sensors

Review Toward Cancer Diagnostics of the Tumor Suppressor p53 by Surface Enhanced Raman Spectroscopy

Anna Rita Bizzarri * and Salvatore Cannistraro Biophysics and Nanoscience Centre, Dipartimento DEB, Università degli Studi della Tuscia, 01100 Viterbo, Italy; [email protected] * Correspondence: [email protected]



Received: 21 November 2020; Accepted: 11 December 2020; Published: 14 December 2020


Abstract: The tumor suppressor p53 protein plays a crucial role in many biological processes. The presence of abnormal concentrations of wild-type p53, or some of its mutants, can be indicative of a pathological cancer state. p53 represents therefore a valuable biomarker for tumor screening approaches and development of suitable biosensors for its detection deserves a high interest in early diagnostics. Here, we revisit our experimental approaches, combining Surface Enhanced Raman Spectroscopy (SERS) and nanotechnological materials, for ultrasensitive detection of wild-type and mutated p53, in the perspective to develop biosensors to be used in clinical diagnostics. The Raman marker is provided by a small molecule (4-ATP) acting as a bridge between gold nanoparticles (NPs) and a protein biomolecule. The Azurin copper protein and specific antibodies of p53 were used as a capture element for p53 (wild-type and its mutants). The developed approaches allowed us to reach a 17 detection level of p53 down to 10− M in both buffer and serum. The implementation of the method in a biosensor device, together with some possible developments are discussed.

Keywords: surface enhanced Raman spectroscopy (SERS); p53; p53 mutants; ultrasensitive detection; cancer biomarkers

1. Introduction The p53 protein, commonly referred as tumor suppressor, is a transcription factor which plays a central role in maintaining the genome integrity of the cell. It is coded by the TP53 gene which belongs to a highly conserved family of genes [1,2]. p53 is controlled by a variety of post-translational modifications and of interactions with several target signaling proteins [3]. In normal conditions, the level of p53 is kept low by a regulatory feedback loop mainly involving mouse double minute 2 homolog (MDM2) and MDMX [4,5]. Under cellular stress conditions, such as DNA damage and oncogenic stresses, p53 is activated giving rise to multiple biological processes that lead to repair genome mutations or to cell-cycle arrest, apoptosis and senescence, if the damage is irreparable [3]. p53 was found to be inactivated in several tumors, through either mutations in the TP53 gene or a deregulation of its associated path [6–9]. On the other hand, p53 is overexpressed in many tumors, as due, e.g., to specific oncogenic stresses or MDM2 gene alterations. These changes yield a deregulation of p53 associated pathways, with drastic effects on cell health [10]. The presence of mutated p53 proteins in human blood, can be put into relationship to a pathological state [11–13]. On the other hand, abnormal concentrations of wild-type p53 (p53wt) were found in some cancers [14]. Accordingly, p53 represents a valuable cancer biomarker and its ultrasensitive detection can be an extremely useful tool for non-invasive screening approach in the field of tumor early diagnostics. Additionally, p53 detection can help in evaluating prognosis also in the perspective of next-generation individualized precision medicines. High efforts are therefore devoted to develop biosensors able to selectively reveal p53 at very low concentrations [15]. With such an aim, different techniques, often

Sensors 2020, 20, 7153; doi:10.3390/s20247153 www.mdpi.com/journal/sensors Sensors 2020, 20, 7153 2 of 19 combined with nanotechnological strategies, were applied. They range from traditional methods, such as enzyme-linked immunosorbent assay (ELISA), electrophoretic immunoassay and mass spectroscopy, to more innovative and advanced techniques, such as surface plasmon resonance (SPR), surface enhanced Raman spectroscopy (SERS), field-effect transistor (FET), chemi-luminescence, calorimetry, quartz crystal microbalance, etc. [16–21]. Detection of p53 presents, however, some more difficulties in comparison to those of other proteins because of its structural disorder [22]. Indeed, p53 belongs to intrinsic disorder protein (IDP) family which are characterized by the presence of some disordered regions giving rise to a multiplicity of configurations, often co-existing in solution and resulting in a structural heterogeneity [23,24]. On one hand, the partial lack of well-defined structure confers to IDP a high adaptability in terms of binding capability to different partners; but, on the other, it may limit the strength of the formed complexes [25]. For an effective biosensing capability, a special care should be then devoted to the choice of the biomolecular capture element, able to form a stable complex with p53, through a specific biorecognition process. In our group, we investigated the interaction of p53, and also of other molecules of p53 family (p63 and p73) with various ligands, such as Azurin (Az), Az-derived peptides, MDM2, E3 ubiquiting-protein ligase COP1 [26–28]. These studies were performed by different techniques, operating in bulk or at single molecule level, such as atomic force microscopy and spectroscopy (AFM and AFS), SPR, fluorescence, Foester resonance energy transfer (FRET) and Raman [29–32]. A particular attention was paid to the interaction of p53 with the redox blue copper-containing protein Az from the bacterium Pseudomonas aeruginosa which has been shown to exert an anticancer action, upon its binding to p53 [33,34]. The study of the p53 binding network was combined with the previous and consolidated experience on ultrasensitive detection of different biomarkers by SERS [35–37], to implement highly specific and sensitive detection of p53 [38–40]. SERS is a Raman-based technique exploiting the huge enhancement of the Raman cross section when molecules are placed in close proximity of a nanostructured metal surface, typically aggregated gold and silver nanoparticles (NPs). Hence, SERS conjugates the richness of chemical and structural specificity of Raman spectroscopy with a very high sensitivity, and then, it is an extremely suitable technique to detect molecules at very low concentration [41–43]. However, to effectively reach notable results, SERS-based approaches should be developed in a close connection with suitably functionalized nanostructures, specifically tailored for the biomarker to be detected. Here, we reviewed and revisit our SERS-based approaches for ultrasensitive detection of wild-type and some mutants of p53, in the perspective to develop biosensor devices to be used in clinical applications [38–40]. Our approaches exploit nanotechnological strategies, with the help of surface chemistry, to prepare suitable gold NPs and substrates; both of them being functionalized with one of the two partners. The Raman marker is provided by a small molecule (4-ATP) acting as a bridge between gold NPs and a protein molecule; the corresponding Raman signal being highly enhanced when 4-ATP molecules are conjugated to gold NPs. Functionalized NPs were fluxed on the substrate to promote a biorecognition process between the molecular partners and then the capture from the substrate. The final system was scanned under the microscope objective for detection. Different molecular architectures, together with the use of different capture elements for targeting p53, were investigated. In particular, we exploited the capability of Az, or specific monoclonal antibodies, to bind both p53wt and some p53 mutants. For each system, the molecular strategy was carefully refined in order to maximize the detection capability. The results were evaluated in terms of sensitivity, selectivity, accuracy and reproducibility; the potentialities of the best detection method to be implemented in clinical biosensors were also analyzed. Finally, our results were discussed in connection with other available SERS-based approaches for ultrasensitive detection of p53 and possible developments were briefly presented. Sensors 2020, 20, 7153 3 of 19

2. Biosensor: Detection Technique, Molecular Components and Procedures

2.1. Biosensors: Basic Principles A biosensor is a device which combines a biological component (capture element) able to recognize a specific analyte (target) and then producing, through a physico-chemical component (transducer), a measurable signal for detection. The main features required for a biosensor are: (i) the accuracy; (ii) the reproducibility; (iii) the stability with respect to the measurement conditions; (iv) the Limit of Detection (LOD), which is the lowest concentration that can be detected; (v) the sensitivity, given by the variation of the biosensor response with respect to the variation of target concentration; (vi) the operating range, providing the concentration range of detection; and finally (vii) the selectivity which should be evaluated with a particular attention with respect to possible competitors of the target. Additionally, a linear response with concentration is commonly required. In the following, our SERS based detection methods will be presented and discussed in connection with the main biosensor features also in the perspective to be implemented in a device for real applications.

2.2. SERS Principles, Equipments and Measurement Conditions As already mentioned, SERS is based on the huge enhancement, up to several orders of magnitude, of the extremely weak Raman cross-section, when molecules are close to a nanostructured metal surface (mainly noble metals) [44]. Such an effect is generally attributed to two main mechanisms: (i) an electromagnetic enhancement (EM) mechanism, associated with the occurrence of large local field caused by surface plasmon resonance; and (ii) a charge transfer chemical mechanism (CT) from the metal to the adsorbed molecules [45]. The real underlying processes are characterized by a rather complex interplay among them, and the signals may fluctuate in time [46,47]. Accordingly, some effort should be devoted to establish stable measurement conditions with a maximized enhancement. Different regions (3 3 mm2) of each substrate sample were manually scanned by the microscope objective in × order to localize sites characterized by extremely high scattering; these sites, called “hot spots”, being usually related to highly Raman enhancement. For other experimental details see refs. [40,48].

2.3. Target Molecules: Wild-Type and Mutated p53 p53 is a tetrameric protein in which each monomer is made of 393 amino acid residues, arranged into four functional regions: the N-terminal transactivation (NT) domain, the DNA binding domain (DBD), the tetramerization domain and the C-terminal regulatory (CT) domain. As due to the presence of some disordered regions, the structure of the full-length p53 is still unresolved [48,49]. In the following, we considered wild-type p53 (p53wt), and three single point mutants: p53R249S, p53C135V and p53R175H, with a mutation inside the DBD portion. All the mutants are valuable biomarkers due to their high relevance in cancer. Figure1a shows a graphical representation of the DBD portion of p53 wt from 1 TUP entry of the protein data bank (PDB) [50]; the positions of amino acids susceptible to be mutated in our investigation are marked (see legend of Figure1). The single mutation in p53 R249S (azur sticks in Figure1a), yields a structural distortion of in the L3 loop (residues 243–249) of DBD affecting then its binding capability to DNA [51]; this mutation being very commonly observed in cancer [52]. In particular, it was found to be strictly related to mutagenesis as induced aflatoxin [53]. Sensors 2020, 20, x FOR PEER REVIEW 4 of 18 Sensors 2020, 20, 7153 4 of 19

