatty acids enable emergence of infectious Staphylococcus aureus resistant to FASII-targeted 594 antimicrobials. Nat Commun. 2016;7:12944. doi: 10.1038/ncomms12944. PubMed PMID: 27703138; 595 PubMed Central PMCID: PMC5059476. 596 35. Wongdontree P, Millan-Oropeza A, Upfold J, Lavergne JP, Halpern D, Lambert C, et al. Oxidative 597 stress is intrinsic to staphylococcal adaptation to fatty acid synthesis antibiotics. iScience. 598 2024;27(4):109505. Epub 2024/04/05. doi: 10.1016/j.isci.2024.109505. PubMed PMID: 38577105; 599 PubMed Central PMCID: PMCPMC10993138. 600 36. Gloux K, Guillemet M, Soler C, Morvan C, Halpern D, Pourcel C, et al. Clinical Relevance of Type 601 II Fatty Acid Synthesis Bypass in Staphylococcus aureus . Antimicrob Agents Chemother. 602 2017;61(5):61:e02515-16. doi: 10.1128/AAC.02515-16. PubMed PMID: 28193654; PubMed Central 603 PMCID: PMC5404599. 604 37. Parsons JB, Frank MW, Subramanian C, Saenkham P, Rock CO. Metabolic basis for the differential 605 susceptibility of Gram-positive pathogens to fatty acid synthesis inhibitors. Proc Natl Acad Sci U S A. 606 2011;108(37):15378-83. doi: 10.1073/pnas.1109208108. PubMed PMID: 21876172; PubMed Central 607 PMCID: PMC3174620. 608 38. Kaplan N, Awrey D, Bardouniotis E, Berman J, Yethon J, Pauls HW, et al. In vitro activity (MICs 609 and rate of kill) of AFN-1252, a novel FabI inhibitor, in the presence of serum and in combination with 610 other antibiotics. Journal of chemotherapy. 2013;25(1):18-25. Epub 2013/02/26. doi: 611 10.1179/1973947812Y.0000000063. PubMed PMID: 23433440; PubMed Central PMCID: 612 PMCPMC3558989. 613 39. Reichmann NT, Picarra Cassona C, Monteiro JM, Bottomley AL, Corrigan RM, Foster SJ, et al. 614 Differential localization of LTA synthesis proteins and their interaction with the cell division machinery 615 in Staphylococcus aureus . Mol Microbiol. 2014;92(2):273-86. Epub 2014/02/19. doi: 616 10.1111/mmi.12551. PubMed PMID: 24533796; PubMed Central PMCID: PMCPMC4065355. 617 40. Léguillier V, Gloux K, Khalife M, Catto M, Péchoux C, Sentic M, et al. Biotope-dependent High 618 Level Resistance to Reactive Oxygen Species, Antibiotic Tolerance, and Virulence of Staphylococcus 619 aureus. BioRxiv. 2025;Submitted. doi: https://doi.org/10.1101/2025.10.31.685911. 620 41. Wang J, Soisson SM, Young K, Shoop W, Kodali S, Galgoci A, et al. Platensimycin is a selective 621 FabF inhibitor with potent antibiotic properties. Nature. 2006;441(7091):358-61. doi: 622 10.1038/nature04784. PubMed PMID: 16710421. 623 42. Schneewind O, Missiakas D. Lipoteichoic acids, phosphate-containing polymers in the envelope 624 of gram-positive bacteria. J Bacteriol. 2014;196(6):1133-42. Epub 2014/01/15. doi: 10.1128/JB.01155-625 13. PubMed PMID: 24415723; PubMed Central PMCID: PMCPMC3957714. 626 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted April 7, 2026. ; https://doi.org/10.64898/2026.04.06.715823doi: bioRxiv preprint 24 43. Lambert C, Poyart C, Gruss A, Fouet A. FabT, a Bacterial Transcriptional Repressor That Limits 627 Futile Fatty Acid Biosynthesis. Microbiol Mol Biol Rev. 2022;86(3):e0002922. Epub 2022/06/22. doi: 628 10.1128/mmbr.00029-22. PubMed PMID: 35726719; PubMed Central PMCID: PMCPMC9491164. 629 44. Richter SG, Elli D, Kim HK, Hendrickx AP, Sorg JA, Schneewind O, et al. Small molecule inhibitor 630 of lipoteichoic acid synthesis is an antibiotic for Gram-positive bacteria. Proc Natl Acad Sci U S A. 631 2013;110(9):3531-6. Epub 2013/02/13. doi: 10.1073/pnas.1217337110. PubMed PMID: 23401520; 632 PubMed Central PMCID: PMCPMC3587227. 633 45. Vickery CR, Wood BM, Morris HG, Losick R, Walker S. Reconstitution of Staphylococcus aureus 634 Lipoteichoic Acid Synthase Activity Identifies Congo Red as a Selective Inhibitor. J Am Chem Soc. 635 2018;140(3):876-9. Epub 2018/01/05. doi: 10.1021/jacs.7b11704. PubMed PMID: 29300473; PubMed 636 Central PMCID: PMCPMC5856125. 