Chlorotoxin

Chlorotoxin comes from the deathstalker scorpion Leiurus quinquestriatus and was originally described as a chloride channel blocker, which is where the name comes from. The tumour-binding behaviour is the interesting part and the mechanism is still not fully settled: binding appears to involve a membrane complex containing MMP-2, annexin A2 and chloride channels that is upregulated on glioma cells and largely absent from healthy brain. Four disulfide bonds in a cystine-stabilised alpha/beta fold make it extraordinarily rugged — it survives serum, heat and pH extremes, which is exactly what you want in a targeting moiety. Conjugated to indocyanine green it becomes tozuleristide, or Tumor Paint, an intraoperative fluorescent agent that lights up tumour margins the surgeon's eye cannot distinguish. Radiolabelled with iodine-131 it was TM-601, a therapeutic attempt that did not progress. Most recently the chlorotoxin sequence has been used as the recognition domain of CAR-T cells directed against recurrent glioblastoma, with early feasibility and safety data published. It is worth saying plainly: chlorotoxin is not an analgesic and does not treat pain. It sits in this class as a venom-derived neuro-targeting peptide.

Targets: MMP-2 / annexin A2 membrane complex on glioma cells, Chloride channels (the original described target), Neuroectodermal tumour cell surface

• · The adult glioma phase 1 study ran 3 mg to 30 mg as a single IV dose with no dose-limiting toxicity. This is a hospital imaging agent given once, not a repeated therapy.
• · In the CAR-T application the peptide is not administered at all — its sequence is engineered into the receptor on the patient's own T-cells, which are then delivered locoregionally into the tumour cavity or ventricle. Dosing is expressed in cells, not micrograms.

Work out your exact syringe units →

• very commonNo dose-limiting toxicity observed across 3-30 mg in phase 1— The tolerability profile has been notably clean in the imaging studies to date.
• uncommonInfusion reactions— Standard for any IV protein or conjugate.
• uncommonTransient greenish skin or urine discolouration— From the indocyanine green component.
• rareHypersensitivity to indocyanine green or iodides— Relevant to the conjugate, not the peptide.

• Known hypersensitivity to indocyanine green or iodinated contrast, for the tozuleristide conjugate.
• Any expectation that this is a pain or performance compound. It is a tumour-targeting imaging agent, and there is no self-administration use for it.

• cautionara-290-neuropathy — No known pharmacological interaction; listed only because tissue-protective, anti-apoptotic signalling is not desirable in the setting of an active tumour.

• · Intraoperative fluorescence signal-to-background ratio, which is the actual performance endpoint.
• · Standard infusion-reaction observation during and after administration.
• · Extent of resection on post-operative imaging, which is what the agent is meant to improve.

Investigational only. Not approved in the US or EU. Native chlorotoxin is sold as a research reagent; tozuleristide exists solely within clinical trials.

• Patil et al. 2019, phase 1 safety, pharmacokinetics and fluorescence imaging study of tozuleristide (BLZ-100) in adults with gliomas (trial)
• PNOC012 — phase 2/3 randomised blinded study of tozuleristide for fluorescence imaging in paediatric CNS tumours (trial)
• Brown et al. 2025, Cell Reports Medicine — chlorotoxin-directed CAR-T cell therapy for recurrent glioblastoma (trial)
• Soroceanu et al., chlorotoxin binding to glioma cells and the MMP-2 complex (preclinical)

It is worth saying at the top of any deep dive on chlorotoxin that it is not an analgesic, has never been tested for pain, and treats nothing. It sits in this class as a venom-derived neuro-targeting peptide, and its interest is entirely in the targeting. The name is a historical artefact: chlorotoxin was originally characterised as a small-conductance chloride channel blocker, and that description has stuck to it in every textbook despite the chloride channel having very little to do with what the molecule is now used for. The behaviour that made it famous is tumour selectivity — chlorotoxin binds glioma and other neuroectodermal tumour cells and largely ignores healthy brain. The molecular basis is still not fully settled, which is unusual for a molecule this far into clinical development. The best-supported model involves a membrane complex containing matrix metalloproteinase-2 and annexin A2, upregulated on glioma cells and largely absent from normal brain parenchyma; chlorotoxin binding appears to require that complex, and it internalises with it. Whether the chloride channel component is part of the binding site or an unrelated early observation remains genuinely contested. Structurally, the four disulfide bonds are the whole engineering story. A cystine-stabilised alpha/beta fold is one of the most robust scaffolds in nature: it survives serum proteases, heat, and pH extremes that would destroy a linear peptide, which is exactly what you need in a moiety that has to circulate for hours, cross into tumour tissue, and stay bound while the background clears. Three applications have been built on that. Conjugated to indocyanine green it becomes tozuleristide — infuse it hours before surgery, and under near-infrared illumination the tumour margin fluoresces where the surgeon's eye sees nothing distinguishable. Radiolabelled with iodine-131 it was TM-601, a therapeutic attempt that did not progress. And most recently, the sequence itself has been used as the antigen-recognition domain of a CAR-T construct, replacing the usual single-chain antibody fragment with a 36-residue scorpion peptide — an unusual design choice that works because chlorotoxin's binding target is present across glioblastoma's notoriously heterogeneous cell populations, which is precisely the problem that has defeated conventional single-antigen CAR-T in this disease. One more mechanistic point that matters for the imaging use: chlorotoxin appears to reach tumour tissue partly because the blood-brain barrier is already disrupted at the tumour, which is why 'partial' is the honest answer on barrier penetration rather than 'yes'.

