{"paper_id":"e4dd61e1-04b0-434c-8df7-c8a188b48041","body_text":"202 | PHILIPPOU ET AL. | MOL MED 20:202-214, 2014\nINTRODUCTION\nThe insulinlike growth factor-I (IGF-I),\nalso called somatomedin C, is a cellular\nand secreted growth factor which is criti-\ncal for normal body growth, develop-\nment and maintenance, and has impor-\ntant roles in multiple biological systems\n(1–3). A variety of cellular responses are\ninduced by IGF-I, including cell prolifer-\nation, differentiation, migration and sur-\nvival (4–8). These cellular responses have\nimplicated IGF-I in several conditions\nsuch as the pathophysiology of several\ncancers (9–11), or the mitogenic and\nmyogenic processes during muscle de-\nvelopment, regeneration or hypertrophy,\nsince, unlike other growth factors, IGF-I\nacts as both a mitogen and a differentia-\ntion factor (12,13).\nIGF-I is produced by many tissues, in-\ndicating that a significant component of\nIGF-I action is due to its autocrine and\nparacrine mode of function, although it\nalso acts as a classical circulating hor-\nmone. In the endocrine mode of action,\nIGF-I acts as a mediator (somatomedin\nC) of the growth-promoting effects of pi-\ntuitary growth hormone (GH, soma-\ntotropin), which induces the synthesis\nand release of IGF-I by the liver (2,14,15).\nCirculating IGF-I is mainly derived from\nthe liver, but also from skeletal muscle\n(3,16–18), and is mostly bound to high\naffinity IGF-binding proteins, which pro-\ntect it from proteolytic degradation and\nmodulate its bioavailability to the IGF-I\nreceptors (2,19).\nDifferent IGF-I mRNA transcripts are\nproduced as a result of the alternative\nsplicing of the IGF1 gene, encoding for\nseveral IGF-I precursor proteins. These\nIGF-I protein isoforms differ by the\nstructure of their extension peptides, or\nE-peptides, on the carboxy-terminal end\nand by the length of their amino-\n terminal signal peptides. However, they\nshare the same mature peptide, which is\nthe common part of all the IGF-I precur-\nsors (20–23). IGF-I mediates its actions\nthrough the binding and activation of\nseveral receptors and the IGF-I domain\nThe Complexity of the IGF1 Gene Splicing, Posttranslational\nModification and Bioactivity\nAnastassios Philippou,1 Maria Maridaki,2 Spiros Pneumaticos,3 and Michael Koutsilieris1\n1Department of Experimental Physiology, Medical School, 2Department of Sports Medicine and Biology of Physical Activity, Faculty\nof Physical Education and Sport Science, and 3Third Department of Orthopaedic Surgery, Medical School, National and Kapodistrian\nUniversity of Athens, Athens, Greece\nThe insulinlike growth factor-I (IGF-I) is an important factor which regulates a variety of cellular responses in multiple biological\nsystems. The IGF1 gene comprises a highly conserved sequence and contains six exons, which give rise to heterogeneous mRNA\ntranscripts by a combination of multiple transcription initiation sites and alternative splicing. These multiple transcripts code for dif-\nferent precursor IGF-I polypeptides, namely the IGF-IEa, IGF-IEb and IGF-IEc isoforms in humans, which also undergo posttransla-\ntional modifications, such as proteolytic processing and glycosylation. IGF-I actions are mediated through its binding to sever al\ncell-membrane receptors and the IGF-I domain responsible for the receptor binding is the bioactive mature IGF-I peptide, which\nis derived after the posttranslational cleavage of the pro-IGF-I isoforms and the removal of their carboxy-terminal E-peptides (that\nis, the Ea, Eb and Ec). Interestingly, differential biological activities have been reported for the different IGF-I isoforms, or for their\nE-peptides, implying that IGF-I peptides other than the IGF-I ligand also possess bioactivity and, thus, both common and unique\nor complementary pathways exist for the IGF-I isoforms to promote biological effects. The multiple peptides derived from IGF-I and\nthe differential expression of its various transcripts in different conditions and pathologies appear to be compatible with the  dis-\ntinct cellular responses observed to the different IGF-I peptides and with the concept of a complex and possibly isoform-specific\nIGF-I bioactivity. This concept is discussed in the present review, in the context of the broad range of modifications that this growth\nfactor undergoes which might regulate its mechanism(s) of action.\nOnline address: http://www.molmed.org\ndoi: 10.2119/molmed.2014.00011\nAddress correspondence to Anastassios Philippou, Department of Experimental Physiol-\nogy, Medical School, National and Kapodistrian University of Athens, 75 Micras Asias,\nGoudi-Athens, 115 27, Greece. Phone: + 30210 7462690; Fax: + 30210 7462690; E-mail: \ntfilipou@med.uoa.gr\nSubmitted January 18, 2014; Accepted for publication March 11, 2014; Epub\n(www.molmed.org) ahead of print March 12, 2014.\n\nREVIEW ARTICLE\nMOL MED 20:202-214, 2014 | PHILIPPOU ET AL. | 203\nwhich is responsible for the receptor\nbinding is the biologically active mature\npeptide. It is derived after the posttrans-\nlational cleavage of the pro-IGF-I\n(iso)forms and the removal of the \nE-peptides (23–26). Interestingly, it has\nbeen proposed that the E-peptides also\npossess bioactivity that is distinct from\nthat of mature IGF-I (20,27).\nThus, during the last decade, many in\nvitro and in vivo studies have investigated\nthe aspect of the differential IGF-I iso-\nforms or their E-peptides actions in vari-\nous conditions and pathologies (28–38).\nThis concept was further supported by\nrecent findings which revealed differen-\ntial, E-peptide– or IGF-I isoform–specific\nsignaling (31,33,34,38–40).\nIn the present review, focus has been\non the propounded concept of the differ-\nential roles and bioactivity of the IGF-I\nisoforms or peptides, in the context of\nthe complexity that characterizes the al-\nternative splicing, posttranscriptional\nregulation and posttranslational modifi-\ncations of this growth factor and which\nmight modulate its mechanism(s) of \naction.\nHUMAN \nIGF1 GENE STRUCTURE AND\nALTERNATIVE SPLICING\nThe IGF1 gene spans a region of over\n80 kb of genomic DNA located on the\nlong arm of chromosome 12 in humans,\nit is a highly conserved sequence in\nmammals and primates (23), and con-\ntains six exons, which give rise to hetero-\ngeneous mRNA transcripts by a combi-\nnation of multiple transcription initiation\nsites (that is, alternative leader se-\nquences), alternative splicing and differ-\nent polyadenylation signals (41,42).\nThese multiple IGF-I transcripts code dif-\nferent precursor polypeptides, which\nalso undergo posttranslational modifica-\ntions (26,43,44), (Figures 1A–C).\nMore specifically, the different leader\nsequences result in two different classes\nof IGF-I mRNA variants: class 1 tran-\nscripts have their initiation sites on exon\n1 (promoter 1), whereas class 2 tran-\nscripts use exon 2 as leader exon (pro-\nmoter 2), and class 1 (exon 1 to exon 3)\nor class 2 (exon 2 to the exon 3) mRNA\ntranscripts are produced by differential\nsplicing of exons 1 and 2 to the common\nexon 3. Alternative splicing of exon 5\nalso results in different mRNA variants\ncontaining exon 5, generally defined as\nclass B (IGF-IEb), or containing exon 6\n(and excluding exon 5) defined as class A\n(IGF-IEa) (26,45,46), (see Figures 1A, B).\nA third variant, the IGF-IEc, which corre-\nsponds to IGF-IEb in rodents, also is gen-\nerated by alternative splicing in the\nhuman IGF1 gene and contains both\nexon 5 and 6 (22), (see Figure 1C). Simi-\nlar to the human IGF1 gene, multiple\nforms of pro-IGF-I mRNA have been de-\nscribed in other species in which IGF-I\ngenomic sequences have been deter-\nmined, such as the designated pro-IGF-I\nEa-1, Ea-2, Ea-3 and Ea-4 in teleosts\n(23,47–50).\nAll possible combinations between\npromoter usage and terminal exon (5 or\n6) can occur in different IGF-I transcripts\n(45,51,52). It has been proposed that the\nuse of promoter 1 could be associated\nwith the synthesis of paracrine IGF-I and\nmay influence interactions with insulin-\nlike growth factor binding proteins (IGF-\nBPs), or promote the formation of the\ntruncated IGF-I peptide (46), (see below:\nIGF-I Processing, Secretion and Glycosy-\nlation). Transcripts initiating at promoter\n1 are widely expressed in many tissues,\nwhereas transcripts initiating at pro-\nmoter 2 are expressed mainly in the liver\n(circulating forms) and kidney (53) and\nare thought to be more GH-dependent\n(23,54–58), or equally GH-responsive\n(59,60). However, the two promoters are\nprobably not mutually exclusive, and\nGH can also stimulate the expression of\ntissue-specific (local) transcripts, al-\nthough the existent evidence is still\nequivocal (48,61–65).\nIGF-IEa transcript derives from the\nsplicing pattern exon 1 or 2–3–4–6 of the\nIGF1 gene, which represents the main\npro-IGF-I mRNA produced in liver (sys-\ntemic IGF-IEa) but also in other tissues\nwith similar exon sequence (22,66), (see\nFigure 1A). IGF-IEb transcript is a splice\nvariant of exon 1 or 2–3–4–5. Its expres-\nsion was firstly detected in the human\nliver (67), while it was also found to be\nexpressed in lung carcinoma cells (20), in\nskeletal muscle (31,62,68) and more re-\ncently in various tissues and cells such as\nprostate, endometrium and lens epithe-\nlial cells (34,35,65,69), (see Figure 1B).