Article Cite This: Langmuir XXXX, XXX, XXX−XXX
pubs.acs.org/Langmuir
Adsorption of Phosphate and Polyphosphate on Nanoceria Probed by DNA Oligonucleotides Xiuzhong Wang,†,‡ Anand Lopez,‡ and Juewen Liu*,‡ †
College of Chemistry and Pharmaceutical Sciences, Qingdao Agricultural University, Qingdao 266109, China Department of Chemistry, Waterloo Institute for Nanotechnology, University of Waterloo, Waterloo, Ontario N2L 3G1, Canada
Downloaded via UNIV OF CALIFORNIA SANTA BARBARA on June 22, 2018 at 11:14:49 (UTC). See https://pubs.acs.org/sharingguidelines for options on how to legitimately share published articles.
‡
ABSTRACT: Phosphate-containing molecules exist in many forms in biology and the environment, and their interaction with metal oxides is an important aspect of their chemistry and biochemistry. In this work, phosphates with different degrees of polymerization (e.g., orthophosphate, pyrophosphate (PPi), sodium triphosphate (STPP), sodium trimetaphosphate (STMP), and polyphosphate with 25 phosphate units) and phosphates with one or two capping groups were studied. CeO2 nanoparticles (nanoceria) were used as a model metal oxide. DNA is also a polyphosphate, and a fluorescently labeled DNA oligonucleotide was mixed with nanoceria. These phosphate species were individually added to displace the adsorbed DNA. Longer phosphate chains were more efficient when each molecule was used at the same molar concentration, whereas PPi and STPP were most efficient at the same total phosphorus atom concentration. By capping the phosphate with organic groups, the affinity was significantly decreased. Isothermal titration calorimetry (ITC) was also performed to quantitatively measure thermodynamic parameters. Although STMP was very slow at displacing DNA, it was still adsorbed very strongly by nanoceria from ITC, indicating kinetic effects likely due to its ring structure. This observation allowed us to use the DNA as a probe to study the hydrolysis of STMP to form STPP. In summary, this study provides a systematic understanding of phosphate species interacting with metal oxides, and interestingly, it demonstrates an analytical application as well.
■
INTRODUCTION Phosphate is a key building block in biology.1,2 Nucleic acids have a phosphate backbone, while most lipids contain a bridging phosphate. Even though phosphate is not present in the natural amino acids, phosphorylated proteins are key for cell signaling. Unlike the bridging phosphates in nucleic acids and lipids, the phosphates in proteins are terminal. Aside from these biomacromolecules, phosphate also exists in many small molecules. The most important energy molecule, adenosine triphosphate (ATP), contains a linear triphosphate. Its hydrolysis to adenosine monophosphate (AMP) releases a pyrophosphate (PPi). Phosphate can also polymerize, forming dimers, trimers, polymers, and cyclized structures.3,4 Finally, the oxygen atoms in phosphate can also be capped by various ligands. Because of its chemical inertness, however, phosphate is often overlooked. Phosphate has a high adsorption affinity for many metal oxides, and this property has resulted in some interesting applications. For example, TiO2 was commonly used to extract phosphorylated proteins for subsequent analysis.5 The phosphate group in lipids was believed to be important for adsorption to some metal oxide surfaces, and engineered lipids with exposed phosphate were found to be even better for this purpose.6,7 DNA oligonucleotides can be stably adsorbed on many metal oxides,8 such as cerium dioxide (CeO2),9 Fe3O4,10,11 ZnO,12−14 and TiO2.15,16 In each case, the phosphate backbone of DNA played the main role for © XXXX American Chemical Society
adsorption, and inorganic phosphate can displace the adsorbed DNA. Interactions between phosphate and minerals are also important. For example, some RNAs such as poly(adenosine diphosphate-ribose) contain PPi linkages,17 and they are believed to be a key factor for organizing cellular architectures and biomineralization.18 Calcium phosphate is the key building block of the hard tissues.19,20 Given the complexity of phosphate chemistry, it is important to have a systematic understanding of its adsorption on metal oxides. Such understandings are useful for biosensor development,21 biomineralization, regulating the catalytic activity of metal oxides (called nanozymes),22,23 and environmental remediation.24−26 Taking advantage of DNA being a polyphosphate and the fluorescence quenching property of many materials,27−30 we reason that it might be possible to use DNA as a probe to understand the adsorption of various phosphates.28,31,32 In this work, we performed a comprehensive study of the interaction between various phosphate species and a model metal oxide, CeO2. Quantitative measurements using isothermal titration calorimetry (ITC) were also carried out. Related findings allowed us to further develop an assay for an important phosphate reaction. Received: May 6, 2018 Revised: June 7, 2018 Published: June 11, 2018 A
