The best magazine
Conceptual Aspects of FNDC5/Irisin and the Controversy Surrounding Its Secretion and Circulating Levels
It is noteworthy that since the discovery of FNDC5/irisin by Boström and colleagues in January 2012, the potential roles attributed to this myokine have still not been consolidated, despite the numerous reports that have been published. The reason for this may lie in certain aspects of the irisin protein that have not been fully clarified, either in the original manuscript or since. In the initial report, FNDC5 was considered to be synthesized as a Type I membrane protein that is cleaved proteolytically, releasing the 12 kDa amino terminal portion into the extracellular space. However, to demonstrate secretion of this myokine, an Abcam antibody directed against the endogenous portion of the protein was used (Abcam 149–178C-terminal, Abcam, Cambridge, UK) (Fig. 1a). Using this antibody (which is no longer available from Abcam), substantial amounts of FNDC5 at a molecular weight of at least 32 kDa were detected in the cell culture media of HEK 293 cells transfected with the C-terminal flag-tagged FNDC5. Utilizing the same antibody, however, a 22-kDa band was also detected in the plasma of mice that were forced to express FNDC5 in the liver through the adenoviral intravenous delivery of a full-length FNDC5 vector. Although the divergence between the theoretical and practical molecular weight could be explained by glycosylation, we must consider that full-length FNDC5/irisin may be secreted, as recently annotated in the Uniprot protein Knowledgebase (UniProtKB Ref. Q8NAU1 Fibronectin Type III domain-containing Protein 5). Reinforcing this idea, and in accordance with Bostrom's results, we described a predominant band of approximately 25 kDa in a rat model using the aforementioned Abcam antibody against the endogenous form of the protein (anti-FNDC5 Abcam 149–178C-terminal), and an additional antibody directed against the theoretically soluble secreted form of 12 kDa (anti-Irisin Phoenix Pharmaceuticals amino acids 42–112, Phoenix Pharmaceuticals, CA, USA) (Fig. 1c).
(Enlarge Image)
Figure 1.
FNDC5 and irisin structure and secretion. Schematic representation of the FNDC5 protein showing the location of its amino acids, including those epitopes recognized by different commercial antibodies (a); schematic representation of the FNDC5 protein located in the cell membrane, encompassing the secretion of irisin portion and two of the truncated isoforms predicted to be expressed in humans by Raschke and collaborators due to a mutation on the start codon of the protein (b); representative immunodetection by Western blot of FNDC5/irisin in the muscle (soleus and gastrocnemius) and adipose tissue (subcutaneous and visceral) secretomes, using the same antibody used in the original paper by Bostrom and collaborators (Abcam) against the intracellular portion of the protein (upper panel) and another antibody against the extracellular portion (Phoenix pharmaceuticals) to recognize the soluble 12-kDa part of the protein (c). SP, signalling peptide; N, n-terminal end; C, c-terminal end; SAT, subcutaneous adipose tissue; VAT, visceral adipose tissue; 1W/3W, one or 3 weeks of voluntary exercise.
At this point, there is no doubt regarding the need to determine which fraction of FNDC5, if any, is cleaved to produce the soluble portion of irisin and to discover its receptor. These data are crucial because all the published results regarding irisin, particularly regarding its circulating levels, have fallen under a cloud of suspicion. The difficulty of identifying which part of FNDC5 is recognized by the commercially available detection kits, which is not indicated in the literature, only exacerbates this situation. Indeed, great uncertainty exists about the accuracy of commercial antibodies in detecting true irisin; huge inconsistencies in irisin levels are exposed in the bibliography. In this regard, our experience has shown that despite the variations in detected irisin concentrations among different ELISA kit lots, the significant differences between the analysed groups of subjects are consistent across independent studies with different cohorts of patients.
On the other hand, studies that predict irisin secretion levels by analysing FNDC5 mRNA expression should be interpreted very cautiously because expression does not always parallel secretion, particularly when it is regulated proteolytically. In this regard, it is still questionable whether a portion or the whole FNDC5 protein is secreted (Fig. 1b), and whether the extracellular portion of irisin is secreted as a dimer (Fig. 1).
Finally, it has to be taken into account that despite the great homology among species, indicated by sequence analysis, a recent study has suggested that the human gene for FNDC5 carries a mutation in the start codon that results in lower translational efficiency and expression of smaller truncated forms (Fig. 1c).
Source: ...