637 46. Park Y, Sisley TA, Stone MC, Corrigan RM, Vickery CR, Grundling A, et al. SpbR controls 638 lipoteichoic acid length by directly inhibiting signal peptidase SpsB in Staphylococcus aureus. Proc Natl 639 Acad Sci U S A. 2025;122(27):e2426464122. Epub 2025/07/01. doi: 10.1073/pnas.2426464122. 640 PubMed PMID: 40587784; PubMed Central PMCID: PMCPMC12260438. 641 47. Ibrahim AM, Azam MS, Schneewind O, Missiakas D. Processing of LtaS restricts LTA assembly 642 and YSIRK preprotein trafficking into Staphylococcus aureus cross-walls. mBio. 2024;15(2):e0285223. 643 Epub 2024/01/04. doi: 10.1128/mbio.02852-23. PubMed PMID: 38174934; PubMed Central PMCID: 644 PMCPMC10865820. 645 48. Wormann ME, Reichmann NT, Malone CL, Horswill AR, Grundling A. Proteolytic cleavage 646 inactivates the Staphylococcus aureus lipoteichoic acid synthase. J Bacteriol. 2011;193(19):5279-91. 647 Epub 2011/07/26. doi: 10.1128/JB.00369-11. PubMed PMID: 21784926; PubMed Central PMCID: 648 PMCPMC3187375. 649 49. Chen SY, Fiedler MK, Gronauer TF, Omelko O, von Wrisberg MK, Wang T, et al. Unraveling the 650 mechanism of small molecule induced activation of Staphylococcus aureus signal peptidase IB. 651 Commun Biol. 2024;7(1):895. Epub 2024/07/24. doi: 10.1038/s42003-024-06575-x. PubMed PMID: 652 39043865; PubMed Central PMCID: PMCPMC11266668 of smartbax limited. All other authors declare 653 no competing interests. 654 50. Le P, Kunold E, Macsics R, Rox K, Jennings MC, Ugur I, et al. Repurposing human kinase inhibitors 655 to create an antibiotic active against drug-resistant Staphylococcus aureus, persisters and biofilms. Nat 656 Chem. 2020;12(2):145-58. Epub 2019/12/18. doi: 10.1038/s41557-019-0378-7. PubMed PMID: 657 31844194; PubMed Central PMCID: PMCPMC6994260. 658 51. Bonini D, Duggan S, Alnahari A, Brignoli T, Strahl H, Massey RC. Lipoteichoic acid biosynthesis by 659 Staphylococcus aureus is controlled by the MspA protein. mBio. 2024;15(8):e0151224. Epub 660 2024/07/22. doi: 10.1128/mbio.01512-24. PubMed PMID: 39037275; PubMed Central PMCID: 661 PMCPMC11323550. 662 52. Wongdontree P, Palmier M, Leguillier V, Louche C, Gloux K, Halpern D, et al. FabF and FadM 663 Cooperate to Recycle Fatty Acids and Rescue Δ plsX Lethality in Staphylococcus aureus. BioRxiv. 2025. 664 doi: https://doi.org/10.1101/2024.12.18.629266. 665 53. Parsons JB, Frank MW, Eleveld MJ, Schalkwijk J, Broussard TC, de Jonge MI, et al. A thioesterase 666 bypasses the requirement for exogenous fatty acids in the plsX deletion of Streptococcus pneumoniae. 667 Mol Microbiol. 2015;96(1):28-41. Epub 2014/12/24. doi: 10.1111/mmi.12916. PubMed PMID: 668 25534847; PubMed Central PMCID: PMCPMC4415690. 669 54. Varman AM, He L, You L, Hollinshead W, Tang YJ. Elucidation of intrinsic biosynthesis yields using 670 13C-based metabolism analysis. Microb Cell Fact. 2014;13(1):42. Epub 2014/03/20. doi: 10.1186/1475-671 2859-13-42. PubMed PMID: 24642094; PubMed Central PMCID: PMCPMC3994946. 672 55. Parsons JB, Frank MW, Jackson P, Subramanian C, Rock CO. Incorporation of extracellular fatty 673 acids by a fatty acid kinase-dependent pathway in Staphylococcus aureus . Mol Microbiol. 674 2014;92(2):234-45. doi: 10.1111/mmi.12556. PubMed PMID: 24673884; PubMed Central PMCID: 675 PMC4007170. 676 56. Fischer W. Bacterial Phosphoglycolipids and Lipoteichoic Acids. In: Kates M, editor. Glycolipids, 677 Phosphoglycolipids, and Sulfoglycolipids Handbook of Lipid Research, Vol 6. 6: Springer, Boston MA.; 678 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted April 7, 2026. ; https://doi.org/10.64898/2026.04.06.715823doi: bioRxiv preprint 25 1990. p. 123-234. 679 57. Koprivnjak T, Zhang D, Ernst CM, Peschel A, Nauseef WM, Weiss JP. Characterization of 680 Staphylococcus aureus cardiolipin synthases 1 and 2 and their contribution to accumulation of 681 cardiolipin in stationary phase and within phagocytes. J Bacteriol. 