DOI: 10.1021/acs.langmuir.8b01482 Langmuir XXXX, XXX, XXX−XXX
Langmuir
■
■
MATERIALS AND METHODS
Article
RESULTS AND DISCUSSION Phosphate Species Studied. In this work, we studied a total of eight phosphate species with different numbers of phosphate units or capping groups (Figure 1). For various
Chemicals. All the DNA samples were purchased from Integrated DNA Technologies (IDT, Coralville, IA, USA) and were purified by standard desalting. The sequence of the carboxyfluorescein (FAM)labeled 24-mer DNA is FAM-ACGCATCTGTGAAGAGAACCTGGG, with the FAM on the 5′-end. Sodium chloride, sodium hydroxide, and 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid (HEPES) were purchased from Mandel Scientific (Guelph, Ontario, Canada). CeO2 nanoparticles (CeO2 NPs), malachite green, sodium molybdate dihydrate, sodium PPi decahydrate, sodium polyphosphate with 25 phosphate units (Pi25), sodium trimetaphosphate (STMP), sodium triphosphate (STPP), and dimethyl phosphate (DMP) were purchased from Sigma-Aldrich. Choline dihydrogen phosphate was purchased from Ionic Liquid Technologies. Milli-Q water was used to prepare all the buffers and solutions. DNA-Based Probing. To study the DNA adsorption and desorption by CeO2, 50 nM FAM-labeled DNA was dissolved in HEPES buffer (10 mM, pH 7.4, with 200 mM NaCl). To optimize CeO2, different concentrations of CeO2 NPs were added to the above solution and incubated for 30 min before the fluorescence was recorded (Eclipse, Varian). The desorption kinetics were recorded after a quick addition of phosphate to the FAM-DNA/CeO2 mixture (final concentration = 5 μg/mL CeO2) in the above HEPES buffer, followed by fluorescence recording for 90 min at room temperature (∼23 °C). The fluorescence was compared with the initial intensity of the free DNA to calculate the adsorbed DNA. The HEPES buffer and CeO2 concentration were for all the experiments unless otherwise indicated. Isothermal Titration Calorimetry. All calorimetric titrations were performed with a microcalorimeter instrument (MicroCal) at 25 °C. All phosphates were prepared using the HEPES buffer (10 mM, pH 7.4, 200 mM NaCl). The CeO2 concentration in the cell was 100 μg/mL (1.4551 mL in the ITC parameter setting), and the phosphate concentration in the injection syringe was 1.0 mM. Each titration consisted of an initial injection of 1 μL, followed by 28 injections of 10 μL spaced 300 s apart. The initial 1 μL injection was affected by diffusion into and out of the injection syringe during the initial equilibration period. Data from this injection were omitted from the analysis. The background heat of buffer injections into the CeO2 solution was subtracted from the corresponding titrations. The enthalpy (ΔH) and binding constant (Ka) were obtained through fitting the titration curves to a one-site binding model using the Origin software. ΔS was calculated from ΔG = ΔH − TΔS. Colorimetric Quantification of Phosphate. The malachite green colorimetric assay procedures were used according to the literature with minor modifications.33 The malachite green reaction solution (R1) was prepared as follows. First, 0.04 g of sodium molybdate dihydrate was dissolved in 5 mL of HCl (4 M). Then, this solution was mixed with 15 mL of 0.045% (w/v) malachite green aqueous solution and stirred for 30 min until it became clear. After centrifugation, the supernatant was stored at 4 °C for use. Eight microliters of various concentrations of phosphate was placed into microcentrifuge tubes and mixed with 8 μL of 0.1 M HCl, 104 μL distilled water, and 80 μL R1 for 10 min. A mixture of the respective phosphate concentrations with distilled water was prepared as controls. A mixture of R1 (80 μL), the respective acidic solution (8 μL), and distilled water (112 μL) was used as a blank for ultraviolet (UV)−visible (vis) spectroscopy. The absorbance was read at 670 nm in a UV spectrometer (Agilent 8453A) at room temperature. The experiments were run with three repeats in series. Hydrolysis of STMP. STMP (10 mM) and STPP (10 mM) were respectively incubated in 2.0, 1.0, 0.5, and 0.1 M NaOH for 60 min, after which the solutions were diluted over 1000-fold in buffer to ensure a neutralized pH. The samples were then used for desorbing DNA from CeO2 as described above. The final STMP (if no hydrolysis occurred) was 10 μM.
Figure 1. Structure of phosphate-containing species studied in this work: (A) phosphates of different degrees of polymerization; (B) capping inorganic phosphate with one or two groups; and (C) structure of a 4-mer DNA oligonucleotide with a polyphosphate backbone (this study used a longer DNA and this is only for showing the structure).
