July 1, 2013 (Vol. 33, No. 13)
John Sterling Editor in Chief Genetic Engineering & Biotechnology News
An increasing number of pharma, biopharma, and academic scientists are utilizing the homogeneous time-resolved fluorescence (HTRF) assay to help them in their drug discovery efforts. This was made obvious at the recent Cisbio meeting, “HTRF in Drug Discovery,” held in Avignon, France. A variety of new drug discovery applications that employ HTRF, along with the benefits and some of the drawbacks associated with the technique, were discussed.
According to Cisbio’s Francois Degorce, conference chairman, Cisbio developed this time-resolved fluorescence resonance energy transfer technology to assess molecular interactions in a mix-and-read, no-wash assay format. A homogeneous, nonradioactive technology, the technique has been applied for over 15 years in drug research and in the investigation of various biological target types, such as G protein-coupled receptors (GPCRs), kinases, and biomarkers.
“HTRF brings fluorescent resonance energy transfer technology (FRET) together with time-resolved measurement (TR). The method is homogeneous because there is no need for separation or liquid removal, and unreacted molecules do not need to be washed away to obtain data, as is the case in a number of other assay procedures,” he explained.
HTRF is based on a pair of specific donor and acceptor dyes which, once coupled to interacting partners (e.g., antibody–antigen, ligand–receptor, etc.), and when brought in close proximity through this interaction, produce a specific fluorescence directly related to the amount of complexes formed. It is time resolved because uninformative and nonuseful signals decay much more rapidly than the signals observed after energy transfer by the acceptor. The result is a selective measurement of the assay specific signal, said Degorce.
Because of the compatibility of HTRF with high-throughput screening (HTS), “we decided to use the technology for the detection of cytokines,” explained Lorena Kallal, Ph.D., manager, biological reagents and assay development, at GlaxoSmithKline in Collegeville, PA. “In addition, you do not have to tag or engineer proteins to detect them if you have antibodies that specifically recognize your protein.”
Dr. Kallal, one of the speakers at the Avignon conference, noted that HTRF reagents tend to be less light sensitive and more stable than some other HTS-compatible detection systems.
“We also found that once reagents were added, we could get the same results if we read the plate a few hours after adding reagent, or overnight,” she continued. “For automated assays, it helps if the incubation times are flexible, and it also gives scientists flexibility in how they want to run the assays.”
Describing her team’s HTRF work with cytokines, she explained that since cytokines are secreted proteins, investigators have the option to test the supernatant only, to avoid interference from the cells in the well.
“We found that there was no difference in the IL-8 and IL-1b assays whether we tested the supernatant only, or whether we ran the assay on plates where the cells remained in the well,” explained Dr. Kallal. “And for one assay, we actually saw better statistics in 1,536-well plates than in 384 well plates.”
She added that although HTRF may not have the same sensitivity as some wash protocols, “if you have enough cytokine present, then there is no issue.”
John Joslin, Ph.D., told the conference attendees about his experience with HTRF and phenotypic screens. Dr. Joslin is a research investigator in the assay development and high-throughput screening group at the genomics institute of the Novartis Research Foundation in San Diego. When running such a screen, his team tries to mimic the disease phenotype in vitro as closely as possible. This often means using primary human cells in their native state.
“HTRF allows us to detect and characterize the endogenous levels of the protein in a high-throughput manner,” he said. “It’s also a technology that can be adapted and customized easily, provided you have good antibody pairs. Although there are other platforms that are more sensitive, these often are lower-throughput or more expensive.”
In addition to classical in vitro biochemical screens, such as peptide phosphorylation assays, Dr. Joslin’s team has run several phenotypic cellular screens with HTRF as the primary readout.
“HTRF allowed us to screen over three million compounds per screen in a short time period. All of the screens yielded promising new leads that are moving toward the clinic,” he pointed out.
