gRAD assay development
gRAD has been developed to fill the gap between standard lateral flow tests and ELISA. gRAD is a tool that allows you to develop your own
rapid lateral flow assays with the benefits of speed (sample to result in under 15 minutes), mobility, simplicity and minimal training of
the end user. But unlike standard lateral flow assays, which are not very precise (CV 25% or more), gRAD with its patented technology
regularly gives CV's below 10% and is in this respect comparable with ELISA.
gRAD is a "generic Rapid Assay Device" supplied as platform for developed assays. The assay developer defines a test reagent that is
mixed with the sample before application to the gRAD. Below is an outline description with links to detailed protocols for developing a
specific test.
Using the other reagents and protocols supplied, you will have developed and produced the biotinylated capture reagent and the conjugated gold colloid
detection reagent. The next major step is to determine the optimal amounts of capture and detection reagents to be included in the test reagent
which is mixed with the sample before application to the gRAD. The steps required and the issues involved in developing a gRAD test are similar
to those involved in developing an ELISA. However, an important consideration with gRAD is that the test is mobile and can be performed outside
the laboratory. Therefore, as a developer you need to define the conditions at which the test is to be performed and how these can be simulated
during development.
The first step is to use a standard assay setup to find a relevant analyte concentration for use in optimizing the test (Experiment 1).
In this experiment you cover a broad range of analyte concentrations with the aim of finding one that can be used to further develop
the assay. The selected analyte concentration should give a good strong signal (e.g. response greater than 250 units in the ESE reader)
but not saturate the assay. You may need to run more than one round of tests to further define the current assay range before selecting
the analyte concentration.
Now that you have an initial assay, you can optimize the test reagent components. The first and most important is the amount of capture
reagent to be used (Experiment 2).
A plot of the response versus capture reagent concentrations at three different times allows you to see how robust the assay is. The major
issue with capillary flow-based tests is variation in the flow rate of the sample. This can have various effects on the test depending on
the way it is set up. With the gRAD system, it is important that the test is read after sufficient capture reagent has passed over the test
line to achieve near-saturation. Once all of the binding sites in the test line have been occupied, the test is independent of differences
in the flow characteristics of the sample. By measuring the response at different time intervals it is possible to find out when this occurs.
The resultant graph should resemble that shown in Fig. 1. If it does not, the capture reagent concentrations need adjustment. If there is
no difference between the responses at 7 and 10 minutes over the whole concentration range, use a lower capture reagent concentration. If there is no
convergence of the responses at 10 and 13 minutes, increase the capture reagent concentration.
Figure 1: Plot of test response versus capture reagent concentrations at 7, 10 and 13 minutes.
A number of factors need to be considered when determining the concentration of the capture reagent to be used. In general, the
more reagent, the more precise is the test. In a standard test setup, the optimal biotinylated antibody concentration applied to the gRAD is
around 5 µg/ml. As seen in Fig. 1, the sensitivity of the test can be improved up to a point by decreasing the concentration of the
capture reagent, but this will also decrease the precision and robustness of the test. In other cases it may be an advantage to
decrease the sensitivity of the test, and this can be done by increasing the concentration of the capture reagent. The sample
applied to the gRAD will run differently depending on the temperature at which it is run, and there will also be sample-to-sample differences
in viscosity. If the tests are carried out in the same laboratory and under controlled conditions, with samples of similar viscosity, then
a lower of capture reagent can be used without a significant decrease in precision. However, if you wish to carry out field studies or the
test is being set up for another user, you will probably want to use a higher concentration of capture reagent. Once the other parameters
are defined, it is a good idea to simulate the variations in the conditions under which the test will be run. In Fig. 1, a concentration of
5 µg/ml read at 10 minutes would work fine in a laboratory and similar samples at warmer temperatures. However, if the samples are likely
to be more viscous or the test is to be run at lower temperatures, then a capture reagent concentration of 10 µg/ml or more may be
required. If the temperatures at which the tests are run are extreme, one option is to read the tests at two different times depending on
the temperature. The amount of capture reagent used may have to be adjusted once the test is fully developed and robustness studies have been
performed.
It is now the turn of the detection reagent (Experiment3). In general, an OD of 0.4-0.6 works best; much below this you lose sensitivity, and much
above this the background becomes a potentially interfering factor. It is also best to add as much detection reagent as possible to reduce the
possibility of the detection reagent becoming saturated at higher analyte concentrations, with a resulting "J curve" in which the response falls
with increaing analyte concentration.
