Tailored pretreatment of serum samples and biomarker extraction afforded by ionic liquids as constituents of aqueous biphasic systems

Prostate cancer serum biomarkers, such as the U.S. Food and Drug Administration (FDA)- approved prostate specific antigen (PSA), are valuable markers of diagnosis, response to treatment and disease progression. Despite the expected benefits of using PSA to better manage the mortality and morbidity associated to prostate cancer, sample complexity and interference within analysis make its quantification in serum a high-cost and strenuous assignment. Here, we make use of the simple preparation, water-rich environment, and selectivity of aqueous biphasic systems (ABS) to develop alternative serum pretreatment strategies. A set of ammonium-and phosphonium-based ionic liquids (ILs) were used as either main phase-forming agents in IL-salt ABS or as adjuvants in polymer-salt ABS, allowing to explore the IL designer solvent nature and versatile role in ABS formation and performance. Serum pretreatment efficiency was evaluated by the capacity of each ABS to simultaneously deplete high abundance serum proteins (immunoglobulin G, IgG, and human serum albumin, HSA) at the systems’ interphase (thus, generating a three-phase partitioning system – ABS-TPP) and extract PSA to a single ABS phase. Regardless of the ABS typology under appraisal (IL-salt or polymer-salt-IL), the proper design of the IL structure is critical for both the ability to deplete high abundance serum proteins and extract PSA. The results demonstrate that, in addition to the proper choice of the ABS typology, a careful adjustment of the IL hydrophobicity is key to achieve the simultaneous depletion of high-abundance proteins and complete PSA extraction in a single step, creating new approaches for PSA analysis and its application in diagnosis/prognosis.


Introduction
The most recent epidemiologic studies confirm that cancer remains one of the major health concerns, with 19.3 million new cases diagnosed worldwide in 2020 [1].Of these, 7.3% corresponded to prostate cancer cases, the second most frequent type of cancer among men [1].
In Portugal, 21% of cancer cases diagnosed in men were prostate cancer [2].In addition to family history, ethnicity and race, there are several risk factors associated with prostate cancer, such as smoking and lack of physical activity [3,4].These, if changed, can contribute to decrease the risk of developing such condition [5].Furthermore, an early detection of prostate cancer improves the chances of treatment and survival [6].
The quantification of serum biomarkers, such as the prostate specific antigen (PSA), may contribute to prostate cancer early detection.PSA (~33 KDa) is a glycoprotein that performs as an enzyme, produced by normal and malignant cells of the prostate gland, and found in human serum in concentrations at the ng•mL -1 range [7].In fact, PSA quantification in human serum was first approved by the Food and Drug Administration (FDA) as a prognosis tool for prostate cancer in 1986 [8].Eight years later, this approach was accepted as a diagnosis strategy in asymptomatic men [8].Clinically, PSA is also accepted as a monitoring and prognostic biomarker, providing information on patients' response to the applied therapy and disease progression [9,10].Overall, PSA testing is critical for determining the need and timing of salvage therapies for prostate cancer.
However, inaccurately measured PSA values (elevated or decreased) can interfere with the evaluation of the disease's status, wrongly influencing treatment recommendations [11].
Therefore, the accurate measurement of PSA levels in patients with prostate cancer should be ensured [12].
Aiming to quantify PSA in human samples, enzyme-linked immunosorbent assay (ELISA) [13,14], liquid chromatography -mass spectrometry (LC-MS) [15,16], electrochemistry [17,18] and electroluminescence (ECL) [19,20] are commonly used.Nevertheless, irrespective of the quantification method, quantifying biomarkers in human serum is a cumbersome task likely to suffer from interferences or masking effects of high abundance serum proteins [21], mainly immunoglobulin G (IgG) and human serum albumin (HSA) [22].Constituting 80% of the human serum proteome, HSA and the various forms of immunoglobulins are normally present at concentrations at least six orders of magnitude higher than PSA (i.e., mg•mL -1 versus ng•mL -1 ) [22].The interferences caused by these proteins may induce sensitivity and robustness losses and lead to false positive or negative results [23][24][25].For instance, in immunoassays, immunoglobulins such as IgG have been reported to cause interferences that lead to negative results in sandwich formats and positive results in competitive ELISA formats [26,27], while the high concentration of albumin and capacity to bind and release large amounts of ligands have shown to contribute to high background noises and reduce ELISA sensitivity [28,29].Hence, to perform an accurate detection and quantification of PSA, there is thus the need to pretreat human serum by removing the high abundance serum proteins and extracting PSA, ideally in a single step.
