Purifying Human Mediator Complex From FreeStyle 293-F Cells

The Mediator complex is a large transcriptional coactivator that connects regulatory transcription factors with RNA polymerase II (Pol II). Its size, modular organization, and dynamic interactions make it difficult to isolate in a homogeneous form suitable for biochemical or structural analysis. In their 2025 Bio-protocol paper, Hui-Chi Tang, Kuang-Lei Tsai, and Ti-Chun Chao present a practical method for purifying the human CDK8 kinase module–core Mediator complex from FreeStyle 293-F suspension cells.

Study Background and Research Question

Human Mediator contains a core complex, commonly termed cMED, and a dissociable CDK8 kinase module (CKM). The core includes head, middle, and tail regions, whereas CKM comprises CDK8, Cyclin C, MED12, and MED13. According to the reference study, the human complex contains approximately 30 subunits, making simultaneous recombinant expression of every component impractical for many laboratories.

cMED generally supports Pol II recruitment and transcriptional activation, while CKM can repress this activity through MED13-mediated steric effects. At the same time, CDK8 has been associated with positive transcriptional functions in some contexts. Because CKM is also an entry point for developmental and oncogenic signals, obtaining intact CKM-bound cMED is important for studying how Mediator architecture and kinase activity influence transcription.

The central research question was therefore methodological: can an intact, homogeneous CKM–cMED complex be recovered from human cells at a scale appropriate for downstream structural and functional experiments, while minimizing contamination by Pol II and avoiding crosslinking-induced changes to the native complex?

Key Innovation from the Reference Study

The study’s main innovation is the use of FLAG-tagged CDK8 as a selective affinity handle for the endogenous Mediator population. Rather than tagging multiple cMED subunits or attempting to reconstitute all Mediator components individually, the authors expressed CDK8 with a C-terminal FLAG epitope in FreeStyle 293-F cells. CKM-bound cMED could then be captured through the tagged kinase-module subunit.

This design exploits the reported mutual exclusivity of CKM and Pol II binding to cMED. Selecting CDK8 as the bait is intended to enrich CKM-associated cMED while reducing recovery of the Pol II-bound Mediator state. Importantly, the tag was placed on CDK8 rather than on a core Mediator subunit, and the authors report that the modification did not compromise CKM–cMED stability or CDK8 kinase activity.

A second innovation is process scalability. The authors had previously used an adherent HEK293 system, but expansion of adherent cultures limited the amount of starting material. FreeStyle 293-F cells grow in suspension and are more readily expanded for large-scale protein-complex purification. The workflow therefore links a stable engineered cell line with nuclear extraction, anti-FLAG capture, and density-based polishing.

Methods and Experimental Design Insights

The procedure begins with a FreeStyle 293-F expression system carrying a pcDNA3.1 construct encoding CDK8-F. Transfection and selection are used to establish a stable population, after which suspension cultures can be expanded to provide sufficient biomass. This is a strategically different use of an epitope tag from a conventional recombinant protein expression workflow: the target is not an isolated CDK8 protein, but an endogenous multiprotein assembly recruited through tagged CDK8.

Cells are harvested and processed to generate nuclear extracts, reflecting the nuclear localization and transcriptional role of Mediator. The extract is applied to anti-FLAG M2 affinity gel, allowing FLAG-tagged CDK8 and associated Mediator subunits to be retained. Because the bait is part of CKM, the capture step is designed to preserve the CKM–cMED assembly rather than selectively purifying CDK8 alone.

The affinity-purified material is then subjected to glycerol-gradient centrifugation. Gradient fractionation provides a second level of separation based on the sedimentation behavior of macromolecular assemblies. In this context, it is used to improve homogeneity and help distinguish intact Mediator-containing fractions from incompletely assembled or dissociated species.

The protocol deliberately avoids chemical crosslinkers. That choice is significant for structural biology because crosslinking can stabilize transient contacts, alter conformational distributions, or complicate interpretation of biochemical activity. A non-crosslinked preparation may be more representative of the native complex, although it also places greater demands on extraction and purification conditions.