\nWhether there is equivalent to human\nIGF-IEb splice pattern in nonhuman pri-\nmates is not known (23,43,70). IGF-IEc\nmRNA transcript is an exon 1 or\n2–3–4–5–6 splice variant (see Figure 1C),\nwhich was initially identified also in\nhuman liver, where, however, it is ex-\npressed approximately at 10% relative to\nthe main IGF-IEa transcript (22). Struc-\nturally, its cDNA differs from the IGF-IEa\nvariant by the presence of the first 49\nbase pairs from exon 5 (52 bp in rodents),\nIt results from a splice acceptor site in\nthe intron preceding exon 6 and, due to a\nreading frame shift, it gives rise to a dif-\nferent carboxy-terminal peptide sequence\nand a premature stop codon in exon 6.\nThis transcript was named mechano-\ngrowth factor (MGF) since it was found\nto be upregulated in response to muscle\nstretch and/or damage (71); for review\nsee (72). However, its expression also has\nbeen identified in various tissues such as\nendometrium (35), normal and cancerous\nprostatic cells (34), as well as in\n osteoblast-like osteosarcoma cells (73).\nThe biological significance of IGF-I\nsplice variants is currently unknown and\nthe physiological and molecular mecha-\nnisms that regulate their expression are\nunclear; however, the presence of distinct\ntranscripts is indicative of diverse re-\nsponses of cells to different stimuli (74)\nand they probably reflect the complexity\nof IGF-I actions mediated via its various\nisoforms (75,76).\nRecent studies in humans have shown\nthat the IGF-I splice variants are differen-\ntially transcribed in response to varying\nconditions and pathologies, such as exer-\ncise-induced muscle damage (31,68), en-\ndometriosis (35), and prostate (34), or\ncervical cancer (69), as well as in some\nhuman cell lines after hormonal treat-\nment (65,73), while their differential tran-\nscription is possibly a function of age\n\n204 | PHILIPPOU ET AL. | MOL MED 20:202-214, 2014\nIGF-I SPLICING, MODIFICATION AND BIOACTIVITY\n(64,72). The differential expression of the\nIGF-I splice variants observed in various\npathologies is of particular interest, as it\ncould indicate distinct regulatory mecha-\nnisms and biological roles of the different\nIGF-I isoforms; however, their particular\nfunctions remain as yet unclear.\nThe different IGF-I mRNA transcripts\nencode the corresponding precursor pro-\nteins IGF-IEa, IGF-IEb and IGF-IEc\n(23,44). The 5′ end, by alternative splic-\ning of exons 1, 2 and 3, encodes for the\nsignal peptide of the IGF-I prohormone.\nFour different transcription start sites are\npresent in exon 1 and their positions rel-\native to the translational initiation codons\n(that is, Met-48 located in exon 1, and\nMet-25 and Met-22 located in exon 3) can\ngive rise to three distinct IGF-I signal\npeptides from class 1 mRNAs (exon 1 to\nexon 3), (23,43). However, translation of\nmRNAs initiated at the four transcription\nstart sites mentioned above is expected\nto produce signal sequences of 48 (tran-\nscription start sites 1, 2 and 3) and 25\n(transcription start site 4) amino acids,\nsince between two translation start sites\ncontained in an mRNA, (for example,\nMet-25 and Met-22, located downstream\nof the transcription start site 4), the pref-\nerence is given to the upstream site\n(42,44,55,66), (see Figures 1A–C). It was\nsuggested that when translation initiates\nat Met-25, the nucleotide sequence corre-\nsponding to the first 21 amino acids (en-\ncoded by exon 1) of the sequence of the\nFigure 1. Gene structure, alternative splicing, amino acid sequence and posttranslational modifications of human IGF-I. Posttranslational\ncleavage of pro-IGF-I polypeptides can occur at a unique pentabasic motif and at sites containing the general PC consensus se-\nquence (panels 1A–C; potential processing sites are indicated by arrows; see text for details). Only the Ea-peptide of human IGF-I con-\ntains an N-linked glycosylation site. Mature IGF-I peptide residues that interact with IGFBPs are shown in orange color and underlined\nwhereas residues that interact with IGF-IR are in purple color and marked with an asterisk (*) (panel 1A).\nContinued on next page\n\nREVIEW ARTICLE\nMOL MED 20:202-214, 2014 | PHILIPPOU ET AL. | 205\n48 amino acids (aa) signal peptide may\nhave a specific function as part of the\npromoter 1, or may play a role in post-\ntranscriptional regulation of IGF-I\nmRNA by GH (23,55,58).\nFrom class 2 (exon 2 to exon 3) tran-\nscripts, three transcription start sites and\ntheir upstream position relative to the\ntranslational initiation codon Met-32 (lo-\ncated in exon 2) give rise to an IGF-I pre-\ncursor polypeptide with 32 aa long signal\nsequence (23,42,44), (see Figures 1A–C).\nThe mature IGF-I peptide is coded by\nexons 3 (25 aa) and 4 (45 aa). The first 16\namino acids of the amino-terminal por-\ntion of the IGF-I E-peptide are coded by\nexon 4. Exons 5 and 6 encode, by alter-\nnative splicing, distinct portions of the\nE-peptide with alternative carboxy-\n terminal sequences that contain also dis-\ntinct termination codons (22,44). Three\ndifferent E-peptides have been identified\nin humans, encoded by three mRNA\nvariants produced by alternative splic-\ning of the 3 ′ end of the pre-IGF-I mRNA.\nExon 4 to exon 6 mRNA splicing en-\ncodes the Ea-peptide, which contains \n35 aa. The first 16 aa, which are common\nin all the E-peptides, are encoded by the\nexon 4 and the remaining 19 are en-\ncoded by exon 6 (44,66). Splice variant of\nexon 4 to exon 5 yields the Eb-peptide\nwhich, apart from the 16 common aa en-\ncoded by the exon 4, contains 61 addi-\ntional aa encoded by exon 5, resulting in\nthe 77 aa long Eb-peptide (44,67). The\nthird mRNA splice variant, which con-\ntains exon 4, only 49 bp from exon 5,\nand then exon 6, produces the Ec-\n peptide with a predicted length of 40 aa,\nthat is, 16 aa from the exon 4, 16 aa from\nthe exon 5 and 8 aa from the exon 6\n(22,44), (see Figures 1A–C). It is noted,\nthat the last 8 aa of Ec-peptide are en-\ncoded by exon 6, however they differ\nfrom the corresponding Ea-peptide se-\nquence, because of a frameshift at the\nsplice point. Ec-peptide is thought to\nFigure 1. Continued.\nContinued on next page\n\n206 | PHILIPPOU ET AL. | MOL MED 20:202-214, 2014\nIGF-I SPLICING, MODIFICATION AND BIOACTIVITY\noccur by use of a cryptic IGF 633 donor\nsplice site, which is located 49 bp down-\nstream from the 5 ′ end of the exon 5.\nWhen this cryptic IGF 633 donor splice\nsite is not used, the alternative splicing\nof exon 4–5 occurs, that is, the Eb pep-\ntide (22,43). The predicted molecular\nmass and the residues of the three differ-\nent isoforms of the human IGF-I precur-\nsor polypeptide as well as of their vari-\nous forms and partial peptides (based on\nthe amino acid sequence derived from\nthe various IGF-I mRNA transcripts) are\nsummarized in Table 1.\nIn general, the complexity introduced\nby the transcriptional and splicing vari-\nants, posttranscriptional regulation and\nposttranslational modifications of the\nIGF1 gene (77), giving rise to various\nIGF-I isoforms, probably indicate their\ndifferent biological roles under various\nconditions or pathologies and following\ndifferent stimuli (78). The development\nof epitope-specific antibodies for distin-\nguishing the different IGF-I E-peptides\n(20,38,79–82) can contribute to a more\ndefinitive analysis of IGF-I isoforms ex-\npression in various tissues and physio-\nlogical or pathophysiological conditions.\nIGF-I PROCESSING, SECRETION AND\nGLYCOSYLATION\nPost translational processing of IGF-I\nprecursor protein may be a regulatory\nmechanism of the IGF-I activity, as indi-\ncated by the unique processing features\nof IGF-I precursor polypeptides that\nhave been described (25,78). Posttransla-\ntional endoproteolysis of those polypep-\ntides produces the signal, the mature and\nthe E-peptides (E domains). It is thought\nthat the signal peptide at the start of a\nprecursor is removed after facilitating the\npassage of the polypeptide into the en-\ndoplasmic reticulum and the secretory\npathway, with possibly no further bio-\nlogical significance (23,83). The se-\nquences of the signal peptides and the \nE-peptides are less strongly conserved\ncompared with mature IGF-I peptide,\nthough to a variable extent (23).\nFigure 1. Continued.\n\nREVIEW ARTICLE\nMOL MED 20:202-214, 2014 | PHILIPPOU ET AL. | 207\nThe mature peptide comprises four\ndomains, that is, the B amino-terminal\ndomain, C and A domain and D carboxy-\nterminal domain, of IGF-I polypeptides\n(25,84). In addition, two other protein\nproducts have been identified in the\nhuman brain; the tripeptide glycyl-\n prolyl-glutamate (GPE) corresponding to\nthe NH\n2-terminal of the B domain of ma-\nture IGF-I and a truncated IGF-I form\n(–3N:IGF-I) that lacks the first three\namino acids of the amino terminal end of\nmature peptide, probably due to alter-\nnate signal peptides or the combined ac-\ntion of some peptidases (78,85,86). Re-\nmoval of the NH\n2-terminal tripeptide\ncould be a mechanism for increasing the\nbiological potency and availability of\nIGF-I, since the truncated –3N:IGF-I has\nless affinity for IGF-binding proteins\nthan mature IGF-I, thus, increasing its\nbioactivity (78), (see Figure 1A) (see\nbelow: IGF-I Receptors and Binding\n Proteins).