Fig 1.
Effects in rats in vivo of plasma anionex fractions from OV-INTACT vs OVX sheep. Post-mortem organ weights expressed as a percentage of body weight for two groups of seven intact adult female Sprague Dawley rats. The zero baseline is provided by rats receiving for four days by intraperitoneal injection a fraction (spin and gel filtered to 10-20 kDa, eluting in HPLC anionex at ~0.2 M NaCl; S1 File Babraham Method) of the jugular vein plasmas pooled from two OVX sheep. The other group of rats received the same fraction in the same amount from ovarian venous plasmas pooled from six OV-INTACT sheep. No significant body weight changes were seen to confound the picture of organ shrinkage with adrenal enlargement. * = P < 0.05, one-tailed t test. (N.B. OV-INTACT versus PBS controls, adrenals +8.53%, significant: S2 File).
Fig 2.
MALDI MS of ovine ovarian follicular fluid.
Anion exchange Fraction 27 (i.e., late eluting), after spin and gel filtration. Matrix sinapinic acid, calibration against carbonic anhydrase (Babraham MS, S3 File). For Candidate 7500 MS peak analysis in regard to Figs 2–5 see Table 1. The peak at m/z 4785 is analysed to be a 42mer N-terminal fragment of sSgII-70 and the 1149 base peak (‘BP’ in header) an 11mer. (See S3 Figs. 1 & 2 in S3 File for equivalent bovine and porcine spectra.).
Table 1.
MALDI MS peak analysis for ovine Candidate 7500 identified as sSgII-70.
Analysis based on C-terminal truncation and water losses. Masses are for the native 70mer polypeptide sSgII-70 down to a 60mer (with predictions generated by Expasy Compute pI/Mw [14] and double-checked via an MS/MS fragment ion calculator: see S3 File). A mean deviation of m/z 2.66 is delivered at a mean error (i.e., difference as a percentage of match) of 0.035%; seven integer or next-integer matches (58%); ∑Observed (n = 12) 91120/∑Calculated Matches 91113 x100 = 100.01%; and chi-squared P = 0.024, significant (S3 Table 5 in S3 File). An extended version of Table 1 is provided as S3 Table 1 in S3 File.
Fig 3.
MALDI MS of ovine jugular vein EDTA plasma anionex Fraction 29.
The Beale 4 Edman reading (xxPxxxxVxxFNxx) was obtained from a gel lacking in visible bands but excised anyway at ~7 kDa. The Beale 4 aa correspond to EPL001 residues 3, 8, 11 and 12, without EPL001’s 10 other aa being discernable, thus: 1MKPLTGKVKEFNNI14 (S1 Table 1 in S1 File, SEQ ID NOs: 3 & 4). MALDI matrix and calibration as Fig 2. The ion at m/z 7805 is an integer match to a predicted sSgII-70 68mer. A peak without mass annotation is ‘c7500’. Minor peaks at 5070 and 4748 are analysed to be 45mer and 43mer fragments of sSgII-70, respectively. The base peak is missing from this record but is 1147 (‘B.P.’ in header), an sSgII-70 11mer.
Fig 4.
MALDI MS of a solid ‘upstream precipitate’ that formed on the 3 kDa membrane during a 3-30 kDa ultrafiltration of bovine ovarian follicular fluid (upper panel) and ovine jugular vein EDTA plasma (lower panel) (S1 in S1 File Sheffield Method). The sought-for mammalian Candidate 7500 flew in MS from a complex bacterially contaminated mixture added without treatment to a protein matrix. Calibration (against insulin) varied due to the use of high laser power (operator comment). The mV figures in the header (1.8 = bovine, 13 = ovine) are weak. Strong samples can show up to 1000 mV. The bovine double peak corresponds to an alanine difference, 67mer vs 66mer bSgII-70 (S3 Table 2 in S3 File). The ovine double peak corresponds to a lysine and four waters difference, 66mer vs 65mer, sSgII-70 (Table 1). All peaks in the lower spectra are predicted to be related to sSgII-70 (S3 Table 1 in S3 File). The item at 4755.85 is a next-integer match to an sSgII-70 predicted 42mer. The 9509.01 item is double this, suggesting artefactual homodimerization. The lower panel pattern has been reproduced on more purified uncontaminated upstream precipitate (S3 Fig. 5 in S3 File), while the upper panel has also been replicated (S3 Fig. 4 in S3 File). MS courtesy of Carolyn Carr, University of Oxford, Oxford, UK.
Fig 5.
