From:  Do new neurons grow in the adult human hippocampus? A review of the evidence

 Common adult hippocampal neurogenesis-related markers and key caveats.

MarkerWhat is it used to inferMajor caveats (alternative explanations/confounds)Best-practice interpretation/controls
DCX (doublecortin)Immature neurons/neuronal lineage progression (late progenitor to immature neuron stage)Epitope sensitivity to post-mortem interval and prolonged fixation; low-level expression can be missed (false negatives). Can also reflect prolonged maturation or plasticity-state changes rather than recent birth.Use with additional markers (e.g., PROX1, NEUROD1, calretinin) and anatomical localization to SGZ/GCL. Report fixation/PMI and antigen retrieval. Validate with multiple antibodies and negative controls.
PSA-NCAMStructural plasticity; often used as an ‘immature neuron’ proxy (especially with DCX)Not specific to newborn neurons; can label plasticity-related states and some mature interneurons. Signal depends on tissue handling and antibody/thresholding.Avoid using alone; interpret as a plasticity marker unless co-localized with neuronal lineage markers and appropriate SGZ distribution. Include isotype controls and evaluate non-SGZ staining.
Ki-67Active cell cycling (proliferation)Marks all cycling cells (not neuron-specific); low event rate in adult human DG increases sampling error. Can be affected by agonal state, hypoxia, inflammation, and tissue quality.Combine with progenitor markers (SOX2, Nestin) and lineage markers; quantify regionally (SGZ) and report counting frame and total cells sampled.
MCM2/Other MCM proteinsLicensing of DNA replication; proliferation competenceNot neuron-specific; can label cells poised for division. Like Ki-67, very sparse events in adult human tissue.Use with cell-type markers and SGZ localization; interpret as evidence of proliferation capacity, not neuron production by itself.
PCNADNA replication/repair-associated protein; sometimes used for proliferationLess specific than Ki-67; can be elevated during DNA repair and in non-dividing contexts, increasing false positives.Prefer Ki-67/MCM2 for proliferation; if used, pair with additional proliferation markers and stringent controls.
SOX2Neural stem/progenitor-like stateAlso expressed outside the DG niche, depending on context, it does not prove ongoing neurogenesis; affected by tissue quality.Interpret as progenitor pool indicator; combine with proliferation markers (Ki-67/MCM2) and SGZ localization.
NestinNeural progenitor/radial glia-like morphology (in some contexts)Not exclusive to neural stem cells; expression depends on antibody and fixation; may label reactive glia after injury.Use with SOX2 and morphology (radial processes) plus SGZ location; consider injury/inflammation status.
TBR2 (EOMES)Intermediate progenitors (transit-amplifying)Very rare in adult human DG; detection sensitive to tissue handling and section sampling; not neuron birth by itself.Use to support the presence of progenitor stages when co-detected with proliferation markers and SGZ confinement.
NEUROD1Neuronal differentiation program (early neuronal lineage commitment)Can be transient and sparse; transcript/protein detection varies by assay sensitivity; interpretation depends on co-expression patterns.Use as part of a lineage panel (with DCX/PROX1) and anatomical context; corroborate with orthogonal methods.
PROX1Dentate granule cell lineage identity (granule-cell fate)Lineage identity marker, not an ‘age’ marker; does not distinguish newborn vs. mature granule cells on its own.Use to confirm granule-cell identity of DCX+ cells; combine with maturity markers (calbindin) and stage markers (calretinin/DCX).
CalretininImmature to early post-mitotic neuron stage (often used with DCX)Not exclusive to newborn granule cells; can be expressed in interneurons and varies by region; sparse signals are prone to sampling error.Interpret as supportive when co-localized with DCX/PROX1 in SGZ/GCL; include cell-type disambiguation where needed.
CalbindinMature granule cell phenotypeReduction can reflect dematuration or stress/injury effects; not direct evidence for or against neuron birth.Use to define maturity gradients and to interpret ‘immature-like’ states; avoid using as a sole proxy for neurogenesis.
snRNA-seq ‘immature neuron’ clusters (e.g., DCX/NEUROD1/PROX1 programs)Rare immature granule-cell states; potential ongoing neurogenesisDropout and sparse capture reduce sensitivity; nuclei assays under-sample cytoplasmic transcripts; clustering thresholds, doublets, and ambient RNA can blur rare states. Immature-like signatures can also reflect prolonged maturation, dematuration, or plasticity.Require convergence: progenitor/proliferation signatures, SGZ localization (spatial), and/or independent birth-dating. Report QC, doublet filtering, batch correction, and sensitivity analyses.
Spatial transcriptomics neurogenic signaturesAnatomical mapping of gene programs to DG nichesLower sensitivity on some platforms; many technologies capture multi-cell spots, causing signal dilution; deconvolution assumptions can bias rare-cell inference.Pair with immunohistochemistry and high-resolution segmentation; interpret rare signatures conservatively; report platform resolution and detection thresholds.

SGZ: subgranular zone; PSA-NCAM: polysialylated neural cell adhesion molecule; DG: dentate gyrus; MCM2: minichromosome maintenance protein 2; PCNA: proliferating cell nuclear antigen; snRNA-seq: single-nucleus RNA sequencing; QC: quality control.