\nThe mature IGF-I is a 70 aa long\nsingle-chain peptide and a highly con-\nserved sequence among primate species\n(23,24). Cleavage of pro-IGF-I removes\nthe carboxyl-terminal E domain and can\noccur at the highly conserved, unique\npentabasic motif K\n65-X-X-K68-X-X-R71-X-\nX-R74-X-X-R77. More specifically, the\nArg71-Ser72 bond is cleaved followed by\nthe removal of the Arg residue by the ac-\ntion a carboxypeptidase (25), (see Fig-\nures 1A–C). In general, proproteins can\nbe processed at this specific motif, usu-\nally residing at the end of their pro re-\ngions, by proprotein convertases (PCs)\nsuch as furin (25,87,88). Furin belongs to\nthe  subtilisin-related PCs (SPCs), a major\nfamily of endoproteolytic processing en-\nzymes of the secretory pathway in mam-\nmals (89,90). Seven mammalian PCs have\nbeen identified, namely PC1, PC2, furin,\nPC4, PC5, paired basic amino acid cleav-\ning enzyme 4 (PACE4) and PC7, and a\nmethod of prediction of the general PC-\nspecific or furin-specific cleavage sites\nhas been proposed (88). PCs process pre-\ncursors at sites usually containing the\nspecific consensus sequence [R/K]-Xn-\n[R/K], where X indicates any amino acid\nresidue, and n, the number of spacer\namino acid residues, which is 0, 2, 4 or 6\n(91), (see Figures 1A–C). However, furin\nappears to have a more stringent speci-\nficity and preferentially recognizes sites\nthat contain the sequence motif R-X-\n[R/K]-R (87), while R-X-X-R is its mini-\nmal cleavage sequence (88). Thus, apart\nfrom the cleavage site Arg\n71-Ser72 for ma-\nture IGF-I, another furin-mediated pro-\ncessing at Arg\n77 has been observed in\noverexpression studies, which produces\nan extended, 76 aa long mature IGF-I\n(25,78), (see Figures 1A–C).\nPC-mediated processing of pro-IGF-I\nto mature peptide has been shown to\noccur intracellularly (92), as expected for\nintracellular convertases such as furin,\nwhich are located in the secretory path-\nway (92,93). Nevertheless, evidence has\nbeen provided that the E domains are\nnot cleaved intracellularly (94), and the\nsecretion of unprocessed pro-IGF-IEa iso-\nform, both glycosylated and nonglycosy-\nlated, has been reported (25,30,95–97),\nwhile there are potential proprotein con-\nvertases that could process pro-IGF-I ex-\ntracellularly (97,98).\nConversion of pro-IGF-I to mature\npeptide cleaves off the E domains of the\nIGF-I precursors and one peptide is pro-\nduced from the human Ea domain after\nposttranslational processing of pro-IGF-\nIEa isoform (23,80), (see Figure 1A).\nHowever, the Eb domain contains po-\ntential processing sites (see Figure 1B),\nand at least two putative peptide prod-\nucts of the human Eb peptide have been\nidentified: the EB\n1 peptide (residues\n103–124 aa) with a C-terminal amide\nand the EB\n2 peptide amide (residues\n129–142) (20). The Ec domain also con-\ntains potential processing sites, from\nwhich one corresponds to the EB\n1 pep-\ntide cleavage site and cleaves off the last\neight residues of the Ec domain encoded\nby the exon 6 (26,82), (see Figures 1B, C).\nHowever, so far only the proform of\nIGF-IEc isoform and not the Ec peptide\nor other cleavage products have been\ndetected (31,33–35,73,82).\nThe Ea-peptide of human IGF-I con-\ntains an N-linked glycosylation site at\nAsn\n92 based on the consensus sequence\nAsn-X-Ser/Thr, where X represents any\nencoded amino acid except proline\n(99,100), (see Figure 1A). However, it has\nbeen found that the human IGF-IEa pre-\ncursor with a signal sequence of 48 aa\n(Met-48) is not glycosylated, implying\nthat the greater the number of the amino\nacids contained in the signal peptide the\nlesser the extent of a glycosylation process\n(101). The human Eb- and Ec-peptides\nTable 1. Predicted molecular mass and residues of the three different isoforms (i.e., IGF-IEa,\nIGF-IEb, IGF-IEc), as well as of their various forms and partial peptides, of the human IGF-I\nprecursor polypeptide.\nMolecular Number of NCBI reference \nClass IGF-I form mass (kDa) residues (aa) sequence\n1 Signal peptide 5.35 or 2.73 48 or 25 CAA24998/AAA52543\n2 Signal peptide 3.51 32 NP_001104754\n1or 2 Mature IGF-I 7.65 70 CAA01954\n1 Pre-pro-IGF-I Ea 17.03 or 14.41 153 or 130 CAA24998/AAA52543\n2 Pre-pro-IGF-I Ea 15.18 137 NP_001104754\n1or 2 Pro-IGF-I Ea 11.69 105 NP_001104754\n1or 2 Ea-peptide 4.05 35 AAA52543/NP_001104754\n1 Pre-pro-IGF-I Eb 21.84 or 19.22 195 or 172 NP_001104755\n2 Pre-pro-IGF-I Eb 19.99 179\n1or 2 Pro-IGF-I Eb 16.51 147 NP_001104755\n1or 2 Eb-peptide 8.87 77 NP_001104755\n1 Pre-pro-IGF-I Ec 17.76 or 15.14 158 or 135 NP_001104753/EAW97695\n2 Pre-pro-IGF-I Ec 15.91 142\n1or 2 Pro-IGF-I Ec 12.43 110 NP_001104753/EAW97695\n1or 2 Ec-peptide 4.79 40 NP_001104753/EAW97695\n\n208 | PHILIPPOU ET AL. | MOL MED 20:202-214, 2014\nIGF-I SPLICING, MODIFICATION AND BIOACTIVITY\nlack this cotranslational modification due\nto the reading frame shift resulted from\nthe inclusion of the exon 5 in these two\nIGF-I isoforms (see Figures 1B, C).\nN-linked glycosylation involves the trans-\nfer of a lipid-linked tetradecasaccharide\n(GlcNAc\n2-Man9-Glc3) to an asparagine\nside chain (100,102), (see Figure 1A). It oc-\ncurs in endoplasmic reticulum and the\nsubsequent diversification of the conju-\ngates occurs both in the endoplasmic\nreticulum and Golgi apparatus (100,102).\nRecently, it has been found that the gly-\ncosylation status of pro-IGF-IEa does not\naffect its processing (97), while the secre-\ntion of a peptide similar or identical to\nEa from a human B-lymphocyte cell line,\nin a partially glycosylated form has been\nreported (80). Although the deglycosyla-\ntion process is expected to occur intracel-\nlularly, whether the deglycosylation of\npro-IGF-IEa or the Ea-peptide also could\nbe clipped extracellularly is not known.\nConsidering the unique role of glycosy-\nlation in the protein biosynthesis process\n(100), it is possible that the Ea-peptide\nglycosylation might play a role in interac-\ntions with chaperones in the endoplasmic\nreticulum (78), or in regulation of the\nbioavailability of the different species of\nthis IGF-I isoform (that is, pro-IGF-I Ea,\nmature IGF-I or Ea-peptide) (78,97,103).\nThus, the existence of an N-linked glyco-\nsylation site in the Ea-peptide, which is\nabsent in the Ec- and Eb-peptide, might\nreflect a differential and specific biologi-\ncal action of the IGF-IEa isoform medi-\nated by this posttranslational modifica-\ntion of its Ea-peptide (25,43,97). Recently,\nit has been shown that the species of \nIGF-I produced by the IGF-IEa isoform\nhas a differential ability to activate IGF-\nIR. Glycosylated pro-IGF-IEa is less effi-\ncient at receptor activation than pro-IGF-I\nand mature IGF-I (97),  resembling the de-\ncreased receptor- binding affinity of pegy-\nlated IGF-I forms (104,105) and implying\nthat glycosylated pro-IGF-IEa may serve\nas a reservoir for IGF-I that can be stored\nuntil needed (97). The strong conserva-\ntion observed in the sequence of the Ea\ndomain also suggests a specific biological\nfunction for the Ea-peptide (23).\nNevertheless, whether the E-peptides\nare more stable and/or bioactive within\ntheir pro-IGF-I forms, or they are\nprocessed to act directly on their targets\nremains to be elucidated (see further dis-\ncussion in next sections). Moreover, it\nwould be essential to verify whether the\nIGF-I isoforms could be released in the\ncirculation as different proforms or\nE-peptides (3,79), or the final peptide that\nenters the circulation after extracellular\nendoproteolysis of the IGF-I prohormone\nis only the mature peptide (103,106).\nIGF-I RECEPTORS AND BINDING\nPROTEINS\nIGF-I actions are mediated through its\nbinding to several receptors, such as \nIGF-IR (or type I IGF receptor) and \nIGF-IIR (or type II IGF receptor), insulin\nreceptor (IR), and some atypical receptors\nsuch as the hybrid IR/IGF-IR (107–109).\nMore specifically, the mature IGF-I\npeptide, which is responsible for binding\nto the receptors, binds IGF-IR with the\nhighest affinity, IGF-IIR with low affinity\nand is also able to interact with IR. The\nIGF-IR exhibits a high degree of homol-\nogy to IR (110) and, given the significant\nstructural similarity between IGF-I and\ninsulin, these ligands can cross-activate\nboth receptors, while the IGF-IR signal-\ning pathways share multiple intracellular\nmediators with the insulin signaling cas-\ncade (19,52). The IGF-IR/IR hybrid re-\nceptor is thought to function predomi-\nnantly as an IGF-I receptor, since its\nbinding affinity for insulin is lower than\nthat for IGF-I, however the functional\nimportance of IGF-IR/IR hybrid receptor\nremains poorly understood (74,109,111).\nIt is widely recognized that most of the\nobserved IGF-I biological effects on cell\ngrowth, differentiation, invasion and sur-\nvival depend on the binding and activa-\ntion of IGF-IR, which is a ligand-activated\nreceptor tyrosine kinase (4,112). Specifi-\ncally, functional epitope mapping of IGF-I\nhas revealed that the IGF-IR interacts with\nthe residues 21, 23, 24, 44 as well as the ty-\nrosines 31 and 60, which are located in the\nC and A domains (113), (see Figure 1A).