Introduction

Reversible phosphorylation of proteins is one of the pervasive and paramount post-translational modifications (PTM) [1]. Phosphorylation serves as irreplaceable part in vital movement processes, such as intercellular signaling, transduction, and neural activity [2, 3]. Some physiological abnormalities and pathologies are thought to be associated with abnormal phosphorylation, and some phosphorylated proteins have become biomarkers of several kinds of diseases [4, 5]. Therefore, understanding the reaction mechanism of protein phosphorylation and phosphorylation sites has a non-negligible effect on understanding the life activities of organisms. Mass spectrometry has become an irreplaceable tool in proteomics/polypeptide omics analysis due to its high sensitivity and resolution. However, due to the influence of the phosphoric acid group, the ionization efficiency of phosphopeptides is poorer than that of non-phosphopeptides, and the abundance of phosphopeptides is lower. Therefore, there is considerable impediment to the direct analysis of actual biological samples by mass spectrometry without pretreatment. Various unfavorable factors require us to develop an efficient method for isolating phosphopeptides from complex biological samples [6, 7].

To date, researchers have made much effort to developing strategies for enrichment of phosphopeptides, including immunoprecipitation [8], reverse liquid chromatography [9], ion-exchange chromatography (IEC) [10, 11], and affinity chromatography [12,13,14]. Affinity chromatography-immobilized metal ion affinity chromatography (IMAC) has become a prevalent strategy due to the simple operation steps along with the diversity of metal cations and chelating ligands [15,16,17,18,19,20,21]. The enrichment mechanism is based on a coordination reaction between metal ions and phosphate groups, so the phosphopeptides can be captured under acidic conditions and eluted under alkaline conditions [22]. Therefore, many high-valent metal ions have been explored and used for the phosphorylated protein separation and enrichment, and most of them have achieved good results. Up to now, there are more than 10 types of metal ions that have been developed for phosphopeptide detection, such as Fe3+, Ti4+, Cu2+, Ni3+, Zr4+, Nb5+, Ga3+, and Zn2+ [23,24,25,26,27]. Because of the diversity of vacancy orbit and positive charge of metal ions, the affinity of different metal ions to phosphopeptides is different, which affects the analysis of global phosphopeptides [28]. For example, Ti4+, Fe3+, and Tb3+ tend to capture mono-phosphopeptides, while Cu2+, Ni3+, Zr4+, Nb5+, Ga3+, and Zn2+ have stronger attraction for multi-phosphopeptides [21]. Therefore, working out a nanocomposite with the capability of enrichment of global phosphopeptides has become a hot spot in the research of phosphorylated proteomics [29].

The selection of chelating ligands is the priority in the study of IMAC materials. In the 1970s, it was observed that cysteine, histidine, and copper ions can form stable complexes in aqueous solutions, and the complexes can enrich some specific proteins in the solution, thus making the application of IMAC method expands to macromolecular protein [30]. In 1986, Andersson further developed this method by combining iminodiacetic acid (IDA) with Fe3+ chelation, which made phosphate amino acids in the solution preserved, while the non-phosphate amino acids did not, indicating that IDA could be used as a chelating ligand of metal cations to enrich phosphopeptides [31]. Subsequently, a variety of chelating ligands have been developed, such as NTA, PDA, and ATP, which can not only chelate metal cations but also increase the hydrophilicity of the material, greatly promoting the development of phosphoproteomics [32,33,34]. For example, Deng research group prepared the material Fe3O4@PDA-Ti/Nb by using polydopamine (PDA) as chelating ligand for loading Ti and Nb. Comparing with only one metal ion-loaded material (Fe3O4@PDA-Ti and Fe3O4@PDA-Nb), Fe3O4@PDA-Ti/Nb displayed better enrichment efficiency than the single metal ion-loaded microspheres [35]. This work has contributed a lot to the development of phosphorylated proteomics. Nevertheless, the Fe3O4 has relatively low surface area. Therefore, development of new ligands and materials with large specific surface area is still an urgent task.

