Reagents and materialsAcetonitrile (ACN), dimethyl sulfoxide (DMSO), HCl, CH3COOH (suprapur®), KH2PO4, and normal human serum were purchased from Merck (Darmstadt, Germany). Crystalline l-Trp (≥ 98%), l-Kyn (≥ 98%), BSA, trichloroacetic acid (TCA), NaCl, KCl, NaOH, Na2HPO4 and LC–MS grade HCOOH, HCOONH4, CH3COONH4, Superclean™ PSA SPE Bulk Packaging, Discovery® DSC-18 SPE Bulk Packaging, and AC were obtained from Sigma-Aldrich (St Louis, MO, USA). PBS solution (0.1 mol·L−1) was prepared as described in40 and further diluted 10 times with ultrapure water produced by the Millipore Direct Q 3UV water purification system (Millipore, UK). pH of the PBS solution was adjusted using HCl or NaOH.Stock solutions (1 g·L−1) of Trp and Kyn were prepared by dissolving a reagent in DMSO, stored at − 20 °C, and used for up to three freeze and thaw cycles. Working solutions at intermediate concentrations were prepared daily by dilution in 0.01 mol·L−1 PBS. Calibration standards of Trp and Kyn were prepared by diluting stock solutions with the appropriate medium: 0.01 mol·L−1 PBS, surrogate matrix (BSA solution at 20 mg·mL−1 prepared in 0.01 mol·L−1 PBS), or serum matrix (normal human serum diluted 2 times in 0.01 mol·L−1 PBS). Standards prepared in a medium containing BSA and serum were deproteinized with 30% (w/v) aqueous TCA (100 µL per 500 µL of samples) before analysis.HPLC–ESI–MS analysisQuantification of Trp and Kyn in experimental samples was carried out using an Agilent Technologies 1200 Series high-performance liquid chromatograph equipped with autosampler (G1329A), quaternary pump with vacuum degasser (G1311A), column thermostat (G1316A) and connected to an Agilent Technologies quadrupole mass spectrometer (G6120A) with API-ESI ionization source. Chromatography was carried out using Zorbax Eclipse Plus-C18 RRHT analytical column (4.6 × 100 mm, 3.5 μm) protected by guard column (2.1 × 12.5 mm, 5 μm) both purchased from Agilent Technologies. For separation, the mobile phase consisted of solvent A (0.01% (v/v) CH3COOH in water) and solvent B (ACN) was used. The following gradient program was applied: 0–10 min: 5% solvent B; 10–12 min: 5–20% solvent B and 12–18 min: 20–40% solvent B in A (post run: 4 min). Other parameters were set as follows: mobile phase flow rate: 0.5 mL·min−1, column temperature: 25 °C, injection volume: 10 µL.Ionization was achieved by adopting settings from 41. Analytes were detected using positive polarity and SIM modes. Up to 7 min, the ions of 209 m/z ([Kyn + H]+), and then ions of 205 m/z ([Trp + H]+) were monitored. Fragmentor was set at 80 V. Position of Kyn and Trp peaks on the chromatogram was presented in Fig. 4.Figure 4Example of a chromatogram acquired for a mixture of kynurenine and tryptophan standards by HPLC–ESI–MS.Adsorption evaluationBatch adsorption experiments were used to evaluate the capture of Trp and Kyn by poly(4VP-co-14DMB) and poly(4VP-co-TRIM) in 0.01 mol·L−1 PBS solution (neat solvent) and BSA matrix (surrogate matrix). Copolymers were obtained and characterized according to the procedure described in19.For experiments, 50 mg of the polymer was mixed with 1 mL of the evaluated matrix fortified with the target compound at 10 µmol·L−1. Then, the samples were agitated using an IKA MS 5 shaker (15,000 rpm, room temperature), and centrifuged using the Eppendorf 5415R centrifuge (16,110×g, 10 min). Before analysis, the supernatants were filtered on 13 mm, 0.22 µm regenerated cellulose syringe filters (for experiments carried out in PBS matrix) or mixed with 30% (w/v) TCA and centrifuged (16,110×g, 10 min) to remove proteins (for BSA matrix). The series of experiments was accomplished by changing the pH (3–8.5) of the PBS solution (adsorption time: 5 h; room temperature). The studied pH range was selected considering the limits for the analytical column used for chromatographic analysis. Kinetic experiments were carried out for the adsorption time from 2 min to 24 h (in a matrix solution of pH 7 and 8.5 for Trp and Kyn, respectively). The isotherms were evaluated by changing the initial concentration of the target molecule within the range of 2–300 µmol·L−1 in a medium of pH 7.0 and 8.5 for Trp and Kyn, respectively. The applied adsorption times during isotherms determination are listed in Table 1. Concentrations of Trp and Kyn in the supernatants were determined by HPLC–ESI–MS. Control HPLC–ESI–MS measurements have confirmed that the PBS and BSA matrices were analyte-free. Each evaluated matrix solution was fortified with Trp or Kyn at 10 µmol·L−1 before running a set of experiments, and their initial concentrations were determined.The equilibrium adsorption capacities (Qe) of target compounds were calculated using Eq. (1) and expressed42 in µmol·g−1:$${Q}_{e}=\frac{{(C}_{i}-{C}_{e})\cdot V}{m}$$
