Fraction collection and how to assay fractions

A fraction is a tube of mobile phase that left the column during a defined slice of the run. The chromatogram suggests which tubes to look at. The assay decides which tubes to keep. This page is the troubleshooting path when those two disagree. The separation that filled the tubes is described in chromatography methods in life-science labs.

Two ways to cut, and the delay between them

You can advance the collector every fixed volume, or every time the detector says a peak started and stopped. Fixed volume misses nothing that elutes, including activity that has no chromophore, and it spends tubes on baseline. Peak triggering saves tubes and will throw away a colourless active protein if the threshold never fires. For a new method, collect by volume across the window where binding and elution should occur, assay, and only then trust a peak trigger.

The detector is upstream of the drip. The volume of tubing between the flow cell and the tube is a delay. At analytical HPLC flows that delay can move the peak by a large part of a narrow fraction. At preparative flows it is a smaller fraction of a big tube and still enough to put the front of a peak into the previous tube. Measure it. A dyed step or a salt step you can see, collected by volume, shows which tube actually received the change. Write that offset into the collector programme. ISO 8655-1 is relevant when a pipette, not the collector, is what you later use to build the assay plate. The collector's own calibration is a separate record.

Column volume keeps the slice honest. A fraction of 0.05 column volumes is a fine analytical cut. A fraction of half a column volume on a polishing column blends the separation you just paid for. Match tube size to peak width. A peak that is three tubes wide can be assayed tube by tube. A peak that is one-tenth of a tube cannot be rescued by a clever threshold.

What to assay, and in which order

Absorbance is the fastest location tool. At 280 nanometres it sees aromatic protein and also sees imidazole and other additives. At peptide-bond wavelengths it sees almost every peptide and a great deal of buffer noise. Use it to choose which tubes are worth a slower assay. Do not use it as the only reason to pool.

A gel asks whether the band you expect is in the tube and what else is with it. Load equal volumes, or equal protein, and say which you did. A thick band from a large volume can be a dilute contaminant. Activity asks whether the function survived. It is the right decider for an enzyme and a poor decider for a structural protein that has no assay yet. A blot asks about an epitope. Mass spectrometry asks about mass and, with fragments, about identity, which is a claim you report with the care HUPO discusses, against a sequence you can check in UniProt.

Run the assay on the load, the flow-through, a baseline tube, the leading edge, the apex, and the trailing edge. That set distinguishes "never bound", "still on the column", "in the shoulder", and "the assay does not work in this buffer". Spiking a known amount of load into a baseline fraction tests inhibition. If the spike dies, the elution salt or the imidazole is the story, and desalting one tube is the next check. Chromatography delivered the liquid. It did not promise the assay would tolerate the liquid.

Symptoms and the next check

Activity only in the flow-through means the column did not bind the target. Stop polishing the elution tubes. Fix pH, salt, solvent, or the tag, and rerun at scout scale. Activity only on the trailing edge, with the ultraviolet apex earlier, means the detector and the function are not the same molecule, or the delay volume put the protein one tube later than the trace. Shift the assignment by the measured delay and reassay before you change the gradient.

Activity split across two regions can be two forms, an aggregate and a monomer, or a collection glitch that skipped a tube. Look at the gel of both regions before you call them isoforms. A gap in tube volumes, or a tube that overflowed into its neighbour, is a hardware symptom. Fix the rack and the flow before you invent a biological story.

Empty tubes under a beautiful HPLC peak usually mean the collector never saw the trigger, the diverter valve stayed in waste, or the peak was so narrow that the delay put it between cuts. Watch one run with the lid open if that is safe for the solvent. A peak in the waste line is a valve problem. A peak in a tube you did not expect is a delay problem.

Pooling too early is the failure that looks like productivity. You combine ten tubes because the trace was above threshold, and the gel of the pool grows a contaminant that was only in tubes eight and nine. The fix is to keep the individuals until the assay sheet says they agree. A pool cannot be unmixed.

Cold tubes and a warm assay

Fractions warm up, evaporate, and grow if they sit in open racks. Cap them, hold them at the temperature the protein prefers, and assay before you leave them over a weekend. A collector in a hot laboratory will concentrate late tubes by evaporation and make the last fraction look richer than it is. Weigh or read the volume of a suspicious tube. Organic fractions lose volume faster than aqueous ones and may belong in a hood, not on an open bench.

A power cut that stops the pump and the collector at different moments creates a tube of unknown volume. Mark that tube and do not average it into the pool. Restart from a saved sample only if you still have one. The half-collected gradient is not a calibration.

Research-use limits

Assays on fractions support a research keep-or-discard decision. They are not a clinical result. The biological risk of the fraction is the risk of the load, until your institutional rules say it has changed. Chemical risk follows the mobile phase. Do not dry a solvent fraction on the bench because the assay volume was small.

Enquiry

State flow, fraction volume, detector-to-tube delay if you know it, tube material, and whether collection is by volume or by peak. Name the assay you will run so the fraction buffer can be discussed against it. Hardware is in the scientific instruments catalogue. Send the requirement with the quote request. If the fractions are a protein purification record, the custom protein expression and purification reference is a place to frame the method question as a discussion. ---