Herein, a brand-new chelating agent, ammonia methyl phosphate (APA), was developed to immobilize two metal ions Ti4+/Nb5+ to the magnetic mesoporous silicon dioxide matrix (denoted as Fe3O4@mSiO2@APA@Ti4+/Nb5+). Ti4+/Nb5+ endowed the newly prepared nanosphere with the capability to capture global phosphopeptides. Ti4+ ion is one of the most commonly used ions in the IMAC method; the maturity of its development and research can ensure that the material has a certain enrichment performance. In addition, after investigation, it is found that Nb5+ has an excellent enrichment effect on polyphosphorylated peptides among many metal cations, and can make up for the deficiency of Ti4+ ion in the enrichment of polyphosphorylated peptides. The introduction of the new chelating ligand APA made the material excellent metal cation chelating ability, which greatly improves the loading of metal ions for the specific adsorption of phosphorylated peptides. Fe3O4@mSiO2 was used as the core, which not only imparts strong magnetism for material separation but also provides excellent water dispersion as well as plenty of loading sites for post-synthesis of the nanosphere. Fe3O4@mSiO2@APA@Ti4+/Nb5+ with the above merits was anticipated to have great latent capacity in phosphoproteomics study.

Experimental section

Chemicals

Iron chloride hexahydrate (FeCl3·6H2O, AR), dithiothreitol (DTT, AR), ammonium bicarbonate (NH4HCO3, AR), iodoacetamide (IAA, AR), hexadecyl trimethyl ammonium bromide (CTAB), sodium hydroxide (NaOH, AR), trifluoroacetic acid (TFA, AR), 3-glycidoxypropyltrimethoxysilane (GLYMO, AR), methylbenzene, and aminomethyl phosphonic acid (APA, AR) were bought from J&K (Shanghai, China, www.jkchemical.com). Albumin from bovine serum (BSA, 98%), ethanol (EtOH, AR), beta-casein (β-casein, 98%), trypsin, DHB, and ethylene glycol were bought from Sigma-Aldrich (www.sigmaaldrich.com). Acetonitrile (ACN, AR), niobium(V) oxalate hydrate (Nd2(C2O4)3, AR), titanic sulfate (Ti(SO4)2, AR), ammonium hydroxide, and sodium acetate (CH3COONa, AR) were bought from Aladdin (Shanghai, China, www.aladdin-e.com). Serum and saliva were obtained from affiliated NBU Hospital (number of volunteers: 1, gender: male, age 25).

Preparation of Fe3O4@mSiO2@APA@Ti4+/Nb5+

Fe3O4 was acquired by the method reported in the previous literature and simply modified [12]. The specific step is to disperse FeCl3·6H2O (2.7 g) into 150 mL of ethylene glycol ((CH2OH)2), stirred the mixture through to yellow and transparent (300 rpm/min), after add ground anhydrous sodium acetate (7.2 g) and stir for 2 h (300 rpm/min) that was devolved to a reaction vessel reacted at 200 °C for 16 h. The shell structure mSiO2 was modified by employing a dry magnetic sphere (100 mg) and 1 g of CTAB dispersed in 100 mL deionized water under ultrasound conditions, then, slow accession 10 mM NaOH (100 mL) and 800 mL of deionized water and reacted for 1 h at 60 °C with mechanically stirred (300 rpm/min). Finally, 2 mL of ethanol and 0.5 mL of TEOS were added and mechanically stirred (300 rpm/min) at 60 °C about 12 h, and the obtained products were cleaned three times with water phase (ultrapure water) and organic phase (ethanol). After vacuum dried, the Fe3O4@mSiO2 was obtained by calcining dry Fe3O4@SiO2 at 350 °C for 4 h.