He added that the power of the approach comes from the design of the screen. “In one example, we collected samples from over 20 healthy human donors for our primary screen and screened approximately three million compounds using HTRF as the readout. There were other formats we could have chosen that may have allowed us to run faster and cheaper.
“However, our goal was to ensure the primary screen was as close to the disease setting as possible—hence the use of primary human cells. HTRF is one technology that enables a phenotypic screening approach.”
Preclinical Assay Development
Patrick Sarmiere, Ph.D, a cell and molecular technologies scientist at Acorda Therapeutics, said the company uses HTRF methods for in-solution ligand binding assays, cellular kinase assays, and for quantitation of biologics in complex matrices. The HTRF assay is routinely applied for measuring the activity of several intracellular kinases in multiple cell types. “Results from these assays allow us to better evaluate how candidate molecules will behave in different tissues and cell types,” he explained.
His group specifically chose HTRF for ligand binding assays because “knowing how much material is bound to a plate for a standard ELISA is difficult to ascertain and, as a result, makes estimating true affinities tricky.”
The main benefits of HTRF, according to Dr. Sarmiere, are no numerous wash or binding steps, time saved, the abilities to perform end-point analysis and to follow the kinetics of binding (which is much more difficult with standard plate-bound assays), and the existence of a long-half-life emission profile of the energy-providing crypates attached to the donor molecule, which allows the end user to avoid typical issues of autofluorescence in biological matrices.
He said the main limitations of HTRF can relate to the detection reagents.
“If we do not have a good antibody that specifically detects our target, then we can have problems developing a good assay,” he explained. “However, this is not necessarily unique to HTRF and is a common concern with all assay formats that use antibodies as a means of detection.
“There are other technologies that can achieve greater sensitivity than HTRF, but typically these technologies require more dedicated and expensive instrumentation platforms.”
A presentation by a MedImmune scientists focused on the application of HTRF for the isolation, optimization, and characterization of antibody drugs. Elizabeth England, who is based in MedImmune’s U.K. facility, specifically discussed a case study involving an anti-IL6 antibody.
A number of assays were developed to identify antibodies with different mechanism of action. Antibodies were found that inhibited the binding of IL-6 to IL-6R or that inhibited the recruitment of the accessory protein gp130.
“The real power of the HTRF technology was demonstrated with the use of the epitope competition assay for lead optimization screening. [The assay’s sensitivity made it] possible to discriminate sub-nanomolar affinity antibodies,” said England. “The development of a second-generation epitope competition assay, whereby a higher-affinity antibody is substituted for the original antibody, can be developed quickly. This is the case because there are fewer variables within the HTRF assay that need to be optimized.”
England and several other conference presenters pointed out a limitation of the HTRF technology that is actually common to all sandwich immunoassays (an antibody pair recognizing the same analyte) and is the consequence of a large excess of analyte with respect to the detecting antibodies. This is the so-called “hook effect,” which can be explained as follows.
While the signal of a sandwich assay increases proportionally (and typically in a linear way) to the analyte concentration, a too-high amount will lead to an increasing quantity of analyte bound to only one of the two conjugates, i.e. not generating any signal. As a consequence, the signal will decrease, and the concentration calculated from the signal measured may seem to correspond to a much lower concentration.
“This can mean that determining the affinity of an interaction becomes more complex as it not simply a case of increasing concentrations until saturation is reached,” explained England.
Compatibility of the toolbox reagents (i.e., reagents individually selected by a researcher as opposed to the use of specific kits) also can sometimes be an issue, she said. For example, antihuman Fc detection of a target protein cannot be used when screening human IgGs. A further limitation is the distance required between donor and acceptor to get energy transfer.
“If very large proteins are used, this distance can be such that there is no excitation of the acceptor molecule,” noted England. “There are also potential issues with steric hindrance from anti-tag detection reagents or from inactivating proteins with direct labeling.”
Nevertheless, HTRF is a very useful technology and researchers just need to be aware of these limitations, she said.