Multibranched gold nanoparticles coated with serum proteins fit for photothermal tumor ablation

Photothermal tumor ablation might be carried out with multibranched gold nanoparticles (MBAuNPs) having maximum absorbance (Amax) in the infrared region and functionalized with ligands that would bind them to the target tumor markers. However, in nanomedicine applications, the nanostructures must reach their target tissues to be effective, but the corona of serum proteins they instantaneously acquire when administered by intravenous injection may affect their activity; for this reason, we decided to analyze the effect that exposing MBAuNPs to bovine serum albumin (BSA) and human serum (HS) have on their protein corona and physical properties. The synthesized spherical Au seeds stoichiometrically generate piñata-like MBAuNPs of 8–20 peaks potentially useful for photothermal tumor ablation since they induce hyperthermia of more than 4 ○C in phantom gels mimicking the skin irradiated with an 808 nm laser at 0.75 W/cm2. The calculated surface area of MBAuNPs ranges from 24 984 nm2 to 40 669 nm2, depending on the number of peaks we use for modeling the NPs. When MBAuNPs are exposed to BSA, they acquire a protein corona with an internal “hard” portion composed by one or two layers of BSA containing ∼1000–4000 molecules covalently bound to their surface, and an external “soft” portion formed by agglomerated BSA molecules linked by non-covalent bonds. Functionalization with BSA decreases the tendency of MBAuNPs to agglomerate and increases their size dispersion. MBAuNPs and MBAuNPs–BSA exposed to HS bind HS albumin and other HS proteins ranging from 25 kDa to 180 kDa that increase their hydrodynamic diameter and decrease their stability. We conclude that MBAuNPs exposed to serum albumin and HS instantaneously acquire a hard and soft protein corona that may affect prior or subsequent functionalization aiming to direct them to specific cell or tissue targets. © 2020 Author(s). All article content, except where otherwise noted, is licensed under a Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/). https://doi.org/10.1063/5.0025368 AIP Advances 10, 125030 (2020); doi: 10.1063/5.0025368 10, 125030-1


INTRODUCTION
Malignant tumors are usually treated by surgical resection, chemotherapy, and radiotherapy. 1 Chemotherapy and radiotherapy are often ineffective and have undesirable side effects because they do not discriminate between healthy and cancerous cells. 2 Although hyperthermia for tumor ablation is an alternative to surgical resection, 3 it is rarely used because the temperature in the target tissue is not homogeneous. 4 Since radiation in the near-infrared region (NIR) penetrates deeply but is hardly absorbed by tissues, [5][6][7] attempts have been made to develop tumor ablation procedures based on nanoparticles (NPs) that convert NIR radiation into heat. 2 Gold NPs (AuNPs) are non-toxic, 8 and since spherical AuNPs absorb NIR waves much less efficiently, nowadays the use of multibranched AuNPs (MBAuNPs) 9 for hyperthermal phototherapy is being intensively explored. 10 Depending on the tumor and its anatomical location, the potential efficacy of photothermal nanotherapy would depend on both the photothermal conversion efficiency 2 and the ability of NPs to reach and penetrate the desired tumor targets. 11 NPs are functionalized with ligands that direct them to specific cell receptors, 12 but when they enter the bloodstream they are instantly modified through their interaction with blood proteins attaching to their surface and forming a structure known as protein corona (PC). 11,13 The PC is formed by the hard corona (HC) consisting of proteins strongly associated with the NP surface, and by the soft corona (SC) that is an outer layer of weakly bound proteins. 14 The PC is also complex and unique for each nanomaterial and changes the physicochemical properties of NPs, such as size, surface charge, and state of aggregation. 15 These changes may interfere with the functionalities provided in biological microenvironments. 16,17 The development of NPs suitable for in vivo photothermal therapy thus depends on controlling PC formation and structure and its effects on the physical and biological properties of NPs. 18,19 In this report, we describe the synthesis of MBAuNPs, their characterization, and the features of the hard and soft protein corona they acquire after being exposed to bovine serum albumin (BSA), to human serum (HS), or BSA followed by HS. The photothermal tumor ablation potential of our MBAuNPs is supported by their high photothermal conversion efficiency leading to a net temperature increase of ∼5 ○ C in phantom gels irradiated at low power (<1 W/cm 2 ) with an 808 nm near-infrared laser.