Fe3O4@mSiO2 (100 mg) was dispersed in 80 mL of toluene solution for ultrasonic dispersion (containing 800 μL of silane coupling agent GLYMO), refluxed at 80 °C for 12 h, cleaned three times with water phase and organic phase and dried under vacuum at 50 °C overnight, and Fe3O4@mSiO2-GLYMO was obtained. The chelating ligand APA was immobilized on Fe3O4@mSiO2 by a simple synthesis reaction. A total of 100 mg of APA was dispersed in 50 mM NH4HCO3 (120 mL) under ultrasonic conditions, then 100 mg of Fe3O4@mSiO2-GLYMO was dispersed into 60 mL of the configured APA solution and reacted at 65 °C for 3 h, the supernatant was removed and the remaining APA solution was added, and the reaction was continued for 3 h, washed obtained product, and freeze drying, achieving secondary product Fe3O4@mSiO2@APA.

Fe3O4@mSiO2@APA@Ti4+/Nb5+ was obtained by dispersing Fe3O4@mSiO2@APA (200 mg) in 10 mL aqueous solution containing 100 mM Ti (SO4)2 and 50 mM Nd2(C2O4)3, shaking for 2 h at room temperature and removed supernatant by magnet separation. The obtained final product was cleaned three times with water phase and organic phase, and dried under vacuum at 50 °C for 12 h to obtain Fe3O4@mSiO2@APA@Ti4+/Nb5+. Nanosphere of the Fe3O4@mSiO2@APA@Ti4+ obtained using the same method was almost the same as above except for Nd2(C2O4)3/Ti(SO4)2.

Characterization

SEM was characterized by a Keol 2012 microscope. X-ray diffraction (XRD) was characterized using Bruker XRD (D4). Infrared (FT-IR) spectroscopy was recorded using Thermo Fisher Scientific 10 infrared spectrometer analysis. Matrix-assisted laser desorption ionization time-of-flight mass spectrometry (MALDI-TOF MS) was used in an autoflex max (Bruker, USA). TEM image was recorded using JEOL 1011 microscopy (Japan), X-ray photoelectron spectroscopy (Axis Ultra DLD), automatic specific surface area, and pore analyzer (HD88, USA Micromeritics).

Standard protein and actual sample pretreatment

β-Casein was primarily pretreated prior to enrichment. A total of 2 mg β-casein was dispersed to 400 μL of ammonium bicarbonate aqueous solution, denatured at 100 °C boiling water for 10 min, after cooling down, enzymatic hydrolysis for 12 h with the help of trypsin (2 mg/mL, 50 μL).

Reduction alkylation of BSA with DTT and IAA was by classic approach. Firstly, 1 mg BSA was dispersed in 200 μL of 50 mM NH4HCO3 solution and denatured at 100 °C boiling water for about 10 min. Five microliters 200 mM DTT and 45 μL 50 mM NH4HCO3 were added, and shook for 1 h under 37 °C. Next, 10 μL 400 mM IAA and 90 μL 50 mM NH4HCO3 were employed of alkylation under the dark of 37 °C about 1 h. Finally, 20 μg of trypsin was added, the final solution was incubated at 37 °C about 16 h.

Saliva comes from healthy adults in the laboratory. Suspended matter in saliva was detached via centrifugation under hypothermia and 14,000 rpm high speed, then, saliva after treatment was used for subsequent experiment. Human serum was used without further treatment.

Enrichment experiment

A total of 500 μg of Fe3O4@mSiO2@APA@Ti4+/Nb5+ was dispersed into 200 μL of enrichment buffer (ACN/TFA/H2O = 50%:/0.1%:/49.9%), after ultrasonic for about 5 min, 2 μL of β-casein was appended, removing the supernatant by a magnet after shaken at 37 °C about 30 min, and the phosphopeptide-loaded nanomaterial was washed several times using enrichment buffer (ACN/TFA/H2O = 50%:/0.1%:/49.9%, about 200 μL). Afterward, incubated for 15 min at 37 °C after 10 μL of desorption buffer ammonia (NH4OH, 0.4 mol/L) was added. Finally, analysis phosphopeptides from the eluate was by MALDI-TOF MS. The enrichment process of practical samples is the same as above, only changing 2 μL β-casein to 2 μL of practical samples (serum, saliva).