Figure 1. (a) Graphical representation of the DBD portion of p53wt from chain B of 1TUP PDB entry [50 ]. Amino acids undergoing single mutations are marked: p53 (azur sticks) p53 (green sticks) Figure 1. (a) Graphical representation of the DBD portion ofR249S p53wt from chain BC135V of 1TUP PDB entry and p53R175H (blue sticks). Zn ion is represented as a yellow sphere. (b) Graphical representation of [50]. Amino acids undergoing single mutations are marked: p53R249S (azur sticks) p53C135V (green the blue copper protein Azurin (Az) from chain B of the 1AZU PDB entry [54]. The main amino acids sticks) and p53R175H (blue sticks). Zn ion is represented as a yellow sphere. (b) Graphical representation expected to be involved in the interaction with DBD are marked in red. Cu ion is represented as an of theorange blue c sphere.opper protein The figures Azurin were created(Az) from by Pymol chain [55B ].of the 1AZU PDB entry [54]. The main amino acids expected to be involved in the interaction with DBD are marked in red. Cu ion is represented as an orangeThe p53sphere.C135V Themutant figures (green were sticks created in Figureby Pymol1a) is[55 a]. temperature-sensitive mutant, involved in neoplastic cell formation, tumor invasiveness, and genotoxic stresses [56]. Below 37 ◦C, p53C135V Thebehaves p53C135V as the mutant wild-type (green p53, sticks while in at higherFigure temperature1a) is a temperature its conformation-sensitive switches mutant, to ainvolved mutant in phenotype. Finally, the p53 mutant (blue sticks in Figure1a) is found in several tumors, such as neoplastic cell formation, tumorR175H invasiveness, and genotoxic stresses [56]. Below 37 °C, p53C135V be- haveslung, as the colon wild and-type rectal p53, and while with aat particularly higher temperature high incidence its conformation in breast tumor switches [57–59]. to a mutant phe- notype.2.4. Finally, Capture the Elements p53R175H for p53 mutant (blue sticks in Figure 1a) is found in several tumors, such as lung, colon and rectal and with a particularly high incidence in breast tumor [57–59]. The capture element of a biosensor is tasked to selectively and stably bind the target (here wild-type or mutated p53) through a biorecognition process. As already mentioned, such a task 2.4. Capture Elements for p53 could be rather difficult for p53, as due to the presence of some unstructured regions which may Theresult capture into some element instability of a biosensor [22]. As capture is tasked molecules to selectively for p53, and we stably selected bind two the diff targeterent kinds(here ofwild- type ormolecules: mutated Az p53) and through specific antibodies.a biorecognition The blue process. copper-containing As already Azmentioned, protein (MW such 14kDa), a task whose could be ratherstructure difficult from for X-rayp53, investigationas due to the is presence represented of in some Figure unstructured1b [ 54], is characterized regions which by a high may structural result into somestability instability and [22] well-defined. As capture mechanical molecules properties for p53, [we60,61 selected]. Notably, two Az different exhibits kinds peculiar of molecules: redox and Az optical properties which make it suitable also for bio-nano electronics applications [62–64]. As already and specific antibodies. The blue copper-containing Az protein (MW 14kDa), whose structure from mentioned, Az exerts an anticancer action, by enhancing apoptosis in tumor cells upon its binding X-ray investigation is represented in Figure 1b [54], is characterized by a high structural stability1 and to p53 [65]. For the Az-p53 complex, we found a dissociation rate value of koff of 0.09 s− and a well-defined mechanical properties6 [60,61]. Notably, Az exhibits peculiar redox and optical proper- dissociation constant KD~6x10− M[65]. Furthermore, a tentative binding region between Az and DBD ties whichwas also make proposed it suitable by applying also for computational bio-nano electronics docking applications with the help [62 of mutagenic–64]. As already data [66 mentioned,,67]; the Az exertsamino an acids anticancer expected action, to be involved by enhancing in the DBD apoptosis/Az complex in tumor being marked cells upon in red its in Figurebinding1b. to Az p53 was [65]. −1 For thealso Az demonstrated-p53 complex, to selectivelywe found binda dissociation both the p53 rateR249S valueand of p53 kC135Voff of 0.09mutants s and [39]. a Thesedissociation features con- stant makeKD~6x10 Az an−6 M extremely [65]. Furthermore, valuable capture a tentative element forbinding p53. As region alternative between capture Az biomolecules,and DBD was we also used pro- posedmonoclonal by applying antibodies, computational specifically docking selected with to target the thehelp p53 ofwt mutagenicor the mutants data of [6 interest.6,67]; the In particular,amino acids expecweted used to be the involved mouse monoclonal in the DBD/ anti-p53Az complexwt PAb 1620 being antibody marked (anti-p53 in redwt in, MWFigure 150 1b kDa). Az and was the also rabbit polyclonal anti-p53R175H antibody (anti-p53R175H, MW 150 kDa) [40]. Finally, suitably designed demonstrated to selectively bind both the p53R249S and p53C135V mutants [39]. These features make Az aptamers can be used. an extremely valuable capture element for p53. As alternative capture biomolecules, we used mono- clonal2.5. antibodies, Description specifically and Characterization selected of to the target 4-ATP-NP the p53 Probewt or the mutants of interest. In particular, we used the mouse monoclonal anti-p53wt PAb 1620 antibody (anti-p53wt, MW 150 kDa) and the rabbit The structure of the 4-aminothiophenol (4-ATP) molecule, used as Raman marker in our detection polyclonalmethod, anti is shown-p53R175H in Figureantibody2. (anti-p53R175H, MW 150 kDa) [40]. Finally, suitably designed ap- tamers can be used.

2.5. Description and Characterization of the 4-ATP-NP Probe The structure of the 4-aminothiophenol (4-ATP) molecule, used as Raman marker in our detec- tion method, is shown in Figure 2.

Sensors 2020, 20, x FOR PEER REVIEW 5 of 18

Sensors 2020, 20, x FOR PEER REVIEW 5 of 18 Sensors 2020, 20, 7153 5 of 19

Figure 2. Sketch of the main steps for probe preparation. 4-ATP reacts with gold NP to form the 4- ATP-NP system which is successively activated to yield the diazonium compound able to reacts with Figure 2. Sketch of the main steps for probe preparation. 4-ATP reacts with gold NP to form the a Figureprotein4-ATP-NP 2. target, Sketch system givingof which the mainrise is successively tosteps the forfinal probe activated probe prepara (4 to-ATP yieldtion.- theNP 4-ATP- diazoniumpro) reacts (see also compoundwith the gold text). NP able to to form reacts the with 4- ATPa protein-NP system target, which giving is rise successively to the final activated probe (4-ATP-NP-pro) to yield the dia (seezonium also thecompound text). able to reacts with Thea protein Raman target, spectrum giving rise of to4- theATP final, as probe obtai (4ned-ATP by-NP an-pro) excitation (see also theof 633text). nm, is shown in Figure 3a. In the 1000The–1500 Raman cm spectrum−1 region, of 4-ATP,the spectrum as obtained exhibits by an excitation three main of 633 bands, nm, is shownwith other in Figure small3a. Inpeaks the being 1 hardly1000–1500The detectable. Raman cm− region,spectrum At 2555 the ofcm spectrum 4-−1ATP, a band,, as exhibits obtai attributedned three by mainan to excitation the bands, stretching withof 633 other nm,vibration smallis shown peaks of –inSH Figure being group hardly 3a. ofIn 4-ATP, 1 is thedetectable.also 1000 pre–sent1500 At (see 2555cm−1 the cmregion, −inset, a band,the in Figspectrum attributedure 3a )exhibits [6 to8] the. The stretchingthree assi maingnment vibration bands, of thewith of –SHmain other group bands small of to 4-ATP,peaks the 4 isbeing-ATP also vibra- present (see the inset in Figure−13 a) [ 68]. The assignment of the main bands to the 4-ATP vibrational tionalhardly modes detectable. is reported At 2555 in cm refs., a [39,6band,8 attributed]. The Raman to the spec stretchingtrum of vibration 4-ATP asof –obtainedSH group by of using4-ATP, an exci- modes is reported in refs. [39,68]. The Raman spectrum of 4-ATP as obtained by using an excitation tationis also wavelength present (see of the 514 inset nm in is Fig almosture 3a )the [68 ]same. The assi(notgnment shown) of [40]the main. bands to the 4-ATP vibra- tionalwavelength modes of is514 reported nm is in almost refs. the[39,6 same8]. The (not Raman shown) spec [40trum]. of 4-ATP as obtained by using an exci- tation wavelength of 514 nm is almost the same (not shown) [40].

Figure 3. (a) Raman spectrum of 4-ATP in water solution obtained with 633 nm laser wavelength. FigureFigure 3. 3. ((aa)) RamanRaman spectrumspectrum of of 4 -4ATP-ATP in inwater water solution solution obtained obtained with with633 nm 633 laser nm wavelength. laser wavelength. (b,c) Raman spectra of 4-ATP conjugated to 50 nm gold NPs, obtained with a 633 nm and 514 nm (b(,bc,)c )Raman Raman spectraspectra ofof 4 4--ATPATP conjugated conjugated to to50 50nm nm gold gold NPs, NPs, obtained obtained with awith 633 nma 633 and nm 514 and nm 514 laser nm laser laser wavelength, respectively. The spectral range corresponding to the S–H stretching mode is shown wavelength, respectively. The spectral range corresponding to the S–H stretching mode is shown in wainvelength, the insets. respectively. The yellow bandThe spectral highlights range the region corresponding where the to Raman the S marker–H stretching should appearmode is (not shown in the insets. The yellow band highlights the region where the Raman marker should appear (not present thepresent insets. here). The yellow band highlights the region where the Raman marker should appear (not present here).here). The Raman-SERS spectra of 4-ATP upon their conjugation to 50 nm gold NPs (4-ATP-NP), excited The Raman-SERS spectra of 4-ATP upon their conjugation to 50 nm gold NPs (4-ATP-NP), ex- at 633The nm Raman and 514-SERS nm, arespectra shown of in 4 Figure-ATP 3uponb,c, respectively. their conjugatio In bothn theto spectra,50 nm gold the band NPs centered (4-ATP at-NP), ex- cited at 6331 nm and 514 nm, are shown in Figure 3b,c, respectively. In both the spectra, the band cited2550 at cm 633− disappears nm and 514 (see nm, the insets),are shown witnessing in Figure the formation 3b,c, respectively. of a covalent In bond both between the spectra, the 4-ATP the band centered at 2550 cm−1 disappears (see the insets), witnessing the formation of a covalent bond between thiol group and gold−1 (S-Au bond). The Raman spectra in Figure3b,c substantially show the same centeredthe 4-ATP at 2550thiol cmgroupdisappears and gold (S (see-Au the bond). insets), The witnessingRaman spectra the information Figure 3b of,c asubstantially covalent bon showd between peaks, although with a slightly different relative intensity. Some of these peaks are also observed in the 4-ATP thiol group and gold (S-Au bond). The Raman spectra in Figure 3b,c substantially show thethe spectrumsame peaks, of 4-ATPalthough alone, with even a slightly if with different a small red-shift, relative andintensity. a diff erentSome relative of these intensity; peaks are for also the theobserved same peaks, in the spectrumalthough of with 4-ATP a slightlyalone, even different if with relativea small redintensity.-shift, and Some a different of these relative peaks in- are also assignment of the main peaks of 4-ATP when conjugated to a gold NP see ref. [68]. tensity; for the assignment of the main peaks of 4-ATP when conjugated to a gold NP see ref. [68]. observedThe in enhancement the spectrum of the of 4-ATP4-ATP signals alone, conjugated even if with to gold a small NP, withred-shift, respect and to 4-ATPa different alone, relative was in- The enhancement of the 4-ATP signals conjugated to gold NP, with respect to 4-ATP alone, was tensity;determined for the from assignment the signal of to the noise main (S/ N)peaks ratio of of 4- theATP intensity when conjugated of a select peakto a gold (reference NP see signal) ref. [68]. determinedThe enhancement from the signal of the to 4 noise-ATP (S/N) signals ratio conjugated of the intensity to gold of aNP, select with peak respect (reference to 4 -signal)ATP alone, to was determined from the signal to noise (S/N) ratio of the intensity of a select peak (reference signal) to

Sensors 2020, 20, 7153 6 of 19 Sensors 2020, 20, x FOR PEER REVIEW 6 of 18 that of the noise signal. As reference signal, we selected the peak of the C-S stretching mode, charac- to that of the noise signal. As reference signal, we selected the peak of the C-S stretching mode, −1 −1 terized by a strong Raman signal and detected at 1089 cm and 1078 cm1 for 4-ATP and1 4-ATP-NP, characterized by a strong Raman signal and detected at 1089 cm− and 1078 cm− for 4-ATP and respectively. The noise signal was determined from the intensity averaged in (1800 ± 10) cm−1 spectral 4-ATP-NP, respectively. The noise signal was determined from the intensity averaged in (1800 10) region, 1where no Raman signals are present [39]. The average of the S/N ratio evaluated from ten± cm− spectral region, where no Raman signals are present [39]. The average of the S/N ratio evaluated spectra independently prepared samples, resulted into a SERS enhancement of about seven orders of from ten spectra independently prepared samples, resulted into a SERS enhancement of about seven magnitude with an excitation at 633 nm, and of six orders with an excitation at 514 nm. orders of magnitude with an excitation at 633 nm, and of six orders with an excitation at 514 nm. The conjugation of 4-ATP with NPs was more closely monitored by following the reaction be- The conjugation of 4-ATP with NPs was more closely monitored by following the reaction between tween 4-ATP and NP with optical spectroscopy. 4-ATP and NP with optical spectroscopy. The absorption of bare gold NP, shown in Figure 4a (black line), reveals a band centered at about The absorption of bare gold NP, shown in Figure4a (black line), reveals a band centered at 533 nm arising from the surface plasmon excitation. Upon binding of 4-ATP to gold, the band under- about 533 nm arising from the surface plasmon excitation. Upon binding of 4-ATP to gold, the band goes a slight broadening together with an upward shift of the peak up to 537 nm (red line in Figure undergoes a slight broadening together with an upward shift of the peak up to 537 nm (red line in 4a); with this witnessing the formation of a bond between 4-ATP and NP (4-ATP-NP) [68,69]. The Figure4a); with this witnessing the formation of a bond between 4-ATP and NP (4-ATP-NP) [ 68,69]. shift was found to progressively increase as far as more 4-ATP molecules are bound. To fulfill the The shift was found to progressively increase as far as more 4-ATP molecules are bound. To fulfill the requirements of a high covering of NPs with 4-ATP, optimized conditions were searched by varying requirements of a high covering of NPs with 4-ATP, optimized conditions were searched by varying the concentration of 4-ATP, the temperature and the reaction time (see also below). the concentration of 4-ATP, the temperature and the reaction time (see also below).

