Nonreducing Terminal Modifications Determine the Chain Length of Polymannose O Antigens of Escherichia coli and Couple Chain Termination to Polymer Export via an ATP-binding Cassette Transporter
Bibliographic record
Abstract
The chain length of bacterial lipopolysaccharide O antigens is regulated to give a modal distribution that is critical for pathogenesis. This paper describes the process of chain length determination in the ATP-binding cassette (ABC) transporter-dependent pathway, a pathway that is widespread among Gram-negative bacteria. Escherichia coli O8 and O9/O9a polymannans are synthesized in the cytoplasm, and an ABC transporter exports the nascent polymer across the inner membrane prior to completion of the LPS molecule. The polymannan O antigens have nonreducing terminal methyl groups. The 3-O-methyl group in serotype O8 is transferred from S-adenosylmethionine by the WbdDO8 enzyme, and this modification terminates polymerization. Methyl groups are added to the O9a polymannan in a reaction dependent on preceding phosphorylation. The bifunctional WbdDO9a catalyzes both reactions, but only the kinase activity controls chain length. Chain termination occurs in a mutant lacking the ABC transporter, indicating that it precedes export. An E. coli wbdDO9a mutant accumulated O9a polymannan in the cytoplasm, indicating that WbdD activity coordinates polymannan chain termination with export across the inner membrane. The chain length of bacterial lipopolysaccharide O antigens is regulated to give a modal distribution that is critical for pathogenesis. This paper describes the process of chain length determination in the ATP-binding cassette (ABC) transporter-dependent pathway, a pathway that is widespread among Gram-negative bacteria. Escherichia coli O8 and O9/O9a polymannans are synthesized in the cytoplasm, and an ABC transporter exports the nascent polymer across the inner membrane prior to completion of the LPS molecule. The polymannan O antigens have nonreducing terminal methyl groups. The 3-O-methyl group in serotype O8 is transferred from S-adenosylmethionine by the WbdDO8 enzyme, and this modification terminates polymerization. Methyl groups are added to the O9a polymannan in a reaction dependent on preceding phosphorylation. The bifunctional WbdDO9a catalyzes both reactions, but only the kinase activity controls chain length. Chain termination occurs in a mutant lacking the ABC transporter, indicating that it precedes export. An E. coli wbdDO9a mutant accumulated O9a polymannan in the cytoplasm, indicating that WbdD activity coordinates polymannan chain termination with export across the inner membrane. Lipopolysaccharide (LPS) 1The abbreviations used are: LPS, lipopolysaccharide; O-PS, O-antigenic polysaccharide; und-P, undecaprenol phosphate; und-PP, undecaprenol pyrophosphate; AdoMet, S-adenosylmethionine; ABC, ATP-binding cassette; PBS, phosphate-buffered saline. is a unique and abundant glycolipid found in the outer membranes of Gram-negative bacteria. LPS has three structural domains (1Raetz C.R. Whitfield C. Annu. Rev. Biochem. 2002; 71: 635-700Crossref PubMed Scopus (3423) Google Scholar). The hydrophobic lipid A forms the outer leaflet of the outer membrane and is responsible for the endotoxic properties of LPS. A short core oligosaccharide extends from lipid A. In many bacteria, the core is capped with a repeating unit glycan polymer known as the O polysaccharide (O-PS; O antigen). Lipid A is structurally conserved among Gram-negative bacteria, whereas limited variability of the core oligosaccharide is often observed within species. For example, five distinct core structures have been identified in different isolates of Escherichia coli (2Heinrichs D.E. Yethon J.A. Whitfield C. Mol. Microbiol. 1998; 30: 221-232Crossref PubMed Scopus (286) Google Scholar). In contrast, O-PS structures vary extensively within a given species because of differences in the number and type of sugars in the repeat unit and the nature of glycosidic linkages within and between repeat units. Variation of the O-PS forms the basis of the O-antigen serotyping scheme. There are ∼170 O serotypes in E. coli (3Ørskov I. Ørskov F. Jann B. Jann K. Bacteriol. Rev. 1977; 41: 667-710Crossref PubMed Google Scholar). LPS preparations from a given isolate contain a spectrum of molecular species with different sizes. The distribution can range from lipid A core molecules devoid of O-PS to LPS molecules with greater than 100 O-PS repeat units. However, most O-PS-substituted LPS molecules in a preparation fall within a limited size range (i.e. a modal distribution). Lipid A-core and O-PS are synthesized separately at the cytoplasmic face of the inner membrane. The two component parts are subsequently ligated at the periplasmic face of the inner membrane prior to export to the cell surface (reviewed in Ref. 1Raetz C.R. Whitfield C. Annu. Rev. Biochem. 