What Is AP-1? Transcription Factor Explained

By Dr. Zubair Khalid, DVM, MS, PhD · 2026-09-20

AP-1 (activator protein 1) is a family of dimeric transcription factors, not a single protein, assembled from basic-region leucine-zipper (bZIP) subunits in the Jun, Fos, Fra, and ATF protein families. Each functional AP-1 complex is a pair of these subunits, and the identity of the pair determines which DNA element it recognizes and which genes it turns on or off.

AP-1 matters because it sits at the convergence point of several major signaling cascades. Growth factors, inflammatory cytokines, phorbol esters, oxidative stress, and ultraviolet radiation all funnel into AP-1 through mitogen-activated protein kinase (MAPK) pathways. The complex then converts those signals into changes in gene expression that decide whether a cell divides, matures, repairs damage, or dies. Because AP-1 controls proliferation and survival genes, its dysregulation appears in cancer, in developmental toxicity, and in chronic inflammatory and oxidative disease states [1][2][3].

Why AP-1 Is a Family, Not a Protein

The name "AP-1" was originally assigned to a DNA-binding activity that recognized a specific enhancer element in the human metallothionein IIA gene and in the SV40 viral genome. When researchers purified that activity, they found it was not one polypeptide. It was a mixture of dimers formed from a pool of related proteins. That is why the precise phrasing matters: AP-1 is an activity, and the proteins that produce it vary by cell type, by developmental stage, and by the signals the cell has received.

Wang and Studzinski described AP-1 as a heterogeneous entity composed in mammalian cells of dimers chosen from a group of at least eight proteins belonging to three families: Jun, Fos, and ATF [4]. Later work expanded the roster to include the Fos-related antigens Fra-1 and Fra-2, plus JunB and JunD, and placed the whole system under the bZIP structural class [5].

The practical consequence is that two cells can both have "active AP-1" and still express completely different target genes, because their AP-1 dimers are built from different subunits. Composition, not just abundance, is the regulatory variable.

The Jun Family

The Jun family has three members in mammals: c-Jun, JunB, and JunD. Jun proteins share two properties that define them as the backbone of the system.

First, Jun proteins can homodimerize. A c-Jun:c-Jun homodimer is a functional AP-1 complex. Second, Jun proteins are the preferred DNA-binding half of most heterodimers. In Fos-Jun heterodimers, Jun binds preferentially to the consensus half-site, while Fos tolerates nonconsensus sequence [6].

c-Jun is the most studied member. It is a classic immediate-early gene product, induced rapidly and transiently after growth factor or stress stimulation, and it is a direct phosphorylation target of c-Jun N-terminal kinase (JNK).

The Fos Family

The Fos family includes c-Fos, FosB, Fra-1 (FOSL1), and Fra-2 (FOSL2). The defining feature of this family is a structural constraint: Fos proteins cannot homodimerize and cannot bind DNA as monomers. They must partner with a Jun protein or an ATF protein to form a complex that binds DNA.

That asymmetry is the single most important fact for students to internalize. Jun can stand alone as a homodimer. Fos cannot. When a cell needs Fos-driven transcription, it must also supply a Jun partner.

Fos proteins also carry transactivation domains that Jun homodimers lack in the same configuration, so adding Fos to a Jun dimer changes both the strength and the character of the transcriptional response.

The ATF and Other bZIP Partners

ATF2 and ATF7 belong to the activating transcription factor branch of the bZIP superfamily. ATF proteins can heterodimerize with Jun proteins, producing dimers such as c-Jun:ATF2 that behave differently from c-Jun:c-Fos [7][8].

ATF2 is a substrate for both JNK and p38 MAPKs, and its role in tumorigenesis has been reviewed alongside its developmental functions [9]. ATF2 and Fos are described as competitive monomers that can occupy the same dimerization pool, meaning the relative levels of Fos and phosphorylated ATF2 shift which dimers form [10].

The Maf family of bZIP proteins also contributes to some AP-1 complexes, and the broader definition of AP-1 includes Jun, Fos, Maf, and ATF family DNA-binding proteins [9].

