Highly Efficient Separation, Enrichment, and Recovery of Peptides by

Sep 27, 2014 - The extraction is mainly promoted by multisite electrostatic interaction, and the hydrophilic and cationic properties of PEI at low pH ...
0 downloads 0 Views 3MB Size
Download PDF
Article pubs.acs.org/Langmuir

Highly Efficient Separation, Enrichment, and Recovery of Peptides by Silica-Supported Polyethylenimine Feng Chen,† Decheng Wan,*,† Zhihong Chang,† Hongting Pu,†,‡ and Ming Jin† †

Institute of Functional Polymers, School of Materials Science and Engineering, Tongji University, 4800 Cao-an Road, Shanghai 201804, PR China ‡ Key Laboratory of Advanced Civil Engineering Materials, Tongji University, Ministry of Education, Shanghai 201804, China S Supporting Information *

ABSTRACT: Highly efficient and charge-selective adsorption and desorption of peptides at trace level by a solid-phase adsorbent is described. The adsorbent of SiO2@PEI is synthesized by covalent immobilization of branched polyethylenimines (PEI) exclusively on the outer surface of the porous silica particles (∼300 μm). For aqueous peptides (Mw = 600−3000 Da), SiO2@PEI can capture the negatively charged ones and leave the positively charged ones intact, and by adjusting pH of the system peptides with different isoelectric points (pIs) can be well separated. Targeted peptide at low abundance (at least as low as 0.1 mol % with respect to the highest one) can be well separated. The association constants of K > 1012 M−1 at pH > pI and K < 104 M−1 at pH < pI are found; that is, selectivity > 108 is generally available. Thus, a peptide even at sub-femtomolar level can be extracted and eluted for analysis, and efficient recovery (79−92%) of the peptides is found. The extraction is mainly promoted by multisite electrostatic interaction, and the hydrophilic and cationic properties of PEI at low pH play a unique role in desorption efficiency and selectivity. The unbiased nature of this method renders the adsorbent applicable to the efficient separation of a broad spectrum of peptides, including those with similar pIs.

■

INTRODUCTION The separation, purification, and analysis of peptides have attracted much attention.1−5 Peptide extraction and enrichment are favorable for purification of a functional peptide, and also for analysis and profiling of a protein. With many methods currently available for the extraction and separation of peptides and proteins, only a few can deal with a substrate at the trace level. Electrophoresis6−8 or chromatography9,10 are well-known for analysis and separation of a peptide at small-scale operation. Solid phase adsorption is operationally simple at either largescale or small-scale operation but is susceptible to low selectivity and low desorption ratio. Recently, some welldesigned nanoscale adsorbents have been developed to highly selectively capture a specific peptide.3,11−18 For example, Wang and co-workers12,13 fabricated a magnetic nanomaterial allowing highly specific extraction of a phosphopeptide at the femtomolar level from a complicated mixture. Lu et al.16 also developed a technique to highly extract glycopeptides from a mixture by the synergistic role of two types of particles, where the boronic-acid-functionalized magnetic particle is responsible for adsorption of glycopeptides and the hydrophobic and nonmagnetic nanoparticle is responsible for adsorption of the other peptides. However, a general method applicable to the highly efficient separation of a broad spectrum of peptides is still rarely reported. Additionally, a well-designed and nanoscale-sized adsorbent is only tediously available, or else the © XXXX American Chemical Society

selectivity may decrease. More importantly, for a bulky peptide (500 < Mw < 5000 Da), sufficient recovery from an adsorbent is of a general challenge. It is known that the adsorption of a peptide is of multisite or multivalent interaction with an adsorbent, a nature of acceptor−ligand interaction. A quantitative analysis19 pointed out that the acceptor-ligand association constant grows exponentially with the interacting valences. That’s why adsorption is easy while desorption is reluctant to occur where a bulky substrate is involved. Therefore, efficient desorption of a multisite substrate remains a central concern in adsorbent-aided separation. Peptide extraction at the trace level is also favorable for peptide analysis. For example, the matrix-assisted laser desorption/ionization time-of-flight (MALDI-TOF) mass spectrometer has become one of the most important tools for analysis of peptides,20 but it should be better combined with peptide extraction.21−25 MALDI-TOF combines a series of advantages such as (1) allowing the direct analysis of a multicomponent mixture and being able to display the absolute molecular weights of each species present in the sample; (2) allowing a fast and accurate characterization of each species; (3) usually permitting the integrity of the species during measureReceived: May 29, 2014 Revised: September 21, 2014

