Letter Cite This: Anal. Chem. 2019, 91, 7500−7504
pubs.acs.org/ac
Facile Counting of Ligands Capped on Nanoparticles via a Titration Chip of Moving Reaction Boundary Electrophoresis Qiang Zhang,‡,†,§ Weiwen Liu,† Muhammad Idrees Khan,‡,†,§ Cunhuai Wang,‡,†,§ Guoqing Li,‡,†,§ Hua Xiao,§ Yuxing Wang,*,∥ and Chengxi Cao*,‡,†,§ ‡
Shanghai Sixth People’s Hospital East, Shanghai Jiao Tong University Medical School, Shanghai 201306, China Department of Instrument Science and Engineering, School of Electronic Information and Electrical Engineering, Shanghai Jiao Tong University, Shanghai 200240, China § School of Life Science and Biotechnology, State Key Laboratory of Microbial Metabolism, Shanghai Jiao Tong University, Shanghai 200240, China ∥ School of Physics and Astronomy, Shanghai Jiao Tong University, Shanghai 200240, China
Downloaded via 79.110.18.71 on August 13, 2019 at 11:21:02 (UTC). See https://pubs.acs.org/sharingguidelines for options on how to legitimately share published articles.
†
S Supporting Information *
ABSTRACT: Absolute quantification of ligand capped on the surface of nanoparticles (NPs) has faced a great challenge without the use of complex inner standards (CIS). Herein, we proposed a facile electrophoresis titration (ET) model, designed an ET device, and developed a relevant method for counting the ligand on NPs without the use of CIS, based on moving reaction boundary (MRB). Furthermore, we conducted the relevant ET runs by using 3-mercaptopropionic acid (MPA) and quantum dots (QDs) as the model ligand and NPs, respectively. The experiments revealed that the ligand content of 1518 ± 295 obtained via an ET was close to the one of 1408 ± 117 determined via NMR, validating the ET model. Moreover, the experiments showed fair stability (RSD < 5.62%) and simplicity of ET without the use of CIS. Evidently, the ET model opens a window for facile assay of ligand capped on NPs.
S
(CIS), complex sample pretreatment, and expensive instrumentation.19,21,28,29 Herein, we outlined the electrophoresis titration (ET) model of counting ligands capped on NPs by relying on the protein ET,30,31 enzyme catalysis ET,32 and photocatalysis ET33 via moving reaction boundary (MRB). Figure 1 and Figure S1 show the NPs-ET model for counting the ligands on the NPs surface. In the ET model, 3-mercaptopropionic acid (MPA) and CdSe/ZnS core−shell QDs are, respectively, used as the model ligand and NPs uniformly immobilized in the channel via polyacrylamide gel (PAG), and the hydroxyl ion in the cathode well is applied as the titrant.
ince nanoparticles (NPs) have fantastic properties, e.g., size,1 shape,2 and electronic,3 optical,4 and surface modifications,5 and are increasingly used in clinical diagnosis,6 drug delivery,7 sensors,8 electronic devices,9 and materials with unique properties,10 numerous efforts have been devoted to the techniques of NPs characterization.11,12 For example, dynamic light scattering,13 NMR,14 Fourier transform-infrared (FT-IR),15 electron microscopy,16 X-ray photoelectron spectroscopy (XPS),17 and ultracentrifugation18 have been developed for identification of NPs. Many analytical methods, e.g., NMR,19 inductively coupled plasma mass spectrometry (ICPMS),20 XPS,21 thermogravimetric analysis (TGA),22 and electrochemistry,23 have been proposed for understanding NPs properties of solubility,24 specificity,25 absorption ability,26 charge,27 and electrochemical activity.28 However, a facile method has been rarely proposed for absolute quantification of the ligand of NPs without the use of complex inner standards © 2019 American Chemical Society
Received: March 1, 2019 Accepted: May 27, 2019 Published: May 27, 2019 7500
DOI: 10.1021/acs.analchem.9b01098 Anal. Chem. 2019, 91, 7500−7504
Letter
Analytical Chemistry
MPA (see eq S15). From eq 3, we have the real ligand number on a single NP (Nreal) after having the content of NPs (cNPs), c̅ Nreal = MPA c NPs
(4)
Clearly, the surface coverage of the ligand (λreal) on the NP surface can be approximately obtained via eq 5 after detecting the NP size (rNPs) and the length of ligand (rMPA, the computational spherical layer is set in the middle length of ligand (see the Supporting Information)) λreal =
Under the electric field, the hydroxyl ion is driven into the channel and titrates the ligand of MPA on the QDs surface, resulting in the following boundary titration, NP@S(CH 2)2 COOH + OH− (I)