\nIGF-IR is a transmembrane protein con-\nsisting of two extracellular α-subunits,\nwhich contain the cysteine-rich ligand\nbinding site, and two transmembrane\nβ-subunits that have a cluster of three ty-\nrosine residues, which undergo phospho-\nrylation and activation upon IGF-I bind-\ning (2,110,114). A structural rearrangement\nin the transmembrane β subunits of the\nreceptor is caused by binding of IGF-I to\nIGF-IR, resulting in transautophosphory-\nlation of the cytoplasmic tyrosine kinase\ndomain of the receptor, as one kinase do-\nmain phosphorylates the other, and thus\ndestabilizing the autoinhibitory conforma-\ntion within the kinase domain (115,116).\nThis conformational change permits unre-\nstricted access to the binding sites for pro-\ntein substrates (117), thus recruiting spe-\ncific cytoplasmic molecules, such as\ninsulin receptor substrate (IRS) proteins,\nand activating specific intracellular path-\nways including Ras/mitogen-activated\nprotein kinase (MAPK)/extracellular\n signal-regulated kinases 1 and 2 (ERK1/2)\nand phosphatidylinositol 3-kinase (PI3K)/\nAkt (118).\nBiological actions of IGF-I are modu-\nlated by a family of at least six IGFBPs\n(52,75,119–21), which interact mainly\nwith the residues 1–3 and 49–51 of ma-\nture IGF-I (113), (see Figure 1A). In gen-\neral, IGFBPs transport IGF-I and in-\ncrease its half-life in the circulation;\nmost of the circulating IGF-I is protected\nfrom proteolytic degradation by form-\ning a ternary complex with IGFBP-3 and\nthe glycoprotein acid-labile subunit\n(ALS) (74,122). IGFBPs also would be\nexpected to modulate and control, both\nin the circulation and in the extracellu-\nlar environment, the extent of IGF-de-\npendent cellular effects via regulation of\nfree IGF-I concentration and its local\nbioavailability in the tissue, since IGF-\nBPs provide tissue specificity for the\nlocal action of IGF-I (74,119,123–126). In\naddition, IGFBPs compete with IGF-IR\nand normally have higher binding affin-\nity to IGF-I than IGF-IR does. Therefore,\nbinding of IGFBPs to IGF-I prevents the\nligand from interacting with the recep-\ntor and, thus, suppresses IGF-I actions\n(74,124,125).\n\nREVIEW ARTICLE\nMOL MED 20:202-214, 2014 | PHILIPPOU ET AL. | 209\nHowever, some IGFBPs can exhibit\nIGF-I potentiating effects and their \nIGF-I-inhibitory or stimulatory activities\nare dictated by factors such as the tissue-\nspecific distribution of particular IGFBPs\nand the ratio between free (active) IGF-I\nand IGFBP-IGF-I bound (4,11). Moreover,\nit has been shown that certain IGFBPs\nhave IGF-independent activities, imply-\ning that they can modulate cell survival\nand apoptosis, or inhibit tumor growth\nin the absence of the ligand (11,127). In\naddition, proteolytic fragments of\nIGFBP-3 were reported to possess mito-\ngenic activity in the peritoneal fluid of\nwomen with endometriosis (128).\nThere is also a group of cysteine-rich\nproteins, known as IGFBP-related pro-\nteins (IGFBP-rPs), that share important\nstructural similarities with the IGFBPs\nbut they have low binding affinity to\nIGFs. It has been proposed that these\nproteins and the IGFBPs constitute an\nIGFBP superfamily (129,130), however\nthe functions of the IGFBP-rPs regarding\nthe IGFs actions are as yet unclear (131).\nIGF-I PEPTIDES ACTIONS AND\nSIGNALING\nAlthough, by the general consensus,\nIGF-I is thought to exert its biological\nactions predominantly through mature\npeptide, differential biological activities\nhave been reported for the different\nIGF-I isoforms (propeptides), or for\ntheir E-peptides, exogenously adminis-\ntrated or overexpressed in various in\nvivo (28,29,36,132–134) and in vitro mod-\nels (20,27,31–33,135,136), implying that\nthere are peptides other than the IGF-I\nligand that also possess bioactivity and,\nthus, both common and unique or com-\nplementary pathways exist for the IGF-I\nisoforms to promote biological effects\n(31,36).\nA differential expression profile of the\nIGF-I isoforms have been shown in vari-\nous conditions or pathologies in humans,\nsuch as skeletal muscle damage (31,68),\nendometriosis (35) or prostate (34), cervi-\ncal (69) and colorectal cancer (137). The\ndiversity and the patterns of differential\nexpression have been proposed to reflect\npotential biological activities associated\nwith the E domain peptides (27,52). Di-\nvergent actions and signaling of the dif-\nferent pro-IGF-I forms or mature IGF-I\nlacking any E-peptide have been re-\nported after viral-mediated expression of\nthe IGF-I isoforms (IGF-IA and IGF-IB) in\nmouse skeletal muscle (36,132). Specifi-\ncally, it has been shown that overexpres-\nsion of mature IGF-I in skeletal muscle\ndid not promote muscle hypertrophy in\nyoung mice, suggesting that the pro-IGF-\nI forms are required for this effect and\nthat E-peptides, either as a part of pro-\nIGF-I or independent of mature IGF-I,\nmay be necessary for IGF-I-mediated\nmuscle hypertrophy (36). Moreover, by\nusing the MKR transgenic mouse model\nwhere functional IGF-IRs are lacking in\nmuscle fibers (as a dominant negative\nIGF-IR is expressed specifically in skele-\ntal muscle), it was shown that, regardless\nof which isoform is overexpressed, IGF-I\nreceptors on muscle fibers are required\nfor IGF-I-mediated hypertrophy (36). In-\nterestingly, however, after viral-mediated\ndelivery of murine IGF-I isoforms into\nskeletal muscle, even though both iso-\nforms caused increased phosphorylation\nof the IGF-IR, increased expression of\nIGF-IB (murine IGF-IEb) drove both\nmain pathways downstream of IGF-IR,\nthat is, the PI3K/Akt pathway and the\nMAPK pathway, whereas IGF-IA\n(murine IGF-IEa) overexpression resulted\nin increased Akt phosphorylation only\n(132). These findings imply distinguish-\ning IGF-I isoform–specific actions, re-\ngardless of any potential receptor(s) acti-\nvated (see also discussion in following\nsections). In addition, constitutively\noverexpression of the IGF-IEa isoform\nspecifically in cardiac muscle was shown\nto protect the heart from oxidative stress\nvia Sirtuin 1 (SirT1)⁄c-Jun N-terminal ki-\nnase 1 (JNK1) activity while, conversely,\nmature IGF-I triggered oxidative stress in\nthe heart and did not affect SirT1 activity\n(138,139). Moreover, it has been shown\nthat although both mature IGF-I and\nIGF-IEa propeptide triggered the phos-\nphorylation of IGF-IR, its downstream\ncanonical PI3K/Akt/mTOR signaling\npathway was not induced in the trans-\ngenic mice overexpressing IGF-IEa\n(138,140). Instead, this specific IGF-I\npropeptide activated alternate signaling\nintermediates 3-phosphoinositide-\n dependent protein kinase-1 (PDK1) and\nserum/glucocorticoid regulated kinase 1\n(SGK1), as well as SirT1. Thus, it was\nsuggested that this downstream of IGF-I\nreceptor(s) signaling activated by IGF-\nIEa employs novel pathways and that\nthe divergent signaling mechanisms be-\ntween the two IGF-I forms (that is, ma-\nture IGF-I and IGF-IEa propeptide) may\naccount for their opposing effects on the\nheart (136,138,140).\nConsidering particularly the bioactiv-\nity of the E-peptides, in vitro studies have\nsuggested that the E-peptides of the\nhuman IGF-I precursors may act as inde-\npendent growth factors, since their syn-\nthetic analogs, generated from unique re-\ngions within the E domains, were\ndemonstrated to possess mitogenic\n(20,27,34,35,73,141), angiogenic (142) and\nmigratory activity (28,141,143), and regu-\nlate cell differentiation (27,28) in various\nhuman cells or cell lines. Antitumor ac-\ntivity of human Eb-peptide also has been\nreported in some cancer cells (144).\nSimilar to the results derived from\nhuman cell lines, studies using animal\ncell lines models have shown that exoge-\nnous administration or overexpression of\nsynthetic peptides, generated from dif-\nferent regions within the E domain of\nhuman (Ec) (28,31,33,145,146), trout (Ea)\n(144,147) or rodent (Ea and Eb) se-\nquences (38) in human, rodent or porcine\ncells in culture showed unique though\ninconsistent effects in promoting cell pro-\nliferation and migration, and in delaying\nor inhibiting cell differentiation. In par-\nticular, actions associated with peptides\nderived from mammalian Ea domain\nhave only recently been established, with\nrodent Ea-peptide reported to possess\nbioactivity (38). Furthermore, it was\nshown that murine Ea- and Eb-peptides\nincrease cell entry of IGF-I from the\nmedia, providing evidence that, in addi-\ntion to having independent activity, they\nmay modulate IGF-I (30).\n\n210 | PHILIPPOU ET AL. | MOL MED 20:202-214, 2014\nIGF-I SPLICING, MODIFICATION AND BIOACTIVITY\nThe differential biological effects of the\nsynthetic Ec-peptide compared with ma-\nture IGF-I peptide, such as cell prolifera-\ntion versus differentiation, and the lack\nof suppression of the synthetic E-peptide\nbioactivity after blocking (mature) IGF-I\nsignaling with IGF-IR neutralizing anti-\nbodies, makes it tempting to postulate\nthat the Ec-peptide acts via a different re-\nceptor (28,135,145). However, concerns\nhave been raised about the effectiveness\nof the IGF-IR neutralizing antibodies to\nblock IGF-I signaling, since they could\ninternalize and, in this way, even activate\nthe receptor, or change its localization,\nthus facilitating an E-peptide action (38).\nNevertheless, the mature IGF-I bioactiv-\nity appeared to be suppressed in those\ncells where IGF-IR neutralizing antibod-\nies were used.