2002; 71: 635-700Crossref PubMed Scopus (3423) Google Scholar). Goldman and Hunt (4Goldman R.C. Hunt F. J. Bacteriol. 1990; 172: 5352-5359Crossref PubMed Google Scholar) first suggested that the modal distribution was established by competition between O-PS polymerization and termination (by ligation). However, subsequent work has implicated specific components of the O-PS biosynthesis systems in regulating the modal distribution of some O-PSs (see below). O-PSs are assembled on a 55-carbon lipid acceptor, undecaprenol phosphate (und-P), and biosynthesis is initiated by transfer of a sugar-1-phosphate residue from its nucleotide diphosphosugar precursor to und-P. Subsequent extension and processing of the und-PP-linked intermediate proceeds through one of three distinct pathways. Although one of these pathways is currently confined to a single example (5Keenleyside W.J. Whitfield C. J. Biol. Chem. 1996; 271: 28581-28592Abstract Full Text Full Text PDF PubMed Scopus (115) Google Scholar), the other two are widespread among Gram-negative bacteria. The two major pathways are termed Wzy (polymerase)-dependent and ATP-binding cassette (ABC) transporter-dependent biosynthesis, respectively. They differ by the mechanisms involved in polymerization of the repeat units and in the process of translocation of the und-PP-linked polymer or intermediates across the inner membrane. In the Wzy-dependent mechanism (reviewed in Ref. 1Raetz C.R. Whitfield C. Annu. Rev. Biochem. 2002; 71: 635-700Crossref PubMed Scopus (3423) Google Scholar), single O-PS repeat units are assembled on an und-P carrier lipid by sequential glycosyltransferase reactions. A translocase (Wzx) then mobilizes und-PP-linked repeat units to the periplasmic face of the inner membrane, where the polymerase (Wzy) assembles those repeat units into und-PP-linked polysaccharide. Chain extension occurs by transfer of the growing glycan from the und-PP carrier to the nonreducing terminus of another und-PP-linked monomer, effectively extending the chain one repeat unit at a time. The extent of polymerization is controlled in a process that is not yet understood by the chain-length regulator protein, Wzz (formerly Rol or Cld) (6Batchelor R.A. Haraguchi G.E. Hull R.A. Hull S.I. J. Bacteriol. 1991; 173: 5699-5704Crossref PubMed Scopus (71) Google Scholar, 7Bastin D.A. Stevenson G. Brown P.K. Haase A. Reeves P.R. Mol. Microbiol. 1993; 7: 725-734Crossref PubMed Scopus (157) Google Scholar, 8Morona R. Van Den Bosch L. Manning P.A. J. Bacteriol. 1995; 177: 1059-1068Crossref PubMed Scopus (159) Google Scholar). Different Wzz proteins confer a characteristic modal distribution of O-PS chain length when expressed in a heterologous Wzy-dependent system (9Klee S.R. Tzschaschel B.D. Timmis K.N. Guzman C.A. J. Bacteriol. 1997; 179: 2421-2425Crossref PubMed Google Scholar, 10Franco A.V. Liu D. Reeves P.R. J. Bacteriol. 1998; 180: 2670-2675Crossref PubMed Google Scholar). The O-PS chains of wzz mutants are composed predominantly of one or two repeat units, and the amount of each molecule is inversely proportional to its size, i.e. fully elongated LPS species are extremely rare (7Bastin D.A. Stevenson G. Brown P.K. Haase A. Reeves P.R. Mol. Microbiol. 1993; 7: 725-734Crossref PubMed Scopus (157) Google Scholar, 11Murray G.L. Attridge S.R. Morona R. Mol. Microbiol. 2003; 47: 1395-1406Crossref PubMed Scopus (172) Google Scholar). In the ABC transporter-dependent pathway, O-PS chains are elongated on the und-PP-linked intermediate by processive glycosyl transfer onto the nonreducing end of nascent polymer (reviewed in Ref. 1Raetz C.R. Whitfield C. Annu. Rev. Biochem. 2002; 71: 635-700Crossref PubMed Scopus (3423) Google Scholar). Polymerization occurs within the cytoplasm, and an ABC transporter is required for export of the nascent polymer to the periplasmic face of the inner membrane, where it is ligated to lipid A-core (12Bronner D. Clarke B.R. Whitfield C. Mol. Microbiol. 