Caulobacter Fur–Lipid A Loss

Lipid A is an essential component of the outer membrane in nearly all Gram-negative bacteria. Caulobacter crescentus can survive complete loss of lipid A only when fur is also inactivated. The published mechanism (Zik et al. 2022, PMID 35649364) shows the rescue requires anionic sphingolipids (CPG) and a ΔsspB co-deletion. This project characterizes the regulatory and proteomic architecture of that rescue using RNA-seq + OM proteome of the rescued and intermediate strains, BERDL Caulobacter RB-TnSeq fitness data (198 experiments), a re-analysis of Leaden 2018 (SRP136695) for a clean Fur-only signature, and a cross-species comparative arm covering the three other Gram-negative species known to tolerate lipid A loss (N. meningitidis, A. baumannii, M. catarrhalis). Four hypotheses are tested — see RESEARCH_PLAN.md.

Hypothesis scorecard

Phase A — orientation findings (NB00) that motivated the analysis plan

NB00 motivated three reframings before formal hypothesis testing: (a) the sphingolipid biosynthesis pathway is not induced (rejecting the initial "Δfur derepresses sphingolipid biosynthesis" framing); (b) the canonical Lpt apparatus is maintained or up, consistent with Uchendu 2026's shared-component model; (c) SdpA at +4.8 log2 OM proteome surfaced peptidoglycan remodeling as a fourth hypothesis (H4) that wasn't in the v1 plan.

NB01 — Leaden 2018 Fur signature

The scatter shows the dramatic asymmetry between our amplified Fur-derepression cohort (top-right quadrant, exceeding Leaden's logFC) and the buffered cohort (cluster near y=0 spanning Leaden's full -9 to -5 range). HutA (the iron-derepressed TBDT) is the extreme top-right point at our +10.5 vs Leaden's +2.2.

NB02 — Fitness data leverage

The Caulobacter compendium covers 95 stress experiments (envelope-, drug-, metal-related stresses), 46 nitrogen-source, 42 carbon-source, 10 PYE control, plus 5 others. Zero pure iron-limitation experiments. The envelope-stress subset (22 experiments) is large enough to compare gene-set phenotype-bearing rates against the genome background (33.25%). The pre-registered ≥10% threshold sits below background — it is essentially a presence test rather than an enrichment test. Hypergeometric enrichment against background is the formal H2 verdict in the revised report (see Limitations). Path A clears the background rate marginally (1.60×, p=0.016); Path B does not (1.04×, p=0.515).

NB03 — ChvI phase partition

ChvI itself (CCNA_00237) is in the both-phases set with logFC +1.45 in 4584-vs-4580 and continued induction in 4599-vs-4584 — direct evidence of ChvG-ChvI autoregulation during the Fur+SspB release phase. (Earlier draft mis-labeled this as "early cohort"; ChvI is in the 10-gene both-phases subset.) Theme distribution shifts from regulator-rich early (6.7% regulator/TCS in the unique-to-early 20) to envelope-structural late (10.2% envelope/OM, 12.2% TBDT in the late 49, with lipid/membrane and PG/cell wall categories at 0% in early). The Fisher exact test of envelope/transport/regulator enrichment in late vs unique-to-early is not significant (p=0.243), so the "regulator → structural" shift is suggestive but not statistically discriminated.

NB04 — Sphingolipid, CtpA, Lpt panel

NB05 — PG remodeling heatmap

NB06 — Comparative species presence/absence (PaperBLAST original) + NB06b NCBI confirmation

The original NB06 used PaperBLAST description-text matching. The adversarial review (I4) correctly flagged ~50% false-negative rate for known Caulobacter essential genes. NB06b (added post-review) re-ran the comparative panel against NCBI's protein database via Biopython Entrez, which uses curated genome annotation rather than text matching:

PaperBLAST false negatives in C. crescentus confirmed (NB06b vs NB06 for genes that PaperBLAST said 0 and NCBI confirms present):

Biological claims CONFIRMED by NCBI annotation (where NB06 and NB06b agree, pattern 1000 = Caulobacter only):

The sphingolipid biosynthesis pathway absence in A. baumannii, N. meningitidis, M. catarrhalis — and the ChvG-ChvI absence — are now independently supported by NCBI annotation in addition to PaperBLAST text matching and the published literature (Olea-Ozuna 2020/2024; Greenwich 2023). The headline claim of Finding 6 is robust.