degrees of polymerization, we included simple inorganic orthophosphate (Pi), PPi, STPP, polyphosphate with 25 phosphate units (Pi25), and STMP (Figure 1A). Among them, STMP is cyclic, whereas the rest are linear. The second group was based on Pi (Figure 1B), and we capped it by one [phosphocholine (PC)] or two groups (DMP). These species are representative in terms of phosphate chemistry and allow a systematic study. Note that at neutral pH, a fraction of the phosphate groups was protonated, depending on the pKa of each group, but these protons were not drawn in Figure 1 for clarity. The structure of a 4-mer DNA oligonucleotide is also shown (Figure 1C). DNA is a polyphosphate, and all of its phosphate groups are in the bridging form with two of the oxygen atoms capped, which is similar to DMP. Fluorescent DNA as a Probe. DNA is a polyphosphate, and it can strongly adsorb on various metal oxides mainly via its phosphate backbone.12,13 The affinity of adsorption can be readily tuned by varying the length of DNA, whereas the DNA sequence is often less important.10,12 For example, CeO2 NPs (nanoceria) strongly adsorb DNA and also quench fluorescence. Previous studies confirmed the importance of the DNA phosphate backbone for adsorption.9,34 CeO2 is also highly important for its catalytic activities,35 and it has been tested for analytical36 and biomedical applications.37−39 For these reasons, we chose CeO2 in this work as a model oxide. Our CeO2 NPs were ∼5 nm [Figure 2A for transmission electron microscopy (TEM)], and adding increasing concentrations of CeO2 to a FAM-labeled DNA gradually quenched the fluorescence (Figure 2B). Even 10 μg/mL of CeO2 fully quenched the emission from 50 nM of a 24-mer DNA. This DNA was chosen as a representative random sequence. As explained above, the effect of a DNA sequence should be B
DOI: 10.1021/acs.langmuir.8b01482 Langmuir XXXX, XXX, XXX−XXX
Article
Langmuir
PPi and STPP samples had a similar high response. Out of the three phosphate groups in STPP, two are terminal phosphates and one is bridged in the middle. We reason that the bridging phosphate might contribute less to adsorption. A previous work on Pi and PPi adsorption by the iron oxide surface has indicated that the number of Fe−O−P linkages is important for the final binding strength,40 and other detailed studies have also been reported at such levels.41,42 Terminal phosphate can establish more linkages with the surface and thus has a higher affinity. PPi has only terminal phosphates, and it showed strong binding even though it should be less efficient in terms of polyvalent binding. 43 On the other hand, the molar concentration of PPi was 1.5-fold higher than that of STPP, and this may also contribute to the higher signal. By comparing Pi and PPi, we can see that polyvalency contributed significantly. Despite the molar concentration of PPi being only half of Pi, PPi had doubled the signal, and this can only be attributed to the polyvalency effect. The Pi25 sample was even slower than the Pi sample, and this can be rationalized by most of its phosphates being bridging ones. As its molar concentration was 25-fold lower, the disadvantage of the bridging phosphate became very obvious. It is quite interesting to notice that the cyclic STMP had a very slow response, although it also has three phosphate units. For STMP, all of its phosphates are bridging. The shape of its kinetic curve was also different (Figure 3A). Initially, its kinetics was very slow, followed by a faster phase, and the overall trend appeared to have an induction period. STMP might have a hard time competing with DNA kinetically. Only when adsorbed at a sufficiently high concentration, it can compete with the DNA. The cyclic geometry of STMP may be the key reason for its kinetic disadvantage. To further understand their adsorption, we studied the effect of phosphate concentration. A typical first-order kinetics was observed for PPi (Figure 3C), whereas STMP at below 100 μM showed a sigmoidal shape (Figure 3D). The PPi sample achieved half of the fluorescence increase with just 5.0 μM,
Figure 2. (A) TEM micrograph of the CeO2 NPs used in this work; (B) quenching of a FAM-labeled 24-mer DNA with a random sequence (50 nM) by different concentrations of CeO2 NPs in buffer (10 mM HEPES, pH 7.45, 200 mM NaCl); and (C) scheme showing adsorbed DNA as a probe to study the adsorption of phosphate species.
small.9 We reason that various phosphorus-containing species might compete with the DNA and displace it from the particle surface (Figure 2C). A stronger adsorbing species may induce faster DNA desorption and thus faster fluorescence enhancement. This method may allow us to conveniently study the adsorption of these phosphate species. Phosphate with Different Degrees of Polymerization. With this DNA-based probe, we first tested the phosphate species in Figure 1A. When 10 μM of each phosphate species was respectively added to the FAM-DNA/CeO2 complex, we observed different kinetics of fluorescence enhancement. The highest enhancement was observed with Pi25, followed by STPP, PPi, and phosphate (Figure 3A). This can be easily explained by the length of the phosphate chain. The longer the phosphate chain, the stronger the adsorption affinity. The only outlier was STMP, which was the slowest among these species despite the fact that it contains a total of three phosphate groups. As different compounds contained different numbers of phosphate units, to make a fair comparison, we then prepared another batch of samples with the same concentration of the total phosphorus atom of 10 μM (Figure 3B). In this case, the
Figure 3. Kinetics of fluorescence enhancement indicative of FAM-labeled DNA displaced by the added phosphate species: (A) 10 μM of each species and (B) 10 μM of the total phosphate unit. Different concentrations of (C) PPi and (D) STMP were added to induce desorption of the DNA. C
DOI: 10.1021/acs.langmuir.8b01482 Langmuir XXXX, XXX, XXX−XXX
Article
Langmuir
Figure 4. ITC traces and integrated heat with fitting of a one-site binding model for titrating (A) PO43−, 1.0 mM; (B) PPi, 1.0 mM; (C) STPP, 0.5 mM; (D) STMP, 0.5 mM; (E) Pi25, 1 mM; and (F) Pi25, 0.1 mM into CeO2 NPs. (G) Pi adsorption isotherm by CeO2 determined using the malachite green colorimetric method. The trend was fitted to a Langmuir isotherm. Inset: Photograph showing the color of these samples with 20, 50, 100, 200, 400, 800, and 1200 μM Pi.
