Evelyn McGown, Ph.D.
Molecular Devices Corporation. 7/99
In 1994, Molecular Devices introduced SPECTRAmax, the first microplate reader equipped with a monochromator, rather than bandpass filters, for wavelength selection. It was also the first microplate reader capable of measuring absorbance below 340 nm. These new features transformed the microplate reader from a limited colorimetric filterbased instrument to a true UVVis spectrophotometer.
Despite the huge popularity of microplates, many people experience difficulty when adapting assays to them. One source of confusion is the fact that samples read in a cuvette and in a microplate do not have identical raw absorbance values because of differing optical pathlengths. Also, microplates are susceptible to surface effects including floating particulates, foaming and variable meniscus formation. The susceptibility to particulates has increased in recent years because modern microplate spectrophotometers such as SPECTRAmax are designed to accommodate microplates with smaller wells and thus have smaller beams than older instruments. All of these factors cause microplates to demand more attention to careful technique to get accurate absorbance results. Especially in the UV spectral range, appropriate procedures should be followed to avoid poor reproducibility and unsatisfactory sensitivity. This Application Note provides guidelines for optimizing UV measurements in SPECTRAmax microplate spectrophotometers, particularly measurements of proteins at 280 nm and nucleic acids at 260 nm.
1. SPECTRAmax microplate spectrophotometer (Molecular Devices Corp.)
2. UVtransparent microplates; e.g.
UVPlate (CorningCostar Cat. No. 3635, Tel: 18004921110)
UVStar (Greiner Cat. No. 655801, E & K Scientific Products, Tel: 408378 2013)
UVMax (Polyfiltronics) fro m Whatman LabSales or distributors
Quartz microplate (Hellma Cells, Inc., Tel: 7185449534, 9166)
CHOICE OF MICROPLATES
Quartz microplates are commercially available, but expensive (> $1000). There areat least three UVtransparent plastic microplates available. Figure 1 shows the spectra of the abovementioned microplates between 200 and 400 nm.
Quartz is transparent throughout the wavelength range. The CorningCostar UV and Greiner UV microplates have good transmission down to ~220 nm. The Polyfiltronics UV Plate is usable down to 240 nm, though it has a higher background than the Costar and Greiner plates. Standard polystyrene microplates can be used down to approximately 300 nm.
The microplates described above have excellent uniformity in the spectral regions where they have high transmission. At wavelengths near their cutoff values, they may have undesirable welltowell variability in optical density (OD). This source of error can be minimized by prereading the plate with water in the wells. (Reference solvent, usually water, is important because its refractive index influences the apparent background OD of the microplate, resulting in 0.004 to 0.014 lower OD values than dry plates.) SOFTmax PRO will make the wellby well background subtractions automatically. Table 1 summarizes the useful spectral ranges and recommended preread ranges for the 5 microplates.
MICROPLATES VERSUSCUVETTES: OPTICAL PATHLENGTH IS THE DIFFERENCE
In traditional spectrophotometers, samples are read horizontally and the optical pathlength is fixed by the physical dimensions of the cuvette (Figure 2). In micro-plate spectrophotometers, samples are read vertically, so the optical pathlength in each well varies, depending on the volume of fluid in the well and the degree of curvature of the meniscus. The variable light path has made it difficult for people to compare results obtained in microplate readers and spectrophotometers.
The BeerLambert Law states that absorbance of a solute in solution is determined by its absorptivity, its concentration, and the optical pathlength (Figure 3). For a given solute under defined solution conditions, absorptivity (sometimes called Extinction Coefficient) is a constant. In traditional cuvettes, the pathlength is also a constant (typically 1 cm), so the only variable is concentration. In microplates, both concentration and optical pathlength are variables. In order to relate absorbance values in microplates with those in cuvettes, one must measure the optical pathlengths. (SPECTRAmax microplate spectrophotometers and SOFTmax PRO can do the measurements and calculations automatically.)
FACTORS DETERMINING ABSORBANCE VALUES IN MICROPLATE WELLS
1. Total amount of solute in the optical lightpath. With different optical pathlengths, different concentrations of DNA can have identical absorbance values as long as the products (Concentration x Pathlength) are identical. For example, wells containing DNA at concentrations of 1000, 500 and 333 ng/mL will have the same absorbance values if their respective sample volumes are 100, 200 and 300 L; i.e., the total amount of DNA in each well is 100 ng (Figure 4).
