TM4SF20

TM4SF20 is a polytopic endoplasmic reticulum membrane protein that controls the regulated intramembrane proteolysis (RIP) of the membrane-bound transcription factor CREB3L1, thereby gating collagen synthesis (PMID:25310401). In its default state TM4SF20 blocks CREB3L1 cleavage; TGF-β relieves this block by suppressing TM4SF20 expression, permitting CREB3L1 cleavage, nuclear translocation, and activation of collagen genes (PMID:25310401). The protein's activity is dictated by its membrane topology, which ceramide reverses through regulated alternative translocation (RAT): in the absence of ceramide the N terminus of the first transmembrane helix is translocated into the ER lumen in a TRAM2-dependent manner, whereas ceramide blocks this translocation so the N terminus faces the cytosol, converting TM4SF20 from an inhibitor into an activator of CREB3L1 RIP (PMID:27499293). RAT depends on specific determinants in the first TM helix—the GXXXN motif residue Asn-26 is essential, with Pro-29 acting together with Leu-25 or Val-17, while the GXXXN motif alone is insufficient (PMID:30808712). Topology is further tuned by ceramide-sensitive retrotranslocation of a glycosylated luminal loop (N132, N148, N163) back to the cytosol independently of ER-associated degradation, which repositions the C terminus and is delayed by ceramide (PMID:36972171). A truncating deletion removing exon 3 produces a stable protein that mislocalizes to the cytoplasm rather than the plasma membrane, consistent with a toxic gain-of-function mechanism (PMID:23810381).

1. 2013 Medium

   Establishing that TM4SF20 mislocalization has pathological consequences, a truncating variant was shown to generate a stable protein that accumulates in the cytoplasm rather than reaching its normal membrane destination.

   Evidence Minigene analysis and subcellular localization of truncated vs. wild-type protein

   PMID:23810381

   Open questions at the time

   • Does not define the normal molecular function of TM4SF20
   • Toxic gain-of-function mechanism inferred from localization, not directly demonstrated
   • No link to the CREB3L1/ceramide pathway established at this point
2. 2014 Medium

   Defining TM4SF20's first known function, it was shown to inhibit RIP of CREB3L1 and to act as the node through which TGF-β relieves this block to drive collagen synthesis.

   Evidence Loss-of-function and expression studies in A549 cells with CREB3L1 cleavage and collagen gene readouts

   PMID:25310401

   Open questions at the time

   • Mechanism by which TM4SF20 blocks CREB3L1 cleavage not resolved
   • Single cell line, single lab
   • Direct physical interaction with CREB3L1 not demonstrated
3. 2016 High

   Explaining how TM4SF20 switches between inhibitor and activator states, ceramide was shown to invert its membrane topology via regulated alternative translocation, with TRAM2 required for luminal translocation in the absence of ceramide.

   Evidence Topology mapping, TRAM2 knockdown, ceramide treatment, and functional CREB3L1 cleavage assays

   PMID:27499293

   Open questions at the time

   • How the inverted topology mechanistically activates CREB3L1 RIP not detailed
   • Source/sensing of ceramide upstream not defined
   • Structural basis of the topology switch not resolved
4. 2019 High

   Defining the sequence determinants of the topology switch, mutagenesis identified Asn-26 of the GXXXN motif as essential and Pro-29 with Leu-25 or Val-17 as required, while showing the GXXXN motif alone is insufficient.

   Evidence Site-directed mutagenesis, TM4SF20–TM4SF4 chimera analysis, and topology assays

   PMID:30808712

   Open questions at the time

   • How these residues mechanistically govern translocation direction unknown
   • No structural model of the first TM helix during RAT
   • Additional non-helix determinants not excluded
5. 2023 High

   Revealing a second layer of topological control, the luminal loop was mapped to glycosylated N132/N148/N163 and shown to retrotranslocate to the cytosol independently of ERAD, a step ceramide delays to stabilize the originally synthesized form.

   Evidence Glycosylation site mapping, fractionation, topology assays, and glycosylation-site mutagenesis with/without ceramide

   PMID:36972171

   Open questions at the time

   • Machinery driving the ERAD-independent retrotranslocation unidentified
   • Functional consequence of C-terminal repositioning for CREB3L1 control not fully linked
   • How ceramide delays retrotranslocation mechanistically unknown
6. 2021 Low

   Linking the topology mechanism to physiology, ceramide-induced RAT of TM4SF20 was reported as important for doxorubicin chemotherapy effectiveness.

   Evidence Functional cellular assay reported in a review/book chapter

   PMID:32986129

   Open questions at the time

   • Reported in a review without detailed experimental methods in the abstract
   • Mechanistic link between RAT and drug response not delineated
   • Not independently confirmed in the timeline

• How the topology-determined inhibitor/activator states physically engage the CREB3L1 RIP machinery, and the identity of the retrotranslocation apparatus, remain unresolved.
• No defined direct interaction between TM4SF20 and CREB3L1 or its proteases
• ERAD-independent retrotranslocation machinery unidentified
• Upstream ceramide-generating signals controlling RAT undefined