Reaction I creates a sharp anodic-moving acid−base neutralized MRB, which can be observed via the phenolphthalein indicator in the channel. In the model, there is always a balance between the number of ligands titrated and the equivalent amount of titrant consumed 28−31 (see the Supporting Information). Figure 1B shows the simple calculation model of the MPA ligand capped on NPs surface. The boundary movement can be described by the equation, dNPs = vMRB(t1 − t0)
(1)
where dNPs is the boundary migration distance from the initial time of t0 and the end time of t1 and vMRB is the velocity of reaction boundary. Assuming a constant hydroxyl ion flux in the two channels of blank (NPs-free) and NPs, we can, respectively, acquire the two MRB movement distances of blank (dBlank) and NPs (dNPs) in the given time (t). The relative difference (Δdrel) of MRB movement of the blank (NPs-free) and NPs runs is Δdrel = dBlank − dNPs
(2)
The content of ligand capped on NPs has the following relation if the content of H+ change is ignored (see eqs S8− S14), Δdrel N c c̅ = MPA = real NPs dNPs cOH− cOH−
4π (rNPs + rMPA /2)2
(5)
Equations 1−5 are very simple but have great significance to the research on micronano science and technology because they indicate that the ligand number (Nreal) and surface coverage (λreal) per NP can be easily achieved via monitoring the ET boundary movement without the use of CIS, complex pretreatment, and expensive equipment as shown below. A titration microfluidic device was designed (Figure 1C) for the experiment on NPs ET. The microdevice had an ET chip, a power supply for the electric field of ET run, a digital microscope for image capture, and a computer for image storage and processing. The ET chip had a cathodic well, two channels (30 mm × i.d. 600 μm) and an anodic well. With the microdevice, the NPs ET run was conducted as described below. First, the NPs capped with ligand of MPA or L-Cys were immobilized in one channel via cross-linked polyacrylamide gel (PAG, 15% T, 4% C) mixed with background electrolyte of 0.1 M KCl, and another channel was filled with the blank gel as control. Second, the anode and cathode wells were injected with 0.1 M KCl and 0.1 M Tris-HCl buffer (pH 10) with 0.1 M KCl, respectively (see the Supporting Information). Third, the NPs ET was conducted by using an electric field of 40 V with a current about 0.4 mA, and the MRB migration distances for the blank and NPs ET runs were d0 and d1, respectively (Figure 1B). Finally, the total concentration of ligand (MPA) could be measured from the ET runs, and the ligand number and coverage of MPA capped on a certain diameter QD could be estimated. To reveal the dynamics of the boundary titration in Reaction (I), the computer simulation was developed based on the diffusion and convection model via the COMSOL software (see the Supporting Information). Figure 2 revealed the titration of MPA ligand capped on CdSe/ZnS QDs via the ET method. Figure 2A showed the 0.1 mg mL−1 QDs on the electromigration boundary of the hydroxyl ion in the PAG-filled channel. After the run of ET, we could achieve the accurate boundary movements (Figure 2A) and the relations between the boundary motion distance (d) and the running time (t) for the NPs ET under different contents of QDs (Figure 2B). The experiments in Figure 2B indicated a series of good linear fittings within the given running time (Table S1). Figure 2C reveals the boundary motion distance in the blank was significantly longer than that in the gels with different contents of QDs. Evidently, the ET motion distances calculated from the analytical model and numerical simulation under the same experimental conditions of Figure 2C were consistent with experiments of QDs ET (Figure 2D). All the comparative data in Figure 2 and Table S1 indicated the retardation of MPA capped on QDs to the
Figure 1. Schematic diagram of ET for counting ligands capped on NPs. (A) Boundary movement distance (d) from the initial time of t0 to the end time of t1; (B) model of NPs ET based on boundary movements in the blank run without NPs in the channel (dblank) and in the real run with NPs immobilized via PAG in the channel (dQDs) at running times of t1; and (C) whole detection system containing the ET chip, power supply, backlight sheet, digital microscope, and computer. The insert gives the details of the ET chip with the cathode, cathode well, channels, anode well, and anode.
