Reversed-phase high-performance liquid chromatography of mouse epidermal growth factor and its congeners: mobile phase optimization with ion-pairing additives

The design of analytical methods for use in topical epidermal growth factor product development

Data from a reverse phase gradient elution HPLC assay for human epidermal growth factor (EGF) was compared and correlated with data obtained from a competitive heterogeneous radioimmunoassay (RIA). The RIA was established to measure very low concentrations of EGF in formulation drug release and compatibility studies. The HPLC assay, capable of resolving parent and possible modified or fragmented forms of EGF isolated from human urine, was studied as a potential development tool for stability and final product evaluation. As independent analytical methods, the HPLC and the RIA procedures produced correlated results when quantifying freshly prepared and certain degraded samples of EGF. The capacity of the HPLC method to serve as a stability indicating assay was examined. Degradation of EGF was induced by storage in 0.05 M phosphate buffer pH 7.4 at 25, 37 or 50 degrees C. The same three degradation products were detected at each temperature by the HPLC method, one of which was identified as L-isoaspartyl EGF.

Purification and characterization of three molecular forms of insulin-like growth factor II from human Cohn paste IV

The somatomedins or insulin-like growth factors (IGFs) are a family of peptides present in human serum. They are bound to specific carrier proteins and are thought to mediate growth-promoting actions of human growth hormone. Starting from Cohn fraction IV of human plasma, we describe here a rapid and highly efficient procedure for the purification to homogeneity, in addition to IGF I. of three forms of insulin-like growth factor II: IGF IIA (10-12 kDa), IGF IIB (the "classical" 7.5 kDa IGF II) and IGF IIC, identified as the IGF II variant of Jansen by fast-atom bombardment mass spectrometry. The procedure is based on ion-exchange chromatography and gel permeation chromatography on Biogel P10. As judged by specific radioimmunoassay methods for IGF I and IGF II, one of the most striking advantages of this process at this stage is the yield of IGF I not contaminated by 7.5 kDa IGF II. Isoelectric focusing or chromatofocusing, which require affinity chromatography to separate proteins from the polybuffers, are not necessary in this procedure. Final purification was directly achieved by preparative, followed by analytical high-performance liquid chromatography. The N-terminal sequence of peptide IGF IIB (39 amino acids) and peptide IGF I (29 amino acids) showed total homology with those previously described by Rinderknecht and Humbel [FEBS Lett., 89 (1978) 283]. The final yields of purified human IGF I and IGF IIB were 15 and 25 micrograms, respectively, from 11 of serum. All peptides interact with specific receptors on human lymphocytes and red blood cells, and are biologically active (stimulation of 35S uptake, increasing [3H]thymidine incorporation in human and chick embryo fibroblasts).

Isolation of multiple biologically and chemically diverse species of epidermal growth factor

We have analyzed several lots of epidermal growth factor (EGF) purified from murine submaxillary glands including "receptor grade" EGF from Collaborative Research and EGF from Boehringer Mannheim Biochemicals. New England Nuclear uses "receptor grade" EGF to produce 125I-labeled EGF. Though these reagents are reported to be homogeneous, we found them to be a mixture of six species. A method was developed to separate this mixture into its component parts. The individual components were chemically characterized and tested for biological potency. N-terminal sequence analysis of the unfractionated EGF-mixture reveals three different sequences starting with residues 1, 2, or 3 of the mature peptide. Each component exhibited different degrees of mitogenic and EGF receptor binding activity indicating that the N-terminal region contributes to the biological response. The species representing the complete EGF peptide is the most active species in all biological assays. A rapid method for purification of homogeneous complete EGF from commercial EGF preparations is described.

Evaluation of advanced silica packings for the separation of biopolymers by high-performance liquid chromatography. III. Retention and selectivity of proteins and peptides in gradient elution on non-porous monodisperse 1.5-microns reversed-phase silicas

Following previous studies of the use of non-porous monodisperse 1.5-microns n-octyl- and n-octadecyl-bonded silicas in gradient elution of proteins, this work was aimed at elucidating further the properties of this novel column material for peptide and protein separations in comparison with wide-pore silicas. First, it is demonstrated that with short columns (e.g., 35 X 8 mm I.D.) packed with these non-porous reversed-phase materials, mixtures of small peptides and mixtures of proteins can be very efficiently resolved. When the chain length of the bonded ligand was varied, the retention of a test set of proteins in gradient elution followed the ligand sequence C18 greater than C8 approximately C4 approximately phenyl greater than C2 under constant elution conditions, and the selectivity remained unchanged. Comparison of the S values of these proteins, as determined from evaluation of the log k' vs. phi dependences with non-porous silicas and with a LiChrospher Si 1000 C8 with identical accessible ligand surface areas per unit column volume, indicated lower values for the non-porous materials (k' = capacity factor; phi = molar fraction of organic solvent; S = slope of the plot of log k' vs. phi). The origin of this behaviour is discussed.