NCBI annotation also has its own false negatives in C. crescentus for genes annotated under non-canonical names. For example, MsbA (NCBI Cc=0) is present as CCNA_00307 ("phospholipid/LPS ABC transporter"); LptA (NCBI Cc=0) is present but not annotated with the "lptA" symbol in Caulobacter NA1000. These are naming-convention false negatives, not true absences — Caulobacter is genome-annotated primarily with CCNA_* locus tags rather than canonical Gram-negative gene symbols. The sphingolipid + ChvI findings are unaffected by this because the corresponding C. crescentus genes ARE annotated with the canonical symbols (NCBI Cc=7 for spt, 5 for ChvG, 6 for cerR/ChvI).

Mechanistic synthesis: what the data support vs what remains hypothesis

The combined evidence supports a multi-layer model for Caulobacter crescentus Δfur ΔsspB-permitted Δlpxc viability. After adversarial review, the layers are stated at their actual evidence levels:

1. Δfur derepresses TBDT and iron-uptake systems (SUPPORTED). The Path A concordant_strong subset (32 genes) is marginally enriched for envelope-stress phenotype-bearing (53% observed vs 33% background; fold 1.60×, hypergeometric p=0.016). Top hits ChvT, HutA, and the FrpB cluster encode OM transport machinery whose normal job is iron acquisition. In the rescued state, this machinery is plausibly available to participate in sphingolipid (CPG) trafficking via shared Lpt apparatus (Uchendu 2026), although that participation is not directly demonstrated here.

2. ΔsspB buffers the cbb3 / cyd / fix-NOPQ micro-aerobic respiratory chain transcript-level decline (HYPOTHESIS, not statistically supported as mechanistically critical). The transcript-level buffering is real (NB01 concordance — 53/93 Leaden Δfur DEGs blunted in our data, dominated by the cbb3/fix operon). But the Path B fitness phenotype-bearing rate (34.6%) is indistinguishable from the genome background (33.25%, hypergeometric p=0.515). The fitness data therefore do not selectively support the cbb3/fix genes as more critical than randomly drawn Caulobacter genes under envelope stress. "Respiratory ATP required to perform envelope-remodeling work" remains a mechanistically plausible model but is not established by the data presented here.

3. ChvI engages in two phases plus a cooperative continuation (PARTIALLY SUPPORTED). Phase structure is real (20 unique-to-early, 10 both-phase, 49 late). ChvI itself sits in the both-phase set with logFC +1.45 in the early contrast — direct autoregulation evidence. The "regulator-rich early → envelope-structural late" theme shift is suggestive but not statistically discriminated (Fisher p=0.243).

4. Sphingolipid biosynthesis is constitutive — rescue does NOT require biosynthesis upregulation (SUPPORTED strongly). Zero biosynthesis genes are significantly UP; spt and sphk are mildly DOWN. The existing CPG pool suffices. CtpA's specific role as the LpxF-equivalent processing step (Zik 2022's prediction) is not supported by our data at the pre-registered bar (CtpA transcript pvalue=0.048 in 4599-vs-4584, FDR=0.109, protein not detected); the cumulative 4599-vs-4580 contrast is significant (FDR=0.035) but confounds Δfur and Δlpxc effects. CtpA's LpxF-substitute hypothesis (Zik 2022) remains untested.

5. Lpt apparatus repurposing in the Δlpxc state shows transcript-protein discordance (MIXED). At the transcript level: MsbA-like CCNA_00307 +0.89 (FDR 0.01) and LptC-related CCNA_03716 +0.56 (FDR 0.005) are significantly upregulated, consistent with Uchendu 2026's shared-component model. At the protein level: the two canonical Lpt proteins actually detected — LptD and LptE — decline in the rescued strain (LptD log2(4672/4659)=−0.47, LptE=−0.78). Whether this reflects reduced demand for LPS-specific transport (canonical Lpt being substrate-limited in absence of LPS) or contradicts the shared-component model cannot be resolved by single-replicate proteomics.