↔ H 2O + NP@S(CH 2)2 COO−
Nreal
(3)
−
where cOH is the content of hydroxyl ion in the cathode buffer, cM ̅ PA is the constitute concentration of MPA, including the dissociated species of MPA− and undissociated ones of 7501
DOI: 10.1021/acs.analchem.9b01098 Anal. Chem. 2019, 91, 7500−7504
Letter
Analytical Chemistry
Figure 2. (A) Boundary motion in ET formed with 0.1 mg mL−1 QDs in the channel and 0.1 M Tris-HCl (pH 10) in the cathode well at different running times; (B) relationships between MRB migration distance (d) and running time (t) during the ET runs with different QDs contents; (C) MRB movements with the different content of QDs and the given running time of 5 min; and (D) theoretical, experimental, and simulative relations between MRB movement difference (Δdrel) and QDs contents. Conditions: titrant solution with 0.1 M Tris-HCl buffer (pH 10) and 0.1 M KCl in the cathode well, 4 μL of gel solution (mixed with QDs and 0.1 M KCl) injected into the channel, 0.1 M KCl in the anodic well, 40 V, and 25 °C. In the theoretical and simulative calculations, the pKa of ligands in QDs was set to 7 and the MRB position was set at pH 8.5. All the runs were repeated three times. The other conditions are given in the Supporting Information.
Figure 3. (A) Comparisons on ligand contents of MPA (cM ̅ PA) on CdSe/ZnS QDs with different contents detected by NPs ET and NMR; (B) comparisons on ligand contents of L-cysteine (cC̅ ys) on AuNPs with different diameters measured by NPs ET and NMR; (C) comparisons on Nreal values of MPA ligand on CdSe/ZnS QDs obtained by the theoretical calculation, simulation, ET and NMR assays; (D) comparisons on Nreal values of Cys ligand on AuNPs obtained by the ET run and NMR methods; (E) comparisons on λreal values of MPA ligand on CdSe/ZnS QDs obtained by the theoretical calculation, simulative computation, ET run, and NMR experiment; and (F) comparisons on λreal values of the Cys ligand on AuNPs detected by ET and NMR. All the ET runs were repeated three times. The analytical and simulative error bars in panel C and E were calculated based on the experimental and simulative data in Figure 2D. The experimental conditions were the same as those in Figure 2.
boundary motion, briefly validating the NPs ET model and analytical method in Figure 1A,B. The ligand contents could be facilely calculated by using the analytical model of Figure 1B based on the boundary movement distances. In Figure 2B, the boundary distances of a 5 min ET run for 0, 0.1, 0.2, 0.3, 0.4, and 0.5 mg mL−1 QD concentrations were, respectively, 6673 ± 48, 6315 ± 109, 5929 ± 189, 5606 ± 137, 5328 ± 139, and 5184 ± 139 μm. The corresponding contents of MPA ligand calculated from the experimental data were 0, 5.7 ± 2.4, 12.6 ± 3.6, 19.1 ± 2.8, 25.3 ± 2.9, and 28.8 ± 2.8 μM (Figure 3A), respectively. To verify the ET of MPA on QDs, the MPA contents in the same samples were further determined by the absolute quantitation method of NMR based on the ratio of triplet of methylene in the MPA and the singlet of maleic acid (Figure S7). Figure 3A revealed that results of NMR determination were in a good agreement with the ones of QDs ET, manifesting the NPs ET model of Figure 1B and eq 3 evidently. To further demonstrate the ET model and eq 3, the ligand of L-Cys capped on AuNPs were titrated with different diameters of 5, 10, and 20 nm. Figure 3B showed that (i) the total ligand contents of L-Cys of 5, 10, and 20 nm AuNPs quantified via the ET method were, respectively, 21.6 ± 8.0, 49.5 ± 5.9 and 21.3 ± 4.9 μM; (ii) the corresponding values of NMR were, respectively, 14.7 ± 0.16, 44.1 ± 1.3, and 18.5 ± 0.3 μM; and (iii) there was no significant difference between the values of ET and the ones of NMR (P > 0.05). Evidently, these comparative results in Figure 3B validated the model and eq 3. To manifest the ET model and eq 4, we conducted the theoretical computation, simulation and experimental analysis