The thing that goes wrong with chlorotoxin is almost entirely a matter of what people think it is. It appears on peptide vendor sites, it comes from a scorpion, it has a dramatic name, and it sits in a pain and neuromodulation category — from which people reasonably but incorrectly conclude it does something for pain or for the nervous system generally. It does not. It has no analgesic activity, has never been tested in any pain indication, and treats no disease. Tozuleristide is a surgical imaging agent that makes tumour margins fluoresce; TM-601 was a radiotherapeutic that did not progress; and the CAR-T application does not involve administering the peptide to anyone at all — the sequence is engineered into a receptor on the patient's own T-cells and the dose is expressed in cells, not micrograms. The second issue is availability framing: native chlorotoxin sold as a research reagent is a laboratory tool for cell-binding studies, and there is no human protocol, no dose, and no reason to inject it. The third, for anyone reading the clinical literature, is over-reading the CAR-T result — four participants with three achieving stable disease is a feasibility signal in a disease where feasibility signals have repeatedly failed to become efficacy, and it should be read as encouraging rather than as evidence of benefit. And the fourth is a handling point in the imaging context: the indocyanine green conjugate photobleaches and is light-sensitive at every stage, so the material is protected from light from preparation through to the operating theatre.

• MMP-2 / annexin A2 membrane complex on glioma cells — No published Kd. Binding is described as high-affinity and highly selective for neuroectodermal tumour cells over normal brain, but the exact molecular partner and stoichiometry remain unsettled.

  Selective binding and internalisation into tumour cells. This is the property everything else depends on — it is a targeting event, not a pharmacological one, and chlorotoxin binding does not itself kill or treat anything.

• Small-conductance chloride channels — The original described target and the source of the name.

  Blockade, in the original electrophysiology. Its relevance to the tumour-binding behaviour is disputed and may be an unrelated early observation. Do not assume the name describes the mechanism.

• Neuroectodermal tumour cell surface generally — glioma, medulloblastoma, neuroblastoma, melanoma, some sarcomas — Broad within that lineage; minimal binding to normal brain, which is what makes the surgical contrast useful.

  The basis of the fluorescence imaging application and of the CAR-T targeting domain. Breadth across heterogeneous tumour cell populations is the specific advantage over single-antigen targeting in glioblastoma.

• Nociceptors, opioid receptors, ion channels involved in pain transmission — None relevant.

  Nothing. Chlorotoxin has no analgesic activity and has never been studied for pain. This row exists because the compound appears in a pain class and the absence should be explicit rather than inferred.

• Patil et al. — phase 1 safety, pharmacokinetics and fluorescence imaging study of tozuleristide (BLZ-100) in adults with gliomas Phase 1, open-label dose escalation · n=17 · 1 weeks · 2019

  Single intravenous doses from 3 mg to 30 mg before craniotomy. No dose-limiting toxicity was observed and no adverse events were considered drug-related. Fluorescence signal was detected in both high-grade and low-grade tumours, with intensity related to dose. The safety result is the important one — this is a remarkably clean profile for a scorpion venom peptide.

• Chlorotoxin-directed CAR T cell therapy for recurrent glioblastoma — interim clinical experience Phase 1, first-in-human · n=4 · 2025

  Interim report of the first participants. Feasibility and safety objectives met with no dose-limiting toxicities; three of four participants had stable disease as best response; CLTX-CAR T cells were detected in tumour cavity fluid at higher levels than in blood, confirming that locoregional delivery does what it is supposed to. Four patients is four patients — this is a feasibility signal, not efficacy.

Tumour fluorescence becomes usable a few hours after infusion, which is why tozuleristide is given the evening before or the morning of surgery. The useful window is wide — bound signal remains imageable for up to about 36 hours while plasma background continues to fall, so the signal-to-background ratio actually improves over the first several hours rather than decaying. That is an unusually forgiving pharmacokinetic profile for an imaging agent and it is what makes the surgical scheduling workable. For the CAR-T application the timeline is entirely different and governed by cell therapy rather than peptide kinetics: locoregional infusion into the tumour cavity or ventricle, with expansion and persistence measured over days to weeks.