A

dx.doi.org/10.1021/la502093k | Langmuir XXXX, XXX, XXX−XXX

Langmuir

Article

ment. With these characters, this technique has gained popularity in the analysis of proteins and nucleotides. Regardless of the high detection sensitivity, it is also found that MALDI-TOF is unlikely to detect peptides present in the sample with a low abundance. In this regard, efficient extraction and separation of peptides at low abundance becomes necessary. Branched polyethylenimine (PEI) appears to be unique for the reversible protonation and pH-dependent hydrophilicity. PEI is a spherelike macromolecule, and is known to exist as a cationic electrolyte in water unless pH > 10.26 The repeat unit of PEI is a short hydrophobic ethylene unit, and the aliphatic amino groups are of high content, so PEI becomes rather hydrophilic due to ready protonation at low pH.26 The pHreversible protonation of PEI has found application in separation of small and ionic species. PEI can be readily transformed into a reverse micellelike nanocapsule by, for example, alkylation, where the PEI-based core acts as a cationic microenvironment. The resulting nanocapsules are able to encapsulate a variety of small and anionic guest molecules by liquid−liquid extraction in a water−oil biphasic system, thus leading to facile separation.27−29 However, such a strategy is rarely applicable to a bulky molecule like a peptide mainly due to two factors: (1) the dense shell of the reverse micellelike nanocapsule is favorable to the capture and phase transfer of small guest species, but in case of a bulky species encapsulation becomes difficult due to the kinetic inhibition by the dense shell; and (2) the nanocapsules show a strong propensity to precipitate along the biphasic interface due to aggregation of the amphiphilic core−shell structure unless the shell is dense enough. Alternatively, PEI was impregnated in or covalently immobilized on porous silica, and the resulting material could act as an adsorbent to remove acidic gases such as CO2 and H2S from air.30,31 And recently, Sun et al.32−34 used supported PEI for adsorption and separation of proteins, where the adsorption kinetics and adsorption mechanism was carefully studied. Here we show that, with PEI exclusively attached to the outer surface of a large silica particle, the resulting adsorbent can highly efficiently and highly selectively extract trace peptide from mixed peptides while high desorption ratio is still available; and the hydrophilicity of PEI at low pH appears to be unique for the property.

■

Table 1. Molecular Parameters of Several Peptides peptide

amino acid sequence

pIa

Mwb

ε/104 (L·mol −1 ·cm−1)c

EE-6 ED-9 SP-20

EDAPA E ELRDD EAGD SDEDS DGDRP QASPG LGPGP EDDYE DHDGD DKDHD IDQKD DDDEL PPGFS P QPLSS PPFF VSEIQ LMHNL GKHLN SME PSKPE DNPGA PAEDA RSALR HINLI TQH GFLRR I

3.0 3.4 3.4

630.61 1018.99 1953.95

1.29/1.50 1.32/1.58 2.33/2.90

3.5

3006.82

3.04/3.91

6.0 6.0 6.1

600.67 1019.16 2080.42

4.20/2.92 3.05/2.01 2.88/1.95

6.1

3035.33

4.46/3.02

12.4

760.94

1.29/1.50

EL-25 PP-6 QF-9 VE-18 PH-28 GI-6 a

pIs were calculated using the software ANTHEPROT 5.0. bMolar mass derived from using a mass spectrometer. cMolar extinction coefficient (ε) of EE-6, ED-9, SP-29, and EL-25 were measured at pH = 5 (AC buffer) and at pH = 2 (PB buffer) for adsorption and desorption, respectively, while those of PP-6, QF-9, VE-18, and PH-28 were measured at pH = 7.4 (PB buffer) and at pH = 4 (AC buffer) for adsorption and desorption, respectively. performance liquid chromatography (HPLC) analysis was carried out on an Agilent 1100 RP-HPLC system which included a Waters NovaPak C18 column (4.6 × 250 mm) with a mobile phase at a flow rate of 1 mL/min. A gradient elute composing of 0.1% trifluoroacetic acid in acetonitrile and 0.1% trifluoroacetic acid in water was adopted, with the former gradually increasing to 100% within 25 min. MALDITOF mass spectrometry analysis was performed in positive reflection mode on a 5800 Proteomic Analyzer (Applied Biosystems, Framingham, MA) with a Nd:YAG laser at 355 nm, a repetition rate of 200 Hz, and an acceleration voltage of 20 kV. The range of laser energy was optimized to obtain good resolution and signal-to-noise ratio and kept constant for further analysis. UV/vis spectra were recorded on a Mapada UV-6300 spectrophotometer (Shanghai Mapada Instruments Co. Ltd.). Molar mass of peptides was derived with a Waters Micromass Quattro Premier XE tandem quadruple mass spectrometer equipped with an electrospray ionization (ESI) interface, the main working parameters were set as follows: capillary voltage, 3.00 kV; extractor voltage, 5.00 V; desolvation temperature, 350 °C; desolvation gas flow, 350 L h−1. Dynamic light scattering (DLS) was recorded on Malvern Zetasizer Nano ZS90 instrument equipped with a 4 mW He−Ne gas laser (λ = 633 nm). Measurements were conducted at a fixed scattering angle of 90°, and molecular diameters were calculated from the computed diffusion coefficient using the Stokes−Einstein equation. The CONTIN analysis method was used. Field emission scanning electron microscopy (SEM) was recorded on a Magellan 400 (FEI Co. Ltd) instrument. Synthesis of SiO2@PEI. The synthesis was based on a literature method,31 but to ensure that the PEI is exclusively attached to the outer surface of the silica, an essential modification was adopted. The silica was activated by treating with HCl (1 M) for 36 h, followed by washing with water, ethanol, and acetone in sequence before drying. A mixture of PEI (18.00 g, eq 0.418 mol repeat units of CH2CH2NH) and GPTMS (3.2 mL, 14.72 mmol) in 250 mL of ethanol was prepared and refluxed under nitrogen atmosphere. The mixture was monitored via DLS, and when the size reached 10 nm, the activated silica (45.00 g) was dropped in and the mixture was kept under refluxing and mild stirring for 72 h. The solid was separated by filtration and fully washed with absolute ethanol before being dried in vacuum at 323 K. Elemental Analysis of the organic part: C, 4.98%; H, 1.33%; N, 2.71%. The fraction of PEI on silica was derived from nitrogen content by the equation (43/14) × N% = 8.3%. Saturated Adsorption of Peptides by SiO2@PEI. Typically, to vials, each charged with 7.0 mg of SiO2@PEI, 6 mL of PP-6 (pI = 6.0) at a diversity of concentrations (1 × 10−5−1 × 10−3 M) in a PB buffer