Nowadays, highly selective sample pretreatment strategies featuring simple execution and low cost remain unavailable.
To overcome the challenges in the pretreatment of human serum and, ultimately, in PSA extraction and analysis, aqueous biphasic systems (ABS) can be used.ABS are liquid-liquid systems formed due to the immiscibility of two aqueous solutions of two polymers or a polymer and a salt, either in the absence (ternary systems) or presence (quaternary systems) of adjuvants [39,40].Existing literature data support the potential of polymeric ABS in the extraction/analysis of proteins [19,41] and extracellular vesicles [42,43] from human fluids, both used in the diagnosis of prostate cancer.First, the high-water content creates a favorable environment to extract biomarkers in an intact form and without mass losses [44].Second, the variety of ABS constituents and their possible conjugations enables to tailor the efficiency and selectivity of extraction processes involving complex samples and the need for the removal of interfering molecules [44].Furthermore, especially when high selectivity is mandatory, ionic liquids (ILs) outperform conventional ABS constituents and adjuvants [45].ILs are easily adjusted to the selectivity needs of the extraction process simply by varying the IL cation/anion structure and combination [46].By using ILs, either as ABS main constituents or adjuvants, the interactions/affinities to be investigated in the development of highly selective extraction processes are significantly widened [47].
We have shown that IL-salt ABS can concentrate PSA from urine, contributing for the differential diagnosis of prostate cancer [48].Although convenient due to the non-invasive and simplicity of the sampling and the sample, the analysis of PSA in urine (instead of serum) still faces controversy.Here, we aim to explore the versatile role of ILs in the development of aqueous biphasic systems (ABS) for serum pretreatment.In the proposed systems, sample pretreatment is performed by depleting HSA and IgG through precipitation at the interphase of the two liquid phases, forming three-phase partitioning (ABS-TPP) systems.The efficiency and molecular-level mechanisms behind the simultaneous depletion of high-abundance serum proteins, HSA and IgG, at the interphase, and the extraction of PSA into a single phase were investigated by using ABS composed of IL-salt and polymer-salt-IL (as adjuvant), which were further benchmarked against conventional polymer-salt ABS.To assist the development of ABS-TPP-mediated serum pretreatment, the effects of the use of polymer-salt versus IL-salt, the presence versus absence of ILs as adjuvants, the use of ILs as main constituents versus adjuvants and the IL structure were assessed.

Materials
To form the polymer-salt, IL-salt and polymer-salt-IL ABS here investigated, polyethylene glycol with a molecular weight of 1000 g.mol -1 (PEG 1000) and the ionic liquids  1 along with their full names, acronyms, purities, and suppliers.These were applied in combination with phosphate buffer at pH of ca. 7, which was chosen according to the stabilization effect over proteins [49,50] and was prepared at ca. 40 wt% of the two following salts: potassium phosphate dibasic, K 2 HPO 4 (purity > 98 wt%) and potassium phosphate monobasic, KH 2 PO 4 (purity 99.5 wt%), both from Sigma-Aldrich.For the ABS-TPP preparation, the human serum used was acquired from Sigma-Aldrich (H4522-Lot # SLCDX6353 and SLCD4040) and kept at -20 °C until used.Pellets to prepare the phosphate buffered saline solution (PBS, pH ca.7.4) were purchased from Sigma-Aldrich and used following the specifications of the supplier.The quantification of PSA was performed using the commercial enzyme-linked immunosorbent assay (ELISA) kits Human PSA-total ELISA Kit RAB0331 (Lot # 0309F0316 and # 0216J316) from Sigma Aldrich.These were kept in storage at -20 ºC.Spiked aqueous and serum solutions were prepared using the PSA supplied in the mentioned ELISA kit, namely lyophilized human-PSA total protein standard resuspended in 400 µL of the sample diluent A (50 ng•mL -1 ).