6. Sphingolipid-specific lptC2 protein induction (PILOT, requires replication). Single-replicate OM proteome shows lptC2 (CCNA_01226) protein log2(4672/4659) = +1.08 despite transcript -0.60 (FDR 0.034). Net change vs WT baseline is more modest (+0.66 log2, ≈1.58×) because lptC2 was already DOWN -0.42 in the intermediate strain. Suggestive of post-transcriptional stabilization but not statistically established; requires the replicated proteomics scheduled for summer 2026.

7. Peptidoglycan reorganization: specific lytic engagement on a backdrop of broad basal shutdown (SUPPORTED). Shutdown dominates (20 of the 28 H4 hits are DOWN — FtsI, PbpZ, MurD, multiple endopeptidases and amidases). Against that backdrop, SdpA at +4.8 log2 OM protein, PleA, PbpX, and the Pal-Tol envelope-integrity factor Pal stand out as specific inductions.

8. Pal-Tol upregulation has a parsimonious mechanistic interpretation distinct from the earlier draft (REFRAMED). Earlier drafts attributed Pal upregulation to "compensation for LPS-mediated Mg²⁺-bridged OM-LPS-Pal stacking" — a claim with no cited support. Tan & Chng 2025 (Nat Commun 16:2293, PMID 40055349) established that the primary function of the Tol-Pal complex is retrograde phospholipid transport to maintain OM lipid homeostasis, not LPS-Pal structural anchoring. A mechanistically supported interpretation: in Δlpxc, OM lipid asymmetry is disrupted (loss of LPS in the outer leaflet); upregulation of the Tol-Pal complex (including Pal itself) is consistent with increased retrograde PL transport to restore OM lipid homeostasis. Direct Pal-PG contacts (Yeh et al. 2010, PMID 20693330) may contribute additionally, particularly because Caulobacter Tol-Pal is essential for OM constriction at cell division.

9. Cross-species check (SUPPORTED at multiple levels). A. baumannii, N. meningitidis, M. catarrhalis do not encode the sphingolipid biosynthesis pathway — established by (a) NCBI annotation in NB06b (spt: 7,0,0,0; cerR: 6,0,0,0; bcerS query failed in C.c. but other markers confirm), (b) phylogenetic distribution (none are alphaproteobacteria or Bacteroidetes), and (c) Olea-Ozuna 2020/2024 characterizing the Caulobacter cluster. ChvG-ChvI is alphaproteobacterial-restricted — confirmed by both PaperBLAST and NCBI annotation (ChvG: 5,0,0,0; ChvI: 6,0,0,0). The earlier "NB06 measurement fragile" caveat is upgraded: NB06b NCBI annotation re-tests the comparative arm and confirms the headline biological claims, while also exposing the ~80% false-negative rate of PaperBLAST text matching for Caulobacter's own lipid A biosynthesis genes (LpxA/C/D/B/K all returned 0 from PaperBLAST but 11-18 each from NCBI). Because the headline claim concerns absence in the comparators (where NCBI confirms), this PaperBLAST issue does not threaten the cross-species finding.

Literature Context

• Zik et al. 2022 (PMID 35649364) — the foundational paper, authored by this project's data provider K.R. Ryan. Establishes that Δlpxc viability in C. crescentus requires both Δfur and anionic sphingolipid (CPG), and that "Fur-regulated processes (not iron status per se)" underlie viability. Our project characterizes the regulatory and proteomic constituents of this rescue at a level Zik 2022 did not. Specific evidentiary extensions: (a) ranking which Fur regulon members are mechanistically critical (NB02 fitness), (b) confirming the sphingolipid pathway is constitutive at the transcript level rather than induced (NB04), (c) demonstrating the canonical Lpt apparatus is maintained or upregulated (NB04), (d) the novel lptC2 protein-level induction.