2011;193(16):4134-42. Epub 682 2011/06/15. doi: 10.1128/JB.00288-11. PubMed PMID: 21665977; PubMed Central PMCID: 683 PMCPMC3147675. 684 58. Beltran-Heredia E, Tsai FC, Salinas-Almaguer S, Cao FJ, Bassereau P, Monroy F. Membrane 685 curvature induces cardiolipin sorting. Commun Biol. 2019;2:225. Epub 2019/06/27. doi: 686 10.1038/s42003-019-0471-x. PubMed PMID: 31240263; PubMed Central PMCID: PMCPMC6586900. 687 59. Santa Maria JP, Jr., Sadaka A, Moussa SH, Brown S, Zhang YJ, Rubin EJ, et al. Compound-gene 688 interaction mapping reveals distinct roles for Staphylococcus aureus teichoic acids. Proc Natl Acad Sci 689 U S A. 2014;111(34):12510-5. Epub 2014/08/12. doi: 10.1073/pnas.1404099111. PubMed PMID: 690 25104751; PubMed Central PMCID: PMCPMC4151746. 691 60. Lee K, Campbell J, Swoboda JG, Cuny GD, Walker S. Development of improved inhibitors of wall 692 teichoic acid biosynthesis with potent activity against Staphylococcus aureus. Bioorg Med Chem Lett. 693 2010;20(5):1767-70. Epub 2010/02/09. doi: 10.1016/j.bmcl.2010.01.036. PubMed PMID: 20138521; 694 PubMed Central PMCID: PMCPMC2844852. 695 61. Fischer W. Molecular analysis of lipid macroamphiphiles by hydrophobic interaction 696 chromatography, exemplified with lipoteichoic acids. Anal Biochem. 1993;208(1):49-56. Epub 697 1993/01/01. doi: 10.1006/abio.1993.1007. PubMed PMID: 8434795. 698 62. Paoletti L, Lu YJ, Schujman GE, de Mendoza D, Rock CO. Coupling of fatty acid and phospholipid 699 synthesis in Bacillus subtilis. J Bacteriol. 2007;189(16):5816-24. Epub 2007/06/15. doi: 700 10.1128/JB.00602-07. PubMed PMID: 17557823; PubMed Central PMCID: PMCPMC1952045. 701 63. Fey PD, Endres JL, Yajjala VK, Widhelm TJ, Boissy RJ, Bose JL, et al. A genetic resource for rapid 702 and comprehensive phenotype screening of nonessential Staphylococcus aureus genes. MBio. 703 2013;4(1):e00537-12. doi: 10.1128/mBio.00537-12. PubMed PMID: 23404398; PubMed Central 704 PMCID: PMC3573662. 705 64. Pathania A, Anba-Mondoloni J, Gominet M, Halpern D, Dairou J, Dupont L, et al. (p)ppGpp/GTP 706 and Malonyl-CoA Modulate Staphylococcus aureus Adaptation to FASII Antibiotics and Provide a Basis 707 for Synergistic Bi-Therapy. mBio. 2021;12(1). Epub 2021/02/04. doi: 10.1128/mBio.03193-20. PubMed 708 PMID: 33531402; PubMed Central PMCID: PMCPMC7858065. 709 65. Lambert C, Gaillard M, Wongdontree P, Bachmann C, Hautcoeur A, Gloux K, et al. The double-710 edged role of FASII regulator FabT in Streptococcus pyogenes infection. Nat Commun. 2024;15(1):8593. 711 Epub 2024/10/05. doi: 10.1038/s41467-024-52637-3. PubMed PMID: 39366941; PubMed Central 712 PMCID: PMCPMC11452403. 713 66. Couto I, Costa SS, Viveiros M, Martins M, Amaral L. Efflux-mediated response of Staphylococcus 714 aureus exposed to ethidium bromide. J Antimicrob Chemother. 2008;62(3):504-13. Epub 2008/05/31. 715 doi: 10.1093/jac/dkn217. PubMed PMID: 18511413. 716 67. Pinamonti D, Manzano M, Maifreni M, Bianco S, Domi B, Ferrin A, et al. Prevalence and 717 Characterization of Staphylococcus aureus Isolated from Meat and Milk in Northeastern Italy. J Food 718 Prot. 2025;88(2):100442. Epub 2024/12/27. doi: 10.1016/j.jfp.2024.100442. PubMed PMID: 719 39725327. 720 68. Schindelin J, Arganda-Carreras I, Frise E, Kaynig V, Longair M, Pietzsch T, et al. Fiji: an open-721 source platform for biological-image analysis. Nat Methods. 2012;9(7):676-82. Epub 2012/06/30. doi: 722 10.1038/nmeth.2019. PubMed PMID: 22743772; PubMed Central PMCID: PMCPMC3855844. 723 69. Rozes N, Garbay S, Denayrolles M. A rapid method for the determination of bacterial fatty acid 724 composition. Lett Appl Microbiol. 1993;17(3):126-31. doi: https://doi.org/10.1111/j.1472-725 765X.1993.tb01440.x. 726 70. Bligh EG, Dyer WJ. A rapid method of total lipid extraction and purification. Can J Biochem 727 Physiol. 1959;37(8):911-7. Epub 1959/08/01. doi: 10.1139/o59-099. PubMed PMID: 13671378. 