\nInterestingly, an autonomous, IGF-IR-\nindependent bioactivity has been re-\nported for specific regions of the human\nEb domain of IGF-I; the synthetic analog\nof the EB\n1 peptide (see above: IGF-I Pro-\ncessing, Secretion and Glycosylation) was\nfound not only to possess mitogenic ac-\ntivity in human bronchial epithelial cells\nbut also to bind to specific high-affinity\nreceptors on those cells. Furthermore,\nneither ligand binding was inhibited by\nrecombinant IGF-I or recombinant in-\nsulin, nor did a monoclonal antibody an-\ntagonist to the IGF-IR suppress the prolif-\nerative response induced by the synthetic\nEB\n1 peptide. It was suggested that IB1 is a\ngrowth factor that mediates its effect\nthrough a specific receptor (20). Similarly,\nin accordance with the well established\nnotion that the molecular action of a\nbioactive peptide initiates through its\nspecific binding to a cell surface receptor,\nit was demonstrated that the human Eb-\npeptide binds to common cell surface\nmolecules on human neuroblastoma cells\n(27). Again, the saturation of mature \nIGF-I or insulin did not displace the bind-\ning of the human Eb-peptide, suggesting\nthe existence of distinct putative receptor\ncomponents on those cells (148).\nFurther evidence for distinct, IGF-IR-\n independent bioactivity of the human Ec\ndomain was provided from its divergent\nsignaling compared with mature IGF-I.\nOur group (31,33,34,65) and others (39)\nhave shown that a synthetic analog of the\nhuman Ec peptide possesses distinct sig-\nnaling compared with IGF-IR ligand. Spe-\ncifically, to distinguish any unique biolog-\nical effect of the Ec domain from a\npotential bioactivity possessed by the\ncommon part (that is, the first 16 residues)\nof the IGF-I E domains, we utilized a syn-\nthetic Ec-peptide corresponding to the \nregion beyond that common sequence,\nthat is, a peptide similar to the C-terminal\n24-residues, of which 16 are encoded by\nexon 5 and the last 8 by exon 6. It was\ndocumented that this sequence of the\nhuman Ec domain possesses distinct sig-\nnaling since, in contrast to mature IGF-I, it\nactivates only ERK1/2 and not Akt\n(31,33,34,65). Moreover, the selective acti-\nvation of only one of the two main signal-\ning pathways downstream of IGF-I/IGF-\nIR and the IGF-IR- and IR- independent\nbioactivity of this synthetic part of the Ec\ndomain was further documented by using\nthe siRNA knock-out model in various\nhuman cell lines (34,35,73).\nMore recently, synthetic E-peptides\ncorresponding to the rodent Ea and Eb\ndomain sequences were utilized in a ki-\nnase receptor activation (KIRA) assay to\ntest IGF-I-dependent and -independent\nactivation of IGF-IR by these E-peptides,\nshowing that they do not directly induce\nIGF-IR phosphorylation in mouse fibrob-\nlasts (P6 cells) (38). Interestingly, how-\never, the presence of either of those E-\npeptides increased IGF-IR activation by\nIGF-I in the murine C2C12 cell line, sug-\ngesting that they may modulate IGF-I ac-\ntivity (38). Moreover, from the findings\nof that model, it was concluded that E-\npeptides signaling, as well as mitogenic\nand motogenic effects are dependent\nupon a functional IGF-IR, and their activ-\nity plausibly reflects actions of pro-IGF-I\n(38). Further evidence has been provided\nrecently, suggesting that IGF-I splice\nvariants may exert their actions through\nmature IGF-I and not the E-peptides\n(149), or by supporting the bioactivity \nof pro-IGF-I forms (97). Furthermore,\nthere is controversial evidence regarding\nthe role of proforms (103,136,149) or \nE-peptides (38) of the murine IGF-I iso-\nforms, particularly in cell differentiation.\nWhether it reflects a differential mode of\naction of the free E-peptides compared\nwith the pro-IGF-I isoforms, or potential\ndifferences between their exogenous ad-\nministration and overexpression (or nat-\nural endogenous production), possibly\nshould be addressed.\nCollectively, the in vitro models utiliz-\ning the sequences of the rodent IGF-I \nE domains have provided specific infor-\nmation regarding the bioactivity and the\nmode of action of those E-peptides in\nmurine cell lines. The findings of these\nstudies suggest that the E-peptides have\nlittle or no independent activity and, in-\nstead, they modulate mature IGF-I (lig-\nand) activity and signaling (see review\nin [150]). Nevertheless, since these do-\nmains are very variable and much less\nconserved, it has been indicated that the\nspecies specificity must clearly be taken\ninto account when assessing the activity\nof the human IGF-I Eb and Ec domains,\nfrom which peptides with important bi-\nological activities have been reported\n(23). Thus, it remains to be further eluci-\ndated whether the autonomous, IGF-IR-\n independent and IR-independent bio -\nactivity, and the specific high-affinity\nreceptor binding observed for human\nEb- and Ec-peptides in various human\ncell lines (20,34,35,73,148) reflect an alter-\nnative, species-specific ligand/  receptor\nmechanism of action for these human\nE domains. Species-specific models, in\nterms of utilizing peptide sequences that\ncorrespond to the IGF-I E domains of\nthe species or the cells used in the\nmodel, are of particular importance to\ninvestigate physiological mechanisms \nof actions, such as a ligand/receptor\nmechanism.\nMoreover, it is still an elusive fact\nwhere the E-peptides signaling diverges\nfrom that of mature IGF-I (given the dis-\ntinct activation of ERK1/2 but not Akt\npathway by the E-peptide). It has been\nproposed that the E-peptides may in-\ncrease the phosphorylation of ERK1/2\nby increasing the IGF-IR internalization,\n\nREVIEW ARTICLE\nMOL MED 20:202-214, 2014 | PHILIPPOU ET AL. | 211\nwhich would act in favor of the MAPK\narm of the IGF-IR downstream signaling\n(38). However, since the possibility that\nthe E-peptides also activate ERK1/2\nthrough an IGF-IR independent mecha-\nnism cannot be excluded, it remains to\nbe further elucidated whether they af-\nfect the ERK1/2 pathway only at the\nlevel of the IGF-IR per se (that is, possi-\nbly by increasing its internalization, or\nby affecting its conformational change\nafter its binding to IGF-I and, thus, acti-\nvating exclusively a specific pathway)\nor if the E-peptide–induced activation of\nERK1/2 occurs at a level downstream of\nthe IGF-IR, or even upstream, in that the \nE-peptide utilizes another “receptor” or\ncellular uptake mechanism, given the\nactivation of ERK1/2 after exogenous\nadministration of the E-peptide. The in-\ntracellular signaling pathways initiated\nby IGF-IR ligation may interact with\nsignaling via G protein-coupled recep-\ntors or other mediators, modulating\nsome responses (151). However, the ex-\nistence of such a putative, canonical or\nnoncanonical receptor or internalization\nmechanism for the E-peptide remains to\nbe determined and characterized. Alter-\nnatively, the exclusive phosphorylation\nof ERK1/2 has been proposed to be a re-\nsult of a possible tuning of the IGF-IR\nsignaling cascade by the E-peptide to-\nward MAPK (38).\nThe identification of the signaling\npathways and effectors upstream and\ndownstream of the E-peptide–induced\nactivation of the extracellular-regulated\nkinases ERK1 and ERK2, the best-\n characterized members of the MAP ki-\nnase family (152), would provide infor-\nmation of particular importance\nregarding the mechanisms of E-peptide\nbioactivity. Furthermore, it is tempting to\nspeculate that the selective activation of\nMAPK/ERK1/2 pathway up to a critical\nlevel by the E-peptide might further\nexert an inhibitory regulation of its com-\npetitive PI3K/Akt pathway, since it has\nbeen shown that the activity of one path-\nway might inactivate portions of the\nother (153,154), thus consisting of a regu-\nlatory mechanism of competitive biologi-\ncal actions such as cell proliferation and\ndifferentiation.\nIt should be also mentioned that, given\nthe subcellular as well as differential local-\nization of the E-peptides (94,141,155,156),\nan intracrine signaling mechanism that\nmediates E-peptides bioactivity, as pro-\nposed for some peptide growth factors\n(157), cannot be excluded.\nCONCLUSION\nIGF-I regulation occurs at multiple lev-\nels due to the alternative splicing of its\npre-mRNA both at the 5′and 3′ end, and\nthe isoform-specific co- and posttransla-\ntional modifications, which appear to\nplay key roles in modulating the bio -\navailabilty and bioactivity of this growth\nfactor. New evidence supports the con-\ncept of IGF-I isoform– or IGF-I various\npeptides–specific functions, possibly via\nspecific and divergent signaling. Never-\ntheless, it remains to be determined where\nthe distinct E-peptide signaling diverges\nfrom that of mature IGF-I, or whether this\nsignaling is induced uniquely and au-\ntonomously by the E-peptides or syner-\ngistically with the mature IGF-I bioactive\nmolecule, and in the form of processed or\npro-IGF-I molecules. Thus, the regulatory\nmechanism(s) of a potentially specific and\napportioned bioactivity of the various\nIGF-I peptides is of particular interest\nwithin the context of a revisited character-\nization of the multiple IGF-I actions.\nDISCLOSURE\nThe authors declare that they have no\ncompeting interests as defined by Molecu-\nlar Medicine, or other interests that might\nbe perceived to influence the results and\ndiscussion reported in this paper.