The bZIP Structure: How AP-1 Dimers Work

Every AP-1 subunit has the same two-domain architecture, and understanding it explains nearly all AP-1 behavior.

The Leucine Zipper

The leucine zipper is a stretch of roughly 30 amino acids in which leucine residues appear at every seventh position along an alpha helix. Because leucine is hydrophobic and the helix has about 3.5 residues per turn, these leucines line up along one face of the helix. Two such helices interdigitate like a zipper, holding the two subunits together.

Dimerization specificity lives here. The zipper determines which subunits can pair with which. This is why Fos cannot pair with Fos, and why the pool of possible dimers is constrained rather than random.

The Basic Region

Immediately N-terminal to the zipper is the basic region, a cluster of positively charged amino acids. When two subunits dimerize, their basic regions align to form a continuous DNA-binding surface that sits in the major groove.

The basic region reads the DNA sequence. The zipper holds the two basic regions in the correct orientation to do so. Disrupt either domain and the complex fails. This is exactly how dominant-negative reagents work: a mutant such as A-Fos carries an intact zipper but a disrupted basic region, so it heterodimerizes with wild-type partners and produces dimers that cannot bind DNA [7].

DNA Recognition

A canonical AP-1 dimer contacts about 7 base pairs of DNA, with each subunit's basic region reading roughly half of that. The consensus element is the TPA response element, or TRE, written 5'-TGAG/CTCA-3'. The slash notation reflects the palindromic nature of the site: the two half-sites are inverted relative to each other so that the dimer can bind with its twofold symmetry axis at the center.

The cAMP response element, or CRE, written 5'-TGACGTCA-3', is closely related. The difference is a single central base pair. Jun homodimers and Jun-Fos heterodimers bind TRE with high affinity. Jun-ATF2 heterodimers bind CRE-like elements preferentially. This is the mechanistic basis for the statement that dimer composition dictates DNA-binding preference.

Binding is not always symmetric. Fos-Jun heterodimers bind nonconsensus AP-1 sites in a preferred orientation, because Fos and Jun recognize the two half-sites nonidentically [6]. Orientation then affects how AP-1 cooperates with neighboring factors such as NFAT1 at composite regulatory elements.

Table: AP-1 Subunit Combinations and DNA-Binding Preferences

The table is a guide, not a rulebook. Context, post-translational modification, and neighboring transcription factors all modulate what a given dimer actually does at a given promoter.

Step by Step: How an AP-1 Signal Becomes a Gene Expression Change

Step 1: Extracellular Signal

A growth factor, cytokine, phorbol ester such as PMA (also called TPA), or a physical stressor such as ultraviolet light or reactive oxygen species engages receptors or damages macromolecules.

Step 2: MAPK Cascade Activation

The signal is relayed through a kinase cascade. Three MAPK modules dominate AP-1 regulation: ERK1/2, JNK, and p38. MEKK1 is a MAPK kinase kinase that regulates both ERK1/2 and JNK, and it also acts as a ubiquitin ligase [11].

Step 3: Phosphorylation of Pre-Existing Subunits

JNK phosphorylates c-Jun on its N-terminal transactivation domain, increasing its transcriptional potency without necessarily changing its abundance. p38 phosphorylates ATF2. These modifications convert a bound but weak complex into a strong activator.

Step 4: Induction of Immediate-Early Genes

ERK and JNK signaling rapidly induces c-fos and c-jun transcription. c-fos is a classic immediate-early gene, and its protein product Fos is widely used as a marker of recent neuronal activity [12].

Step 5: Dimer Assembly and Nuclear Translocation

Newly synthesized Fos and Jun proteins enter the nucleus, where the leucine zipper drives dimerization. In bovine endometrial epithelial cells, knockdown of the protein IFI6 prevented c-Jun/c-Fos heterodimerization and nuclear translocation, which blocked proliferation [13]. That experiment shows that dimerization and localization are regulated steps, not automatic ones.

