Discussion
In this study, we isolated a new mua-3 mutation that shows defects in connective tissue-like tissues that lead to death specifically during the L4 molt in a temperature-sensitive manner. The sequence homology and the similar phenotypes we and others found in mua-3 mutants suggest that MUA-3 is a homolog of FBN1 and that the role of fibrillin genes in C. elegans and humans could be conserved. mua-3(uy19) lethality that is specific during the L4 molt (but not in other molts) could suggest that the mechanical strain may be most stressful during the L4 molt.
The temperature dependence shown in two independent alleles of mua-3 is interesting. Despite the wide use of temperature-sensitive (ts) mutations, especially for studying essential genes that cause lethality when they are deleted, the molecular basis of ts phenotypes is largely unknown. It is generally assumed that the lethality results from thermal inactivation of gene products. However, there are also several lines of evidence suggesting that the exacerbated phenotypes at the nonpermissive temperature come from the changes in conditions under which the molecules function that coincide with the temperature shift (Van Voorhies and Ward 1999; Masuda et al. 2013), such as rates of metabolism or development processes. Our result that a null mutation of dbl-1 has different body size depending on temperature also supports this idea. Therefore, higher temperature may exacerbate mua-3 phenotypes by increasing metabolic rate and/or by speeding the developmental processes. One of the current treatments for Marfan syndrome is β-adrenergic blockers that lower heart rate and slow aortic growth. However, this only ameliorates the deleterious symptoms of progressive aortic root enlargement and aortic dissection. If metabolic rate plays any role in regulating disease progression, then a combined treatment with β-adrenergic blockers and a drug that decreases metabolic rate may help further to alleviate Marfan syndrome.
We identified dpy-17 mutations as suppressors of mua-3 from two independent genetic screens, suggesting that DPY-17 and MUA-3 genetically interact. DPY-17 encodes a cuticle collagen. C. elegans genome contains approximately 154 collagen genes, many of which are essential and developmental stage-specific (Johnstone 1994, 2000; Kramer 1997). In Marfan syndrome models, the tunica media of the aorta, which normally contains elastin sheets and collagen, is fragmented, disorganized, and lost (Cui et al. 2014). Also, it has been reported that collagen metabolism is abnormal in Marfan patients and overexpression of a collagen gene results in reduction of flexibility of extracellular matrix (ECM), thus contributing to Marfan pathology (Priest et al. 1973; Tajima 1995; Robinson et al. 2006). This might suggest that excessive collagen could negatively affect the structure of ECM when there is not enough Fibrillin-1, and that removal of DPY-17 when mua-3 function is reduced might prevent the formation of abnormal ECM and rescue the lethality during the L4 molt.
DPY-17 is known to interact with DPY-31 (Novelli et al. 2006). DPY-31 is a homolog of bone morphogenic protein-1 metalloprotease. The human bone morphogenic protein-1 is a tolloid-like gene whose proteolytic activity is required for TGFβ activation (Ge and Greenspan 2006). In humans, increased activity of other metalloproteases such as MMP2 (metalloprotease 2) and MMP9 has clear implications in Marfan pathology; in Marfan syndrome, these endopeptidases cleave ECM components, including microfibrils, elastic fibers, and collagen, resulting in loss of ECM integrity and aggravation of the syndrome (Chung et al. 2007, 2008). Our study suggests that DPY-31 in C. elegans and BMP-1 in humans may be involved in TGFβ dysregulation, a process that has been implicated in Marfan syndrome disease progression (Ge and Greenspan 2006).
Together, our results suggest that Fibrillin-1, TGFβ, and a Tolloid-like protein may act in concert to modulate TGFβ availability and connective tissue-like tissue integrity in C. elegans. These results further suggest the molecular conservation among known genetic causes for Marfan or Marfan-related syndromes between humans and C. elegans, providing a useful genetic model for Marfan syndrome.