One afternoon of titrations that closes a 1,300-fold question: tagless 1147-1203 against tagless 1147-1244, matched in every other respect
also called the patch titration · the missing control in a 24-year literature
Nobody has ever compared the short and long versions of the fusion's DNA-gripping module in the same tube on the same day. Making both proteins and titrating them against the same DNA at the same salt would settle the disagreement that has stood since 2002.
The same thing, for a scientist
The residual gap in the CXXC-DNA affinity ladder is 19-fold between Risner's GST-1147-1203 at 0.64 uM and Birke's GST-1147-1244 at 33 nM, confounding construct length, CpG count, duplex length and assay geometry. The experiment is a 2x2x2 matched titration: two constructs (1147-1203 and 1147-1244), both tagless after cleavage of the same tag from the same vector, against two DNA ligands (a 12-mer with two CpGs and the 52-bp six-CpG CpG52 duplex from Birke), at two ionic strengths (150 and 300 mM NaCl), all by ITC or MST in one session with one protein preparation per construct. Eight titrations, one afternoon, standard reagents. The single most informative pair is tagless 1147-1203 versus tagless 1147-1244 on the 12-mer at 150 mM. No published experiment contains this comparison and the whole affinity literature for this domain would be interpretable if it existed.
Design
Constructs. KMT2A Q03164 1147-1203 and 1147-1244, both in the same vector with the same cleavable tag (His-SUMO or His-TEV, not GST - the point is to remove the valency variable, and Erfurth already showed thrombin-cleaved material behaves). Verify monomeric by SEC-MALS: this is the control Birke could not have run and Risner did not, and it prices the GST term directly if a GST-tagged arm is included as a fifth sample.
DNA. (a) GGGTCGCGGGAG, the Hoxa9 CpG1-I 12-mer used by both Erfurth and Risner, which ties the new numbers into the existing ladder at two points; (b) Birke's CpG52 52-bp six-CpG duplex, which ties it in at the other end. Include the methylated CpG52 as the specificity control.
Salt. 150 mM NaCl and 300 mM NaCl, so the salt term is priced for the first time. If budget allows, five points from 100 to 400 mM on one construct-ligand pair gives the slope of log Kd against log[Na+] and hence the number of ion pairs released - the quantity that determines how far any of these numbers can be extrapolated to a nucleus.
Readout. ITC is preferred because it gives stoichiometry, which is the variable that betrayed the multi-CpG ligands in Erfurth's hands (N = 1.93 on a three-CpG 16-mer). MST or fluorescence anisotropy is an acceptable substitute at lower protein cost and does not give N.
Matrix. 2 constructs x 2 ligands x 2 salts = 8 titrations, plus the methylated control and the GST-tagged arm. Roughly 10 mg of each protein for ITC, far less for MST.
What each result means, decided in advance
| result on the 12-mer at 150 mM | interpretation |
|---|---|
| 1147-1244 within 2-fold of 1147-1203 | the patch hypothesis is dead. Birke's 19-fold is ligand valency and surface geometry; the minimal CXXC is the operative module; every structure is of the right thing. |
| 1147-1244 tighter by 5- to 20-fold | the patch carries a substantial part of the grip. The operative DNA-binding module of the fusion is the MT/BCB domain and the structural literature has been working on a truncation. Strengthens the address hypothesis - affinity and specificity live in different pieces. |
| 1147-1244 tighter by more than 20-fold | the patch dominates, and the CXXC fold is closer to a specificity filter bolted onto a generic polycation. Would require re-reading the entire structure-based-design case. |
| either construct on CpG52 reaching 33 nM | Birke's number is reproducible in solution and is not an SPR surface artefact at all - which would be the most consequential single outcome, because it makes the tight number the physiological one. |
And the negative control that decides whether any of it is specific: the methylated CpG52. Birke reports greater than 1 mM, i.e. a discrimination of over 30,000-fold; if the tagless construct reproduces even three orders of magnitude of that, the measurement is sound.