FigureFigure 4. (a 4.) Absorption(a) Absorption spectrum spectrum of NPs of NPsbefore before (black (black line) line)and after and after(red line) (red conjugation line) conjugation with 4 with- ATP.4-ATP. The maximum The maximum absorption absorption peaks peaks are marked are marked by dashed by dashed lines lines.. (b) Raman (b) Raman spec spectrumtrum of the of theprobe probe (4-ATP(4-ATP-NP-pro),-NP-pro), sketched sketched in the in theinset inset (not (not in scale) in scale),, charged charged with with a protein a protein (here (here antip53). antip53). The The yellow yellow 1 bandband highlights highlights the theRaman Raman marker marker peaked peaked at 1328 at 1328 cm−1 cm. − .

TheThe 4-ATP 4-ATP-NP-NP system system was was charged charged with with one one of of the the biomolecules biomolecules of interest (Azurin,(Azurin,p53 p53 antibody anti- bodyor or p53) p53) through through a diazoa diazo coupling coupling reaction, reaction, as as sketched sketched in in Figure Figure2 [2 [35,735,700].]. Briefly, Briefly, the free end end of + 4-ATP-NP (the -NH group) was first activated to give rise to the diazonium compound ( +N N), of 4-ATP-NP (the -NH22 group) was first activated to give rise to the diazonium compound (−N−≡N),≡ ableable to react to react with with the theexposed exposed histidine histidine or tyrosine or tyrosine residues residues of a of protein a protein molecule, molecule, forming forming a diaz a diazoo bondbond (N=N) (N= [6N)8] [.68 Such]. Such a reaction a reaction leads leads to the to the4-ATP 4-ATP-NP-NP tightly tightly linked linked a protein, a protein, constituting constituting the the probe,probe, named named 4-ATP 4-ATP-NP-pro.-NP-pro. To avoid To avoid the presence the presence of unbound of unbound reactive reactive sites on sites the onNP the surface, NP surface, the 4-ATPthe- 4-ATP-NP-proNP-pro was further was further treated treated with ETA with (for ETA more (for details more details see ref. see [40] ref.). [40]). FigureFigure 4b 4representativelyb representatively shows shows the the SERS SERS spectrum spectrum of ofthe the 4-ATP 4-ATP-NP-NP probe, probe, as ascharged charged with with antianti-p53.-p53. A comparison A comparison with withthe SERS the SERS spectrum spectrum of 4-ATP of 4-ATP-NP-NP alone alone(see Figure (see Figure 3) indicates3) indicates the ap- the 1 pearanceappearance of an ofadd anitional additional strong strong Raman Raman band band centered centered at about at about 1328 1328 cm−1 cm (highlighted− (highlighted in Figure in Figure 4b).4 b). SuchSuch a band a band can can be assigned be assigned to the to thestretching stretching vibration vibration modes modes of the of thediazo diazo bond bond and and it indicates it indicates the the formationformation of a of covalent a covalent bond bond between between 4-ATP 4-ATP and and the thebiomolecule. biomolecule. Notably, Notably, such such a band a band is the is thesame same irrespectivelyirrespectively of the of the speci specificfic protein protein conjugated conjugated to to the the 4- 4-ATP-NP;ATP-NP; in in our our procedures procedures the the protein protein can can be be Az, p53 p53 antibody antibody or or p53 p53.. The The presence presence of of a well a well distinguishable distinguishable vibrational vibrational fingerprint, fingerprint, combined combined 1 withwith a very a very high high enhanc enhancementement by bythe the plasmonic plasmonic effect, effect, make make the the 1328 1328 cm cm−1 band− band an appropriate an appropriate and and veryvery efficient efficient Raman Raman marker marker (called (called also also Raman Raman fingerprint) fingerprint) to be to monitored be monitored in SERS in SERS-based-based detection detection methodsmethods [68 []68. ].

Sensors 2020, 20, x FOR PEER REVIEW 7 of 18 Sensors 2020, 20, 7153 7 of 19 The number of biomolecules effectively conjugated to 4-ATP-NP, useful to find out optimized conditionsThe for number the probe, of biomolecules can be estimated effectively by fluorescence conjugated to experiments 4-ATP-NP, useful [71]. to The find initial out optimized solution con- tainingconditions the protein for the (here probe, anti can-p53 bewt estimated), used in bythe fluorescence probe assembly experiments and the [71 supernatants,]. The initial solutionafter succes- sive containingincubation the steps, protein were (here investigated anti-p53 wtby), static used fl inuorescence the probe emission. assembly The and signal the supernatants, obtained by an excitationafter successive at 280 nm incubation and collected steps, at were 340 investigatednm, was followed. by static Fluorescence fluorescence emission.emission intensity The signal of the supernatantsobtained by progressively an excitation at decreases 280 nm and as collectedconsequence at 340 of nm, the was removal followed. of Fluorescence unbound molecules. emission The intensity of the supernatants progressively decreases as consequence of the removal of unbound amount of the molecules conjugated to 4-ATP-NP (Cconjugated), can be then evaluated from the follow- ing expression:molecules. The amount of the molecules conjugated to 4-ATP-NP (Cconjugated), can be then evaluated from the following expression: − = − = F0 F (1) Cconjugated = C0 Cunbound = C0 − (1) − F0 where C0 and F0 are the concentration and fluorescence intensity of the initial solution, respectively; where C0 and F0 are the concentration and fluorescence intensity of the initial solution, respectively; unbound whilewhile C Cunbound andand F areF are the the concentration concentration and and the the fluorescence fluorescence intensity intensity of the supernatant,of the supernatant, respectively. respec- 3 3 7 −7 tively.We We found found that about that 2.5about10 2.5molecules × 10 molecules of anti-p53 ofwt ,anti corresponding-p53wt, corresponding to 1.9 10 M, to were 1.9 conjugated× 10 M, were × × − conjugatedto a single to NP.a single By applying NP. By the applying same procedure, the same we procedure, found that about we found 250 and that 980 about molecules 250 ofand p53 980 and mole- culesAz, of respectively,p53 and Az, were respectively, conjugated were to a single conjugated NP. to a single NP.

2.6. Description and Characterization of the Capture Substrate 2.6. Description and Characterization of the Capture Substrate The functionalization procedure of the glass substrates for capturing the 4-ATP-NP probe was The functionalization procedure of the glass substrates for capturing the 4-ATP-NP probe was done by following the steps sketched in Figure5 and reported in ref. [40]. done by following the steps sketched in Figure 5 and reported in ref. [40].

Figure 5. Sketch of the main steps followed for the substrate functionalization. An UV ray treatment to increase the amount of silanol groups is followed by incubation with APTES and GLUTA and then by a Figurefunctionalization 5. Sketch of the with main the biomolecules. steps followed Unreacted for the sitessubstrate were blocked functionalization. with GST. Finally, An UV the ray substrate treatment to increaseis incubated the amount with the of probe silanol (4-ATP-NP-pro) groups is followed (see also by the incubation text). with APTES and GLUTA and then by a functionalization with the biomolecules. Unreacted sites were blocked with GST. Finally, the substrateBriefly, is incubat after aed cleaning with the procedure, probe (4-ATP the-NP glass-pro substrates) (see also werethe text). irradiated by ultraviolet (UV) light to increase the number of silanol groups on the glass surface. They were then treated with 3-aminopropyltriethoxy-silaneBriefly, after a cleaning procedure, (APTES) the dissolved glass substrates in chloroform were and irradiated reacted with by 1% ultraviolet glutaraldehyde (UV) light to increasesolution the (GLUTA). number The of substrates silanol grou wereps then on incubated the glass with surface a drop. They (40 µ L)were of solution then treated containing with the 3-ami- biomolecules at the required concentration; with this promoting the formation of a self-assembled nopropyltriethoxy-silane (APTES) dissolved in chloroform and reacted with 1% glutaraldehyde so- monolayer through the solvent-exposed amino groups. In the approach using antibodies as capture lution (GLUTA). The substrates were then incubated with a drop (40 µL) of solution containing the elements, the substrates were also treated with the glutathione S-transferase (GST) protein to block biomoleculeseventually at unreacted the required sites. Finally,concentration; the functionalized with this substrates promoting were the incubated formation with of a a drop self (10-assembledµL) monolayer through the solvent-exposed amino groups. In the approach using antibodies as capture elements, the substrates were also treated with the glutathione S-transferase (GST) protein to block eventually unreacted sites. Finally, the functionalized substrates were incubated with a drop (10 µL) of the probe at 4 °C, to favor the capture of the molecule conjugated to the 4-ATP probe through a biorecognition process. The incubation times were in the 2–10 h range in order to maximize the num- ber of hot spots showing the features of the 4-ATP-NP probe. The substrates were finally rinsed to

Sensors 2020, 20, 7153 8 of 19

of the probe at 4 ◦C, to favor the capture of the molecule conjugated to the 4-ATP probe through a biorecognition process. The incubation times were in the 2–10 h range in order to maximize the number of hot spots showing the features of the 4-ATP-NP probe. The substrates were finally rinsed to remove uncaptured probes. Blank reference substrates were prepared by following the same procedure, but skipping the incubation step with the biomolecules. For more details see ref. [40].

3. Detection of p53 by SERS

3.1. Overview Different methods for ultrasensitive detection by SERS of wild-type and mutated p53, in buffer and in serum, were described and analyzed in the perspective of an implementation into a biosensor. All the methods exploit the active molecule (4-ATP) conjugated, on one end to gold NPs, and, on the other, to a protein molecule (target or capture element); the resulting system, named 4-ATP-NP-pro, 1 constituting the SERS probe. Indeed, the Raman signal, at 1328 cm− , emerging from the conjugation of 4-ATP to the protein, provides the vibrational fingerprint to be followed for detection. The SERS probes are fluxed onto a substrate on which the biomolecular partners were previously immobilized. The final sample was then scanned by the microscope objective to reveal the Raman fingerprint arising from probes captured by the substrate through a biorecognition process.

