Editorial: The Fall of Mechanogrowth Factor?

Scientists want to learn about the inner workings of the natural world; as biomedical research scientists, our goals are to understand the mechanisms that regulate normal human physiology, to discern how things may go awry in disease, and to devise effective and specific treatments. When a potential new discovery is made that may facilitate achieving these objectives, it is an admittedly natural tendency to favor the supporting information and dismiss confounding or contradictory data. This psychological phenomenon is termed confirmation bias (1) and is implicit in attempts to prove rather than test hypotheses. The saga of the mechanogrowth factor (MGF) is instructive because it illustrates the dangers of confirmation bias.

Many hormones and growth factors play critical roles in skeletal muscle growth, maintenance, and repair and in mediating responses to mechanical and other stresses (2–4). For example, IGF-1 functions as an endocrine hormone in promoting overall somatic growth, including skeletal muscle (5), and also acts as a paracrine/autocrine factor in muscle tissue, particularly in response to injury and mechanical strain (6, 7). As befits an agent with multiple distinct physiological actions, IGF-1 is under multifactorial regulation. The two IGF-1 precursor proteins [there are three in humans (8)] represent the translation products of distinct mRNAs that arise from alternative RNA splicing (9). Each contains the same 70-residue mature IGF-1 molecule, which is located at the NH₂ terminus and is released by posttranslational proteolytic processing, but differs in the predicted amino acid sequence at the COOH terminus (9). More than 15 years ago, it was noted that soon after muscle injury, the fold increase of one of the minor classes of IGF-1 transcripts was greater than the major mRNA species (absolute transcript levels were not measured) (10, 11). This result and subsequent observations led to formulation of the MGF hypothesis, in which it was postulated that a segment of the COOH-terminal extension peptide encoded by the mRNA that increased most exuberantly and most quickly in injured muscle enhanced repair by promoting proliferation of muscle stem cells (termed satellite cells) (for review, see reference 12). Despite skepticism about the correlative nature of the studies (8), the MGF hypothesis has gained traction in the regenerative medicine field (13, 14).

Against this background, a recent report by Fornaro et al (15) stands as a strongly negative challenge to the MGF concept. The authors, representing muscle research groups from three pharmaceutical companies, used synthetic versions of MGF and compared their actions with effects of equimolar concentrations of 70-amino acid mature IGF-1 in human and mouse muscle cells. In key experiments they showed that added IGF-1 was able to promote proliferation of the C2C12 muscle cell line and human myoblasts and mouse satellite cells in primary culture but that human or mouse MGF was not (15). The latter two cell types also were able to proliferate after exposure to a 110-amino acid human IGF-I precursor protein that contained the 24-residue MGF peptide (15). The authors' interpretation of their results was that IGF-I was biologically active but that MGF on its own, even when modified to enhance its half-life (16), was inert.

What can we learn from these admittedly negative data and from the entire MGF story? First, as scientists we must fight against confirmation bias. In the case of MGF, it appears that confirmation bias was evident in the initial idea that correlation (increase in mRNA levels after muscle injury) was equivalent to causation (biological effects of endogenous MGF peptide on satellite cell proliferation) when this had not been established. Second, independently performed follow-up studies are critically needed to confirm or refute novel observations and the resultant hypotheses, and the scientific literature needs to provide a forum for publication of such studies. And third, as acknowledged by the title of this editorial, the last word about MGF has not been written yet. We should maintain an open mind and a healthy skepticism about the possibility that a peptide exists with the postulated biological properties of MGF and as a scientific community demand experiments that critically and objectively test the MGF hypothesis.