Targeting LIN28B - the essential-function and wrong-sign risks are established by the SAME paper that makes it interesting, and the sign may reverse with lineage
also called LIN28B target liability · bite-back on LIN28B · why LIN28B may be undruggable in infants
The paper that shows infant leukaemia cannot start without LIN28B also shows that a baby's own B cells and NK cells cannot be made without it. And other papers show that in myeloid leukaemia, adding LIN28B makes the cancer grow up and die - so a drug that blocks it could make that kind worse.
The same thing, for a scientist
LIN28B is being discussed as a target in this graph on the strength of one human study. **Four liabilities are already on the published record, and three of them come from that study or its direct counterparts rather than from speculation.** (1) **Essential-function overlap:** LIN28B is essential in human fetal liver CD34+ cells to maintain proliferation and stem-like properties and to support B- AND NK-lymphopoiesis - the same cells, the same paper [E1]. (2) **Wrong sign in myeloid:** enforced LIN28B in established mouse MLL::AF9 AML extended survival with 80% of recipients alive beyond four months and induced monocytic differentiation, so in myeloid disease LIN28B is the *therapeutic agent*, not the target [E2]. (3) **Phase mismatch:** the demonstrated requirement is for **initiation**, which has already happened in every diagnosed patient; the therapeutically load-bearing arm reports loss of a proliferative *advantage*, not regression [E1]. (4) **No chemical matter:** there is no clinical-grade LIN28B inhibitor, and the target is a small RNA-binding protein of a class repeatedly nominated and rarely drugged. **Confidence in the underlying result is additionally limited by read depth**: the primary source is paywalled and absent from PMC, so its controls have not been inspected by this graph [E3].
The four liabilities, with severity and what would retire each
| # | kind | what is on the record | severity | what would retire it |
|---|---|---|---|---|
| 1 | essential-function toxicity | LIN28B essential in human FL CD34+ for proliferation, stem-like properties, B- and NK-lymphopoiesis 1 | high - the target tissue and the toxic tissue are the same cell | a therapeutic index measured in matched normal and transformed fetal progenitors; or a degrader with leukaemia-restricted delivery |
| 2 | wrong sign by lineage | enforced LIN28B differentiates and eradicates mouse MLL::AF9 AML 2; LIN28B suppresses MLL-ENL initiation | high - KMT2A-r disease switches lineage under immunotherapy, so a patient can cross the sign boundary mid-treatment | showing the differentiation effect is mouse-specific, or absent in human myeloid PDX |
| 3 | phase mismatch | the strong result is an INITIATION requirement; patients present post-initiation. The maintenance arm reports lost proliferative advantage, not regression 1 | moderate-high - decides whether this is a target at all versus an aetiological insight | a regression (not growth-rate) readout on LIN28B knockdown in an established human KMT2A::AFF1 xenograft |
| 4 | no chemical matter / delivery | no clinical-grade LIN28B inhibitor; RNA-binding proteins of this class are hard | moderate | a selective tool compound with demonstrated target engagement, and paralogue selectivity over LIN28A |
Liability 1 deserves emphasis because of where it comes from. It is not an extrapolation from a knockout mouse phenotype in another tissue. It is the same figure panel set, in the same human cells, in the same paper, as the result that makes LIN28B attractive. A target whose therapeutic window is contradicted by its own discovery paper is unusual and the graph should say so plainly.
The self-limiting-target problem, which is specific to fetal-restricted factors
A fetal-restricted factor has an attractive property - it is absent from healthy older tissue, so an agent against it should be selective in an older patient - and an unattractive one that follows immediately: the target disappears as the patient ages.
For KMT2A::AFF1 infant ALL this cuts badly. The disease's main cause of death is early relapse, often within the first two years. If LIN28B expression falls with postnatal age in the leukaemic clone as it does in normal tissue, then:
- the drug's target is highest exactly when the disease is newly diagnosed and chemotherapy is working best, and
- lowest at the relapse the drug would be needed for.