1994; 14: 505-519Crossref PubMed Scopus (84) Google Scholar). This system requires neither a Wzy polymerase enzyme nor a Wzz chain-length regulator. However, the resulting O-PSs still exhibit a modal chain-length distribution. Recent structural analysis has identified novel residues at the nonreducing end of some O-PSs synthesized by the ABC transporter-dependent pathway (Ref. 13Vinogradov E. Frirdich E. MacLean L.L. Perry M.B. Petersen B.O. Duus J.O. Whitfield C. J. Biol. Chem. 2002; 277: 25070-25081Abstract Full Text Full Text PDF PubMed Scopus Google and the for the first the by these nonreducing terminal in O-PS chain The polymannan O-PSs of E. coli and O9a for the ABC transporter-dependent structural and conserved biosynthesis to the O-PS biosynthesis The polymannans have nonreducing terminal groups E. Frirdich E. MacLean L.L. Perry M.B. Petersen B.O. Duus J.O. Whitfield C. J. Biol. Chem. 2002; 277: 25070-25081Abstract Full Text Full Text PDF PubMed Scopus Google Scholar, J. G. PubMed Scopus Google Scholar). O-PS structures and are found in coli and K. coli as a of transfer 1997; PubMed Google Scholar). and for the E. coli and K. are The of the E. coli O8 and O9a O-PSs on the lipid carrier have not been fully but a has been on the structural and E. Frirdich E. MacLean L.L. Perry M.B. Petersen B.O. Duus J.O. Whitfield C. J. Biol. Chem. 2002; 277: 25070-25081Abstract Full Text Full Text PDF PubMed Scopus Google Scholar, K. Jann K. J. Bacteriol. 1995; 177: PubMed Google Scholar). O-PS with the conserved of a by transfer of a residue to und-P. This is by the of the G.L. K. J. Bacteriol. 1994; PubMed Google Scholar). activity is not confined to O-PS biosynthesis, and the structural is of the O-PS biosynthesis Escherichia coli and and Scholar). the lipid intermediate as an acceptor, an is assembled by the and K. Jann K. J. Bacteriol. 1995; 177: PubMed Google Scholar). the lipid intermediate to the O-PS The polymer the repeating unit is assembled by the processive activity of one or both of the and The E. coli and O9a structural from the and a single is to the of the O-PS repeat unit and serotype to O9a J. Bacteriol. PubMed Scopus Google Scholar). The and domains are identified in structural of the E. Frirdich E. MacLean L.L. Perry M.B. Petersen B.O. Duus J.O. Whitfield C. J. Biol. Chem. 2002; 277: 25070-25081Abstract Full Text Full Text PDF PubMed Scopus Google Scholar). In this that the WbdD proteins the chain length of the E. coli O8 and O9a polymannans by the nonreducing end of nascent und-PP-linked that these terminal with the export of nascent O-PS across the inner membrane. and bacterial used in this are in I. at in in Scholar) or J. A Scholar). with or or The and added when or on lacking to the of that the by polymannan or or coli K. J. Biochem. PubMed Scopus Google coli P.A. Whitfield C. Mol. Microbiol. 1997; PubMed Scopus Google coli C. G. L. J. Microbiol. Google in a of in by the from the E. coli into and transfer was by and O8 and by for a to the The was by the of on and by that O-PS was dependent on the of in the from and used in for each was by bacterial from a single into the polymerase was used to and was used for of mutants and and from with the or from the with the was with the and as by the enzyme was by the of and of the WbdDO8 and WbdDO9a on by from the of E. coli and respectively. in the used to the in D. J. J. Bacteriol. 1995; 177: PubMed Scopus Google Scholar) to and The for and for wbdDO9a and The WbdDO9a by a the the and and with the of polymerase The and into the used to the resulting at the end of the wbdDO9a of by wbdDO9a was by a from the of the in and it with a the from 1993; Google Scholar) to The cassette is when in the as A the wbdDO9a was from and into the D. J. Bacteriol. 