The GDF (growth differentiation factor) family is an important branch of the TGF-β superfamily. Its members are widely involved in embryonic development, osteochondral formation, skeletal muscle homeostasis, reproductive regulation, metabolic stress, tissue injury repair, and remodeling of the tumor microenvironment.

Keywords: GDF family; TGF-β superfamily; growth differentiation factor; BMP; SMAD; GDF5; GDF8; GDF9; GDF11; GDF15; ActRIIB; BMPR; cell differentiation; tissue development

1 Positioning of the GDF Family within the TGF-β Superfamily

1.1 Basic Structure of the TGF-β Superfamily

(1) Superfamily composition

The TGF-β superfamily includes multiple classes of secreted cytokines, such as TGF-β, BMP, Activin, Nodal, GDF, and AMH. Members of this superfamily are usually synthesized as precursor proteins, processed by proteases to form mature peptides, and often exert biological activity as dimers.

(2) Signal transduction mode

Most TGF-β superfamily ligands transmit signals through type I and type II serine/threonine kinase receptor complexes. After ligand binding, the type II receptor phosphorylates the type I receptor, which then activates receptor-regulated SMAD proteins and mediates transcriptional regulation in the nucleus through SMAD complexes.

(3) Pathway branches

Canonical TGF-β superfamily signaling can generally be divided into the SMAD2/3 axis and the SMAD1/5/8 axis. TGF-β, Activin, Nodal, and some GDF members tend to activate the SMAD2/3 pathway, whereas BMP and BMP-like GDF members mostly tend to activate the SMAD1/5/8 pathway. Some members can also regulate cell behavior through non-SMAD pathways such as MAPK, PI3K-AKT, and Rho GTPase pathways.

1.2 Branch Attributes of the GDF Family

(1) Naming characteristics

GDF family members are named “growth differentiation factors,” but their functions are not limited to promoting cell growth or differentiation. Some members have BMP-like features, some are closer to Activin/TGF-β-like signaling, and some have special receptor dependence and stress-response properties.

(2) Functional breadth

The GDF family is involved in multiple processes, including embryonic axis establishment, osteoarticular development, neurodevelopment, negative regulation of skeletal muscle growth, follicular development, tissue remodeling, metabolic stress, and inflammatory responses. The tissue expression and receptor compatibility of different GDF members vary, which determines that their functions are strongly dependent on cell type and pathological context.

(3) Research boundaries

The GDF family should not be regarded as a functionally homogeneous cytokine group. GDF5, GDF6, and GDF7 are mostly associated with BMP-like osteochondral and neurodevelopmental signaling; GDF8 and GDF11 are biased toward Activin-like SMAD2/3 signaling; GDF15 mainly reflects GFRAL/RET-dependent stress and metabolic regulatory features.

Table 1 Major GDF Family Members and Functional Positioning

2 Molecular Structure and Maturation Process of the GDF Family

2.1 Precursor Protein Structure

(1) Signal peptide

GDF members are usually synthesized as secreted precursor proteins with an N-terminal signal peptide, which guides the protein into the endoplasmic reticulum and secretory pathway. After removal of the signal peptide, the precursor protein continues to undergo folding, disulfide bond formation, glycosylation, and other processing events.

(2) Propeptide region

The propeptide region participates in mature ligand folding, secretion stability, and activity regulation. In some TGF-β superfamily members, after release of the mature peptide, the propeptide may still remain non-covalently associated with it, forming latent complexes or influencing the diffusion range of the ligand.

(3) Mature peptide region

The C-terminal mature peptide is the core region through which GDF members bind receptors and activate signaling. The mature peptide usually contains conserved cysteine residues, forms characteristic disulfide bond structures, and acts on cell surface receptors in the form of homo- or heterodimers.

2.2 Dimer Formation and Activation

(1) Dimeric forms

Most mature GDF ligands exert their functions as dimers. Dimerization not only enhances receptor-binding ability, but also determines the matching relationship between the ligand and type I/type II receptor combinations.

(2) Proteolytic processing

GDF precursor proteins usually need to be cleaved by furin-like proteases or other proprotein convertases to generate mature active fragments. Cleavage efficiency directly affects the level of mature ligand, secretion efficiency, and biological activity.

(3) Activity regulation

The effective activity of GDF ligands is not determined solely by expression level. It is also affected by precursor processing, dimer formation, binding to antagonistic proteins, retention in the extracellular matrix, and receptor accessibility. Therefore, increased GDF mRNA cannot be directly equated with enhanced pathway activation.

