FGFRL1: A Kinase-Dead FGF Receptor-like Protein — Mechanism of Signaling, Adhesion, and Fusion

Research question. How does human FGFRL1, an atypical kinase-dead fibroblast growth
factor receptor-like protein, modulate FGF/FGFR signaling and cell adhesion/fusion in vivo?
Which mechanistic model(s) — ligand/heparin-binding decoy, inhibitory FGFR-complex component,
cytoplasmic Sprouty/Spred scaffold, or adhesion/fusion molecule — are supported, and should
FGFRL1 receive the canonical Gene Ontology term fibroblast growth factor receptor activity?

Evidence provenance. Findings below are drawn from direct FGFRL1-specific experiments
(mostly the Trueb laboratory, Bern). Where I extend from canonical FGFR-family biology I flag it
as [FGFR-family inference]. PMIDs are given for every claim; DOIs that I could not verify
from the cached abstract metadata are flagged [DOI uncached — verify].

1. Summary (answer)

FGFRL1 is best described as a multifunctional, kinase-dead ectodomain receptor whose in vivo
activity is carried almost entirely by its extracellular Ig domains, not by intracellular
signal transduction. Direct evidence supports three overlapping ectodomain-based roles — a
ligand/heparin-binding decoy (ligand sink), a constitutively dimeric, HSPG-dependent
adhesion molecule, and an active cell–cell fusogen — while the cytoplasmic Sprouty/Spred
scaffold is real biochemically but genetically dispensable in vivo. Because FGFRL1 lacks the
tyrosine kinase domain and does not transduce FGF signals by transphosphorylation, assigning it
the canonical GO term fibroblast growth factor receptor activity (which denotes kinase-dependent
signal transduction) overstates its function; FGF binding + heparin binding + cell
adhesion / cell–cell fusion / negative regulation of FGFR signaling is the accurate annotation.

2. Key findings with statistical / experimental evidence

2.1 Ligand-sink / decoy model — strongly supported (direct evidence)

• The recombinant ectodomain and the membrane-bound receptor bind FGF2, FGF3, FGF4, FGF8,
  FGF10, FGF22 with high affinity (ligand dot-blot, cell-based binding, surface plasmon
  resonance). The ectodomain is proteolytically shed near the membrane, releasing a soluble
  ligand-binding fragment, and ectopic FGFRL1 antagonizes FGFR signaling in Xenopus embryos
  (PMID 19920134; Steinberg et al., 2010, J Biol Chem — [DOI uncached — verify]).
• Original characterization: FGFRL1 binds heparin and FGF2 specifically and exerts a
  negative effect on proliferation in MG-63 cells (PMID 12813049; Trueb et al., 2003,
  J Biol Chem — [DOI uncached — verify]).
• FGF8 binds via the Ig2 domain with very high affinity (rapid on-rate, slow off-rate;
  ELISA + Biacore); constructs lacking Ig2 bind poorly (PMID 33019532; Zhuang et al., 2020,
  Int J Mol Sci — [DOI uncached — verify]).

2.2 Constitutive dimerization + HSPG-dependent adhesion — strongly supported (direct)

• FRET and co-precipitation of differentially tagged polypeptides show constitutive homodimers
  at the cell surface, enriched at cell–cell contacts. The ectodomain promotes adhesion but
  not spreading; adhesion is mediated by cell-surface heparan sulfate, is specifically
  blocked by soluble heparin (not other GAGs), is reduced by heparin-binding-site mutagenesis,
  and is neutralized by a synthetic heparin-binding-site peptide. The authors note FGFRL1
  "resembles the nectins" (PMID 18061161; Rieckmann et al., 2008, J Biol Chem —
  [DOI uncached — verify]).
• Mechanistic significance: unlike canonical FGFRs, which dimerize upon ligand + heparan
  sulfate engagement to trigger kinase activation [FGFR-family inference], FGFRL1 is
  pre-dimerized and uses HSPGs for adhesion in trans — decoupling dimerization from any
  kinase output.