There is no stacking context for chlorotoxin in the sense this site usually means, because there is no self-administration use and the material is not something a person takes. In its clinical settings the relevant combinations are surgical and oncological rather than pharmacological: tozuleristide is used alongside 5-aminolevulinic acid fluorescence and intraoperative MRI as complementary margin-detection tools, and the CAR-T construct is used within a standard lymphodepletion and cell-therapy framework. The one interaction worth flagging is conceptual and points the other way: if someone is running tissue-protective, anti-apoptotic peptides such as ARA-290 in the setting of an active tumour, that is a reasonable thing to worry about — not because chlorotoxin interacts with them, but because the innate repair receptor's anti-apoptotic signalling is not something you want acting on tumour tissue while you are trying to resect it. Native chlorotoxin sold as a research reagent has no described human use, no dose and no protocol.

Chlorotoxin belongs on this page for the same reason ziconotide does — both are venom peptides that made it into humans — and comparing them is the fastest way to understand what venom pharmacology actually offers. Ziconotide is a cone snail toxin that became an approved analgesic by blocking the channel it evolved to block; chlorotoxin is a scorpion toxin whose named target turned out to be largely beside the point, and whose value lies in an accidental binding selectivity nobody designed. Ziconotide has phase 3 data and a label; chlorotoxin has a 17-patient phase 1, a 4-patient CAR-T interim report, and a pivotal paediatric imaging trial that completed enrolment. Against the broader field of tumour-margin imaging, tozuleristide competes with 5-aminolevulinic acid, which is already approved for high-grade glioma — chlorotoxin's proposed advantages are that it also lights up low-grade tumours, where 5-ALA performs poorly, and that it works in paediatric CNS tumours. Against conventional CAR-T targeting, using a 36-residue scorpion peptide instead of a single-chain antibody fragment is a genuinely novel structural approach, and the argument for it — coverage of heterogeneous tumour cell populations — addresses the specific reason single-antigen CAR-T has failed repeatedly in glioblastoma. Whether that argument holds up in more than four patients is the entire open question.

No monthly cost exists. Tozuleristide is a single pre-operative hospital dose given within a clinical trial, not a marketed product; chlorotoxin-directed CAR-T is an experimental cell therapy delivered in an academic centre; and native chlorotoxin research reagent has no human dosing protocol to cost out. Any figure here would imply a repeating use pattern that does not exist.

• ?The molecular target of chlorotoxin's tumour selectivity is genuinely unsettled. The MMP-2 / annexin A2 complex model is the best supported, but the role of the chloride channel that gave the molecule its name remains disputed, and no Kd has been published for any proposed partner.
• ?No formal human pharmacokinetic parameters have been published: no volume of distribution, no protein binding, no clearance figure, no half-life number. The 'hours in plasma, up to 36 hours of tumour signal' description is qualitative.
• ?Blood-brain barrier penetration is recorded as partial, which reflects that tozuleristide reaches tumour tissue where the barrier is already disrupted rather than crossing an intact barrier. This distinction is not well quantified.
• ?The status of the pivotal paediatric CNS tumour imaging trial was not verified in this session; the Core record states enrolment completed with FDA fast-track designation, and I did not independently confirm current status or results.
• ?Duration and follow-up length for the CLTX-CAR T interim report are left null; the publication is an interim clinical experience without a fixed study duration.
• ?The UniProt entry P45639 gives the 36-residue sequence beginning with methionine; some sources describe a 35-residue mature form lacking that N-terminal methionine, and I did not resolve which form is used in tozuleristide.
• ?Native chlorotoxin has no human dosing data of any kind. Everything human in this record relates to the ICG conjugate or the CAR-T construct.

• Phase 1 Safety, Pharmacokinetics, and Fluorescence Imaging Study of Tozuleristide (BLZ-100) in Adults With Newly Diagnosed or Recurrent Gliomas Patil CG, Walker DG, Miller DM, et al., Neurosurgery, 2019 · PMID 31069381

  The human safety and imaging study. 17 patients, 3 mg to 30 mg single IV dose, no dose-limiting toxicity, fluorescence detected in both high- and low-grade tumours. The source of the dose range quoted in the Core record.

• Chlorotoxin-directed CAR T cell therapy for recurrent glioblastoma: Interim clinical experience demonstrating feasibility and safety Barish ME, Aftabizadeh M, Hibbard J, et al., Cell Reports Medicine, 2025 · PMID 40818458

  The first-in-human CLTX-CAR T report. Four participants, feasibility and safety met, three of four with stable disease as best response, CAR-T cells detectable in tumour cavity fluid above blood levels. This is where the compound's future actually is.

• Chlorotoxin-directed CAR T cells for specific and effective targeting of glioblastoma Wang D, Starr R, Chang WC, et al., Science Translational Medicine, 2020 · PMID 32132216

  The preclinical paper that made the CAR-T application possible. CLTX-CAR T cells produced tumour regression in orthotopic xenograft glioblastoma models without off-target activity against normal tissue, and the argument for why a 36-residue peptide beats a single-chain antibody fragment against a heterogeneous tumour is laid out here.

• Chlorotoxin (UniProtKB P45639) — Leiurus quinquestriatus UniProt Consortium, UniProt Knowledgebase

  Source of the verified 36-residue sequence MCMPCFTTDHQMARKCDDCCGGKGRGKCYGPQCLCR and of the approximately 4 kDa molecular weight quoted in the Core record.