AIGR Gene Hypothesis Deep Research — LysB (Q08694, DROME)

Focus type: computational_prediction · Hypothesis slug: prediction-chitinase-immune Term under evaluation: chitinase activity (GO:0004568) Reference context: doi:10.64898/2026.03.19.712954

Summary

The BioReason-Pro SFT model predicts that Drosophila melanogaster LysB (UniProt Q08694) has chitinase activity (GO:0004568) and two immune roles — negative regulation of innate immune response (GO:0045824) and defense response to Gram-negative bacterium (GO:0050829). Independent evaluation from sequence classification, domain architecture, a custom catalytic-site alignment, primary expression literature, and the current GO annotation record indicates that all three predictions are either wrong or unsupported by the physiological evidence. The overall verdict is REFUTED / OVER-ANNOTATED.

LysB is a glycoside hydrolase family 22 (GH22) C-type lysozyme (muramidase, EC 3.2.1.17) — not a chitinase (EC 3.2.1.14; GH18/GH19). Its molecular function is peptidoglycan hydrolysis. The "chitinase" prediction is best explained as a fold/activity misassignment arising from the well-documented, weak chito-oligosaccharide side-activity that C-type lysozymes possess, together with the general confusion between lysozyme and chitinase glycosidase folds. Physiologically, the Drosophila LysB–E cluster comprises constitutive, midgut-restricted, acidic digestive lysozymes that are transcriptionally repressed (not induced) by bacterial infection and are not expressed in fat body or hemocytes — the opposite of what an induced humoral immune effector or an immune regulator would look like. This makes GO:0045824 unsupported and biologically implausible for LysB, and it substantially weakens the case for GO:0050829, which exists in the GO record only as an ISS "by similarity" carry-over with no experimental basis.

For a curator, the actionable conclusion is: do not add GO:0004568 or GO:0045824; treat the existing GO:0050829 as low-confidence (ISS-only, conflicting with digestive biology) and consider flagging it for review or "non-core" status; and retain the well-supported core terms — GO:0003796 (lysozyme activity, MF) and GO:0005576 (extracellular region, CC) — ideally complemented by a BP term reflecting digestion / bacteriophagy rather than immunity.

Executive Judgment

Verdict: REFUTED / OVER-ANNOTATED.

All three predicted terms fail independent evaluation:

• GO:0004568 (chitinase activity): Refuted. LysB is a GH22 muramidase with a conserved lysozyme catalytic dyad and no GH18/GH19 chitinase domain or active-site motif. The only "chitinase-like" property is the weak chitodextrin side-activity intrinsic to C-type lysozymes, which is captured by the canonical lysozyme catalytic-activity description and does not warrant a chitinase MF term.
• GO:0045824 (negative regulation of innate immune response): Refuted/unsupported. No experimental or credible computational basis; contradicts the constitutive, infection-repressed, midgut-restricted expression profile.
• GO:0050829 (defense response to Gram-negative bacterium): Weakly supported at best; present only as a single ISS "by similarity" annotation, in tension with the digestive role.

The most important caveat is that a minority of insect c-type lysozymes are genuinely dual-function (digestive + mild immune effector), so a low-level bactericidal activity for LysB cannot be fully excluded — but even the strongest reading of that literature would never justify a negative regulator of immunity assignment, and the Drosophila-specific expression data argue against an immune-effector role.

Key Findings

Finding 1 — LysB is a GH22 C-type lysozyme (EC 3.2.1.17), not a chitinase

UniProt Q08694 carries the recommended name "Lysozyme B", EC number 3.2.1.17 (lysozyme/muramidase), and the family SIMILARITY statement "Belongs to the glycosyl hydrolase 22 family." The domain architecture is exclusively C-type lysozyme with no chitinase signature whatsoever:

No GH18 or GH19 chitinase domain (e.g., Pfam PF00704 / Glyco_hydro_18, or Glyco_hydro_19) is present. The GO molecular-function annotation for Q08694 is GO:0003796 lysozyme activity, not GO:0004568 chitinase activity.

Critically, the mechanistic origin of the model's error is visible in the enzyme's own catalytic-activity description. The canonical C-type lysozyme reaction hydrolyzes the β-1,4 linkage between N-acetylmuramic acid (MurNAc) and N-acetylglucosamine (GlcNAc) in peptidoglycan — "and between N-acetyl-D-glucosamine residues in chitodextrins." This weak activity on chito-oligosaccharides is an intrinsic, well-known side-property of C-type lysozymes and is almost certainly the textual/feature signal that led a machine-learning model to over-predict full endochitinase activity. A weak in-vitro side-activity on short chitodextrins is not equivalent to biological chitinase activity (processive hydrolysis of insoluble chitin polymer), which is carried out by structurally distinct GH18/GH19 enzymes.