EXPERIMENTAL SECTION

Materials. Silica particles (300 μm) were purchased from Sinopharm Chemical Reagents Company (SCRC, China), and the small-sized fraction was removed by stirring and decanting. 3Glycidyloxypropyltrimethoxysilane (GPTMS) was purchased from Aladdin. Hyperbranched polyethylenimine (PEI, Mn = 1 × 104, Mw/ Mn = 2.5, degree of branch = 60%) was purchased from Aldrich, and the fraction with low molecular weight was removed by dialysis (MWCO 8000−14 000). The peptides of EE-6, ED-9, SP-20, EL-25, PP-6, QF-9, VE-18, PH-28, and GI-6 (Table 1) were purchased from GL Biochem Co. Ltd (Shanghai, China). Other chemicals were used as received except where stated otherwise. Mainly two types of buffers were used unless stated otherwise: the phosphate buffers (PB) and acetate buffer (AC), pH = 2 (NaH2PO4·2H2O, 0.0107 M + H3PO4, 0.0143 M); pH = 4 (NaAc·3H2O, 0.0072 M + HAc, 0.0491 M); pH = 5 (NaAc·3H2O, 0.014 M + HAc, 0.009 M); pH = 7.4 (Na2HPO4· 12H2O, 0.0162 M + NaH2PO4·2H2O, 0.0056 M). Measurements. Thermogravimetric analysis (TGA) was performed using a Netzsch STA 449C thermogravimetric analyzer with air purging and a heating rate of 10 °C/min over the temperature range of ca. ambient to 1000 °C. Elemental analysis was carried out with an Elementar-vario EL III microelemental analyzer. High B

dx.doi.org/10.1021/la502093k | Langmuir XXXX, XXX, XXX−XXX

Langmuir

Article

(pH = 7.4, 0.01 M) was added, and the resulting mixture was subjecting to mild stirring. After sufficient equilibration (typically 14 h), the decanted layer was separated for UV/vis measurement, and a neat buffer was used as a reference. In the case where the absorbance was larger than 1, the solution was diluted to allow the absorbance to fall within 0.10−1.0, and the measured data were amplified to the equivalent concentration. Desorption of Peptides by SiO2@PEI. Typically, SiO2@PEI with adsorbed PP-6 was immersed in an AC buffer (pH = 4, 0.04 M). The release of PP-6 could be monitored with a UV/vis spectrometer, with the neat pH buffer as a reference. The desorption ratio was available when compared with the amount of PP-6 adsorbed. Separation of Peptides with Different pIs by Selective Adsorption of SiO2@PEI. Typically, EE-6 (pI = 3.0) and QF-9 (pI = 6.0) (both at 5.0 × 10−5 M, 6 mL) in a buffered water with pH = 5 (AC buffer, 0.01 M) was prepared, and SiO2@PEI (30 mg) was dropped into the solution with stirring at room temperature. After 14 h, the solid adsorbent was removed and the residual aqueous layer was concentrated by freeze-drying (from about 6 mL to 300 μL) and then further subjected to HPLC or MALDI-TOF analysis. Other mixed peptides could be similarly separated based on their pIs. MALDI-TOF Analysis. An aqueous peptide (5 μL) and the matrix solution (5 μL, α-cyano-4-hydroxycinnamic acid in tetrahydrofuran/ H2O/trifluoroacetic acid at volume ratio of 70:30:0.3) were mixed by keeping α-cyano-4-hydroxycinnamic acid at 4.5 g L−1. An aliquot of 1 μL was spotted on a stainless-steel target, and the solvent was allowed to evaporate for MALDI-TOF analysis. Determination of the Association Constant K. According to the definition, K = [adsorbent − peptide]/([peptide][adsorbent]). Since the activity of the solid adsorbent with or without a captured peptide should be regarded as 1,35 the equation is simplified as K = 1/[peptide]

To favor the desorption, the PEI should better be exclusively attached to the outer surface of the silica; if PEI is also attached to the inner surface of the silica, kinetic inhibition of the release of peptides may occur. Although synthesis of SiO2@PEI was reported recently, PEI was attached to either outer or inner surface of the silica.31 In this route, GPTMS was coupled to PEI by oxirane-amino reaction, and the resulting GPTMS-PEI was further attached to silica. Here, a small but essential modification of this route is adopted: the GPTMS-PEI is allowed to grow large (about 10 nm) via water-promoted oligmerization so that entering the small pores of the silica becomes impossible. The successful attachment of PEI to silica is supported by thermogravimetric analysis and elemental analysis of the fully purified adsorbent. After the functionalization of silica with PEI (Figure s1, Supporting Information), the combustible part makes up about 10% of the total weight. With nitrogen content derived from elemental analysis, it can be further determined that PEI makes up 8.3% of the adsorbent. As mentioned, one of our concerns is the localization of PEI on the porous silica. Treatment of nitrogen gas sorption/desorption isotherms with the Brunauer− Emmett−Teller (BET) equation shows that the silica bears pores with an average diameter of 3.5 nm and the surface area is as large as 437 m2/g. The oligmerization of GPTMS-PEI is monitored with DLS, and when the size is larger than 10 nm, attachment to the silica is further carried out. As shown in Figure 1, for SiO2@PEI, the surface area is significantly reduced to 27.6 m2/g, and the pores of the silica are hardly detectable (beyond the detection limit of the instrument). An optical micrograph shows the SiO2@PEI remains as irregular particles

(1)

where [peptide] represents the residual concentration of a peptide at equilibration.