728 71. Abreu S, Solgadi A, Chaminade P. Optimization of normal phase chromatographic conditions for 729 lipid analysis and comparison of associated detection techniques. J Chromatogr A. 2017;1514:54-71. 730 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted April 7, 2026. ; https://doi.org/10.64898/2026.04.06.715823doi: bioRxiv preprint 26 Epub 2017/08/05. doi: 10.1016/j.chroma.2017.07.063. PubMed PMID: 28774713. 731 72. Covas G, Vaz F, Henriques G, Pinho MG, Filipe SR. Analysis of Cell Wall Teichoic Acids in 732 Staphylococcus aureus. Methods Mol Biol. 2016;1440:201-13. Epub 2016/06/18. doi: 10.1007/978-1-733 4939-3676-2_15. PubMed PMID: 27311674. 734 73. Freeman DJ, Falkiner FR, Keane CT. New method for detecting slime production by coagulase 735 negative staphylococci. J Clin Pathol. 1989;42(8):872-4. Epub 1989/08/01. doi: 10.1136/jcp.42.8.872. 736 PubMed PMID: 2475530; PubMed Central PMCID: PMCPMC1142068. 737 74. Ferreira AA, Tette PAS, Mendonça RCS, Soares AdS, Maria De Carvalho M. Detection of 738 exopolysaccharide production and biofilm-related genes in Staphylococcus spp. isolated from a poultry 739 processing plant. Food Science and Technology. 2014; 34(4):710-6. doi: 740 http://dx.doi.org/10.1590/1678-457X.6446. 741 75. Kurokawa K, Kim MS, Ichikawa R, Ryu KH, Dohmae N, Nakayama H, et al. Environment-mediated 742 accumulation of diacyl lipoproteins over their triacyl counterparts in Staphylococcus aureus . J 743 Bacteriol. 2012;194(13):3299-306. Epub 2012/04/03. doi: 10.1128/JB.00314-12. PubMed PMID: 744 22467779; PubMed Central PMCID: PMCPMC3434734. 745 76. Fischer W, Koch HU. Alanyl lipoteichoic acid of Staphylococcus aureus: functional and dynamic 746 aspects. Biochem Soc Trans. 1985;13(6):984-6. Epub 1985/12/01. doi: 10.1042/bst0130984. PubMed 747 PMID: 4092831. 748 77. Pereira MP, D'Elia MA, Troczynska J, Brown ED. Duplication of teichoic acid biosynthetic genes 749 in Staphylococcus aureus leads to functionally redundant poly(ribitol phosphate) polymerases. J 750 Bacteriol. 2008;190(16):5642-9. Epub 2008/06/17. doi: 10.1128/JB.00526-08. PubMed PMID: 751 18556787; PubMed Central PMCID: PMCPMC2519377. 752 78. Cuypers MG, Subramanian C, Gullett JM, Frank MW, White SW, Rock CO. Acyl-chain selectivity 753 and physiological roles of Staphylococcus aureus fatty acid-binding proteins. J Biol Chem. 754 2019;294(1):38-49. Epub 2018/11/16. doi: 10.1074/jbc.RA118.006160. PubMed PMID: 30429218; 755 PubMed Central PMCID: PMCPMC6322867. 756 79. Parsons JB, Broussard TC, Bose JL, Rosch JW, Jackson P, Subramanian C, et al. Identification of a 757 two-component fatty acid kinase responsible for host fatty acid incorporation by Staphylococcus 758 aureus. Proc Natl Acad Sci U S A. 2014;111(29):10532-7. Epub 2014/07/09. doi: 759 10.1073/pnas.1408797111. PubMed PMID: 25002480; PubMed Central PMCID: PMCPMC4115530. 760 761 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted April 7, 2026. ; https://doi.org/10.64898/2026.04.06.715823doi: bioRxiv preprint 27 Figure legends 762 763 Figure 1. The metabolic dilemma of producing four possible lipids from one substrate. 764 Phosphatidylglycerol (PG) gives rise to four alternative lipid products. The factors that may favor the 765 flux towards one of these pathways are unknown. 766 767 Figure 2. Morphological changes in the S. aureus envelope accompany anti-FASII adaptation. The JE2 768 strain was grown in SerFA (BHI containing 10% serum and an equimolar mixture of 3 FAs) or in SerFA 769 supplemented with the anti-FASII (AFN-1252, 0.5 µg/ml; SerFA-AFN). A. Transmission electron 770 microscopy: left, exponential (3 h) growth of non-treated S. aureus in SerFA to OD600 = ~3; middle, 6 h 771 growth post anti-FASII-treatment; right, 10 h post-anti-FASII-treatment, OD600 = ~3. White bar, 200 nm. 772 A zoom of the surface within the dotted box highlights morphological differences in septal regions. 