2.1. Pulse Radiolysis of Albumin Solutions
The main purpose of our research was to study the reactions induced by ionizing radiation in protein aqueous solutions under reductive radical stress. To study the reaction of hydrated electrons with human serum albumin, we applied pulse radiolysis.
Figure 1 shows transient absorption spectra recorded at various times after the electron pulse irradiation of N
2-saturated aqueous buffer solution containing 300 µM HSA. ΔA represents the difference in absorbance compared to the non-irradiated sample.
A broad absorption band of
with a maximum at ~720 nm was observed immediately after irradiation of the buffer solution of HSA (300 μM) containing
t-BuOH (0.1 M). This band disappeared within a microsecond time scale [
1]. New absorption in the visible region with a maximum near 420 nm is due to the reduction product of HSA. The main process of reaction of a hydrated electron with HSA is the formation of the disulfide radical anion (CyS-SCy
•−). Referring to our previous work (and references therein), we have proven that disulfide bridges are the most reactive residue in the capture of hydrated electrons by HSA [
1]. We conducted pulse radiolysis experiments separately for aqueous solutions containing amino acids, including cysteine or cystine, and for the solution of HSA. The analysis of the shape of the transient absorption spectra and the reactivity of
with HSA and amino acids suggests that electron attachment to disulfide bonds (CyS-SCy
•−) is responsible for the spectrum with a maximum at 420 nm that was recorded in the case of pulse radiolysis of the protein solution [
1].
The kinetic absorption pattern at 420 nm is complex. Due to the strong overlapping of the absorption bands of the hydrated electrons (~420 nm) and CyS-SCy
•− radicals (λ
max = 420 nm), the formation of disulfide radical anions cannot be directly used to calculate the rate constant of
scavenging by HSA. For this reason, we corrected the rate constant of the CyS-SCy
•− radical formation (a decay at λ = 420 nm), taking into account the decay of the
at 720 nm where the reduction product of HSA does not absorb light (
Figure 2). The normalized decay of the hydrated electron was subtracted from the absorption trace recorded at 420 nm. After correction, it can be concluded that the decay of
and formation of CyS-SCy
•− radicals evolve at the same rate (k
decay = (1/232 ns) = 1.43 10
6 s
−1; k
formation = (1/228 ns) = 1.46 10
6 s
−1).
We also observed that the absorbance of hydrated electrons (λ
max~720 nm) produced in the albumin solution, even at high concentrations (above 300 μM), was only about 2% lower compared to the neat buffer solution for the same dose. The difference in the initial hydrated electron absorbance value can be attributed to dry electron capture by HSA. In our opinion, the percentage of dry electrons scavenged by proteins in pulse radiolysis is insignificant and may be omitted in the analysis of the HSA reduction process. In the case of other proteins, the contribution of the dry electron to the albumin scavenging process may be different [
26]. In the case of pulse radiolysis of protein solutions isolated from hen egg whites, the contribution of dry electrons in the scavenging was much higher than in the case of the studied albumins and amounted to 23%.
We estimated that at pH 7.2, about 50% of all electrons produced in pulse radiolysis (15.0 µM) contributed to the formation of 7.7 µM CyS-SCy
•− radicals (
Figure 3). The reduction process was monitored at 420 nm at the maximum of the absorbance band of the product of one electron-reduced HSA. This strongly indicates that not all electrons are attached to the disulfide residue. The remaining 50% of electrons reacted with the residues of the HSA polypeptide chain, and the primary capture products were not visible in the pulsed radiolysis measurements. It has been proven that these reactions of the hydrated electron with HSA mainly lead to the modification of several specific methionine residues [
24].
Based on the literature data [
24,
25,
26,
27] and our measurements, it was proved that the modifications ascertained in the protein structure were highly dependent on the system under investigation and the experimental conditions used during sample irradiation. To accurately determine the yield of CyS-SCy
•− radicals, the concentration of scavengers should exceed the hydrated electron concentration.