• Uchendu, Isom, Klein 2026 (bioRxiv 10.1101/2026.04.12.717747) — identifies the Caulobacter sphingolipid IM transporters CCNA_01213/01214/01226 (lptG2/F2/C2) and shows they share the canonical LptB ATPase with LPS transport. Our project provides the first regulatory and proteomic evidence that this shared-component model operates in a Δlpxc strain: the canonical Lpt apparatus is maintained, MsbA-like and LptC-related components are upregulated, and the Uchendu lptC2 protein accumulates >2-fold despite transcript downregulation.

• Leaden et al. 2018 (PMID 30210482) — published Caulobacter Δfur RNA-seq. Our NB01 re-analyzed their supplementary Table 2 to provide a clean Fur-only DEG signature. The Spearman ρ = 0.315 concordance with our 4584-vs-4580 confirms Fur derepression as a major component of our signal. The buffered cbb3/fix respiratory operon is a new observation made possible only by comparing our Δfur ΔsspB combined to Leaden's Δfur-alone.

• Stein et al. 2021 (PMID 34124942) and Quintero-Yanes et al. 2022 (PMID 36480504) — published Caulobacter ChvI regulons used as reference sets for the H1 partition test. Greenwich et al. 2023 (PMID 37040790) — alphaproteobacterial ChvG-ChvI conserved-circuit review, consistent with our cross-species finding that ChvG-ChvI is absent in A.b./N.m./M.c.

• da Silva Neto et al. 2009 (PMID 19520766) — earlier Caulobacter Fur ChIP/microarray. Cross-validates the concordant_strong gene set (HutA, FrpB-like cluster, bacterioferritin) as canonical Fur targets.

• Kang et al. 2021 (PMID 33402533) — A. baumannii Δlpxc via PBP1A loss + LdtJ/LdtK. Mechanistically distinct from but biologically analogous to our finding that Caulobacter Δlpxc engages SdpA/PleA + Pal-Tol PG-remodeling. Both organisms reorganize PG when LPS/LOS is removed, but via different enzyme-family specifics.

• Steeghs et al. 2001 (PMID 11742971) — N. meningitidis ΔlpxA via capsule substitution. Consistent with our NB06 finding that N. meningitidis has 9 capsule-biosynthesis PaperBLAST hits (vs 2-3 in the others).

• Gao et al. 2008 (PMID 18795947) — M. catarrhalis late-acyltransferase truncation. Consistent with our NB06 finding that the lpxX/lpxL family is present in A.b. (8) and N.m. (3) but absent in C.c. (0).

• Olea-Ozuna et al. 2020/2024 (PMIDs 33063925, 39093898) — independent confirmation that Caulobacter encodes a sphingolipid biosynthesis pathway absent in the other three Gram-negatives.

• Tan WB & Chng SS 2025 (PMID 40055349, Nat Commun 16:2293) — establishes the primary function of the Tol-Pal complex as retrograde phospholipid transport for OM lipid homeostasis, not LPS-Pal structural anchoring as earlier draft of this report supposed. This 2025 paper grounds the revised mechanistic interpretation of the Pal upregulation finding (Mechanistic Synthesis §8).

• Yeh YC, Comolli LR, Downing KH, et al. 2010 (PMID 20693330, J Bacteriol 192:4847) — establishes that the Caulobacter Tol-Pal complex is essential for OM constriction at cell division (Caulobacter Tol-Pal is essential unlike E. coli Tol-Pal). Relevant context for the Pal upregulation finding, particularly because the rescued strain must complete division with no LPS.

Novel Contribution

After adversarial review, the substantive contributions are reframed to match the actual evidence level:

1. The ΔsspB co-deletion buffers a specific transcript-level program that Δfur alone would otherwise repress — chiefly the cbb3/fix-NOPQ micro-aerobic respiratory operon (NB01: 53 of 93 Leaden Δfur DEGs blunted; dominated by CCNA_01466–01476). This is a clean novel transcriptomic observation that depended on comparing Δfur ΔsspB to Leaden's Δfur-alone. Mechanistic significance for the Δlpxc rescue remains hypothesis: the Path B fitness data do not statistically discriminate these genes from the genome background under envelope stress.