MALDI MS analysis

All detections are carried out in reflected positive ion mode. The instrument uses an improved Nd:YAG laser with a detection frequency of 1000 Hz, a flight tube acceleration voltage of 20 kV, a detector voltage of approximately 18 kV, and detection range of mass charge ratio is 1000–3500 Da. The concentration of matrix DHB is 20 mg/mL, (solvent with ACN/H2O = 30%/70%).

Results and discussion

Preparation of Fe3O4@mSiO2@APA@Ti4+/Nb5+ nanosphere

The synthesis process of Fe3O4@mSiO2@APA@Ti4+/Nb5+ nanosphere was presented in Scheme 1. In short, Fe3O4 was first prepared by the hydrothermal method. Next, a mesoporous silica magnetic material Fe3O4@mSiO2 was prepared to employ CTAB as the templating agent. Then, the APA chelating agent was successfully modified onto the surface of Fe3O4@mSiO2. Finally, Ti4+ and Nb5+ were modified on APA to prepare Fe3O4@mSiO2@APA@Ti4+/Nb5+ nanosphere. In order to verify the successful preparation of the material and the performance, we performed a series of characterization and enrichment experiments on the material.

Scheme 1
scheme 1

The preparation step for Fe3O4@mSiO2@APA@Ti4+/Nb5+nanosphere

Full size image

The micromorphic status and particle size of Fe3O4@mSiO2@APA@Ti4+/Nb5+ were firstly observed through SEM and TEM. As shown in SEM (Fig. 1a, b), the average particle size of the Fe3O4@mSiO2@APA@Ti4+/Nb5+ is about 200 nm in diameter, and the wrinkles on the surface of the material also indicate that the modifier has been successfully coated on the surface of the material. The TEM image (Fig. 1c, d) shows that the total thickness of silica and other modifications is approximately 30 nm. Ti and Nb observed from EDX analysis show that Ti4+ and Nb5+ were faultlessly modified onto the material (Fig. S1). XPS analysis was further tested for the elemental composition of the material (Fig. S2). The phase structure of Fe3O4@mSiO2@APA@Ti4+/Nb5+ was analyzed by XRD ray diffraction. As shown in Fig. S3, the diffraction peaks of 30.1° (220), 43.1° (400), and 57.1° (511) belong to the Fe3O4 crystal, and 36.1° (220) and 63.2° (422) belong to the characteristic peak of silicon dioxide. The material was then subjected to Fourier infrared detection to prove successful preparation. Comparing the three curves of Fig. S4, the strong peak at 575 cm−1 attributed to the Fe–O, peak at 1100 cm−1 belongs to the stretching vibration of Si–O–Si, the infrared peak caused by the stretching vibration of N–H is at 1650 cm−1. It is found in actual experiments that the nanosphere has excellent water dispersibility as well as an outstanding magnetic response (Fig. S5), which makes the separation of materials extremely simple. Nitrogen adsorption-desorption isotherm indicates the presence of mesoporous channels. As shown in Fig. S6, the BET surface area is 151.1 m2/g, the total pore volume is 0.17 cm3/g, and the average pore width (4 V/A by BET) is 4.45 nm. The data shows that Fe3O4@mSiO2@APA@Ti4+/Nb5+ may have the ability to exclude large volumes of proteins when enriching phosphopeptides.