\nREFERENCES\n1. Liu JP , Baker J, Perkins AS, Robertson EJ, \nEfstratiadis A. (1993) Mice carrying null muta-\ntions of the genes encoding insulin-like growth\nfactor I (Igf-1) and type 1 IGF receptor (Igf1r).\nCell. 75:59–72.\n2. Laviola L, Natalicchio A, Giorgino F. (2007) The\nIGF-I signaling pathway. Curr. Pharm. Des. 13:663–9.\n3. Barton ER, et al. (2012) Deletion of muscle GRP94\nimpairs both muscle and body growth by inhibit-\ning local IGF production. FASEB J. 26:3691–702.\n4. Jones JI, Clemmons DR. (1995) Insulin-like\ngrowth factors and their binding proteins: bio-\nlogical actions. Endocr. Rev. 16:3–34.\n5. Koutsilieris M, Mitsiades C, Sourla A. (2000) In-\nsulin-like growth factor I and urokinase-type\nplasminogen activator bioregulation system as a\nsurvival mechanism of prostate cancer cells in\nosteoblastic metastases: development of anti-\n survival factor therapy for hormone-refractory\nprostate cancer. Mol. Med. 6:251–67.\n6. Siddle K, et al. (2001) Specificity in ligand bind-\ning and intracellular signalling by insulin and\n insulin-like growth factor receptors. Biochem. Soc.\nTrans. 29:513–25.\n7. LeRoith D, Roberts CT, Jr. (2003) The insulin-like\ngrowth factor system and cancer. Cancer Lett.\n195:127–37.\n8. Kooijman R. (2006) Regulation of apoptosis by\ninsulin-like growth factor (IGF)-I. Cytokine\nGrowth Factor Rev. 17:305–23.\n9. Werner H, LeRoith D. (1996) The role of the in-\nsulin-like growth factor system in human cancer.\nAdv. Cancer Res. 68:183–223.\n10. Reyes-Moreno C, Sourla A, Choki I, Doillon C,\nKoutsilieris M. (1998) Osteoblast-derived survival\nfactors protect PC-3 human prostate cancer cells\nfrom adriamycin apoptosis. Urology. 52:341–7.\n11. Werner H, Bruchim I. (2009) The insulin-like\ngrowth factor-I receptor as an oncogene. Arch.\nPhysiol. Biochem. 115:58–71.\n12. Ewton DZ, Roof SL, Magri KA, McWade FJ,\nFlorini JR. (1994) IGF-II is more active than IGF-I\nin stimulating L6A1 myogenesis: greater mito-\ngenic actions of IGF-I delay differentiation. J. Cell\nPhysiol. 161:277–84.\n13. Florini JR, Ewton DZ, Coolican SA. (1996) Growth\nhormone and the insulin-like growth factor sys-\ntem in myogenesis. Endocr. Rev. 17:481–517.\n14. Daughaday WH, Rotwein P . (1989) Insulin-like\ngrowth factors I and II. Peptide, messenger ri-\nbonucleic acid and gene structures, serum, and\ntissue concentrations. Endocr. Rev. 10:68–91.\n15. Ohlsson C, et al. (2009) The role of liver-derived in-\nsulin-like growth factor-I. Endocr. Rev. 30:494–535.\n16. Naranjo WM, et al. (2002) Protein calorie restric-\ntion affects nonhepatic IGF-I production and the\nlymphoid system: studies using the liver-specific\nIGF-I gene-deleted mouse model. Endocrinology.\n143:2233–41.\n17. Klover P , Hennighausen L. (2007) Postnatal body\ngrowth is dependent on the transcription factors\nsignal transducers and activators of transcrip-\ntion 5a/b in muscle: a r\nole for autocrine/\nparacrine insulin-like growth factor I. Endocrinol-\nogy. 148:1489–97.\n18. Stratikopoulos E, Szabolcs M, Dragatsis I, Kli-\nnakis A, Efstratiadis A. (2008) The hormonal ac-\ntion of IGF1 in postnatal mouse growth. Proc.\nNatl. Acad. Sci. U. S. A. 105:19378–83.\n19. Duan C, Ren H, Gao S. (2010) Insulin-like growth\nfactors (IGFs), IGF receptors, and IGF-binding\nproteins: roles in skeletal muscle growth and dif-\nferentiation. Gen. Comp. Endocrinol. 167:344–51.\n20. Siegfried JM, et al. (1992) A mitogenic peptide\namide encoded within the E peptide domain of\n\n212 | PHILIPPOU ET AL. | MOL MED 20:202-214, 2014\nIGF-I SPLICING, MODIFICATION AND BIOACTIVITY\nthe insulin-like growth factor IB prohormone.\nProc. Natl. Acad. Sci. U. S. A. 89:8107–11.\n21. Gilmour RS. (1994) The implications of insulin-\nlike growth factor mRNA heterogeneity. J. En-\ndocrinol. 140:1–3.\n22. Chew SL, Lavender P , Clark AJ, Ross RJ. (1995)\nAn alternatively spliced human insulin-like\ngrowth factor-I transcript with hepatic tissue ex-\npression that diverts away from the mitogenic\nIBE1 peptide. Endocrinology. 136:1939–44.\n23. Wallis M. (2009) New insulin-like growth factor\n(IGF)-precursor sequences from mammalian\ngenomes: the molecular evolution of IGFs and\nassociated peptides in primates. Growth Horm.\nIGF. Res. 19:12–23.\n24. Rotwein P , Pollock KM, Didier DK, Krivi GG.\n(1986) Organization and sequence of the human\ninsulin-like growth factor I gene. Alternative\nRNA processing produces two insulin-like\ngrowth factor I precursor peptides. J. Biol. Chem.\n261:4828–32.\n25. Duguay SJ, Lai-Zhang J, Steiner DF. (1995) Mu-\ntational analysis of the insulin-like growth fac-\ntor I prohormone processing site. J. Biol. Chem.\n270:17566–74.\n26. Barton ER. (2006) The ABCs of IGF-I isoforms:\nimpact on muscle hypertrophy and implications\nfor repair. Appl. Physiol. Nutr. Metab. 31:791–7.\n27. Kuo YH, Chen TT. (2002) Novel activities of pro-\nIGF-I E peptides: regulation of morphological dif-\nferentiation and anchorage-independent growth in\nhuman neuroblastoma cells. Exp. Cell Res. 280:75–89.\n28. Mills P , Dominique JC, Lafreniere JF, Bouchen-\ntouf M, Tremblay JP . (2007) A synthetic mechano\ngrowth factor E Peptide enhances myogenic pre-\ncursor cell transplantation success. Am. J. Trans-\nplant. 7:2247–59.\n29. Dluzniewska J, et al. (2005) A strong neuroprotec-\ntive effect of the autonomous C-terminal peptide\nof IGF-1 Ec (MGF) in brain ischemia. FASEB J.\n19:1896–8.\n30. Pfeffer LA, Brisson BK, Lei H, Barton ER. (2009)\nThe insulin-like growth factor (IGF)-I E-peptides\nmodulate cell entry of the mature IGF-I protein.\nMol. Biol. Cell. 20:3810–7.\n31. Philippou A, et al. (2009) Expression of IGF-1 iso-\nforms after exercise-induced muscle damage in\nhumans: characterization of the MGF E peptide\nactions in vitro. In Vivo. 23:567–75.\n32. Quesada A, Ogi J, Schultz J, Handforth A. (2011)\nC-terminal mechano-growth factor induces heme\noxygenase-1-mediated neuroprotection of SH-\nSY5Y cells via the protein kinase C/Nrf2 path-\nway. J. Neurosci. Res. 89:394–405.\n33. Stavr\nopoulou A, et al. (2009) IGF-1 expression in\ninfarcted myocardium and MGF E peptide actions\nin rat cardiomyocytes in vitro. Mol. Med. 15:127–35.\n34. Armakolas A, et al. (2010) Preferential expression of\nIGF-1Ec (MGF) transcript in cancerous tissues of\nhuman prostate: evidence for a novel and au-\ntonomous growth factor activity of MGF E peptide\nin human prostate cancer cells. Prostate. 70:1233–42.\n35. Milingos DS, et al. (2010) Insulinlike growth factor-\n1Ec (MGF) expression in eutopic and ectopic en-\ndometrium: characterization of the MGF E-peptide\nactions in vitro. Mol. Med. 17:21–8.\n36. Barton ER, DeMeo J, Lei H. (2010) The insulin-\nlike growth factor (IGF)-I E-peptides are required\nfor isoform-specific gene expression and muscle\nhypertrophy after local IGF-I production. J. Appl.\nPhysiol. (1985). 108:1069–76.\n37. Gentile MA, et al. (2010) Androgen-mediated\nimprovement of body composition and muscle\nfunction involves a novel early transcriptional\nprogram including IGF1, mechano growth fac-\ntor, and induction of {beta}-catenin. J. Mol. En-\ndocrinol. 44:55–73.\n38. Brisson BK, Barton ER. (2012) Insulin-like growth\nfactor-I E-peptide activity is dependent on the\nIGF-I receptor. PLoS One. 7:e45588.\n39. Quesada A, Micevych P , Handforth A. (2009) \nC-terminal mechano growth factor protects\ndopamine neurons: a novel peptide that induces\nheme oxygenase-1. Exp. Neurol. 220:255–66.\n40. Deng M, et al. (2011) Mechano growth factor E pep-\ntide promotes osteoblasts proliferation and bone-\ndefect healing in rabbits. Int. Orthop. 35:1099–106.\n41. Adamo ML, et al. (1993) Distinct promoters in the\nrat insulin-like growth factor-I (IGF-I) gene are\nactive in CHO cells. Endocrinology. 132:935–937.\n42. Yang H, et al. (1995) Alternative leader sequences\nin insulin-like growth factor I mRNAs modulate\ntranslational efficiency and encode multiple sig-\nnal peptides. Mol. Endocrinol. 9:1380–95.\n43. Shavlakadze T, Winn N, Rosenthal N, Grounds\nMD. (2005) Reconciling data from transgenic\nmice that overexpress IGF-I specifically in skele-\ntal muscle. Growth Horm. IGF. Res. 15:4–18.\n44. Philippou A, Maridaki M, Halapas A, Koutsilieris\nM. (2007) The role of the insulin-like growth factor\n1 (IGF-1) in skeletal muscle physiology. In Vivo.\n21:45–54.\n45. Okazaki R, Durham SK, Riggs BL, Conover CA.\n(1995) Transforming growth factor-beta and\nforskolin increase all classes of insulin-like growth\nfactor-I transcripts in normal human osteoblast-like\ncells. Biochem. Biophys. Res. Commun. 207:963–70.\n46. Bloor CA, Knight RA, Kedia RK, Spiteri MA,\nAllen JT\n. (2001) Differential mRNA expression of\ninsulin-like growth factor-1 splice variants in pa-\ntients with idiopathic pulmonary fibrosis and\npulmonary sarcoidosis. Am. J. Respir. Crit. Care\nMed. 164:265–72.\n47. Shamblott MJ, Chen TT. (1993) Age-related and\ntissue-specific levels of five forms of insulin-like\ngrowth factor mRNA in a teleost. Mol. Mar. Biol.\nBiotechnol. 2:351–61.\n48. Duguay SJ, Swanson P , Dickhoff WW. (1994) Dif-\nferential expression and hormonal regulation of\nalternatively spliced IGF-I mRNA transcripts in\nsalmon. J Mol. Endocrinol. 12:25–37.\n49. Faulk CK, Perez-Dominguez R, Webb KA, Jr.,\nHolt GJ. (2010) The novel finding of four distinct\nprepro-IGF-I E domains in a perciform fish, Sci-\naenops ocellatus, during ontogeny. Gen. Comp.\nEndocrinol. 169:75–81.