This finding is anchored in the review-level statement that "Lysozymes from family 22 of glycoside hydrolases are usually part of the defense system against bacteria" (PMID: 17825580), confirming that C-type/digestive insect lysozymes belong to GH22 — a family distinct from the chitinase families.

Finding 2 — Drosophila LysB is a constitutive midgut digestive lysozyme, not an induced immune effector

The defining expression study of the Drosophila lysozyme locus (PMID: 8159165, Daffre et al. 1994) reports that "four closely related genes, LysB, C, D and E, are all strongly expressed in the midgut of larvae and adults" and that "None of the genes is expressed in the fat body or haemocytes." In insects, the fat body and hemocytes are the principal tissues of the systemic humoral immune response; the complete absence of LysB expression there is direct evidence against a humoral/immune-effector role. The same work notes that all of these genes except LysP encode acidic proteins "highly reminiscent" of the acidic digestive lysozymes of mammalian foregut fermenters — a hallmark of enzymes optimized for acidic gut lumen conditions and bacteriolytic digestion.

An independent study of the same locus (PMID: 1588905, Kylsten et al. 1992) found that lysozyme transcription decreases after bacterial injection and concluded that these genes "have been recruited for the digestion of bacteria present in fermenting food." A gene whose transcription is repressed by infection and whose product acts in the gut lumen to digest dietary bacteria is behaving as a digestive enzyme, not as an inducible immune effector and certainly not as a negative regulator of the innate immune response.

Consistent with the primary literature, UniProt Q08694 states in its FUNCTION field that LysB is "Unlikely to play an active role in the humoral immune defense. May have a function in the digestion of bacteria in the food," and its TISSUE SPECIFICITY field localizes expression to the midgut with peak expression in the 3rd larval instar.

The digestive-lysozyme model is reinforced by comparative work across arthropods: housefly (Musca domestica) digestive lysozymes MdL1/MdL2 are acidic-optimum midgut enzymes (PMID: 19099150), the first crystal structure of a digestive lysozyme was solved from housefly (PMID: 17825580), and synanthropic mites use acidic digestive lysozymes to exploit bacteria as a food source, with maximal activity at the acidic pH of the gut (PMID: 18357505). These establish "acidic midgut digestive lysozyme" as a recurrent, well-characterized functional class into which Drosophila LysB clearly falls.

Finding 3 — Sequence analysis confirms the C-type lysozyme catalytic dyad and the absence of a chitinase active site

To test the classification directly rather than relying on database labels, an in-run global alignment (Needleman–Wunsch, implemented in NumPy) was performed between the mature LysB sequence (122 aa, after removing the signal peptide, residues 1–18) and hen egg-white lysozyme (HEWL), the archetypal GH22 C-type lysozyme:

The conservation of both HEWL catalytic residues (the Glu general acid and the Asp that stabilizes/participates in catalysis) together with the eight-cysteine (four-disulfide) C-type lysozyme fold is diagnostic of a functional GH22 muramidase. The absence of the GH18 chitinase catalytic signature (the DxxDxDxE motif that defines the GH18 active site) is direct negative evidence against the chitinase prediction: LysB does not possess the active-site chemistry required for a genuine chitinase.

Finding 4 — The current GO record contains no chitinase or immune-regulation term; the sole defense term is ISS-only

A QuickGO retrieval (performed in-run) returned 6 annotations for Q08694 spanning three distinct terms:

Two facts are decisive for the curation decision:

1. GO:0004568 (chitinase activity) is entirely absent from the current annotation set, and GO:0045824 (negative regulation of innate immune response) is entirely absent. Both predicted terms would therefore be new additions, and neither is justified by any experimental or well-founded computational evidence.
2. The one immune-adjacent term that does exist — GO:0050829 — rests on a single ISS "by similarity" annotation with no experimental support, and it is in tension with the digestive, infection-repressed expression profile documented in Finding 2. It should be treated as low-confidence, not as corroboration of the seed hypothesis.