2.3 Extracellular Antagonists and Binding Proteins

(1) BMP antagonists

BMP antagonists such as Noggin, Chordin, and Gremlin can bind some BMP-like GDF members and block their interaction with receptors. In studies involving GDF5, GDF6, and GDF7, antagonist levels can significantly affect the strength of SMAD1/5/8 signaling.

(2) Activin-like pathway regulatory proteins

Follistatin, FLRG, and related proteins can bind Activin-like ligands and affect the signaling output of GDF8, GDF11, and other members. This regulatory mechanism is important in skeletal muscle homeostasis, tissue repair, and metabolic research.

(3) Local microenvironmental influence

The extracellular matrix, proteases, inflammatory factors, mechanical stress, and local pH can all alter the availability of GDF ligands. In tissue samples, GDF protein levels may not correspond directly to actual receptor signaling, and should be interpreted together with p-SMAD readouts and functional endpoints.

3 Receptor Combinations and Signaling Pathways of the GDF Family

3.1 Canonical Receptor Model

(1) Type II receptors

GDF members can bind different type II receptors, such as BMPR2, ACVR2A, and ACVR2B. Type II receptors are responsible for recruiting and phosphorylating type I receptors, representing a key step in ligand-initiated intracellular signal transduction.

(2) Type I receptors

Type I receptors include ALK1, ALK2, ALK3, ALK4, ALK5, ALK6, and ALK7. Different type I receptors determine downstream SMAD signaling bias. For example, ALK3/ALK6 are mostly associated with BMP-like SMAD1/5/8 signaling, whereas ALK4/ALK5/ALK7 are more closely associated with SMAD2/3 signaling.

(3) Receptor combinations determine function

The same GDF member may produce different effects in different cells because of differences in receptor expression. Whether target cells express the corresponding type I and type II receptors is a key prerequisite for determining whether GDF signaling can occur.

3.2 SMAD1/5/8 Pathway

(1) BMP-like GDF members

GDF5, GDF6, GDF7, and related members often display BMP-like signaling features and commonly activate SMAD1/5/8 through receptor combinations such as BMPR2 and ALK3/ALK6. This pathway is closely associated with osteochondral differentiation, joint formation, tendon and ligament development, and some neurodevelopmental processes.

(2) Transcriptional output

After phosphorylation, SMAD1/5/8 forms a complex with SMAD4 and enters the nucleus to regulate developmental and differentiation-related genes such as ID1, ID2, RUNX2, SOX9, COL2A1, and ACAN. The specific transcriptional output depends on cell type, co-transcription factors, and microenvironmental status.

(3) Detection indicators

When studying BMP-like GDF pathways, p-SMAD1/5/8, ID1/ID2, RUNX2, SOX9, cartilage matrix proteins, and osteogenic differentiation indicators should be examined. Detection of GDF ligand levels alone is insufficient to prove effective activation of this pathway.

3.3 SMAD2/3 Pathway

(1) Activin/TGF-β-like GDF members

GDF8, GDF11, GDF9, and related members are more commonly associated with the SMAD2/3 pathway. GDF8 and GDF11 often activate signaling through ACVR2A/ACVR2B and type I receptors such as ALK4/ALK5. GDF9 shows clear cell-type specificity in the local ovarian microenvironment.

(2) Transcriptional output

SMAD2/3 signaling can affect cell proliferation, differentiation, matrix remodeling, inhibition of muscle mass, and support functions for germ cells. GDF8 can inhibit skeletal muscle growth programs, GDF11 is associated with developmental axial patterning and tissue homeostasis, and GDF9 participates in paracrine regulation between oocytes and granulosa cells.

(3) Detection indicators

Studies of the SMAD2/3 axis should detect p-SMAD2, p-SMAD3, SMAD4 nuclear translocation, target gene expression, and functional endpoints. For GDF8, indicators such as MYOD, MYOG, MuRF1, and Atrogin-1 related to myogenesis and muscle atrophy can be included. For GDF9, follicular development and granulosa cell function indicators can be included.

3.4 Non-Canonical Pathways

(1) MAPK pathway

GDF signaling may activate ERK, p38, or JNK pathways in some cells and participate in the regulation of cell migration, stress responses, and differentiation. Non-SMAD signaling often works together with extracellular matrix conditions, mechanical environment, and inflammatory status to influence functional output.