\n50. Zhang J, et al. (2013) Cloning and characteriza-\ntion of new transcript variants of insulin-like\ngrowth factor-I in Sika deer (Cervus elaphus).\nGrowth Horm. IGF. Res. 23:120–7.\n51. Jansen E, Steenbergh PH, LeRoith D, Roberts CT, Jr.,\nSussenbach JS. (1991) Identification of multiple tran-\nscription start sites in the human insulin-like\ngrowth factor-I gene. Mol. Cell. Endocrinol. 78:115–25.\n52. Philippou A, Armakolas A, Koutsilieris M. (2013)\nEvidence for the Possible Biological Significance\nof the igf-1 Gene Alternative Splicing in Prostate\nCancer. Front Endocrinol. (Lausanne). 4:31.\n53. Adamo ML, Ben-Hur H, LeRoith D, Roberts CT,\nJr. (1991) Transcription initiation in the two\nleader exons of the rat IGF-I gene occurs from\ndisperse versus localized sites. Biochem. Biophys.\nRes. Commun. 176:887–93.\n54. D’Ercole AJ, Stiles AD, Underwood LE. (1984)\nTissue concentrations of somatomedin C: further\nevidence for multiple sites of synthesis and\nparacrine or autocrine mechanisms of action.\nProc. Natl. Acad. Sci. U. S. A. 81:935–9.\n55. Rotwein P , Pollock KM, Watson M, Milbrandt JD.\n(1987) Insulin-like growth factor gene expression\nduring rat embryonic development. Endocrinol-\nogy. 121:2141–4.\n56. Adamo M, Lowe WL, Jr., LeRoith D, Roberts CT,\nJr. (1989) Insulin-like growth factor I messenger\nribonucleic acids with alternative 5′-untranslated\nregions are differentially expressed during devel-\nopment of the rat. Endocrinology. 124:2737–44.\n57. W\nang X, Yang Y, Adamo ML. (1997) Characteri-\nzation of the rat insulin-like growth factor I gene\npromoters and identification of a minimal exon 2\npromoter. Endocrinology. 138:1528–36.\n58. O’Sullivan DC, Szestak TA, Pell JM. (2002) Regu-\nlation of IGF-I mRNA by GH: putative functions\nfor class 1 and 2 message. Am. J. Physiol. En-\ndocrinol. Metab. 283:E251–8.\n59. Woelfle J, Chia DJ, Rotwein P . (2003) Mechanisms\nof growth hormone (GH) action. Identification of\nconserved Stat5 binding sites that mediate GH-\ninduced insulin-like growth factor-I gene activa-\ntion. J. Biol. Chem. 278:51261–6.\n60. Chia DJ, Young JJ, Mertens AR, Rotwein P . (2010)\nDistinct alterations in chromatin organization of\nthe two IGF-I promoters precede growth hor-\nmone-induced activation of IGF-I gene transcrip-\ntion. Mol. Endocrinol. 24:779–89.\n61. Lund PK. (1994) Insulin-like growth factor I: molec-\nular biology and relevance to tissue-specific expres-\nsion and action. Recent Prog. Horm. Res. 49:125–48.\n62. Hameed M, et al. (2004) The effect of recombi-\nnant human growth hormone and resistance\ntraining on IGF-I mRNA expression in the mus-\ncles of elderly men. J. Physiol. 555:231–40.\n63. Imanaka M, et al. (2008) Growth hormone stimu-\nlates mechano growth factor expression and acti-\nvates myoblast transformation in C2C12 cells.\nKobe J. Med. Sci. 54:E46–54.\n64. Aperghis M, et al. (2009) Serum IGF-I levels and\nIGF-I gene splicing in muscle of healthy young\nmales receiving rhGH. Growth Horm. IGF. Res.\n19:61–7.\n65. Moschos MM, et al. (2011) Expression of the in-\nsulin-like growth factor 1 (IGF-1) and type I IGF\n\nREVIEW ARTICLE\nMOL MED 20:202-214, 2014 | PHILIPPOU ET AL. | 213\nreceptor mRNAs in human HLE-B3 lens epithe-\nlial cells. In Vivo. 25:179–84.\n66. Jansen M, et al. (1983) Sequence of cDNA encod-\ning human insulin-like growth factor I precursor.\nNature. 306:609–11.\n67. Rotwein P . (1986) Two insulin-like growth factor\nI messenger RNAs are expressed in human liver.\nProc. Natl. Acad. Sci. U. S. A. 83:77–81.\n68. McKay BR, O’Reilly CE, Phillips SM, Tarnopolsky\nMA, Parise G. (2008) Co-expression of IGF-1 fam-\nily members with myogenic regulatory factors\nfollowing acute damaging muscle-lengthening\ncontractions in humans. J. Physiol. 586:5549–60.\n69. Koczorowska MM, Kwasniewska A, Gozdzicka-\nJozefiak A. (2011) IGF1 mRNA isoform expres-\nsion in the cervix of HPV-positive women with\npre-cancerous and cancer lesions. Exp. Ther. Med.\n2:149–56.\n70. Shimatsu A, Rotwein P . (1987) Sequence of two rat\ninsulin-like growth factor I mRNAs differing within\nthe 5′untranslated region. Nucleic Acids Res. 15:7196.\n71. Goldspink G. (2005) Mechanical signals, IGF-I\ngene splicing, and muscle adaptation. Physiology\n(Bethesda). 20:232–8.\n72. Matheny RW, Jr., Nindl BC, Adamo ML. (2010)\nMinireview: Mechano-growth factor: a putative\nproduct of IGF-I gene expression involved in tissue\nrepair and regeneration. Endocrinology. 151:865–875.\n73. Philippou A, et al. (2011) IGF1Ec expression in\nMG-63 human osteoblast-like osteosarcoma cells.\nAnticancer Res. 31:4259–65.\n74. Yu H, Rohan T. (2000) Role of the insulin-like\ngrowth factor family in cancer development and\nprogression. J. Natl. Cancer Inst. 92:1472–89.\n75. Mourkioti F, Rosenthal N. (2005) IGF-1, inflam-\nmation and stem cells: interactions during mus-\ncle regeneration. Trends Immunol. 26:535–42.\n76. Temmerman L, Slonimsky E, Rosenthal N. (2010)\nClass 2 IGF-1 isoforms are dispensable for viabil-\nity, growth and maintenance of IGF-1 serum lev-\nels. Growth Horm. IGF. Res. 20:255–63.\n77. Zhang J, Whitehead RE, Jr., Underwood LE.\n(1997) Effect of fasting on insulin-like growth fac-\ntor (IGF)-IA and IGF-IB messenger ribonucleic\nacids and prehormones in rat liver. Endocrinology.\n138:3112–8.\n78. Duguay SJ. (1999) Post-translational processing\nof insulin-like growth factors. Horm. Metab. Res.\n31:43–9.\n79. Powell DR, Lee PD, Chang D, Liu F, Hintz RL.\n(1987) Antiserum developed for the E peptide region\nof insulin-like growth factor IA prohormone recog-\nnizes a serum protein by both immunoblot and ra-\ndioimmunoassay. J. Clin. Endocrinol. Metab65:868–75.\n80. W\nilson HE, Westwood M, White A, Clayton PE.\n(2001) Monoclonal antibodies to the carboxy-\n terminal Ea sequence of pro-insulin-like growth fac-\ntor-IA (proIGF-IA) recognize proIGF-IA secreted by\nIM9 B-lymphocytes. Growth Horm. IGF. Res. 11:10–7.\n81. Kravchenko IV , Furalyov VA, Khotchenkov VP ,\nPopov VO. (2006) Monoclonal antibodies to\nmechano-growth factor. Hybridoma (Larchmt).\n25:300–5.\n82. Philippou A, et al. (2008) Characterization of a\nrabbit antihuman mechano growth factor (MGF)\npolyclonal antibody against the last 24 amino\nacids of the E domain. In Vivo. 22:27–35.\n83. Lingappa VR, Lingappa JR, Blobel G. (1980) Sig-\nnal sequences for early events in protein secre-\ntion and membrane assembly. Ann. N. Y. Acad.\nSci. 343:356–61.\n84. Denley A, Cosgrove LJ, Booker GW, Wallace JC,\nForbes BE. (2005) Molecular interactions of the\nIGF system. Cytokine Growth Factor Rev. 16:421–39.\n85. Carlsson-Skwirut C, et al. (1986) Isolation and\ncharacterization of variant IGF-1 as well as IGF-2\nfrom adult human brain. FEBS Lett 201:46–50.\n86. Sara VR, et al . (1993) The biological role of trun-\ncated insulin-like growth factor-1 and the tripep-\ntide GPE in the central nervous system. Ann. N. Y.\nAcad. Sci. 692:183–91.\n87. Nakayama K. (1997) Furin: a mammalian subtil-\nisin/Kex2p-like endoprotease involved in pro-\ncessing of a wide variety of precursor proteins.\nBiochem. J. 327 ( Pt 3): 625–35.\n88. Duckert P , Brunak S, Blom N. (2004) Prediction\nof proprotein convertase cleavage sites. Protein\nEng. Des. Sel. 17:107–12.\n89. Rouille Y, et al. (1995) Proteolytic processing\nmechanisms in the biosynthesis of neuroen-\ndocrine peptides: the subtilisin-like proprotein\nconvertases. Front Neuroendocrinol. 16:322–61.\n90. Steiner DF. (1998) The proprotein convertases.\nCurr. Opin. Chem. Biol. 2:31–9.\n91. Seidah NG, Chretien M. (1999) Proprotein and\nprohormone convertases: a family of subtilases\ngenerating diverse bioactive polypeptides. Brain\nRes. 848:45–62.\n92. Duguay SJ, Milewski WM, Young BD, Nakayama\nK, Steiner DF. (1997) Processing of wild-type and\nmutant proinsulin-like growth factor-IA by sub-\ntilisin-related proprotein convertases. J. Biol.\nChem. 272:6663–70.\n93. Thomas G. (2002) Furin at the cutting edge: from\nprotein traf\nfic to embryogenesis and disease. Nat.\nRev. Mol. Cell. Biol. 3:753–66.\n94. Tan NS, Ho B, Ding JL. (2002) Engineering a\nnovel secretion signal for cross-host recombinant\nprotein expression. Protein Eng. 15:337–45.\n95. Conover CA, Baker BK, Hintz RL. (1989) Cul-\ntured human fibroblasts secrete insulin-like\ngrowth factor IA prohormone. J. Clin. Endocrinol.\nMetab. 69:25–30.\n96. Conover CA, et al. (1993) Human hepatoma cells\nsynthesize and secrete insulin-like growth factor\nIa prohormone under growth hormone control.\nRegul. Pept. 48:1–8.\n97. Durzynska J, Philippou A, Brisson BK, Nguyen-\nMcCarty M, Barton ER. (2013) The pro-forms of\ninsulin-like growth factor I (IGF-I) are predomi-\nnant in skeletal muscle and alter IGF-I receptor\nactivation. Endocrinology. 154:1215–24.\n98. Tsuji A, et al. (2003) Secretory proprotein conver-\ntases PACE4 and PC6A are heparin-binding pro-\nteins which are localized in the extracellular ma-\ntrix. Potential role of PACE4 in the activation of\nproproteins in the extracellular matrix. Biochim.\nBiophys. Acta. 1645:95–104.\n99. Marshall RD. (1972) Glycoproteins. Annu. Rev.\nBiochem. 41:673–702.\n100. Imperiali B, O’Connor SE, Hendrickson T, Kel-\nlenberger C. (1999) Chemistry and biology of\nasparagine-linked glycosylation. Pure Appl.\nChem. 71:777–87.\n101. Simons G, et al. (1993) Overproduction of\nbovine beta-casein in Escherichia coli and engi-\nneering of its main chymosin cleavage site. Pro-\ntein Eng. 6:763–70.