This is testable on banked material and nobody has done it. gene-lin28b already poses the question - is LIN28B expression retained in diagnostic KMT2A::AFF1 infant ALL blasts at a level that would support targeting, and does it fall at relapse? - and this risk entry exists partly to raise its priority: it is a go/no-go question for the entire target rationale, answerable on diagnostic-relapse pairs, with no new experiment needed.
| kind | essential-function |
|---|---|
| severity | high - the therapeutic window is contradicted by the target's own discovery paper, and the sign of benefit may reverse with lineage |
| self limiting target problem | A fetal-restricted target is highest at diagnosis and lowest at the late relapse it would be needed for. Whether LIN28B is retained in diagnostic blasts and whether it falls at relapse is answerable on banked diagnostic-relapse pairs and has never been measured - this entry rates it the go/no-go question for the whole rationale. |
| mitigation | Before any target-dossier work: (a) measure LIN28B in diagnostic and relapse infant KMT2A::AFF1 blasts; (b) obtain a therapeutic index in matched normal and transformed fetal progenitors; (c) establish whether the differentiation effect seen in mouse myeloid AML occurs in human PDX. None requires new chemistry. |
| evidence quality risk | The founding experiment is paywalled and absent from PMC, so its controls have not been inspected by this graph (verified 2026-09-23). This is a risk to the assessment, not to the biology, and it is recorded separately for that reason. |
| shown for fusion | Liabilities 1 and 3: KMT2A::AFF1, human. Liability 2: KMT2A::MLLT3 and KMT2A::MLLT1, mouse. Liability 4: not fusion-specific. |
secondary kinds
[ "compensation", "toxicity", "delivery", "wrong-sign / lineage-conditional" ]
affects
[ "any future LIN28B-directed agent", "let-7 mimetic approaches by extension, since they act on the same axis" ]
liabilities — the numbers behind the chart
[
{
"id": 1,
"kind": "essential-function toxicity",
"statement": "LIN28B is required for normal human fetal B- and NK-lymphopoiesis and for fetal progenitor proliferation and stem-like properties, in the same cells and the same paper as the transformation requirement",
"severity": "high",
"source_pmid": "42213620",
"retired_by": "a measured therapeutic index in matched normal and transformed fetal progenitors, or leukaemia-restricted delivery"
},
{
"id": 2,
"kind": "wrong sign by lineage",
"statement": "enforced LIN28B differentiates and may eradicate mouse MLL::AF9 AML; LIN28B suppresses MLL-ENL initiation. In myeloid disease LIN28B is the agent, not the target",
"severity": "high",
"source_pmid": "38321106",
"retired_by": "showing the differentiation effect is absent in human myeloid PDX",
"clinical_trigger": "lineage switch under blinatumomab or CAR-T, documented in KMT2A-rearranged disease"
},
{
"id": 3,
"kind": "phase mismatch",
"statement": "the strong result is an INITIATION requirement and patients present post-initiation; the maintenance arm reports a lost proliferative advantage, not regression",
"severity": "moderate-high",
"source_pmid": "42213620",
"retired_by": "a regression readout on LIN28B knockdown in an established human KMT2A::AFF1 xenograft"
},
{
"id": 4,
"kind": "delivery / chemical matter",
"statement": "no clinical-grade LIN28B inhibitor exists; the target class is a small RNA-binding protein with a poor drugging record, and paralogue selectivity over LIN28A would be required",
"severity": "moderate",
"source_pmid": null,
"retired_by": "a selective tool compound with demonstrated engagement and LIN28A counter-screening"
}
]- The target's discovery paper establishes the toxicity in the same cells as the efficacy. That is not a risk to be mitigated later; it is a property of the biology.
- In myeloid KMT2A-rearranged disease the published direction of benefit is LIN28B RESTORATION. A single agent cannot serve both lineages, and KMT2A-rearranged disease crosses between them under treatment.
- An initiation requirement is close to unusable clinically, because initiation is complete before diagnosis. The only therapeutically meaningful arm reports a slower growth rate, which is a weak basis for a paediatric development programme.
- The whole rationale may be self-limiting: the target declines with the patient's age while the disease's lethality is concentrated at relapse, when the target would be lowest.
- Because the founding paper cannot be read in full here, this risk assessment itself may be wrong in either direction - the controls that would strengthen or weaken it are behind a paywall.
- Is LIN28B retained in diagnostic infant KMT2A::AFF1 blasts at a targetable level, and does it fall at relapse? Answerable on banked diagnostic-relapse pairs, and it is the go/no-go for the entire rationale.
- What is the therapeutic index - how much more does LIN28B loss cost a transformed fetal progenitor than a normal one? The discovery paper's abstract states both effects and compares neither.