Figure 3 shows the kinetics of the formation of the electron–protein reaction product for several HSA concentrations. An electron beam was used for the measurements, which produced approximately 15 µM hydrated electrons in each of the tested solutions. For HSA concentrations below 150 µM, not all electrons were captured by the protein. Moreover, when analyzing the rate constant of hydrated electron decay for various proteins, the measurements must be carried out with the same buffer solution and the ratio of protein concentration to
concentration, preferably in the same measurement series. In our opinion, a protein concentration ten times higher than the concentration of the hydrated electron produced during radiolysis measurement should be used. According to the mentioned rules, it can be concluded that the decay of hydrated electron in HSA or BSA solution evolves at the same rate (
Figure 4).
In the buffer solution, the reaction of the hydrated electron with HSA or BSA is diffusion-controlled. The rate of the reaction was determined from the hydrated electron decay in the vacuum-deaerated aqueous solution of BSA containing 0.1 M
t-BuOH (absorbance at 720 nm, at room temperature) for various protein concentrations. The decay curves, absorbance vs. time, were fitted by the pseudo-first-order kinetic equation. The dependence of the rate constant for the decay of the absorption band of hydrated electrons in the reaction with BSA is linear over the entire range of protein concentrations (Inset
Figure 4). From the slope of the line k
obs vs. BSA concentration, the second-order rate constant for scavenging was calculated as 1.3 × 10
10 dm
3 mol
−1 s
−1. This value is in good agreement with the value we previously obtained for the rate constant of the reaction
with HSA: 1.1 × 10
10 dm
3 mol
−1 s
−1 [
1,
2,
3,
4,
5,
6,
7,
8,
9,
10,
11,
12,
13,
14,
15,
16,
17,
18,
19,
20,
21,
22,
23,
24,
25,
26,
27,
28].
The reaction of radicals with albumins takes place in the physiological environment of a living cell. Free radicals are involved in many biological processes, including apoptosis or aging. For this reason, measurements under aerobic conditions were carried out to study the effect of oxygen on the HSA reduction process. It is well known that the energy barrier for the diffusion of oxygen molecules through the albumin structure is much higher than the analogous barrier in water. Similarly, the diffusion of reducing agents (
, H
•,
) in the internal albumin matrix is also inhibited.
Figure 5 shows the transient absorption spectra recorded after the electron pulse irradiation of N
2- or O
2-saturated buffer solutions containing human serum albumin (30 or 300 μM) and
t-BuOH (0.1 M). The oxygen concentration in an O
2-saturated solution at room temperature is about 1.25 mM. The superoxide anion radical is formed as a result of one-electron reduction of oxygen by the hydrated electron. The value of absorption of the transient recorded during pulse radiolysis of 30 µM HSA is about 0.025 (λ
max = 420 nm) for a 40 Gy dose in the case of reaction with hydrated electrons (
Figure 5). For a concentrated solution of albumin (300 μM HSA), this value increases almost 2.5 times for the same dose. Saturation of the albumin solution with oxygen does not lead to the formation of the HSA reduction product. This is probably due to the low value of the reduction potential of superoxide anion radical
(E
0 = −0.33 V) compared to
(E
0 = −2.9 V).
The reaction of the hydrogen atom with HSA leads to the formation of very small amounts of CyS-SCy
•− based on the intensity of the band with a maximum of 420 nm (
Figure 5). In steady-state measurements of HSA solutions irradiated with the same dose in the case of the reaction of
or H
• with albumin, we recorded identical absorption spectra. It should be emphasized that in the case of a solution containing HSA and saturated with oxygen, after irradiation, the radical anion
is formed as well as small numbers of hydrogen atoms.