Step 6: DNA Binding at TRE or CRE

The dimer binds its cognate element in the regulatory region of a target gene. Binding can be cooperative with adjacent factors. NFAT and Fos-Jun cooperate on composite NFAT-AP-1 elements, which integrates calcium and protein kinase C/Ras signaling in T cells [14].

Step 7: Coactivator Recruitment and Transcription

Bound AP-1 recruits coactivators. The LIM domain protein Trip6, in its nuclear isoform, is a selective coactivator that interacts only with Fos family members, so it activates Fos-containing dimers but not c-Jun:ATF2 dimers [8]. This is a clean example of how dimer composition determines which cofactors can be recruited.

Step 8: Feedback and Compositional Change

AP-1 regulates its own components. After CRH stimulation of corticotroph cells, AP-1 complexes initially recruited c-Fos and JunB, then later shifted to Fra-2 and JunD. The later Fra-2/JunD dimers suppressed CRH induction of c-fos mRNA and the recruitment of c-Fos/JunB, creating an autoregulatory loop that prevents overexpression of the POMC gene [15]. MEKK1 similarly controls the AP-1 repertoire by suppressing JunB mRNA and by promoting Fra-2 ubiquitination and degradation [11].

What AP-1 Actually Does in Cells

Proliferation

AP-1 drives entry into the cell cycle through targets that include cyclins and components of the replication machinery. In bovine endometrial epithelial cells, IFI6 sustains proliferation by acting upstream of c-Jun/c-Fos and maintaining ERK1/2-dependent heterodimerization and nuclear translocation [13]. When AP-1 components were suppressed in areca nut extract treated KB carcinoma cells, proliferation cell nuclear antigen (PCNA) and cyclin dependent kinase 4 (CDK4) protein levels fell, alongside reduced c-Jun and c-Fos [1].

Differentiation

AP-1 is required for at least some differentiation programs, and its composition changes as cells mature. In HL60 leukemia cells induced toward monocytic differentiation by 1,25-dihydroxyvitamin D3, functional AP-1 was required, and the subunits occupying the TRE changed as differentiation progressed. Early-stage cells used c-jun, ATF-2, fos-B, fra-1, and fra-2. Cells with a more established monocytic phenotype used c-jun, ATF-2, jun-B, and fos-B [4].

In osteoblasts, all seven AP-1 members peak during proliferation and decline during matrix maturation. In fully differentiated osteoblasts, Fra-2 and JunD predominate, and only the Fra-2/JunD pair stimulated an osteocalcin reporter construct when coexpressed, while other pairs down-regulated it [16].

Apoptosis

AP-1 can promote or oppose apoptosis depending on context and dimer identity. NFAT cooperation with Fos-Jun is essential for the activation-induced cell death program in T cells, and blocking that cooperation prevented the apoptotic program even though NFAT could still drive other genes [14]. In areca nut extract treated KB cells, AP-1 subunit expression fell while the pro-apoptotic protein Bax and the cell cycle inhibitors p16 and p21 increased, and p53, retinoblastoma protein (Rb), and CDK4 decreased [1].

Stress Responses

AP-1 is an oxidant-responsive transcription factor. In rats maintained on a vitamin E-deficient diet for six months, adrenal AP-1 DNA-binding activity dropped by roughly 55 percent, and this loss was not due to a change in dimeric composition but to a general down-regulation of Fos and Jun family protein levels [17]. In human aldosterone-producing adenomas, oxidative stress induced FOS and JUN upregulation and activation, and co-overexpression of FOS and JUN suppressed steroidogenic genes including StAR, CYP11B1, and CYP11B2 while reducing aldosterone and cortisol secretion and increasing reactive oxygen species [3].

The Fosl2/c-Jun dimer links AP-1 to endoplasmic reticulum stress. In ischemic cerebral infarction models, Fosl2 and c-Jun were elevated, the Fosl2/c-Jun dimer transcriptionally activated Lcn2 by binding its promoter, and knockdown of either subunit reduced oxidative stress, infarct extent, and ER stress markers [18].