2. The canonical Lpt apparatus shows transcript upregulation with protein-level discordance in the Δlpxc state (NB04). MsbA-like CCNA_00307 +0.89 (FDR=0.01) and LptC-related CCNA_03716 +0.56 (FDR=0.005) at transcript level support Uchendu 2026's shared-component prediction. The two canonical Lpt proteins actually detected (LptD, LptE) decline at the protein level in the rescued strain (-0.47, -0.78 log2). Replicated proteomics are needed to resolve whether the apparatus is functionally maintained, substrate-limited, or downregulated at the protein level.

3. The sphingolipid biosynthesis pathway is constitutive rather than induced in the Δlpxc rescue (NB04). spt and sphk are mildly DOWN, not UP, at transcript level. The cell does not respond to lipid-A loss by making more CPG-biosynthesis transcript — rescue must be flux-driven or post-transcriptional. This rules out the "Δfur derepresses sphingolipid biosynthesis" framing.

4. The ChvI envelope-stress regulon engages in two phases with cooperative continuation (NB03). The partition: 20 unique-to-early, 10 both-phase, 49 late. ChvI itself sits in the both-phase set with logFC +1.45 in the early contrast — autoregulation evidence. The full disjoint partition was not previously characterized.

5. Suggestive (single-replicate) post-transcriptional sphingolipid-transporter induction. lptC2 (CCNA_01226) shows transcript -0.60 (FDR 0.034) and protein log2 = +1.08 vs intermediate (+0.66 vs WT) — single replicate, requires the proteome replication scheduled for summer 2026.

6. Pal-Tol upregulation in the Δlpxc state, with revised mechanistic interpretation. Pal CCNA_00784 +2.08 transcript + 2.84 protein (NB05 + NB03 late cohort). Following Tan & Chng 2025 (PMID 40055349), the parsimonious interpretation is that Pal upregulation supports increased retrograde phospholipid transport for OM lipid homeostasis (Tol-Pal's primary function), not LPS-Pal structural anchoring as earlier draft proposed.

7. A general methodological lesson (added in adversarial review): pre-registered fitness-phenotype thresholds in BERIL Fitness Browser projects must be calibrated against the genome-wide background rate rather than a fixed percentage. A ≥10% threshold against a 33% background passes any random gene set. The plan's H2 threshold structure should be revised in future projects to use fold-enrichment over background.