3.2. Detection of p53 in Buffer Using Az The procedure, called 4-ATP-NP-p53/Az, to detect p53 in buffer involves a probe functionalized with the target (p53) and a substrate charged with the capture molecule Az. Figure6a shows representative SERS spectra of the substrate before and after incubation with the probe; the substrate having been 11 13 charged at progressively lower concentrations of p53wt (from 5 10 M to 5 10 M). The absence × − × − of any detectable Raman signal in the analysed region in the substrate before incubation with the probe indicates that the substrate itself does not interfere with our Raman analysis. At variance, different peaks appear in the spectra corresponding to samples containing p53wt. In particular, it is evident the 1 presence of the Raman marker at 1328 cm− (marked in yellow in Figure6); this reflecting the capture of the probes from the substrate through a biorecognition process between Az and p53wt. At visual inspection, the intensity of the Raman fingerprint decreases by lowering the concentrations of p53, in agreement with a reduction of hot spots. Such an aspect was, however, quantitatively analysed by evaluating the S/N ratio which exhibited a progressive decrease as far as lower p53 concentrations were taken into consideration. Additionally, the S/N was found to become less than 3 (at which the signal is assumed to be undetectable over the noise), for concentrations below 5 10 13 M (not shown); such a × − value being then assumed as the LOD for p53wt detection in buffer. As a control, we performed an experiment in which 4-ATP-NPs (at a concentration of 1 10 12 M) not charged with p53, were deposed × − on the Az-coated substrate. In this case no signal over the noise was detected, with this indicating that no capture events occur without the target. The same method was applied for detection of p53R249S and p53C135V mutants. Figure6b shows the SERS spectra of the substrate, before and after deposition of the 4-ATP-NPs charged with p53R249S at concentrations from 5 10 11 M to 5 10 13 M. The Raman fingerprint is clearly evident down to × − × − 5 10 13 M, as confirmed by a S/N noise analysis. At lower concentrations, it is no more detectable over × − the noise. Similar results were obtained for the p53C135V mutant (not shown). For both the p53R249S and p53 mutants in buffer, a LOD of 5 10 13 M was found. Accordingly, it can be inferred C135V × − that Az is a suitable capture element for p53wt as well as for the two p53 mutants. Such a finding is consistent with the fact that the analyzed mutations do not interfere with the involved Az-p53 binding site. Indeed, the capability of Az to bind p53, irrespectively of the presence of these two mutations, hinders the possibility of discrimination among p53wt, p53R249S and p53C135V. When such a task is required, other approaches should be used. Sensors 2020, 20, x FOR PEER REVIEW 8 of 18 remove uncaptured probes. Blank reference substrates were prepared by following the same proce- dure, but skipping the incubation step with the biomolecules. For more details see ref. [40].

3. Detection of p53 by SERS

3.1. Overview Different methods for ultrasensitive detection by SERS of wild-type and mutated p53, in buffer and in serum, were described and analyzed in the perspective of an implementation into a biosensor. All the methods exploit the active molecule (4-ATP) conjugated, on one end to gold NPs, and, on the other, to a protein molecule (target or capture element); the resulting system, named 4-ATP-NP-pro, constituting the SERS probe. Indeed, the Raman signal, at 1328 cm−1, emerging from the conjugation of 4-ATP to the protein, provides the vibrational fingerprint to be followed for detection. The SERS probes are fluxed onto a substrate on which the biomolecular partners were previously immobilized. The final sample was then scanned by the microscope objective to reveal the Raman fingerprint aris- ing from probes captured by the substrate through a biorecognition process.

3.2. Detection of p53 in Buffer Using Az The procedure, called 4-ATP-NP-p53/Az, to detect p53 in buffer involves a probe functionalized with the target (p53) and a substrate charged with the capture molecule Az. Figure 6a shows repre- sentative SERS spectra of the substrate before and after incubation with the probe; the substrate hav- ing been charged at progressively lower concentrations of p53wt (from 5 × 10−11 M to 5 × 10−13 M). The absence of any detectable Raman signal in the analysed region in the substrate before incubation with the probe indicates that the substrate itself does not interfere with our Raman analysis. At variance, different peaks appear in the spectra corresponding to samples containing p53wt. In particular, it is evident the presence of the Raman marker at 1328 cm−1 (marked in yellow in Figure 6); this reflecting the capture of the probes from the substrate through a biorecognition process between Az and p53wt. At visual inspection, the intensity of the Raman fingerprint decreases by lowering the concentrations of p53, in agreement with a reduction of hot spots. Such an aspect was, however, quantitatively ana- lysed by evaluating the S/N ratio which exhibited a progressive decrease as far as lower p53 concen- trations were taken into consideration. Additionally, the S/N was found to become less than 3 (at which the signal is assumed to be undetectable over the noise), for concentrations below 5 × 10−13 M (not shown); such a value being then assumed as the LOD for p53wt detection in buffer. As a control, we performed an experiment in which 4-ATP-NPs (at a concentration of 1 × 10−12 M) not charged with p53, were deposed on the Az-coated substrate. In this case no signal over the noise was detected, with Sensors 2020, 20, 7153 9 of 19 this indicating that no capture events occur without the target.

Figure 6. (a) Raman spectra of the capture substrate (functionalized with Az and after the incubation Figure 6. (a) Raman spectra of the capture substrate (functionalized with Az and after the incubation with the 4-ATP-NP-p53 probe charged with p53wt at different concentrations. (b) Raman spectra of withthe capturethe 4-ATP substrate-NP-p53 (functionalized probe charged with with Az p53 andwt afterat different the incubation concentrations. with the (b 4-ATP-NP-p53) Raman spectra probe of the capture substrate (functionalized with Az and after the incubation with the 4-ATP-NP-p53 probe charged with p53R249S at different concentrations. The yellow band highlights the Raman marker 1 peaked at 1328 cm− .

Finally, we explored a reversed strategy, in which the substrate was functionalized with p53, while the 4-ATP-NP was charged with Az molecules. Although in principle such a strategy should be better because a higher number of Az molecules can be bound to a single 4-ATP-NP (980 for Az against 250 for p53), almost the same results in terms of LOD were obtained. This supports that the reached detection level of the method depends on the contribution from different factors whose optimization requires a careful global investigation.

3.3. Detection of p53 in Serum Using Az

The capability of Az to detect p53wt or p53R249S in serum was then investigated. In this case, the large presence of the other molecules (lipids phosholipids, and other proteins, such as albumin and globulins, etc.), might interfere with the Az-p53 interaction, limiting the p53 detection. We selected the configuration in which Az was conjugated with 4-ATP-NP and p53 charged on the substrate, since it resulted more efficient with respect to the reversed one. Accordingly, substrates were functionalized with pure human healthy serum and serum spiked with p53wt or p53R249S at different concentrations. From Figure7, it is evident that no Raman fingerprint over the noise was detected for samples obtained incubating the probe with a substrate functionalized only with pure human serum; with this indicating that Az does not significantly interact with the molecules deposed on the substrate. 1 At variance, the peak at the Raman fingerprint at 1328 cm− appears in the spectra obtained by flowing 10 the probe on substrates functionalized with serum spiked with p53wt or p53 , at 5 10 M, R249S × − 5 10 11 M and 5 10 12 M; spectra related to p53 being shown in Figure7. Other Raman bands × − × − R249S corresponding to 4-ATP-NP can be also observed in these spectra, confirming therefore its presence within the substrate. At lower concentrations, no signal over the noise is detected. Accordingly, a LOD of 5 10 12 M can be derived for detection of p53 in serum. Similar results were obtained for the × − R249S p53wt (not shown). Therefore, this method is able to specifically identify both p53wt and the p53R249S mutant in a serum-like environments, where many possible interference components are present, with a LOD of 5 10 12 M. × − Sensors 2020, 20, 7153 10 of 19 Sensors 2020, 20, x FOR PEER REVIEW 10 of 18 Sensors 2020, 20, x FOR PEER REVIEW 10 of 18

Figure 7. Raman spectra of the capture substrate functionalized with human serum and human se- Figure 7. Raman spectra of the capture substrate functionalized with human serum and human Figurerum/p53 7. RamanR249S spectra mixtures of at the different capture concentrations substrate functionalized of p53R249S and with incubat humaned with serum the 4and-ATP human-NP-Az se- serum/p53 mixtures at different concentrations of p53 and incubated with the 4-ATP-NP-Az probe. TheR249S yellow band highlights the Raman marker peakedR249S at 1328 cm−1. rum/p53R249S mixtures at different concentrations of p53R249S and incubat1ed with the 4-ATP-NP-Az probe. The yellow band highlights the Raman marker peaked at 1328 cm− . probe. The yellow band highlights the Raman marker peaked at 1328 cm−1. 3.4. Detection Detection of of p53 p53 in in Buffer Buffer Usin Usingg Abp53 Abp53

3.4. DetectionThe of detectionp53 in Buffer procedure Using Abp53 4-ATP -NP-Abp53/p53wt, exploits a specific monoclonal antibody The detection procedure 4-ATP-NP-Abp53/p53wt, exploits a specific monoclonal antibody (Abp53) (Abp53) as a capture element for p53. In this case, we used a configuration in which the 4-ATP-NP Theas a detection capture element procedure for p53. 4-ATP In this-NP case,-Abp53/p53 we usedwt a, configuration exploits a specific in which monoclonal the 4-ATP-NP antibody was was functionalized with Abp53, while the substrate was charged with p53 at different concentrations. (Abp53functionalized) as a capture with element Abp53, for while p53. theIn this substrate case, waswe used charged a configuration with p53 at di infferent which concentrations. the 4-ATP-NP Figure8 8aa showsshows thethe SERSSERS spectrum spectrum from from the the capture capture substrate substrate functionalized functionalized with with p53 p53wt(at (at 1010−1122 M)M) was functionalized with Abp53, while the substrate was charged with p53 at differentwt concentrations.− before i incubationncubation with with the the probe. probe. No No Raman Raman signal signal over over the the noise noise can can be observed be observed.. Figure Figure 8a also8a −12 Figureshows 8a shows the SERSthe SERS spectra spectrum from substrates from the functionalized capture substrate with p53functionalizedwt at different with concentrations p53wt (at 10 (the M) also shows the SERS spectra from substrates functionalized with p53wt at different concentrations before10 incubation−12, 10−1125 and with 1015−17 theM) afterprobe.17 incubation No Raman with signalthe 4-ATP over-NP the-Abp53 noisewt probe.can be T observedhe Raman. fingerprint Figure 8a atalso (the 10− , 10− and 10− M) after incubation with the 4-ATP-NP-Abp53wt probe. The Raman shows1328 the cmSERS−1 (marked spectra in from yellow1 substrates), can be clea functionalizedrly detected in allwith the p53spectrawt at witnessing different theconcentrations presence of the (the fingerprint at 1328 cm− (marked in yellow), can be clearly detected in all the spectra witnessing the −12 4-−1ATP5 -NP-Abp53−17 wt probe as due to a biorecognition process between Abp53 and p53wt. More specif- 10 , 10presence and of10 the M 4-ATP-NP-Abp53) after incubationwt probe with as the due 4- toATP a biorecognition-NP-Abp53wt process probe. between The Raman Abp53 fingerprint and p53wt. at 1328 cmically,More−1 (marked specifically,the Raman in yellow fingerprint the Raman), can is fingerprintbe characterized clearly isdetected characterized by a S/N in signalall the by higher aspectra S/N signalthan witnessing 3, higher for a target than the concentration 3,presence for a target of the down to 10−17 M, while it becomes17 practically undetectable over the noise, at lower concentrations. 4-ATPconcentration-NP-Abp53wt downprobe to as 10 due− M,to a while biorecognition it becomes practicallyprocess between undetectable Abp53 over and the p53 noise,wt. M atore lower specif- Accordingly, a LOD of 10−17M was assumed17 for detection of p53wt by this method. ically, concentrations.the Raman fingerprint Accordingly, is characterized a LOD of 10− byM a was S/N assumed signal higher for detection than 3, of for p53 awt targetby this concentration method. down to 10−17 M, while it becomes practically undetectable over the noise, at lower concentrations. Accordingly, a LOD of 10−17M was assumed for detection of p53wt by this method.