- Does enforced LIN28B differentiate a human KMT2A::AFF1 PDX? If it does, the correct intervention is restoration and every assumption in this entry inverts.
- Would a LIN28B degrader spare LIN28A, and does LIN28A compensate? LIN28A uses a different, cytoplasmic mechanism, so paralogue compensation is not guaranteed - but it has not been tested in a haematopoietic cell.
- E1strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
LIN28B was essential in human fetal liver CD34+ cells to maintain proliferation and stem-like properties and to support B- and NK-lymphopoiesis; in the same system human FL CD34+ cells fail to transform upon induction of the KMT2A::AFF1 translocation in the absence of LIN28B, and LIN28B-expressing CRISPR-KMT2A::AFF1 leukaemias were more proliferative in vitro and in vivo with that advantage lost upon LIN28B knockdown.
The fetal specific gene LIN28B is essential for human fetal B-lymphopoiesis and initiation of KMT2A::AFF1 infant leukemia. · Ling RE, Welsh AM, Hamer L, Jackson T, Neil E, Elliott N, et al. · Blood 2026What this evidence cannot show
ABSTRACT ONLY and paywalled. No effect sizes, no donor numbers, no therapeutic-index measurement. Notably, the abstract gives NO comparison of the magnitude of the normal-cell effect against the leukaemic-cell effect, which is precisely the number a target assessment needs. The toxicity concern is therefore qualitative: the same gene is required in both, with no statement of margin.
Evidence that argues against this (1)
- In mouse, Lin28b deletion in fetal progenitors shifts lineage output toward adult-like myeloid rather than abolishing progenitor function, and conditional fetal deletion was compatible with normal survival in a leukaemia model - so the severity of the human normal-tissue phenotype may be overstated by the word 'essential'. J Exp Med 2016
- E2strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
Enforced LIN28B in established mouse MLL::AF9 AML significantly extended survival, with 80% of secondary recipients surviving beyond 4 months and GFP+ AML cells undetectable in the marrow of most survivors; it induced partial monocytic differentiation by 6 days and severely depleted colony-forming potential in 5 biologically independent AMLs without increasing programmed cell death; and conditional Lin28b deletion in fetal progenitors did not accelerate MLL::ENL AML initiation.
LIN28B promotes differentiation of fully transformed AML cells but is dispensable for fetal leukemia suppression. · Li Y, Mendoza-Castrejon J, Patel RM, Casey EB, Denby E, Bryder D, Magee JA · Leukemia 2024What this evidence cannot show
FULL TEXT READ (PMC10912017). Mouse, transgenic fusions, MYELOID, and LIN28B levels exceeded physiological levels by the authors' own statement. The authors themselves note that whether LIN28B promotes AML differentiation in human patient-derived xenografts remains to be established, citing potential species differences - so the wrong-sign risk is a real published direction, not an established human liability.
Evidence that argues against this (1)
- The authors themselves state that whether LIN28B promotes AML differentiation in human patient-derived xenografts remains to be established, citing potential species differences in LIN28B expression and function. The wrong-sign liability is therefore a published mouse direction, not a demonstrated human liability, and grading it 'high' is a precautionary judgement rather than an evidenced one. Leukemia 2024
- E3established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
The primary source for the LIN28B requirement is not open access and is absent from PubMed Central and Europe PMC as of 2026-09-23, returning isOpenAccess 'N', inEPMC 'N', inPMC 'N' and a single fullTextUrl of availability 'Subscription required'.
What this evidence cannot show
A statement about availability on one date; Blood deposits to PMC after an embargo. Recorded as a risk item because a target assessment that cannot inspect its own founding experiment's controls is carrying an evidence-quality risk distinct from any biological one.
Evidence that argues against this (1)
- Availability is a moving target, not a fact about the paper. Blood deposits to PMC after an embargo, so this record's 'not open access' status will become false and any bounded statement resting on 'the full text could not be read' silently expires. A later reader must re-check rather than inherit this. source 2026
structured target-liability dossier for LIN28B
liability 1 - the normal state that shares the target
liability 2 - the sign dossier
liability 3 - the requirement demonstrated is for initiation, not maintenance
the self-limiting-target problem: target falls with age, lethality concentrates at relapse
let-7 mimetic approaches inherit these liabilities by acting on the same axis