on the ligand number capped on a single QD (Nreal value). Figure 3C and Table S2 displayed that (i) the analytical and simulative values of Nreal were respectively 1424 ± 62 and 1382 ± 54; (ii) the ET and NMR experimental values of Nreal were 1518 ± 295 and 1408 ± 117, respectively The comparisons in Figure 3C and Table S2 indicated that (i) the Nreal value of ET was in concordance with the one of NMR; (ii) the ET experiment of Nreal was close to the analytical and simulative results based on the QDs ET model; and (iii) the NMR and ET experiments indicated the validity of analytical model in Figure 1B and S1 and simulative model. The indication of Figure 3C was directly demonstrated by the comparative results in Figure 3D showing that the Nreal values of AuNPs ET were in coincidence with the ones of AuNPs NMR. Evidently, the comparisons in Figure 3C and 3D certified the NPs ET model in Figure 1 and eq 4. To manifest eq 5, we performed the calculation of NPs λreal based on the ET boundary motion distances. Figure 3E and Table S2 further exhibited that (i) the λreal values of ET theoretical analysis and simulation were 2.97 ± 0.13 and 2.88 ± 0.11, respectively, and (ii) the λreal values of ET and NMR were 3.16 ± 0.61 and 2.92 ± 0.24, respectively. Clearly, the comparisons in Figure 3E and Table S2 implied that (i) the NMR value was highly near to the one of ET experiment and (ii) the ET determination was very close to the ET theoretical 7502
DOI: 10.1021/acs.analchem.9b01098 Anal. Chem. 2019, 91, 7500−7504
Letter
Analytical Chemistry
be directly measured by distances. Furthermore, we developed the microfluidic device and method of ET and performed the real determination of ligand number on NPs. The experiments demonstrated that the quantitative analysis of ligands caged on the surface of NPs could be achieved by the NPs ET method. Particularly, the experiments showed for the first time that the coverage of ligand on NPs could be simply detected based on a convenient measurement of boundary motion in an ET run without use of CIS. The developed ET microdevice and method were extremely facile and have low cost and low labor in contrast to the classic NMR method, which uses CIS compounds and deuterium oxide. The developed method has potential for the quantitative analysis of ligand capped on micronanomaterials, which are facing great challenges.
and simulative values. The similar results could be found in the detection of L-Cys capped on the AuNPs experiments (Figure 3F). Hence, the NMR results further proved the ET experiment, and the NMR and ET experiments validated the theoretical and simulative calculations and finally the ET model and eq 5. Figure 3A,B,D,F showed that the ligand values obtained via ET method were always more than that via the NMR method, implying the existence of system deviation between the two methods. The system deviation might be induced by the following reasons. In the ET method, the sample of NPs was directly immobilized in the channel of ET without loss of ligand capped on the NP surface. Whereas, in the sample pretreatment of NPs for NMR analysis, the NP samples had to be repeatedly centrifugated and washed by ultrapure water to remove impure compounds, dried by freezing drying, and finally redistributed by deuterium oxide, leading to the possible loss of ligand from the surface of NPs. In an ET, the NPs added in the gel may change the liquid environment causing the tensile force to increase, which may slow down the boundary migration and finally may enlarge the content of the ligand. In addition, there might be some residue of free ligand in NPs sample which could be sensed by the ET method (due to no pretreatment), but not the NMR, as discussed above and indicated by the comparisons of theoretical, simulative, and experimental