773 Envelope thickness in the three conditions are shown in the graph at right. P-values were determined 774 using a two-sided T-test based on 73, 89, and 60 measurements respectively on at least 15 individual 775 bacteria (Fig. 2A Source data). Additional images are in Supplementary Fig. S2. B. Ethidium bromide 776 (EtBr) retention was compared in non-treated (NT) and anti-FASII-adapted (AD) S. aureus in 777 exponential (exp, N=3) and stationary (stat; N=6) phase cultures. EtBr retention is presented at 20 778 minutes; see Fig. 2B Source data for the full data set. P-values were determined using a two-sided T-779 test on biological triplicates for exponential and six replicates for stationary cultures. NS, non-780 significant. 781 782 Figure 3. FASII bypass leads to LTA depletion in S. aureus and S. agalactiae. S. aureus and numerous 783 streptococci compensate FASII inhibition, deletion, or repression, by incorporating exogenous FAs in 784 membranes [32, 33, 43, 65] . A. Antibiotic- or fabD-mutation- generated FASII inhibition leads to LTA 785 depletion in S. aureus. WT JE2 was grown in SerFA without (NT, non-treated) or with FabI inhibitor 786 AFN-1252 (AD, anti-FASII-adapted) [38]. fabD is a FASII auxotroph ([34, 40]), and was grown in SerFA 787 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted April 7, 2026. ; https://doi.org/10.64898/2026.04.06.715823doi: bioRxiv preprint 28 (N=3). B. Both anti-FabI AFN-1252 (AFN) and anti-FabF platensimycin (PTM, [41]) lead to LTA depletion 788 in S. aureus (N=3). C. Inhibition of FASII by exogenous FAs in the streptococcus species S. agalactiae 789 leads to LTA depletion. The S. agalactiae FASII pathway is repressed by FabT when bound to exogenous 790 FAs [32] . Here, S. agalactiae strain NEM316 was grown in BHI plus 0.025% FA-free bovine serum 791 albumin without or with C17:1 (100 µM); exogenous FA addition represses FASII but allows bacterial 792 growth. LTA was monitored by immunoblotting using anti-LTA antibody (N=5). Samples were taken 793 from exponential phase OD600 = ~2-3 (A and B) and OD600 = ~1 (C). 794 795 Figure 4. LtaS overproduction does not restore LTA during FASII bypass. A. Congo red inhibits LtaS 796 activity [45]. Non-treated (NT) and anti-FASII-adapted (AD) S. aureus JE2 overnight cultures were 797 adjusted to the same OD600 and dilutions were plated on solid SerFA medium without and with Congo 798 Red (N=3). AD cultures show greater resistance than NT to Congo Red, possibly suggesting less reliance 799 on LtaS. Black zones surrounding AD colonies on Congo Red indicates exopolysaccharide production 800 [73, 74]. B, C. LtaS is required for growth but does not restore LTA in anti-FASII-adapted S. aureus. B. 801 The IPTG inducible locus iltaS [9] established in the USA300 LAC derivative strain ANG2505 was grown 802 overnight with 1 mM IPTG . Growth of ANG2505 NT (orange) and AD (treated by AFN-1252; green) 803 cultures was followed without (dashed line) or with 1 mM IPTG (solid line). Results show the mean ± 804 SD (standard deviation) of 4 biological replicates. C, D, E. LTA was detected by immunoblotting using 805 anti-LTA antibody. C. Increasing LtaS expression does not restore LTA synthesis in FASIIbypass conditions. 806 The WT parental LAC strain and ANG2505 iltaS were grown in NT and AD conditions in the presence of 807 1 mM IPTG. N=3. D. PK150 stimulates SpsB, which enhances LtaS cleavage and reportedly increases 808 LTA [50, 51]. PK150 was added as indicated in NT and AD conditions and LTA production was assessed. 809 E. Mutations in mspA derepress SpsB protease, leading to increased LTA [51]. LTA levels in JE2 WT and 810 mspA mutants (USA300_FPR3757 mutant positions 2379899 [ mspA1] or 2380097 [ mspA2][63]) were 811 examined in exponential phase SerFA (NT) or SerFA-AFN (AD) cultures. Samples at right were also 812 treated with 1 µM PK150 during growth (N=3). 