1997; 179: PubMed Scopus Google Scholar) to transfer the into the E. coli O9a onto and and at from the and to for on and at For of a a both and was by from E. coli the and The was by a from the into the used to the into the of the to A the and was with a the from to A the was into the of the to The of for is R.A. 1998; PubMed Scopus Google Scholar). of the on and at in and by analysis of the of to The by at for in and in of The by with on The cell was by at for The membranes from the by at for in of and at In of into of from into polymannan was in of membranes of The reaction was initiated by The at at and the by with of The membranes onto and with of The and in of and the was by The membranes in For of O-PS the in of membranes of of and In one of the was with the and The at for and by the of of The in a for and by with and a with residues by on and as the of membrane preparations with the as the LPS was for by of cell Brown J. Bacteriol. PubMed Google Scholar). to the at 100 for was in PubMed Scopus Google Scholar). LPS was by Biochem. PubMed Scopus Google Scholar). of with of E. coli was with of both E. coli and was used to the O9a in with and a in at of was added to the and was for to of by at in of in and at for by in of PBS, and in of The cell was to the of a with and at for by a modification of a J. J. Bacteriol. 2003; PubMed Scopus Google Scholar). The cell was at for in of in and then at for in in of the O9a O-PS was by the at for with in extensively in PBS, and with in at for The extensively in and with in The in and in on a a and the in the WbdD in the of to the O8 and O9a WbdD proteins from E. coli serotypes O8 and O9a differ in size at and respectively. Although these proteins only limited contain the conserved WbdDO8 and WbdDO9a a of that and domains are to have a of structures as by the A. Van J. 1991; PubMed Scopus Google Scholar). the characteristic repeat B. K. J. Mol. Biol. PubMed Scopus Google Scholar) with hydrophobic residues at a and and or residues at and a and the between and different of are An is found in WbdDO9a only The of the WbdD proteins with The WbdD proteins the most with a from the of that are involved in biosynthesis conserved and have been identified in most and are implicated in and methyl transfer Clarke Biochem. 1994; PubMed Scopus Google Scholar). and identified in the of The WbdDO9a and but was not The WbdDO9a an with the from (Ref. J. 1995; PubMed Scopus Google and The and residues of the kinase and the conserved identified in WbdDO9a The of a kinase in WbdDO8 is in its size with structural of the E. coli O8 and the structurally have the of at the nonreducing of these E. Frirdich E. MacLean L.L. Perry M.B. Petersen B.O. Duus J.O. Whitfield C. J. Biol. Chem. 2002; 277: 25070-25081Abstract Full Text Full Text PDF PubMed Scopus Google Scholar, J. G. PubMed Scopus Google Scholar). The by WbdDO8 and known is with the that WbdD the nonreducing end of the growing und-PP-linked In the K. coli the of the 3-O-methyl group was identified by E. Frirdich E. MacLean L.L. Perry M.B. Petersen B.O. Duus J.O. Whitfield C. J. Biol. Chem. 2002; 277: 25070-25081Abstract Full Text Full Text PDF PubMed Scopus Google Scholar). However, the terminal of the identified methyl group in the K. coli was not that WbdDO9a and kinase the terminal modification both and of WbdD O-PS Chain of from E. coli and exhibit of molecular LPS modal O chain-length and In to the of the in the the of and in of wbdDO9a the of serotype O9a LPS from E. coli that most LPS molecules short chain O-PS The amount of a given LPS species by on as the number of repeating units In the in the E. coli O8 of in E. coli in O-PS chains the that WbdD as an O-PS chain-length regulator. O-PS chain length was not by the heterologous WbdD proteins not The WbdD proteins are specific for a given The LPS resulting from WbdD a on O-PS chain length or an of short O-PS chains for onto lipid these E. coli membrane used to the of wbdDO9a on the of from into polymannan in The membranes from the WbdDO9a a in with membranes from und-PP-linked O-PS intermediates synthesized in by the ABC transporter-dependent mechanism are not ligated to lipid and of und-PP not the observed in a single of extension of intermediates K. Jann B. of I. Scholar). the observed in membranes from wbdDO9a is with of O-PS chain length by WbdD at the of chain Chain of E. coli O8 methyl groups transferred to und-PP-linked in from E. coli with and and the by molecular identified with a of O-PS chain The in size and distribution to the from a reaction and The of in lacking not that the molecular polymannan of in an in reaction with a in O-PS chain length In the of added AdoMet, the A in chain length was in the of AdoMet, and a when the amount of was to in polymannan chain length was The of for the critical of WbdD and in the of chain