Oncofusion

start / graph / protein-sec-l3-complex

Proteinmoderate confidence 3 of 5: One strong study, or several smaller consistent ones.Q187-the-sec-family-and-yeats-selectivity

SEC-L3 — the AFF3-scaffolded complex, and the only SEC-family member with a known physiological job: genomic imprinting and X inactivation

also called SEC-like 3 · SEC-L3 · AFF3 complex · LAF4 complex

SEC-L3 is a transcription-speeding machine built on AFF3. It carries the same two mark-reading proteins, ENL and AF9, as the machine the infant-leukaemia fusion hijacks. What makes it different is its day job: AFF3 sits on imprinted genes and on the switch that silences one X chromosome, so this is the copy of the machine that helps keep parent-of-origin gene silencing in place.

The same thing, for a scientist

SEC-L3 is the AFF3-scaffolded member of the super elongation complex family: AFF3 with P-TEFb (CDK9, cyclin T1/T2) and the YEATS-domain readers ENL/MLLT1 and AF9/MLLT3, isolated by Flag-AFF3 purification with MudPIT, endogenous co-IP and Superose 6 chromatography, and — as for SEC-L2 — **without detectable ELL1, ELL2 or ELL3** [E1]. It has Pol II CTD kinase activity comparable to SEC, is not required for HSP70 induction, and its RNAi signature is 423 genes of which 61.7% are unique [E2]. The reason SEC-L3 deserves its own entry rather than a line in a table is what the same laboratory found afterwards: **AFF3 is an allele-specific factor at imprinted loci.** It binds gametic differentially methylated regions and the Meg3 enhancer in the Dlk1-Dio3 imprinted cluster, is recruited to the IG-DMR by ZFP57 and to the Meg3 enhancer by ZFP281, and is required for expression of the ~200 kb Meg3 polycistron [E3][E4]. It also binds the DMR downstream of the XIST promoter, and AFF3 knockdown de-represses the inactive allele of XIST in terminally differentiated cells [E5]. So of the four AFF scaffolds, the one with the best-defined normal function is doing parent-of-origin and X-inactivation epigenetics — in a complex that shares its reader subunits with the machinery this project proposes to drug.

Go deeper
Composition, with the same ELL negative as SEC-L2
subunitin SEC-L3?how established
AFF3 (P51826)scaffoldFlag bait
CDK9 / cyclin T1-T2yesMudPIT, endogenous co-IP, size-exclusion WB 1
ENL / MLLT1yesMudPIT, endogenous co-IP, size-exclusion WB 1
AF9 / MLLT3yesMudPIT, size-exclusion WB 1
ELL1, ELL2, ELL3NOT DETECTEDabsent from MudPIT; ELL2/ELL3 absent from endogenous anti-AFF3 IP while present in anti-AFF1, anti-AFF4 and anti-CDK9 1
EAF1 / EAF2not testedexpected from conservation, not confirmed 1

The Flag-AFF3 complex elutes in Superose 6 fractions 10-14, overlapping AFF4/SEC, attributed to AFF3 homodimerisation.

The physiological function, which is not elongation-in-general

Three papers from the same laboratory, over four years, make AFF3 the SEC-family scaffold with the sharpest known normal job, and all three name SEC-L3 explicitly:

  • Dlk1-Dio3, gametic DMRs and the Meg3 enhancer. AFF3 binds both gDMRs and enhancers within imprinted loci allele-specifically; the heterochromatic environment at the gDMR reinforces silencing of the related enhancer by controlling AFF3 binding and activity; AFF3 at the Meg3 enhancer is required for the ~200 kb Meg3 polycistron 3.
  • The recruiting factor at the enhancer is ZFP281. AFF3 associates with ZFP281 in mouse ES cells; ZFP281 recruits AFF3 to the Meg3 enhancer; genome-wide, ZFP281 generally associates with AFF3 at enhancers and the two together regulate a subset of genes. The gDMR route is separate and uses ZFP57 4.
  • XIST. AFF3 binds the DMR downstream of the XIST promoter; AFF3 knockdown de-represses the INACTIVE allele of XIST in terminally differentiated cells; the binding depends on DNA methylation and also feeds back on methylation level at the DMR 5.

None of these papers measures ENL or AF9 at those loci. That is the gap, and it is the gap that decides whether a YEATS inhibitor touches imprinting.