Table 1. Thermodynamic Parameters from ITC Fitting Ka (×105 M−1)
N Pi PPi STPP STMP Pi25 PC DMP
95.0 60.0 21.4 26.8 2.0 90.7
± ± ± ± ± ±
8.0 2.0 3.7 0.2 0.42 3.44
ΔH (kcal mol−1) −7.190 −9.607 −10.79 −22.27 −79.21 −7.535
1.33 33.4 16.6 49.6 97.8 0.328
± ± ± ± ± ±
0.34 0.33 1.82 0.25 12.5 0.11
ΔS (cal K−1 mol−1) −0.674 −2.37 −18.5 −44.1 −234 −4.61
To have an accurate fitting, we need to find the number of adsorption sites on each CeO2 NP, and we used a colorimetric assay for this purpose. Malachite green forms a strongly absorbing species at around 670 nm (molar absorption coefficient, 90 000 M −1 cm −1 ) after complexing with phosphomolybdate and changes its color from light green to deep blue.33 The inset of Figure 4G shows the color of the samples containing molybdate, malachite green, and different concentrations of Pi. A nice color gradient was observed. These samples were used to build a calibration curve for subsequent quantification of Pi adsorption. We then added Pi of different concentrations to a fixed concentration of CeO2, and the samples were centrifuged to precipitate CeO2 with adsorbed Pi. Then, the nonadsorbed Pi was measured to construct an adsorption isotherm (Figure 4G). From this, we fitted a Langmuir adsorption model and obtained the saturated adsorption capacity. Taking the average size of each CeO2 NP to be 5 nm, we calculated its molar concentration to be 0.69 μM for a 100 μg/mL dispersion. Therefore, each CeO2 NP adsorbed 160 Pi ions. With the number of adsorption sites determined, we quantitatively fitted the data and obtained all the thermodynamic parameters (Table 1). The binding ratio (N) of phosphate to CeO2 decreased with the increasing length of the phosphate chain. Pi has a ratio of ∼107, and this is similar
whereas over 200 μM STMP was needed to induce the same increase in 90 min. The PPi sample reached equilibrium quickly within 10 min even with submicromolar PPi concentrations, whereas STMP displaced the DNA much slower, and equilibrium was not reached for most samples even after 2 h. Note that after a long incubation, the STMP signal caught up gradually. Therefore, PPi displaced the DNA much faster, whereas the adsorption affinity was hard to judge by this DNA probe. ITC Analysis. To quantitatively understand the adsorption of these phosphate species, we then performed ITC by gradually titrating each phosphate sample into a dispersion of CeO2 (no DNA added), and the rate of heat release was recorded (Figure 4A−F, top panels). ITC is a powerful biophysical technique that allows quantitative measurement of binding reactions, including molecular adsorption on surfaces.44−47 All these reactions were initially exothermic as indicated by the downward spikes of the thermograms. With increasing length of the phosphate chain, lesser injections were required to reach saturation. In particular, the Pi25 sample saturated after just two injections (Figure 4E). This was explained by the limited surface area of CeO2, and the phosphate chains were adsorbed lengthwise. We then integrated the heat (bottom panels) and fitted it with a single binding site model. D
DOI: 10.1021/acs.langmuir.8b01482 Langmuir XXXX, XXX, XXX−XXX
Article
Langmuir
magnetic resonance.48,49 A simpler biosensor method could allow more researchers to study this important reaction, which is related to the chemical synthesis of ATP and other phosphorylated compounds.50 We incubated STMP in NaOH solutions of different concentrations (and thus different pHs) for 60 min, and then the mixture was diluted ∼1000-fold in buffer to ensure neutralized pH. These samples were then added to the FAM-DNA/CeO2 mixture (Figure 5A). With
to the 160 determined from the colorimetric method. The difference could be attributed to the two different techniques used, including possible insufficient separation of CeO2 in the colorimetric method (e.g., centrifugation at 15 000 rpm may not fully remove all the CeO2 NPs). PPi had the ratio of around 60 which was about half that of Pi, suggesting that it can also fully occupy the surface of CeO2. STPP and STMP were adsorbed at around 30 molecules on each CeO2. Multiplying this number by 3 reached around 90 phosphate groups, and this number was also close to 107 for Pi adsorption. Therefore, it is likely that all these species can almost fully cover the surface adsorption sites. The only exception was the Pi25 chain, which saturated after just the first injection. It had a much lower total phosphate unit of just 20 (not shown). This number is unlikely to be accurate as only one main peak was available, and we could not obtain an accurate fitting with these data. We then decreased the concentration of Pi25 by 10-fold and performed a new titration. In this case, although the number of peaks increased, the data did not follow a simple trend (Figure 4F). We still fitted them to the same model and obtained a binding ratio of around 50 phosphate units. We attributed the lower number of adsorbed phosphate here to its polymeric nature, which may cross-link the CeO2 NPs to mask some adsorption sites. Although the initial injections all released heat, later injections showed heat absorption for many samples. It appears that the longer the phosphate chain, the more the heat was absorbed for the later injections (e.g., comparing Figure 