length in serotype O8 A Chain of E. coli O9a of kinase and domains in the WbdDO9a suggested a for both and in the of the O9a und-PP-linked E. coli O9a O-PS was membranes from E. coli with in with and of molecular only in those both and but not in in was indicating that was dependent on both polymer and phosphorylation. The from as the of chain length was in the of and Chain length was in and but not in those with only The chain length was in an with at a of as the of to and the to between a specific residue and the methyl group structural of the polymannan E. Frirdich E. MacLean L.L. Perry M.B. Petersen B.O. Duus J.O. Whitfield C. J. Biol. Chem. 2002; 277: 25070-25081Abstract Full Text Full Text PDF PubMed Scopus Google Scholar). The nature of the for chain termination and and the in in suggested that these in the prior to export of nascent that this was the and into the of the of on chain length was in The of the ABC transporter that the und-PP-linked and that only those preceding export The in chain length of in reaction synthesized by membranes from the mutant was from that with membranes from the and that chain-length determination export and that chain length was not controlled by the of export of through the ABC the of WbdDO9a in chain-length a coli was by in membranes from E. coli still to O9a polymannan in the chain length was to that by the However, the was not by or and nor was the chain length to in the reaction of the O9a O-PS by WbdD for from the of the wbdDO9a mutant coli was in the of i.e. for O9a polymannan in of the with a wbdDO9a properties not indicating that the was only to the wbdDO9a and not an or of the on the in E. coli and O-PS E. coli and to in and was then added to polymannan The LPS then by analysis of cell The E. coli molecular O9a LPS. In contrast, not exhibit lipid O9a O-PS a wbdDO9a was in on of the molecular O9a LPS was The of membranes from E. coli to O9a polymannan was established by analysis of the in in the preceding analysis of the activity in in the of that the wbdDO9a membranes only activity than those of the The for this are but to resulting from the or of because of of the in the the of O-PS-substituted LPS in E. coli only by an to polymannan in or an to export the polymer for to lipid A-core at the periplasmic face of the membrane. these the of polymannan was by specific for the O9a The E. coli at the cell as for surface LPS. In contrast, E. coli surface a with E. coli E. coli of cell and that the and wbdDO9a mutants synthesized O9a the and was the and the of wbdDO9a mutants to O-PS-substituted LPS is to a in export. The of O-PS chain length is critical for of Gram-negative bacteria. are required for the of an modal chain length distribution and for that most lipid A-core is capped with to and Annu. Rev. Microbiol. PubMed Scopus Google Scholar). The of Wzz (formerly Rol or Cld) in regulating the chain length of O-PS synthesized by Wzy-dependent pathways is in wzz G.L. Attridge S.R. Morona R. Mol. Microbiol. 2003; 47: 1395-1406Crossref PubMed Scopus (172) Google Scholar), E. coli Hull S.I. 1998; PubMed Google Scholar), and serotype R. C. Van Den Bosch L. 2003; PubMed Scopus Google Scholar) to and mutants with chain length are to by and are for E. A. J.A. Microbiol. 2003; PubMed Scopus Google Scholar). In Wzz modal O-PS chain length that is critical for the surface of and and R. C. Van Den Bosch L. 2003; PubMed Scopus Google Scholar, Mol. Microbiol. 1997; PubMed Scopus Google Scholar, Bosch L. Manning P.A. Morona R. Mol. Microbiol. 1997; PubMed Scopus Google Scholar, R. Van Den Bosch L. Microbiol. 2003; PubMed Scopus Google Scholar, R. Van Den Bosch L. Microbiol. 2003; PubMed Scopus Google Scholar). Although the of O-PS chain length vary in different Gram-negative it is that Wzz is in other bacterial O-PS is synthesized in a Wzy-dependent In the Wzy-dependent pathway, the polymerization of O-PS occurs at the periplasmic face of the inner membrane, and of are ligated to lipid the of Wzz on O-PS chain length is in of LPS from a The critical for O-PS chain length in the of bacterial is not to confined to those with Wzy-dependent O-PS biosynthesis, but the differences in the mechanism and of polymerization that a different process involved for chain-length determination in the ABC transporter-dependent The give the first into the mechanism of a novel nonreducing