■

RESULTS AND DISCUSSION Preparation and Characterization of SiO2@PEI. The concept of peptide extraction and desorption (Scheme 1) is based on the different pIs of the peptides, where the adsorbent is PEI supported on large-sized and porous silica (SiO2@PEI). Scheme 1. Schematic Presentation of Peptide Separation, Enrichment, and Desorption Based on pH-Reversible Interaction between the Adsorbent and the pI-Different Peptides

Figure 1. Nitrogen sorption/desorption isotherms (77 K) of the activated silica (A) and SiO2@PEI (B). C

dx.doi.org/10.1021/la502093k | Langmuir XXXX, XXX, XXX−XXX

Langmuir

Article

The adsorption should be mainly attributed to electrostatic interaction. Experimentally, in the case of PP-6 (pI = 6.0), control experiments show that pH > 6 is essential for the adsorption of PP-6 by the adsorbent, as judged from UV/vis monitoring, and it takes about 8 h for equilibration of adsorption at pH 7.4 (PB buffer, Figure s2A, Supporting Information). When PH-28 (pI = 6.1) is used in place of PP-6 for a similar test, it takes about 12 h for equilibration. If the pH of the system is switched to 5, adsorption of EE-6 (pI = 3.0) or EL-25 (pI = 3.5) by the adsorbent is found (Figure s2B, Supporting Information). The other peptides were similarly tested (data not shown), and the equilibration time falls within 6−12 h as well, and the trend is the same: the larger the molecular weight of a peptide is, the longer the time the equilibration takes. Further experiments (data not shown) show that the adsorption rate is independent of the pH value of the system but can be accelerated by stirring, indicating the adsorption is a diffusion-controlled process. With UV/vis spectral monitoring, the absorbing capacity of the adsorbent is measurable. Typically, after adsorption equilibration in a buffer, the residue of the peptide in the solution, if any, can be quantified with a UV/vis spectrometer. As shown in Figure 3A, no UV/vis signal of PP-6 is detectable in the aqueous phase unless the mass ratio of PP-6/adsorbent is above certain value. And from the reflection point shown in Figure 3B, it can be derived that the saturated absorbing capacity is 32.8 mg/g. The adsorption pattern also suggests a Langmuir adsorption mechanism (monolayer molecular adsorption). However, the association constant (K) should be very large that the measuring error does not allow a precise K value to be derived (K can be derived otherwise, as will be mentioned later). Other peptides were similarly tested, and the results are collected in Table 2. The absorbing capacity usually increases with Mw of a peptide (but decreases in terms of molar ratio) per unit of adsorbent, as shown in Table 2. In the view that the adsorption is Langmuir type and an aqueous peptide usually tends to exist in a spherelike shape, this phenomenon seems to be reasonable (refer to Scheme 1, compare B and C). The adsorbent can highly efficiently capture a peptide at pH > pI. To quantify the ability of the adsorbent to capture a peptide, it is necessary to measure the adsorbent−peptide association constant K. Although no precise K can be derived

(Figure 2a). SEM micrograph shows that the silica (Figure 2b), after derivation with GPTMS-PEI, becomes no longer porous (Figure 2c and d), agreeing with the BET measurement.

Figure 2. Micrograph of SiO2@PEI under optical microscope (a) and SEM micrograph of the surface of SiO2 (b) and SiO2@PEI (c, d). Scale bar is 250 μm (a), 300 nm (b and c), and 100 nm (d).

Efficient Adsorption and Desorption of Peptides by SiO2@PEI. The adsorption of an aqueous peptide can be monitored with a UV/vis spectrometer. The synthetic peptides with different pIs as well as other molecular parameters are listed in Table 1. Any of the peptides show UV/vis absorbance only within the region of 190−200 nm, where most chemicals show absorbance as well. However, two factors render the peptide detection possible: (1) upon dropping into the aqueous solution of a peptide, the adsorbent resides exclusively at the bottom of the cuvette and exerts no influence on UV/vis signal due to a peptide, as supported by a control experiment; (2) though the measurement has to be carried out in a pH buffer (due to pH-dependent peptide absorbance), the molar extinction coefficient of peptides is typically in the magnitude of 104 L·mol−1·cm−1 (Table 1), much larger than that of the buffers (either AC buffer or PB buffer, typically in the magnitude of 10−102 L·mol−1·cm−1); thus, when trace buffer is replaced by a peptide, the measuring error is within 1%.