2. Meniscus curvature and symmetry. The optical pathlength through a sample in a microplate well is affected by the curvature of the meniscus (which in turn is dependent on the composition of the solution and surface treatment of the microplate). For example, 100 L of water or buffer in an untreated micro-plate (flat meniscus) has a pathlength of 0.32 cm, versus 0.25 cm for 100 L of 0.1% Tween (curved meniscus) a difference of 23% (Figure 5A). Meniscus formation in DNA and protein solutions can range between the two extremes. Thus variable meniscus formation can be a source of betweenwell or betweenplate error in microplate analyses, unless steps are taken to ensure that meniscus curvature is identical in all wells. Alternatively, the pathlength may be measured in each well with PathCheck and the absorbance values normalized.
For best precision, the meniscus, whether flat or curved, must be symmetrical, and not skewed up one side of the well (Figure 5B). A sample with an asymmetrical meniscus has a variable and unpredictable optical pathlength. Also an irregular meniscus acts like lens and can deflect the light beam away from the detector, resulting in falsely high absorbance readings.
3 Particulates and bubbles. Dust particles are becoming much more problematic as the beam size and the detector size in the typical instrument become smaller. A particle is not noticeable unless it happens to be in the optical path at the moment of measurement, in which case it has a huge effect. Such particles can cause artifactual spikes in the data up to 0.3 OD above the sample absorbance.
4 Dirty microplates. Microplates can appear perfectly clean, yet have high optical density at wavelengths below 400 nm (the lower limit of detection by the human eye). Fingerprints and residual contamination in reused plates cause errors in UV absorbance measurements.
5 Sample turbidity. Colloidal suspensions scatter light and therefore increase apparent optical density. If the particles are much smaller than the wavelength of the light, the scattering is proportional to (1/lambda)4 and is known as Rayleigh scattering1. Thus the optical density of such a suspension is 10 times greater at 250 nm than at 450 nm. If the particles are larger than the wavelength, light scattering is due to ordinary diffuse reflection (Tyndall scattering). Turbid samples analyzed in the typical laboratory likely contain various particle sizes so that their optical densities are a a combination of Rayleigh and Tyndall scattering (Figure 6). It is a common practice to correct A260 measurements for background turbidity by subtracting an absorbance value taken at a longer wavelength (e.g. 320 nm). Although such a subtraction may be better than nothing at all, Figure 6 indicates that it is likely an undercorrection.
6 Adsorption to well surfaces. Adsorption of analyte to the walls of the microplate wells removes the analyte from the interrogating light beam and thus decreases absorbance. Adsorption can be an especially severe problem with low concentrations of analyte.
7 Well geometry. Microplates with halfarea wells have twice the optical path-length per unit volume than standard flatbottom microplates. Samples in roundbottom wells have a slightly longer pathlength than the same volume in flatbottom wells. In quartz plates, the wells are cylindrical and the walls are vertical, whereas in plastic microplates, the walls are tapered slightly outward from bottom to top. Because of differences in well geometry, a given sample may have different absorbance values in different microplates. All of these differences can be eliminated by measuring the pathlengths and normalizing the results with PathCheck.
SUGGESTIONS FOR BEST QUALITY UV ABSORBANCE RESULTS
1. Use clean microplates that are UVtransmissible.
Protect the bottom surface from fingers and hard surfaces.
Do not reuse disposable microplates.
Blow out dry microplate with clean dry air.
2. Avoid dust particles in samples:
Filter buffers (e.g. 5 filter) to remove ambient particulate contamination.
Cover micropl ate if samples are not read immediately.
Run the blanks (buffercontaining wells) in triplicate, at least, and note that their absorbance values fall within the expected range (to avoid accidental use of dirty plates or incorrect plates).
3. Automix briefly before reading to encourage symmetrical menisci.
4. Avoid bubbles. If samples are prone to foaming (e.g. contain detergent, protein, or nucleic acids) use reverse pipetting technique.
5. If samples have low absorbance, maximize optical pathlength by using increased sample volumes (e.g. 250 L to 300 L in 96well microplates).
6. Run samples in duplicate or triplicate.
Absorbance measurements made through microplate wells are subject to path-length variability and are vulnerable to interference from surface effects. The need to avoid bubbles has long been obvious and the effect of meniscus curvature on pathlength is predictable and may be normalized by employing PathCheck. Microplates are essentially open vessels exposed to dust and packing debris even in the cleanest of laboratories. Dust particles have not been a serious problem in the past because microplate readers had large optical beams and therefore were relatively insensitive to occasional particles in the light path. Modern microplate spectrophotometers, however, have optics designed for microplates with smaller wells and are very susceptible to particles in the light path. All of the above factors contribute to the need for careful attention to technique, especially in the UV spectral range, in order to get accurate, reproducible absorbance results in microplates.
1. Jenkins, F.A. And H.W. White. 1976. Fundamentals of Optics. McGrawHill, New York. Pp. 466468.