4A,B,C,F). Heat absorption is indicative of entropydriven reactions. Adsorption of a molecule is typically accompanied with entropy loss, whereas an increase of entropy is often associated with release of structured water molecules. As our entropy change was proportional to the phosphate chain length, it was likely from the release of water molecules associated with the bridging phosphates. We then compared the association constants (Ka), which is a measurement of adsorption affinity. Among all the species, PPi had the highest adsorption affinity, followed by STTP. This is consistent with their excellent ability to displace DNA. Although the affinity of STMP was slightly weaker than that of PPi and STTP, it was 4-fold stronger than that of Pi. On the other hand, STMP was not as effective as Pi in displacing the DNA. Therefore, STMP was indeed kinetically disadvantageous in the presence of preadsorbed DNA. With a free CeO2 surface, its adsorption was quite strong. This is an interesting observation, and it might be related to its ring structure, which makes it more difficult to quickly achieve stable adsorption in the presence of a competing DNA. The enthalpy of the reactions increased gradually with the chain length of the phosphate species, indicating that most phosphate units in these molecules contributed to adsorption. The fitting was mainly based on the first few titrations, from which the entropy was all negative. A longer phosphate chain had a higher initial entropy value, and after adsorption its entropy also decreased more. This is consistent with a typical binding reaction of hydrophilic species in water. As discussed above, a second stage of binding driven by entropy could also exist at higher phosphate concentrations. Monitoring of STMP Hydrolysis To Form STPP. The very different DNA desorption kinetics in the presence of STMP and STPP may allow us to use this system to monitor the hydrolysis of STMP. Typically, this reaction needs to be monitored by mass spectrometry, chromatography, or nuclear
Figure 5. Kinetics of fluorescence enhancement of the FAM-DNA/ CeO2 complex mixed with neutralized (A) STMP and (B) STTP solutions preincubated with various concentrations of NaOH for 60 min. The initial STMP and STTP concentrations were 10 mM, and in the final mixture, their concentrations were 10 μM.
increasing NaOH concentrations, higher fluorescence enhancement was observed. This can be explained by the faster hydrolysis of STMP in basic solutions. We also compared the response of the same mixture of STPP mixed in the same base solutions (Figure 5B). In this control case, all the samples showed a similar response. The change observed in Figure 5A was indeed due to the hydrolysis reaction. Therefore, this simple DNA-based assay system could be used to monitor this difficult reaction. Phosphate with Capping Groups. The above studies all focused on polyphosphates. Another important aspect is to cap the simple inorganic phosphate with different groups. This is often seen in biological molecules. For example, the phosphate group in phospholipids has two capping groups and so has the phosphate in nucleic acids. Many small molecules such as PC, AMP, and flavin mononucleotide have one capping group. In this study, we compared three molecules as shown in Figure 1B to understand the capping effect. The same DNAbased method was used to study the displacement reaction (Figure 6A). Interestingly, Pi and PC had a very similar response, and thus, a single capping group did not seem to affect the affinity to CeO2. This can probably be explained by the geometry of the molecule. With a tetrahedral shape, not all the four oxygen atoms in the Pi can bind to the surface at the same time. On the other hand, DMP had almost no response, demonstrating that two capping groups have significantly decreased binding affinity. To further study this system, we also performed ITC (Figure 6B−D). Indeed, the adsorption profiles were quite similar for Pi and PC in terms of adsorption capacity (Table 1, last entries), whereas almost no heat was produced from the DMP sample. Quantitative fitting also showed that Pi had around 4fold higher adsorption affinity (Ka) compared to PC, and thus, capping even one oxygen was still detrimental for the adsorption affinity. Interestingly, Pi and PC had very similar enthalpy of adsorption, whereas the weaker affinity of PC was mainly from its higher entropy decrease. It is likely that PC is a larger molecule and has a higher entropy. After adsorbing on the surface, the flexibility of the choline part was also affected, causing more entropy loss. E
DOI: 10.1021/acs.langmuir.8b01482 Langmuir XXXX, XXX, XXX−XXX
Article
Langmuir
Figure 6. (A) Kinetics of DNA displacement from CeO2 by phosphate with capping groups (10 μM) in HEPES buffer. ITC traces and integrated heat with fitting of a one-site binding model for (B) PO43−; (C) PC; and (D) DMP (1.0 mM into 0.10 mg/mL CeO2).
■
ACKNOWLEDGMENTS We thank Dr. S. Jain for proofreading the paper. Funding for this work was from The Natural Sciences and Engineering Research Council of Canada (NSERC). X.W. was supported by the Shandong Provincial Government Scholarship (China) to visit the University of Waterloo. We also thank the Natural Science Foundation of Shandong Province, China (no. ZR2018MB030).