Mechanistic Model / Interpretation

The seed hypothesis effectively bundles two distinct claims, and both fail for the same underlying reason — conflation of superficially related categories:

  BioReason-Pro prediction                Reality (this analysis)
  ─────────────────────────               ────────────────────────────────
  MF: chitinase activity      ──►  GH22 C-type lysozyme (muramidase, EC 3.2.1.17)
      (GO:0004568)                  · Glu35/Asp52 dyad conserved (41% id to HEWL)
                                     · 8 Cys / 4 disulfides, lysozyme fold
                                     · NO GH18/GH19 chitinase domain or motif
                                     · "chitinase" = weak chitodextrin SIDE-activity
                                       intrinsic to C-type lysozymes → misread as
                                       full endochitinase

  BP: neg. reg. innate         ──►  Constitutive DIGESTIVE lysozyme
      immune response                · midgut-restricted; NOT in fat body/hemocytes
      (GO:0045824)                   · transcription REPRESSED by infection
  BP: defense vs Gram-neg.           · "recruited for digestion of bacteria in food"
      (GO:0050829, ISS-only)         · immune terms = digestive-vs-immune conflation

Molecular axis (lysozyme vs chitinase). Both lysozymes and chitinases are glycoside hydrolases that cleave β-1,4 glycosidic bonds in GlcNAc-containing polymers, so they share a substrate chemistry motif. But they are different enzyme families with different folds and active sites: GH22 (lysozyme) versus GH18/GH19 (chitinase). C-type lysozymes have a documented, weak ability to hydrolyze short chitodextrins — this is even written into the UniProt/Rhea catalytic-activity string for Q08694. A model trained to associate GlcNAc-hydrolysis language with "chitinase" will over-predict GO:0004568 for a lysozyme. The direct evidence (conserved lysozyme dyad, correct fold, absent chitinase motif) refutes a genuine chitinase assignment.

Physiological axis (digestive vs immune). Insect lysozymes occupy two functionally divergent niches: (i) inducible immune-effector lysozymes expressed in fat body/hemolymph and up-regulated by infection, and (ii) constitutive digestive lysozymes expressed in an acidic midgut and used to lyse and digest dietary/gut bacteria (the "bacteriophagy" strategy shared with ruminants, foregut-fermenting mammals, and acaridid mites). Drosophila LysB belongs unambiguously to class (ii): midgut-restricted, acidic, constitutive, and infection-repressed. A negative regulator of innate immunity (GO:0045824) would be a signaling/regulatory role entirely unlike a secreted gut hydrolase; nothing in the evidence supports it, and the model likely generated it by over-generalizing "lysozyme → immunity" and then inverting it to fit a repression signal.

The coherent model is therefore: LysB is a secreted, extracellular (GO:0005576), acidic midgut muramidase (GO:0003796) whose biological process is the digestion of bacteria in food (bacteriophagy), not immune defense or immune regulation.

Evidence Matrix

Evidence Base (literature summary)

GO Curation Implications

Lead (requires curator verification):

1. GO:0004568 (chitinase activity, MF) — DO NOT ADD. The prediction is a fold/activity misassignment. LysB is a GH22 muramidase with a conserved lysozyme catalytic dyad and no GH18/GH19 chitinase domain or active-site motif. The only "chitinase-like" property is the intrinsic weak chitodextrin side-activity of C-type lysozymes, which is already captured by the lysozyme catalytic-activity description and does not warrant a chitinase MF term (in vitro-only, side-activity, not the biological function).

2. GO:0045824 (negative regulation of innate immune response, BP) — DO NOT ADD. No experimental or credible computational support. The term is biologically implausible for a secreted digestive hydrolase and contradicts the constitutive, infection-repressed, midgut-restricted expression profile.

3. GO:0050829 (defense response to Gram-negative bacterium, BP) — FLAG / treat as non-core. Present in the record only as a single ISS "by similarity" to HEWL-type P00697, with no experimental basis, and in tension with digestive biology. Recommend curator review; at minimum it should not be treated as core function, and it should not be cited as corroboration of the immune hypothesis.

4. GO:0003796 (lysozyme activity, MF) — RETAIN (core). Well supported by four independent evidence lines (ISS, IBA, IEA/InterPro-EC, and ISS to 8159165) and by the in-run catalytic-dyad analysis. This is the correct primary molecular function.

5. GO:0005576 (extracellular region, CC) — RETAIN. Consistent with a secreted gut lysozyme.

6. Consider ADDING a digestion-focused BP term (curator lead): a term reflecting digestion / catabolism of bacterial cell wall in the gut / bacteriophagy would more accurately capture the biological process than any immune term. Candidate directions include a peptidoglycan catabolic process or digestion term; the exact choice should be curator-verified against GO structure and the 8159165 / 1588905 evidence.

