Automated Protein Biomarker Analysis: on-line extraction of clinical samples by Molecularly Imprinted Polymers

Robust biomarker quantification is essential for the accurate diagnosis of diseases and is of great value in cancer management. In this paper, an innovative diagnostic platform is presented which provides automated molecularly imprinted solid-phase extraction (MISPE) followed by liquid chromatography-mass spectrometry (LC-MS) for biomarker determination using ProGastrin Releasing Peptide (ProGRP), a highly sensitive biomarker for Small Cell Lung Cancer, as a model. Molecularly imprinted polymer microspheres were synthesized by precipitation polymerization and analytical optimization of the most promising material led to the development of an automated quantification method for ProGRP. The method enabled analysis of patient serum samples with elevated ProGRP levels. Particularly low sample volumes were permitted using the automated extraction within a method which was time-efficient, thereby demonstrating the potential of such a strategy in a clinical setting.

1. Optimization of the polymer synthesis:

Solubility tests.
In order to adapt the experimental approaches described elsewhere 1 to the precipitation polymerization procedure, solubility tests for the functional monomer EAMA.HCl were performed. Moreover, the solubility tests had the aim to determine the amount of DMSO needed to bring all the precipitation polymerization components into a homogenous solution. Different mole ratio of EAMA.HCl and DVB-80 (crosslinker) together with different combinations of MeCN (solvent) and DMSO (co-solvent) were tested, as presented in Table S-1. The use of EAMA.HCl in the same 1 / 5 molar ratio of EAMA.HCl / DVB as used in the synthesis performed in the previous work 1 resulted in monomer insolubility. This problem was overcome by decreasing the functional monomer concentrations. In solubility test #4, the amount of EAMA.HCl was reduced ten times and was dissolved completely in 4 % of DMSO and 96 % of MeCN. Thus the mole ratio of EAMA.HCl to DVB was set at 0.1:5. The mole ratio of DVB to N-3,5-bis(trifluoromethyl)-phenyl-N'-4-vinylphenylurea was set at 5:0.02.

Choice of reaction vessel and synthesis conditions.
In order to optimize the synthetic protocol, polymers without the addition of the template (non-imprinted polymers) were pre-tested as shown in Table S-2. A polymer with only the crosslinker (DVB-80) was prepared as control and polymers providing the use of the selected functional monomers (EAMA.HCl and N-3,5-bis(trifluoromethyl)-phenyl-N'-4vinylphenylurea) were prepared in two different reaction vessels.
As the table shows, yields are lower for the polymers synthesized in Polyethylene Nalgene bottles. Thus Borosilicate Kimax tubes were used to perform the syntheses. PMP and TBA.HO were used to bring the various functional groups (of EAMA and N-3,5-bis(trifluoromethyl)-phenyl-N'-4-vinylphenylurea respectively) into appropriate ionization states for noncovalent interactions between functional monomers and the template (which here was not added). Moreover, for the synthesis of Molecularly Imprinted Polymers, where the template will be added at the beginning, it was decided to increase the amount of PMP and TBA.HOfrom 0.006 mmol to 0.01 mmol since the template has two sites able to bind the functional monomers: the carboxylic acid group in the glutamic acid (E) residue and C-terminus of Z-NLLGLIEA[Nle] as shown in Figure 2 of the main text. The incubation time was extended to 48 hours in order to increase the yield of the polymerization as Table S-3 demonstrates.

Characterization of the polymers:
3.1 SEM analysis.

Binding isotherms.
The

S6
The Freundlich model commonly describes site distributions well in MIPs. The model implies a heterogenous distribution of sites continuously ranging from low to high binding energies and absence of homogenous populations of binding sites The parameter m is of particular importance and here confirms an heterogeneous population of molecularly imprinted binding sites arising from the non-covalent molecular imprinting strategy adopted.

Imprinting Factors.
Imprinting factors of the polymers were calculated as described by Manesiotis et al. 3 for both MIP/NIP pairs based on the retention times of a non-retained peptide (LSAPGSQR) and the target analyte (NLLGLIEAK) after the isocratic elution with 5% MeCN from the cartridges according the equation 4: IF = k′ MIP /k′ NIP Eq.4 where k′ MIP and k′ NIP are the respective retention factors defined as: k′ = (t R − t 0 )/t 0 Eq.5 with t R the retention time of the analyte (NLLGLIEAK) and t 0 the retention time of a not-retained peptide (LSAPGSQR).

Verification of NIP failure:
The extraction on a NIP cartridge (NIP A) of a serum samples spiked with 1 nM of ProGRP isoform 1 was performed in order to compare the performance with the extracted samples from the calibration curve. Addition of a solution of NLLGLIEA[K_ 13 C6 15 N2] 10 nM was performed before the injection in the chromatographic system in order allow a correct peak identification. The presence of the internal standard only show the impossibility of the NIP in enriching the targeted peptide within the serum sample after the optimized sample preparation, while the MIP could enrich such concentration with similar intensities for both target peptide and internal standard.