Individual DMP is too weak to measure, and we did not try to fit the data. It is interesting to note that the phosphate groups in DNA were similar to that in DMP with two caps. Therefore, polyvalent binding must be important for the phosphate in DNA to bind. Studying the adsorption of phosphate species by metal oxides has been extensively carried out, but previous work focused on simple Pi, PPi, or their derivatives.40,51 Various spectroscopic methods and density functional theory calculations have been performed to achieve atomic level understanding. In our work here, we focused on the thermodynamic and kinetic aspects of the reactions as a function of phosphate chain length and capping groups. Such understanding is complementary to the prior knowledge on this system.
■
CONCLUSIONS
■
AUTHOR INFORMATION
■
REFERENCES
(1) Schwartz, A. W. Phosphorus in Prebiotic Chemistry. Philos. Trans. R. Soc., B 2006, 361, 1743−1749. (2) Westheimer, F. Why Nature Chose Phosphates. Science 1987, 235, 1173−1178. (3) Achbergerová, L.; Nahálka, J. Polyphosphate - An Ancient Energy Source and Active Metabolic Regulator. Microb. Cell Fact. 2011, 10, 63. (4) Penczek, S.; Pretula, J.; Kubisa, P.; Kaluzynski, K.; Szymanski, R. Reactions of H3PO4 forming polymers. Apparently simple reactions leading to sophisticated structures and applications. Prog. Polym. Sci. 2015, 45, 44−70. (5) Zhou, H.; Ye, M.; Dong, J.; Han, G.; Jiang, X.; Wu, R.; Zou, H. Specific Phosphopeptide Enrichment with Immobilized Titanium Ion Affinity Chromatography Adsorbent for Phosphoproteome Analysis. J. Proteome Res. 2008, 7, 3957−3967. (6) Reimhult, E.; Höök, F.; Kasemo, B. Intact Vesicle Adsorption and Supported Biomembrane Formation from Vesicles in Solution: Influence of Surface Chemistry, Vesicle Size, Temperature, and Osmotic Pressure. Langmuir 2003, 19, 1681−1691. (7) Wang, F.; Zhang, X.; Liu, Y.; Lin, Z. Y. W.; Liu, B.; Liu, J. Profiling Metal Oxides with Lipids: Magnetic Liposomal Nanoparticles Displaying DNA and Proteins. Angew. Chem., Int. Ed. 2016, 55, 12063−12067. (8) Liu, B.; Ma, L.; Huang, Z.; Hu, H.; Wu, P.; Liu, J. Janus DNA Orthogonal Adsorption of Graphene Oxide and Metal Oxide Nanoparticles Enabling Stable Sensing in Serum. Mater. Horiz. 2018, 5, 65−69. (9) Pautler, R.; Kelly, E. Y.; Huang, P.-J. J.; Cao, J.; Liu, B.; Liu, J. Attaching DNA to Nanoceria: Regulating Oxidase Activity and Fluorescence Quenching. ACS Appl. Mater. Interfaces 2013, 5, 6820− 6825. (10) Liu, B.; Liu, J. DNA Adsorption by Magnetic Iron Oxide Nanoparticles and Its Application for Arsenate Detection. Chem. Commun. 2014, 50, 8568−8570. (11) Ghaemi, M.; Absalan, G. Study on the adsorption of DNA on Fe3O4 nanoparticles and on ionic liquid-modified Fe3O4 nanoparticles. Microchim. Acta 2014, 181, 45−53. (12) Ma, L.; Liu, B.; Huang, P.-J. J.; Zhang, X.; Liu, J. DNA Adsorption by ZnO Nanoparticles near Its Solubility Limit: Implications for DNA Fluorescence Quenching and DNAzyme Activity Assays. Langmuir 2016, 32, 5672−5680.
In summary, we have systematically studied the adsorption of various phosphate species on CeO2 NPs using both a DNAbased fluorescent displacement assay and ITC. In general, the longer the phosphate chain, the higher the adsorption affinity and the faster the kinetics of displacing DNA. The exception was STMP, which has a circular structure. Although STMP can tightly adsorb on CeO2, it cannot efficiently displace the adsorbed DNA likely because of kinetic issues related to its ring structure. At the same molar concentration of the total phosphate unit, the highest desorption was observed with PPi and STTP, and this can be explained by the higher displacing power of terminal phosphate and its balance with the polyvalent effect. ITC provided a more quantitative binding picture, and the similar total phosphate adsorption capacity for most species has indicated a tight lengthwise packing of these molecules. The difference in STMP and STTP for DNA displacement was used to monitor the hydrolysis of STMP. By capping the phosphate with organic groups, the adsorption affinity was only slightly affected by one cap but significantly affected by two caps. Therefore, DNA and lipids can tightly adsorb on CeO2 and other oxides relying on polyvalency.