\n102. Freeze HH. (2001) Endoglycosidase and gly-\ncoamidase release of N-linked oligosaccharides.\nCurr. Protoc. Mol. Biol. Chapter 17:Unit17 13A.\n103. Hede MS, et al. (2012) E-peptides control\nbioavailability of IGF-1. PLoS One. 7:e51152.\n104. Metzger F, et al. (2011) Separation of fast from\nslow anabolism by site-specific PEGylation of\ninsulin-like growth factor I (IGF-I). J. Biol. Chem.\n286:19501–10.\n105. Sivaramakrishnan M, et al. (2013) PEGylation of\nlysine residues reduces the pro-migratory activ-\nity of IGF-I. Biochim. Biophys. Acta. 1830:4734–42.\n106. Rinderknecht E, Humbel RE. (1978) The amino\nacid sequence of human insulin-like gr\nowth fac-\ntor I and its structural homology with proin-\nsulin. J. Biol. Chem. 253:2769–76.\n107. Federici M, et al. (1997) Increased abundance of\ninsulin/IGF-I hybrid receptors in adipose tissue\nfrom NIDDM patients. Mol. Cell. Endocrinol.\n135:41–7.\n108. Nakae J, Kido Y, Accili D. (2001) Distinct and\noverlapping functions of insulin and IGF-I re-\nceptors. Endocr. Rev. 22:818–35.\n109. Taguchi A, White MF. (2008) Insulin-like signal-\ning, nutrient homeostasis, and life span. Annu.\nRev. Physiol. 70:191–212.\n110. De Meyts P , Whittaker J. (2002) Structural biol-\nogy of insulin and IGF1 receptors: implications\nfor drug design. Nat. Rev. Drug. Discov. 1:769–83.\n111. Kristensen C, Wiberg FC, Andersen AS. (1999)\nSpecificity of insulin and insulin-like growth\nfactor I receptors investigated using chimeric\nmini-receptors. Role of C-terminal of receptor\nalpha subunit. J. Biol. Chem. 274:37351–6.\n112. Baserga R, Peruzzi F, Reiss K. (2003) The IGF-1\nreceptor in cancer biology. Int. J. Cancer 107:873–7.\n113. Manes S, et al. (1997) Functional epitope map-\nping of insulin-like growth factor I (IGF-I) by\nanti-IGF-I monoclonal antibodies. Endocrinology.\n138:905–15.\n114. Philippou A, Halapas A, Maridaki M, Koutsilieris\nM. (2007) Type I insulin-like growth factor recep-\ntor signaling in skeletal muscle regeneration and\nhypertrophy. J. Musculoskelet. Neuronal Interact.\n7:208–18.\n115. Wu J, et al. (2008) Small-molecule inhibition and\nactivation-loop trans-phosphorylation of the\nIGF1 receptor. EMBO. J. 27:1985–94.\n116. Ozkan EE. (2011) Plasma and tissue insulin-like\ngrowth factor-I receptor (IGF-IR) as a prognostic\nmarker for prostate cancer and anti-IGF-IR agents\nas novel therapeutic strategy for refractory cases:\na review. Mol. Cell. Endocrinol. 344:1–24.\n117. Hubbard SR. (1997) Crystal structure of the acti-\n\nvated insulin receptor tyrosine kinase in com-\nplex with peptide substrate and ATP analog.\nEMBO J. 16:5572–81.\n118. Baserga R. (1999) The IGF-I receptor in cancer\nresearch. Exp. Cell. Res. 253:1–6.\n119. Baxter RC. (2000) Insulin-like growth factor\n(IGF)-binding proteins: interactions with IGFs\nand intrinsic bioactivities. Am. J. Physiol. En-\ndocrinol. Metab. 278:E967–76.\n120. Duan C. (2002) Specifying the cellular re-\nsponses to IGF signals: roles of IGF-binding\nproteins. J. Endocrinol. 175:41–54.\n121. Cohen P . (2006) Overview of the IGF-I system.\nHorm. Res. 65 Suppl 1:3–8.\n122. Baxter RC, Martin JL, Beniac VA. (1989) High\nmolecular weight insulin-like growth factor\nbinding protein complex. Purification and prop-\nerties of the acid-labile subunit from human\nserum. J. Biol. Chem. 264:11843–8.\n123. Baxter RC, Martin JL. (1989) Structure of the Mr\n140,000 growth hormone-dependent insulin-like\ngrowth factor binding protein complex: deter-\nmination by reconstitution and affinity-labeling.\nProc. Natl. Acad. Sci. U. S. A. 86:6898–902.\n124. Kelley KM, et al. (1996) Insulin-like growth factor-\nbinding proteins (IGFBPs) and their regulatory\ndynamics. Int. J. Biochem. Cell Biol. 28:619–37.\n125. Clemmons DR. (1997) Insulin-like growth factor\nbinding proteins and their role in controlling\nIGF actions. Cytokine Growth Factor Rev. 8:45–62.\n126. Firth SM, Baxter RC. (2002) Cellular actions of\nthe insulin-like growth factor binding proteins.\nEndocr. Rev. 23:824–54.\n127. Oh Y. (1998) IGF-independent regulation of\nbreast cancer growth by IGF binding proteins.\nBreast Cancer Res. Treat. 47:283–93.\n128. Koutsilieris M, Lavergne E, Lemay A. (1997) As-\nsociation of protease activity against IGFBP-3\nwith peritoneal fluid mitogens: possible impli-\ncations for the ectopic growth of endometrial\ncells in women with endometriosis. Anticancer\nRes. 17:1239–44.\n129. Hwa V , Oh Y, Rosenfeld RG. (1999) The insulin-\nlike growth factor-binding protein (IGFBP) su-\nperfamily. Endocr. Rev. 20:761–87.\n130. Hwa V , Oh Y, Rosenfeld RG. (1999) Insulin-like\ngrowth factor binding proteins: a proposed su-\nperfamily. Acta. Paediatr. Suppl. 88:37–45.\n131. Lopez-Bermejo A, et\nal. (2006) Insulin resistance\nis associated with increased serum concentra-\ntion of IGF-binding protein-related protein 1\n(IGFBP-rP1/MAC25). Diabetes. 55:2333–9.\n132. Barton ER. (2006) Viral expression of insulin-\nlike growth factor-I isoforms promotes different\nresponses in skeletal muscle. J. Appl. Physiol.\n(1985) 100:1778–84.\n133. Carpenter V , et al. (2008) Mechano-growth fac-\ntor reduces loss of cardiac function in acute my-\nocardial infarction. Heart Lung Circ. 17:33–9.\n134. Rae FK, et al. (2012) Proximal tubule overex-\npression of a locally acting IGF isoform, Igf-1Ea,\nincreases inflammation after ischemic injury.\nGrowth Horm. IGF. Res. 22:6–16.\n135. Ates K, et al. (2007) The IGF-I splice variant MGF\nincreases progenitor cells in ALS, dystrophic,\nand normal muscle. FEBS Lett 581:2727–32.\n136. Poudel B, et al. (2011) Increased cardiogenesis in\nP19-GFP teratocarcinoma cells expressing the\npropeptide IGF-1Ea. Biochem. Biophys. Res. Com-\nmun. 416:293–9.\n137. Kasprzak A, et al. (2012) Differential expression\nof IGF-1 mRNA isoforms in colorectal carci-\nnoma and normal colon tissue. Int. J. Oncol.\n42:305–16.\n138. Vinciguerra M, Musaro A, Rosenthal N. (2010)\nRegulation of muscle atrophy in aging and dis-\nease. Adv. Exp. Med. Biol. 694:211–33.\n139. Vinciguerra M, et al. (2012) mIGF-1/JNK1/SirT1\nsignaling confers protection against oxidative\nstress in the heart. Aging Cell. 11:139–49.\n140. Santini MP , et al. (2007) Enhancing repair of the\nmammalian heart. Circ. Res. 100:1732–40.\n141. Durzynska J, Wardzinski A, Koczorowska M,\nGozdzicka-Jozefiak A, Barton ER. (2013) Human\nEb peptide: not just a by-product of pre-pro-\nIGF1b processing? Horm. Metab. Res. 45:415–22.\n142. Deng M, et al. (2012) New proangiogenic activ-\nity on vascular endothelial cells for C-terminal\nmechano growth factor. Acta. Biochim. Biophys.\nSin. (Shanghai). 44:316–22.\n143. Collins JM, Goldspink PH, Russell B. (2010)\nMigration and pr\noliferation of human mes-\nenchymal stem cells is stimulated by different\nregions of the mechano-growth factor prohor-\nmone. J. Mol. Cell Cardiol. 49:1042–5.\n144. Chen MJ, et al. (2007) Suppression of growth\nand cancer-induced angiogenesis of aggressive\nhuman breast cancer cells (MDA-MB-231) on\nthe chorioallantoic membrane of developing\nchicken embryos by E-peptide of pro-IGF-I. \nJ. Cell. Biochem. 101:1316–27.\n145. Yang SY, Goldspink G. (2002) Different roles of\nthe IGF-I Ec peptide (MGF) and mature IGF-I in\nmyoblast proliferation and differentiation. FEBS\nLett 522:156–60.\n146. Qin LL, et al. (2013) Mechano growth factor\n(MGF) promotes proliferation and inhibits dif-\nferentiation of porcine satellite cells (PSCs) by\ndown-regulation of key myogenic transcrip-\ntional factors. Mol. Cell. Biochem. 370:221–30.\n147. Tian XC, Chen MJ, Pantschenko AG, Yang TJ,\nChen TT. (1999) Recombinant E-peptides of pro-\nIGF-I have mitogenic activity. Endocrinology.\n140:3387–90.\n148. Kuo YH, Chen TT. (2003) Specific cell surface\nbinding sites shared by human Pro-IGF-I Eb-\npeptides and rainbow trout Pro-IGF-I Ea-4-\n peptide. Gen. Comp. Endocrinol. 132:231–40.\n149. Matheny RW, Jr., Nindl BC. (2011) Loss of IGF-\nIEa or IGF-IEb impairs myogenic differentia-\ntion. Endocrinology. 152:1923–34.\n150. Brisson BK, Barton ER. (2013) New Modulators\nfor IGF-I Activity within IGF-I Processing Prod-\nucts. Front Endocrinol. (Lausanne) 4:42.\n151. Malbon CC, Karoor V . (1998) G-protein-linked\nreceptors as tyrosine kinase substrates: new\nparadigms in signal integration. Cell Signal.\n10:523–7.\n152. Blenis J. (1993) Signal transduction via the MAP\nkinases: proceed at your own RSK. Proc. Natl.\nAcad. Sci. U. S. A. 90:5889–92.\n153. Samuel DS, et al. (1999) Raf-1 activation stimu-\nlates proliferation and inhibits IGF-stimulated\ndifferentiation in L6A1 myoblasts. Horm. Metab.\nRes. 31:55–64.\n154. Rommel C, et al. (1999) Differentiation stage-\nspecific inhibition of the Raf-MEK-ERK path-\nway by Akt. Science. 286:1738–41.\n155. Zhang B, Xian C, Luo Y, Wang Y. (2009) Expres-\nsion and subcellular localization of mechano-\ngrowth factor in osteoblasts under mechanical\nstr\netch. Sci. China C. Life Sci. 52:928–34.\n156. Peng Q, et al. (2012) The nuclear localization of\nMGF receptor in osteoblasts under mechanical\nstimulation. Mol. Cell Biochem. 369:147–56.\n157. Re RN. (2003) The intracrine hypothesis and in-\ntracellular peptide hormone action. Bioessays.\n25:401–9.\nIGF-I SPLICING, MODIFICATION AND BIOACTIVITY\n214 | PHILIPPOU ET AL. | MOL MED 20:202-214, 2014","source_license":"CC-BY-4.0","license_restricted":false}