Figure 3. Determination the peptide-adsorbing capacity of SiO2@PEI. (A) UV/vis spectra of residual PP-6 in water after equilibrated adsorption by SiO2@PEI (conditions: 6 mL samples of PP-6 at 1−100 × 10−5 M subjected to adsorption with 7 mg of SiO2@PEI at pH 7.4 (PB buffer), and the aqueous solution is measured after equilibration); (B) Dependence of absorbing capacity of SiO2@PEI on concentration of PP-6 (data derived from (A)). D

dx.doi.org/10.1021/la502093k | Langmuir XXXX, XXX, XXX−XXX

Langmuir

Article

Table 2. Saturated Adsorbing Capacity and Desorption Ratios of SiO2@PEI against Different Peptides adsorptiona c

peptide

Mw (g/mol)

charge

EE-6 ED-9 SP-20 EL-25 PP-6 QF-9 VE-18 PH-28

630.61 1018.99 1953.95 3006.82 600.67 1019.16 2080.42 3035.33

−2.637 −3.518 −3.610 −10.956 −0.005 −0.005 −0.928 −0.928

(mg/g) 45.1 66.2 94.3 75.9 32.8 50.5 77.1 73.2

desorptionb (10

−2

mol/g)

7.16 6.5 4.83 2.52 5.46 4.97 3.71 2.41

charge

c

+0.586 +1.555 +1.545 +4.393 +0.016 +0.016 +2.276 +3.05

desorbed (%) 92.4 91.2 88.7 85.5 89.2 87.4 84.3 78.8

a pH 5 (AC buffer, 0.01M) for EE-6, ED-9, SP-20, and EL-25; pH 7.4 (PB buffer) for PP-6, QF-9, VE-18, and PH-28. bpH 2 for EE-6, ED-9, SP-20, and EL-25 (PB buffer); pH 4 (AC buffer) for PP-6, QF-9, VE-18, and PH-28. cCalculated with ANTHEPROT 5.0 software.

from the Langmuir adsorption pattern, K is measurable otherwise. MALDI-TOF is known for ultrahigh sensitivity, and if adsorption and elution of trace peptide really occurs, it can be monitored with MALDI-TOF. Experimentally, a stock solution containing PP-6, QF-9, VE-18, and PH-28 (any of them with pI around 6.0) in water (pH 7.4, PB buffer) is incubated with the adsorbent, and after equilibration the adsorbent is separated and eluted with water at pH 4 (AC buffer). The eluant is detected by uisng a MALDI-TOF mass spectrometer. It is found that even when a peptide in the stock solution is decreased to 5.0 × 10−13 M, the eluant still show well resolved signals (signal/noise > 85) of any peptide species, as shown in Figure 4. This result shows the high affinity of the

the milliliter level, peptide at the sub-femtomolar level can be extracted. Regardless of the high affinity to the adsorbent, high desorption ratio is available at pH < pI. It is known that a bulky substrate is usually reluctant to be desorbed from an adsorbent due to a strong multivalent complement.19 However, our adsorbent can release most of the adsorbed peptide at pH < pI. The releasing can be monitored with a UV/vis spectrometer as well. As shown in Figure s3 (Supporting Information), the equilibration for releasing several peptides takes a time within 12 h. Other peptides were similarly tested (data not shown), and it was found that the smaller the Mw of a peptide is, the quicker and the more sufficient the release is. The reason is not well understood yet. Most peptides can be well recovered (7992%), especially for those with lower molar mass, as shown in Table 2. A small fraction of peptide molecules are reluctant to be released, which can stem from either thermodynamic or kinetic factor. A control experiment shows, that under similar condition, adsorption hardly occurs, indicating that kinetic factor should be responsible for the incomplete release. Under the condition of pH < pI and [peptide] = 10−4 M, adsorption hardly occurs, and efficient release can occur (Table 2), so an association constant K′ can be similarly estimated from eq 1 as K′ < 104 M−1 for any peptide here tested. Compared with the K at pH > pI, K/K′ > 108 can be derived, ensuring the high selectivity of the adsorbent upon peptide adsorption. Noticing the generality of this strategy, this is the highest selectivity ever reported for common peptides. Interestingly, even in the case that the peptide is poorly charged, reversible sorption is observed. One can notice that, in the case of PP-6 and QF-9, the peptide is almost neutral at pH 7.4 (notice that no precipitation occurs at this pH, as monitored with a UV/vis spectrometer during pH switch), but adsorption still occurs. On the other hand, desorption can readily occur by decreasing the pH of the system. It seems the pH-dependent hydrophilicity of PEI26 should be responsible for this phenomenon. It is known that a peptide, even in an ionic state, is rather hydrophobic, while for PEI it is hydrophobic at high pH and still shows certain affinity to a peptide, but at low pH, being cationic and rather hydrophilic, it shows poor affinity to a neutral or positively charged peptide. Such an excellent ability is further observed in later experiments on peptide separation. In this aspect, this PEI-based adsorbent appears to be unique. Highly Charge-Selective Extraction of Peptides by SiO2@PEI. Since the adsorption is mainly based on electrostatic adsorbent−peptide complement and each peptide has a unique pI, charge-selective extraction (Scheme 1) and separation of

Figure 4. Aqueous peptides at extremely low concentration can be extracted by the adsorbent, as evidenced by MALDI-TOF detection of the eluant from the adsorbent, where signals due to any stock peptide of PP-6, QF-9, VE-18, and PH-28 can be observed. Conditions: aqueous peptides (each at 5.0 × 10−13 M) are subjected to extraction by the adsorbent at pH 7.4 (PB buffer) and followed by elution (with equal volume of eluant) at pH 4 (AC buffer).