(2) PI3K-AKT pathway

In skeletal muscle, osteochondral tissues, and the tumor microenvironment, GDF-related signals may intersect with the PI3K-AKT axis. When GDF8 inhibits muscle growth, AKT-mTOR, FOXO, the ubiquitin-proteasome system, and autophagy-related pathways often need to be evaluated simultaneously.

(3) Special pathway of GDF15

GDF15 is a special stress factor within the GDF family. Its classical effects do not primarily depend on the common TGF-β superfamily receptor-SMAD axis, but instead act through the GFRAL/RET receptor system to influence appetite, energy metabolism, and stress responses. When analyzing GDF15, the receptor and SMAD patterns of GDF5 or GDF8 should not be mechanically applied.

Table 2 Major Signaling Axes and Detection Indicators of the GDF Family

4 Functional Differentiation of Major GDF Members

4.1 GDF5, GDF6, and GDF7

(1) Osteochondral development

GDF5, GDF6, and GDF7 are often classified as BMP-like GDF branches and are closely related to the development of bone, cartilage, joints, tendons, and ligaments. GDF5 is relatively typical in endochondral ossification, joint cavity formation, and cartilage matrix maintenance.

(2) Tissue specificity

GDF5 is more commonly associated with joint and cartilage structures, GDF6 is related to osteochondral, ocular, and neural development, and GDF7 is more common in tendon, ligament, and nervous system development research. Although all three tend toward BMP-like signaling, they cannot be functionally substituted for one another in interpretation.

(3) Experimental interpretation

When studying these members, GDF ligand expression should be combined with p-SMAD1/5/8, SOX9, ACAN, COL2A1, RUNX2, and matrix deposition indicators. Detecting ligand expression without downstream signaling is insufficient to determine whether it truly drives osteochondral differentiation.

4.2 GDF8 and Skeletal Muscle Regulation

(1) Inhibition of muscle growth

GDF8, also known as myostatin, is a classical negative regulator of skeletal muscle growth. Enhanced GDF8 signaling is usually associated with suppressed myogenesis, decreased protein synthesis, muscle fiber atrophy, and reduced muscle mass.

(2) SMAD2/3 axis

GDF8 often activates the SMAD2/3 pathway through type II receptors such as ACVR2B and type I receptors related to ALK4/ALK5. It suppresses myogenic programs such as MYOD and MYOG and affects myocyte differentiation and muscle fiber size.

(3) Mechanistic extension

The role of GDF8 is not limited to suppression of myogenic factors. It can also intersect with AKT-mTOR, FOXO, the ubiquitin-proteasome system, and autophagy pathways. In muscle atrophy studies, muscle mass, muscle fiber cross-sectional area, p-SMAD2/3, and protein degradation-related indicators should be detected together.

4.3 GDF9 and Follicular Development

(1) Oocyte-derived signal

GDF9 is mainly secreted by oocytes and is an important paracrine factor in oocyte-granulosa cell interactions. It helps support follicular development, granulosa cell proliferation and differentiation, and maintenance of the microenvironment required for oocyte maturation.

(2) Granulosa cell response

GDF9 can regulate the functional state of granulosa cells, enabling them to support oocyte development and follicular structure formation. This process shows synergy or interaction with other oocyte-derived factors such as BMP15.

(3) Research focus

GDF9 research should focus on follicular stage, oocyte quality, granulosa cell markers, hormone synthesis-related genes, and SMAD2/3 signaling. Functional strength and reproductive phenotypes of GDF9 vary among species, and species background should be considered when interpreting results.

4.4 GDF11 and Development/Tissue Homeostasis

(1) Developmental pattern regulation

GDF11 participates in embryonic development, including regulation related to axial patterning, the nervous system, and skeletal development. Its function has some homology with GDF8, but its tissue expression and physiological roles are not identical.

(2) Tissue homeostasis

GDF11 has attracted attention in studies of adult tissue homeostasis, metabolism, and aging. Because experimental systems, detection methods, and sample backgrounds differ greatly, conclusions related to GDF11 require strict distinction between protein detection specificity and functional validation results.

(3) Experimental requirements

GDF11 and GDF8 are structurally similar, and some antibodies or detection systems may show cross-reactivity. In research, detection methods with well-validated specificity should be prioritized, and interpretation should be combined with receptor signaling and functional endpoints.

