TNIK in Pain Relief — Target-Disease Evaluation

Dated 2026-08-19

The evidence

Current state of the evidence

TNIK x Pain evaluationCautionMonitor — a highly druggable kinase whose pain rationale is pathway-level only.

“TNIK is the kinase that licenses this final transcriptional step: it is recruited to Wnt target-gene promoters and phosphorylates TCF4/TCF7L2, the same transcription factor shown to maintain neuropathic pain (PMID 25963533).”

HTML

The reasoning

TNIK is an exceptionally tractable Wnt-activating kinase that sits directly upstream of the TCF4/TCF7L2 transcriptional node implicated in neuropathic pain, but no study has tested the kinase itself in pain, expression evidence in the disease proxy is absent, and close kinase paralogs raise a selectivity burden; the recommendation is to monitor and de-risk with a proof-of-mechanism study.

Table 1Each line of evidence is set out with the ground it covers and the position it supports.
Evidence	Scope	Position
TNIK x Pain evaluation	One target-disease evaluation of TNIK for pain relief, scored in PandaOmics via complex regional pain syndrome as a proxy.	argues for, with conditions

Positions record where each line of evidence lands on the question this page asks rather than on the target in general.

Evidence

The data on this pair

Not analyzedexpression evidenceThe gene appears in none of the one datasets available for this disease, so its expression score records the absence of an analysis rather than the absence of a change.

0.859strongest dimensionThe dimension is Heterogeneous graph walk, whose score counts as strong evidence on that dimension’s own scale.

11dimensions scoredSix of the dimensions carry too little signal to point either way.

Expression evidence: not analyzed

The gene appears in none of the one datasets available for this disease, so its expression score records the absence of an analysis rather than the absence of a change.

Direction of the evidence

11 dimensions shown

Points to modifying the targetToo little signal to point either way

Figure 1Each bar is one dimension on which this pair is scored, colored by what that dimension implies about drugging the target.

A dimension scoring below 0.2 carries too little signal to imply a direction, and is drawn as pointing neither way rather than being left off.

Safety

The cost of inhibiting the target

Moderateon-target riskAcross four dimensions the grades run three moderate and one low, and the summary takes the worst of them because a patient meets all of them at once.

1tissues at high riskOf the three tissues in which the gene’s level has been reported, a tissue counts as at risk where the gene is present and its job there is one that nothing else does.

2close relativesA close relative is a protein similar enough to the target that a molecule designed against the target may bind it as well.

Table 2Each safety dimension is set out with the evidence it rests on and the measures that would reduce it.
Dimension	Risk	Underlying evidence	Mitigation
KO lethality	Moderate	Tnik KO mice viable but with cognitive/synaptic deficits (PMID 23035106)	Titrate exposure; monitor CNS pharmacodynamics.
Essential tissue function	Moderate	Broad expression; Wnt roles in intestinal/stem-cell homeostasis, cf. Wnt-pathway (tankyrase) intestinal toxicity (PMID 26692561)	Intermittent dosing; GI monitoring; therapeutic-index studies.
Immune system effects	Moderate	TRAF2 interaction (STRING 0.762) links TNIK to TNF/JNK signaling	Immune-safety pharmacology; cytokine monitoring.
Reproductive toxicity	Low	No reported reproductive phenotype in Tnik KO	Standard reproductive-toxicology package.

The four graded dimensions stand at three moderate and one low.

Table 3Each tissue in which the gene is expressed is listed with the cost that blocking it there would carry.
Tissue	Reported levelas a tissue expression atlas reports it	Function in that tissue	Risk if inhibitedjudged here, not measured
Brain / neurons	High	Dendritic arborization, synaptic and cognitive function	High
Intestine / stem-cell compartments	Medium	Wnt/β-catenin-dependent proliferation and homeostasis	Moderate
Broad somatic tissues	Medium	Wnt signaling and cytoskeletal regulation (tissue-specificity 0.054)	Moderate

The two right-hand columns are deliberately on different scales, because one records a level that was measured while the other states the conclusion drawn from that level here.