Depletion of abundant proteins
All the phase diagrams of the ABS studied in this work were already reported in the literature [51][52][53][54][55] bearing ILs as adjuvants, the biphasic mixture compositions were selected considering the binodal curve of the corresponding ternary system, i.e., PEG 1000, K 2 HPO 4 /KH 2 PO 4 and water [51].The required amount of IL (1, 3 and 5 wt%) as adjuvant was added as part of the total water of the system.ABS-TPP were prepared by weighing the appropriate amount of each component, within initially present in the serum and in the top, inter or bottom phases according to the concentrations (mg•mL -1 ) determined by SE-HPLC.It should be remarked that the top phase corresponds to the PEG-rich phase in both polymer-salt and polymer-salt-IL systems, and to the IL-rich phase in the IL-salt systems.In all systems, the bottom phase is salt-rich and the interphase corresponds to precipitated serum proteins.

Biomarker extraction
To evaluate the ABS ability to extract PSA to one of the liquid phases, while avoiding its depletion, only the previously identified best performing systems were studied, namely Additionally, PEG 1000 + KH 2 PO 4 /K 2 HPO 4 was used for comparison purposes.Extraction studies were performed using the same compositions as well as equilibrium/phase separation conditions described above, with few modifications.At this stage, instead of human serum, one out of the two following solutions was added to each system: PSA aqueous solution (to appraise the partitioning of PSA) and PSA spiked human serum (to evaluate the ABS-TPP ability to extract PSA to one of the liquid phases while preserving HSA and IgG depletion efficiencies).PSA concentration in both solutions was of 4 ng•mL -1 , according to the cut-off values used for prostate cancer diagnosis [56].It should be mentioned that in the systems prepared with the PSA aqueous solution (without serum), no interphase was formed.For quantification purposes, the liquid phases were diluted using the provided sample diluent buffer A, which is appropriate for human serum samples.Depending on the ABS typology and to improve analysis accuracy, different dilution factors were applied: (i) IL-salt, 1:20 (top phase) and 1:5 (bottom phase); (ii) polymer-salt-IL, 1:30 (top phase) and 1:5 (bottom phase); and (iii) polymer-salt, 1:10 (top phase) and 1:3 (bottom phase).The amount of PSA precipitating at the interphase was assessed through mass balance.All systems were prepared in triplicate to determine the average values and the corresponding standard deviation.PSA concentration in each liquid phase was determined following the ELISA kit manufacturer's instructions and using a calibration curve previously determined within the 10.24 -400 pg•mL -1 range (r 2 value ≥ 0.98).The interference of the ABS components and/or the human serum matrix with the quantification method was also addressed using blank controls, i.e., the same system compositions containing pure water (for the systems containing PSA aqueous solution) or non-spiked serum (for the systems containing spiked serum).
In the studies involving PSA aqueous solutions, the extraction efficiency of PSA (EE PSA %) and recovery yield (RY PSA %) to the top phase was determined using Equations 3 and 4, respectively: In the studies involving spiked serum, PSA recovery yield (RY PSA, %) in the top, inter and bottom phase, i.e., the percentage ratio between total weight of PSA in a given phase to that initially spiked in the serum, was defined according to equation 5: where represent the mass of PSA in the top, inter or bottom phase, whereas  //  is the mass of PSA contained in the spiked serum added to the system.   Circular dichroism (CD) spectra were finally recorded to evaluate the conformation stability of PSA through any secondary structure changes.CD spectra of PSA in the top phases of two representative ABS, namely 30 wt% of [N 4444 ]Cl/30 wt% of PEG 1000 + 12 wt% of KH 2 PO 4 /K 2 HPO 4 + 48 wt% of water + 10 wt% of PSA aqueous solution (25 ng•mL -1 ) were determined.CD analysis of a PSA standard consisting of a solution of 4 ng.mL - of PSA in phosphate-buffered saline (PBS) was used as reference.CD analyses were also ran for blank controls to address any influence arising from the ABS phase.The CD spectra were recorded using a JASCO J-815 spectropolarimeter (JASCO, Easton, MD, USA) at 25 °C in rectangular quartz Suprasil CD cuvettes of 0.1 cm.Each spectrum results from the average of three scans acquired by collecting data in a range of 190-280 nm at a scan speed of 50 nm/min.