| design | 2 constructs (Q03164 1147-1203, 1147-1244; identical cleavable non-oligomerising tag, both verified monomeric by SEC-MALS) x 2 DNA ligands (Hoxa9 CpG1-I 12-mer GGGTCGCGGGAG; Birke CpG52 52-bp six-CpG duplex) x 2 ionic strengths (150, 300 mM NaCl), plus a methylated-CpG52 specificity control and an uncleaved-GST arm to price the tag directly. Optional five-point salt series on one pair to obtain the slope of log Kd against log[Na+]. Add an EMSA arm at 150 mM NaCl on the short construct to tie the new numbers to Liang 2021's published 5.7 uM comparator. |
|---|---|
| readout | ITC dissociation constant and stoichiometry; MST or fluorescence anisotropy as a lower-material substitute without stoichiometry |
| controls | SEC-MALS on every protein preparation to establish oligomeric state before any titration; methylated CpG52 as the specificity negative; the uncleaved-GST arm reproduces Risner's and Birke's tag state within the same session; an oligonucleotide devoid of CG as the Erfurth-style blank |
| cost tier | low |
| timeline | two weeks including cloning and expression; the titrations themselves are one afternoon |
| what it closes | the only unresolved step in ana-cxxc-dna-affinity-ladder, a 19-fold gap confounding four variables |
| why nobody has done it | the two constructs belong to two literatures that do not cite each other and are reachable by disjoint PubMed queries - see concept-mt-domain-and-cxxc-domain-are-two-names-for-one-region-and-the-literatures-do-not-intersect |
| shown for fusion | NO - this is an isolated-domain biophysics experiment by design; the fusion-context version is the CUT and RUN arm listed under the hypothesis's predictions. |
decision table
[
{
"result": "1147-1244 within 2-fold of 1147-1203 on the 12-mer at 150 mM",
"conclusion": "hyp-the-post-cxxc-basic-patch-carries-part-of-the-address is falsified; Birke's 19-fold is ligand valency plus surface geometry"
},
{
"result": "1147-1244 tighter by 5-20 fold",
"conclusion": "the operative DNA-binding module is the MT/BCB domain; the structural literature has been working on a truncation; strengthens the address hypothesis"
},
{
"result": "1147-1244 tighter by more than 20-fold",
"conclusion": "the patch dominates and the CXXC fold is a specificity filter on a generic polycation; the structure-based design case must be re-read"
},
{
"result": "either construct reaches 33 nM on CpG52 in solution",
"conclusion": "Birke's number is not an SPR surface artefact and the tight number is the physiological one - the most consequential outcome available"
}
]existing comparator
{
"kd_nM": 5700,
"conditions": "tagless synthesised CXXC, 12-bp single-CpG palindrome, 10 mM Tris pH 7.4, 150 mM NaCl, 50 uM ZnCl2, EMSA, on ice",
"pmid": "34280370",
"use": "the short arm at 150 mM already has a published value; the experiment must reproduce it before the long arm can be interpreted"
}- A clean answer here does not tell you what the fusion does. Both constructs are isolated domains on naked DNA; if the operative species is a multivalent oncoprotein on a nucleosomal CpG island then this experiment prices a variable that may not be rate-limiting in cells.
- If 1147-1244 turns out to be insoluble or aggregation-prone without GST, the experiment returns no answer and the literature's use of GST will look less like a convenience and more like a necessity - which is itself informative but does not close the gap.
- Should a 1147-1300 construct be included, given that the UniProt disordered region runs to 1375 and 1244 is a PCR boundary rather than a structural one?
- Is a synthetic 1204-1244 peptide titration worth adding as a direct test of whether the patch binds DNA at all on its own?
- Could the same session price the menin comparison by running MBM1-menin at 150 and 300 mM, which would convert ana-the-menin-affinity-comparison-was-made-at-one-sixth-the-salt from a caution into a correction?