Table 4Each relative listed here is close enough to the target that a molecule aimed at the target might bind it as well.
Relative	Identitypercent, over the region named beside it	Region compared	Overlapping function	Selectivity risk
MINK1	89	N-terminal kinase domain	Close GCK-family paralog; shared kinase fold and overlapping neuronal signaling	High
MAP4K4	91	N-terminal kinase domain	Close GCK-family paralog (HGK); shared kinase fold and stress/metabolic signaling	High

Identity is quoted over a named region rather than the whole protein: two proteins can share little overall and almost everything where a drug would bind.

Consequences of losing the gene, in two models

Mouse, Germline knockout: Viable. Impaired postsynaptic/nuclear signaling and cognitive function (PMID 23035106).
Human, Loss-of-function / structural variants: Not established. Cognitive disability and chromosome-3 structural disorders; no pain phenotype.

Competition

Other programs against this target

0programs in this diseaseNo compound against this target is in development for pain relief, so the table below is empty because the field is empty rather than because the search was narrow.

4programs elsewherePrograms aimed at the same target in other diseases show that the target can be drugged, although they are not evidence that drugging it helps here.

3patent filingsThree filings were made against this target, running from 2009 to 2022.

Nothing is in development for pain relief

No compound against this target is in development for pain relief, so the table below is empty because the field is empty rather than because the search was narrow.

Table 5The programs listed here act on the same target but are aimed at other diseases.
Compounds and biologics	Stage	Status	Developer	Evidence	Disease
rentosertib	Phase III	Phase III initiated	Insilico Medicine	PMID 40461817	IPF
TNIK inhibitor series	Preclinical / Lead	Lead	Academic	PMID 39422731	Cancer
mebendazole	Tool Compound	Preclinical	Academic	PMID 27650168	Oncology
osimertinib	Approved	Approved	AstraZeneca	PMID 40999821	IPF

They are listed so that it is clear who holds what, although none of it is evidence about this disease.

Table 6These are the filings that a program in this disease would have to work around.
Filing	Filed	Holder	Scope of the claims	Approach
US11530197B2	2022-12-20	Not disclosed in retrieved record	Composition-of-matter — TNIK inhibitors with anti-fibrotic (collagen/extracellular-matrix) use.	Anti-fibrotic small molecule
WO2019156439A1	2019-01-31	Korea Research Institute of Chemical Technology / Yonsei University IACF	Composition-of-matter and method-of-treatment (cancer).	Oncology small molecule
WO2010064111A1	2009-11-30	Yamada Tesshi and co-inventors (academic)	Composition-of-matter (aminothiazoles) and method-of-treatment (colorectal/solid cancer).	Oncology small molecule

Three filings were made against this target, running from 2009 to 2022.

Verdict

Recommendation and next steps

MonitorrecommendationA highly druggable kinase mechanistically upstream of a pain-relevant transcriptional node, but with no direct evidence for the target in pain; monitor and de-risk with a proof-of-mechanism study.

4routes consideredEach route is scored for feasibility, novelty and selectivity against this target in pain relief.

Rank of this disease among the target’s indications

Three of the four indications ranked for this target score above pain relief, namely cancer, neoplasm and nervous system disease.