ABS-TPP-mediated pretreatment of serum by the depletion of highabundance proteins
Due to the multipurpose character of ILs, ILs were here investigated as main constituents of IL-salt ABS, as well as adjuvants in polymer-salt systems, to develop alternative sample pretreatment strategies.Regardless of the ABS typology, along this work, the IL cation/anion chemical structure was adjusted to manipulate the properties of the systems.Thereby, different sample pretreatment and biomarker extraction efficiencies are expected to be achieved.Variations at the level of the cation alkyl chain length (i.e., [N 2222 ] + ,[N 3333 ] + and [N 4444 ] + ) and nature (i.e., [Ch] + , [C 4 mim] + , [N 4444 ] + and [P 4444 ] + ) promote distinct hydrophilic-hydrophobic effects as well as other possible interactions with proteins (e.g., π-π stacking).Anions with opposite hydrogenbond strength (i.e., Cl -versus Br -) allow adjusting the polarity of the systems [57].Being a common ABS/ABS-TPP constituent, the salt selected was K 2 HPO 4 /KH 2 PO 4 (due to its saltingout strength and buffering capacity) and the polymer chosen was PEG (a protein precipitating agent).To carry out IgG and HSA depletion from human serum samples, the same mixture composition localized at the biphasic region of all ABS phase diagrams was used.Upon the addition of serum to these mixtures, ABS-TPP bearing a solid interphase composed of precipitated high-abundance serum proteins were formed.The depletion efficiency of both proteins was determined by the fraction of serum HSA and IgG precipitating at the interphase following protein quantification by SE-HPLC.

Ionic-liquid-salt ABS-TPP
The first group of systems under investigation comprises IL-salt ABS-TPP.Using a mixture composition of 30 wt% of IL + 12 wt% of K 2 HPO 4 /KH 2 PO 4 (pH ≈ 7) buffer + 10 wt% of human serum + 48 wt% of water, it was possible to evaluate the depletion efficiencies of IgG and HSA (DE IgG and DE HSA ) at the interphase of the systems.Figure 1 shows the depletion efficiency results obtained using IL-salt ABS-TPP (please refer to the Supplementary Information, Table S1, for the detailed numerical data together with the recovery yields of HSA and IgG to the top and bottom phase).
The depletion efficiencies obtained using IL-salt ABS-TPP ranged between 0% and 99.98% for IgG and between 0% and 98.8% for HSA.Considering the data depicted in Figure 1 The formation of IL-salt ABS depends mainly on the competition between the IL and salt ions for hydration [47], while the partition/precipitation of proteins in ABS and TPP is mainly associated with the increasing hydrophobicity and salting-out intensity of the phase-forming agents, complemented by protein-protein interactions [58].Besides the salting-out effect of K 2 HPO 4 /KH 2 PO 4 , our results correlate well with the octanol-water partition coefficients (logK ow ) of ILs.ILs exhibiting negative logK ow (i.e., -2.15 for [C 4 mim]Br/Cl, -2.13 for [N 2222 ]Br/Cl, -0.12 for [N 3333 ]Br/Cl) [59], thus the more hydrophilic, do not present the capability to promote the depletion of HSA and IgG at the interphase.On the contrary, [N 4444 ]Br/Cl and [P 4444 ]Br/Cl, which present higher logK ow (2.01 and 4.07, respectively [59]) and thus higher hydrophobicity, translate into ABS top phases of higher hydrophobicity and reduced water content.Such environments are propitious for protein-protein interactions, intensifying protein depletion events.
As for the effect of the anion, i.e., by comparing Br -versus Cl -based ILs, Br -has a lower hydrogen bond basicity than Cl -and induces a higher protein depletion [57].Accordingly, Br -is less prone than Cl -to accept hydrogen bonds, thus limiting protein solubility.Nevertheless, no anion effect is apparent using lighter quaternary ammonium ILs (i.e., [N 2222 ]Br/Cl), and highly polar aromatic ILs, i.e., [C 4 mim]Br/Cl.Based on the exposed, the depletion efficiency trends are in agreement with the rank of ABS formation ability in the presence of salts, where phosphonium-and heavier tetraalkylammonium-based ILs outperform the remaining ILs under investigation.It is thus possible to establish structure-properties-performance relationships in serum pretreatment using ABS-TPP.Overall, under the conditions tested, the higher the interaction of the ILs ions with water molecules or the IL hydrophilicity, the harder is the ABS-TPP formation through protein depletion at the interphase.In all IL-based ABS-TPP developed, higher depletion efficiencies were obtained for IgG than HSA.This agrees with the IgG higher molecular weight (ca.160 KDa versus ca.66 KDa), being more prone to be depleted [34].Moreover, proteins tend to precipitate at their isoelectric point (pI) [60,61].Considering that IgG has a pI of 7-9.5 [62], it is more prone to precipitate at the pH of the ABS-TPP than HSA (pI = 4.7) [63].HSA presents an overall negative charge at pH ≈ 7, being thus more soluble, and thus less prone to precipitate [64].Besides the role of the IL cation hydrophobicity, anion hydrogen bond basicity and the salting-out effect of K 2 HPO 4 /KH 2 PO 4 , electrostatic interactions seem to play a role in the IL-salt ABS-TPP developed [65].