Neuronal Plasticity

Fos combines with Jun to form AP-1, and this complex has been implicated in neuronal plasticity, apoptosis, and regeneration [12]. In the auditory brainstem, electrical intracochlear stimulation increased Fos expression, and Fos expression correlated with a locally matching decrease in phosphorylated ATF2 in the anteroventral cochlear nucleus and lateral superior olive [10]. Because Fos and pATF2 compete for the same dimerization pool, a rise in one shifts the composition of the complexes that form.

How AP-1 Is Studied in the Lab

Electrophoretic Mobility Shift Assay (EMSA)

A radiolabeled or fluorescently labeled TRE oligonucleotide is incubated with nuclear extract. If AP-1 dimers are present and active, they bind the probe and the complex migrates more slowly through a polyacrylamide gel than free probe. Supershift assays add antibodies against specific subunits, and a shifted band identifies which protein is in the complex. This is how the changing composition of AP-1 during HL60 differentiation was tracked [4].

Reporter Assays

A promoter containing AP-1 sites drives a luciferase or chloramphenicol acetyltransferase reporter. Induction by PMA, by c-Fos and c-Jun coexpression, or by a constitutively active MEKK1 truncation all increase reporter output, and dominant-negative bZIP mutants such as A-Fos and A-ATF2 reduce it [7]. This design lets you test which subunit is required for a given promoter.

Chromatin Immunoprecipitation (ChIP)

ChIP fixes protein-DNA contacts in living cells, shears chromatin, immunoprecipitates with an antibody against a specific AP-1 subunit, and identifies the bound DNA by quantitative PCR or sequencing. In the ischemic infarction study, ChIP and dual-luciferase assays together confirmed that Fosl2/c-Jun binds the Lcn2 promoter [18].

Computational Docking and Molecular Dynamics

When no small-molecule inhibitor is available, structure-based screening can predict whether a compound binds the DNA-bound AP-1 complex. Docking of Satureja hortensis phytochemicals against the DNA-bound AP-1 structure (PDB entry 1FOS) predicted that cedrelanol and allo-aromadendrene bound with affinities of -7.7 kcal/mol each, compared with -7.3 kcal/mol for the reference inhibitor SR11302. A 100-nanosecond molecular dynamics simulation of the cedrelanol-1FOS complex kept the protein backbone RMSD between 0.8 and 1.2 angstroms and the ligand RMSD below 1.3 angstroms, and MM/GBSA analysis gave a binding free energy of -75.04 kcal/mol [19]. These are computational predictions, not measured cellular effects, and should be read that way.

Expression Profiling and Survival Analysis

RNA sequencing of tumor cohorts combined with Kaplan-Meier statistics can link individual AP-1 subunit levels to clinical outcomes. In 373 ovarian carcinoma patients from The Cancer Genome Atlas, high expression of JUNB alone or in combination with FOSB, FOSL1, or FOSL2 correlated with shorter survival, while combinations with JUN, JUND, or FOS did not [2]. This is a strong illustration of the central principle: which subunit is elevated matters more than whether "AP-1" is elevated.

Comparative and Clinical Relevance

Cancer

AP-1 sits downstream of the most frequently deregulated signaling events in cancer, and MAPK pathway components are among the most commonly altered in tumors [9]. AP-1 plays a pivotal role in initiating and maintaining expression of the human papillomavirus oncoproteins E6 and E7 during HPV-linked cervical carcinogenesis [5]. In non-tumorigenic HeLa-fibroblast hybrids, Jun family members heterodimerized mainly with Fra-1, and c-Fos was completely absent from AP-1 complexes. In tumorigenic segregants and HeLa cells, Fra-1 was low and c-Fos was abundant. Ectopic expression studies in that system linked the shift in AP-1 composition to conversion to invasive growth [5].

Developmental Toxicity

Valproic acid is a widely used antiepileptic drug and a known teratogen that induces developmental defects in part through aberrant activation of the AP-1 (c-Fos/c-Jun) pathway. Computational screening has proposed plant terpenoids as plausible AP-1 modulators in this context, providing a molecular rationale for further investigation [19].