813 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted April 7, 2026. ; https://doi.org/10.64898/2026.04.06.715823doi: bioRxiv preprint 29 814 Figure 5. LTA production relies on FASII activity independently of membrane FA composition. S. 815 aureus non-treated (NT), anti-FASII-adapted (AD), and fabD (disabled for FASII) cultures were grown 816 in A, 250 µM “Natural Mix”, which mimics endogenously produced FAs. B. The same strains and 817 conditions as in A, plus a plsX strain (deleted for plsX, see Supplementary Fig. 1; [52] were grown in 818 250 µM ai15, the major FA synthesized by S. aureus; ai15 is elongated to ai17 and ai19 only in the non-819 treated WT strain. A, B , upper: FA composition of membrane extracts determined by gas 820 chromatography. Peak heights correspond to relative responses (mV) of each FA in a sample. 821 Predominant FAs are indicated; N=2. A, B , lower: Cell extracts were prepared and submitted to 822 immunoblotting using anti-LTA antibody N=3. Samples in B were run on the same gel and subjected to 823 the same exposure time. 824 825 Figure 6. ATP and GroP costs of LTA synthesis, starting with FASII and FASIIbypass pathways. A. Multiple 826 biosynthetic pathways underly synthesis of LTA, and carry an energy cost. LTA production relies on 3 827 major biosynthetic pathways (FASII and FASII bypass, ①a and ①b; phosphatic acid synthesis, ②; and 828 phosphatidylglycerol synthesis, ③) before branching to LTA synthesis. FAs are in green; lipoprotein 829 sometimes comprises a third FA (dashed line) [75]. 830 ATP and GroP or Gro expenses (boxed in red), gains (in green), or neutral changes (in grey) during 831 production of the 4 possible PG products, including LTA ( Fig. 1). For a single LTA molecule, one PG is 832 used to produce the anchor, while ~25 PGs donate GroP to produce the polymer. The GroP donors are 833 then recycled to regenerate PGs, which requires ATP. Synthesis of a single LTA molecule costs about 834 150 ATPs [1], 25 GroPs, and 26 PGs. In contrast, production of the three other lipid products leads to 835 positive or neutral ATP and GroP footprints. PG product distribution in the inner and outer membrane 836 leaflets is asymmetric: cardiolipin, but not the other lipid products, is preferentially enriched in the 837 membrane inner leaflet [58]. B and C, bacteria were grown in SerFA (non-treated) or in SerFA-AFN 838 (FASIIbypass). B. ATP pools are greater in FASIIbypass conditions, as FASIIbypass uses ~7-10 times less ATP per 839 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted April 7, 2026. ; https://doi.org/10.64898/2026.04.06.715823doi: bioRxiv preprint 30 FA molecule compared to FASII (see A). N=5. C. FASIIbypass causes depletion of GroP pools. These 840 measurements do not include GroP net loss associated with depletion of the GroP polymer attached 841 to LTA. N=6. Data in B and C are shown as mean values ± SD normalized to cognate measurements in 842 non-treated samples; P-values were determined using the two-sided T-test. Full data sets for B and C 843 is in Fig. 6B and 6C Source data . 844 845 Figure 7. FASII bypass leads to lipid species redistibution. A. Lipid extractions were performed on non-846 treated (NT) and anti-FASII-adapted (AD) S. aureus JE2, and the plsX derivative where indicated, from 847 OD600 = ~3 cultures prepared in BHI plus 10% delipidated serum, supplemented by ‘Natural Mix’ (FA 848 Mix), ai15 and C18:0 (125 µM each), or ai15 (250 µM). Relative mass proportions (% lipid (mass)) of 849 monoglucosyl diacylglycerol (MG-DAG), diglucosyl diacylglycerol (DG-DAG; the LTA lipid anchor), 850 phosphatidylglycerol (PG), and cardiolipin (CL) are shown, with data points (black dots), ranges, and 851 average of biological duplicates. Of note, the plsX strain produces LTA (Fig. 5) and its lipid distribution 852 is similar to that of the WT NT strain. See Fig. 7A Source data for original data readouts. B and C. WT 853 S. aureus JE2 and LAC strains, and the cls1cls2 (cls12) LAC derivative [57] were grown in SerFA 854 without and with anti-FASII (AFN-1252; AFN). B. Growth of WT LAC and cls12 strains was compared 855 in NT and AD cultures, and was comparable in each condition (N=3). C. LTA detection by 856 immunoblotting was performed on WT JE2 and LAC, and cls12 cultures using anti-LTA antibody (N=3). 