CATCHFIRE: the reversible, self-reporting protein dimerizer

CATCHFIRE: fluorogenic, reversible chemically induced dimerization

Chemically induced dimerization (CID) uses a small molecule to bring two proteins together inside living cells, so that an interaction, a location or an activity can be switched on at a chosen moment. CATCHFIRE (Chemically Assisted Tethering of CHimera by Fluorogenic-Induced Recognition) is a fluorogenic CID technology: one small molecule brings two chosen proteins together and makes their new assembly fluorescent, so you can trigger a protein interaction, watch it happen at the same moment, and reverse it by washing the molecule out.

CATCHFIRE was introduced in 2023 (Bottone et al., Nature Methods 2023) by the teams of Arnaud Gautier and Franck Perez, co-founders of The Twinkle Factory. The journal’s News & Views described it as the first fluorogenic, reversible CID system built on a small peptide tag (Wang and Shen, Nature Methods 2023).

The CATCHFIRE principle: a fluorogenic match glue recruits FIREmate- and FIREtag-fused proteins into a reversible, fluorescent ternary assembly (left), using dyes that span green to far-red (right).

How CATCHFIRE works

CATCHFIRE comes from splitting the 14 kDa protein tag pFAST into two pieces: FIREmate, a 114-amino-acid domain (12 kDa), and FIREtag, a peptide of just 11 amino acids (1.4 kDa). Each is fused to one of the two proteins you want to bring together, and on their own the two pieces stay apart. A fluorogenic “molecular glue” of the match family binds both at once, assembles them and lights up: the match glues brighten more than 100-fold when the three-part complex forms, reaching about 80% of the brightness of full-length pFAST (Bottone et al., Nature Methods 2023).

• Fast: the two proteins come together with a half-time of about 25 seconds after the glue is added, and half-maximal dimerization needs only about 100 to 500 nM of match.
• Reversible, again and again: washing the glue out separates them with a similar half-time, so full assembly and full release each take about 100 seconds, and the catch-and-release cycle can be repeated.
• Self-reporting: the glue is the fluorophore, so recruitment is read out directly, with no separate reporter to build.
• Several colours: ^match540 (green, emission 540 nm), ^match550 (green-yellow, 550 nm) and ^match600 (orange-red, 600 nm); ^matchDark, non-fluorescent, drives proximity without adding a signal. A near-infrared version, nirCATCHFIRE, pairs FIREtag with a dedicated nir-FIREmate and is lit by ^match715 (emission 715 nm) (El Hajji et al., Nat. Commun. 2025).
• Small and cell-permeant: the match glues enter cells readily, and the original paper describes CATCHFIRE as the chemically induced proximity tool with the smallest dimerizing domains.
• A zero-background option: when expressed at high levels, FIREtag and FIREmate can pair weakly on their own. A caged FIREtag masks the tag until the glue arrives, for experiments where even a trace of basal pairing matters.
• Plays well with others: CATCHFIRE runs alongside rapamycin FKBP-FRB in the same cell, and its coloured glues belong to the same hydroxybenzylidene rhodanine family as the FAST ^TFFluorogens.

How CATCHFIRE compares with other proximity tools

Controlling protein proximity is a cornerstone of modern cell biology, and CATCHFIRE fills a gap the older tools leave open:

• vs. classical chemical dimerizers (rapamycin FKBP-FRB, gibberellin GID1-GAI, abscisic acid PYL-ABI1): these induce proximity but do not report it, so a separate fluorescent tag or split reporter is needed to see the event. Association is often fast, but dissociation is not: rapamycin binds FKBP with a K_D of about 0.2 nM, which makes the assembly practically irreversible, and rapamycin’s own biological activity limits its use, which is why rapalogs were developed (Shen et al., Chemistry-Methods 2024). CATCHFIRE reports the event itself and reverses by simple washout, repeatedly.
• vs. optogenetic dimerizers (CRY2-CIB1, iLID, eMags): light gives millisecond switching and subcellular precision that a diffusing small molecule cannot match. The costs are illumination hardware, controls for light-induced side effects, and blue light that penetrates tissue poorly. CATCHFIRE needs no special hardware; the two approaches are complementary.
• vs. split fluorescent proteins (split-GFP, BiFC): these report an interaction but complement slowly and irreversibly, which stabilises the very assembly they are meant to observe. splitFAST reports interactions reversibly; CATCHFIRE is the member of the family engineered to induce them.