Table 7The recommendation rests on the judgments set out here.
Assessment	Rating	Grounds for the rating
Overall Druggability	High	Highly tractable kinase with 11 PDB structures, multiple selective inhibitor series, and a clinical-stage inhibitor (rentosertib).
On-Target Safety	Moderate	TNIK knockout is compatible with life, but non-redundant roles in cognition and Wnt-dependent tissue homeostasis, plus close kinase paralogs, mean CNS and GI on-target liabilities to manage under chronic dosing.
Expression Evidence	Not analyzed	TNIK is absent from the significant DEGs in the available PandaOmics pain-proxy datasets; its Expression omics-dimension score is 0.520.
Opportunity	Moderate	A transcription-level analgesic mechanism upstream of TCF4/TCF7L2, distinct from ion-channel and opioid approaches, but unproven for TNIK itself.
Risk	High	The pain hypothesis is validated only at the pathway level (Wnt/TCF4), not for the kinase; paralog selectivity and chronic-dosing CNS liability compound it.
Competitive Intensity	Low	No TNIK compound is in development for pain; all human TNIK experience is in idiopathic pulmonary fibrosis.
IP White Space	High	No patent claims TNIK for pain; existing filings cover fibrosis and oncology.

The expression rating is derived from the datasets themselves, so it is identical to the one reported in the evidence section above.

Table 8Each route is scored out of ten on three separate counts.
Route	Feasible	Novel	Selective	Requirements
Small molecule	8	8	5	Lead strategy — repurpose an existing TNIK inhibitor for a proof-of-mechanism pain study
PROTAC	5	8	6	Exploratory — only if scaffolding function proves important and catalytic inhibition is insufficient
Antibody / biologic	1	3	6	Not feasible — intracellular kinase
Antisense/siRNA	3	6	7	Backup — contingent on solving CNS/DRG delivery

Ten is the top of each scale, and a route scoring well on all three is not thereby the one to take, because the scores say what is possible rather than what this program is set up to do.

The work each route would require

Each route scored above is set out with the result that would settle whether it works.

Small molecule
Repurpose a clinical-stage or tool TNIK inhibitor to test whether kinase inhibition reproduces the analgesia seen when the Wnt/TCF4 pathway is blocked.
What would rule it out: Reversal of mechanical and thermal hypersensitivity in a rodent nerve-injury model with an existing TNIK inhibitor.Proof-of-mechanism rodent readout in 6–12 months; IND-enabling package in 2–3 years if positive; Phase II proof-of-concept in 4–6 years.
PROTAC
A degrader removes both catalytic and scaffolding TNIK functions, relevant if the STRIPAK/interaction role matters; but complete removal may amplify the cognitive on-target liability seen in knockouts (PMID 23035106).
What would rule it out: Selective TNIK degradation demonstrated without paralog loss.Degrader hit identification 1–2 years; IND-enabling 4–5 years.
Antibody / biologic
TNIK is an intracellular kinase, so an antibody cannot reach the target; listed for completeness.
What would rule it out: Not pursued.Not applicable.
Antisense/siRNA
Oligonucleotide knockdown is selective but depends on solving delivery to CNS and dorsal-root-ganglion neurons.
What would rule it out: TNIK knockdown and analgesia demonstrated in DRG/spinal tissue after feasible delivery.Lead oligonucleotide 2–3 years; IND-enabling 4–6 years.

Table 9The obstacles this program would meet are listed with the measures that would reduce them.
Challenge	Severity	Mitigation
Pain hypothesis is unvalidated for TNIK itself — every supporting study targets the Wnt/TCF4 pathway, not the kinase (PMID 25963533)	High	Run a proof-of-mechanism study with a selective TNIK inhibitor in a rodent nerve-injury model before committing to a pain program.
Paralog cross-reactivity with MINK1/MAP4K4 (STRING 0.774/0.665) can confound both efficacy and CNS tolerability	High	Structure-guided selectivity design and broad kinase-panel counter-screening.
On-target CNS liability — TNIK is required for cognition (PMID 23035106) — in a chronic-dosing pain setting	Moderate	Monitor CNS pharmacodynamics; titrate exposure; consider intermittent dosing.
No pain intellectual-property position — existing filings claim fibrosis and oncology uses only	Moderate	File method-of-treatment claims for pain if proof-of-mechanism is positive.

Four challenges are raised in all, graded two high and two moderate.

Literature cited