Limitations

• CtpA is REJECTED at the pre-registered bar. The 4599-vs-4584 transcript test yielded pvalue=0.048 (fails BORDERLINE lower bound of 0.05), FDR=0.109 (fails PASS bar of 0.05), and CtpA was not detected in the single-replicate OM proteome. The cumulative 4599-vs-4580 contrast FDR=0.035 is significant but conflates Δfur + ΔsspB and Δlpxc effects; the pre-registered 4599-vs-4584 contrast was designed to isolate the Δlpxc-specific response. CtpA's hypothesized LpxF-substitute role (Zik 2022) remains untested by this work. Earlier drafts mis-labeled this as BORDERLINE; corrected here.
• OM proteome is single-replicate per strain — no per-protein statistics; only direction is interpretable. The lptC2 protein-induction finding, the Pal-Tol upregulation, the LptD/LptE protein-level decline, and the CCNA_01217 protein increase all need the replicated proteomics scheduled for summer 2026 before being claimed at publication rigor.
• Path B (SspB-buffered) fitness phenotype-bearing rate is indistinguishable from genome background. The "respiratory ATP required" arm of the dual-release switch model is therefore a working hypothesis, not an established finding. The pre-registered ≥10% threshold sat below the 33.25% background rate — a methodological miscalibration that the adversarial review surfaced and that has been recorded as a learned pattern (see Novel Contribution §7).
• Single growth condition (PYE rich-medium routine) for the transcriptome — the Fur signal is constitutive Δfur derepression, not a real iron-limitation response. BERDL fitness data supplied multi-condition resolution for H2's envelope arm but iron-limitation experiments specifically are unavailable in the Caulobacter FB compendium.
• Pal upregulation interpretation depends on Tan & Chng 2025, not on the project's data alone. The data show Pal up; the mechanistic interpretation (retrograde PL transport for OM lipid homeostasis) is grounded in the cited paper's redefinition of Tol-Pal function and remains an inference, not a direct measurement. Earlier drafts proposed an "LPS-Mg²⁺-bridged stacking" mechanism that was uncited and inconsistent with current Tol-Pal biology; this has been replaced.
• SigU regulon literature gap — Caulobacter SigU (CCNA_02977) is uncharacterized in PaperBLAST. The H1 SigU-as-driver test was reframed from a literature overlap to a functional coherence check on the late cohort; even the relaxed criterion did not pass (Fisher p=0.243). Targeted SigU-induction RNA-seq would resolve the cohort attribution.
• NB06 PaperBLAST has ~80% false-negative rate for known Caulobacter essential lipid A biosynthesis genes (LpxA, LpxC, LpxD, LpxK all 0 hits in PaperBLAST; NCBI confirms 11-18 each). NB06b (NCBI annotation re-test, added post-review) re-confirmed the headline biological claims — sphingolipid pathway absent in A.b./N.m./M.c., ChvG-ChvI alphaproteobacterial-restricted — and explicitly diagnosed the PaperBLAST false-negative pattern. The original NB06 should now be read as a screen; NB06b is the authoritative comparative measurement for the headline claim.
• M. catarrhalis is under-annotated in PaperBLAST (162 genes total). Treat its presence/absence rows as advisory only.
• Fitness-browser iron-limitation experiments are absent in the Caulobacter compendium (zero of 198 experiments). H2's iron-limitation arm was descoped to exploratory per the plan's preflight rule. Iron-axis testing requires either additional RB-TnSeq experiments under bipyridyl chelation or a cross-walk to Leaden 2018's published Δfur phenotype.
• NB03 cohort double-counting in earlier drafts: the previous report described an "early cohort of 30 genes" that included the 10 both-phase genes. Corrected here as a disjoint partition (20 unique-to-early + 10 both-phase + 49 late). The phase-structure verdict is unchanged.
• NB05 PG-remodeling regex false positives: CCNA_00565 (γ-glutamyltranspeptidase, a glutathione enzyme) and CCNA_01833 (glucosylceramidase, a ceramide enzyme) entered the pre-curated set via description-text matching. They are present in the union of significant hits but do not change the H4 verdict (well above the ≥3 threshold). Flagged here so downstream interpretation does not lean on them.

1. Replicate the OM proteome (already planned by the data provider for summer 2026). The two novel post-hoc findings — lptC2 protein induction (transcript-protein decoupling) and Pal-Tol upregulation — need replication before publication. Targeted Western blots against lptC2-tagged and Pal-tagged strains would also strengthen these claims.

2. Characterize the Caulobacter SigU regulon via SigU-induction RNA-seq (ectopic vanillate-inducible SigU expression) followed by genome-wide DE. The published literature contains no Caulobacter SigU regulon; this is a clear gap. The late ChvI cohort from NB03 provides a candidate target list to validate.

3. Test the dual-release-switch model genetically. The model predicts: (a) Δlpxc should NOT be viable on a Δfur alone background (sspB intact) because the cbb3/fix respiratory chain would shut down; (b) Δlpxc should NOT be viable on a ΔsspB alone background (fur intact) because Fur would repress the transport machinery needed for sphingolipid trafficking. Both predictions are testable in the existing Ryan-lab strain background.

4. Lipidomics of the rescued strain. The post-transcriptional flux model predicts: in 4599 (Δlpxc Δfur ΔsspB), the sphingolipid pool size should be maintained or increased despite no biosynthesis induction, because LpxC competition for shared substrates is gone. The CtpA-mediated processing should produce a specific CPG headgroup (the LpxF substitute step). Targeted lipidomics with intact lipid A precursor quantification would test both predictions.

5. Cross-species engineering test. The "structural unavailability" model predicts that introducing the Caulobacter sphingolipid biosynthesis pathway into A. baumannii (which lacks it) should be neutral or modestly protective under colistin selection. The PBP1A/Ldt route in A. baumannii works independently of sphingolipids, so a sphingolipid-engineered A. baumannii should have two alternative routes to colistin resistance. Predicts increased rates of colistin-resistant escapers.