adsorbent toward the peptides, and the unbiased nature of the adsorbent to any negatively charged peptide. The association constant K can be derived from eq 1. Since it is certain that the residual peptides are 1012 M−1 can be derived for any of the tested peptides. For a common peptide, this is the highest value ever reported, comparable or better than that for a specific peptide.13,16 The highly strong interaction should be attributed to multisite electrostatic complement. And since the adsorption can be carried out at E

dx.doi.org/10.1021/la502093k | Langmuir XXXX, XXX, XXX−XXX

Langmuir

Article

Figure 5. HPLC detection of aqueous mixtures of (A) EE-6 and QF-9 or (B) PP-6 and GI-6 before (solid line) and after (dotted line) the solidphase extraction with excess SiO2@PEI. Conditions: the pH is 5 (AC buffer) for (A), while it is 7.4 (PB buffer) for (B); all stock solutions of peptides are at 5.0 × 10−5 M, but each is concentrated by freeze-drying by a factor of about 20 before HPLC measurement; gradient elute of 0.1% CF3COOH in CH3CN/0.1% CF3COOH in water with the ratio raising from 18% to 100% within 25 min for (A) and from 15% to 100% within 25 min for (B); X represents the signal due to CF3COOH.

peptides should be possible. In a typical experiment, the aqueous solution of EE-6 (pI = 3.0) and QF-9 (pI = 6.0) was subjected to extraction with excess SiO2@PEI in a buffer (pH = 5, AC buffer). The residual aqueous phase was concentrated for HPLC analysis. For comparison, the stock solution is also similarly concentrated and tested by HPLC. As shown in Figure 5A, the signal due to the negatively charged EE-6 completely disappeared after extraction, indicating a highly efficient separation. To show the versatile separating ability of SiO2@ PEI, another pair of PP-6 (pI = 6.0) and GI-6 (pI = 12.4) was also tested at pH 7.4, and the separation is of high efficiency as well, as supported by HPLC patterns shown in Figure 5B. On the other hand, MALDI-TOF detection (figure not shown) of the eluant shows no signal of QF-9 or GI-6, supporting the high selectivity of the adsorbent. A further test shows that even peptides with close pIs can be well separated. When an aqueous mixture of EE-6 (pI = 3.0) and EL-25 (pI = 3.5) in a buffer (pH = 3.2, phosphate/citric acid buffer) is subjected to extraction with the excess adsorbent, only EE-6 is detectable in the residual solution by MALDITOF, as shown in Figure 6. Therefore, separation based on the fine difference of peptide pIs is feasible. The peptide extraction and enrichment ability of the adsorbent can favor profiling of a digested protein. Regardless of the ultrahigh sensitivity of a MALDI-TOF spectrometer, it has been found that a peptide at low abundance is not detectable along with one at high abundance. As shown in Figure 7A, peptides EE-6, ED-9, SP-20, and EL-25 at low abundance are not detectable along with PP-6 at high abundance (molar ratio at 1:1000). However, after extraction and elution, they are well detectable (Figure 7B). Similarly, when switched to another pH, low abundance of peptide PP-6, QF-9, VE-18, and PH-28 along with GI-6 at high abundance can also be extracted and eluted for detection (Figure 7C and D), indicating the generality of the selectivity. Such ability is favorable for mapping and profiling of proteins, and also for the extraction of a functional peptide. Interestingly, in both cases, the peptide at high abundance is hardly detectable in the eluant, once again supporting the high selectivity of the adsorbent. In summary, the adsorbent of SiO2@PEI, with PEI exclusively populating on the outer surface of silica, can well separate mixed peptides solely based on their pIs through a

Figure 6. EE-6 and EL-25 with similar pIs can be well separated by the adsorbent, where only the signal of EE-6 is detectable by MALDI-TOF after extraction. Conditions: extracted with excess SiO2@PEI at pH = 3.2 (phosphate/citric acid buffer, 0.01M) and the residual solution is subjected to MALDI-TOF test.

conventional procedure of incubation−separation−elution. The uniqueness of this adsorbent lies on the highly efficient capture of a peptide, high desorption ratio, and high peptide selectivity. The PEI usually exists as a cationic electrolyte (unless pH > 10), while peptides can be switched to either the anionic state for adsorption or cationic state for desorption. The association constant for adsorption is K > 1012 M−1, rendering highly efficient extraction of peptide at sub-femtomolar levels possible. The desorption ratio is as high as 79−92%. The selectivity is larger than 108 that peptides at low abundance (e.g., at 1:1000 molar ratio) can be well extracted, as detected by MALDI-TOF analysis. The unbiased extraction nature of this material renders it a general yet highly efficient adsorbent for comprehensive peptide analysis, and this method is applicable for the extraction of a functional peptide. The cationic and high hydrophilic properties of PEI at low pH should be responsible for the high selectivity and high desorption ratio. F

dx.doi.org/10.1021/la502093k | Langmuir XXXX, XXX, XXX−XXX

Langmuir

Article

Figure 7. Low abundance peptides of EE-6, ED-9, SP-20, and EL-25 (each at 4.0 × 10−7 M) along with high abundance peptide of PP-6 (4.0 × 10−4 M) are not detectable by MALDI-TOF mass spectra (A), but after incubation−separation−elution treatment, they are well detectable (B); similarly, low abundance of PP-6, QF-9, VE-18, and PH-28 (each at 4.0 × 10−7 M) along with high abundance GI-6 (4.0 × 10−4 M) are not detectable (C), but after incubation−separation−elution treatment, they are well detectable (D). Conditions: with excess adsorbent of SiO2@PEI; incubation at pH 5 (AC buffer) and recovered by elution at pH 2 (PB buffer) for (B); and incubation at pH 7.4 (PB buffer) and recovered at pH 4 (AC buffer) for (D).