Additionally, it was found that neither reagent (
or H
•) led to the formation of HSA aggregates. This was evidenced by the absence of an emission spectrum typical of protein aggregates (
Figure 6). Before irradiation, the emission spectrum of HSA showed one band with a maximum at 430 nm (λ
exc = 345 nm). After irradiation of the solution containing HSA (30 µM) and
t-BuOH (0.1 M), the formation of a fluorescent product with maximum emission at 420 nm was observed (λ
exc = 345 nm). Saturation of the same solution with oxygen before radiation did not lead to emission centered at 420 nm (
Figure 6). The emission spectrum of HSA recorded after irradiation of an O
2-saturated solution containing
t-BuOH is almost identical to that before irradiation. We recorded identical spectra after the reaction of the hydrogen atom with HSA and, in order not to complicate
Figure 6, we omitted them. Therefore, an important conclusion can be drawn regarding the process of obtaining albumin nanostructures by irradiation, namely, the need to remove oxygen from the HSA or BSA solution before irradiation. Under the influence of air, radiolysis leads to a reduction in the effectiveness of the aggregation process and, even more unfavourably, to the scission of peptide chains. Most likely,
and H
• attack other -S-S- moieties within HSA than the hydrated electron. These initially reduced disulfide bridges separate and are reconnected, but intramolecularly, this leads to modification of the albumin without aggregation. In the case of albumin reduction by hydrated electrons, selected cystine groups are modified so that the best possible conditions for intermolecular -S-S- exchange are met. In the case of an O
2-saturated solution containing HSA, t-BuOH (to remove hydroxyl radicals) and HCOONa, no new emission was observed after irradiation. This means that in a sample saturated with oxygen, the generated
reacts mainly with O
2.
The rate constant for the reaction of the hydrated electron with HSA in buffer solution has been measured for the temperature range 278–318 K. The rate constant at various temperatures was determined based on the decay of absorbance at 720 nm.
The Arrhenius expression describes the observed rate constant of electron scavenging by has relatively well. Nonlinear regression was conducted using OriginPro 2022 software to calculate the activation energy. For many years, the paradigm of radiation chemistry was that the activation energies for hydrated electron reactions are equal to the self-diffusion of water and tend to be in the 12–15 kJ mol
−1 range. Most of the scavenging reactions of a hydrated electron are diffusion-controlled reactions and, as for HSA (60 µM), the energy of activation oscillates around the value 16.8 ± 1.0 kJ mol
−1, as calculated by us (
Figure 7). We performed the same measurements for a neat HSA solution with a greater concentration than 60 μM. Regardless of the protein concentration, we always obtained the same activation energy value. This result suggests that the thermodynamics of the reduction of amino acid residues in HSA is determined by the diffusion of the hydrated electron through bulk water to the HSA surface.
It is known from our experiments and from the literature that the reaction of HSA with a hydrated electron leads mainly to the reduction of the cystine residue [
1,
2,
3,
4,
5,
6,
7,
8,
9,
10,
11,
12,
13,
14,
15,
16,
17,
18,
19,
20,
21,
22,
23,
24,
25,
26,
27].
Figure 8 shows the emission spectra of the HSA solution (300 µM) containing 0.1 M
t-BuOH recorded before (a) and after (b) pulse radiolysis. The solution was saturated with N
2 and irradiated with a dose of 25.2 kGy.
Spontaneous aggregation occurs in albumin solutions, and light-excited aggregates exhibit visible emission spectra before irradiation. The new fluorescence is associated with electron delocalization through the amino acid network of neighbouring HSA molecules and is not dependent on the presence of aromatic amino acids in albumin. In addition to hydrogen bonds that stabilize such aggregates, electrostatic forces, van der Waals forces and hydrophobic interactions are most likely also involved in the formation of new chromophore groups [
17]. Emission spectra of spontaneous (non-covalent) HSA aggregates can be easily recorded for concentrations as low as 30 µM.
As a result of the reaction of a hydrated electron with HSA, aggregates of albumin are formed, which are characterized by emission centered at about 420 nm after excitation of the irradiated solution with 320 nm light. Aggregates are formed as a result of the intermolecular recombination reaction of disulfide radicals. The excitation of HSA solutions with light above 330 nm leads to the bathochromic shift of the emission band with a maximum at 420 nm. Subtracting the HSA emission spectra before and after irradiation shows that as a result of the reaction of a hydrated electron with albumin, one type of aggregate stabilized by covalent S-S bridges is formed (after subtracting, we observed emission centered near 420 nm for all excitation wavelengths). Although fluorescence (emission) is not an additive quantity, a procedure of subtracting fluorescence data from the sample background (in our case, the native fluorescence of concentrated albumin solutions) is often used [
29].
