The comparison that founded 'the CXXC contact is too weak' puts a 300 mM NaCl protein-DNA measurement against a 50 mM NaCl protein-peptide measurement
also called 44 uM versus 53 nM is not like for like · salt mismatch in the menin comparison
This project says the DNA grip is a thousand times weaker than the menin grip. The two numbers were measured in salt concentrations six times apart, and salt hurts a DNA grip far more than it hurts a protein-protein grip, so the comparison is biased in exactly the direction the argument needed.
The same thing, for a scientist
The load-bearing comparison in this graph is Kd 44 uM for KMT2A CXXC-DNA against Kd 53 nM for menin-MBM1, reported as roughly three orders of magnitude. Reading both Methods sections: the CXXC figure was measured in 50 mM phosphate pH 7.1 with 300 mM NaCl (Cierpicki 2010), and the menin figure in 50 mM Tris pH 7.5 with 50 mM NaCl (Grembecka 2010). That is a six-fold difference in monovalent ionic strength, applied to two interactions with opposite salt sensitivities: the CXXC-DNA contact is entropically driven and counterion-release dominated, with three of its most costly residues being Arg1154, Lys1185 and Lys1193, while the menin-MBM1 contact is dominated by a hydrophobic patch whose single hot spot is Phe9 (alanine substitution costs more than 2,000-fold on IC50). Neither measurement is at physiological ionic strength and they err in opposite directions. The nearest-to-physiological tagless CXXC measurement, 5.7 uM at 150 mM NaCl (Liang 2021), puts the ratio at 108-fold rather than roughly 1,000-fold, and the menin side is still measured at a third of physiological salt. The comparison is not like for like, and no entry in this graph states the buffers.
The two Methods sections, side by side
| CXXC-DNA | menin-MBM1 | |
|---|---|---|
| value | 44 uM 1 | 53 +/- 4.2 nM 2 |
| method | NMR chemical-shift titration | fluorescence polarisation |
| construct A | KMT2A 1147-1203, isolated 57-aa domain | full-length recombinant menin |
| construct B | 12-bp palindrome, one CpG | 12-residue peptide, KMT2A 4-15 |
| buffer | 50 mM phosphate pH 7.1 | 50 mM Tris pH 7.5 |
| monovalent salt | 300 mM NaCl | 50 mM NaCl |
| temperature | 25 C | not stated for FP; ITC at 27 C |
| driving force | entropic, counterion release 3 | hydrophobic patch, Phe9 hot spot, >2,000-fold on IC50 2 |
Why the mismatch is not neutral. A protein-DNA association that works by displacing condensed counterions from the DNA backbone has a Kd that rises steeply with salt; a peptide binding into a hydrophobic pocket does not. Raising one measurement to 300 mM and lowering the other to 50 mM therefore widens the apparent gap on both sides at once, and does so by an amount nobody has measured for either pair.
The thermodynamics are in the literature and confirm the mechanism. Erfurth's five ITC titrations of this exact domain are all entropically driven, with small favourable enthalpies (-1,069 to -1,776 cal/mol) and dS between 15.9 and 19.5 cal/mol/K 3 - the signature of counterion release. Grembecka's alanine scan of MBM1 finds the dominant contribution in a hydrophobic patch, Phe9 above all 2.
What the comparison looks like when the salt is matched as well as it can be
There is now one measurement of each side at or near a defensible ionic strength, and they are still not matched.
- CXXC-DNA at 150 mM NaCl, tagless: 5.7 +/- 0.2 uM (Liang 2021, EMSA, 12-bp single-CpG palindrome, chemically synthesised untagged domain) 4
- CXXC-DNA at 150 mM KCl, GST-tagged: 0.64 uM (Risner 2013, fluorescence polarisation, 12-bp two-CpG Hoxa9 site)
- menin-MBM1 at 50 mM NaCl: 53 nM (Grembecka 2010, fluorescence polarisation, 12-residue peptide) 2
Ratio 108-fold on the tagless pair, 12-fold on the tagged pair - and the menin number still enjoys a third of physiological salt in both comparisons. Taking Birke's MT-domain SPR at 150 mM NaCl instead, the CXXC-DNA contact at 33 nM is tighter than menin-MBM1.
The defensible statement is therefore: no published pair of measurements supports a stated fold-difference between the menin and CXXC contacts, because no two of them share an ionic strength, a construct scale or a method. Anything between 'comparable' and 'a hundred-fold weaker' is consistent with the data; 'three orders of magnitude weaker' is not, and it is what this graph has been saying.
