Assessing the presence or absence of the C-terminal sequence in insoluble elastin has been problematic. It was originally proposed that this region served a pro-peptide function, similar to the pro-peptides of collagen, and was removed concurrent with assembly (Chipman et al., 1985
; Franzblau et al., 1989
). Other evidence, however, suggested that it is retained in the crosslinked polymer. Several biochemical studies have tried to use the cysteine residues unique to this region as markers during peptide purification (Price et al., 1993
), but contaminants from cysteine-rich microfibrils make this approach difficult. Another complicating factor is suggested by our finding that purification of elastin using hot NaOH modifies or removes the C-terminal sequence. Thus, the nature of the starting material one chooses to study is perhaps more important than previously appreciated. Our findings described in this report suggest that the domain-36 sequence is, in one form or another, present in mature elastin that has not be prepared using NaOH. However, the low levels that we detect using a quantitative immunological approach suggest that the majority of domain-36 sequence is either removed at some stage of elastin assembly or that the antigenic epitope is altered by posttranslational modification. It is also important to note that low levels of CTP were detected in unprocessed ligamentum nuchae, a tissue highly enriched in elastin. This suggests that CTP modification occurs coincident with the incorporation of tropoelastin into the mature polymer. As mentioned above, domain-36 of tropoelastin has properties that are important for the biological activity of the molecule. The results shown here indicate that there is little or no intact, unmodified domain-36 in mature elastin. Thus, the cell and matrix interaction properties found in this region of tropoelastin are lost as elastin matures.
The quantitative recovery of CTP activity when tropoelastin or CT-25 is treated with elastase () confirms that the estimate of CTP levels in insoluble elastin are not artifactually low due to inactivation of the antigenic epitope by elastase digestion. The question then becomes whether something other than removal of the CTP during assembly can explain the low levels. The presence of the CTP epitope in Starcher-purified elastin is particularly informative in this regard. Starcher elastin, which utilizes trypsin treatment in the purification protocol, contains the same amount of CTP epitope as neutral salt purified elastin even though trypsin digestion is capable of destroying the CTP epitope. Resistance to trypsin digestion suggests that one or more of the lysine residues normally recognized by trypsin is modified in the domain-36 sequence once tropoelastin is incorporate into the insoluble polymer. The most likely modification is crosslinking to another lysine residue, which is supported by studies of Mithieux et al (2005)
who demonstrated that lysines in the RKRK (domain-36) sequence of purified tropoelastin are particularly sensitive to lysyl oxidase modification and are capable of forming a bifunctional cross-link with lysine elsewhere in the molecule. A high crosslinking frequency associated with lysines in domain-36 was also demonstrated using bifunctional crosslinkers (Wise et al., 2005
). Lysine alteration through cross-link formation would likely alter the antigenicity of the CTP epitope and covalently lock the CTP into the mature polymer. Hence, epitope modification and low extractability would explain the low levels of CTP reactivity found in the elastase digests. It is also possible that the C-terminal region is removed from some molecules during assembly, while that which remains participates in a crosslink.
Our studies are similar to those of Rosenbloom et al. (1986)
who, using an antibody generated to a similar C-terminal peptide, found reactivity in neutrophil elastase digests of insoluble elastin. Other than demonstrating the presence of the C-terminal epitope amongst the solubilized peptides, the amount of recovered peptide was not quantified. It is interesting to note that these authors identified CTP reactivity in digests of NaOH-treated elastin whereas we did not. Nevertheless, both studies support the presence of C-terminal sequences in insoluble elastin. Also consistent with retention of the CTP in mature elastin are mass spectrometry studies that have identified fragments of the CTP in proteolytic digests. Getie et al. (2005)
, for example, identified a peptide in an elastase digest of skin elastin that contains the residues N-terminal to the first cysteine in domain-36, which is consistent with our finding of an elastase cleavage site at this sequence (). An analysis of a thermolysin digest of the same elastin identified sequence immediately upstream of RKRK, including the two cysteine residues (Schmelzer et al., 2005
). In neither study, however, were the lysine residues in domain-36 identified, which is consistent with their modification to form cross-links.
The presence of a cross-link in domain-36 suggests how this domain may play a critical role in elastin assembly. Expression of tropoelastin deletion constructs in vitro
and as transgenes in mice have identified sequences encoded by exon 30 as being critical for the self-association of tropoelastin molecules (Kozel et al., 2003
). Tropoelastin constructs lacking domain-36, however, were able to associate with existing elastic fibers, but crosslinking was greatly attenuated (Hsiao et al., 1999
; Kozel et al., 2003
; Sato et al., 2007
). These studies support a multistep process for elastin assembly that involves tropoelastin self-association (mediated by domain 30) followed by crosslinking and maturation mediated by domain-36 (Czirok et al., 2006
; Kozel et al., 2004
; Kozel et al., 2006
; Sato et al., 2007
). The data from the current study are consistent with this model and suggest that domain-36 may function to facilitate fiber maturation by forming an initial cross-link that serves to help register the multiple tropoelastin crosslinking sites for efficient oxidation by lysyl oxidase. A crosslinking function would explain why mutations that alter the domain-36 sequence, such as frame shift mutations associated with dominant cutis laxa (Szabo et al., 2006
; Zhang et al., 1999
), have detrimental effects on elastic fiber assembly.