Figure 8. (a) Raman spectra of the capture substrate, functionalized with p53wt at different Figure 8. (a) Raman spectra of the capture substrate, functionalized with p53wt at different concentra- concentrations, before and after incubation with the 4-ATP-NP-pro probe charged with antip53wt. tions, before and after incubation with the 4-ATP-NP-pro probe charged with antip53wt. (b) Raman (b) Raman spectra of the capture substrate functionalized with p53 , at different concentrations, spectra of the capture substrate functionalized with p53R175H, at differentR175H concentrations, before and before and after the incubation with the 4-ATP-NP-pro probe charged with antip53R175. The yellow after the incubation with the 4-ATP-NP-pro probe charged with antip53R175. The yellow band high- 1 band highlights the Raman marker peaked at−1 1328 cm− . Figurelights 8. (a) the Raman Raman spectra marker of peaked the capture at 1328 subs cm trate,. functionalized with p53wt at different concentra-

tions, beforeSimilar and results after were incubation obtained with by the incubating 4-ATP-NP the-pro 4-ATP-NP-Abp53 probe charged withR175H antiprobep53 withwt. (b) substrates Raman Similar results were obtained12 by incubating17 the 4-ATP-NP-Abp53R175H probe with substrates spectracharged of the with capture p53R175H substrate, in the functionalized 10− –10− M with concentration p53R175H, at range different and concentrations, shown in Figure before8b. Again, and charged with p53R175H, in the 10−12–10−17 M concentration range and shown in Figure 8b. Again, the after the incubation with the 4-ATP-NP-pro probe charged with antip53R175. The yellow band high- characteristic Raman fingerprint at 1328 cm−1 (marked in yellow in Figure 8b) can be clearly observed lights the Raman marker peaked at 1328 cm−1. down to 10−17 M; such a concentration indicating the LOD of this method. As a further control, the 4-

Similar results were obtained by incubating the 4-ATP-NP-Abp53R175H probe with substrates charged with p53R175H, in the 10−12–10−17 M concentration range and shown in Figure 8b. Again, the characteristic Raman fingerprint at 1328 cm−1 (marked in yellow in Figure 8b) can be clearly observed down to 10−17 M; such a concentration indicating the LOD of this method. As a further control, the 4-

Sensors 2020, 20, x FOR PEER REVIEW 11 of 18

ATP-NP-Abp53wt probe (or the 4-ATP-NP-Abp53R175H probe) was incubated on the blank substrate. No Raman signal over the noise were found in both cases (not shown). This means that the probes do not undergo a significant interacting with the blank substrate. The possibility of a cross-reactions between p53wt and p53R175H was evaluated by analyzing the following samples: (i) the 4-ATP-NP-Abp53wt probe combined with the substrate functionalized with p53R175H; and (ii) the 4-ATP-NP-Abp53R175H probe combined with the substrate functionalized. In both the case the substrate was charged the target at a concentration of 10−15 M. The characteristic Raman fingerprint at 1328 cm−1 was not detected over the noise level from the Raman spectra of both the samples (not shown). These results therefore demonstrate that this method is able to well discrimi- nate between the p53wt and p53R175H. Remarkably, the use of specific antibodies of p53wt allowed us to reach a lower detection level of p53wt in comparison to that using Az. Such a sensitivity improvement in the for p53wt detection can be mainly attributed to a more specific recognition by the antibody with respect to Az; however, a contribution from other aspects, such as the optimized immobilization Sensors 2020, 20, 7153 11 of 19 strategy, cannot be ruled out. In particular, the use of GST to block unreacted sites could help in maximizing biorecognition events. 1 the characteristic Raman fingerprint at 1328 cm− (marked in yellow in Figure8b) can be clearly observed down to 10 17 M; such a concentration indicating the LOD of this method. As a further 3.5. Detection of p53 in Serum− Using Abp53 control, the 4-ATP-NP-Abp53wt probe (or the 4-ATP-NP-Abp53R175H probe) was incubated on the blankTo apply substrate. the developed No Raman method signal over to the the noisedetection were of found p53 in bothserum, cases we (not preliminarily shown). This analysed means the substratethat the functionalized probes do not undergowith bare a significant healthy human interacting serum. with The the blankRaman substrate. spectrum, shown in Figure 9, reveals theThe presence possibility of oftwo a cross-reactionsbroad bands at between about 1050 p53wt andand 1175 p53 R175Hcm−1. wasThese evaluated bands were by analyzing not observed in thethe substrate following function samples:alized (i) the only 4-ATP-NP-Abp53 with p53 (seewt Figureprobe combined 8), they are with likely the substrate attributable functionalized to the various biomoleculeswith p53R175H present; and (ii)in serum the 4-ATP-NP-Abp53 (lipids phosholipidR175H probes, and combined other proteins, with the such substrate as albumin functionalized. and globu- 15 lins,In etc. both), and the casedeposed the substrate on the wassubstrate. charged However the target, atthe a concentration absence of Raman of 10− signalsM. The characteristicin the 1328 cm−1 Raman fingerprint at 1328 cm 1 was not detected over the noise level from the Raman spectra of region allowed us to use our probe− for detection. Notably, these bands were not detected by the Az- both the samples (not shown). These results therefore demonstrate that this method is able to well based method (see Figure 7) likely due to the slightly different functionalization procedures (see Sec- discriminate between the p53wt and p53R175H. Remarkably, the use of specific antibodies of p53wt tion 2.6. allowed us to reach a lower detection level of p53wt in comparison to that using Az. Such a sensitivity wt improvementThen, we analysed in the for the p53 sampleswt detection obtained can bebymainly incubating attributed the 4- toATP a more-NP- specificAbp53 recognitionprobe (or by4-ATP- NP-Abp53the antibodyR175H) on with the respect substrates to Az; previously however, a functionalized contribution from with other human aspects, serum such spiked as the optimized with p53wt (or p53R175Himmobilization) at different strategy, concentrati cannotons. be ruledWe found out. Inthat particular, the Raman the usefingerprint of GST to at block 1328 unreacted cm−1 is clearly sites de- tectablecould for help both in maximizingp53wt and p53 biorecognitionR175H samples, events. down to 10−15 M; representative SERS spectra for p53wt being shown in Figure 9. In all the cases the S/N ratio was found to be higher than 3, while at lower 3.5. Detection of p53 in Serum Using Abp53 concentrations, the Raman fingerprint is no more detectable over the noise. Consequently, the 10−15 M concentrationTo apply constitutes the developed the method LOD for to thethedetection p53wt and of p53 p53R175H in serum, detection we preliminarily in human serum. analysed The the cross- substrate functionalized with bare healthy human serum. The Raman spectrum, shown in Figure9, reaction between our probes (Abp53wt or Abp53R175H) and the healthy human serum constituents, was 1 reveals the presence of two broad bands at about 1050 and 1175 cm− . These bands were not observed checked. The spectrum obtained by incubating the 4-ATP-NP-Abp53wt, probe on substrate function- in the substrate functionalized only with p53 (see Figure8), they are likely attributable to the various alized with healthy human serum, does not reveal any significant signal at 1328 cm−1 over the noise. biomolecules present in serum (lipids phosholipids, and other proteins, such as albumin and globulins, This means that the probe was not captured by the substrate. A similar spectrum was 1obtained by etc.), and deposed on the substrate. However, the absence of Raman signals in the 1328 cm− region droppingallowed the us 4 to-ATP use our-NP probe-Abp53 forR175H detection. probe Notably, on the barethese healthybands were human not detected serum- byfunctionalized the Az-based sub- strate.method (see Figure7) likely due to the slightly di fferent functionalization procedures (see Section 2.6).

Figure 9. Raman spectra of the capture substrate (functionalized bare healthy human serum) and of 15 functionalized with human serum spiked with p53wt or p53R175H at 10− M and incubated with the 4-ATP/NP probe charged with anti-p53wt. The yellow band highlights the Raman marker peaked at 1 1328 cm− .

Then, we analysed the samples obtained by incubating the 4-ATP-NP-Abp53wt probe (or 4-ATP-NP-Abp53R175H) on the substrates previously functionalized with human serum spiked with 1 p53wt (or p53R175H) at different concentrations. We found that the Raman fingerprint at 1328 cm− 15 is clearly detectable for both p53wt and p53R175H samples, down to 10− M; representative SERS spectra for p53wt being shown in Figure9. In all the cases the S /N ratio was found to be higher Sensors 2020, 20, 7153 12 of 19 than 3, while at lower concentrations, the Raman fingerprint is no more detectable over the noise. 15 Consequently, the 10− M concentration constitutes the LOD for the p53wt and p53R175H detection in human serum. The cross-reaction between our probes (Abp53wt or Abp53R175H) and the healthy human serum constituents, was checked. The spectrum obtained by incubating the 4-ATP-NP-Abp53wt, probe on substrate functionalized with healthy human serum, does not reveal any significant signal 1 at 1328 cm− over the noise. This means that the probe was not captured by the substrate. A similar spectrum was obtained by dropping the 4-ATP-NP-Abp53R175H probe on the bare healthy human serum-functionalized substrate. Additionally, the possibility of some interference between p53wt and p53R175H, was investigated by taking into consideration the samples in which the 4-ATP-NP-Abp53wt probe was used in connection with serum spiked with both p53wt and p53R175H; with both the targets were at a concentration of 15 10− M. We found a spectrum almost the same as that obtained from the serum sample in which only p53wt was present (Figure9). The same results were obtained when the 4-ATP-NP-Abp53 R175H probe was used with serum spiked with both p53wt and p53R175H. These results indicate that our approach is able to well discriminate between the p53wt and p53R175H. In summary, the presented method is able 15 to selectively detect both p53wt and p53R175H proteins down to 10− M in human serum.

3.6. Comparison with Other SERS-Based Detection Methods for p53 As already mentioned, several different approaches were applied for detection of p53 (wild-type and mutated forms). An overview of these methods, together with the related advantages and limitations, were recently presented (see ref. [15]). On the other hand, although a large variety of SERS-based approaches were developed for detection of several different biomolecules, a very few approaches were developed for p53. Here, we briefly reviewed some SERS-based methods to detect p53 available in literature. The main features and the detection results of these methods are summarized in Table1; for comparison data from ELISA, which represents a benchmark, being also reported. Practically, almost all the methods use antibodies as capture elements, however, they are conjugated with rather different molecular strategies; all of them having been able to reach a rather low detection level. As emerging from Table1, among other methods, our approaches combine the capability of a very low LOD value, with a wide concentration range. In particular, the antibody-based method can reach a very competitive detection level even with respect to the other techniques.

Table 1. SERS-based biosensing methods for detection of p53.