Something else follows that is more interesting than the correction. Grembecka also reports MLL160 at 9.8 nM and MLL46 at 6.8 nM by ITC - the longer KMT2A N-terminal fragments bind menin roughly 6- to 8-fold tighter than the 12-residue MBM1 peptide alone, because MBM2 adds a second contact. So both sides of this comparison get tighter as the construct gets longer and more multivalent, and in a fusion protein both are on the same polypeptide at once. The quantity the project actually wants - the avidity of the intact N-terminus for a menin-plus-LEDGF-plus-CpG-island surface - is not any of these numbers and has never been measured.
| script | none - a reading of two Methods sections |
|---|---|
| status | run |
| result summary | The graph's headline comparison, 44 uM CXXC-DNA against 53 nM menin-MBM1, puts a measurement at 300 mM NaCl against one at 50 mM NaCl - a six-fold ionic-strength mismatch on two interactions with opposite salt sensitivities. With the nearest-to-physiological CXXC measurement (0.64 uM at 150 mM KCl) the ratio is 12-fold, not 1,000-fold; with Birke's MT-domain SPR at 150 mM NaCl the CXXC contact is tighter than menin-MBM1. No published pair of measurements supports any stated fold-difference between these two contacts. |
| entries quoting 53 NM without naming the buffer | 9 |
| recommendation | Any entry stating a fold-difference between the menin and CXXC contacts must name the ionic strength of both measurements, or state the difference as unquantified. Prefer the second. |
inputs
[ "PMID 20010842 Methods (PMC2908503)", "PMID 20961854 Methods (PMC3003368)", "PMID 18483194 Table 1 (PMC2396713)", "PMID 23990460 Methods (PMC3795288)" ]
comparison
{
"cxxc_dna": {
"kd_nM": 44000,
"salt_mM": 300,
"salt_species": "NaCl",
"buffer": "50 mM phosphate pH 7.1",
"method": "NMR titration",
"pmid": "20010842"
},
"menin_mbm1": {
"kd_nM": 53,
"kd_sd_nM": 4.2,
"salt_mM": 50,
"salt_species": "NaCl",
"buffer": "50 mM Tris pH 7.5",
"method": "fluorescence polarisation on FITC-KMT2A 4-15",
"pmid": "20961854"
},
"salt_ratio": 6,
"as_quoted_in_graph": "roughly 1,000-fold",
"best_salt_matched_pair_ratio": 107.5,
"best_salt_matched_pair": "5.7 uM tagless at 150 mM NaCl (Liang 2021, EMSA) against 53 nM at 50 mM NaCl (Grembecka 2010, FP) = 108-fold; with Risner's GST-tagged 0.64 uM at 150 mM KCl it is 12-fold",
"birke_geometry_ratio": 0.62,
"cxxc_dna_nearest_physiological_tagless": {
"kd_nM": 5700,
"salt_mM": 150,
"salt_species": "NaCl",
"method": "EMSA",
"tag": "none",
"pmid": "34280370"
}
}other menin affinities in the same paper — the numbers behind the chart
[
{
"ligand": "MLL160 (KMT2A 1-160)",
"kd_nM": 9.8,
"method": "ITC"
},
{
"ligand": "MLL46 (KMT2A 1-46)",
"kd_nM": 6.8,
"method": "ITC"
},
{
"ligand": "MLL15 (KMT2A 1-15)",
"kd_nM": 90,
"kd_sd_nM": 27,
"method": "ITC"
},
{
"ligand": "MBM1 (KMT2A 4-15) peptide",
"kd_nM": 53,
"kd_sd_nM": 4.2,
"method": "FP"
},
{
"ligand": "MBM2 (KMT2A 23-40) peptide",
"kd_nM": 1400,
"kd_sd_nM": 420,
"method": "FP"
},
{
"ligand": "MBM1 F9A peptide",
"kd_nM": 80000,
"method": "FP, estimated"
}
]- If the menin and CXXC contacts are comparable rather than a thousand-fold apart, blocking menin removes a smaller fraction of the total binding energy than this project assumes, which is a mechanistic reason a menin inhibitor might relocate the fusion rather than evict it.
- Conversely, if a drug displaces one of several comparable contacts, the dose-response will be shallow and incomplete - which is what is clinically observed with menin inhibitors, and this analysis makes that expected rather than surprising.
- What is the salt dependence of the KMT2A CXXC-DNA interaction? A five-point salt series gives the slope of log Kd against log [Na+] and converts this analysis from a mechanism-based prediction into a number.
- What is the menin-MBM1 Kd at 150 mM NaCl? If it is genuinely salt-insensitive the comparison can be rescued in one direction; nobody has measured it.
- What is the avidity of the intact KMT2A N-terminus - MBM1 plus MBM2 plus AT-hooks plus CXXC - for a surface presenting menin, LEDGF and a CpG island simultaneously? That is the quantity every argument in this project actually needs and it has never been measured in any form.
- E1established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.paper
The 44 uM CXXC-DNA dissociation constant was determined by NMR titration in 50 mM phosphate pH 7.1 with 300 mM NaCl and 2 mM DTT at 25 C.