MALDI MS of a solid ‘upstream precipitate’ that formed on the 3 kDa membrane during a 3−30 kDa ultrafiltration (with sodium azide) of ovine jugular vein EDTA plasma (S1 in S1 File Sheffield Method) – without the bacterial contamination of Fig 4. Matrix and calibration as Fig 2. Acquisition mass range 2500−25000 Da. Electroeluted SDS-PAGE Band 1 at ~7 kDa showed cell-inhibitory bioactivity in vitro. Edman sequencing of Band 1 from a prior (contaminated, azide free) purification provided the 14-residue N-terminal sequence MKPLTGKVKEFNNI (EPL001). This matched anionex Beale 4, the minimal sequence obtained from maximally purified (bacterially uncontaminated) material. Gel electroelution, as used here, provides a cleaner spectrum than solvent extraction (S3 Fig. 5 in S3 File), albeit with the same base peak, an sSgII-70 65mer, lacking –KANNI (Table 1).
Fig 6.
MALDI mass spectrum of bovine ovarian follicular fluid, analysed for variations in bSgII-70 C-terminal truncation and water differences. Anionex Fraction 24 (i.e., late eluting), after spin and gel filtration. The statistical significance of proposed matches are at the > 99% confidence level (chi-squared test), as in the sheep (S3 in S3 File). Matrix and calibration as Fig 2.
Fig 7.
Tryptic digest. sSgII-70 glimpsed in a MALDI MS involving ovine ovarian follicular fluid, anionex Fraction 28 containing Candidate 7500, digested with porcine trypsin.
Positive ion, reflectron mode, delayed extraction, CHCA matrix, high laser power. Prominent are matrix cluster peaks in the m/z 800s and other irrelevancies (e.g., keratin at 1060 & 1134: Mascot Contaminants [19]). Predicted fragments from a tryptic digestion in silico of sSgII-70 find minor peak matches, as per black blocks: (i) Observed m/z 906.3169, Predicted 906.0693, Difference +0.2476 (0.03% above Predicted), sequence (3K)TGEKPVFK(R12), no missed cleavage; (ii) Observed 1278.4464, Predicted 1278.5954, Difference −0.1490 (0.01% below Predicted), 1MLKTGEKPVFK(R12), one missed cleavage; (iii) Observed 1298.0373, Predicted 1294.4123, Difference +3.6250 (0.28% above Predicted), being either (38K)LTGPNNQKHER(A50), as the Predicted, or the MSO version of 1278 (at 1294.5948), both one missed cleavage; and (iv) Observed 2094.4897, Predicted 2091.1458, Difference +3.3439 (0.16% above Predicted), (49R)ADEEQKLYTDDEDDIYK(A67), one missed cleavage. Items (i) and (ii) relate to a proposed reverse splice 9 + 61 junction at 9VF10. A fifth match is moot, as being at higher (but still low) intensity within the matrix ion group: Observed 872.2498, Predicted 871.9682, Difference +0.2816 (0.03% above Predicted), (38K)LTGPNNQK(H47), no missed cleavage. (S3 File, Digest 5.).
Fig 8.
IN = immunoneutralization, IP/LC-MS = immunoprecipitation/liquid chromatography-mass spectrometry, IHC = immunohistochemistry, OV + /OVX– = ovary intact material presenting positively/ovariectomised material presenting negatively.
Fig 9.
Sheep secretogranin II preprohormone.
UniProt W5QEU8, FASTA format [27]. Includes sSgII signal sequence (M1-F30; Courel et al, 2008), which is the ‘pre’ in ‘preprohormone’. sSgII is postulated to give rise to a secreted proteoform, sSgII-70, the product of a 9 + 61 reverse peptide splicing (DB in an A-E modularisation, with residues emphasized). Yet SgII is self-aggregating, admitting the possibility that the 9mer D and 61mer B units of sSgII-70 derive from different SgII molecules.
Fig 10.
In black, N-terminal aa residues M1-K11 and in red C-terminal residues D59-I70. ‘sSgII-14’ (putative Edman N-terminal input) is plotted along the top against EPL001 (actual Edman N-terminal output) vertically. Following the bold line yields the aa sequence of sSgII-14 across the grid, column by column. The faint track reads out EPL001 downwards, row by row. For lysine grid placements see [22]. Green ladders represent potential salt bridges. Overlapping ends are implied.
Fig 11.
sSgII-70 primary structure deduction.
sSgII-70 superscripts are positions in the sSgII master protein (Fig 9). Helix predictions (shaded) by PSIPRED [38] uphold ChemSW [39] modelling in silico (S5 Fig. 1 in S5 File), but only the central prediction is supported by AlphaFold2 [40] (S5 Fig 10 in S5 File). EPL001 residues emphasized are anionex Beale 4. Edman superscripts denote spiralised reading order of residues within sSgII-70 to obtain EPL001. Bracketed Edman items (together with TG, unbracketed) appear thus in Harwell 1 (SEQ ID NO: 6 in S1 Table 1 in S1 File): MKV/IT/G QY S/V G?K?. A quadruple doubleton correspondence between Harwell I and sSgII-70 is unlikely to be due to chance (with Harwell 1 additionally sharing MK & TG with EPL001). Stepped-down gapped doubleton readings connote spirality.