The consequence this corpus did not have

This project's standing therapeutic interest in the ENL/AF9 YEATS pocket (target-phase2-enl-yeats-dossier, drug-sgc-imllt) has been argued almost entirely against leukaemia biology and healthy-HSPC toxicity. The SEC-L3 literature adds a different class of liability: the shared reader subunits sit in a complex whose documented normal function is allele-specific silencing at imprinted loci and at XIST, in a disease whose patients are infants. Whether YEATS occupancy perturbs that is unmeasured — see absence-q187-no-yeats-inhibitor-has-been-tested-against-sec-l2-or-sec-l3 and risk-q187-yeats-inhibition-touches-an-imprinting-complex-in-an-infant.

Facts and figures
complex nameSEC-L3 (super elongation complex-like 3)
scaffoldAFF3
documented physiological functionallele-specific regulation at imprinted loci (Dlk1-Dio3 IG-DMR and Meg3 enhancer) and at the XIST DMR; recruitment by ZFP57 and ZFP281
epigenetic readoutAFF3 binding is DNA-methylation-dependent and allele-specific at gDMRs and at the XIST DMR; the complex additionally carries two YEATS acyl-lysine readers
shown for fusionNO — AFF3 is a recurrent 5'-KMT2A fusion partner in its own right (t(2;11)) but no experiment places SEC-L3 on any KMT2A fusion
system of function evidencemouse embryonic stem cells and terminally differentiated human cells; the composition evidence is HEK293
subunits
{
  "AFF3": {
    "accession": "P51826",
    "length_aa": 1226,
    "role": "scaffold; Flag bait in the defining purification"
  },
  "CDK9": {
    "accession": "P50750",
    "role": "P-TEFb kinase"
  },
  "CCNT1": {
    "accession": "O60563",
    "role": "P-TEFb cyclin"
  },
  "MLLT1": {
    "accession": "Q03111",
    "role": "YEATS-domain acyl-lysine reader; shared with SEC and SEC-L2"
  },
  "MLLT3": {
    "accession": "P42568",
    "role": "YEATS-domain acyl-lysine reader; shared with SEC and SEC-L2"
  }
}
subunits looked for and not found
[
  "ELL",
  "ELL2",
  "ELL3"
]
subunits not tested
[
  "EAF1",
  "EAF2"
]
recruiting factors
[
  "ZFP57 (IG-DMR)",
  "ZFP281 (Meg3 enhancer)"
]
marks named
[
  "DNA methylation (gDMR, IG-DMR, XIST DMR)",
  "H3K9 methylation (KAP1 pathway, reported NOT to govern XIST)",
  "H3K27ac",
  "H3K9ac"
]
measurements
[
  {
    "quantity": "differentially_expressed_genes",
    "subject": "AFF3 shRNA knockdown, HEK293T (SEC-L3 proxy)",
    "value": 423,
    "units": "genes",
    "n": 2,
    "universe": "HEK293T, RefSeq via TopHat/Cufflinks, FDR-adjusted p<0.05, 2 shRNA replicates vs 4 non-targeting replicates",
    "null": null,
    "note": "338 downregulated; 61.7% unique to AFF3 knockdown"
  },
  {
    "quantity": "shared_genes",
    "subject": "AFF2 shRNA DEGs intersected with AFF3 shRNA DEGs",
    "value": 76,
    "units": "genes",
    "n": 2,
    "universe": "same HEK293T experiment; AFF2 set 226, AFF3 set 423",
    "null": null,
    "note": "76 is 33.6% of the AFF2 set and ~18% of the AFF3 set; Jaccard 76/573 = 0.133"
  }
]
uniprot members
[
  "P51826",
  "P50750",
  "O60563",
  "Q03111",
  "P42568"
]
How this could bite us back
  • SEC-L3's documented physiological job is allele-specific silencing at imprinted loci and at XIST. If the shared ENL/AF9 reader is needed for SEC-L3 function, a YEATS inhibitor given to an infant is acting on imprinting machinery, and loss of imprinting at Dlk1-Dio3 has a known growth and developmental phenotype. This liability is absent from every YEATS-inhibitor entry in this corpus and from the published probe and in vivo papers.
  • The escape hatch is real and should be tested before the alarm is believed: AFF3's recruitment to those elements is driven by ZFP57 and ZFP281, which are sequence-specific and have nothing to do with acyl-lysine reading. A YEATS inhibitor may simply not reach them.
  • AFF3 is itself a recurrent KMT2A fusion partner. Any argument that AFF-scaffold paralogues are a safe comparator arm has to survive the fact that one of them is also an oncogenic partner.
Still open
  1. Are ENL and AF9 present at the AFF3-bound gametic DMRs, the Meg3 enhancer and the XIST DMR? Nobody has ChIPped a YEATS reader at an imprinted control region. That single experiment decides whether the imprinting liability of a YEATS inhibitor is real or imaginary.
  2. Does AFF3 do its imprinting job inside SEC-L3 or outside it? ENL and AF9 are known to exist outside SEC when bound to DOT1L, so scaffold membership is not a given at every locus, and no imprinted-locus experiment has blotted a single other SEC-L3 subunit.
  3. Is AFF3 protein present in the fetal pro-B cell where KMT2A::AFF1 arises? AFF3 is thymus-maximal by transcript; gene-aff3 already asks this and it remains open.
  4. Does a KMT2A::AFF3 fusion, which is recurrent, behave like KMT2A::AFF1 in menin and DOT1L dependence? If the AFF moiety only supplies a generic SEC, it should; if the scaffold sets the locus, it should not.
Every source behind this entry (5)
  1. E1strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper

    In HEK293 Flp-In-TRex cells, Flag affinity purification of AFF3 recovered CDK9, cyclin T1/2, ENL/MLLT1 and AF9/MLLT3 by MudPIT with confirmation by endogenous co-IP and Superose 6 chromatography (elution fractions 10-14), while ELL1, ELL2 and ELL3 were not reproducibly detected and ELL2/ELL3 appeared only in AFF1-, AFF4- and CDK9-directed immunoprecipitations; the complex was named SEC-like 3 (SEC-L3).

    The super elongation complex family of RNA polymerase II elongation factors: gene target specificity and transcriptional output. · Luo Z, Lin C, Guest E, Garrett AS, Mohaghegh N, Swanson S, Marshall S, Florens L, Washburn MP, Shilatifard A · Molecular and Cellular Biology 2012
    PMID 22547686DOI
    What this evidence cannot show

    HEK293/293T with an over-expressed tetracycline-inducible bait; no haematopoietic or leukaemic system. The ELL negative is confounded with the roughly 10-fold lower AFF3 expression in this cell type, which the authors state. Stoichiometry of ENL/AF9 is in a figure panel (dNSAF values) not recoverable from text.

    Evidence that argues against this (1)
    • The paper's own Figure 1F schematic states that ELL- and EAF-related proteins are 'expected based on sequence conservation among AFF family members' even though its blots did not find them, so the paper contradicts itself on this point and a reader could legitimately record ELL as present in SEC-L3. Luo Z et al. 2012
  2. E2strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper

    AFF3 shRNA in HEK293T changes 423 genes (338 downregulated), 61.7% of which are unique to AFF3 depletion; 76 genes are shared with the AFF2 knockdown, which is ~18% of the AFF3 set and ~33.6% of the AFF2 set; AFF3 depletion does not compromise heat-shock induction of HSP70 and AFF3 is not detectably recruited to HSP70.

    The super elongation complex family of RNA polymerase II elongation factors: gene target specificity and transcriptional output. · Luo Z, Lin C, Guest E, Garrett AS, Mohaghegh N, Swanson S, Marshall S, Florens L, Washburn MP, Shilatifard A · Molecular and Cellular Biology 2012
    PMID 22547686DOI
    What this evidence cannot show

    HEK293T only, FDR<0.05, two shRNA replicates against four non-targeting controls. A knockdown-response set is not an occupancy set. The DEG counts scale with scaffold abundance (AFF4 >> AFF3 > AFF2 in this cell type), so the ranking of set sizes is partly a power ranking.

    Evidence that argues against this (1)
    • No AFF3 ChIP-seq was performed in this paper — only AFF4 ChIP-seq. So the statement that SEC-L3 'controls a different subset of genes' rests entirely on a knockdown transcriptome, and the direct-binding half of the claim is supplied later and in a different system (mouse ES cells, imprinted loci) by the same group's subsequent work. Luo Z et al. 2016
  3. E3strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper

    In mouse embryonic stem cells, AFF3 — described by the authors as the central factor of SEC-L3 — binds both gametic differentially methylated regions and enhancers within imprinted loci in an allele-specific manner, and AFF3 at an enhancer is required for expression of the ~200 kb Meg3 polycistron within the imprinted Dlk1-Dio3 locus; the heterochromatic environment at the gDMR reinforces silencing of the related enhancer by controlling AFF3 binding and activity.

