thebench.bio #1 Ecosystem for the Bio-Tech Industry
All Immunology questions

Question 11 of 20 Intermediate Troubleshooting

An ELISA and an intracellular cytokine assay disagree. What would you investigate first?

Short answer

First check that they measure the same thing. ELISA measures cytokine secreted into the supernatant by all cells over the whole culture, while intracellular staining counts cells holding cytokine trapped by a secretion blocker during a short window. Timing, cell source, the blocker and consumption of cytokine often explain the gap before any assay fault does.

Why

The two assays answer different questions. Intracellular cytokine staining uses brefeldin A or monensin to block Golgi transport, so cytokine builds up inside the producing cell. The result is a frequency of producing cells within a few hours.

ELISA sums everything secreted into the medium, minus what cells consumed or proteases degraded, over the full culture. A few strong producers give a high ELISA value but a low frequency. Many weak producers can do the reverse.

Monocytes and other non-T cells add to the supernatant but may be gated out of a T-cell readout. The optimization study by Mandala and colleagues found 5 hours of whole blood stimulation best for T cells and 4 hours for monocytes, and brefeldin A worked better than monensin. Kinetics differ by cytokine.

TNF can peak early and be degraded, while other cytokines accumulate for a day. IL-2 can be consumed by the cells themselves.

How to work it through

  1. Write down what each assay measures: which cells, which time window, and per-cell frequency versus total secreted amount.
  2. Compare stimulation time, cell numbers, culture volume and whether the same cells were used.
  3. Check the intracellular assay's unstimulated and positive controls and the secretion blocker.
  4. Check the ELISA's standard curve, sample dilution, spike recovery and whether the samples sat above the top standard.
  5. Run a time course with matched supernatant and stained cells from the same wells.

An example

A compound raises IFN-gamma in supernatant threefold at 48 hours, but the frequency of IFN-gamma positive CD8 T cells at 6 hours is unchanged. A time course from the same wells shows the frequency rises only after 24 hours, and NK cells, not gated in the T-cell panel, produce half the cytokine. The assays agreed once the panel included NK cells and both were read at 24 hours.

Where the answer stops holding

Even matched, the readouts need not correlate, because secretion per cell varies. Secretion blockers are toxic over long incubations, so intracellular staining cannot cover a 48 hour window. Some cytokines bind soluble receptors or the plate matrix, which lowers the ELISA value without any cellular change.

A strong answer covers

  • Frequency of producing cells versus total secreted amount
  • Different time windows and secretion blockers
  • Cell populations gated in one assay but not the other
  • Consumption and degradation of cytokine in culture
  • Matched time course from the same wells

Common mistakes

  • Assuming one assay is broken because the two numbers disagree
  • Comparing a 6 hour stained sample with a 48 hour supernatant
  • Ignoring non-T cells that contribute cytokine to the supernatant