The size of generated aggregates of albumin was monitored by dynamic light scattering (DLS). Analysis of DLS measurements after irradiation of N2-saturated solution containing HSA (300 μM) confirmed the formation of protein aggregates. Size measurements were made with a buffer solution of HSA (300 μM) and t-BuOH (0.1 M) before and after irradiation. HSA solutions (300 µM) were γ-irradiated with doses from 0 to 90 kGy. In the native solution, the mean size of albumin particles is about 7 nm. The average size of the HSA aggregates after irradiation was in the range of 17–38 nm. The conducted experiment confirmed that the size of HSA aggregates increases with the increase in the radiation dose (the size of protein aggregates depends linearly on the radiation dose). Using ionizing radiation, the formation of HSA nanostructures allows for control of aggregate size based on radiation dose and protein concentration.
The impact of radiation type (electron beam or gamma radiation) on the aggregation process of HSA was investigated. Emission excitation spectra of irradiated (electron beam or gamma radiation at 50 and 90 kGy, respectively) solutions containing HSA (420 μM) and
t-BuOH (0.1 M) were recorded for different emission wavelengths (
Figure S1 in Supplementary Materials). Spectroscopic measurements suggest the presence of two types of aggregates. Our previous spectroscopic measurements have shown that HSA undergoes reversible self-aggregation through protein–protein interactions [
17]. This is probably a consequence of the formation of the hydrogen bond network and/or π–π interactions (van der Waals force) between HSA molecules. Disulfide bridges produced by the intermolecular recombination of CyS-SCy
•− radicals as a result of radiolysis of an aqueous protein solution are the main cause of HSA aggregation through crosslinking. The formation of aggregates stabilized by hydrogen bonds is preferable when the solution is irradiated by γ-radiation. The emission excitation spectra show two bands with a maximum at 364 and 383–396 nm (γ-radiation) or 382 nm and 382–396 nm (electron pulse). The intensity of emission of the red-shifted bands depends on the type of radiation used; the emission intensity is 1.5 times lower in the solution irradiated with an electron beam.
In our further work, we compared the process of HSA reduction by chemical and radiation methods. Dithiothreitol (DTT) was used as a chemical reducing agent for disulfide bonds [
30]. The redox potential of DTT is −0.33 V at pH 7 [
31].
Figure 9 shows the emission spectra of the HSA solution (150 or 300 µM) containing DTT (5 mM) or
t-BuOH (0.1 M). The recommended DTT concentration for protein reduction is 1–10 mM [
31]. The albumin solution with the addition of
t-BuOH was deaerated and irradiated with a dose of 50 kGy. In the reaction with DTT, mainly S-S bonds within HSA, which are exposed to an aqueous solution, were reduced. Analysis of the HSA reduction process using chemical and radiation methods allowed us to confirm that -S-S- (CyS-SCy) bonds play the main role in both cases. The reaction of DTT or
with HSA led to the formation of similar products. We observed new emissions in the same spectral range in cases for
and DTT after the excitation of albumin solutions with 350 or 380 nm light (
Figure 9). Recombination of the CyS-SCy
•− radicals led to the formation of protein aggregates stabilized by intermolecular disulfide bonds. However, considering the intensity of the HSA emission bands, a better method for obtaining protein aggregates is the radiation technique. For example, the intensity of the HSA emission band after excitation of the solution with 350 nm light was 3.5 times higher for
as a protein reductant than for DTT. The radiation method does not require the use of chemical crosslinkers, and it is also possible to simultaneously synthesize protein nanostructures and sterilize the final protein product. Moreover, as we mentioned above, by using different radiation doses, the size of the generated HSA aggregates can be controlled.
