FNDC5

Fibronectin type III domain-containing protein 5, the precursor of irisin, is a is_associated_with::protein that is encoded by the FNDC5 is_associated_with::gene.

Fibronectin domain-containing protein 5 is a is_associated_with::membrane protein comprising a short cytoplasmic domain, a transmembrane segment, and an ectodomain consisting of a ~100 kDa fibronectin type III (FNIII) domain. The ectodomain has been proposed to be cleaved to give a soluble is_associated_with::peptide hormone named irisin. Boström et al proposed that irisin is secreted from muscle in response to exercise, and may mediate some beneficial effects of exercise in humans; potential for generating weight loss and blocking diabetes has been suggested. Several other research groups have reproduced these results  while others have vigorously questioned the findings.

Discovery
FNDC5 was initially discovered in 2002 during a genome search for is_associated_with::fibronectin type III domains and also independently in a search for peroxisomal proteins. Irisin was found to be a cleaved version of FNDC5. Boström and coworkers named the cleaved product irisin, after the Greek messenger goddess Iris.

Biosynthesis and secretion
The FNDC5 gene encodes a prohormone, a single-pass type I membrane protein (human, 212 amino acids; mouse and rat, 209 amino acids) that is upregulated by muscular exercise and undergoes post-translational processing to generate irisin. The sequence of the protein includes a signal peptide, a single fibronectin type III (FNIII) domain, and a C-terminal hydrophobic domain that is probably anchored in the cell membrane. The production of irisin is similar to the shedding and release of other hormones and hormone-like polypeptides, such as EGF and TGFalpha, from transmembrane precursors. After the N-terminal signal peptide is removed, the peptide is proteolytically cleaved from the C-terminal moiety, glycosylated, and released as a hormone of 112 amino acids (in human, amino acids 32-143 of the full-length protein; in mouse and rat, amino acids 29-140) that comprises most of the FNIII repeat region. The sequence of irisin, the cleaved and secreted portion of FNDC5, is highly conserved in mammals; the human and murine sequences are identical. However, the start codon of human FNDC5 is mutated to ATA, which causes it to be expressed at only 1% the level of other animals with the normal ATG start

A difference in the nucleotide sequence of human FNDC5 from that of mouse Fndc5 creates a different initiation codon, potentially generating a protein that begins at methionine-76 (Met-76). A protein initiated at Met-76 would be missing the signal peptide and would be trapped in the cytoplasm. Human FNDC5 could be initiated by the mutant ATA start codon, but Raschke et al showed that this reduces the level of expression to 1% of that with the normal ATG start codon.

Function
Exercise causes increased expression in muscle of peroxisome proliferator-activated receptor gamma coactivator 1 alpha (PGC-1alpha, encoded by the gene is_associated_with::PPARGC1A), which is involved in adaptation to exercise. In mice, this causes production of the FNDC5 protein which is cleaved to give a new product irisin. Due to its production through a mechanism initiated by muscular contraction, Pedersen and Febbraio class irisin as a is_associated_with::myokine.

Boström et al. proposed that irisin promotes the conversion of white fat to brown fat in humans which would make it a health promoting hormone. Their proposal was based on evidence that FNDC5 induces is_associated_with::thermogenin expression in fat cells, that overexpression of FNDC5 in the liver of mice prevented diet-induced weight gain, and that FNDC5 is_associated_with::mRNA levels were elevated in a small number of human muscle samples after exercise.

However this proposal has been challenged by several groups. Timmons et al. noted that over 1,000 genes are upregulated by exercise and examined how expression of FNDC5 was affected by exercise in ~200 humans. They found that it was upregulated only in highly active elderly humans, casting doubt on the conclusions of Boström et al. Raschke et al. reported that there is a mutation in the is_associated_with::start codon of the human gene for FNDC5 resulting in an alternative start codon, which reduced the protein level to 1% of that produced with the normal start codon when expressed in an artificial CMV-driven system. It was already known that FNDC5 used an alternative start codon together with many other expressed genes. Raschke et al. concluded that any function of FNDC5 and irisin in mice might be lost in humans. More than 30 reports have found irisin levels in human plasma using different, validated methods including Huh et al. Wrann et al. show that hippocampal expression of Fndc5 in mice is induced by endurance exercise; peripheral delivery of Fndc5 to the liver by adenoviral vectors, which increases circulating irisin levels, activates a neuroprotective gene program in the brain, including expression of brain-derived neurotrophic factor (is_associated_with::BDNF). Endurance exercise, which is known to improve cognitive function, and the important metabolic mediators PGC-1alpha and FNDC5 are therefore linked to expression of BDNF in the brain.

However Timmons noted that his prior screen did not see any increase in FNDC5 gene expression in response to exercise. Erickson noted that the original cell culture discovery used a truncated protein, missing three of the seven beta strands. He also noted that the antibody used in the that study was made against a peptide in the membrane portion of FNDC5, and so should not recognize the irisin portion. Raschke et al. noted that the mutant start codon of human irisin reduced its expression to 1% of that in other animals with the normal start codon; the miniscule amounts of FNDC5 and irisin in humans are too low to have any physiological role. Albrecht et al. examined the ELISA's used to quantify irisin levels in blood in 50-100 followup studies, and concluded that they had strong cross-reaction to non-specific blood proteins. They questioned the validity of all of these published studies, and called irisin "a myth, rather than an exercise-inducible myokine."