Polymer-salt-IL ABS-TPP
The second group of ABS studied entails polymer-salt as main components and ILs as adjuvants [40].This typology of ABS allows to reduce the amount of IL used and better control the properties of conventional polymeric systems [40].The depletion studies were performed resorting to a biphasic mixture composition containing 30 wt% of PEG 1000 + 12 wt% of K 2 HPO 4 /KH 2 PO 4 (pH ≈ 7) + 10 wt% of human serum + 5 wt% of IL + 43 wt% of water.Aiming to understand the effect of adding ILs over the depletion efficiencies of HSA and IgG, a system composed of 30 wt% of PEG 1000 + 12 wt% of K 2 HPO 4 /KH 2 PO 4 (pH ≈ 7) + 10 wt% of human serum + 48 wt% of water was analyzed for comparison purposes.Figure 2 shows the depletion efficiency results obtained using polymer-salt-IL ABS (please refer to the Supplementary Information, Table S2, for the detailed numerical data together with the recovery yields of HSA and IgG to the top and bottom phase).
The polymer-salt ABS (i.e., without any IL) reached depletion efficiencies of 23% and 95% for HSA and IgG, respectively.Remarkably, the addition of ILs as adjuvants led to higher depletion efficiencies, especially of HSA, in most ABS studied (except for [N 2222 ]Cl).Thereby, the addition of ILs as adjuvants, if well designed, can improve the efficiency of serum pretreatment.As for IL-salt ABS, the depletion efficiencies reached with polymer-salt-IL ABS are governed by effects related to the IL structure, the salting-out effect of K 2 HPO 4 /KH 2 PO 4 , proteins molecular weight and pH of the system.Indeed, more hydrophobic ILs, particularly those bearing [N 4444 ] + and [P 4444 ] + cations, are more efficient adjuvants, leading to depletion efficiencies higher than 90% for both proteins.Despite the higher hydrophilicity of its cation, cholinium-based ILs (i.e., logK ow values of -4.66 for [Ch]Br/Cl [59]) exhibited a similar depletion performance, particularly if comprising the Br -anion (DE IgG = 99% and DE HSA = 95%).In the case of choliniumbased ILs, it seems that specific Ch-proteins interactions are contributing to the depletion of IgG and HSA. Figure 3 illustrates the effect of IL concentration on the depletion efficiencies of IgG and HSA for each ABS-TPP (please refer to the Supplementary Information, Table S2, for the detailed numerical data together with the recovery yields of HSA and IgG to the top and bottom phase).
Higher IL concentration resulted in increased depletion efficiencies of IgG and HSA for all the ABS-TPP studied, this trend being more accentuated for the depletion of HSA along the entire IL concentration range evaluated (1 to 5 wt%).These results highlight again the importance of hydrophobic effects and electrostatic interactions to induce depletion at the interphase.Overall, comparing systems of distinct ABS typologies but sharing the same IL, polymersalt-IL ABS seem to be more efficient strategies for the depletion of high abundance serum proteins.The ABS comprising [P 4444 ]Br as the main constituent together with K 2 HPO 4 /KH 2 PO 4 , however, represents an exception to this behavior, highlighting its usefulness as a protein depletion approach.The joint effect of PEG and salts seems to determine the ability of ABS-TPP for serum protein depletion, in which the protein solubility in the polymer-rich top phase is reduced.Depletion can be further improved/adjusted by the addition of ILs as adjuvants, which can modulate the polarity of the phases or assist the salting-out effect, impacting protein solubility/stability [40,66].Due to their outstanding role as either ABS main constituents and/or adjuvants, ABS formed by [P 4444 ]Br, [P 4444 ]Cl, [N 4444 ]Cl and [Ch]Br were adopted in the PSA partition and extraction studies presented below.

