Signal Detection

As should be obvious from the above sections, as an assay is miniaturized from a 96-well (100 |L) format to a 1536-well (3-5 |L format), the signal from each individual well will decrease proportionately. In most cases, 20- or 30-fold less signal would doom a 96-well assay to failure, since a suitable signal window could not be achieved [34]. Using this analysis, it would be easy to conclude that it is impossible to perform an assay in a 1536-well format. However, new technologies as well as modifications and improvements to old technologies have overcome this potential problem.

It is impossible to avoid the fact that when the volume of an assay is reduced from 100 |L to 3 |L the resultant signal will be 33-fold less. However, the signal per unit surface area is not necessarily much different between the 96-well plate and the 1536-well plate. For example, assume that in a 96-well plate the total signal is 100. Then, for a standard well, the signal per unit area is 2. The same assay run in a 1536-well plate will generate a total signal of about 3, but the signal per unit area is about 1. Thus, upon miniaturization, there is only about a 50% loss in signal if the signal can be measured as signal per unit area rather than the total signal.

Charge coupled devices, CCD cameras, are devices that are capable of taking images of assay plates based on signal per unit area rather than on total signal. This is possible because the silicon chip within the camera is sectored into an array of about 1000 X 1000 pixels. Light (photons) striking each of the pixels is measured independently of the other pixels. If an entire 1536-well plate is imaged at once, then the signal generated by each well of the plate will be analyzed by approximately 150 pixels. As long as the volume imaged per pixel remains constant, identical signal should be obtained independently of the format. In this case, the theoretical limit for assay miniaturization would be the size of an individual pixel.

One of the hurdles in developing a CCD-based system for HTS was the development of a telecentric lens for the camera. Typical lenses such as those found on 35 mm cameras exhibit spherical aberration. That is, the light entering the lens is bent more with increasing distance from the center of the lens (Fig. 3). The consequence of this for imaging an assay plate would be that the signal from each well would decrease from the center of the plate toward the edges, leading to an unacceptably high plate %CV. This hurdle was overcome with the development of a large telecentric lens. In a telecentric lens, the angle of bending of the light is nearly independent of the distance from the center of the lens. Consequently, a nearly uniform image can be taken of the plate, leading to acceptably small %CVs.

Figure 3 The image on the left was taken with a standard 35 mm lens, while the one on the right was taken with a telecentric lens on the Tundra imaging system. As can be observed, the amount of light captured by the 35 mm lens from each well decreases as the distance from the center of the plate increases. However, with the telecentric lens, the amount of light capture is nearly equal across the entire surface. (Courtesy of Amersham Pharmacia Biotech, Cardiff, Wales.)

Figure 3 The image on the left was taken with a standard 35 mm lens, while the one on the right was taken with a telecentric lens on the Tundra imaging system. As can be observed, the amount of light captured by the 35 mm lens from each well decreases as the distance from the center of the plate increases. However, with the telecentric lens, the amount of light capture is nearly equal across the entire surface. (Courtesy of Amersham Pharmacia Biotech, Cardiff, Wales.)

Another advantage of CCD technology is that it images an entire plate at once. Detectors based on photomultiplier technology are usually ganged together to increase throughput. For 96-well plates, photomultiplier-based detectors usually come in some multiple of 8 or 12. Increasing the number beyond 12 usually is not practical due to size and cost constraints. This means that between 8 and 12 separate ''detections'' need to be performed for each individual 96-well plate or from 128 to 192 detections for a 1536-well plate. Since evaporation is a significant issue for these miniaturized formats, whichever system can image the plate faster has a significant advantage. Head to head, a CCD-based system can image a plate approximately 10 to 100 times faster than a photomultiplier-based system. However, a photomultiplier-based system is approximately 10 to 100 times more sensitive than a CCD-based system. What you gain in speed you lose in sensitivity and vice versa.

The earlier section reviewed the advantages, disadvantages, and pitfalls of miniaturizing assays along with the technologies, i.e., hardware that is currently available to address some of those problems and challenges. The following section will deal with the assay formats that have either already been successfully used in miniaturized format or can be used, at least theoretically, in the miniaturized format.

Was this article helpful?

0 0

Post a comment