■

■

ASSOCIATED CONTENT

S Supporting Information *

(1) Syka, J. E. P.; Coon, J. J.; Schroeder, M. J.; Shabanowitz, J.; Hunt, D. F. Peptide and Protein Sequence Analysis by Electron Transfer Dissociation Mass Spectrometry. Proc. Natl. Acad. Sci. U. S. A. 2004, 101, 9528−9533. (2) Link, A. J.; Eng, J.; Schieltz, D. M.; Carmack, E.; Mize, G. J.; Morris, D. R.; Garvik, B. M.; Yates, J. R. Direct Analysis of Protein Complexes Using Mass Spectrometry. Nat. Biotechnol. 1999, 17, 676− 682. (3) Oda, Y.; Nagasu, T.; Chait, B. T. Enrichment Analysis of Phosphorylated Proteins as a Tool for Probing the Phosphoproteome. Nat. Biotechnol. 2001, 19, 379−382. (4) Wolters, D. A.; Washburn, M. P.; Yates, J. R. An Automated Multidimensional Protein Identification Technology for Shotgun Proteomics. Anal. Chem. 2001, 73, 5683−5690. (5) Gomez-Escudero, A.; Azagarsamy, M. A.; Theddu, N.; Vachet, R. W.; Thayumanavan, S. Selective Peptide Binding Using Facially Amphiphilic Dendrimers. J. Am. Chem. Soc. 2008, 130, 11156−11163. (6) Gorg, A.; Weiss, W.; Dunn, M. J. Current Two-Dimensional Electrophoresis Technology for Proteomics. Proteomics 2004, 4, 3665−3685. (7) Chu, Y. H.; Dunayevsiiy, Y. M.; Kirby, D. P.; Vouros, P.; Karger, B. L. Diffusion-Edited NMR−Affinity NMR for Direct Observation of Molecular Interactions. J. Am. Chem. Soc. 1996, 118, 7827−7835. (8) Yoshida, T. Hydrophilic Interaction Chromatography. J. Biochem. Biophys. Methods 2004, 60, 265−280. (9) Opiteck, G. J.; Jorgenson, J. W. Two-Dimensional SEC/RPLC Coupled to Mass Spectrometry for the Analysis of Peptides. Anal. Chem. 1997, 69, 2283−2291.

Thermogravimetric analysis of activated SiO2 (a) and SiO2@ PEI; adsorption kinetics of SiO2@PEI against peptides of PP-6 or PH-28 at pH = 7.4 (PB buffer) and against EE-6 or EL-25 at pH = 5 (AC buffer); desorption kinetics of adsorbent against peptide EE-6 and EL-25 at pH = 2 (PB buffer) and against PP6 and PH-28 at pH = 4 (AC buffer). This material is available free of charge via the Internet at http://pubs.acs.org.

■

REFERENCES

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. Author Contributions

The manuscript was written through contributions of all authors. All authors have given approval to the final version of the manuscript. Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS The Natural Science Foundation of China (No. 21074094 and 51273149), the Open Measuring Fund for Large Instrument and Equipment, Tongji University (No. 0002013005) support this work. G