Concentration Description Capture Element LOD References Range Au NPs Azurin 5 10 13 M 10 13–10 12 M [39] × − − − Au NPs Antibody 1 10 17 M 10 17–10 12 M [40] × − − − Ag substrate Antibody 5 10 10 M 10 10–10 6 M [72] × − − − Au-Ag Nanorods Antibody 1 pg/mL - [73] ELISA Antibody 5 10 12 M - Commercial ELISA × −

4. Analysis of the Performance Parameters in the Perspective to Build a Biosensor for p53 The biosensing approach using antibodies combined with an optimized immobilization strategies, resulted to be the most efficient one to detect p53 in terms of detection level. Indeed, we evaluated, 17 15 a LOD of 10− M in buffer and of 10− M in serum; with these values being comparable with the lowest values available in literature for p53 even by other techniques (see e.g., ref. [15]). Furthermore, the method is also characterized by a good selectivity and accuracy. On such a basis, it could be a suitable candidate to be implemented in a biosensor for detection of wild-type and mutated p53 for clinical uses. In the following, the most relevant performance parameters (LOD, reproducibility, sensitivity, linearity, etc.) of this method were summarized in the perspective of applicative developments. Sensors 2020, 20, 7153 13 of 19

To test the reproducibility of our detection method, we analysed the peak of the Raman fingerprint by considering five independently prepared samples; for each of them, the integrated area of the 1 SensorsRaman 2020 band, 20, xpeaked FOR PEER at 1328REVIEW cm− , of ten spectra from different regions, was evaluated. The average, 13 of 18 15 and the corresponding standard deviation of this area, for the five samples of p53wt at 10− M and 15 areafor values p53R175H areat the 10− saM,me in within buffer the and errors. in serum, This are assesses shown in the Figure reproducibility 10a. For each of case, the the detection five area method values are the same within the errors. This assesses the reproducibility of the detection method for both for both p53wt and p53R175H in the different environment. Notably, the values for p53 in serum are p53wt and p53 in the different environment. Notably, the values for p53 in serum are significantly significantly lowerR175H (about 400 a.u.) than those obtained in buffer at the same concentration (about 600 lower (about 400 a.u.) than those obtained in buffer at the same concentration (about 600 a.u.). Such a a.u.).decrease Such coulda decrease be ascribed could be to ascribeda reduction to a of reduction the number of the of specific number binding of specific sites binding available sites in the available in substratesthe substrates functionalized functionalized with serum, with whose serum, molecules whose may molecules partially may hinder partially the interaction hinder the between interaction betweenp53 and p53 the 4-ATP-NPand the 4- probe.ATP-NP probe.

1 Figure 10. (a) Average integrated area of the Raman peak detected at 1328 cm− for p53wt (top), −1 Figurein buff 10.er (green)(a) Average and in integrated serum (red), area and of for the p53 R175HRaman(bottom) peak detected in buffer (azur) at 1328 and cm in serumfor p53 (blue)wt (top), in bufferfor five (green) different and p53 in samples.serum (red), The averageand for and p53 theR175H corresponding(bottom) in buffer standard (azur) deviations and in (bars) serum were (blue) for fivederived different from p53 ten samples. spectra acquired The average at different and regions the corresponding of the sample. standard (b) Calibration deviations plot obtained(bars) were de- 1 rivedfrom from the average ten spectra integrated acquired area of at thedifferent Raman regions peak at 1328 of the cm sample.− , as a function (b) Calibration of the logarithm plot obtained of the from p53wt concentration (full squares) and of p53R175H concentration (open circles). The average and the the average integrated area of the Raman peak at 1328 cm−1, as a function of the logarithm of the p53wt corresponding standard deviation were obtained from ten spectra acquired at different regions of the concentration (full squares) and of p53R175H concentration (open circles). The average and the corre- sample. The continuous lines represent the fit by the equation Y = A + BX (for the extracted values see sptheonding text); standard the resulting deviation adjusted were R2 values obtained being from reported. ten spectra The red acquired stars mark at different the integrated regions area of asthe sam- ple. The continuous lines represent the fit by the equation Y = A + BX (for the extracted values see the obtained for two unknown samples containing p53R175H (see also ref. [40]). text); the resulting adjusted R2 values being reported. The red stars mark the integrated area as ob-

tainedTo obtain for two the unknown sensitivity samples of our approach,containing a p53 calibrationR175H (seeplot also wasref. [40] derived). by using the integrated 1 area of the Raman marker, peaked at 1328 cm− . The averages and the corresponding standard 10 deviationsTo obtain of thisthe areasensitivity are plotted of our as aapproach, function of a thecalibration logarithm plot of thewas concentration derived by us(froming 10 the− integratedM 17 areaand of 10 the− RamanM), for p53marker,wt (squares) peaked and at p53 1328R175H cm(open−1. The circles) averages (see Figureand the 10b). corresponding We found a very standard good devi- linear relationship in the whole concentration range for both the cases. In Figure 10b, the resulting ations of this area are plotted as a function of the logarithm of the concentration (from 10−10 M and fitting curves (continuous lines) by the equation Y = A + BX, where Y is the averaged area intensity 10−17 M), for p53wt (squares) and p53R175H (open circles) (see Figure 10b). We found a very good linear and X is the logarithm of the corresponding p53 concentration, are also shown. The goodness of the fit relationshipwas assessed in bythe the whole chi-square concentration test. The parameters range for both extracted the cases.from the In fitFigure are: A 10b,= (2520 the resulting110) a.u. fitting curves (continuous lines) by the equation Y = A + BX, where Y is the averaged area intensity± and X is and B = (133 10) a.u./decade for p53wt and A = (2300 180) a.u. and B = (120 14) a.u/decade for ± ± ± thep53 logarithmR175H; with of the the slope corresponding B of the calibration p53 concentrat plot providingion, theare sensitivity also shown. of the The method. goodness Although of the the fit was assessedslope average by the appears chi-square slightly test. higher The parameters for p53wt with extracted respect tofrom p53 R175hthe fit, the are: sensitivity A = (2520 is, ± within 110) a.u. the and B = (133error, ± almost 10) a.u./decade the same. for p53wt and A = (2300 ± 180) a.u. and B = (120 ± 14) a.u/decade for p53R175H; with the slope B of the calibration plot providing the sensitivity of the method. Although the slope average appears slightly higher for p53wt with respect to p53R175h, the sensitivity is, within the error, almost the same. The calibration plot was then used to evaluate the capability for our biosensor to quantify p53 samples at unknown concentration. With such an aim, we determined the integrated area of the 1328 cm−1 peak from the SERS signal from two samples of p53R175H at the unknown concentrations (named US1 and US2). Upon evaluating the average of the area, the corresponding concentration were ex- tracted from the calibration plot (see stars in Figure 11b). We found CUS1 = (2.0 ± 0.6)10−12 M and CUS2 = (16.2 ± 0.8)10−12 M. The same samples were tested by applying the photometric one-step-enzyme immunoassay p53 ELISA kit for the quantification of unknown samples of p53R175H (100 µl, diluted Sensors 2020, 20, x FOR PEER REVIEW 14 of 18

1:5 in sample diluent) (see ref. [40]). We found: CUS1(ELISA) = (1.8 ± 0.4)10−12 M and CUS2(ELISA) = (15 ± 3)10−12 M; with both of them resulting statistically consistent with the values obtained by SERS ex- periments. Such an agreement validates the accuracy of our SERS-based method in the quantification of p53 concentration. In summary, our SERS-based method satisfies a set of requirements which make it suitable for applications in clinical diagnostics.

5. Possible Development: Use of Aptamers as Capture Elements for p53 As a possible development of our biosensing approach for detecting p53, we focused our atten- Sensors 2020, 20, 7153 14 of 19 tion to decrease the LOD for the p53R175H mutant, whose presence, even at very small amount, could signal a disease at very early stage [59]. With such an aim, we preliminarily investigated the feasibility of using aptamers,The calibration instead plot of was antibodies then used (or to Az), evaluate as capture the capability molecules for our for biosensorp53. Aptamers to quantify are synthet p53 ic molecules,samples often at unknown oligonucleotides, concentration. designed With to such specifically an aim, we bind determined a target upon the integrated undergoing area a of bior theecog- 1 nition1328 process cm− peak [74]. from In the the following, SERS signal a from preliminarily two samples study of p53 finalizedR175H at the to unknownthe implementation concentrations of ap- (named US1 and US2). Upon evaluating the average of the area, the corresponding concentration tamers in our SERS-based approach is outlined. For detection of the p53R175H mutant, we can12 take were extracted from the calibration plot (see stars in Figure 11b). We found CUS1 = (2.0 0.6)10− advantage of the results as obtained12 by a SELEX procedure which identified a specific aptamer± (called M and CUS2 = (16.2 0.8)10− M. The same samples were tested by applying the photometric here APT175) (see ref. [7±5]), having a strong affinity for p53R175H and a much less for p53wt. Such an one-step-enzyme immunoassay p53 ELISA kit for the quantification of unknown samples of p53R175H aptamer(100 µisl, also diluted able 1:5 to in restore sample the diluent) anticancer (see ref. activity [40]). We of found: mutated C p53.(ELISA) Starting= (1.8 from0.4)10 the 12availableM and se- US1 ± − quenceC of(ELISA) APT175= (15 (ATTAGCGCATTTTAACATAGGGTGC3)10 12 M; with both of them resulting statistically) from ref. consistent [75], the with corresponding the values 3D US2 ± − structureobtained was by modelled SERS experiments. by following Such the an agreementsame procedure validates described the accuracy in ref. of our[76] SERS-based, using the RNAFOLD, method [77] inand the the quantification SIMRNA softwa of p53re concentration. [78]. The extracted In summary, best model our SERS-based for APT175 method is shown satisfies in Figure a set 11a). of requirements which make it suitable for applications in clinical diagnostics.

Figure 11. (a) Graphical representation of the best model for the APT175 aptamer. (b) Graphical Figure 11. (a) Graphical representation of the best model for the APT175 aptamer. (b) Graphical rep- representation of the two predicted best arrangements for the APT175 with respect to DBDR175H in the resentationcomplex. of The the 3 0twoend predicted of APT175 isbest marked arrangements in red. The for figures the wereAPT175 created with by respect Pymol [55to ].DBDR175H in the complex. The 3′ end of APT175 is marked in red. The figures were created by Pymol [55]. 5. Possible Development: Use of Aptamers as Capture Elements for p53

For Asdetection, a possible we developmentare intended ofto ouruse biosensingthe strategy approach in which for the detecting target (p53 p53,R175H we) focused is immobilized our on theattention substrate, to decrease while the the capture LOD for element the p53R175H is conjugatedmutant, whose with presence, 4-ATP-NP even to atform very the small probe amount,. Accord- ingly,could a His signal tag is a diseaseto be added at very to early one stageof the [ 59APT175]. With end such. T anhe aim, choice we preliminarilyof which APT175 investigated end should the be chargedfeasibility with ofthe using His aptamers,tag, was insteaddone by of antibodiespreliminarily (or Az), performing as capture a molecules computational for p53. docking Aptamers between are synthetic molecules, often oligonucleotides, designed to specifically bind a target upon undergoing a the APT175 and the DBDR175H portion of p53R175H using HDOCK [79]. The structure of the DBDR175H biorecognition process [74]. In the following, a preliminarily study finalized to the implementation of portion was derived from the B chain of DBD in complex with a consensus DNA (1 TUP entry from aptamers in our SERS-based approach is outlined. For detection of the p53R175H mutant, we can take PDB) [51]), by replacing the Arg175 residue with His175 through the I-TASSER suite [80]. Docking advantage of the results as obtained by a SELEX procedure which identified a specific aptamer (called was carried out by also requiring the further conditions that the His175 residue of p53 is to be in- here APT175) (see ref. [75]), having a strong affinity for p53R175H and a much less for p53wt. Such an volvedaptamer at the is binding also able int toerface restore. The the first anticancer ten complex activity models of mutated with p53.the best Starting docking from scores the available were taken into sequenceconsideration of APT175 for further (ATTAGCGCATTTTAACATAGGGTGC) analysis. In these models, the ligand from (APT175) ref. [75], thewas corresponding found to cluster 3D into two structuremain groups was modelled around bythe following receptor the (DBD sameR1 procedure75H), as representatively described in ref. [76 shown], using in the Figure RNAFOLD 11b). [ 77In] both the cases,and the the SIMRNA 3′ end softwareof the aptamer [78]. The results extracted to be best availabl model fore for APT175 hosting is shown a His intag Figure without 11a. substantial interferingFor with detection, the p53 we are bin intendedding. The to use capability the strategy of in the which approach the target using (p53 R175H APT175) is immobilized-His-tag to on detect the substrate, while the capture element is conjugated with 4-ATP-NP to form the probe. Accordingly, p53R175H is under investigation and preliminarily results are rather encouraging. a His tag is to be added to one of the APT175 end. The choice of which APT175 end should be charged with the His tag, was done by preliminarily performing a computational docking between the APT175

and the DBDR175H portion of p53R175H using HDOCK [79]. The structure of the DBDR175H portion was derived from the B chain of DBD in complex with a consensus DNA (1 TUP entry from PDB) [51]), by replacing the Arg175 residue with His175 through the I-TASSER suite [80]. Docking was carried out Sensors 2020, 20, 7153 15 of 19 by also requiring the further conditions that the His175 residue of p53 is to be involved at the binding interface. The first ten complex models with the best docking scores were taken into consideration for further analysis. In these models, the ligand (APT175) was found to cluster into two main groups around the receptor (DBDR175H), as representatively shown in Figure 11b). In both the cases, the 30 end of the aptamer results to be available for hosting a His tag without substantial interfering with the p53 binding. The capability of the approach using APT175-His-tag to detect p53R175H is under investigation and preliminarily results are rather encouraging.