Adrenal and Endocrine Biology

FOS and JUN coordinate a redox-steroidogenesis axis in the adrenal cortex. Phosphorylated FOS and JUN were detected exclusively in adrenal cortex adjacent to functional adenomas, with negligible levels in cortex adjacent to non-functional adenomas and in normal adrenal cortex [3]. This spatial restriction suggests AP-1 activation is tied to the functional state of the adjacent tissue rather than to adenoma presence alone.

Yeast as a Model

The budding yeast AP-1 homolog Yap1 is required for the oxidative stress response and also confers caffeine tolerance. Yap1 overexpression rendered cells resistant to high caffeine concentrations, and low hydrogen peroxide concentrations induced Yap1 activation that restored viability against caffeine toxicity. Notably, oxidative stress-mediated adaptation to caffeine required Yap1 but not its target Flr1 [20]. The core logic of an oxidant-responsive bZIP factor is conserved from yeast to mammals.

Common Mistakes and Limitations

Treating AP-1 as one protein. The most frequent error is writing "AP-1 binds the promoter" as if AP-1 were a single entity. Always specify the dimer when the identity is known. A c-Jun:c-Fos dimer and a Fra-2:JunD dimer bind overlapping but distinct elements and produce different outcomes [16][8].

Assuming Fos can homodimerize. It cannot. Any statement implying a Fos:Fos complex is incorrect. Fos must partner with Jun or ATF.

Assuming all TRE binding means the same thing. The TRE is a consensus element, and many functional AP-1 sites are nonconsensus. Fos-Jun heterodimers bind nonconsensus sites in a preferred orientation, and orientation affects cooperativity with neighboring factors such as NFAT1 [6].

Ignoring post-translational modification. JNK phosphorylation of c-Jun and p38 phosphorylation of ATF2 change transcriptional output without changing protein levels. Measuring subunit abundance alone can miss an activated complex.

Confusing correlation with mechanism. Survival correlations between a subunit and patient outcome, such as the JUNB finding in ovarian carcinoma [2], identify an association. They do not establish that the subunit causes the outcome.

Overreading computational predictions. Docking and molecular dynamics scores describe predicted binding energetics in silico [19]. They are hypotheses for bench testing, not evidence of cellular activity.

Assuming AP-1 always promotes proliferation. AP-1 is required for monocytic differentiation of HL60 cells [4] and for osteoblast differentiation programs [16], and it can drive apoptosis through NFAT cooperation [14]. Direction depends on context.

Individual experimental systems and clinical scenarios vary, and any interpretation of AP-1 data in a specific disease context requires expert evaluation.

Quick Review

1. AP-1 is a family of dimeric bZIP transcription factors, not a single protein.
2. Jun proteins homodimerize. Fos proteins cannot and must partner with Jun or ATF.
3. Dimer composition determines DNA-binding preference: TRE for Jun-Fos, CRE-like elements for Jun-ATF2.
4. The leucine zipper controls which subunits pair, and the basic region reads the DNA sequence.
5. MAPK cascades (ERK, JNK, p38) converge on AP-1 through phosphorylation and immediate-early gene induction.
6. AP-1 composition changes over time after stimulation, creating autoregulatory feedback loops [15][11].
7. AP-1 contributes to proliferation, differentiation, apoptosis, oxidative stress responses, and neuronal plasticity, with the outcome set by context and dimer identity.

Frequently Asked Questions

Is AP-1 a single protein?

No. AP-1 is a family of dimeric transcription factors assembled from Jun, Fos, Fra, and ATF subunits. The functional unit is always a pair of bZIP proteins.

Can Fos form a homodimer?

No. Fos proteins cannot homodimerize or bind DNA alone. They must pair with a Jun or ATF protein to create a DNA-binding complex.

What DNA sequence does AP-1 bind?

The canonical AP-1 site is the TPA response element, 5'-TGAG/CTCA-3'. Jun-ATF2 dimers prefer the related cAMP response element, 5'-TGACGTCA-3', which differs by one central base pair.

Why does AP-1 dimer composition matter?

Because the two subunits jointly determine which DNA element is recognized, which coactivators can be recruited, and whether the target gene is activated or repressed. A Fos-containing dimer can recruit the coactivator Trip6, while a c-Jun:ATF2 dimer cannot [8].

