Co-immunoprecipitation (Co-IP) is the classic way to ask whether two proteins share a complex in cells. Its strength is also its limitation: it preserves native interactions better than pull-downs with denatured bait, but it cannot by itself distinguish direct binding from indirect binding through a third protein. People who forget this write "Co-IP proves A binds B" and then get corrected by biochemistry for the next decade.
Lysis conditions
The experiment begins with lysis under non-denaturing conditions. RIPA, NP-40, Triton X-100, or CHAPS buffers are chosen by how much of the complex you want to keep versus how much membrane/aggregate junk you want to remove. Too harsh and transient interactions dissolve; too mild and you pull the entire insoluble proteome. Protease inhibitors are mandatory; phosphatase inhibitors matter if you care whether the interaction is phosphorylation-dependent. Keep everything cold, because complexes that are stable at 4 °C may rearrange at room temperature.
Bait preparation: endogenous vs. tagged
Bait preparation has two modes. Endogenous Co-IP uses the cell's own protein and an antibody against it—this is the most physiologically honest version, but signal is weak and antibody quality decides success. Tagged Co-IP uses Flag/HA/Myc/GFP constructs and an anti-tag antibody; it is easier and cleaner, but overexpression can create artificial complexes that never exist at physiological concentration. A credible story usually combines both: show the interaction endogenously, then map domains or mutants with tags.
Pre-clearing and controls
Pre-clearing reduces background. Lysate is incubated with protein A/G beads and irrelevant IgG so that proteins sticking to beads or antibody Fc regions are removed before the real IP. Then the specific antibody is added: matched isotype IgG as negative control, empty-vector/knockout lysate when possible, and input lane loaded from the same lysate. These controls are not optional decoration. An IgG lane that also pulls the "prey" means the interaction is bead/antibody artifact. An input with no target protein means downstream bands are contamination.
Washing and elution
Washing is where specificity is won or lost. Low-stringency washes keep weak interactions but admit background; high-stringency washes clean the blot but may erase the biology. Experienced analysts wash in graded stringency and watch prey disappearance across washes. Elution can be denaturing (SDS sample buffer, heating) for Western detection, or milder (glycine pH shift, competitive peptide) if the complex will be mass-spec'd or functionally assayed. Hard elution kills native conformation; gentle elution risks co-eluting stuck proteins.
Detection and IP-MS
Detection is usually Western blot with antibodies against prey and bait. But Co-IP can feed mass spectrometry (IP-MS), giving the whole neighborhood of a protein rather than one hypothesized partner. IP-MS also exposes the central Co-IP problem: beads bind proteins nonspecifically, tags attract chaperones, and overexpression recruits stress-response factors. Distinguishing real interactors from passengers requires replicate IPs, stringent controls, spectral-count or intensity-ratio logic, and often orthogonal validation by reciprocal IP, GST/pull-down, FRET, SPR, or in vitro reconstitution.
Stoichiometry and dynamics
A subtle issue is complex stoichiometry. Co-IP enriches what the antibody catches; it does not tell you that A and B are 1:1, or that B only joins when C is phosphorylated. Temporal dynamics matter too: many interactions are stimulus-dependent. Serum shock, receptor activation, DNA damage, mitosis—each can switch a complex on or off. A Co-IP done in resting cells may miss the only biologically interesting state.
What Co-IP actually claims
The honest framing is: Co-IP demonstrates that under these lysis and wash conditions, prey co-enriches with bait. That sentence is powerful. It says the proteins inhabit a shared biochemical space in cells. It does not say they touch. Follow-up with domain mutants, truncation mapping, in vitro binding, crosslinking, and biophysical measurement converts "shared complex" into "direct interaction." Co-IP is the front door; it is not the whole house.
References
- Phizicky EM, Fields S. Protein–protein interactions: methods for detection and analysis. Microbiol Rev. 1995;59(1):94–123.
- Golemis EA, ed. Protein–Protein Interactions: A Molecular Cloning Manual. Cold Spring Harbor Laboratory Press; 2002.
- Ohh M, Park CW, Ivan M, et al. The von Hippel–Lindau tumor suppressor protein is required for proper assembly of an extracellular fibronectin matrix. Mol Cell. 1998;1(7):959–968.
- Free RB, Hazelwood LA, Sibley DR. Identifying novel protein–protein interactions using co-immunoprecipitation and mass spectroscopy. Curr Protoc Neurosci. 2009;Chapter 5:Unit 5.28.
- Dunham WH, Mullin M, Gingras AC. Affinity-purification coupled to mass spectrometry: basic principles and strategies. Proteomics. 2012;12(10):1576–1590.
- Mellacheruvu D, Wright Z, Couzens AL, et al. The CRAPome: a contaminant repository for affinity purification–mass spectrometry data. Nat Methods. 2013;10(8):730–736.