Auric seeds
Scanning electron microscope (SEM) images showed that seeds were spherical, with a hydrodynamic diameter (D H ) of 18 nm [Figs. 1(a) and 1(b) and Table I], and a surface area of 1020 nm 2 , and maximum absorbance at 520 nm [ Fig. 1(c)].
The ζ-potential of the seeds was −36.8 mV, with a polydispersity index (PDI) of 0.602 (Table I). The concentration determined by inductively coupled plasma atomic emission spectroscopy (ICP-OES) was 60 mg/l, corresponding to 1 × 10 15 seeds/l. The ζ-potential of MBAuNPs was −24.4 mV, with a PDI of 0.455 (Table I). The MBAuNP concentration determined by ICP-OES was 23.5 mg/l, corresponding to 1 × 10 13 MBAuNPs/l.

Hyperthermia induced by irradiation of auric seeds and MBAuNPs in aqueous and phantom gel suspensions
The temperature was monitored in aqueous solutions containing auric seeds (60 mg/l), or MBAuNPs (23.5 mg/l), and in phantom gels containing MBAuNPs (2 mg/l or 3 mg/l). All samples were irradiated continuously for 10 min with an 808 nm laser at irradiation powers of 1 W/cm 2 , 0.75 W/cm 2 , and 0.45 W/cm 2 .

Surface area and BSA corona in MBAuNPs
TEM and high-resolution transmission electron microscope (HRTEM) images of functionalized MBAuNPs showed a 10 nm-15 nm thick layer covering their surface that was absent in pure MBAuNPs (Fig. 4). D H values increased from 125 nm for MBAuNPs to 146 nm for MBAuNPs-BSA. The ζ-potential of MBAuNPs was −72.7 mV, with a PDI of 0.929 (Table I).
The UV-VIS spectra of pure BSA in solution and MBAuNP suspensions functionalized with BSA were determined using

ARTICLE
scitation.org/journal/adv phosphate-buffered saline (PBS) as a vehicle. The absorbance peaks observed were as follows: pure BSA, a single peak at 280 nm; pure MBAuNPs, a single peak at 775 nm; MBAuNPs-BSA, two peaks, a major one at 280 nm and a minor one at 825 nm (Fig. 5).
The average number of BSA molecules in a suspension containing 7.6 × 10 11 MBAuNPs/ml was calculated by UV spectroscopy. The BSA concentration in MBAuNP-BSA sedimented by centrifugation and resuspended in PBS was 4.5 × 10 16 molecules/ml, corresponding to 59 210 BSA molecules per MBAuNP. MBAuNPs-BSA suspensions washed six times with PBS contained 1.14 × 10 15 BSA molecules/ml, corresponding to 1500 BSA molecules per MBAuNP (2.5% of the total BSA molecules initially bound).
Through mathematical estimation of the surface area of MBAuNPs and taking into account the area of individual BSA molecules, we calculated 3.5 × 10 12 molecules/cm 2 , assuming that the hard protein corona formed by a molecular monolayer would amount to ∼900, ∼1000, and ∼1500 BSA molecules per MBAuNP with 8, 12, and 20 tips, respectively. Assuming a protein bilayer, the number of BSA molecules per MBAuNP would be ∼1900, ∼2500, and ∼3880 (Table II).
XPS spectra of MBAuNPs-BSA. The high-resolution x-ray photoelectron spectroscopy (XPS) spectrum of gold in pure MBAuNPs had three peaks (including their doublets) at 82.37 eV, 82.87 eV, and 85.06 eV, corresponding to Au 0 , Au 1+ , and Au 3+ , respectively. In contrast, the high-resolution spectrum of gold in MBAuNPs-BSA had two peaks: the first at 81.31 eV identified as Au 0 and the second at 82.11 eV identified as a C-S-Au (I) bond [Figs. 7(a) and 7(b) and Table III]. The high-resolution spectrum of sulfur had the main peak at 166.21 eV, corresponding to the S 2p 3/2 position identified as an S-Au bond [ Table III 9. SDS-polyacrylamide gel electrophoresis of proteins bound to MBAuNPs and MBAuNPs-BSA exposed to human serum. Lane 1, protein molecular markers. Lane 2, BSA in solution (10 g/l). Lane 3, pellet from the third wash of MBAuNPs-BSA. Lane 4, human serum. Lane 5, pellet from MBAuNPs exposed to human serum. Lane 6, pellet of MBAuNPs-BSA exposed to human serum.  After having been exposed to human serum, MBAuNPs and MBAuNPs-BSA that were extensively washed by centrifugation and then subjected to denaturing conditions released protein bands with apparent molecular weights of 180 kDa, 156 kDa, 120 kDa, 84 kDa, 75 kDa, and 25 kDa (Fig. 9).