6. Conclusions Detection of p53 at very low concentrations represents an important goal for early diagnostics and prognosis of cancer diseases. SERS, combined with suitably devoted nanotechnological architectures, provides a valuable biosensing approach for ultrasensitive detection of p53. The presented methods 17 15 were able to detect wild-type and single point mutated p53, down to 10− M in buffer and to 10− M in human serum. Such a detection capability deserves a high interest for signaling abnormal cancer states or other diseases, with high potentialities in nanomedicine. Among the various approaches that have been reviewed, that one using a specific antibody as capture element of p53, combined with an optimized biomolecular immobilization strategy, was found to exhibit the highest sensitivity. At the same time, it is also characterized by a high reproducibility, accuracy, linearity in a rather wide concentration range, and even a high selectivity between p53wt and the p53R175H mutant. All these features make it extremely suitable for its implementation as a biosensing platform for clinical applications. Among possible improvements, the use of aptamers suitably designed for targeting the p53R175H specific mutants, as alternative capture elements for p53, could be envisaged. From a more general point of view, the method can be also adapted to detect other protein biomarkers upon selecting suitable capture molecules and defining the architecture and functionalization procedures for an optimized detection.

Author Contributions: Conceptualiation, investigation, writing—review and editing, A.R.B. and S.C. All authors have read and agreed to the published version of the manuscript. Funding: This research received no external funding. Conflicts of Interest: The authors declare no conflict of interest.

References

1. Lane, D.P. p53, guardian of the genome. Nature 1992, 358, 15–16. [CrossRef][PubMed] 2. Levine, A.J.; Oren, M. The first 30 years of p53: Growing ever more complex. Nat. Rev. Cancer 2009, 9, 749–758. [CrossRef][PubMed] 3. Vogelstein, B.; Lane, D.; Levine, A.J. Surfing the p53 network. Nature 2000, 408, 307–310. [CrossRef][PubMed] 4. Freedman, D.A.; Wu, L.; Levine, A.J. Functions of the MDM2 oncoprotein. Cell. Mol. Life Sci. 1999, 55, 96–107. [CrossRef] 5. Moll, U.M.; Petrenko, O. The MDM2-p53 interaction. Mol. Cancer Res. 2003, 1, 1001–1008. 6. Vogelstein, B.; Sur, S.; Prives, C. p53: The most frequently altered gene in human cancers. Nat. Educ. 2010, 3, 6. 7. Joerger, A.C.; Fersht, A.R. The p53 Pathway: Origins, inactivation in cancer, and emerging therapeutic approaches. Annu. Rev. Biochem. 2016, 85, 375–404. [CrossRef] 8. Liu, D.P.; Song, H.; Xu, Y. A common gain of function of p53 cancer mutants in inducing genetic instability. Oncogene 2010, 29, 949–956. [CrossRef] 9. Joruiz, S.M.; Beck, J.A.; Horikawa, I.; Harris, C.C. The ∆133p53 isoforms, tuners of the p53 pathway. Cancers 2020, 12, 3422. [CrossRef] 10. Goldstein, I.; Marcel, V.; Olivier, M.; Oren, M.; Rotter, V.; Hainaut, P. Understanding wild-type and mutant p53 activities in human cancer: New landmarks on the way to targeted therapies. Cancer Gene Ther. 2011, 18, 2–11. [CrossRef] Sensors 2020, 20, 7153 16 of 19

11. Oh, M.; Choi, J.; Ho, Y.; Kwan, J.; Young, J.; Kyun, Y.; Wan, K.; Yong, S.; Saw, H. Mutant p53 protein in the serum of patients with cervical carcinoma: Correlation with the level of serum epidermal growth factor receptor and prognostic significance. Cancer Lett. 2004, 203, 107–112. [CrossRef][PubMed] 12. Olivier, M.; Hollstein, M.; Hainaut, P. TP53 mutations in human cancers: Origins, consequences, and clinical use. Cold Spring Harb. Perspect. Biol. 2010, 2, a001008. [CrossRef][PubMed] 13. Balogh, G.A.; Mailo, D.; Nardi, H.; Corte, M.M.; Vincent, E.; Barutta, E.; Lizarraga, G.; Lizarraga, P.; Montero, H.; Gentili, R. Serological levels of mutated p53 protein are highly detected at early stages in breast cancer patients. Exp. Ther. Med. 2010, 1, 357–361. [CrossRef][PubMed] 14. Attallah, A.M.; Abdel-Aziz, M.M.; El-Sayed, A.M.; Tabll, A.A. Detection of serum p53 protein in patients with different gastrointestinal cancers. Cancer Detect. Prev. 2003, 27, 127–131. [CrossRef] 15. Pundir, S.; Pundir, C.S. Detection of tumor suppressor protein p53 with special emphasis on biosensors: A review. Anal. Biochem. 2020, 588, 113473. 16. Tong, Z.; Liu, N.; Lin, L.; Guo, X.; Yang, D.; Zhang, Q. miR-125a-5p inhibits cell proliferation and induces apoptosis in colon cancer via targeting BCL2, BCL2L12 and MCL1. Biomed. Pharmacother. 2015, 75, 129–136. [CrossRef] 17. Zhang, B.; Zhang, X.; Yan, H.; Xu, S.; Tang, D.; Fu, W. A novel multi-array immunoassay device for tumor markers based on insert-plug model of piezoelectric immunosensor. Biosens. Bioelectron. 2007, 23, 19–25. [CrossRef] 18. Han, S.H.; Kim, S.K.; Park, K.; Yi, S.Y.; Park, H.J.; Lyu, H.K.; Kim, M.; Chung, B.H.; Hee, S.; Kyu, S.; et al. Detection of mutant p53 using field-effect transistor biosensor. Anal. Chim. Acta 2010, 665, 79–83. [CrossRef] 19. Jiang, T.; Minunni, M.; Wilson, P.; Zhang, J.; Turner, A.P.F.; Mascini, M. Detection of TP53 mutation using a portable surface plasmon resonance DNA-based biosensor. Biosens. Bioelectron. 2005, 20, 1939–1945. [CrossRef] 20. Li, H.; Wu, Z.; Qiu, L.; Liu, J.; Wang, C.; Shen, G.; Yu, R. Ultrasensitive label-free amplified colorimetric detection of p53 based on G-quadruplex MBzymes. Biosens. Bioelectron. 2013, 50, 180–185. [CrossRef] 21. Bantz, K.C.; Meyer, A.F.; Wittenberg, N.J.; Im, H.; Kurtulu¸s,Ö.; Lee, S.H.; Lindquist, N.C.; Oh, S.-H.; Haynes, C.L. Recent progress in SERS biosensing. Phys. Chem. Chem. Phys. 2011, 13, 11551–11567. [CrossRef] [PubMed] 22. Wells, M.; Tidow, H.; Rutherford, T.J.; Markwick, P.; Jensen, M.R.; Mylonas, E.; Svergun, D.I.; Blackledge, M.; Fersht, A.R. Structure of tumor suppressor p53 and its intrinsically disordered N-terminal transactivation domain. Proc. Natl. Acad. Sci. USA 2008, 105, 5762–5767. [CrossRef][PubMed] 23. Habchi, J.; Tompa, P.; Longhi, S.; Uversky, V.N. Introducing protein intrinsic disorder. Chem. Rev. 2014, 114, 6561–6588. [CrossRef][PubMed] 24. Minde, D.P.; Halff, E.F.; Tans, S. Designing disorder. Intrinsically Disord. Proteins 2013, 1, e26790. 25. Fung, H.Y.J.; Birol, M.; Rhoades, E. IDPs in macromolecular complexes: The roles of multivalent interactions in diverse assemblies. Curr. Opin. Struct. Biol. 2018, 49, 36–43. [CrossRef] 26. Coppari, E.; Yamada, T.; Bizzarri, A.R.; Beattie, C.W.C.W.; Cannistraro, S. A nanotechnological, molecular-modeling, and immunological approach to study the interaction of the anti-tumorigenic peptide p28 with the p53 family of proteins. Int. J. Nanomed. 2014, 9, 1799–1813. 27. Moscetti, I.; Bizzarri, A.R.; Cannistraro, S. Imaging and kinetics of the bimolecular complex formed by the tumor suppressor p53 with ubiquitin ligase COP1 as studied by atomic force microscopy and surface plasmon resonance. Int. J. Nanomed. 2018, 13, 251. [CrossRef] 28. Moscetti, I.; Cannistraro, S.; Bizzarri, A.R. Surface plasmon resonance sensing of biorecognition interactions within the tumor suppressor P53 network. Sensors 2017, 17, 2680. [CrossRef] 29. Funari, G.; Domenici, F.; Nardinocchi, L.; Puca, R.; D’Orazi, G.; Bizzarri, A.R.; Cannistraro, S. Interaction of p53 with Mdm2 and azurin as studied by atomic force spectroscopy. J. Mol. Recognit. 2010, 23, 343–351. [CrossRef] 30. Yamada, T.; Christov, K.; Shilkaitis, A.; Bratescu, L.; Green, A.; Santini, S.; Bizzarri, A.R.; Cannistraro, S.; Gupta, T.K.D.K.D.; Beattie, C.W.W. p28, A first in class peptide inhibitor of cop1 binding to p53. Br. J. Cancer 2013, 108, 2495–2504. [CrossRef] 31. Coppari, E.; Santini, S.; Bizzarri, A.R.; Cannistraro, S. Kinetics and binding geometries of the complex between β2-microglobulin and its antibody: An AFM and SPR study. Biophys. Chem. 2016, 211, 19–27. [CrossRef][PubMed] Sensors 2020, 20, 7153 17 of 19

