When using BCA to quantify serum samples, low concentrations are always inaccurate. Are there any optimized conditions?

Low-concentration samples show weak BCA color development, and 37°C for 30 minutes often gives unstable low readings. You can try 37°C for 60 minutes, or 60°C for 30 minutes, but do not directly copy this, as formulations differ among manufacturers. It is recommended to perform a small-scale temperature-time gradient, such as 37°C for 45/60 minutes and 60°C for 20/30 minutes, and compare the absorbance values and replicate well CV at the 0, 25, and 50 ug/mL points. Prolonging incubation will increase the signal, but the background will also rise, so a matrix-matched blank is essential. Choose a condition with the best low-point linearity and acceptable background, then fix it for routine experiments.

Large replicate CV mainly comes from pipetting errors, sample heterogeneity, and adsorption loss. Serum/exosome samples have low protein concentrations, and after dilution, proteins readily adsorb to tube walls, pipette tips, and well walls. It is recommended to use low-binding tubes and low-binding tips, and after adding the sample to the working solution, use a multichannel pipette to add the working solution to reduce time differences. During incubation at 37°C, cover the plate to prevent evaporation, do not leave surrounding wells empty, and add water or blank working solution to reduce edge effects. Before reading, briefly centrifuge or gently tap to remove bubbles. If one replicate is clearly an outlier, it can be recorded and excluded, but at least two replicates should be retained. Routine monitoring of replicate stability with a known low-concentration quality control sample is also recommended.

The default sample volume in the instructions is, for example, 20 uL. It is mentioned that 25 uL can improve sensitivity, but attention should be paid to the working solution ratio. Generally, the working solution volume should be at least 8–10 times that of the sample; do not increase only the sample volume while reducing the working solution. It is best to also uniformly change the standards to 25 uL loading, and the blank should likewise be 25 uL. After the loading volume is increased, the liquid level in the ELISA plate wells is higher, so mixing should be thorough but without generating bubbles. Low-concentration proteins tend to adsorb to the inner wall of the pipette tips; it is recommended to aspirate and dispense slowly, use low-binding tips, and include at least 3 replicate wells per sample. After changing the volume, the standard curve linearity should first be validated, and the formula in the instructions should not be followed directly.

The main differences between standard 96-well plates and UV plates lie in well-bottom flatness, background absorbance, and light transmission consistency. Low-concentration samples produce weak signals, and high plate-bottom background or scratches can significantly amplify replicate-well CV. UV plates typically have low background and transparent, uniform well bottoms, making them more suitable for low-signal readings, but they are more expensive. It is recommended to use plates of the same brand and same lot for both the standard curve and samples; do not mix plates. Before reading, check for bubbles in the wells and wipe the plate bottom with lint-free paper. Edge wells are prone to temperature gradients; it is recommended to place low-concentration standard curve points in the center of the plate and place blanks or water in the edge wells.

For low-concentration samples, the low points of the standard curve should be densified, and the conventional 0–2000 ug/mL range should not be copied directly. It is recommended to set 0, 25, 50, 75, 100, and 200 ug/mL. When the sample concentration is below 50 ug/mL, the linearity of the lower segment of the curve is more important than that of the upper segment. For fitting, weighted linear or quadratic fitting may be used; do not force a full-range linear fit, as drift at the low points will bias all samples. Blank wells should use the same buffer matrix as the samples, because differences in background among different buffers will be amplified by low-concentration signals. If the results are still unsatisfactory after optimization, a micro BCA or fluorescent protein quantification method may be used for parallel verification to see whether the trends are consistent.