DISCUSSION
The MBAuNPs described in this paper are potentially useful for photothermal ablation of tumors since they induce hyperthermia by irradiation with a near-infrared laser at minimum power both in aqueous suspensions and in phantom gels mimicking the human skin.
The auric seeds, synthesized as precursors of MBAuNPs, are spherical and stable as expected, 20 they present maximum absorbance at 520 nm and have excellent stability and dispersion as shown by dynamic light scattering (DLS) analysis. Their concentration was determined by the ICP-MS method of Allabashi et al., 21 used also to control the amount of seeds used for the synthesis of anisotropic MBAuNPs, that was carried out with a modification of the method of Maiorano et al. 22 The resulting MBAuNPs had maximum absorbance at 775 nm, indicating an increased efficiency to produce hyperthermia by NIR irradiation, whereas the loss of the 520 nm absorbance peak suggests that each seed gives rise to a single MBAuNP. The larger shift of the absorption peak toward the infrared by our MBAuNPs may be due to the relatively high HEPES concentration (50 mM) used for their synthesis, leading to an increased length of the cones attached to the seed surface.
The highest photothermal efficiency was attained with suspensions containing 3 mg/l of MBAuNPs in phantom gels irradiated with the 808 nm laser at 0.75 W/cm 2 , whose temperature was increased by 4.8 ○ C, well above the results obtained in previous studies. [23][24][25][26] Our finding that the hyperthermia of phantom gels depends on both the MBAuNP concentration and irradiation power may be further tested in cell cultures and animal models.
Functionalization with specific ligands to direct MBAuNPs to diseased tissues is expected to improve their therapeutic efficacy when administered systemically through intravenous injection. However, such functionalization may be affected and even nullified by the protein corona that instantaneously forms on their surface, even in MBAuNPs previously functionalized with BSA to prevent their agglomeration and to direct them to the albumin receptors that abound on the surface of the endothelial cells forming the inner wall of blood vessels. 27 HRTEM images of extensively washed MBAuNPs-BSA have a 10 nm-15 nm thick halo absent from naked MBAuNPs and resolved individual BSA molecules appearing to form a tapestry of one or two molecular layers. The thickness of the protein layer is consistent with the 8 × 8 × 3.5 nm 3 dimensions of the native BSA molecule. 28 The 50 nm displacement of the absorbance peak from 775 nm in MBAuNPs to 825 nm in MBAuNPs-BSA is higher than the 7 nm displacement reported by Nghiem et al. 29 This difference may be due to a difference in the number of BSA molecules bound per nanoparticle since the magnitude of the displacement depends on the initial BSA concentration used for functionalization.
Our mathematical estimation of the surface area of MBAuNPs and the number of BSA molecules bound to their surface matches the number of BSA molecules determined by UV spectrophotometry. Taking into account the HRTEM images, the increase of the hydrodynamic diameter, the mathematical estimation of the surface area, and the experimental quantification results, it appears that the PC of MBAuNPs-BSA has two parts: an internal film of one or two layers of BSA molecules covalently bound to the MBAuNP surface, and a lax external layer formed by agglomerated BSA molecules linked by non-covalent bonds that are released by extensive washing. This model is consistent with that of Kokkinopoulou et al. 30 who proposed that the protein corona consists not only of multiple layers covering each NP but also a three-dimensional network of proteins linked to the NP surface.