Fig. 1
figure 1

a, b SEM image and c, d TEM image of Fe3O4@mSiO2@APA@Ti4+/Nb5+

Full size image

Investigation on the enrichment of phosphopeptides from β-casein digests

The procedure for enriching phosphopeptides from β-casein digests was displayed in Scheme 2. First of all, Fe3O4@mSiO2@APA@Ti4+/Nb5+ was dispersed into the loading buffer (ACN/TFA/H2O = 50%: 0.1%: 49.9%) containing β-casein digest to capture phosphopeptides, then washed several times with loading buffer after enrichment in order to elute the non-phosphopeptides that adhere to the surface of the material. Finally, desorption buffer (0.4 M NH4OH) was utilized to elute the enriched phosphorylated peptides, and the eluant was analyzed by MALDI-TOF MS.

Scheme 2
scheme 2

The preparation step for Fe3O4@mSiO2@APA@Ti4+/Nb5+

Full size image

The enrichment time was firstly investigated under the same conditions except for the enrichment time. The results showed that the two phosphopeptide peaks increased with the increase of enrichment time, but the change range tended to be gentle after 30 min (Fig. S7). Therefore, after comprehensive consideration, we finally selected 30 min as the final enrichment time.

For effective evaluation of the capacity of Fe3O4@mSiO2@APA@Ti4+/Nb5+ towards phosphopeptides, β-casein was used as a phosphorylated peptide standard protein. Before enrichment, the baseline of the graph was higher and the peak intensity was relatively weak, and no phosphorylated peptides were observed in virtue of the shielding effect of the non-phosphopeptides (Fig. 2a). After enriched with Fe3O4@mSiO2@APA@Ti4+ (Fig. 2b), 8 phosphopeptides and dephosphopeptides were detected, and the peak strength and the number of hetero-peaks have greatly improved compared to Fig. 2a. Especially, the peak intensity of monophosphopeptides (2061 and 2556) was significantly higher than that of multiphosphopeptides (3122), indicating the enrichment bias of Ti4+ to monophosphopeptides. The result enrichment with Fe3O4@mSiO2@APA@Nb5+ (Fig. 2c) showed Nb5+ had a strong affinity for multiphosphopeptides. Figure 2d shows the spectrum obtained after hiring Fe3O4@mSiO2@APA@Ti4+/Nb5+, and the typical 2062, 2556, and 3122 phosphopeptides displayed extremely high peak intensity contrast to the precursor material, revealing that the introduction of Ti4+ and Nb5+ dual ions prominently improved the enrichment efficiency towards global phosphopeptides (the detail information was shown in Table S1). Besides, we also observed the cycle stability of the prepared nanosphere, as shown in Fig. S8, both Fe3O4@mSiO2@APA@Ti4+ and Fe3O4@mSiO2@APA@Ti4+/Nb5+ showed good reusability. We also investigated the stability of the material by comparing the performance of the material in different periods (Fig. S9). It can be seen that even if the material was stored for 6 months, it still has good enrichment capacity, which exhibits that the material has a good storage lifetime.

Fig. 2
figure 2

MALDI-TOF mass spectra of β-casein digests (0.1 pmol).Before enrichment (a). Afterenriched with Fe3O4@mSiO2@APA@Ti4+(b), Fe3O4@mSiO2@APA@Nb5+(c), and Fe3O4@mSiO2@APA@Ti4+/Nb5+(d). Phosphopeptide peaks areflagged as “*”, dephosphorylated peaks are flagged as“#”