4.5 GDF15 and Stress Responses

(1) Stress-induced factor

GDF15 is often elevated in inflammation, hypoxia, mitochondrial stress, tissue injury, tumors, and metabolic abnormalities. Its expression is more similar to a systemic stress signal than to a typical tissue development factor.

(2) GFRAL/RET axis

The classical metabolic effects of GDF15 are mainly mediated through the GFRAL/RET receptor system in the hindbrain, affecting appetite, body weight, and energy metabolism. This feature makes GDF15 clearly distinct from most GDF members that act through the TGF-β receptor-SMAD axis.

(3) Disease biomarker property

GDF15 is often used as a biomarker of disease severity, tissue injury, metabolic stress, or tumor-related status. When GDF15 elevation is detected, it should be interpreted together with inflammatory, metabolic, mitochondrial stress, and tissue injury indicators, rather than being simply attributed to a single differentiation pathway.

5 Roles of the GDF Family in Tissue Development and Disease

5.1 Embryonic Development and Organ Formation

(1) Axial pattern establishment

Some GDF members participate in early embryonic axial patterning, left-right axis establishment, and tissue regionalization. GDF1, GDF3, and related members often act together with Nodal-related signaling networks and are important in early development.

(2) Skeletal and joint formation

GDF5, GDF6, and GDF7 participate in the formation of osteochondral tissues, joints, tendons, ligaments, and related connective tissues. Abnormal signaling intensity or spatial distribution may affect skeletal morphology, joint structure, and cartilage matrix stability.

(3) Neural and sensory organ development

GDF6, GDF7, and related members have research value in the development of the nervous system, ocular structures, and sensory organs. These roles are usually associated with BMP-like signaling gradients and tissue boundary establishment.

5.2 Muscle, Osteoarticular, and Metabolic Diseases

(1) Muscle atrophy

Enhanced GDF8 signaling is closely related to muscle growth inhibition and muscle atrophy. In studies of chronic disease, aging, malnutrition, and cancer cachexia, GDF8 is often analyzed together with muscle protein degradation, muscle fiber atrophy, and functional decline.

(2) Osteoarticular diseases

GDF5 is closely related to cartilage matrix homeostasis, joint development, and osteoarticular disease research. In cartilage injury or osteoarthritis models, GDF5 signaling may interact with cartilage repair, matrix synthesis, and the inflammatory microenvironment.

(3) Metabolic stress

GDF15 is frequently detected in obesity, diabetes, mitochondrial diseases, cancer cachexia, and systemic inflammation. Its elevation may reflect either a protective stress response or increased disease burden, and should be interpreted according to the specific model.

5.3 Reproduction, Tumors, and Tissue Repair

(1) Follicular development

GDF9 is important for follicular development and female reproductive function. Abnormal GDF9 may affect oocyte quality, ovulation, and ovarian functional status.

(2) Tumor microenvironment

Multiple GDF members can participate in tumor microenvironment regulation. GDF15 is often associated with tumor stress, immune suppression, metabolic changes, and prognosis. GDF8/GDF11 and related members also have research value in tumor-related wasting and tissue remodeling.

(3) Tissue repair

GDF5, GDF6, GDF7, and related members can be used in research on tissue regeneration, osteochondral repair, and soft tissue repair. Their effects depend on local inflammatory status, stem cell source, mechanical environment, and receptor expression background.

6 Experimental Detection, Intervention Strategies, and Related Reagent Selection

6.1 Ligand Expression Detection

(1) mRNA detection

qPCR can be used to detect the transcriptional levels of GDF family members and is suitable for analyzing tissue expression differences, stimulus responses, and time dynamics. However, increased mRNA does not necessarily indicate increased mature ligand levels, nor does it directly prove that the receptor pathway has been activated.

(2) Protein detection

Western blot, ELISA, immunohistochemistry, and immunofluorescence can be used to detect GDF proteins. Because GDF precursors, propeptides, and mature peptides may coexist, the antibody recognition region directly affects result interpretation.

(3) Distinguishing mature peptide from precursor

When studying GDF activity, precursor protein, mature peptide, and dimeric forms should be distinguished as much as possible. Detecting total protein alone is usually insufficient to determine whether the ligand has receptor-activating capacity.

6.2 Receptor and SMAD Signal Detection

(1) Receptor expression

The response of different cells to GDF varies significantly and depends critically on whether the corresponding type I and type II receptors are expressed. GDF5-related studies should focus on BMPR2, ALK3, and ALK6; GDF8/GDF11 studies should focus on ACVR2A, ACVR2B, ALK4, and ALK5.

