S. Hodgkinson et al., PROTEINS IN TISSUE-EXTRACTS WHICH BIND INSULIN-LIKE GROWTH-FACTOR BINDING PROTEIN-3 (IGFBP-3), Journal of Endocrinology, 145(1), 1995, pp. 1-6
Membrane associated IGFBP-3 is now known to play a role in the modulat
ion of IGF at the cellular level, but mechanisms involved in cell memb
rane binding are far from certain. In this study we report the identif
ication and initial structural characterisation of proteins in a range
of sheep and rat tissues which specifically bind recombinant human no
n-glycosylated IGFBP-3. Tissues were homogenised in Tris HCl (0.1 M, p
H 7.4), containing proteolytic enzyme inhibitors and the residues re-e
xtracted in buffer containing SDS and Triton X-100 (both 1% w/v) prier
to analysis. These were subjected to SDS-PAGE, electro-blotted onto n
itrocellulose and subjected to ligand blot analysis (LBA) using radioi
odinated IGFBP-3 as ligand. LBA revealed a major band of binding activ
ity migrating at 60 kDa in extracts of rat muscle while sheep muscle c
ontained forms of 52 and 40 kDa as the principal species and ovine pan
creatic extracts an abundance of the 40 kDa variant alone. Distributio
n of the binding activity appears tissue specific. Apart from skeletal
muscle, pancreas and a small amount of the 52 kDa form in pituitary,
analysis revealed no evidence of IGFBP-3 binding in a range of other s
heep tissues including liver, kidney, spleen, intestine, adrenal, brai
n, mammary, uterus, ovary and plasma. The binding activity is TCA prec
ipitable, trypsin digestible, dose responsive and relatively specific
for IGFBP-3 since the related proteins recombinant human IGFBP-2 and p
urified caprine IGFBP-4 failed to bind. Additionally, IGFBP-3 binding
to these species appeared unaffected by presaturation of IGFBP-3 with
IGF-L nor by their transient acidification which together with molecul
ar weight estimates and their pattern of tissue distribution suggests
they are not related to the acid labile subunit of the ternary 150 kDa
IGFBP complex. IGEBP-3 binding and molecular mobility of the species
were unaffected by sample reduction but binding was suppressed in a co
ncentration dependent fashion by heparin. Finally LBA using I-125-Conc
anavalin A revealed the 40 kDa species to be glycosylated while the 52
kDa form was not. Functions of these novel proteins in targeting or m
odulating IGF activity at the cellular level remain to be clarified.