Brewer et al. 31 estimated that the number of BSA molecules bound to spherical AuNPs ranges from 2.0 × 10 12 cm −2 to 3.3 × 10 12 cm −2 depending on the orientation in which BSA is bound. Assuming that the area of each BSA molecule to be 28 nm 2 and applying it to auric seeds and MBAuNPs, we obtained a similar range: 3.5 × 10 12 BSA molecules per cm 2 .
To calculate the surface area of MBAuNPs, we used Eq. (3) to estimate the area that would bind the hard corona and a slightly modified Eq. (4) for the area available for the soft corona. Our results were similar to the values of the area for MBAuNPs calculated by Tsoulos et al., 32 and for the number of molecules attached to a given Au surface published by Brewer et al. 31 NPs with ζ-potential values ≥+30 or ≤−30 are considered stable. 33 Our spheric seeds (ζ-potential = −36.8 mV) are quite stable and do not agglomerate for months, whereas our MBAuNPs (ζpotential = −24.4 mV) tend to agglomerate. Since our MBAuNPs-BSA have a ζ-potential = −72.7 mV, we confirmed that BSA decreases their tendency to agglomerate 34 and turn them into a stable system.
High-resolution XPS analyses allowed us to unambiguously determine the presence of C-S-Au bonds between BSA molecules and gold atoms on the MBAuNP surface; 35 they are probably Au-S cystine disulfide coordination bonds as those described by Wang et al. 36 Each BSA molecule contains 34 cysteine residues and a free thiol at Cys34 that most likely is involved in the covalent binding of BSA to the gold surface, making MBAuNPs-BSA stable.
Assessment of the secondary BSA structure in solution and that bound to MBAuNPs revealed a decrease in the proportion of α-helices on the functionalized protein. Disulfide bonds are the main form of sulfur in BSA that stabilizes α-helices to maintain the protein structure. XPS analysis indicates that BSA binding to the MBAuNP surface through S-Au bonds may decrease the proportion of α-helices. Similar changes in the secondary protein structure have already been reported, 36,37 and it is not known if these changes affect the affinity of BSA to the receptor-binding domain located on the surface of endothelial cells and some cancer cells. 27 We also found that exposure of MBAuNPs and MBAuNPs-BSA to human serum increased their ζ-potential values close to those of uncoated MBAuNPs, and also increased D H values of 23 nm for MBAuNPs and 39 nm for MBAuNPs-BSA. These effects may be explained by the presence of human serum molecules other ARTICLE scitation.org/journal/adv than BSA or HSA and led us to characterize the electrophoretic patterns of the proteins bound to MBAuNPs and MBAuNPs-BSA exposed to human serum. HSA and other unidentified proteins with apparent molecular weights of 180 kDa, 156 kDa, 120 kDa, 84 kDa, 75 kDa, and 25 kDa bound both to MBAuNPs and MBAuNPs-BSA were associated with the higher increase of ζ-potential and D H in MBAuNPs-BSA, suggesting that previous functionalization with BSA affects the binding of human serum proteins. This study reinforces the potential of MBAuNPs for photothermal tumor ablation. Their use in medical application, however, depends greatly on the successful delivery to their biological targets through the fine control and tuning of the protein corona, a process that requires research by oncologist/molecular biologist/physicist teams.

