|Home | About | Journals | Submit | Contact Us | Français|
Synthetic peptides become more and more important as drug candidates in the treatment of a variety of diseases. A particular therapeutic focus for synthetic peptides is cancer treatment.1,2 In order to keep pace with the growing number of newly synthesized peptides, peptide purification should not represent the bottleneck in the drug discovery process. Since the target masses of synthetic peptides are well known, mass-based fraction collection represents an efficient technique for their purification. In contrast to fraction triggering with less specific detectors, employing a mass selective detector leads in each run only to the purification of the target mass. Consequently, it is not necessary to pick the compound of interest out of a series of redundant fractions. In this article we demonstrate mass-based purification of a variety of crude synthetic peptides by reversed phase high-performance liquid chromatography. The peptides were in the mass range from less than 1 kDa to more than 10 kDa and covered a pI range from 4 to 13. We particularly focused on some technical aspects of the system that were prerequisite for reliable compound purification with high recoveries.
Today, solid phase peptide synthesis is the method of choice for large-scale peptide production.3,4 In order to remove scavengers and reaction by-products a subsequent purification step is indispensable. Usually preparative reversed-phase high-performance liquid chromatography (HPLC) is applied to purify crude peptides after synthesis.5,6 In such an approach, target peptides are fractionated UV-based and their identity and purity is determined off-line by mass spectrometry (MS), e.g., matrix-assisted laser desorption/ionization time-of-flight mass spectrometry.7 However, this approach has some weak points: Since UV detection is not selective in each chromatographic run, several peaks might be collected and the target compound has to be picked out of a series of redundant fractions (Fig. 11,, upper panel). Moreover, fractions are collected even though synthesis might not have been successful. After chromatography some preparation steps are required for MS investigation of the fractions. In summary, UV-based fraction collection with subsequent off-line MS-analysis is time consuming, wastes valuable resources in the fraction collector, and does not provide any on-line MS information.
Another very efficient approach for the purification of compounds with well-known masses is mass-based fraction collection. The benefits of this technique are illustrated in Figure 11.. In contrast to a UV detector, an electrospray ionization mass spectrometer (ESI/MS) is employed to observe the occurrence of specific masses during chromatographic runs. When those masses are detected, fraction collection is triggered automatically. As a consequence only the compounds of interest—the target compounds—are collected (Fig. 11,, lower panel). Since the masses of the target molecules are known in peptide synthesis, each crude peptide sample ends up in only one fraction containing the purified compound.
HPLC-grade acetonitrile (Merck, Darmstadt, Germany), water (Millipore, Bedford, MA), and HPLC-grade trifluoracetic acid (TFA) and formic acid (Sigma-Aldrich, Milwaukee,) were used in the preparation of the mobile phases. Lysozyme was purchased from Sigma-Aldrich. For filtering of aqueous sample solutions 0.45-μm filters (Schleicher & Schuell, Dassel, Germany) were used. The HPLC system was fully composed from Agilent 1100 Series modules and controlled by ChemStation A.09.03 and the Purification/HighThruput software A.01.02 (Waldbronn, Germany). Preloaded Wang resins were obtained from Calbiochem-Novabiochem, UK. Protected amino acids, solvents, and HBTU [2-(1H-benzotriazol-1-yl)-1,1,3,3-tetramethyluronium hexafluorophosphate] were obtained from Applied Biosystems, UK. Other reagents were obtained from Sigma-Aldrich.
Twelve different peptides covering a pI range from 4.3 to 12.5 with molecular weights between 1 and 10 kDa were synthesized. Sequences and pI values of the all peptides are given in Table 11.
The peptides were synthesised on a Model 433A Applied Biosystems Solid Phase Synthesiser on preloaded Wang resins using basic feedback monitoring cycles and HBTU as a coupling reagent. Temporary α-amino group protection was achieved using 9-fluorenylmethyloxycarbonyl which was removed using piperidine.
Cleavage from the resin and deprotection of the peptides were achieved by treating the peptidyl-resin with 10 mL of a mixture containing 9.25 mL TFA, 0.25 mL ethanedithiol, 0.25 mL triisopropylsilane and 0.25 mL water at 20°C for 2 h. The peptide was precipitated using ice-cold diethylether and then filtered on a fine sintered glass filter funnel under light vacuum. The peptide precipitate was dissolved in 10% acetic acid/water solution and lyophilized.
Appropriate solvent conditions (water, acetonitrile, TFA) for each sample were chosen according to pI and hydrophobicity of the corresponding peptides. The final peptide concentrations were between 5 and 10 mg/mL. The lyophilized samples were then re-dissolved and filtered prior to chromatography. Between 200 μL and 5 mL of these sample solutions were injected into the system for each purification run.
The purification system was set up as illustrated in Figure 22 comprising two preparative pumps, preparative autosampler, diode array detector (with 10-mm flow cell for analytical scale purification and re-analysis and 0.06-mm flow cell for preparative scale purification), preparative scale fraction collector, isocratic pump, and LC/MSD equipped with an ESI interface and active splitter.
One of the two flow paths, the main flow, led from the preparative pumps to the autosampler, the detector and finally to the fraction collector. Since the mass selective detector (MSD) is a destructive detector and the flow rate of the main flow is too high to route it directly into the electrospray source, a make-up flow was sustained by the isocratic pump. This make-up flow led from the isocratic pump directly to the MSD. The Agilent active splitter connected the two flow paths. In principle, this splitter functions by transferring aliquots from the main flow into the make-up flow.8 The desired split ratio could be easily adjusted through the software and optimized in order to prevent the MSD from overloading but still obtaining signals with a good signal to noise. In contrast to a traditional passive splitter design, an active splitter offers several advantages: It adds almost no additional backpressure and delay volume to the system and sustains a constant split despite changes in temperature and mobile phase composition.