2.3 Active cell–cell fusion via Ig3 + transmembrane domain — supported (direct)

• FGFRL1 is the first mammalian protein shown to actively fuse cells into large syncytia;
  cytoplasmic-mixing reporters map fusion to Ig3 + transmembrane domain (necessary and
  sufficient), and FGFRL1-transfected HEK293/HeLa fuse with untransfected CHO cells —
  implying a partner on the opposing membrane (PMID 20851884; Steinberg et al., 2010,
  J Cell Sci — [DOI uncached — verify]).
• Fusing contact sites form a net-like structure with ~1 µm pores, proposed fusion-pore
  precursors (PMID 21980560; Trueb & Steinberg, 2011 — [DOI uncached — verify]).
• In vivo correlate: FgfrL1 is upregulated during myoblast→myotube fusion; FgfrL1-null mice die
  at birth from a malformed diaphragm and specifically lack slow muscle fibers (PMID
  25172430; Amann et al., 2014, Dev Biol — [DOI uncached — verify]).
• Open question: the identity of the Ig3 trans-fusion partner is unknown.

2.4 Cytoplasmic Sprouty/Spred scaffold — real biochemically but genetically dispensable

• The C-terminal histidine-rich domain binds Spred1 and other Sprouty/Spred proteins (via
  their shared SPR domain; yeast two-hybrid + co-precipitation + membrane co-distribution), and
  Spred1 increases FGFRL1 plasma-membrane retention (PMID 21616146; Zhuang et al., 2011 —
  [DOI uncached — verify]).
• But knock-in mice lacking the three conserved intracellular motifs (FgfrL1ΔC-GFP) are
  viable, fertile, and phenotypically normal (normal diaphragm and kidney) (PMID 25126760;
  Bluteau et al., 2014 — [DOI uncached — verify]).
• Systematic domain-deletion mice: intracellular-domain deletion → normal; Ig3 deletion →
  complete absence of metanephric kidneys; Ig2 deletion → hypoplastic kidneys. The
  principal function is attributed to Ig3 (with Ig2 contributing to ligand binding) (PMID
  31923383; Gerber et al., 2020, Dev Biol — [DOI uncached — verify]).

2.5 GO annotation verdict — canonical FGF receptor activity overstates FGFRL1

• Every primary source states FGFRL1 lacks the tyrosine kinase domain required for FGF-mediated
  signal transduction by transphosphorylation (PMID 12813049, 19920134, 40699684).
• GO:0005007 fibroblast growth factor receptor activity denotes combining with FGF and
  transmitting the signal across the membrane — a kinase-dependent transduction activity FGFRL1
  does not perform. Recommended annotations: fibroblast growth factor binding (GO:0017134),
  heparin binding (GO:0008201), cell-cell adhesion, cell-cell fusion / myoblast fusion,
  and negative regulation of fibroblast growth factor receptor signaling pathway.

3. Supported vs. refuted hypotheses

Integrating view. The decoy, adhesion, and fusion activities are not mutually exclusive;
they are three read-outs of the same ligand/HSPG/partner-binding ectodomain. Notably the
Ig3 domain is required for both kidney development and fusion, and Ig2 carries
FGF8 binding — suggesting a domain-partitioned mechanism: Ig2 = ligand sink, Ig3 = trans
adhesion/fusion partner engagement.

4. Experiments that would distinguish the four models

1. Ligand-sink vs. receptor-complex (separation of function).
   Compare an Ig2-only "ligand-trap" knock-in (binds FGF8, cannot fuse) against an
   Ig3-only knock-in (fuses, weak ligand binding) in mice; if kidney rescue tracks with Ig3
   not with FGF sequestration, the essential role is adhesion/fusion, not ligand sink.
   Complement with quantitative FGF8 gradient imaging (reporter) in wild-type vs. FGFRL1-null
   metanephros to test whether FGFRL1 sharpens/limits the FGF8 field (ligand-sink prediction).

2. Physical FGFRL1·FGFR complex.
   Co-IP / proximity-ligation / BioID or split-nanoluc between endogenous FGFRL1 and FGFR1–4
   in nephrogenic cells ± FGF8/heparin; cross-linking mass spec on purified ectodomains.
   Functionally, test whether FGFRL1 cis-inhibits FGFR autophosphorylation in a defined
   co-expression system (pERK / pFRS2 dose–response) — distinguishes a true inhibitory-complex
   component from a purely extracellular ligand sink.