dx.doi.org/10.1021/la502093k | Langmuir XXXX, XXX, XXX−XXX

Langmuir

Article

(10) Nuhse, T. S.; Stensballe, A.; Jensen, O. N.; Peck, S. C. Largescale Analysis of in Vivo Phosphorylated Membrane Proteins by Immobilized Metal Ion Affinity Chromatography and Mass Spectrometry. Mol. Cell. Proteomics 2003, 2, 1234−1243. (11) Cerny, M.; Skalak, J.; Cerna, H.; Brzobohaty, B. Advances in Purification and Separation of Posttranslationally Modified Proteins. J. Proteomics 2013, 92, 2−27. (12) Ma, W. F.; Zhang, Y.; Li, L. L.; Lu, H. J.; Wang, C. C. Ti4+Immobilized Magnetic Composite Microspheres for Highly Selective Enrichment of Phosphopeptides. Adv. Funct. Mater. 2013, 23, 107− 115. (13) Zhang, Y. T.; Li, L. L.; Ma, W. F.; Zhang, Y.; Yu, M.; Guo, J.; Lu, H. J.; Wang, C. C. Two-in-One Strategy for Effective Enrichment of Phosphopeptides Using Magnetic Mesoporousγ-Fe2O3 Nanocrystal Clusters. ACS Appl. Mater. Interfaces 2013, 5, 614−621. (14) Najam-ul-Haq, M.; Jabeen, A. S. F.; Maya, F.; Ashiq, M. N.; Sharif, A. Newly Developed Poly(Allyl Glycidyl Ether/Divinyl Benzene) Polymer for Phosphopeptides Enrichment and Desalting of Biofluids. ACS Appl. Mater. Interfaces 2014, 6, 3536−3545. (15) Cheng, G.; Wang, Z. G.; Liu, Y. L.; Liu, Y. L.; Zhang, J. L.; Sun, D. H.; Ni, J. Z. Magnetic Affinity Microspheres with Meso-/ Macroporous Shells for Selective Enrichment and Fast Separation of Phosphorylated Biomolecules. ACS Appl. Mater. Interfaces 2013, 5, 3182−3190. (16) Wang, Y. L.; Liu, M. B.; Xie, L. Q.; Fang, C. Y.; Xiong, H. M.; Lu, H. J. Highly Efficient Enrichment Method for Glycopeptide Analyses: Using Specific and Nonspecific Nanoparticles Synergistically. Anal. Chem. 2014, 86, 2057−2064. (17) Xu, Y. W.; Wu, Z. X.; Zhang, L. J.; Lu, H. J.; Yang, P. Y.; Webley, P. A.; Zhao, D. Y. Highly Specific Enrichment of Glycopeptides Using Boronic Acid-Functionalized Mesoporous Silica. Anal. Chem. 2008, 81, 503−508. (18) Li, H.; Shan, Y. H.; Qiao, L. Z.; Dou, A. B.; Shi, X. Z.; Xu, G. W. Facile Synthesis of Boronate-Decorated Polyethyleneimine-Grafted Hybrid Magnetic Nanoparticles for the Highly Selective Enrichment of Modified Nucleosides and Ribosylated Metabolites. Anal. Chem. 2013, 85, 11585−11592. (19) Mammen, M.; Choi, S. K.; Whitesides, G. M. Polyvalent Interactions in Biological Systems: Implications for Design and Use of Multivalent Ligands and Inhibitors. Angew. Chem., Int. Ed. 1998, 37, 2754−2794. (20) Aebersold, R.; Goodlett, D. R. Mass Spectrometry in Proteomics. Chem. Rev. 2001, 101, 269−295. (21) Wong, V. N.; Fernando, G.; Wagner, A. R.; Zhang, J.; Kinsel, G.R.; Zaucher, S.; Dyer, D. J. Separation of Peptides with Polyionic Nanosponges for MALDI-MS Analysis. Langmuir 2009, 25, 1459− 1465. (22) Mitrovic, B.; Eastwood, S.; Wong, V.; Dyer, D.; Kinsel, G.; Scott, C. Peptide/Protein Separation with Cationic Polymer Brush Nanosponges for MALDI-MS Analysis. Langmuir 2013, 29, 696−700. (23) Rodthongkum, N.; Chen, Y.; Thayumanavan, S.; Vachet, R. W. Selective Enrichment and Analysis of Acidic Peptides and Proteins Using Polymeric Reverse Micelles and MALDI-MS. Anal. Chem. 2010, 82, 8686−8691. (24) Rodthongkum, N.; Chen, Y.; Thayumanavan, S.; Vachet, R. W. Matrix-Assisted Laser Desorption Ionization-Mass Spectrometry Signal Enhancement of Peptides after Selective Extraction with Polymeric Reverse Micelles. Anal. Chem. 2010, 82, 3686−3691. (25) Zhao, M.; Deng, C.; Zhang, X.; Yang, P. Facile Synthesis of Magnetic Metal Organic Frameworks for the Enrichment of LowAbundance Peptides for MALDI-TOF MS. Proteomics 2013, 13, 3387−3392. (26) Borkovec, M.; Kiper, G. J. M. Proton Binding Characteristics of Branched Polyelectrolytes. Macromolecules 1997, 30, 2151−2158. (27) Wan, D. C.; Pu, H. T.; Cai, X. Y. Separation Promoted by Molecular Recognition of a Core Engineered Macromolecular Nanocapsule. Macromolecules 2008, 41, 7787−7789.

(28) Wan, D. C.; Wang, G. C.; Pu, H. T.; Jin, M. Can Nonspecific Host−Guest Interaction Lead to Highly Specific Encaspulation by a Supramolecular Nanocapsule? Macromolecules 2009, 42, 6448−6456. (29) Wan, D. C.; Yuan, J. J.; Pu, H. T. Macromolecular Nancapsule Derived from Hyperbranched Polyethylenimine (HPEI): Mechanism of Guest Encapsulation versus Molecular Parameters. Macromolecules 2009, 42, 1533−1540. (30) Wang, X. X.; Ma, X. L.; Sun, L.; Song, C. S. A Nanoporous Polymeric Sorbent for Deep Removal of H2S from Gas Mixtures for Hydrogen Purification. Green Chem. 2007, 9, 695−702. (31) Kassab, H.; Maksoud, M.; Aguado, S.; Pera-Titus, M.; Albela, B.; Bonneviot, L. Polyethylenimine Covalently Grafted on Mesostructured Porous Silica for CO2 Capture. RSC Adv. 2012, 2, 2508−2516. (32) Yu, L. L.; Tao, S. P.; Dong, X. Y.; Sun, Y. Protein Adsorption to Poly(ethylenimine)-Modified Sepharose FF: I. A Critical Ionic Capacity for Drastically Enhanced Capacity and Uptake Kinetics. J. Chromatogr. A 2013, 1305, 76−84. (33) Yu, L. L.; Sun, Y. Protein Adsorption to Poly(ethylenimine)Modified Sepharose FF: II. Effect of ionic strength. J. Chromatogr. A 2013, 1305, 85−93. (34) Hong, Y.; Liu, N.; Wei, W.; Yu, L. L.; Ma, G. H.; Sun, Y. Protein Adsorption to Poly(ethylenimine)-modified Sepharose FF: III. Comparison between Different Proteins. J. Chromatogr. A 2014, 1342, 30−36. (35) Arkas, M.; Tsiourvas, D.; Paleos, C. M. Functional Dendrimeric “Nanosponges” for the Removal of Polycyclic Aromatic Hydrocarbons from Water. Chem. Mater. 2003, 15, 2844−2847.

H

dx.doi.org/10.1021/la502093k | Langmuir XXXX, XXX, XXX−XXX