Extraction of prostate specific antigen from serum
The extraction of PSA from human serum was investigated resorting to the ABS-TPP systems that provided the highest depletion efficiencies for IgG and HSA among each typology, namely [ performed along the depletion studies, the impact of the concentration of ILs as adjuvants (i.e., 1, 3 and 5 wt%) on the PSA extraction was also investigated.
Initially, studies involving an aqueous solution of PSA at 4 ng•mL -1 were carried out to infer on the partition of PSA in ABS without the influence of the real sample.In all systems, PSA completely partitioned to the top IL-or PEG1000-rich phase without significant losses (EE PSA > 99% and RY PSA > 98%, cf.Table S3 in the Supplementary Information), which is mainly due to the salting-out effect exerted by K 2 HPO 4 /KH 2 PO 4 .These results agree with the PSA partition in ABS composed of other ILs and salts [48].
Studies involving PSA spiked human serum followed to understand the influence of the human serum proteins in PSA partition.PSA recovery yields (RY PSA , %) are shown in Figure 4 and further detailed in Table S4 in the Supplementary Information.Figure 4 clearly shows that the extraction of PSA is achieved at the top phase of all systems (72% < RY PSA < 100%).These results show that the previously observed partition trend was maintained in the presence of a more complex sample -human serum.]Br.In these systems, no significant PSA precipitation occurred at the interphase.Despite being expected to precipitate due to its pI of 6.9 [67] (pI = pH), PSA is at a much lower concentration and has a smaller size (~33 KDa), these seeming to be the leading causes minimizing PSA precipitation [8,68].Unlike the concentration-dependent behaviour observed for the depletion of HSA and IgG, the recovery yields obtained for PSA were not significantly affected by the concentration of ILs as shown in Figure 4. Based on this, 5 wt% of IL adjuvant is the ideal concentration under study to simultaneously deplete high-abundance serum proteins and extract PSA using polymer-salt-IL ABS-TPP.

Declaration of interests
☒ The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
☐ The authors declare the following financial interests/personal relationships which may be considered as potential competing interests:

± 10 -
4 g, and mixing until all components were dissolved.Human serum was the last component added.Each mixture was gently mixed, centrifuged for 10 min at 3500 rpm, and left to equilibrate for 10 min at 25 ºC to reach the total separation of the liquid phases and the formation of the solid interphase.Both liquid phases were then carefully separated, collected with syringes to avoid any contamination from the interphase, and further diluted with the SE-HPLC mobile phase for analysis.At least two replicates were performed for all systems to determine the average values and the corresponding standard deviations.The HSA and IgG in each liquid phase and in the commercial human serum was quantified by SE-HPLC using a Chromaster HPLC system (VWR Hitachi) equipped with the analytical column Protein KW-802.5 (8 mm × 300 mm) coupled with a Protein KW-G 6B guard column, both from Shodex.The mobile phase consisted of 50 mM sodium phosphate buffer and 0.3 M NaCl solution and run under isocratic mode at 0.5 mL•min -1 for 40 min.Both column oven and autosampler temperatures were operated at controlled temperatures of 25 °C and 10 °C, respectively.The injection volume was 25 µL and the wavelength was set to measure at 280 nm.Under these analytical conditions, the retention times of IgG and HSA were approximately 15.21 min and 16.60 min, respectively.HSA and IgG quantification in each phase resorted to calibration curves previously determined within the 0.1 -1.0 mg•mL -1 range (r 2 ≥ 0.99).Due to the impossibility of analysing the ABS-TPP interphases due to solubilization issues, the amount of IgG and HSA in the interphase was obtained through mass balance considering the amount of each protein quantified in the liquid phases and that present in the commercial human serum added to each ABS.IgG and HSA recovery yields (RY IgG and RY HSA , %, respectively) in the top and bottom phases and depletion efficiencies (DE IgG , % and DE HSA , %), corresponding to their precipitation at the interphase of each system, were determined according to Equations 1 and 2, the mass of IgG or HSA (mg)   /  // / of PSA in the PEG 1000-rich or the IL-rich phase and       in the salt-rich phase, respectively, and is the mass of PSA contained in the aqueous    solution added to the system.