High-performance liquid chromatography of amino acids, peptides and proteins. LXVII. Evaluation of bandwidth relationships of peptides related to human beta-endorphin, separated by gradient-elution reversed-phase high-performance liquid chromatography

The bandwidths of several polypeptides related to human beta-endorphin have been investigated with different n-alkylsilica stationary phases and different elution gradients of 0.1% trifluoroacetic acid-water-acetonitrile mobile phases. In particular, we have examined the influence of changes in the gradient steepness parameter, b, on peakwidth with five different octadecylphases, chemically bonded to porous spherical silica particles of nominally 4 micron and 6 micron average particle diameter, respectively. The effect on the zone dispersion of these polypeptide solutes as the average pore diameter of the silica matrix was increased from 7.3 nm to 30 nm with stationary phases of similar ligand densities packed into columns of identical configuration has been further documented. The experimental data on solute bandwidths and peak capacities are comparable with the corresponding bandwidth and peak capacity values calculated from analytical equations, derived from the general plate height theory and from gradient elution theory. These comparisons clearly demonstrate that anomalous bandbroadening phenomena may occur when polypeptides are eluted with steep gradients, i.e. with gradients of large b values. Moreover, as the relative chromatographic residence times of beta-endorphin peptides capable of forming a C-terminal amphiphilic secondary structures is increased, i.e. as the dwell times and median capacity factors, k, for such peptides are increased, significant divergencies arise between the observed peakwidth behaviour and the behaviour predicted by analytical relationships which describe either the dependency of peak bandwidth (as 4 sigma v) on the gradient steepness parameter, b, or the dependency of peak capacity on gradient time, tG, median capacity factor, k, and the Knox parameter, C, respectively. The importance of these divergences from the predicted bandwidth and peak capacity behaviour for polypeptides separated on reversed phases, and for resolution optimisation in particular, is evaluated. These investigations thus enable further assessment of the quantitative relevance of current models that describe polypeptide zone migration under gradient elution reversed-phase chromatographic conditions in which solute-dependent slow equilibria, mediated by conformational or solvation effects, may still occur.

Rat epidermal growth factor: complete amino acid sequence. Homology with the corresponding murine and human proteins; isolation of a form truncated at both ends with full in vitro biological activity

Epidermal growth factor (EGF) isolated from the submaxillary gland of the rat (rEGF) is missing the COOH-terminal five residues present in both mouse and human EGF. rEGF competes for the binding of 125I-labelled mEGF to human carcinoma cells with the same affinity as mEGF. rEGF and mEGF have identical mitogenic activities on mouse 3T3 fibroblasts, thus the C-terminal region of the sequence is not necessary for the in vitro activity of EGF. Using reversed-phase high-performance liquid chromatography, four molecular forms of EGF have been extracted from rat submaxillary glands. These forms represent rEGF, rEGF(2-48), rEGF(3-48) and rEGF(4-48); all forms appear to be equipotent in both the receptor binding and mitogenic assays. The isoelectric points of these rEGFs are in the range of pH 5.1 to 5.2. The primary structure of rEGF was determined from approximately 10 micrograms protein by sequence analysis of the intact molecule and fragments obtained from the reduced and alkylated protein by chemical cleavage with CNBr and enzymic cleavage with chymotrypsin and a proline-specific endopeptidase. Subnanomole amounts of generated peptides were purified to homogeneity by reversed-phase microbore high-performance liquid chromatography and analysed by automated Edman degradation in a gas-phase sequencer. There are 48 amino acid residues in the complete polypeptide chain which lacks alanine, phenylalanine, lysine and tryptophan. The amino acid sequence of rat epidermal growth factor is: Asn-Ser-Asn-Thr-Gly-Cys-Pro-Pro-Ser-Tyr-Asp-Gly-Tyr-Cys-Leu-Asn- Gly-Gly-Val-Cys-Met-Tyr-Val-Glu-Ser-Val-Asp-Arg-Tyr-Val-Cys-Asn-Cys -Val-Ile-Gly-Tyr-Ile-Gly-Glu-Arg-Cys-Gln-His-Arg-Asp-Leu-Arg. The calculated relative molecular mass from the sequence analysis is 5377.

Mass spectral analysis of murine epidermal growth factor

Fast atom bombardment mass spectrometry has been used to characterize epidermal growth factor isolated from mouse submaxillary glands. The preparation is found to consist of two peptides, one of which has the average molecular weight predicted for the familiar gene product. The molecular weight of the second component is found to be reduced by the mass of one asparagine residue. These observations are discussed in light of previous reports of heterogeneity.