- E1strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
No published experiment compares the 1147-1203 and 1147-1244 KMT2A constructs under matched conditions; the only measurement of the longer construct is GST-tagged, on a different DNA, by a different method, on an immobilised ligand.
The MT domain of the proto-oncoprotein MLL binds to CpG-containing DNA and discriminates against methylation. · Birke M, Schreiner S, García-Cuéllar MP, et al. · Nucleic Acids Res 2002What this evidence cannot show
This is an absence claim about the experimental record, established by reading the Methods of the five papers that report an affinity for this domain (PMIDs 11842107, 16990798, 18483194, 20010842, 23990460) rather than by a bounded query - and one further paper, PMID 29276034, could not be read this cycle and may contain a relevant construct comparison. The absence is therefore bounded by five papers read, not by the literature.
Evidence that argues against this (1)
- PMID 29276034 (Xu et al., Structure 2018) systematically measured DNA-binding selectivities of all human CXXC domains by several assays and could not be retrieved through any of the five documented routes. If it contains a construct-length comparison for MLL, part of this experiment is already done and the design should be revised before anyone buys protein. Xu C et al. 2018
- E2established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.paper
Erfurth et al. successfully expressed KMT2A 1147-1203 as a GST fusion in pGEX-4T-2 and removed the tag with thrombin before ITC, establishing that tagless material of this domain is obtainable and behaves in calorimetry.
MLL protects CpG clusters from methylation within the Hoxa9 gene, maintaining transcript expression. · Erfurth FE, Popovic R, Grembecka J, et al. · Proc Natl Acad Sci U S A 2008What this evidence cannot show
Demonstrated for 1147-1203 only. The 98-residue 1147-1244 construct has only ever been made as an uncleaved GST fusion, and the 41-residue disordered basic extension could well reduce solubility or make the protein a protease substrate; that is the main technical risk in this design and a His-SUMO tag is suggested for that reason.
Evidence that argues against this (1)
- Allen et al. needed a lipoyl-domain fusion partner and TEV cleavage to get 1146-1214 behaved enough for NMR and ITC, and used a different construct (1136-1208) for gel shifts, which suggests this domain is fussier than the single successful thrombin cleavage implies. Budget for two or three tag strategies. Allen MD et al. 2006
- E3strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
A tagless MLL1 CXXC domain obtained by chemical synthesis binds a 12-bp single-CpG palindrome with Kd 5.7 +/- 0.2 uM at 150 mM NaCl, so tagless material at near-physiological ionic strength is obtainable and the short-construct arm of this experiment already has a published comparator.
Dynamics and inhibition of MLL1 CXXC domain on DNA revealed by single-molecule quantification. · Liang L, Ma K, Wang Z, et al. · Biophys J 2021What this evidence cannot show
Chemical synthesis is feasible for a 57-residue domain and is NOT feasible for the 98-residue MT construct, so the long arm still requires recombinant expression and cleavage. EMSA rather than a solution-equilibrium method, and the titration was on ice; to tie into this comparator the proposed experiment should include an EMSA arm alongside ITC or MST. The construct span is in an unretrieved supplementary table, so the comparator's boundaries are unverified.
Evidence that argues against this (1)
- Having a published comparator at 150 mM makes the short arm cheaper and also raises the bar: if the new tagless 1147-1203 measurement does not land near 5.7 uM, the experiment has a reproducibility problem before it has an answer, and the 1147-1244 result cannot be interpreted. Erfurth FE et al. 2008
The matched tagless titration is the decisive test.
Closes the only unresolved step in the ladder.
Prices the 1204-1244 basic tail directly.
- ← proposed byTen affinity measurements for this contact and not one uses a fusion protein, full-length KMT2A or a nucleosome
The matched titration is a first step out of the absence, though it too uses isolated domains.