The literature describes that the hydrated electron may react by tunneling over long distances within biomolecules [
32,
33]. Experimental studies of electron tunneling in proteins were initiated by DeVault and Chance in 1966 [
34]. In our research, we attempted to explain whether electrons enter BSA directly or via subsequent scavenging by disulfide bridges. Steady-state radiolysis using an electron beam was performed at low temperatures. The ethylene glycol–water glassy matrix, containing BSA, was irradiated at 77 K. Under these conditions, many primary species and their reactions can be investigated. The absorption spectra of the 1 mM BSA at 77 K were recorded before and after irradiation with a dose of 6700 Gy and are presented in
Figure 10. After irradiation, the absorption spectrum of the glassy sample shows two bands with a maximum around 410 and 590 nm (curve 1). The band centered at 590 nm can be attributed to the trapped electrons. The absorption spectrum of irradiated glassy BSA after subtraction of absorption of the unirradiated sample is presented as a dotted line (
Figure 10). Curve 3 represents the spectrum of BSA obtained after bleaching with the light of a sample. The optical bleaching was carried out with visible light. The bleaching causes the disappearance of the spectrum of trapped electrons. After removing the trapped electron spectrum, a band with a maximum around 400 nm is still present. In the same spectral range, a band for reduced HSA (product of electron attachment to disulfide bonds) was observed in pulse radiolysis measurements. For this reason, we assume that disulfide radical anions (CyS-SCy
•−) derived from the electron attachment are probably responsible for the absorption spectrum with a maximum at around 400 nm at 77 K. It appears that electrons can reduce disulfide bridges which are in the BSA interior both by diffusion (in the case of a hydrated electron in an aqueous sample) or tunneling through the BSA at matrices at 77 K.
As a result of measurements of the scavenging electrons by CyS-SCy bridges inside albumins (HSA or BSA), we believe that their transfer is possible in the tunneling process. This is indicated by low-temperature radiolysis measurements (77 K) of the EG/H
2O matrix containing HSA or BSA. In a matrix irradiated at liquid nitrogen temperature, electron diffusion is completely inhibited. However, the formation of reduced CyS-SCy bridges is observed (
Figure 10 shows a pronounced shoulder around 400 nm). The electron can also react in its presolvated state (“dry” electron). For the albumin concentrations used (about 1 mM) in the experiments, dry electron capture does not play a significant role. This was also confirmed by pulse radiolysis measurements of liquid solutions used in low-temperature measurements.
In the next stage of our studies, an HSA solution (30 or 300 µM) containing 0.1 M NaCOOH was studied using the pulse radiolysis technique and the absorption spectrum was recorded. The results from pulse radiolysis experiments show that
radical anions react relatively fast (see comment below) and selectively with CyS-SCy groups (
Figure 11). After electron pulse irradiation of the HSA solution, a new transient absorption peak with a maximum at 420 nm is formed (the redox potential of the
radical anion is −2.0 V). BSA and HSA possess homologous amino acid sequences [
35], resulting in almost identical secondary and tertiary structures. Results for the reaction of hydrated electrons or
with both albumins are the same (almost the same rate constants, transient absorption spectra and aggregation process). When
radicals react with HSA, similar transient absorption spectra were obtained, as in the case of the reaction of azide radicals with BSA, but their intensity is different [
16]. For BSA, the absorption band is about twice as intense as the tryptophan radical in HSA. The higher absorbance in the transition spectra for BSA after one-electron oxidation is due to the presence of two tryptophan residues in the BSA structure, as opposed to one tryptophan residue in HSA. These observations confirm our previous measurements [
1] that tryptophan residues in albumin are protected from attack by reducing agents.
Temperature measurements (
Figure 12) of pulse radiolysis show that heating the HSA solution to 70 °C does not support the process of albumin reduction by
anion radicals (the absorbance with a maximum at 420 nm decreases 1.75 times after heating the HSA solution compared to the sample irradiated at room temperature).
The value of the rate constant for the reaction of
radical anions with HSA in buffer solution is around 5.7·10
8 dm
3 mol
−1 s
−1, in agreement with the value published in the literature of 7.0·10
8 dm
3 mol
−1 s
−1 (pH = 7.6) [
34]. We are aware that this value is only an estimate. When analyzing the reactions of
radical anions with a protein, the disproportionation reaction of these radicals should be taken into account, the rate constant of which is 6.5·10
8 dm
3 mol
−1 s
−1 at pH 7.0 [
36].
While examining the formation of albumin aggregates induced by reactions with
or
radical anions, we observed the formation of gels several times. These hydrogels are two-component systems consisting of a three-dimensional network of polypeptides chains and water that is trapped between macromolecules. Currently, hydrogels are used, among other applications, as dressings, supersorbents in the treatment of obesity, drug delivery systems, contact lenses and artificial cartilages. Since the pioneering work of Wichterle and Lim [
37], ionizing radiation has been used to create hydrogels, which has resulted in the commercial production of many products [
38].