857 858 Figure 8. S. aureus wall teichoic acid (WTA) is produced at greater yields and is required for FASIIbypass 859 growth and survival. A and B . Detection of WTA by Alcian blue staining ( A) and LTA by 860 immunodetection ( B) in non-treated (NT) and anti-FASII-adapted (AD) S. aureus JE2 (N=6 and N=5 861 respectively). Representative gels, and means ± SDs of the independent samples are presented; P-862 values were determined using the two-sided T-test. C. Growth and survival of S. aureus upon bitherapy 863 treatment with anti-FASII (AFN-1252, 0.5 µg/ml) and anti-WTA (Targocil, 15 µg/ml for growth, and 5-864 15 µg/ml for survival). Bacteria were precultured in SerFA prior to simultaneous antibiotic addition. 865 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted April 7, 2026. ; https://doi.org/10.64898/2026.04.06.715823doi: bioRxiv preprint 31 Growth was monitored on a Tecan plate reader. Survival after 15 h growth was determined by serial 866 dilution plating of cultures (5 µL per spot) that were treated with one or two antibiotics as indicated 867 above each column. 868 869 Supplementary Figure Legends 870 871 Supplementary Figure S1. S. aureus has two ways to produce FAs for phospholipid synthesis: FASII 872 and FASIIbypass. A. Both FASII and FASIIbypass provide FAs for phospholipid synthesis, but with different 873 outcomes. The FASII pathway is energetically costly, and provides acyl-acyl carrier protein (FA-ACP) for 874 phospholipid synthesis. FASII bypass incorporates environmental FAs (eFA) that produce FA-ACP via 875 reverse PlsX activity. LPA, lysophosphatidic acid; PA, phosphatidic acid; GroP is represented by a 876 burgundy line joined to a P representing the phosphate group. B. Growth of S. aureus via FASII in SerFA 877 (non-treated, orange), or upon exposure to a FASII inhibitor (AFN-1252 0.5 µg/ml , green curve), when 878 it incorporates exogenous FAs. An initial lag period is followed by robust and sustained growth in a 879 non-mutational response [33, 35]. Growth curves are representative of >20 determinations. Refers to 880 information from Introduction. 881 882 Supplementary Figure S2. Transmission electron microscopy of S. aureus JE2 in non-treated and anti-883 FASII-adapted growth at 6 and 10 hours. Bar, 200 nm. Micrographs complement those shown in Fig. 884 2. 885 886 Supplementary Figure S3. Kinetic heatmap of S. aureus envelope synthesis proteins whose levels are 887 altered during FASII bypass. Proteomics analyses were performed previously on S. aureus USA300 JE2 888 strain grown in SerFA, and treated or not with the FabI inhibitor triclosan (0.5 µg/ml), performed on 889 biological quadruplicates (from [35]). Sampling times correspond to 2, 4, 6, 8, and 10 h post anti-FASII-890 treatment. The heat map shows changes in detected LTA and WTA biosynthetic enzymes, and D-891 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted April 7, 2026. ; https://doi.org/10.64898/2026.04.06.715823doi: bioRxiv preprint 32 alanylation enzymes, which mediate decoration of both structures [76]. Changes in protein expression 892 were determined relative to weighted values for each protein (scale at left: navy, down-represented 893 and yellow, up-represented). WTA biosynthetic enzymes TarL’ (also called TarK) and TarL are 894 redundant [77]. Data is derived from [35]. 