, the ABS-TPP composed of[P 4444  ]Br achieved the highest depletion efficiency for both proteins (DE IgG = 99.98% and DE HSA = 98.8%).The performance of this system was followed by those composedof [P 4444 ]Cl > [N 4444 ]Cl > [N 4444 ]Br > [N 3333 ]Br > [C 4 mim]Br > [N 3333 ]Cl ≈ [N 2222 ]Br ≈ [N 2222 ]Cl≈ [C 4 mim]Cl.In the presence of phosphonium-based ILs, the majority of HSA and IgG content precipitated at the interphase.In turn, [N 4444 ]Br and [N 4444 ]Cl promote an equivalent partition of both proteins between the interphase and the IL-rich top phase as shown in Table S1 in the Supplementary Information.In the presence of [C 4 mim]Br/Cl, [N 2222 ]Br/Cl, [N 3333 ]Br/Cl, no appropriate conditions for IgG and HSA precipitation at the interphase were met, since these partition mostly to the IL-rich top phase.
Here, we have successfully taken advantage of the IL versatile application as ABS-TPP constituents while pursuing alternative human serum pretreatment and biomarker extraction strategies.Two typologies of IL-based ABS, namely IL-salt and polymer-salt-IL were screened and further benchmarked against conventional polymer-salt systems regarding their ability to simultaneously deplete high-abundance serum proteins by precipitation and extract PSA with no mass and structural losses.The depletion efficiency of HSA and IgG resorting to ABS-TPP was initially investigated using commercial human serum, with [P 4444 ]Br, [P 4444 ]Cl, [N 4444 ]Cl and [Ch]Br being among the best performing ILs either as main components of ABS (DE IgG > 67% for IgG and DE HSA > 34%) or as adjuvants (DE IgG > 90% and DE HSA > 93%).It has been shown that the depletion efficiencies can be maximized by tailoring the IL chemical structure independently of the ABS typology studied and the IL concentration as adjuvant in PEG-salt ABS.The potential of ABS-TPP as a depletion strategy has been proved by the depletion efficiencies obtained.A maximum of depletion efficiency of 99.98% of IgG and 98.8% of HSA was granted when using [P 4444 ]Br as main constituent of IL-salt systems.Despite the primordial role of the IL, also the salting-out effect of K 2 HPO 4 /KH 2 PO 4 , proteins molecular weight and medium pH must be considered in the design of ABS-TPP-mediated depletion strategies.In turn, recovery yields of PSA from serum higher than 87% were obtained for almost all systems, showing the adequacy of ABS-TPP for PSA extraction.However, to achieve the desired goal, i.e., the simultaneous depletion of high-abundance serum proteins and complete PSA extraction in a single step, ABS containing ILs as adjuvants are more promising candidates than polymer-salt or IL-salt ABS.Particularly if containing a well-designed IL such as [P 4444 ]Br as adjuvant, ABS-TPP can lead to the depletion of almost all the IgG and HSA content at the interphase (DE IgG and DE HSA > 95%) and the extraction of 97% of PSA in the top phase in onestep.Based on the results here reported, ABS-TPP can be adapted to fulfil the performance requirements of high-abundance serum protein depletion, biomarker extraction or both simultaneously.Depending on the final goal of the application, ABS-TPP hold potential to improve the detection of the prostate cancer biomarker PSA, eventually providing better diagnosis/prognosis.Furthermore, to improve the possibility of achieving the simultaneous depletion and biomarker extraction, other IL structures could be investigated especially considering the IL-salt ABS typology.The scope of application of these systems could be additionally broadened by investigating other possible interfering serum molecules and target biomarkers.Particularly, compared to the commonly used analytical techniques, it is expected that ABS-TPP could assist PSA analysis by more cost-effective and expedite approaches or even point-of-care devices, improving the reliability and widespread application of biomarker analysis.CRediT authorship contribution statement M.E.R.: Methodology, Validation, Formal analysis, Investigation, Data Curation, Writingoriginal draft, Visualization; M.S.M.M.: Methodology, Validation, Formal analysis, Investigation, Data Curation, Writing -review & editing; E.C.: Investigation, Data Curation; J.P.C.: Writing -review & editing, Supervision; J.A.P.C.: Writing -review & editing, Supervision; M.G.F.: Conceptualization, Funding Acquisition, Supervision, Writing -review & editing; F.A.eS: Conceptualization, Writing -review & editing, Supervision, Funding Acquisition, Project Administration.

Table 1 .
Name, acronym, chemical structure, purity, and supplier of the ILs and polymer used as ABS constituents.
S4Extraction efficiencies and recovery yields of prostate specific antigen