3. Adhesion vs. fusion (is adhesion a fusion intermediate?).
   Use heparin-binding-site mutants and HSPG-deficient (e.g., Ext1/2-null or heparinase-
   treated) target cells in the CHO trans-fusion assay: if HSPG loss abolishes adhesion but
   fusion persists, adhesion and fusion are separable; if both fail, HSPG bridging is the fusion
   trigger. Live-imaging of the ~1 µm net-like pores with membrane/cytoplasmic dyes to order
   adhesion → pore → mixing.

4. Identify the Ig3 fusion partner (the central unknown).
   Recombinant FGFRL1-Ig3 ectodomain pulldown + mass spectrometry from myoblast/diaphragm
   membranes; CRISPR loss-of-function screen in the CHO trans-fusion reporter (FGFRL1+ cells
   fuse with a genome-wide-KO CHO library — dropout identifies the required partner);
   cell-surface expression cloning for a factor conferring fusion competence on non-fusing
   cells. Validate candidate by conditional KO → expect diaphragm slow-fiber / syncytium defects
   phenocopying FgfrL1-null.

5. Cytoplasmic contribution (already largely answered).
   The ΔC and intracellular-deletion mice (25126760, 31923383) already show the tail is
   dispensable in vivo; a targeted test would be a Spred1;FgfrL1 double-mutant to ask whether
   Spred recruitment fine-tunes ERK output in a sensitized background without being essential.

5. Limitations and future directions

• Much of the mechanistic work is from one laboratory and relies on overexpression and
  heterologous (CHO/HEK/Xenopus) systems; endogenous-level, human, in-vivo confirmation is
  limited.
• A physical FGFRL1·FGFR heterocomplex is inferred but not directly demonstrated in the
  cached literature — the "inhibitory receptor-complex" model remains open.
• The Ig3 fusion partner is unidentified, leaving the fusion mechanism molecularly incomplete.
• Cancer studies (SCLC/ENO1–PI3K/Akt PMID 31957179; rectal/ovarian/ESCC MAPK PMID 32098557,
  29675438, 29963148; miR-210-3p PMID 37062273) report FGFRL1-linked signaling changes but are
  correlative/knockdown and may reflect context-specific or indirect effects rather than a
  cell-autonomous kinase activity.
• DOIs were not machine-verifiable from the available metadata and are flagged accordingly;
  PMIDs are authoritative.

6. Cited literature (PMIDs)

• 12813049 Trueb et al. 2003 — first characterization; heparin/FGF2 binding; decoy hypothesis.
• 18061161 Rieckmann et al. 2008 — constitutive dimers; HSPG-dependent adhesion; nectin-like.
• 19920134 Steinberg et al. 2010 — ectodomain shedding; binds FGF2/3/4/8/10/22; Xenopus antagonism.
• 20851884 Steinberg et al. 2010 — active cell fusion; Ig3+TM necessary/sufficient.
• 21080029 Trueb 2011 — review; high-affinity FGF/heparin binding; human craniosynostosis frameshift.
• 21616146 Zhuang et al. 2011 — Spred1/Sprouty binding via His-rich C-terminus.
• 21980560 Trueb & Steinberg 2011 — net-like pore structure during fusion.
• 23112089 Trueb et al. 2013 — kidney development; FGFRL1 interacts mainly with Fgf8.
• 25126760 Bluteau et al. 2014 — ΔC-GFP mice viable/normal (tail dispensable).
• 25172430 Amann et al. 2014 — required for slow muscle fibers.
• 31923383 Gerber et al. 2020 — domain-deletion mice; Ig3 essential for metanephric kidney.
• 33019532 Zhuang et al. 2020 — FGF8 binds via Ig2 (SPR/Biacore).
• 40699684 Guan et al. 2025 — recent review; kinase-dead, multifunctional.
• Cancer/context: 31957179, 32098557, 29675438, 29963148, 37062273 (correlative).