Does AP-1 promote cancer?

AP-1 is deregulated in many cancers and contributes to proliferation and survival gene expression, and specific subunit combinations correlate with patient outcomes [2][9]. It also participates in normal differentiation and can promote apoptosis, so its role is context dependent.

How do researchers measure AP-1 activity?

Common methods are electrophoretic mobility shift assays with TRE probes, promoter reporter assays, chromatin immunoprecipitation, and expression profiling of individual subunits. Each method answers a different question about abundance, DNA binding, or transcriptional output.

Related Articles

• Transcription Factors: Mechanisms, Regulation, and Roles in Gene Expression
• Cooperative Binding of Transcription Factors: How to Detect and Quantify Synergy in DNA Binding
• Autoregulation of Transcription Factors: How to Identify and Validate Negative and Positive Feedback Loops
• Transcription Factor Activity Inference from Single-Cell Data: From SCENIC to Decoupling
• How Activators and Repressors Fine-Tune Gene Expression: A Practical Guide to Transcription Factor Function
• Elongation Factors and the Control of Transcription: How P-TEFb and NELF Regulate RNA Polymerase II Pausing and Release

Sources

1. Regulation of Jun and Fos AP-1 transcription factors by JNK MAPKs signaling cascade in areca nut extract treated KB cells.
2. Prognostic Relevance of FOS and JUN Family Members and Immune Cell Infiltration for the Survival of Patients With Ovarian Carcinoma.
3. FOS and JUN regulate oxidative stress and steroidogenesis in human aldosterone-producing adenomas.
4. The requirement for and changing composition of the activating protein-1 transcription factor during differentiation of human leukemia HL60 cells induced by 1,25-dihydroxyvitamin D3.
5. Conversion of HPV 18 positive non-tumorigenic HeLa-fibroblast hybrids to invasive growth involves loss of TNF-alpha mediated repression of viral transcription and modification of the AP-1 transcription complex.
6. Asymmetric recognition of nonconsensus AP-1 sites by Fos-Jun and Jun-Jun influences transcriptional cooperativity with NFAT1.
7. Regulation and composition of activator protein 1 (AP-1) transcription factors controlling collagenase and c-Jun promoter activities.
8. Restriction to Fos family members of Trip6-dependent coactivation and glucocorticoid receptor-dependent trans-repression of activator protein-1.
9. The roles of ATF2 (activating transcription factor 2) in tumorigenesis.
10. Counter-regulation of the AP-1 monomers pATF2 and Fos: Molecular readjustment of brainstem neurons in hearing and deaf adult rats after electrical intracochlear stimulation.
11. MEKK1 regulates the AP-1 dimer repertoire via control of JunB transcription and Fra-2 protein stability.
12. The pattern of Fos expression in the rat auditory brainstem changes with the temporal structure of binaural electrical intracochlear stimulation.
13. Interferon-tau-induced IFI6 sustains bovine endometrial epithelial cell proliferation by activating AP-1 via the JAK-STAT signaling pathway†.
14. Gene expression elicited by NFAT in the presence or absence of cooperative recruitment of Fos and Jun.
15. Long-term variations of AP-1 composition after CRH stimulation: consequence on POMC gene regulation.
16. Developmental expression and activities of specific fos and jun proteins are functionally related to osteoblast maturation: role of Fra-2 and Jun D during differentiation.
17. Suppression of steroidogenesis and activator protein-1 transcription factor activity in rat adrenals by vitamin E deficiency-induced chronic oxidative stress.
18. Fosl2/c-Jun dimer induces oxidative stress and endoplasmic reticulum stress by enhancing Lcn2 transcription to promote ischemic cerebral infarction.
19. Computational Studies of Satureja hortensis Phytochemicals as Inhibitors of AP-1 (FOS/JUN) Activation in Valproic Acid Teratogenicity.
20. Yap1-mediated Flr1 expression reveals crosstalk between oxidative stress signaling and caffeine resistance in Saccharomyces cerevisiae.