Another benefit of this system setup is illustrated in Figure 33.. Here the effect of TFA on MS signal intensity becomes clearly visible: Figure 3A3A shows the tremendous suppression of ion formation in the MS caused by TFA and the poor signal intensities that result. On the other hand, TFA is known to be a suitable ion-pairing agent for the separation of peptides by RP-HPLC and is therefore preferred by many labs for synthetic peptide purification. In order to take advantage of the superior separating capabilities associated with TFA and to overcome its limitations regarding ion formation, different mobile phases for main-flow and make-up flow were chosen. Since the active splitter transfers only a small portion of the main-flow to the make-up flow, remaining TFA in this aliquot becomes immediately diluted. As a result, TFA does not impact ion formation in the MS. A great improvement in signal intensity is seen in Figure 3B3B.. Here, 0.1% formic acid was added to the make-up flow, since it was demonstrated in a recent study to provide superior MS signal intensities.9
Mass-based fraction collection was either triggered on the singly charged state of the target peptide or on the most abundant multiply charged state. In order to yield a pure product for peak triggering, the MS signal was connected to a UV-signal via a logical and. In particular, this means that a peak was only collected when the predefined trigger conditions regarding threshold and slope for both detectors were fulfilled. In order to avoid collection of invalid peaks due to fluctuations in the baselines of both trigger signals, threshold values for both signals were specified.
Mass-based fraction collection was performed at 1 (analytical scale, split ratio 100:1) and 25 mL/min (preparative scale, split ratio 20,000:1), respectively. Zorbax SB300-C8 4.6 × 250, 5 μM and 21.2 × 250, 7 μM columns were used for peptide separation. For the mobile phases, water and acetonitrile each containing 0.1% TFA were used. The solvent for the make-up flow was composed of 75% water, 25% acetonitrile, and 0.08% formic acid. After sample injection the column was washed for 2 min with 30% organic phase to remove unbound components. For peptide elution the fraction of organic phase was linearly increased to 70% within 8 min. Re-analyses of the purified peptides were performed at 1 mL/min choosing a split ratio of 50:1.
Mass spectrometry detection was performed under the same conditions for analytical and preparative scale: positive polarity, fragmentor voltage 70V, drying gas flow 11 L/min, drying gas temperature 350°C, and nebulizer pressure 40 psig. Mass spectra were registered in full-scan mode (m/z 400 to 1800, step size 0.1).
In order to take into account the delay times between the detectors and the fraction collector, delay volume calibrations were performed at 1 mL and 25 mL. The precise determination of these delay times is of particular importance for a reliable fraction collection process with high recoveries. With the instrument used for our experiments, such a delay volume calibration could be performed fully automated.10
Figure 44 shows the chromatogram of a preparative scale purification of an acidic (pI 4.3) 56-amino acid 6021-Da peptide (Table 11).). Fraction collection was triggered on the UV signal at 214 nm with a threshold of 15 mAU and an m/z of 1205.3 (the total ion current [TIC] chromatogram is shown in the inset). The vertical lines indicate the beginning and end of fraction collection. Additionally, the position of the collected fraction in the fraction collector and the target mass are given. The horizontal line visualizes the UV threshold value for fraction collection. All conditions for fraction collection were fulfilled as soon as the UV-signal at 214 nm exceeded the threshold value of 15 mAU and the target m/z of 1205.3 was detected by the MSD.
Figure 55 shows the signals that were generated by the MSD. The grey curve is the TIC chromatogram which is a superposition of all ions being detected by the MSD in the specified range. In contrast, the extracted ion chromatogram (black curve) just shows the signal that is solely produced by the target m/z of 1205.3. Figure 55 clearly points out that start and stop of fraction collection correspond precisely to the occurrence of the target compound.
After purification the collected fractions were re-analyzed. Figure 66 shows the TIC chromatogram of the re-analyzed purified peptide. The insert shows the ESI mass spectrum that has been extracted from the TIC chromatogram. Deconvolution of the mass spectrum results in a mass of 6021 Da. The chromatogram and the mass spectrum indicate that purification was successful. For all peptides, recoveries between 90% and 98% could be achieved.
Mass-based fraction collection is the logical choice for compound purifications when the target masses are known. In addition to the purification of combinatorial chemistry libraries,11 mass-based fraction collection is also an elegant method for the purification of synthetic peptides. We were able to demonstrate that this technique can cope with a wide variety of peptides. Hydrophobic and hydrophilic peptides that differ in pI between 4 and 13 and cover a mass range from 1 to more than 10 kDa were successfully purified. Furthermore, we found that fraction collection triggered by predefined masses is advantageous over conventional, less-specific detectors. Since only the compounds of interest were collected in each run, no additional time was spent picking target compounds out of a series of redundant fractions. Mass-based fraction collection therefore is highly efficient and saves valuable resources. Furthermore, when coupling ESI-MS to HPLC for data evaluation, no additional preparation steps were needed as is the case for off-line MS techniques. MS information can easily be analyzed on-line during a chromatographic run. Peptide characterization with an ESI-MSD therefore is an alternative to off-line MS methods, e.g., matrix-assisted laser desorption/ionization time-of-flight mass spectrometry.