Corresponding Author
*E-mail:
[email protected]. ORCID
Juewen Liu: 0000-0001-5918-9336 Notes
The authors declare no competing financial interest. F
DOI: 10.1021/acs.langmuir.8b01482 Langmuir XXXX, XXX, XXX−XXX
Article
Langmuir
Amounts of Phosphate Dissolved in Acidic Solutions. Clin. Oral Investig. 2005, 9, 203−207. (34) Yu, T.; Liu, B.; Liu, J. Adsorption of Selenite and Selenate by Metal Oxides Studied with Fluorescent DNA Probes for Analytical Application. J. Anal. Test. 2017, 1, 2. (35) Montini, T.; Melchionna, M.; Monai, M.; Fornasiero, P. Fundamentals and Catalytic Applications of CeO2-Based Materials. Chem. Rev. 2016, 116, 5987−6041. (36) Liu, B.; Huang, Z.; Liu, J. Boosting the oxidase mimicking activity of nanoceria by fluoride capping: rivaling protein enzymes and ultrasensitive F-detection. Nanoscale 2016, 8, 13562−13567. (37) Celardo, I.; Pedersen, J. Z.; Traversa, E.; Ghibelli, L. Pharmacological Potential of Cerium Oxide Nanoparticles. Nanoscale 2011, 3, 1411−1420. (38) Xu, C.; Bing, W.; Wang, F.; Ren, J.; Qu, X. Versatile Dual Photoresponsive System for Precise Control of Chemical Reactions. ACS Nano 2017, 11, 7770−7780. (39) Xu, C.; Qu, X. Cerium Oxide Nanoparticle: A Remarkably Versatile Rare Earth Nanomaterial for Biological Applications. NPG Asia Mater. 2014, 6, No. e90. (40) Ta, H. T. T.; Tieu, A. K.; Zhu, H.; Yu, H.; Ta, T. D.; Wan, S.; Tran, N. V.; Le, H. M. Chemical Origin of Sodium Phosphate Interactions on Iron and Iron Oxide Surfaces by First Principle Calculations. J. Phys. Chem. C 2018, 122, 635−647. (41) Luengo, C.; Brigante, M.; Avena, M. Adsorption Kinetics of Phosphate and Arsenate on Goethite. A Comparative Study. J. Colloid Interface Sci. 2007, 311, 354−360. (42) Antelo, J.; Avena, M.; Fiol, S.; López, R.; Arce, F. Effects of pH and Ionic Strength on the Adsorption of Phosphate and Arsenate at the Goethite-Water Interface. J. Colloid Interface Sci. 2005, 285, 476− 486. (43) 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, 2755−2794. (44) Varghese, N.; Mogera, U.; Govindaraj, A.; Das, A.; Maiti, P. K.; Sood, A. K.; Rao, C. N. R. Binding of DNA Nucleobases and Nucleosides with Graphene. ChemPhysChem 2009, 10, 206−210. (45) Zhang, Z.; Morishita, K.; Lin, W. T. D.; Huang, P.-J. J.; Liu, J. Nucleotide Coordination with 14 Lanthanides Studied by Isothermal Titration Calorimetry. Chin. Chem. Lett. 2018, 29, 151−156. (46) Lyngsie, G.; Penn, C. J.; Hansen, H. C. B.; Borggaard, O. K. Phosphate Sorption by Three Potential Filter Materials as Assessed By isothermal Titration Calorimetry. J. Environ. Manage. 2014, 143, 26−33. (47) Huang, R.; Lau, B. L. T. Biomolecule-nanoparticle interactions: Elucidation of the thermodynamics by isothermal titration calorimetry. Biochim. Biophys. Acta 2016, 1860, 945−956. (48) Lack, S.; Dulong, V.; Picton, L.; Le Cerf, D.; Condamine, E. High-Resolution Nuclear Magnetic Resonance Spectroscopy Studies of Polysaccharides Crosslinked by Sodium Trimetaphosphate: A Proposal for the Reaction Mechanism. Carbohydr. Res. 2007, 342, 943−953. (49) Kura, G.; Nakashima, T.; Oshima, F. Study of the Acidic Hydrolysis of Cyclic Trimetaphosphate by Liquid Chromatography. J. Chromatogr. A 1987, 219, 385−391. (50) Ferris, J. P. Catalysis and Prebiotic Synthesis. Rev. Mineral. Geochem. 2005, 59, 187−210. (51) Wagstaffe, M.; Thomas, A. G.; Jackman, M. J.; Torres-Molina, M.; Syres, K. L.; Handrup, K. An Experimental Investigation of the Adsorption of a Phosphonic Acid on the Anatase TiO2(101) Surface. J. Phys. Chem. C 2016, 120, 1693−1700.