(2) p-SMAD detection

p-SMAD1/5/8 or p-SMAD2/3 is an important readout for determining activation of canonical pathways. Combining ligand detection with p-SMAD detection allows more accurate distinction between “expression change” and “signal activation.”

(3) Nuclear translocation and target genes

SMAD nuclear translocation and target gene expression can further confirm pathway output. BMP-like signaling can focus on ID1, ID2, RUNX2, SOX9, and related genes; SMAD2/3 signaling can select corresponding target genes according to the specific cell type.

6.3 Functional Experimental Design

(1) Ligand stimulation experiments

Stimulation of cells with exogenous recombinant GDF proteins can be used to analyze receptor activation, SMAD phosphorylation, and functional output. Dose gradients and time gradients should be included to distinguish early signal activation from later transcriptional and functional changes.

(2) Blocking experiments

Neutralizing antibodies, receptor inhibitors, siRNA, shRNA, or CRISPR interference can be used to verify the causal role of GDF signaling. If both p-SMAD and functional endpoints decrease after blockade, the mechanistic evidence is stronger.

(3) Cell type selection

GDF5-related studies are suitable for chondrocytes, mesenchymal stem cells, osteoblasts, or chondrogenic differentiation models. GDF8 studies are suitable for myoblasts, myotubes, skeletal muscle tissue, or muscle atrophy models. GDF9 studies should use ovarian- or granulosa cell-related models. GDF15 studies should be combined with stress models and GFRAL/RET-related systems.

Table 3 Experimental Design Points in GDF Family Research

6.4 Reagents and Detection Tools Related to GDF Family Cytokine Research

Table 4 Reagents and Detection Tools Related to GDF Family and TGF-β Superfamily Cytokine Research

7 Result Interpretation and Common Misconceptions

7.1 Member Differences Cannot Be Ignored

(1) Belonging to the GDF family does not mean having the same function

GDF5, GDF8, GDF9, and GDF15 differ significantly in receptor systems, tissue sources, and functional directions. Interpreting all GDF members uniformly as “pro-differentiation” or “pro-growth” factors can lead to mechanistic misinterpretation.

(2) Different signaling axes

GDF5-related members are more biased toward BMP-like SMAD1/5/8 signaling, GDF8/GDF11 are more biased toward SMAD2/3 signaling, and GDF15 has special GFRAL/RET dependence. Experimental design must select corresponding detection indicators according to the specific member.

(3) Tissue environment determines function

The same GDF member may play different roles during development, injury, inflammation, and within the tumor microenvironment. Increased expression may be either a functional driving factor or a concomitant response after tissue stress.

7.2 Increased Expression Does Not Equal Signal Activation

(1) Limitations of precursor processing

GDF ligands must undergo processing, dimerization, and secretion before they acquire effective biological activity. Increased mRNA or precursor protein does not necessarily indicate increased mature active ligand.

(2) Receptor limitation

Target cells must express the corresponding receptor combination to respond to GDF signaling. If receptors are absent or insufficiently expressed, downstream SMAD may not be effectively activated even when the ligand is elevated.

(3) Influence of antagonistic proteins

Noggin, Follistatin, Gremlin, and other antagonistic proteins can restrict GDF ligand binding to receptors. In tissue samples, the balance among ligand, receptor, and antagonist should be considered simultaneously.

7.3 Detection Specificity Issues

(1) Cross-recognition between GDF8 and GDF11

GDF8 and GDF11 have high homology, and some antibodies or detection systems may show cross-recognition. When studying these two members, antibody specificity and recognition regions should be confirmed first.

(2) Recognition of precursor and mature peptide

Different antibodies may recognize the propeptide, mature peptide, or full-length precursor. If functional activity is being studied, mature ligand or downstream signaling should be detected as much as possible, rather than relying only on total protein.

(3) Interpretation of ELISA results

ELISA is suitable for batch quantification, but it cannot directly prove that the ligand is in an active state. Functional studies should be validated together with p-SMAD, receptor blockade, ligand stimulation, and cellular functional endpoints.

The GDF family is a group of highly functionally differentiated cytokines within the TGF-β superfamily. Its members cover multiple tissues and disease contexts, ranging from BMP-like osteochondral developmental signals and Activin-like muscle and developmental regulatory signals to GDF15 stress-metabolic signals.

For more related articles, please see below:

[1] Structural Basis, Transduction Branches, and Research Applications of the TGF-β Signaling Pathway