Aspect summary: MF = lysozyme activity (retain), not chitinase (reject). BP = digestion/bacteriophagy (add, curator-verify), not immune regulation (reject) and not defense-response as core (downgrade). CC = extracellular region (retain).

Mechanistic Scope

The immediate molecular function under test is glycoside-hydrolase activity on a GlcNAc-containing polymer. The direct gene-product activity of LysB is muramidase/lysozyme activity — hydrolysis of the MurNAc–GlcNAc β-1,4 bond in bacterial peptidoglycan — as established by family classification, domain content, the conserved Glu/Asp catalytic dyad, and EC 3.2.1.17. This is distinct from chitinase activity, which requires the GH18/GH19 fold and active site that LysB lacks.

Everything in the immune framing of the seed hypothesis is either downstream/indirect or absent: - "Defense response to Gram-negative bacterium" is an inferred phenotype-level consequence, imported by similarity, not a demonstrated function of LysB. - "Negative regulation of innate immune response" is a regulatory/signaling process category with no mechanistic link to a secreted digestive hydrolase; there is no evidence LysB participates in immune signaling. - The digestive role (lysing dietary bacteria in the acidic midgut) is the direct physiological deployment of the muramidase activity — a nutritional/digestive process, not an immune-defense process, even though the underlying chemistry (bacteriolysis) is shared with immune lysozymes.

Conflicts and Alternatives

• Paralog / family carry-over confusion. The Drosophila lysozyme locus contains a tandem cluster (LysB–E, plus LysP, LysS, etc.). Annotations and predictions can bleed across paralogs and across the broader "lysozyme/glycosidase" family. The GO:0050829 ISS to HEWL (P00697) is a concrete example of defense-response carry-over from a canonical immune lysozyme to a digestive one.

• Dual-function orthologs are real but do not rescue the immune prediction. Housefly MdLys (PMID: 19161941) is expressed in midgut and fat body, is mildly infection-induced, and its recombinant protein inhibits bacteria — i.e., some insect c-type lysozymes are genuinely dual digestive/immune. This is a legitimate competing consideration. However, it does not apply to Drosophila LysB, which — unlike MdLys — is not expressed in fat body/hemocytes and is repressed (not induced) by infection (PMID: 8159165; PMID: 1588905). Even the strongest immune interpretation would be a mild bactericidal effector activity, never negative regulation of the immune response (GO:0045824).

• Weak in-vitro chitodextrin activity vs biological chitinase. C-type lysozymes can slowly hydrolyze short chito-oligosaccharides. Treating this in-vitro side-activity as a biological chitinase function would be an over-annotation; the physiological substrate is peptidoglycan.

• Organism/tissue specificity. Evidence supporting digestion comes partly from orthologs (housefly, mites). While the Drosophila-specific expression data are decisive against the immune-effector/regulator interpretation, direct enzymatic assays of purified Q08694 protein are not in the reviewed literature set (see Knowledge Gaps).

Limitations and Knowledge Gaps

1. No direct enzymatic assay of purified LysB (Q08694). Checked: the primary literature reviewed here characterizes the locus expression (8159165 1588905) and orthologous digestive lysozymes (housefly, mites), but not a purified recombinant LysB kinetic assay. This matters because the definitive discriminator between lysozyme and chitinase, and between digestive and immune roles, is a direct activity/pH-profile assay on the LysB protein. Resolution: recombinant expression + muramidase vs chitinase activity assays and an acidic-pH-optimum profile.

2. Alignment used a single reference (HEWL). The in-run NW alignment (41% identity, conserved dyad) is robust for classification but is a custom implementation against one archetype. Resolution: a multiple-sequence alignment across GH22 members and a structural superposition (e.g., AlphaFold model vs HEWL and vs a GH18 chitinase) would strengthen the fold assignment. Structural tools (Phenix/superpose) were available but not required given the concordant sequence/domain evidence.

3. GO:0050829 provenance. It is ISS-by-similarity to P00697; the exact rationale and whether FlyBase intends it as core is not fully resolved from the record alone. Resolution: curator inspection of GO_REF:0000024 and the FlyBase gene report.

4. No in-run structural confidence metrics (pLDDT/PAE). An AlphaFold confidence check was not run; it would add independent support for the lysozyme fold but is not necessary to reach the verdict.