Once the amounts of capture and detection reagents are optimized, means of preserving them have to be defined. A number of options are
available depending on whether the test is to be transported or used in-house. The simplest way is to just freeze ready-to-use aliquots
of the test reagent, which can be thawed prior to adding the sample. This would be fine for in-house testing or for one- or two-day trips in the
field. However, if the test is to be shipped or supplied to someone else, it is probably best to dry the test reagent. This can be done by
freeze-drying or drying in an oven. The test reagent should be made up in the preserving buffer supplied with the gold conjugation kit and
for oven drying, the volume should be as small as possible, ideally 20 µl. The tubes containing the test reagent are dried at 37ºC,
usualy for about 1 or 2 hours. The lids of the vials must be sealed while the reagent is still warm to reduce the amount of moisture
present. The dryer the reagent is, the better its stability. It is also possible to freeze-dry reagent for a number of tests in a single
vial; the reconstituted reagent is generally stable for one day in the field.
The stability of the gRAD has been defined, however you will have to define the stability of your test reagent. Accelerated studies are
not possible to the frozen reagent however are for the dried reagent. The general rule for accelerated stability studies is the rate of
decay is doubled for every increase of 10ºC. So if the test reagent will be stored at 4ºC then storage at 37ºC will accelerate decay by
a factor of 10 fold and if the reagent is to be stored at room temperature then storage at 50 ºC is used for a 10 fold increase i decay
rate. Accelerated stability studies are however just an estimate. In most cases the test is much more stable than found with the accelerated
. For example storage at 50ºC may cause some melting of the preserving sugars collapsing the pellet and changing the environment for the
proteins. It is generally a good idea to run forced stability studies at both 37ºC and 50ºC this way you can see if it can shipped and
stored at room temperature or it can be shipped at room temperature with extended storage 4ºC.
Now that the test reagent has been defined, it is the turn of the measuring range. This normally extends over a 50-fold concentration range.
From the first set of experiments you will have an idea of the range, but now it can be defined for the developed test reagent. The resultant
curve should be simular that shown in Fig. 2. The upper and lower ends will be defined by the precision required. It is not necessary at this
point to determine the full calibration curve in triplicate, but triplicate assays should be performed at three or four concentration points
to check the precision of the test.
(a)
(b)
Figure 2: gRAD assay calibrations curves for (a) anti-tetanus toxoid antibody using tetanus toxoid as the
capture reagent, and (b) mouse IgG using two goat anti-mouse-Ig polyclonal antibodies. The error bars represent 1 SD.
Now that the assay range has been defined, you can calculate how much the sample needs to be diluted before mixing it with the test
reagent. Remember that it is possible to adjust the amount of capture reagent to fine-tune the assay range. In general, the sample can
be diluted in running buffer or a PBS to which BSA and/or a surfactant such as Tween 20 can be added. In most cases you will already
know the relevant sample concentration range, but if not, it would be a good idea to run a number of samples with the test "as is"
to establish the normal range. You may first need to make a serial dilution of a pooled normal sample to define the required dilution.
It is a good idea to have pre-aliquoted dilution buffer, and if possible arrange the protocol so that the same pipette can be used for all
three steps, e.g. 100 µl of sample are mixed into the dilution or running buffer, 100 µl of the resulting dilution are
transferred to the test reagent and mixed and using the same pipette, whereupon all of the resultant mixture is applied to the gRAD.
CongRADulations! You now have a fully developed gRAD test. You now have to see if it is sufficiently robust for your needs. If the test is to be
used in the laboratory as a fast routine test, there is probably no need to test how it runs at 10ºC or 37ºC, but if you
intend to use it in the field, it is a good idea to investigate how it stands up to field conditions. As stated earlier, temperature is the
major variability-causing factor in this type of test, chiefly because it has a marked effect on the viscosity of the sample. Humidity can also
have an effect; for example, if the environment is too hot and dry, the running buffer can evaporate to such an extent as to invalidate the test.
Fig. 3 shows how the response usually changes with run time. Once the sample mixture reaches the test line, the response rises dramatically.
As the binding sites in the test line are filled the curve levels off. However, after some time the signal starts to rise again because of the
evaporation of the sample mixture. This causes the flow to stop, and after some time to reverse. This can be postponed by increasing the amount
of sample applied. There is, however, a relatively long plateau interval during which the test can be read without a significant change in the
response (the "reading window"). Fig. 3 also shows how the temperature can have a marked effect on the flow rate, moving the reading window
backwards and forwards. If your test is sufficiently robust, there is an overlap of the reading windows over the required temperature range.
If there is no overlap of the reading windows for the temperature extremes at which you wish to use the test, you either need to redevelop
the test increasing the capture reagent concentration, or define different reading times for use at the extremes of temperature.
Figure 3: Response changes with run time and the effects of temperature on the resulting response curves.
|
|