Synthesis of auric seeds
Synthesis of spherical seeds was performed using the Turkevich method in mixtures containing 25 mM HAuCl 4 and 0.75 mM sodium citrate, pH 7.4, as a reducing agent. 20,38,39 Briefly, 19.8 ml of 0.75 mM sodium citrate pH 7.4 was added to a 30 ml beaker with a magnetic bar, and the mix was stirred on a magnetic hot plate at 70 ○ C; 0.2 ml of 25 mM HAuCl 4 was immediately added, and the mixture was kept under the same conditions for 10 min until the color was changed from light gray to red, with the appearance of red wine. When the reaction ended, the UV-VIS spectrum was recorded in a Cary model 60 Agilent spectrophotometer. To determine the shape and size of NPs, the samples were also analyzed by dynamic light scattering (DLS) with a zetameter (Zetatrac, Microtrac) and with a Dual Beam FEI-Helios Nanolab 600 scanning electron microscope (SEM) from which images were obtained at 5 kV, 88 pA, 4 mm working distance, and 300 000× magnification.

MBAuNP synthesis
MBAuNPs were synthesized in mixtures containing 0.8 mM HAuCl 4 , 1 × 10 15 auric seeds per l (60 mg/l of Au), 100 mM hydroxylamine (NH 2 OH), and 50 mM Hepes pH 7.4. The spectrophotometer probe (A 300 -A 1000 , average reading speed) was kept immersed in a 30 ml beaker containing a magnetic bar and placed in a container with ice to keep the mixture at 4 ○ C, and the absorbance measurements were recorded at the start and every 2 min while the reaction occurred; 12.5 ml of 50 mM Hepes pH 7.4 l, 150 l of seed suspension, 400 l of 100 mM hydroxylamine, and 2.2 ml of 0.8 mM HAuCl 4 (one drop every 8 s) were successively added.

Gold content in auric seeds and MBAuNPs
To determine the amount of gold present in auric seeds and MBAuNPs, an aliquot from each sample was digested with HCl and HNO 3 in a water bath for 1 h. Inductively coupled plasma atomic emission spectroscopy (ICP-OES) of samples was performed in a Varian, model 730-ES instrument.

Concentration of auric seeds
The Au concentration obtained by ICP-OES was used to determine the number of auric seeds per liter, 21 dividing the total number of gold atoms in the sample by the number of atoms per NP, with the following formulas: where Nat = Average number of Au atoms per nanoparticle, NA (Avogadro's number) = 6.022 × 10 23 atoms/mol, ρ (fcc density of the seeds) = 1.93 × 10 −20 g/nm 3 , D (diameter of the nanoparticles) = 18 nm, M (gold atomic weight) = 197 g/mol, and At (total number of atoms in the sample) = 1.834 11 × 10 20 atoms/l, obtained from the ICP-OES analysis.

Mathematical approach to estimate the surface area of MBAuNPs
To determine the surface area of MBAuNPs, their shape was modeled as a "piñata," i.e., a structure formed by a spherical core to whose surface are attached the bases of cones with blunt tips. The total surface area of each MBAuNP would be given by the following contributions: (i) the total area of the spherical core, minus (ii) the area of the sphere caps that occupy the N bases of the cones, plus (iii) the area of N cones, plus (iv) N times the area of the half-sphere that forms the tip of the cones. 32 The terms appear in the same order in the following equation for the total area: where core average radius: a = 34 nm, average tip radius: ri = 4.67 nm, average tip length: hi = 29.16 nm, and average radius at the base of the tip: Ri = 15.1 nm. The values assigned to these variables originate from the statistics of the SEM micrographs of our MBAuNPs. Figure 10 illustrates "piñata-like" NPs with 8 and 20 tips, where N is the number of peaks per NP. Considering that MBAuNPs have different numbers of tips, a first geometric approximation was used with N values of 8, 12, and 20. 40 Hyperthermia assays with auric seeds or MBAuNPs in aqueous suspensions and phantom gels in phantom gels, both of which were irradiated for 10 min with an 808 nm laser (Lasermate model IML-808) at different powers (0.45 W/cm 2 , 0.75 W/cm 2 , and 1 W/cm 2 ). The temperature of each sample was measured with an infrared thermal camera (Flexcam R2 model IR) from time zero until the end of the irradiation period.
To prepare phantom gels resembling human tissues, 41 we prepared a suspension containing MBAuNPs (2 mg/l or 3 mg/l of Au), 100 ml of ultrapure water, 1.5 g of agar, and 2.5 g of TX-151, all added to a vacuum flask. The mixture was heated in a microwave oven for 15 s periods until complete agar and TX-151 dissolution was reached. To the mixture, kept under stirring at 50 ○ C, 5 ml of an MBAuNP suspension (containing 2 g or 3 g of gold, equivalent to 1.3 × 10 10 or 1.9 × 10 10 MBAuNPs per ml, respectively), 2 g of polyethylene powder was added. After polyethylene dissolution, air bubbles were eliminated by aspiration with a vacuum pump (Welch, Model 2522). The mixtures were poured onto 35 × 10 mm 2 Petri dishes and left to solidify overnight at room temperature.