Full size image

To investigate the effect of Ti/Nb ions on the enrichment of phosphopeptides, enrichment experiments on precursor were carried out and the results were shown in Fig. S10. It was found that the precursor materials had almost no enrichment effect on the phosphopeptides, which indicates that Ti(IV) and Nb(V) ions play a decisive role in the enrichment of phosphopeptides and their amounts have a direct impact on the enrichment performance. Then, three kinds of low concentration solutions of the β-casein were employed to investigate the enrichment LOD of Fe3O4@mSiO2@APA@Ti4+/Nb5+ (standard is based on the signal-to-noise ratio SNR = 3, and the RSD was lower than 7%, Table S2). As shown in Fig. 3, Fe3O4@mSiO2@APA@Ti4+ and Fe3O4@mSiO2@APA@Ti4+/Nb5+ can still enrich phosphopeptides when the dilution of β-casein was as low as 1 fmol, and when the standard protein β-casein was diluted to 0.5 fmol, only three phosphorylated peptides could be enriched by Fe3O4@mSiO2@APA@Ti4+, and the multiphosphopeptide 3122 failed to be enriched. Nevertheless, comparing with the mass spectrum after enrichment by Fe3O4@mSiO2@APA@Ti4+/Nb5+, the three typical phosphopeptide peaks are all enriched with higher S/N and peak intensity. Further diluting the standard protein concentration to 0.05 fmol, three phosphorylated peptides can still be enriched by Fe3O4@mSiO2@APA@Ti4+/Nb5+, but Fe3O4@mSiO2@APA@Ti4+ cannot (Fig. S11). It displayed that the immobilization of dual ions improved material both global phosphopeptide affinity and trace detection capability. The above results showed that our new prepared nanosphere Fe3O4@mSiO2@APA@Ti4+/Nb5+ had a low detection limit for phosphopeptides, which was conducive to the detection of low-level phosphopeptides in actual complex samples. The loading capacity of the Fe3O4@mSiO2@APA@Ti4+/Nb5+ was investigated by introducing different amounts of the nanosphere into 10 μL β-casein digest. After enrichment, the eluant was detected by MALDI at the same detection conditions.

Fig. 3
figure 3

MALDI-TOF mass spectra analysis of phosphopeptidesfrom weak solutions of the β-casein digests after enriched by 10 fmol (a), 1 fmol (b), and 0.5 fmol (c)Fe3O4@mSiO2@APA@Ti4+.and by 10 fmol(d), 1 fmol (e), and 0.5 fmol (f) Fe3O4@mSiO2@APA@Ti4+/Nb5+.Phosphopeptide peaks are flagged as “*”, dephosphorylated peaks are flaggedas“#”

Full size image

The capacity of the Fe3O4@mSiO2@APA@Ti4+/Nb5+ was estimated by observing changes of peak intensity of the three typical phosphopeptides. As shown in Fig. 4, the load capacity of phosphopeptides on Fe3O4@mSiO2@APA@Ti4+ was increased from 150 to 200 μg, and reached adsorption saturation at 200 μg; the saturation load capacity was calculated to be 333 μg/mg. The maximum load capacity of Fe3O4@mSiO2@APA@Ti4+/Nb5+ was calculated to be 500 μg/mg using the same method.

Fig. 4
figure 4

Calculation of the loading capacity forphosphopeptides. Fe3O4@mSiO2@APA@Ti4+(a). Fe3O4@mSiO2@APA@Ti4+/Nb5+(b)

Full size image

The anti-interference performance of the material was detected by adding a certain ratio of BSA digest to the β-casein digest. It can be seen from Fig. 5 and Fig. S12 that both materials have strong anti-interference ability. However, under the same ratio, the dual-ion material enriches more phosphorylated peptides than the single-ion material, further expanding the ratio to 1:1500 and four phosphorylated peptides were still detected after being enriched by Fe3O4@mSiO2@APA@Ti4+/Nb5+ (Fig. S13). In addition, we directly analyzed the mixture of β-casein:BSA = 1:100 by mass spectrometry without material enrichment, and the results showed that no phosphorylated peptides were detected (Fig. S14). It shows that the introduction of Nb5+ ion strengthens the affinity of phosphopeptides, and enhances the specificity towards phosphopeptides. The experimental results showed that Fe3O4@mSiO2@APA@Ti4+/Nb5+ had excellent enrichment performance on phosphopeptides.

