Animal Science Abstract
Leptin quantification - on the steep part of the learning curve
We tend to follow a common pattern in the processes we use in the pursuit of scientific truth. We hypothesize, test, acquire, assimilate, rationalize, distill and then begin to form objective and subjective views based upon our interrogation of the facts. Later, in the light of further knowledge, we review and hopefully with a totally open mind, reassess our dogma, and adjust our paradigms to be more consistent with the established and newly revealed facts. Perhaps it is not surprising that we find our patterns of discovery are repeated over time. We knew of the existence of a growth factor (sulphation factor, now known as insulin-like growth factor-1,IGF-1), first described in 1957, which was highly mitogenic; but we were unable to isolate, characterize, or assay it for many years. As is the case for many scientific pursuits, it was the revelation of new knowledge which clarified the reasons for the difficulties associated with the characterization and assay of IGF-1. The discovery of specific binding proteins for IGF-1 which interfered with biological, chemical and immunological assays allowed progress to be made in understanding its biology and in the development of reliable assays. Variations of this theme can be traced in the case of growth hormone and others.
It appears that history may again be repeated, as one of the more recently discovered growth factors, leptin, provides yet another example about which we presently know relatively little. However, in the last eight years, a huge number of reports have begun to unravel the leptin story. It is now clear that serum from a range of species contains specific leptin binding proteins (LBP). Furthermore, as outlined below, non-specific interactions of leptin with other serum components is likely to be problematic in the development of reliable assays for leptin as it was for earlier growth factors.
The development of good quality tools is an essential step in the process of discovery. A properly validated and rigorously tested assay is essential to our understanding of the biology of any bio-molecule. Despite this, sometimes the tools which are available early in the evolution of the discovery process are imperfect. They may be imperfect simply because we have not yet acquired the knowledge to fit the overall puzzle, and the picture has not yet become clear. Is this the case for leptin?
It seems that throughout the world, funding bodies are reluctant to fund research for the development of assays, despite their fundamental necessity to further our understanding. Nonetheless, the literature contains at least fifteen reports of development of leptin assays for a range of species, with widely varying degrees of validation. It is particularly pleasing to note that reports of assays for ruminant leptin are amongst the most rigorously validated, although there are also examples of these where validation has been less convincing. This, of course does not imply that those for which validation has not been as carefully reported, are wrong. However, those which have been rigorously validated certainly provide greater confidence in their value.
There are a number of tests we may apply to demonstrate the validity of an assay. Parallelism provides a test of specificity. A sample is diluted such that the analyte concentration is distributed across the predicted assay range. The values obtained should be parallel to the standards, showing that interactions from other serum components are not causing deviation. Recovery of added exogenous analyte is another useful test for detecting the activity of interfering substances in the sample. If available, similar analytes (analogues/agonists/antagonists/species variants) which might cross-react should also be included to test for specificity. Ideally, these tests should be applied to samples taken from (animals in) a range of physiological states - pre-, post-pubertal, male, female, pregnant, non-pregnant, nutritionally stressed, and from animals of closely and distantly related species, to test for species specificity. Of course, testing for assay repeatability, precision, accuracy, and the ability to transport the assay to other laboratories and obtain similar results are also essential elements.
There appears to be a range of treatments used in current leptin assays to improve 'specific' binding of leptin to antibodies and to reduce the non-specific binding of leptin to hydrophobic sites. The use of relatively high concentrations of detergents may reduce non-specific binding. Given the high hydropathy values for leptin, it is not surprising that hydrophobic interactions are a problem. Evidence from a number of laboratories suggests that inclusion of blocking agents in the buffers is of limited value. For example, leptin appears to bind non-specifically to low affinity, hydrophobic sites on serum albumin, making its inclusion in an assay problematic.
Theoretically, the provision of an excess of high affinity anti-leptin antibody in the familiar protocols such as radio-immunoassay (RIA) or enzyme-linked immunosorbant assay (ELISA) will force the equilibrium towards leptin binding to antibody, rather than binding to LBP in the sample, or low affinity hydrophobic sites; the aim being to approximate first order kinetics. In practice, if low affinity binding sites, or LBP are abundant, it may be difficult to provide sufficient antibody to achieve ideal leptin binding partitioning. In addition, using a higher concentration of antiserum will likely provide other serum proteins (in the antiserum) including LBP, increasing both binding to LBP and hydrophobic/non-specific binding. Thus, the use of affinity purified antibodies may provide a great advantage.
A real danger with inclusion of detergent in an assay buffer is that antibody-hormone binding will be disrupted. Thus, for such treatments it is essential to demonstrate similar binding affinity of tracer and/or standards to the antibody in the presence or absence of detergent. Again, it appears that some of the reported assays have been validated to show that antibody affinity in the presence of detergent is not affected or may be enhanced.
The presence of a high affinity LBP, if similar in affinity to antibody binding, will cause more serious problems in an assay protocol. As for low affinity, hydrophobic interactions, the degree to which LBP may interfere will depend upon the relative concentrations and affinities for leptin of the LBP and the particular antibody. In all cases, the higher the proportion of leptin binding to antibody, the closer the assay is to the ideal.
Historically, in the case in which the hormone bound with very high affinity, (IGF-1 in the IGF-1/IGFBP-3/acid labile subunit complex ~ 10 pM, varying across species), it was necessary to develop a gold standard assay which relied upon acid dissociation of the hormone complex and assay of the fractions following HPLC. However, this is an extreme case due to both the high affinity and the high abundance of the trimolecular complex.
It is interesting that recent studies of groups of lean and obese human subjects have reported that there is an apparent negative correlation between serum leptin and the soluble form of the leptin receptor (sLR) which appears to act as a circulating leptin binding protein. Assays for the sLR indicate that as its concentration increases, so does its interference in conventional leptin assays. The implication is that total leptin may be underestimated, especially in states in which sLR is increased, as in human pregnancy.
The apparent increase in serum leptin in mid- to late-term ruminant pregnancy may also indicate that a circulating LBP is involved, but this will require further investigation. However, if this proves to be the case, care will be required in interpreting data derived from leptin assays in pregnancy and in other physiological states in which LBP are increased. This is not only a task for editors to embrace, but rather for all of the scientific community to participate in thus ensuring the rigor of our pursuit of scientific truth.
Leptin biology has many intriguing aspects which are still unclear. Plasma leptin concentrations from apparently similar animals in apparently similar physiological states/levels of nutrition, reported using the different assays appear to be vastly different, and with no system of inter-assay/inter-laboratory quality control in place, we cannot be sure whether these differences actually exist or if they are real, or are related to the assay procedure. Perhaps it is time to establish a reference bank of serum samples for livestock species, to provide all laboratories with some bench marks against which to gauge each of the assays. For a system such as this to be effective, absolute confidentiality between the co-ordinator and the test laboratory must be respected.
In the light of these problems, it is now the policy of the Growth, Development and Meat Science section of Animal Science that studies which report leptin concentrations must have used an assay for which substantial validation data are either included in the manuscript or cited in peer-reviewed publications. Accurate quantification, using valid immunoassays is essential to enhance our understanding of the physiology of leptin in farm animals.
| Search for papers on a specific subject | |
|---|---|
 |
|