5. Reference doi:10.64898/2026.03.19.712954 (the source of the BioReason-Pro prediction) was not independently retrieved; the evaluation treats the prediction as stated in the seed context.

Discriminating Tests

The following would most efficiently separate the seed hypothesis from the digestive-lysozyme interpretation:

1. Enzymatic assay of recombinant LysB. Measure muramidase activity on Micrococcus lysodeikticus cell walls vs endochitinase activity on a fluorogenic chitin substrate (e.g., 4-MU-chitotrioside) and on insoluble/colloidal chitin. Prediction: strong acidic-pH muramidase activity, negligible/only weak short-chitodextrin activity, no processive chitin degradation. This directly refutes GO:0004568.

2. pH–activity profile. Digestive insect lysozymes have acidic optima (~4.8, cf. housefly MdL1/MdL2, PMID: 19099150). An acidic optimum supports a gut-digestive role.

3. Infection-challenge transcriptomics. RNA-seq/qPCR of LysB after septic injury vs oral bacterial feeding, in gut vs fat body. Prediction: no fat-body induction, repression in gut — inconsistent with an immune effector and with a negative regulator of immunity.

4. Loss-of-function phenotype. LysB knockdown/knockout scored for gut bacterial digestion/clearance and nutritional phenotype vs systemic immune-response readouts (AMP induction, survival to infection). Prediction: digestive/nutritional phenotype, not altered immune signaling.

5. Structural superposition. AlphaFold model of LysB superposed on HEWL (GH22) and on a GH18 chitinase (TIM-barrel). Prediction: RMSD-close to HEWL, fold-incompatible with the chitinase TIM-barrel.

6. Comparative catalytic-residue mapping across the Drosophila Lys cluster and canonical GH18/GH19 chitinases to confirm the absence of the chitinase catalytic glutamate context in all LysB-family members.

Proposed Follow-up Actions (Curation Leads)

All items are leads requiring curator verification.

Candidate action changes - Reject proposed GO:0004568 (chitinase activity) and GO:0045824 (negative regulation of innate immune response) — do not add. - Downgrade / flag GO:0050829 (defense response to Gram-negative bacterium) as ISS-only, non-core, conflicting with digestive expression data. - Retain GO:0003796 (lysozyme activity, MF) and GO:0005576 (extracellular region, CC) as core. - Consider adding a digestion/bacteriophagy BP term (curator to select the exact term; e.g., a peptidoglycan catabolic / digestion term).

Candidate references with snippets to verify - PMID: 8159165: "four closely related genes, LysB, C, D and E, are all strongly expressed in the midgut of larvae and adults" and "None of the genes is expressed in the fat body or haemocytes." → supports digestive, refutes immune-effector. - PMID: 1588905: "have been recruited for the digestion of bacteria present in fermenting food" → supports digestive; note transcription decreases after bacterial injection. - PMID: 17825580: "Lysozymes from family 22 of glycoside hydrolases are usually part of the defense system against bacteria" → confirms GH22 family placement (distinct from chitinase families). - UniProt Q08694 FUNCTION: "Unlikely to play an active role in the humoral immune defense. May have a function in the digestion of bacteria in the food."

Suggested curator questions - Should GO:0050829 remain given it is ISS-only and conflicts with midgut-digestive, infection-repressed biology? - Is there an appropriate GO BP term for gut bacteriophagy/digestion that should replace the immune framing? - Does FlyBase hold any direct enzymatic data on Q08694 that would confirm the acidic muramidase profile?

Suggested experiments — see Discriminating Tests above (recombinant muramidase-vs-chitinase assay; acidic pH profile; challenge transcriptomics; LOF phenotype; structural superposition).

Bottom Line

LysB (Q08694) is a GH22 C-type digestive lysozyme (muramidase, EC 3.2.1.17). The BioReason-Pro predictions of chitinase activity (GO:0004568) and negative regulation of innate immune response (GO:0045824) are refuted / over-annotated: the chitinase call is a fold/side-activity misassignment, and the immune-regulation call conflates a constitutive, infection-repressed, midgut-restricted digestive enzyme with an immune regulator. The pre-existing GO:0050829 annotation is ISS-only and should be treated as low-confidence, non-core. Curators should retain GO:0003796 (lysozyme activity) and GO:0005576 (extracellular region) and consider adding a digestion/bacteriophagy BP term rather than any immune term.