(13) Wahab, R.; Kim, Y.-S.; Hwang, I. H.; Shin, H.-S. A NonAqueous Synthesis, Characterization of Zinc Oxide Nanoparticles and Their Interaction with DNA. Synth. Met. 2009, 159, 2443−2452. (14) Das, S.; Chatterjee, S.; Pramanik, S.; Devi, P. S.; Kumar, G. S. A New Insight into the Interaction of ZnO with Calf Thymus DNA through Surface Defects. J. Photochem. Photobiol., B 2018, 178, 339− 347. (15) Zhang, X.; Wang, F.; Liu, B.; Kelly, E. Y.; Servos, M. R.; Liu, J. Adsorption of DNA Oligonucleotides by Titanium Dioxide Nanoparticles. Langmuir 2014, 30, 839−845. (16) Monti, S.; Walsh, T. R. Molecular Dynamics Simulations of the Adsorption and Dynamical Behavior of Single DNA Components on TiO2. J. Phys. Chem. C 2011, 115, 24238−24246. (17) Leung, A. K. L. Poly(Adp-Ribose): An Organizer of Cellular Architecture. J. Cell Biol. 2014, 205, 613−619. (18) Chow, W. Y.; Rajan, R.; Muller, K. H.; Reid, D. G.; Skepper, J. N.; Wong, W. C.; Brooks, R. A.; Green, M.; Bihan, D.; Farndale, R. W.; Slatter, D. A.; Shanahan, C. M.; Duer, M. J. NMR Spectroscopy of Native and in vitro Tissues Implicates PolyADP Ribose in Biomineralization. Science 2014, 344, 742−746. (19) Dorozhkin, S. V.; Epple, M. Biological and Medical Significance of Calcium Phosphates. Angew. Chem., Int. Ed. 2002, 41, 3130−3146. (20) Wang, L.; Zhang, Z.; Liu, B.; Liu, Y.; Lopez, A.; Wu, J.; Liu, J. Interfacing DNA Oligonucleotides with Calcium Phosphate and Other Metal Phosphates. Langmuir 2017, DOI: 10.1021/acs.langmuir.7b03204. (21) Kim, S. K.; Lee, D. H.; Hong, J.-I.; Yoon, J. Chemosensors for Pyrophosphate. Acc. Chem. Res. 2009, 42, 23−31. (22) Xu, C.; Liu, Z.; Wu, L.; Ren, J.; Qu, X. Nucleoside Triphosphates as Promoters to Enhance Nanoceria Enzyme-Like Activity and for Single-Nucleotide Polymorphism Typing. Adv. Funct. Mater. 2014, 24, 1624−1630. (23) Liu, B.; Liu, J. Surface Modification of Nanozymes. Nano Res. 2017, 10, 1125−1148. (24) Liu, J.; Cao, J.; Hu, Y.; Han, Y.; Zhou, J. Adsorption of phosphate ions from aqueous solutions by a CeO2 functionalized Fe3O4 @SiO2 core-shell magnetic nanomaterial. Water Sci. Technol. 2017, 76, 2867−2875. (25) Xu, J.; Luu, L.; Tang, Y. Phosphate Removal Using AluminumDoped Magnetic Nanoparticles. Desalin. Water Treat. 2017, 58, 239− 248. (26) You, X.; Farran, A.; Guaya, D.; Valderrama, C.; Soldatov, V.; Cortina, J. L. Phosphate Removal from Aqueous Solutions Using a Hybrid Fibrous Exchanger Containing Hydrated Ferric Oxide Nanoparticles. J. Environ. Chem. Eng. 2016, 4, 388−397. (27) Liu, B.; Liu, J. Comprehensive Screen of Metal Oxide Nanoparticles for DNA Adsorption, Fluorescence Quenching, and Anion Discrimination. ACS Appl. Mater. Interfaces 2015, 7, 24833− 24838. (28) Li, F.; Pei, H.; Wang, L.; Lu, J.; Gao, J.; Jiang, B.; Zhao, X.; Fan, C. Nanomaterial-Based Fluorescent DNA Analysis: A Comparative Study of the Quenching Effects of Graphene Oxide, Carbon Nanotubes, and Gold Nanoparticles. Adv. Funct. Mater. 2013, 23, 4140−4148. (29) Lu, C.-H.; Yang, H.-H.; Zhu, C.-L.; Chen, X.; Chen, G.-N. A Graphene Platform for Sensing Biomolecules. Angew. Chem., Int. Ed. 2009, 48, 4785−4787. (30) Wang, Y.; Tang, L.; Li, Z.; Lin, Y.; Li, J. In situ simultaneous monitoring of ATP and GTP using a graphene oxide nanosheet-based sensing platform in living cells. Nat. Protoc. 2014, 9, 1944−1955. (31) Wang, H.; Yang, R.; Yang, L.; Tan, W. Nucleic Acid Conjugated Nanomaterials for Enhanced Molecular Recognition. ACS Nano 2009, 3, 2451−2460. (32) Liu, B.; Salgado, S.; Maheshwari, V.; Liu, J. DNA Adsorbed on Graphene and Graphene Oxide: Fundamental Interactions, Desorption and Applications. Curr. Opin. Colloid Interface Sci. 2016, 26, 41− 49. (33) Attin, T.; Becker, K.; Hannig, C.; Buchalla, W.; Wiegand, A. Suitability of a Malachite Green Procedure to Detect Minimal G
DOI: 10.1021/acs.langmuir.8b01482 Langmuir XXXX, XXX, XXX−XXX