Protocol Parameters

• Cell platform: Use suspension-adapted FreeStyle 293-F cells to support expansion of the stable CDK8-F producer line, as described in the reference protocol.
• Affinity handle: Express CDK8 with a C-terminal FLAG epitope; the study uses the tagged kinase-module subunit rather than tagging cMED subunits individually.
• Extraction material: Prepare nuclear extracts before immunoaffinity capture because Mediator is a nuclear transcriptional complex.
• Capture step: Apply nuclear extract to anti-FLAG M2 affinity gel to enrich CDK8-associated cMED.
• Polishing step: Use a glycerol gradient after affinity purification to improve the homogeneity of the recovered complex.
• Crosslinking: Do not introduce chemical crosslinkers when the goal is to preserve material for native structural or functional analysis; this is a defining feature of the reported workflow.

Core Findings and Why They Matter

The protocol provides a route to intact human CKM–cMED from mammalian cells without requiring expression of all Mediator subunits. The recovered complex is designed to be free of Pol II because the CKM-bound and Pol II-bound states are mutually exclusive at cMED. This distinction matters experimentally: a preparation containing both forms would make it difficult to attribute observed transcriptional or enzymatic behavior to the CKM-associated state.

The authors also report that the C-terminal FLAG modification on CDK8 preserves complex stability and kinase activity. That observation supports the use of a small epitope tag as an affinity handle when native subunit interactions must be maintained. The suspension-cell format further improves the practical feasibility of generating larger quantities of material than an adherent-cell workflow may permit.

Notably, the paper presents a reproducible purification strategy rather than a universal yield or purity benchmark for every laboratory. Its principal contribution is the integration of cell expansion, selective capture, and gradient fractionation into a workflow that can supply material for structural studies, biochemical reconstitution, kinase assays, and investigations of Mediator regulation.

Comparison with Existing Internal Articles

Existing internal discussions of FLAG-based purification, including an overview of epitope-tag performance and a mechanistic discussion of tagged-protein workflows, focus primarily on tag–antibody recognition, elution behavior, solubility, and recombinant protein handling. The Tang et al. protocol addresses a different problem: using a small epitope as a selective handle to isolate a native, multimeric transcriptional complex from nuclear extracts.

The relationship between these topics is useful but limited. Antibody specificity is necessary for capture, yet it does not by itself establish that a large complex remains intact. In the Mediator workflow, the biological placement of the tag, the choice of CDK8 as bait, the absence of crosslinkers, and glycerol-gradient purification are equally important determinants of sample quality.

Limitations and Transferability

The method is specialized for laboratories studying human Mediator and related nuclear complexes. Its success depends on stable CDK8-F expression, preservation of endogenous subunit interactions during nuclear extraction, and sufficient culture scale. Results may differ with other cell backgrounds, growth conditions, expression levels, or epitope positions.

Affinity capture through CDK8 also defines the composition of the sample. The method preferentially enriches CKM-associated Mediator and is not intended to recover every Mediator state equally. Weak or transient interactions may be lost during extraction, washing, or gradient centrifugation, whereas overexpression of the tagged bait could potentially shift complex abundance or assembly equilibria. These issues should be evaluated with controls such as analysis of representative CKM and cMED subunits, assessment of Pol II contamination, and functional testing of kinase activity.

Another consideration is that a gradient improves biochemical homogeneity but does not automatically prove structural uniformity. Fractions should be characterized using orthogonal methods appropriate to the intended application. For cryo-electron microscopy, for example, particle quality and compositional consistency must still be assessed independently. For functional experiments, it is important to distinguish activity intrinsic to the purified complex from effects caused by residual nuclear factors.

Despite these boundaries, the design is transferable in principle to other tagged endogenous or semi-endogenous complexes when three conditions are met: the tagged subunit is an informative entry point, the tag does not disrupt assembly or activity, and the affinity interaction is compatible with native extraction. The paper’s most general lesson is that complex purification should be designed around biological state selection, not only around antibody binding.

Research Support Resources

For related recombinant protein workflows, researchers can use the FLAG tag Peptide (DYKDDDDK) (SKU A6002) as a soluble competitor in appropriate anti-FLAG M1 and M2 affinity resin elution workflows. The product information describes this eight-amino-acid DYKDDDDK peptide as an enterokinase cleavage site peptide that supports gentle elution of compatible FLAG-tagged proteins and recombinant protein detection. It is not a replacement for the tagged CDK8 construct used in the Mediator protocol, and 3X FLAG fusion proteins require a corresponding 3X FLAG peptide for elution.