results in Figure 3C,E. The developed NPs ET method had special merits. First, the NPs ET method had high simplicity. The ET model was extremely simple as shown in Figure 1A. The microfluidic device was facile (Figure 1C). In the ET method, the NPs was simply immobilized by PAG without other operations (Figure 1A,B), and the quantitation could be easily obtained by just a measurement of the boundary motion without use of CIS (Figure 1B and Figure S1). On the contrary, the NMR method was extremely complex.18,20 The NMR model was a little complex as shown in eq S28 and ref S7 in the Supporting Information. The NMR instrument was extremely complicated, and ultrapure compounds of CIS had to be used for obtaining accurate data.18 In addition, the freeze drying, highspeed centrifugation, and ultrapure deuterium oxide had to be used for sample pretreatment shown in section S2.4.5 of the Supporting Information, indicating complex performance of NMR.18,20 Evidently, the developed method was extremely low in cost in contrast to NMR and CE thanks to expensive instruments (e.g., NMR, high-speed centrifugation, and freeze drying as well as CE29) and reagents (e.g., ultrapure CIS compounds and deuterium oxide).18,20 Third, the ET method was much lower labor, whereas the NMR and CE methods were highly labor-consuming.18,20,29 Fourth, the ET method had fair accuracy as shown in Figure 3 and Tables S3 and S4. Fifth, the NPs ET had fair stability due to the intraday and interday RSDs within 5.62% as revealed in Tables S3 and S4. Of course, the ET method had the following limitations. First, the method had only a fair precision (Tables S3 and S4) thanks to poor control of the temperature in an ET run. Second, the method always suffered from interference of residual free ligands in a NPs sample due to no pretreatment of sample. In summary, we proposed the ET model and theory for counting the ligand on NPs without the use of CIS based on MRB. The theoretical and simulation models mechanistically revealed the acid−base titration could convert to a controllable visual MRB, and the titrant of unknown concentration could
■
ASSOCIATED CONTENT
* Supporting Information S
The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.analchem.9b01098. NPs ET model, experimental section, QDs size detection, NPs immobilization and titration conditions, AuNPs concentration calculation, ligands number and coverage calculation, and performance of NPs ET (PDF)
■
AUTHOR INFORMATION
Corresponding Authors
*E-mail:
[email protected]. *E-mail:
[email protected]. ORCID
Hua Xiao: 0000-0002-2831-0436 Chengxi Cao: 0000-0002-3873-5112 Author Contributions
Q.Z. and W. L. contributed equally. Notes
The authors declare no competing financial interest.
■
ACKNOWLEDGMENTS The authors are grateful for financial support from the NSFC (Grants 21475086, 21675067, and 31727801), the National Research and Development program (Grant 2017YFC1200204), and the National High-Tech R&D Program of China (863 Program, Grant 2014AA020545).
■
REFERENCES
(1) Choi, H. S.; Liu, W. H.; Misra, P.; Tanaka, E.; Zimmer, J. P.; Ipe, B. I.; Bawendi, M. G.; Frangioni, J. V. Nat. Biotechnol. 2007, 25, 1165−1170. (2) Manna, L.; Milliron, D. J.; Meisel, A.; Scher, E. C.; Alivisatos, A. P. Nat. Mater. 2003, 2, 382−385. (3) Conklin, D.; Nanayakkara, S.; Park, T. H.; Lagadec, M. F.; Stecher, J. T.; Therien, M. J.; Bonnell, D. A. Nano Lett. 2012, 12, 2414−2419. (4) Olson, J.; Dominguez-Medina, S.; Hoggard, A.; Wang, L. Y.; Chang, W. S.; Link, S. Chem. Soc. Rev. 2015, 44, 40−57. (5) Mout, R.; Moyano, D. F.; Rana, S.; Rotello, V. M. Chem. Soc. Rev. 2012, 41, 2539−2544. (6) Han, M. Y.; Gao, X. H.; Su, J. Z.; Nie, S. M. Nat. Biotechnol. 2001, 19, 631−635. (7) Ferrari, M. Nat. Nanotechnol. 2008, 3, 131−132. (8) Kneipp, K.; Kneipp, H.; Itzkan, I.; Dasari, R. R.; Feld, M. S. Chem. Rev. 1999, 99, 2957−2976.