(1)
where Ci [µmol·L−1] means the initial concentration of analytes, Ce [µmol·L−1] refers to the analyte concentration remaining in the solution after sorption, V [mL] is the solution volume, and m [g] is the polymer amount.The partitioning coefficient (Kd) calculated based on the Eq. (2)20:$${K}_{d}=\frac{{C}_{s}}{{C}_{d}}$$
(2)
where Cs and Cd mean the concentration of adsorbate on the solid and remaining in solution at equilibrium, respectively.Adsorption kinetics were predicted using PFO, PSO and IPD models represented by Eqs. (3), (4), and (5), respectively31:$${\text{ln}(q}_{1}-{q}_{t})= {lnq}_{1}-{k}_{1}t$$
(3)
$$\frac{t}{{q}_{t}}=\frac{1}{{k}_{2}{q}_{2}^{2}}+\frac{t}{{q}_{2}}$$
(4)
$${q}_{t}={k}_{id}{t}^{1/2}+C$$
(5)
where q1 and q2 [µmol·g−1] refer to the amount of target biomolecule adsorbed at equilibrium (for PFO and PSO models, respectively), qt [µmol·g−1] means the amount of target biomolecule adsorbed at time t, k1 [min−1], k2 [g·µmol−1·min−1] and kid [µmol·g−1·min1/2] are the rate constants of PFO, PSO and IPD models, respectively. For the PFO model, the values of q1 and k1 were determined from the intercept and slope of the plot of ln(q1 − qt) versus t. For the PSO model, the values of q2 and k2 were calculated from the intercept and slope of the plot of t/qt versus t. For IPD model, the slope and intercept of the straight line of qt versus t1/2 were used to calculate the values of kid and C, respectively.Langmuir and Freundlich models were used to plot equilibrium isotherms. Each point was repeated twice, and each sample was analyzed in triplicate. The Langmuir and Freundlich equations were represented by Eqs. (6) and (7), respectively43:$$\frac{1}{{Q}_{e}}= \frac{1}{{K}_{L}\cdot {Q}_{max}}\cdot \frac{1}{{C}_{e}}+\frac{1}{{Q}_{max}}$$
(6)
$${logQ}_{e}={logK}_{F}+\frac{1}{n}{\cdot logC}_{e}$$
(7)
where Ce [µmol·L−1] refers to the equilibrium concentration of the target biomolecule, Qmax[(µmol·g−1] is the maximum binding capacity, Qe [µmol·g−1] is the equilibrium adsorption capacity; KL and KF [L·mg−1] is the Langmuir and Freundlich constant; 1/n is a degree of heterogeneity. The values of KL and KF were determined from the plot of Ce/Qe versus Ce, and logQe versus logCe, respectively.For the Langmuir isotherm model, the dimensionless constant (separation factor, RL) was calculated using the Eq. (8)31:$${R}_{L}=\frac{1}{1+{{K}_{L}C}_{i}}$$
(8)
Values 0 < RL < 1 suggest favorable adsorption, RL > 1—unfavorable adsorption, RL = 1—linear adsorption, and RL = 0—irreversible adsorption31.Recovery and matrix effectThe test was carried out in PBS, BSA, and human serum matrices. For experiments, BSA solution was prepared in 0.01 mol·L−1 PBS (pH 7) and human serum was diluted 1:1 with 0.01 mol·L−1 PBS (pH 7). Samples were pre-treated as was described for adsorption evaluation. BSA sample contained 4 µmol·L−1 and 80 µmol·L−1 Kyn and Trp, respectively. Human serum initially contained 3.5 µmol·L−1 and 130 µmol·L−1 Kyn and Trp, respectively. Each experimental batch was carried out by adding of 50 mg of the proper sorbent (poly(4VP-co-TRIM), poly(4VP-co-14DMB), C18, AC, PSA) to 1 mL of sample. The agitation time was 1 h. After dSPE clean-up and sorbent removal, samples containing human serum were treated like a BSA matrix. Control samples were prepared without the dSPE step.Recoveries were determined as follows:$$\%_{remained}= \frac{Analyte\, Peak\, Area\, In\, Sample\, After\, Clean-up}{Analyte\, Peak\, Area\, In\, Initial\, Sample}\cdot 100\%$$
(9)
Matrix effect (ME) was calculated using the formula:$$ME = \frac{(A-B)}{D}\cdot 100\%$$
(10)
where A is the analyte peak area in a sample at the post-preparation spiking with analytes, B is the peak area in a blank sample, and D is the peak area in the blank solvent at the post-extraction spiking with analytes. In the case of the PBS matrix, a blank solvent was ultrapure water. For other evaluated matrices (BSA and serum), 0.01 mol·L−1 PBS (pH 7) was used as a blank solvent.Values of ME > 100% and ME < 100% mean signal enhancement and suppression, respectively.
4-Vinylpyridine copolymers for improved LC–MS tryptophan and kynurenine determination in human serum