895 896 Supplementary Figure S4. LTA remains depleted after prolonged incubation in anti-FASII-adapted S. 897 aureus. A. S. aureus JE2 was grown in SerFA medium without and with AFN-1252 and samples were 898 harvested at 6 or 10 hours, or after overnight (ON) growth. The ON anti-FASII-adapted cultures were 899 sub-cultured (Subc 1) into fresh SerFA-AFN or SerFA medium (indicated by curved arrows) and grown 900 3 h. A second subculture (Subc1 to Subc 2) was prepared in the same condition. Whole cell extracts 901 were prepared and LTA was detected by immunoblotting using anti-LTA antibody. LTA remains 902 depleted even after long term adaptation (N=3). B. Samples from A were extracted for FA analyses at 903 the indicated OD 600 or time. Red, exogenous FAs; black, major endogenous FAs (N=2). Supports data 904 from Fig. 3. 905 906 Supplementary Figure S5. LTA detection in S. aureus NCTC 8325 derivatives. S. aureus RN4220 and 907 HG001 are both from the NCTC 8325 lineage, which lacks the functional fakB1 gene required for 908 complete exogenous FA incorporation [78]. Repair of fakB1 generated RN-R and HG1-R respectively 909 [64]. Cultures were grown in SerFA without (non-treated, NT) and with AFN-1252 (anti-FASII-adapted, 910 AD). LTA production was detected by immunoblotting using anti-LTA antibody (N=3). HG1-R samples 911 are from a single gel, immunoblot, and exposure time. Supports data from Fig. 3. 912 913 Supplementary Figure S6. Model linking FASII synthesis arrest to GroP metabolite depletion. FASII 914 inhibition due to antibiotics, mutation, and/or exogenous FA inhibition, is compensated by 915 incorporation of exogenous FAs in membranes, schematized here for S. aureus [32-34, 43, 65]. 916 Incorporation requires FakAB for FA phosphorylation [79], and reverse PlsX activity to provide acyl-917 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted April 7, 2026. ; https://doi.org/10.64898/2026.04.06.715823doi: bioRxiv preprint 33 ACP [33, 34]. This rerouted cycle is proposed here to desynchronize acyl-ACP and lysophosphatidic acid 918 (LPA) availability for PlsC-mediated PA synthesis. Accumulated LPA intermediates are unstable, leading 919 to degradation of their Gro-P moieties (salmon color zone; [62]), consistent with Gro-P depletion in 920 anti-FASII-adapted S. aureus (Fig. 6). We propose that FASII blockage leads to GroP breakdown, such 921 that the high demand for GroP from PG turnover to produce LTA cannot be met. Short arrows indicate 922 the coordinate LTA decrease and cardiolipin increase when FASII is inhibited. Strong black arrows, 923 favored reactions; thin black arrows, reduced or inhibited reactions; dashed arrow, multi-step 924 reactions. Red arrows and pink zone highlight reactions leading to Gro-P depletion. The upper part of 925 the figure applies information from [62]. The model supports results from Fig. 6. 926 927 Source data is available for results from Fig. 2A, 2B, Fig. 6B, 6C, and Fig. 7A. 928 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted April 7, 2026. ; https://doi.org/10.64898/2026.04.06.715823doi: bioRxiv preprint Fig. 1 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted April 7, 2026. ; https://doi.org/10.64898/2026.04.06.715823doi: bioRxiv preprint Fig. 2 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted April 7, 2026. ; https://doi.org/10.64898/2026.04.06.715823doi: bioRxiv preprint Fig. 3 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted April 7, 2026. ; https://doi.org/10.64898/2026.04.06.715823doi: bioRxiv preprint Fig. 4 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted April 7, 2026. ; https://doi.org/10.64898/2026.04.06.715823doi: bioRxiv preprint Fig. 5 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted April 7, 2026. ; https://doi.org/10.64898/2026.04.06.715823doi: bioRxiv preprint Fig. 6 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted April 7, 2026. ; https://doi.org/10.64898/2026.04.06.715823doi: bioRxiv preprint Fig. 7 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted April 7, 2026. ; https://doi.org/10.64898/2026.04.06.715823doi: bioRxiv preprint Fig. 8 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted April 7, 2026. ; https://doi.org/10.64898/2026.04.06.715823doi: bioRxiv preprint