7503
DOI: 10.1021/acs.analchem.9b01098 Anal. Chem. 2019, 91, 7500−7504
Letter
Analytical Chemistry (9) Cui, Y.; Lieber, C. M. Science 2001, 291, 851−853. (10) Collier, C. P.; Vossmeyer, T.; Heath, J. R. Annu. Rev. Phys. Chem. 1998, 49, 371−404. (11) Hinterwirth, H.; Kappel, S.; Waitz, T.; Prohaska, T.; Lindner, W.; Lämmerhofer, M. ACS Nano 2013, 7, 1129−1136. (12) Boles, M. A.; Ling, D.; Hyeon, T.; Talapin, D. V. Nat. Mater. 2016, 15, 141−153. (13) Dai, Q.; Liu, X.; Coutts, J.; Austin, L.; Huo, Q. J. Am. Chem. Soc. 2008, 130, 8138−8139. (14) Lamm, J. H.; Niermeier, P.; Mix, A.; Chmiel, J.; Neumann, B.; Stammler, H. G.; Mitzel, N. W. Angew. Chem., Int. Ed. 2014, 53, 7938−7942. (15) Juárez, R.; Parker, S. F.; Concepción, P.; Corma, A.; García, H. Chem. Sci. 2010, 1, 731−738. (16) Azubel, M.; Koivisto, J.; Malola, S.; Bushnell, D.; Hura, G. L.; Koh, A. L.; Tsunoyama, H.; Tsukuda, T.; Pettersson, M.; Häkkinen, H.; Kornberg, R. D. Science 2014, 345, 909−912. (17) Shi, W.; Zhang, B. S.; Lin, Y. M.; Wang, Q.; Zhang, Q.; Su, D. S. ACS Catal. 2016, 6, 7844−7854. (18) Bekdemir, A.; Stellacci, F. Nat. Commun. 2016, 7, 13121. (19) Liu, X.; Yu, M.; Kim, H.; Mameli, M.; Stellacci, F. Nat. Commun. 2012, 3, 1182. (20) Montoro Bustos, A. R.; Encinar, J. R.; Fernández-Argüelles, M. T.; Costa-Fernández, J. M.; Sanz-Medel, A. Chem. Commun. 2009, 21, 3107−3109. (21) Walker, B. J.; Nair, G. P.; Marshall, L. F.; Bulović, V.; Bawendi, M. G. J. Am. Chem. Soc. 2009, 131, 9624−9625. (22) Mansfield, E.; Tyner, K. M.; Poling, C. M.; Blacklock, J. L. Anal. Chem. 2014, 86, 1478−1484. (23) Zhou, Y. G.; Rees, N. V.; Compton, R. G. Angew. Chem. 2011, 123, 4305−4307. (24) Li, L.; Chakrabarty, S.; Jiang, J.; Zhang, B.; Ober, C.; Giannelis, E. P. Nanoscale 2016, 8, 1338−1343. (25) McIntosh, C. M.; Esposito, E. A.; Boal, A. K.; Simard, J. M.; Martin, C. T.; Rotello, V. M. J. Am. Chem. Soc. 2001, 123, 7626− 7629. (26) Zheng, M.; Davidson, F.; Huang, X. Y. J. Am. Chem. Soc. 2003, 125, 7790−7791. (27) Kalsin, A. M.; Kowalczyk, B.; Smoukov, S. K.; Klajn, R.; Grzybowski, B. A. J. Am. Chem. Soc. 2006, 128, 15046−15047. (28) Rojas, M. T.; Koeniger, R.; Stoddart, J. F.; Kaifer, A. E. J. Am. Chem. Soc. 1995, 117, 336−343. (29) Vorácǒ vá, I.; Klepárník, K.; Lišková, M.; Foret, F. Electrophoresis 2015, 36, 867−874. (30) Zhang, L. X.; Cao, Y. R.; Xiao, H. X.; Liu, X. P.; Liu, S. R.; Meng, Q. H.; Fan, L. Y.; Cao, C. X. Biosens. Bioelectron. 2016, 77, 284−291. (31) Wang, C. H.; Zhang, Q.; Liu, X. P.; Li, G. Q.; Kong, H.; Khan, M. I.; Xiao, H.; Wang, X. Y.; Liu, W. W.; Cao, C. X. Lab Chip 2019, 19, 484−492. (32) Cao, X. Y.; Kong, F. Z.; Zhang, Q.; Liu, W. W.; Liu, X. P.; Li, G. Q.; Zhong, R.; Fan, L. Y.; Xiao, H.; Cao, C. X. Lab Chip 2018, 18, 1758−1766. (33) Li, W. L.; Kong, F. Z.; Zhang, Q.; Liu, W. W.; Kong, H.; Liu, X. P.; Khan, M. I.; Wahid, A.; Saud, S.; Xiao, H.; Cao, C. X.; Fan, L. Y. Anal. Chem. 2018, 90, 6710−6717.
7504
DOI: 10.1021/acs.analchem.9b01098 Anal. Chem. 2019, 91, 7500−7504
