Previously, we described a mass-spectrometric method for monitoring Insulin-like growth factor-1 (IGF-1) variants by using only 4 mass-to-charge ratios (m/z) comprising variant groups (VG). For each variant, the method used a concept we defined as “the isotopic peak index” (IPi), as well as relative retention time (rRT). VG, IPi, and rRT are unique values that help identify IGF-1 variants and detect novel ones. In addition, by monitoring a signature y-ion using MS/MS, we could distinguish isobaric A67T and A70T IGF-1 variants. More recently, we have used y5 ions to verify the following IGF-1 variants (corresponding amino acid sequence): P66A (AAKSA), A67V (PVKSA), and A67S (PSKSA) in specimens suspected to contain those variants (as assessed by their VG, IPi, and rRT). In several patients, a potentially recurring variant was detected (IP1@VG1, rRT [average] = -0.19 mins) that did not match previously identified IGF-1 variants. Herein we describe how this variant was identified.
MS Analysis and Results
In the first specimen, manually analyzing m/z differences in y-ions of the MS/MS of the unknown variant, a partial sequence at the C-terminus of was identified as PAK matching the WT IGF-1 sequence, but with a shift to 214 Da higher mass indicating a change at the C-terminus (eg, amino acid substitution or modification). However, the total mass of this variant was 22 Daltons lighter than the WT IGF-1. A subsequent specimen provided a better signal for b- and y-ions for sequence information.
In the next specimen, partial sequences were deduced using b- and y-ions derived from amino acids near the C- and N-termini. The N-terminus sequence (ETL) was identified, but b-ions were 420 Da heavier that they should be if derived from the IGF-1 sequence. The C-terminal sequence identified (ATPAK) was inconsistent with IGF-1 (LKPAK) and, as in the earlier specimen, the y-ions were 214 Da heavier than predicted. A BLAST search across the human proteome using C-terminal sequence indicated a sequence consistent with human IGF-2, a 67-amino acid polypeptide. The N-terminus of IGF-2 starts with the sequence AYRPSETL, which includes the (ETL) fragment, but at much heavier m/z due to the presence of the remaining amino acids (the IGF-1 sequence is just GPETL). However, the C-terminus of the IGF-2 variant still had an additional 156 Da to be accounted for (now measured against IGF-2 sequence).
This difference in mass corresponds to a C-terminal arginine (R) extension to the mature IGF-2 sequence (1-67). An R68 immediately follows C-terminus of the mature IGF-2 polypeptide sequence and is the first amino acid of the E-domain of the prohormone. If this possibility is considered, all inconsistencies are resolved: all b- and y-ions match the correct m/z ratios, and the total m/z of the protein accurately matches the theoretical value within 0.6 ppm. The conclusion: this is a case of the IGF-2 protein with an R extension at the C-terminus (ATPAKSER).
In the IGF-2 prohormone, the E-domain (68-156), which starts with R, is cleaved off when the protein matures, leaving only the WT IGF-2 sequence. It seems that in the present case the R was retained with the IGF-2 sequence. To our knowledge, IGF-2 (1-68) has never been previously observed. It may have some relationship to so-called “big IGF-2”, which has been described as an IGF-2 protein in which incorrect cleavage results in extra amino acids from the E-domain of the pro IGF-2 sequence. Both big IGF-2 polypeptides described in the literature, 1-87 and 1-104, are biomarkers of non-islet cell tumor hypoglycemia (NICTH).
Our workflow for routine monitoring of IGF-1 variants by high-resolution mass spectrometry resulted in the detection of an unusual IGF-2 species (1-68) indicative of unexpected prohormone processing with unknown clinical significance and unknown frequency of occurrence.