Functionalization of MBAuNPs with BSA
For functionalization, MBAuNPs (7.6 × 10 11 MBAuNP/ml) were washed five times by centrifugation with 0.1% Nonidet P40 aqueous solution. For the last wash, phosphate-buffered saline (PBS: 137 mM NaCl, 27 mM KCl, 10 mM Na 2 HPO 4, 18 mM KH 2 PO 4 pH 7.4) was used, and the UV-VIS spectrum was recorded before and after functionalization. After the sixth wash, the supernatant was removed by aspiration, and the pellet was sonicated in an ultrasonic bath (Branson, Model 2800) for 4 min, and vortexed for 15 s; 1 ml of BSA (10 g/l) was added, and the mixture was vortexed immediately to prevent MBAuNP agglomeration. The MBAuNPs-BSA suspension was kept at 4 ○ C for 24 h and then washed four times by centrifugation with PBS to eliminate the BSA in solution.

BSA quantitation
Calculation using UV-VIS spectra Since the maximum absorption peak of BSA is at 280 nm (A 280 ), the BSA concentration in solution and MBAuNPs was determined by measuring A 280 . The molar extinction coefficients of BSA in solution and MBAuNP-BSA suspensions were calculated by comparing the slopes of standard curves of BSA in solution and MBAuNP-BSA suspensions, using a 0.2 cm optical path quartz cell. To determine the average number of BSA molecules per MBAuNP, seven consecutive washes by centrifugation with PBS were carried out after 24 h of exposure to BSA, by dividing the number of MBAuNPs by the number of BSA molecules (estimated from A 280 and the molar extinction coefficient).

Mathematical estimation of the number of BSA molecules per MBAuNP
The mathematical approximation to calculate the number of BSA molecules per MBAuNP was performed by dividing the total available surface area of the MBAuNP (calculated as explained above) by the area of native BSA molecules. Each native BSA molecule has a tridimensional structure with an area of 28 nm 2 (Fig. 11). 28,42 Assuming a bimolecular BSA layer, the total area of the protein corona on the surface of each MBAuNP-BSA would be For a uniform 4-nm-thick protein hard corona, the following values were used: average tip radius: ri = 8.67 nm, average tip length: hi = 33.16 nm, and average radius of the tip base: Ri = 18.1 nm. The new total area available will then be divided by the area of native BSA molecules.

Scanning electron microscopy (SEM)
To a 1.5 ml microcentrifuge tube containing 500 l of ethanol, 1 ml of the MBAuNP reaction mixture was added. The tube was immersed in an ultrasonic bath for 5 min, vortexed for 20 s, and centrifuged for 10 min at 13 000 rpm (15 304 × g) and 4 ○ C. The supernatant was removed by careful aspiration, and 1 ml of ultrapure water was added to the pellet that was further washed by centrifugation six more times. On an aluminum sample holder (pin), 75 l-samples were deposited and left to dry overnight before observing them in a scanning electron microscope (Dual Beam FEI-Helios Nanolab 600) at 150 000× magnification, 5 kV, 88 pA, and 4 mm working distance.

Transmission electron microscopy (TEM, HRTEM)
Pure and functionalized MBAuNP samples were stained with uranyl acetate in Petri dishes covered with Parafilm. On a grid (Lacey Carbon), a drop of a concentrated sample was applied and allowed to sediment for 2 h at room temperature. After adding uranyl acetate (10 g/l), the samples were left to dry for 15 min and then were observed with a Jeol 200 CX 100 keV transmission electron microscope (TEM) and a Jeol ARM200F high-resolution transmission electron microscope (HRTEM).

X-ray photoelectron spectroscopy (XPS)
One drop of either MBAuNPs or MBAuNPs-BSA suspension was placed on 1 × 1 cm 2 silicon plates. The plates were left to dry at room temperature for 24 h and analyzed in an ultra-high vacuum with an XPS PHI-5000 spectrometer (Physical Electronics, VersaProbe II Model), and the results were processed with the XPS Multipak software.