6. Iron-axis fitness experiments. The Caulobacter RB-TnSeq compendium has zero iron-limitation experiments, so the iron-flux arm of H2 was untestable here. A small targeted set of RB-TnSeq fitness experiments under bipyridyl chelation, ferric salt supplementation, and hemin would close the H2 iron axis and test whether the same Fur-released subset that scores phenotype-bearing under envelope stress also scores under iron limitation.

7. Tol-Pal retrograde phospholipid-transport assay in the rescued strain. The revised Pal interpretation (Tan & Chng 2025) predicts that the Tol-Pal complex in the Δlpxc state should have increased retrograde phospholipid-transport activity to maintain OM lipid homeostasis. Pulse-chase lipidomics following exogenous fluorescent phospholipid loading would directly test this. Tol-Pal essentiality in Caulobacter (Yeh 2010) complicates depletion experiments but is testable in conditional-depletion strains.

8. Sequence-based homology search for the comparative arm (deeper than NB06b annotation). NB06b's NCBI annotation re-test confirmed the headline biological claims, but annotation-based search misses unannotated paralogs. For full rigor, run Pfam HMM searches (PF00155 serine palmitoyltransferase, PF03331 LpxC, PF01475 Fur) against named RefSeq proteomes (GCF_000196795.1 A. baumannii ATCC 17978, GCF_000008805.1 N. meningitidis MC58, GCF_000092265.1 M. catarrhalis BBH18) using hmmsearch or pyhmmer. This would close the residual concern that an unannotated spt paralog in one of the comparators was missed. Estimated 2-4 hours via CTS.