Good to know: reversing CATCHFIRE means washing the glue out (about 100 seconds), slower than light-cleavable inducers and harder to do in tissue; the glue has to stay in the medium, at low-micromolar levels, for as long as you want the proteins together; and in vivo use is so far limited to the chick embryo.

What you can do with CATCHFIRE

A sampling of CATCHFIRE applications: membrane and organelle recruitment, transport, translocation and biosensing (from Bottone et al., Nat. Methods 2023).

Because a single reagent both actuates and reports proximity, CATCHFIRE covers a broad range of uses (Bottone et al., Nature Methods 2023):

• Targeted recruitment: reroute a protein to the ER, Golgi, plasma membrane or mitochondrial surface within minutes, and watch it arrive.
• Organelle transport: hook lysosomes onto a kinesin motor and drive them to the cell periphery in 30 to 45 minutes, then let them return by washout.
• Nuclear trapping and export: hold a shuttling protein in the nucleus on a chromatin anchor, or force it out, and release it again by washout.
• Secretion assays: retain a cargo in the ER, then release it by washout and follow it to the Golgi within 10 minutes and to the plasma membrane by 60.
• Biosensors: switch on a signal when caspase-3 cleaves its substrate during apoptosis, or when NF-κB p65 enters the nucleus.
• Mitophagy on demand: recruit PINK1 to mitochondria to trigger their removal.

CATCHFIRE in recent research

Since its 2023 debut, CATCHFIRE has moved from a method to a platform, used well beyond its original lab:

• Neurogenesis in a living embryo. In the chick embryo neural tube, a reversible CATCHFIRE tether between mitochondria and a kinesin motor, held for about one cell division, made mitochondria segregate unequally and raised asymmetric progenitor-neuron divisions from 20% to 53%, linking mitochondrial inheritance to neuronal fate (Bunel et al., Nat. Commun. 2025).
• Near-infrared dimerization. nirCATCHFIRE moves the glue signal into the far-red and near-infrared, leaving the green and red channels free for other reporters (El Hajji et al., Nat. Commun. 2025).
• Chemically controlled genes and enzymes. CATCH-ON switches gene expression on by up to 160 to 200-fold with ^match550, in a dose-dependent way and with low leakiness; the same design gives match-controlled split proteases, CRE recombinase and NanoLuc (Franco Pinto et al., ACS Chem. Biol. 2026).
• Secretion screening in bacteria. Adding FIREmate and ^match540 to Bacillus subtilis culture supernatant turns secreted FIREtag fusions into a one-step fluorescence reading; library screens raised nanobody secretion from 26 to 112 mg/L (Gao et al., J. Agric. Food Chem. 2025).
• Dissecting a transcription factor. Forcing chosen STAT5 homodimers and heterodimers showed that a single intact transactivation domain per dimer is enough for nuclear import (Ernst et al., Commun. Biol. 2026).
• Recording cell history. CATCHFiber uses a CATCHFIRE nuclear trap to write 30-minute time stamps into a protein fibre growing inside the cell, so a single image recovers hours of history (El Hajji and Gautier, bioRxiv 2026, preprint).

CATCHFIRE shares its chemistry with the FAST family, which is why the two technologies interoperate so naturally. Get started with our match molecular glues and kits.

Explore the peer-reviewed evidence

CATCHFIRE and its applications are documented in a growing body of peer-reviewed work. Browse the searchable database below, filter by year or publication type, with a direct link to every paper.