Human serum
Human blood obtained with informed consent from healthy donors following the Declaration of Helsinki was left to clot spontaneously. Clotted blood was centrifuged for 10 min at 13 000 rpm, and the supernatant serum was transferred to 15 ml plastic tubes. Pooled serum from four volunteers was stored at −80 ○ C until it was used.

Dynamic light scattering (DLS)
The ζ-potential and size of MBAuNPs and MBAuNPs-BSA were determined by DLS at 22 ○ C and 90 ○ dispersion angles with a zetameter (Zetatrac, Microtrac). 0.1% Nonidet P40, PBS, BSA dissolved in PBS, and human serum were used as vehicles to prepare MBAuNPs and MBAuNPs-BSA suspensions.

Sodium dodecyl sulfate-polyacrylamide gel electrophoresis
The interaction of MBAuNPs with BSA and with human serum was assessed using sodium dodecyl sulfate (SDS)-polyacrylamide gel electrophoresis (SDS-PAGE) of suspension samples washed three times by centrifugation after a 24 h exposure at 37 ○ C. 10-l from each sample was mixed with 10 l of Laemmli buffer (10% 2-mercaptoethanol, 0.004% Bromphenol blue, 20% glycerol, 4% sodium dodecyl sulfate, 125 mM Tris-HCl, pH 6.8), and the mixtures were boiled for 5 min at 100 ○ C. 20-l volumes were applied to 12% polyacrylamide slab gels, run for 2 h at 120 V, stained with Coomassie Blue for 30 min, and destained overnight in a 5:4:1 methanol-water-glacial acetic acid mixture.

Circular dichroism (CD)
To determine the proportion of α-helices from BSA both in solution and bound to MBAuNPs, circular dichroism (CD) measurements were performed at 4 ○ C in a 0.2 cm quartz cell with a 205 nm-330 nm reading range in a MOS-500 spectropolarimeter (BioLogic Science Instruments). In mixtures with a 7.6 × 10 11 MBAuNPs/ml fixed concentration dissolved in PBS, the BSA concentrations tested were 1.0 mg/ml, 0.5 mg/ml, and 0.25 mg/ml. The proportion of α-helices from BSA in solution or bound to MBAuNPs was calculated with the following formulas: 43,44 α − Helix(%) = (MRE 222 − 400033 000 − 4000) × 100, (6) where MRE = mean residual ellipticity, Observed DC (mdeg) = ellipticity obtained directly from DC, Cp = molarity of the protein, n = number of amino acid residues in BSA, and l = path length of the cell in centimeters.
The 4000 and 33 000 values denote the total ellipticity values of the β-sheet form and the pure α-helix form of the protein at 208 nm and 222 nm.

CONCLUSIONS
From spherical Au seeds (DH = 18 nm, surface = 1020 nm 2 , Amax at 520 nm, ζ-potential = −36.8 mV, and PDI = 0.602), piñatalike MBAuNPs (8-20 peaks, D H = 125 nm, Amax at 775 nm, ζpotential = −24.4 mV, and PDI = 0.455) are stoichiometrically synthesized. The surface area of MBAuNPs ranges from 24 984 nm 2 to 40 669 nm 2 . MBAuNPs are potentially useful for photothermal tumor ablation since they induce hyperthermia of more than 4 ○ C in phantom gels irradiated with an 808 nm laser source with a power of 0.75 W/cm 2 . MBAuNPs exposed to BSA acquire a protein corona with an internal "hard" portion composed by one or two layers of BSA molecules covalently bound to the surface of each nanoparticle (1500 molecules per MBAuNP), and an external "soft" portion formed by non-covalent bound agglomerated BSA molecules. BSA functionalization decreases the tendency of MBAuNPs to aggregate but increases the polydispersity index, indicative of greater size dispersion. MBAuNPs and MBAuNPs-BSA exposed to human serum bind HSA and other human serum proteins ranging from 25 kDa to 180 kDa, which decrease their stability and increase their DH, especially in MBAuNPs-BSA.