    Regulation of the imprinted Dlk1-Dio3 locus by allele-specific enhancer activity. · Luo Z, Lin C, Woodfin AR, Bartom ET, Gao X, Smith ER, Shilatifard A · Genes & Development 2016
    PMID 26728555DOI
    What this evidence cannot show

    Mouse embryonic stem cells, not human and not haematopoietic. The paper establishes AFF3 occupancy and requirement; it does NOT show that ENL or AF9 are present at these sites, and does not test any SEC-family subunit other than AFF3 at the imprinted locus. So 'SEC-L3 is an imprinting complex' is an inference from AFF3's complex membership, not a demonstrated co-occupancy.

    Evidence that argues against this (1)
    • The same group's own biochemistry undercuts the complex-level reading. AFF3 was later shown to be recruited to two different imprinted elements by two different zinc-finger proteins (ZFP57 to the IG-DMR, ZFP281 to the Meg3 enhancer), and the authors conclude that 'different zinc finger proteins can recruit AFF3 to different regulatory elements and differentially regulate the function of AFF3 in a context-dependent manner'. That is a description of AFF3 as a recruited module, and it leaves open that AFF3 does this work OUTSIDE SEC-L3 altogether — exactly as ENL/AF9 are known to exist outside SEC when bound to DOT1L. Wang Y et al. 2017
  4. E4strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper

    AFF3, named as the central component of SEC-L3, associates with the Kruppel-like zinc-finger protein ZFP281 in mouse ES cells; ZFP281 recruits AFF3 to the Meg3 enhancer within the imprinted Dlk1-Dio3 locus and thereby regulates allele-specific expression of the Meg3 polycistron, while localisation of AFF3 to the IG-DMR requires ZFP57 instead; genome-wide, ZFP281 generally associates with AFF3 at enhancers.

    A permissive chromatin state regulated by ZFP281-AFF3 in controlling the imprinted Meg3 polycistron. · Wang Y, Shen Y, Dai Q, Yang Q, Zhang Y, Wang X, Xie W, Luo Z, Lin C · Nucleic Acids Research 2017
    PMID 28180295DOI
    What this evidence cannot show

    Mouse ES cells. No SEC-family subunit other than AFF3 is assayed, so this is evidence about AFF3 and only indirectly about SEC-L3. Abstract read in full; body not read for this item.

    Evidence that argues against this (1)
    • If AFF3's imprinted-locus recruitment is driven by sequence-specific zinc-finger proteins rather than by anything intrinsic to the SEC-L3 particle, then blocking the shared YEATS reader may not displace AFF3 from those sites at all, and the imprinting liability this corpus is worried about would not materialise. The recruitment logic described here is the strongest available argument AGAINST the risk model in risk-q187-yeats-inhibition-touches-an-imprinting-complex-in-an-infant. Wang Y et al. 2017
  5. E5moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper

    AFF3, previously identified as the central factor of SEC-L3, binds the differentially methylated region downstream of the XIST promoter in a DNA-methylation-dependent way, and knockdown of AFF3 leads to de-repression of the INACTIVE allele of XIST in terminally differentiated cells.

    AFF3-DNA methylation interplay in maintaining the mono-allelic expression pattern of XIST in terminally differentiated cells. · Zhang Y, Wang C, Liu X, Yang Q, Ji H, Yang M, Xu M, Zhou Y, Xie W, Luo Z, Lin C · Journal of Molecular Cell Biology 2019
    PMID 30535390DOI
    What this evidence cannot show

    One laboratory, terminally differentiated cells, knockdown rather than degradation. Abstract read in full; body not read for this item. The paper is about AFF3, not about ENL, AF9 or the SEC-L3 particle as such, and it does not test whether the YEATS readers are at the XIST DMR.

    Evidence that argues against this (1)
    • The paper's own finding narrows its generality: the KAP1-H3K9-methylation machinery that regulates the imprinted loci is reported NOT to play a major role in maintaining mono-allelic XIST, so the authors themselves describe the XIST DMR and the gDMRs as being governed by different mechanisms that merely share a requirement for AFF3 and DNA methylation. That argues against treating 'AFF3 maintains allele-specific silencing' as a single unified function. Zhang Y et al. 2019
Connected to