Our observed ability to easily produce albumin hydrogels using radiation opens the way to preparing materials for therapeutic purposes. This is a very important issue to which we intend to devote our next publication. Key methods for irradiating human or bovine albumin solutions to obtain hydrogels will be presented below. The use of an electron accelerator is more effective than a gamma radiation source. To obtain gels using the radiation method, the concentrations of albumin solutions must be high, greater than or equal to 300 µM. Both small and large doses of ionizing radiation can be used to effectively produce hydrogels. Small doses require heating the irradiated solutions to a high temperature of 70 °C. When
radical anions were used to reduce albumin, even with a few electron pulses at a dose of 820 Gy in HSA solution (300 µM, 70 °C), a gel was generated. When the hydrated electron was used to initiate the HSA aggregation process (under the conditions described above for
radical anions), a gel was also obtained for only a slightly higher dose. This may be due to the fact that sodium formate changes the structure of HSA to some extent and facilitates the reduction of CyS-SCy bridges. For practical reasons, it is essential that both sodium formate and
t-BuOH are non-toxic and suitable for use in cosmetics [
39,
40]. We also obtained hydrogels at room temperature using approximately 10 times higher doses compared to measurements at 70 °C. Irradiation of HSA solutions at room temperature is more promising in obtaining hydrogels. The key to the effective preparation of hydrogels is the method of delivering the electron beam to the starting solutions. The time interval between electron pulses from LINAC should correspond to the decay time of the reduced CyS-SCy
•− bridges in HSA, which, as we have shown, is several seconds [
1]. In our measurements, we used a pulse repetition rate of approximately 0.5 Hz. To develop a procedure for obtaining commercial gels, this parameter must be optimized. The rate of energy delivery of ionizing radiation to albumin solutions under reducing conditions affects the rheological properties of the gels and their degree of transparency. In some experiments we managed to obtain clear, transparent gels; in others, they were yellow and opalescent. We presented the absorption spectrum of the clear gel in
Figure S2 in the Supplementary Materials. The emission spectra of the hydrogel differ from those of the spontaneous aggregates, as shown in
Figure S2. At the same time, the emission spectra of hydrogels are very similar to analogous spectra of aggregates obtained as a result of irradiation of solutions in conditions that do not lead to gelation. This observation suggests that intermolecular disulfide bridges are formed both in hydrogels and in aggregates obtained as a result of irradiation of HSA solutions (subsequent electron pulses are delivered every several dozen seconds or later).
Measurements of the decay times of light emission generated in albumin solutions using short laser pulses provide helpful information on the spontaneous and ionizing radiation-induced aggregation of HSA or BSA [
17]. This technique should be used with great caution for complex systems such as protein solutions. Differences in the emission lifetime of aggregates depend on the initial concentrations of albumin solutions, pH, measurement temperature and hysteresis of temperature changes. Valuable information was provided from the analysis of the emission time measurements (after excitation with a 337 nm pulse in room temperature for 420 nm) of the solutions: a) the initial BSA solution (300 µM, 0.1 M t-BuOH), b) solution “a” after irradiation at 70 °C with a dose of 1300 Gy (30 pulses every 10 s) and c) the solution that absorbed the same total dose, but with high repeatability (1300 Gy; 30 pulses in 15 s). The BSA aggregates induced by ionizing radiation are characterized by specific fluorescence spectra compared to the emission of non-irradiated solutions (
Figure S3 in Supplementary Materials). The excitation of BSA solutions with the laser light with a wavelength of 337 nm allows registration of the emission of excited protein aggregates only because this light is not absorbed by the monomeric form of BSA. The emission decay at 420 nm is characterized by a two-exponential function with an average time of 2.29 ns (t
1 = 0.49 ns; f
1 = 0.51; t
2 = 4.16 ns; f
1 = 0.49) for neat solution and 3.21 ns (t
1 = 0.52 ns; f
1 = 0.36; t
2 = 4.75 ns; f
1 = 0.64) for both the irradiated solution and the hydrogel. It can therefore be postulated that the properties of albumin aggregates generated by hydrated electrons are the same in solution and in hydrogel.