Structure of the MLL CXXC domain-DNA complex and its functional role in MLL-AF9 leukemia. · Cierpicki T, Risner LE, Grembecka J, et al. · Nat Struct Mol Biol 2010What this evidence cannot show
Quoted verbatim from the paper's Methods, read this cycle via the PMC HTML route. 300 mM NaCl is a standard and defensible choice for NMR of a small basic domain with DNA; the criticism here is of how the resulting number has been used downstream, not of the experiment.
Evidence that argues against this (1)
- Ionic strength is not the only mismatch and may not be the largest one. The CXXC number is a 57-residue domain against a 12-bp duplex; the menin number is a full-length protein against a 12-residue peptide. Attributing the difference chiefly to salt is itself an untested attribution, and no salt titration of either interaction has been published. Grembecka J et al. 2010
- E2established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.paper
The 53 +/- 4.2 nM menin-MBM1 dissociation constant was determined by fluorescence polarisation on a fluorescein-labelled KMT2A 4-15 peptide in 50 mM Tris pH 7.5 with 50 mM NaCl and 1 mM DTT; the same paper gives MLL160 at 9.8 nM and MLL46 at 6.8 nM by ITC, MBM2 at 1.4 +/- 0.42 uM, and identifies Phe9 as the hot spot with an alanine substitution costing more than 2,000-fold on IC50.
Molecular basis of the mixed lineage leukemia-menin interaction: implications for targeting mixed lineage leukemias. · Grembecka J, Belcher AM, Hartley T, et al. · J Biol Chem 2010What this evidence cannot show
Full text read this cycle via the PMC HTML route (PMC3003368); the FP buffer is quoted verbatim from the Fluorescence Polarization Experiments section. FP temperature is not stated in the text read. The 53 nM figure is quoted in nine entries in this graph and none of them states the buffer.
Evidence that argues against this (1)
- At 50 mM NaCl the menin-MBM1 figure is itself not a physiological measurement, so this analysis does not establish that menin binds more weakly than reported either - only that the two numbers cannot be divided by one another. If the menin contact is largely hydrophobic it may be nearly salt-independent, in which case the mismatch penalises only the CXXC side and the correction is smaller than the six-fold salt ratio suggests. Erfurth FE et al. 2008
- E3strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
All five ITC titrations of the KMT2A CXXC domain with CpG oligonucleotides are entropically driven, with enthalpies between -1,069 and -1,776 cal/mol and entropies between 15.9 and 19.5 cal/mol/K, the thermodynamic signature of a counterion-release-dominated protein-DNA interaction and therefore of steep ionic-strength dependence.
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
The inference from an entropically driven, low-enthalpy signature to steep salt dependence is a standard reading of polyelectrolyte binding thermodynamics, not a measurement in this paper: NO salt titration of the CXXC-DNA interaction has been published, so the magnitude of the salt effect for this specific domain is unquantified. The values themselves are from Table 1, read this cycle.
Evidence that argues against this (1)
- An entropically favourable signature can also arise from hydrophobic burial or from conformational release, so counterion release is the likely but not the proven mechanism. The decisive experiment - Kd across a salt series, giving the slope of log Kd against log [Na+] and hence the number of ion pairs released - has not been done for any CXXC domain, and until it is, the salt argument in this entry is a mechanism-based prediction rather than a measured correction. Allen MD et al. 2006
- E4strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
A tagless, chemically synthesised MLL1 CXXC domain binds a 12-bp single-CpG palindrome with Kd 5.7 +/- 0.2 uM by EMSA in 10 mM Tris pH 7.4 with 150 mM NaCl, 50 uM ZnCl2 and 1 mM DTT, which is the only measurement of this contact combining no oligomerising tag with near-physiological ionic strength.
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
EMSA is not a solution-equilibrium method - gel caging commonly yields tighter Kd values than ITC on the same pair - and the titration was on ice rather than at 25 C, so the 7.7-fold difference from Cierpicki's 44 uM confounds salt with method and temperature. The construct span is in a supplementary table that was not retrieved. Full text read via the PMC HTML route (PMC8392069).
Evidence that argues against this (1)
- Because EMSA and the menin FP measurement are different assay classes, substituting 5.7 uM for 44 uM does not make the comparison like-for-like - it makes it differently unlike. The only clean fix is to measure both interactions at one salt in one assay format, which is what this entry's open questions propose and which nobody has done. Grembecka J et al. 2010
The salt mismatch analysis supplies the reason the 44 uM to 53 nM comparison is biased rather than merely imprecise.
The menin-MBM1 53 nM figure this analysis re-reads was measured in 50 mM Tris with 50 mM NaCl by fluorescence polarisation on a fluorescein-KMT2A 4-15 peptide.
Bears on every entry quoting the 53 nM menin-MBM1 affinity, none of which states the buffer.
The entry's 'three orders of magnitude weaker than the menin contact' compares a 300 mM NaCl measurement with a 50 mM NaCl one.
A worked instance of the ionic-strength term in the five-choices concept.