September 15, 1999;
Mammalian BMP-1/Tolloid-related metalloproteinases, including novel family member mammalian Tolloid-like 2, have differential enzymatic activities and distributions of expression relevant to patterning and skeletogenesis.
morphogenetic protein 1 (BMP-1) and Drosophila Tolloid
) are prototypes of a family of metalloproteases with important roles in various developmental events. BMP-1 affects morphogenesis, at least partly, via biosynthetic processing of fibrillar collagens, while TLD
patterning by releasing TGFbeta
-like ligands from latent complexes with the secreted protein Short Gastrulation (SOG). Here, in a screen for additional mammalian members of this family of developmental proteases, we identify novel family member mammalian Tolloid
-like 2 (mTLL-2) and compare enzymatic activities and expression domains of all four known mammalian BMP-1/TLD
-like proteases [BMP-1, mammalian Tolloid
(mTLD), mammalian Tolloid
-like 1 (mTLL-1), and mTLL-2]. Despite high sequence similarities, distinct differences are shown in ability to process fibrillar collagen precursors and to cleave Chordin
, the vertebrate orthologue of SOG. As previously demonstrated for BMP-1 and mTLD, mTLL-1 is shown to specifically process procollagen C-propeptides at the physiologically relevant site, while mTLL-2 is shown to lack this activity. BMP-1 and mTLL-1 are shown to cleave Chordin
, at sites similar to procollagen C-propeptide cleavage
sites, and to counteract dorsalizing effects of Chordin
upon overexpression in Xenopus embryos. Proteases mTLD and mTLL-2 do not cleave Chordin
. Differences in enzymatic activities and expression domains of the four proteases suggest BMP-1 as the major Chordin
antagonist in early mammalian embryogenesis and in pre- and postnatal skeletogenesis.
[+] show captions
FIG. 3. Characterization of the sites at which BMP-1 and mTLL-1 cleave mouse Chordin. (A) Electrophoretic patterns are compared for
purified Flag-tagged mouse Chordin incubated for 24 h in the absence of enzyme (lane 1) or in the presence of BMP-1 (lane 2) or mTLL-1
(lane 3). Samples were run under reducing conditions on a 4–15% acrylamide SDS–PAGE gel and visualized by staining with zinc. (B)
Alignment of Chordin cleavage sites with the physiological cleavage sites for removal of the C-propeptides of fibrillar procollagens I and
III (Kessler et al., 1996) and for proteolytic activation of prolysyl oxidase (Panchenko et al., 1996). (C) Positions of cleavage sites are shown
in relation to the four cysteine-rich (CR) domains in a schematic of mouse Chordin. (D) Alignment of human (Pappano et al., 1999), chick
(Streit et al., 1998), zebrafish (Schulte-Merker et al., 1997), and Xenopus (Sasai et al., 1994) Chordin sequences in regions corresponding to
the mouse Chordin sites cleaved by BMP-1 and mTLL-1.
FIG. 4. BMP-1 and mTLL-1 inhibit Chordin-mediated secondary axis formation, but mTLD and mTLL-2 do not. (A) A schematic depicts
the route of injection of RNA into the single blastomere at the ventral equatorial region of four-cell-stage embryos (An, Vg, V, and D denote
animal and vegetal poles and ventral and dorsal aspects of the embryo, respectively). Representative examples are shown below of embryos
injected with mRNA for Chordin alone, or injected with mRNA for Chordin plus mRNA for human BMP-1 (1huBMP-1), human mTLD
(1huTLD), human mTLL-1 (1huTLL-1), or human mTLL-2 (1huTLL-2). Note that some embryos coinjected with mRNAs for Chordin and
mTLL-1 retained secondary axes, although the majority of these were much reduced in size. Representatives of such “weak” residual axes
are denoted by arrows in the 1huTLL-1 panel. (B) Histogram showing the induction of secondary axes by Chordin, Noggin, a dominant
negative BMP type I receptor (DN-BR), or a “cleavage mutant” BMP-7 (Cm-XBMP7) in the presence or absence of human or Xenopus mTLD
or BMP-1 (huTLD, XTLD, huBMP-1, or XBMP-1, respectively), human mTLL-2 (huTLL-2), or human mTLL-1 (huTLL-1). The frequency of
induction of secondary axes was calculated from the total number (“n,” above the bars) of injected embryos. Human and Xenopus BMP-1
strongly inhibit secondary axis induction by Chordin, whereas mRNA encoding the alternatively spliced product huTLD or XTLD does not.
Human mTLL-1 had intermediate effects, in that 37% of embryos retained secondary axes, although the majority of these were greatly
reduced in size, as illustrated in (A). The lack of effect of BMP-1 or mTLL-1 on secondary axes induced by noggin or by BMP pathway
inhibitors DN-BR or Cm-BMP7 shows that these enzymes specifically act through effects on Chordin.