Sources

Generated Data

• Zik JJ, Yoon SH, Guan Z, et al., Klein EA, Ryan KR. (2022). "Caulobacter lipid A is conditionally dispensable in the absence of fur and in the presence of anionic sphingolipids." Cell Reports 39(11):110888. PMID 35649364. DOI: 10.1016/j.celrep.2022.110888
• Uchendu CG, Isom GL, Klein EA. (2026, preprint). "Homologues of the inner-membrane LPS transport proteins are required for sphingolipid transport in Caulobacter crescentus." bioRxiv 2026.04.12.717747. DOI: 10.1101/2026.04.12.717747
• Dhakephalkar T, Guan Z, Klein EA. (2025). "CpgD is a phosphoglycerate cytidylyltransferase required for ceramide diphosphoglycerate synthesis." PMID 39829823.
• Dhakephalkar T, Stukey GJ, Guan Z, Carman GM, Klein EA. (2023). "Characterization of an evolutionarily distinct bacterial ceramide kinase from Caulobacter crescentus." J Biol Chem 299:104894. PMID 37286040.
• Olea-Ozuna RJ, Poggio S, et al., Geiger O. (2020). "Five structural genes required for ceramide synthesis in Caulobacter and for bacterial survival." Environ Microbiol 23(5):2741. PMID 33063925.
• Olea-Ozuna RJ, Poggio S, et al., Geiger O. (2024). "Genes required for phosphosphingolipid formation in Caulobacter crescentus contribute to bacterial virulence." PLoS Pathog 20(8):e1012401. PMID 39093898.
• Hummels KR. (2024). "mSphere of Influence: Celebrating exceptions to the rule of lipid A essentiality." mSphere 9(3):e00633-23. PMID 38421175.
• Stein BJ, Fiebig A, Crosson S. (2021). "The ChvG-ChvI and NtrY-NtrX Two-Component Systems Coordinately Regulate Growth of Caulobacter crescentus." J Bacteriol 203(15):e00199-21. PMID 34124942.
• Quintero-Yanes A, Mayard A, Hallez R. (2022). "The two-component system ChvGI maintains cell envelope homeostasis in Caulobacter crescentus." PLoS Genet 18(12):e1010465. PMID 36480504.
• Greenwich JL, Heckel BC, Alakavuklar MA, Fuqua C. (2023). "The ChvG-ChvI Regulatory Network: A Conserved Global Regulatory Circuit Among the Alphaproteobacteria with Pervasive Impacts on Host Interactions and Diverse Cellular Processes." Annu Rev Microbiol 77:339. PMID 37040790.
• Leaden L, Silva LG, et al., Marques MV. (2018). "Iron Deficiency Generates Oxidative Stress and Activation of the SOS Response in Caulobacter crescentus." Front Microbiol 9:2014. PMID 30210482.
• da Silva Neto JF, Braz VS, Italiani VCS, Marques MV. (2009). "Fur controls iron homeostasis and oxidative stress defense in the oligotrophic alpha-proteobacterium Caulobacter crescentus." Nucleic Acids Res 37(14):4812. PMID 19520766.
• Moffatt JH, Harper M, et al., Boyce JD. (2010). "Colistin resistance in Acinetobacter baumannii is mediated by complete loss of lipopolysaccharide production." Antimicrob Agents Chemother 54(12):4971. PMID 20855724.
• Kang KN, Kazi MI, et al., Boll JM. (2021). "Septal Class A Penicillin-Binding Protein Activity and ld-Transpeptidases Mediate Selection of Colistin-Resistant LOS-Deficient Acinetobacter baumannii." mBio 12(1):e02185-20. PMID 33402533.
• Steeghs L, de Cock H, et al., van der Ley P. (2001). "Outer membrane composition of a LPS-deficient Neisseria meningitidis mutant." EMBO J 20(24):6937. PMID 11742971.
• Gao S, Peng D, et al., Gu XX. (2008). "Identification of two late acyltransferase genes responsible for lipid A biosynthesis in Moraxella catarrhalis." FEBS J 275(20):5201. PMID 18795947.
• Bhat NH, Vass RH, Stoddard PR, Shin DK, Chien P. (2013). "Identification of ClpP substrates in Caulobacter crescentus reveals a role for regulated proteolysis in bacterial development." Mol Microbiol 88(6):1083. PMID 23647068.
• Flynn JM, Levchenko I, Sauer RT, Baker TA. (2004). "Modulating substrate choice: the SspB adaptor delivers a regulator of the extracytoplasmic-stress response to the AAA+ protease ClpXP for degradation." Genes Dev 18(18):2292.
• Price MN, Wetmore KM, Waters RJ, Callaghan M, Ray J, Liu H, Kuehl JV, Melnyk RA, Lamson JS, Suh Y, et al. (2018). "Mutant phenotypes for thousands of bacterial genes of unknown function." Nature 557(7706):503. PMID 29769718. — primary citation for the BERDL kescience_fitnessbrowser data source.
• Price MN, Deutschbauer AM, Arkin AP. (2021). "PaperBLAST: Text Mining Papers for Information about Homologs." mSystems 6(1):e00185-19. PMID 33531404. — primary citation for the BERDL kescience_paperblast data source.
• Arkin AP, Cottingham RW, Henry CS, Harris NL, et al. (2018). "KBase: The United States Department of Energy Systems Biology Knowledgebase." Nature Biotechnology 36(7):566-569. PMID 29979655. — primary citation for the KBase infrastructure.
• Tan WB, Chng SS. (2025). "Primary role of the Tol-Pal complex in bacterial outer membrane lipid homeostasis." Nature Communications 16(1):2293. PMID 40055349. DOI: 10.1038/s41467-025-57630-y. — basis for the revised Pal-upregulation interpretation (Mechanistic Synthesis §8); identified by the adversarial review as a missing citation.
• Yeh YC, Comolli LR, Downing KH, Shapiro L, McAdams HH. (2010). "The Caulobacter Tol-Pal complex is essential for outer membrane integrity and the positioning of a polar localization factor." Journal of Bacteriology 192(19):4847-4858. PMID 20693330. DOI: 10.1128/JB.00607-10. — Caulobacter-specific Tol-Pal context (essentiality and division-associated role).

See references.md for the full bibliography with PMCIDs, DOIs, and theme-grouping.

• Adam Arkin (University of California, Berkeley) — ORCID 0000-0002-4999-2931. Lead analyst.
• Kathleen R. Ryan (University of California, Berkeley) — data provider and scientific collaborator; co-authorship pending publication discussion.