terminal residue a process to the of processive glycosyltransferase However, the process or but an to (i.e. when the chain length the modal In in with the E. coli O8 and O9a this can by the amount of or This a in However, of WbdD in the a in chain and of the enzyme is in Chain termination is to ABC and the enzyme is not the component involved in the the for and in a growing bacterial a The of a conserved with the to one to in a a mechanism of for and They are involved in and and as for molecular in (reviewed in Ref. J. Biol. Full Text Full Text PDF PubMed Scopus Google Scholar). In bacteria, domains are for within as those involved in type systems G. J. Biol. Chem. Full Text Full Text PDF PubMed Scopus Google Scholar, F. F. G. PubMed Scopus Google Scholar, C. J. Brown J. Biol. Chem. 2002; 277: Full Text Full Text PDF PubMed Scopus Google Scholar). In the of the biosynthesis WbdD proteins are the only O-PS biosynthesis components with the are involved in between WbdD proteins to a or or between WbdD and another The two in E. coli structures in K. However, novel nonreducing terminal residues are found in other For example, terminal residues with different linkages are in the and serotypes of E. Frirdich E. MacLean L.L. Perry M.B. Petersen B.O. Duus J.O. Whitfield C. J. Biol. Chem. 2002; 277: 25070-25081Abstract Full Text Full Text PDF PubMed Scopus Google Scholar). O has a unique at the nonreducing terminus E. Perry M.B. J. Biochem. PubMed Scopus Google Scholar). a in the as O is with a residue E. 2002; PubMed Scopus Google Scholar). In the of B. the O-antigen biosynthesis of and A. J. R. K. J. B. PubMed Google Scholar), but the of the terminal residues has not been O is by or J. Biol. Chem. Full Text Full Text PDF PubMed Scopus Google Scholar). The O biosynthesis to two with one to the I. C. J. Van C. A. J. Biol. Chem. Full Text Full Text PDF PubMed Scopus Google Scholar, I. C. J. A. J. Mol. 2003; PubMed Scopus Google Scholar, A. J. Bacteriol. PubMed Scopus Google Scholar), but the O can in the of the terminal residue A. J. Bacteriol. PubMed Scopus Google Scholar). groups are found in K. 1994; PubMed Scopus Google Scholar). Although mutants lacking this residue are not in the to O the with from type to Morona R. Manning P.A. A. PubMed Scopus (159) Google Scholar). An is the to other cell surface by ABC transporter-dependent pathways. Gram-negative have by this type of pathway, and the E. coli group (reviewed in C. Mol. Microbiol. PubMed Scopus Google and Mol. Microbiol. 1996; PubMed Scopus Google Scholar). The are from O antigens by the of terminal lipid A-core have a pathway for translocation to the cell surface that is distinct from LPS C. Mol. Microbiol. PubMed Scopus Google Scholar, C. A. 2003; PubMed Scopus Google Scholar). from ABC transporter-dependent pathways are by processive but the mechanisms involved in termination and chain-length are currently have on the repeat unit and a conserved at the terminus Jann K. Microbiol. 14: Scopus Google Scholar). There is currently structural for nonreducing terminal in the of B. cell an ABC transporter-dependent pathway D. Mol. Microbiol. 1995; PubMed Scopus (115) Google Scholar), as is the with the structural at the of the nonreducing terminal structures have not been An ABC transporter-dependent system is involved in the of of the R. C. J. PubMed Scopus Google Scholar). are found on many bacterial C. PubMed Scopus Google Scholar), and those in G. are with glycan The polymer in is a of repeat units and is with C. A. J. Biol. Chem. 2002; 277: Full Text Full Text PDF PubMed Scopus Google Scholar). The a terminal to the in the E. coli O8 and O9a that the glycan to and residues in the The for biosynthesis of this glycan an ABC transporter and a with a R. C. J. PubMed Scopus Google Scholar). The between this system and ABC transporter dependent O-PS biosynthesis are and that the chain system that is for the first widespread in bacteria. The of is J. for to the
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.000 | 0.000 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.000 | 0.000 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.000 |
| Open science | 0.000 | 0.000 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.000 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one teacher head, not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".