32. Signorelli, S.; Cannistraro, S.; Bizzarri, A.R. Structural characterization of the intrinsically disordered protein p53 using raman spectroscopy. Appl. Spectrosc. 2016, 71, 823–832. [CrossRef][PubMed] 33. Yamada, T.; Goto, M.; Punj, V.; Zaborina, O.; Chen, M.L.; Kimbara, K.; Majumdar, D.; Cunningham, E.; Das Gupta, T.K.; Chakrabarty, A.M. Bacterial redox protein azurin, tumor suppressor protein p53, and regression of cancer. Proc. Natl. Acad. Sci. USA 2002, 99, 14098–14103. [CrossRef][PubMed] 34. Signorelli, S.; Cannistraro, S.; Bizzarri, A.R. Raman evidence of p53-DBD disorder decrease upon interaction with the Anticancer Protein Azurin. Int. J. Mol. Sci. 2019, 20, 3078. [CrossRef] 35. Bizzarri, A.R.; Cannistraro, S. SERS detection of thrombin by protein recognition using functionalized gold nanoparticles. Nanomedicine 2007, 3, 306–310. [CrossRef] 36. Bizzarri, A.R.; Cannistraro, S. Surface-enhanced resonance Raman spectroscopy signals from single myoglobin molecules. Appl. Spectrosc. 2002, 56, 1531–1537. [CrossRef] 37. Bizzarri, A.R.; Cannistraro, S. Surface-enhanced Raman spectroscopy combined with atomic force microscopy for ultrasensitive detection of thrombin. Anal. Biochem. 2009, 393, 149–154. [CrossRef] 38. Domenici, F.; Bizzarri, A.R.; Cannistraro, S. SERS-based nanobiosensing for ultrasensitive detection of the p53 tumor suppressor. Int. J. Nanomed. 2011, 6, 2033–2042. 39. Domenici, F.; Bizzarri, A.R.; Cannistraro, S. Surface-enhanced Raman scattering detection of wild-type and mutant p53 proteins at very low concentration in human serum. Anal. Biochem. 2012, 421, 9–15. [CrossRef] 40. Bizzarri, A.R.; Moscetti, I.; Cannistraro, S. Surface enhanced Raman spectroscopy based immunosensor for ultrasensitive and selective detection of wild type p53 and mutant p53R175H. Anal. Chim. Acta 2018, 1029, 86–96. [CrossRef] 41. Kneipp, K.; Kneipp, H.; Itzkan, I.; Dasari, R.R.; Feld, M.S. Ultrasensitive chemical analysis by Raman spectroscopy. Chem. Rev. 1999, 99, 2957–2976. [CrossRef][PubMed] 42. Lombardi, J.R.; Birke, R.L. A unified view of surface-enhanced Raman scattering. Acc. Chem. Res. 2009, 42, 734–742. [CrossRef][PubMed] 43. Cialla, D.; März, A.; Böhme, R.; Theil, F.; Weber, K.; Schmitt, M.; Popp, J. Surface-enhanced Raman spectroscopy (SERS): Progress and trends. Anal. Bioanal. Chem. 2012, 403, 27–54. [CrossRef][PubMed] 44. Moskovits, M. Surface-enhanced spectroscopy. Rev. Mod. Phys. 1985, 57, 783–826. [CrossRef] 45. Campion, A.; Kambhampati, P.; Weaver, M.J.; Natan, M.J.; Shalaev, V.M.; Suh, J.S.; Botet, R. Surface-enhanced Raman scattering. Chem. Soc. Rev. 1998, 27, 241–250. [CrossRef] 46. Kneipp, K.; Wang, Y.; Kneipp, H.; Perelman, L.T.; Itzkan, I. Single molecule detection using surface-enhanced Raman scattering (SERS). Phys. Rev. Lett. 1997, 78, 1667–1670. [CrossRef] 47. Bizzarri, A.R.; Cannistraro, S. Lévy statistics of vibrational mode fluctuations of single molecules from surface-enhanced Raman scattering. Phys. Rev. Lett. 2005, 94, 068303. [CrossRef] 48. Bell, S.; Hansen, S.; Buchner, J. Refolding and structural characterization of the human p53 tumor suppressor protein. Biophys. Chem. 2002, 96, 243–257. [CrossRef] 49. Dawson, R.; Müller, L.; Dehner, A.; Klein, C.; Kessler, H.; Buchner, J. The N-terminal domain of p53 is natively unfolded. J. Mol. Biol. 2003, 332, 1131–1141. [CrossRef] 50. Cho, Y.; Gorina, S.; Jeffrey, P.D.; Pavletich, N.P. Crystal structure of a p53 tumor suppressor-DNA complex: Understanding tumorigenic mutations. Science 1994, 265, 346–355. [CrossRef] 51. Link, T.; Iwakuma, T. Roles of p53 in extrinsic factor-induced liver carcinogenesis. Hepatoma Res. 2017, 3, 95–104. [CrossRef][PubMed] 52. Friedler, A.; Hansson, L.O.; Veprintsev, D.B.; Freund, S.M.V.; Rippin, T.M.; Nikolova, P.V.; Proctor, M.R.; Rüdiger, S.; Fersht, A.R.; Rudiger, S.; et al. A peptide that binds and stabilizes p53 core domain: Chaperone strategy for rescue of oncogenic mutants. Proc. Natl. Acad. Sci. USA 2002, 99, 937–942. [CrossRef][PubMed] 53. Villar, S.; Ortiz-Cuaran, S.; Abedi-Ardekani, B.; Gouas, D.; Nogueira da Costa, A.; Plymoth, A.; Khuhaprema, T.; Kalalak, A.; Sangrajrang, S.; Friesen, M.D.; et al. Aflatoxin-induced TP53 R249S mutation in hepatocellular carcinoma in Thailand: Association with tumors developing in the absence of liver cirrhosis. PLoS ONE 2012, 7, e37707. [CrossRef][PubMed] 54. Nar, H.; Messerschmidt, A.; Huber, R.; van de Kamp, M.; Canters, G.W. Crystal structure analysis of oxidized Pseudomonas aeruginosa azurin at pH 5 5 and pH 9 0: A pH-induced conformational transition involves a · · peptide bond flip. J. Mol. Biol. 1991, 221, 765–772. [CrossRef] 55. Guex, N.; Peitsch, M.C. SWISS-MODEL and the Swiss-PdbViewer: An environment for comparative protein modeling. Electrophoresis 1997, 18, 2714–2723. [CrossRef] Sensors 2020, 20, 7153 18 of 19

56. Milner, J.; Medcalf, E.A. Temperature-dependent switching between “wild-type” and “mutant” forms of p53-Val135. J. Mol. Biol. 1990, 216, 481–484. [CrossRef] 57. Bullock, A.N.; Henckel, J.; DeDecker, B.S.; Johnson, C.M.; Nikolova, P.V.; Proctor, M.R.; Lane, D.P.; Fersht, A.R. Thermodynamic stability of wild-type and mutant p53 core domain. Proc. Natl. Acad. Sci. USA 1997, 94, 14338–14342. [CrossRef] 58. Koulgi, S.; Achalere, A.; Sonavane, U.; Joshi, R. Investigating dna binding and conformational variation in temperature sensitive p53 cancer mutants using QM-MM simulations. PLoS ONE 2015, 10, e0143065. [CrossRef] 59. Stindt, M.H.; Muller, P.A.J.; Ludwig, R.L.; Kehrloesser, S.; Dötsch, V.; Vousden, K.H. Functional interplay between MDM2, p63/p73 and mutant p53. Oncogene 2015, 34, 4300–4310. [CrossRef] 60. Beedle, A.E.M.; Lezamiz, A.; Stirnemann, G.; Garcia-Manyes, S. The mechanochemistry of copper reports on the directionality of unfolding in model cupredoxin proteins. Nat. Commun. 2015, 6, 7894. [CrossRef] 61. Yadav, A.; Paul, S.; Venkatramani, R.; Ainavarapu, S.R.K. Differences in the mechanical unfolding pathways of apo- and copper-bound azurins. Sci. Rep. 2018, 8, 1989. [CrossRef][PubMed] 62. Winkler, J.R.; Gray, H.B. Long-Range Electron Tunneling. J. Am. Chem. Soc. 2014, 136, 2930–2939. [CrossRef] [PubMed] 63. Bostick, C.D.; Mukhopadhyay, S.; Sheves, M.; Cahen, D.; Lederman, D. Protein bioelectronics: A review of what we do and do not know. Rep. Prog. Phys. 2017, 81, 026601. [CrossRef][PubMed] 64. Baldacchini, C.; Kumar, V.; Bizzarri, A.R.; Cannistraro, S. Electron tunnelling through single azurin molecules can be on/off switched by voltage pulses. Appl. Phys. Lett. 2015, 106, 183701. [CrossRef] 65. Taranta, M.; Bizzarri, A.R.; Cannistraro, S. Probing the interaction between p53 and the bacterial protein azurin by single molecule force spectroscopy. J. Mol. Recognit. 2008, 21, 63–70. [CrossRef] 66. De Grandis, V.; Bizzarri, A.R.; Cannistraro, S. Docking study and free energy simulation of the complex between p53 DNA-binding domain and azurin. J. Mol. Recognit. 2007, 20, 215–226. [CrossRef] 67. Punj, V.; Das Gupta, T.K.; Chakrabarty, A.M. Bacterial cupredoxin azurin and its interactions with the tumor suppressor protein p53. Biochem. Biophys. Res. Commun. 2003, 312, 109–114. [CrossRef] 68. Jiao, L.S.; Niu, L.; Shen, J.; You, T.; Dong, S.; Ivaska, A. Simple azo derivatization on 4-aminothiophenol/Au monolayer. Electrochem. Commun. 2005, 7, 219–222. [CrossRef] 69. Tian, J.H.; Liu, B.; Li, X.; Yang, Z.L.; Ren, B.; Wu, S.T.; Tao, N.; Tian, Z.Q. Study of molecular junctions with a combined surface-enhanced raman and mechanically controllable break junction method. J. Am. Chem. Soc. 2006, 128, 14748–14749. [CrossRef] 70. Phillips, J.H.; Robrish, S.A.; Bates, C. High efficiency coupling of diazonium ions to proteins and amino acids. J. Biol. Chem. 1965, 240, 699–704. 71. Sotnikov, D.; Zherdev, A.; Dzantiev, B. Development and application of a label-free fluorescence method for determining the composition of gold nanoparticle–protein conjugates. Int. J. Mol. Sci. 2014, 16, 907–923. [CrossRef][PubMed] 72. Owens, P.; Phillipson, N.; Perumal, J.; O’Connor, G.M.; Olivo, M. Sensing of p53 and EGFR biomarkers using high efficiency SERS substrates. Biosensors 2015, 5, 664–677. [CrossRef][PubMed] 73. Wu, L.; Wang, Z.; Zong, S.; Chen, H.; Wang, C.; Xu, S.; Cui, Y. Simultaneous evaluation of p53 and p21 expression level for early cancer diagnosis using SERS technique. Analyst 2013, 138, 3450–3456. [CrossRef] [PubMed] 74. Zhou, W.; Jimmy Huang, P.-J.; Ding, J.; Liu, J. Aptamer-based biosensors for biomedical diagnostics. Analyst 2014, 139, 2627–2640. [CrossRef][PubMed] 75. Chen, L.; Rashid, F.; Shah, A.; Awan, H.M.; Wu, M.; Liu, A.; Wang, J.; Zhu, T.; Luo, Z.; Shan, G. The isolation of an RNA aptamer targeting to p53 protein with single amino acid mutation. Proc. Natl. Acad. Sci. USA 2015, 112, 10002–10007. [CrossRef][PubMed] 76. Bizzarri, A.R.; Cannistraro, S. Investigation of a direct interaction between miR4749 and the tumor suppressor p53 by fluorescence, FRET and molecular modeling. Biomolecules 2020, 10, 346. [CrossRef] 77. Gruber, A.R.; Lorenz, R.; Bernhart, S.H.; Neuböck, R.; Hofacker, I.L. The Vienna RNA websuite. Nucleic Acids Res. 2008, 36, W70–W74. [CrossRef] 78. Boniecki, M.J.; Lach, G.; Dawson, W.K.; Tomala, K.; Lukasz, P.; Soltysinski, T.; Rother, K.M.; Bujnicki, J.M. SimRNA: A coarse-grained method for RNA folding simulations and 3D structure prediction. Nucleic Acids Res. 2015, 44, e63. [CrossRef] Sensors 2020, 20, 7153 19 of 19

79. Yan, Y.; Zhang, D.; Zhou, P.; Li, B.; Huang, S.-Y. HDOCK: A web server for protein-protein and protein-DNA/RNA docking based on a hybrid strategy. Nucleic Acids Res. 2017, 45, W365–W373. [CrossRef] 80. Yang, J.; Yan, R.; Roy, A.; Xu, D.; Poisson, J.; Zhang, Y. The I-TASSER suite: Protein structure and function prediction. Nat. Methods 2014, 12, 7–8. [CrossRef]

Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).