Fig. 5
figure 5

MALDI-TOF mass spectra of enriched phosphopeptidesfrom a mixture of β-casein and BSA with a mass ratio of 1:100 (a), 1:250 (b), 1:500 (c), and 1:1000(d) with Fe3O4@mSiO2@APA@Ti4+/Nb5+.Phosphopeptide peaks are flagged as “*”, dephosphorylated peaks are flaggedas“#”

Full size image

Subsequently, we added the size-exclusion experiment by unhydrolyzed BSA to the hydrolyzed β-casein. The mixed solution was enriched by Fe3O4@mSiO2@APA@Ti4+/Nb5+; both the eluent and the supernatant were analyzed by mass spectrometry. As shown in Fig. S15, the phosphopeptide peak was only observed in the elution, and there was no signal of BSA intact protein, and BSA protein peak was observed in the supernatant. The above results showed that the material Fe3O4@mSiO2@APA@Ti4+/Nb5+ had a significant size exclusion effect on bulk materials.

The enrichment recovery rate of Fe3O4@mSiO2@APA@Ti4+/Nb5+ on the phosphopeptides was detected by the isotope labeling method. In short, β-casein was treated with trypsin in H216O and H218O to obtain 16O- and 18O-labeled β-casein digests, which produced a mass difference of 4 Da in mass spectrometry (introducing two 18O atoms). Take two equal amounts of the digests with heavy label and light label. Firstly, the light label β-casein fraction was enriched with materials, and the eluent was mixed with the heavy label fraction, and the mixture materials were enriched with Fe3O4@mSiO2@APA@Ti4+/Nb5+. The eluent was analyzed by mass spectrometry. The recovery of the material was obtained by calculating the ratio of the peak intensity of the lightly isotope-labeled phosphopeptide to the peak intensity of the corresponding heavy isotope-labeled phosphopeptide. After repeating the experiment three times, the average recovery rate of the material is 87% (Fig. S16).

Study enrichment of biological samples

Actual samples of human serum and saliva were hired to examine whether the material can be used to analyze complex samples (Fig. 6). Four phosphopeptide peaks were detected from human serum (Table S3) and 24 phosphopeptides were analyzed in the saliva treated with Fe3O4@mSiO2@APA@Ti4+/Nb5+. The above results indicated that Fe3O4@mSiO2@APA@Ti4+/Nb5+ nanomaterial had surprising application value in actual research. Based on the perfect chelation of metal ions and phosphate groups, we believe the compounds containing phosphate and carboxyl groups can be exacted using our prepared material. In addition, in order to prove the excellent enrichment efficiency of Fe3O4@mSiO2@APA@Ti4+/Nb5+ materials, we selected several Ti4+ IMAC nano-adsorbed materials previously reported; comparing with the other materials in Table 1, Fe3O4@mSiO2@APA@Ti4+/Nb5+ is superior in selectivity and detection limits.

Fig. 6
figure 6

MALDI-TOF mass spectra of phosphopeptides from humanserum: before enrichment (a), afterenrichment by Fe3O4@mSiO2@APA@Ti4+/Nb5+(b), mass spectra of phosphopeptidesfrom saliva: before enrichment (c), afterenrichment by Fe3O4@mSiO2@APA@Ti4+/Nb5+(d). Phosphopeptide peaks areflagged as “*”

Full size image
Table 1 Enrichment efficiency comparison with previously reported materials
Full size table

Conclusions

Based on the IMAC enrichment strategy, we successfully developed a novel material for phosphopeptide enrichment by a simple synthesis method. Fe3O4@mSiO2@APA@Ti4+/Nb5+ nanosphere greatly improves the characteristics of single metal ion IMAC material biased to phosphopeptide; this material enhances the affinity of IMAC nanomaterials for global phosphopeptides. The employment of the brand-new chelating agent APA enormously ameliorates the immobilization efficiency of metal ions, so that the enriched phosphopeptides has a higher signal. Although Fe3O4@mSiO2@APA@Ti4+/Nb5+ has good selectivity and specificity towards phosphopeptides, the material still has some limitations, for example, the surface area is relatively small, and there are many synthetic steps.