1
Endogenous carbon nanoparticles as source of blue autofluorescence in biological fluids: eventual use in clinical praxis
A.Kuznetsov1, A.Frorip1, M.Ots-Rosenberg2, A.Sünter2, S.Patsaeva3
1- AS Ldiamon, TartuSciencePark, Riia 185, 51014, Tartu;
2 – TartuUniversity; 3- MoscowUniversity
Blue (auto)fluorescence (BF) exсmax/ emmax =3205/4205 nmin biofluids (serum, hemodialysate, hemofiltrate, urine) has enhanced intensity for chronic kidney disease patients (CKD Pts) and is therefore of clinical interest [1-8]. Nevertheless the BF substance is not finally identified yet.
Our approach to indentify the BF is based on the strong similarity in absorption and BF spectra of biofluids, carbon nanodots (CND) and fulvic acid (FA) solutions.
The aim of this work is to present the most expressive examples of this sustainable similarity and explain its nature on the united basis. The search for possibility to use BF for sensing purposes in the clinical praxis was the second aspect of this study.
Methods and samples.Biofluids were collected in the nephrology department of TartuUniversityHospital. CNDs were synthesized by microwave assisted method using sugar as precursor material [9]. FA 2S103F and 1R105F were purchased fromthe International Humic Substances Society ( FA is a ubiquitous end product of organic decay in the nature and is present, e.g.,in drinking water when it is supplied from surface basins (lakes etc.). Tap water is a proper sample for comparative experiments.
Measurements of absorption and fluorescence have been carried out in the range 200-700 nm using mostly 3D registration (NarTest NTX 2000, LDI Tallinn).
Results
1) Equal red shift. All spectra of BF in biofluids, CND and FA solutions (e.g.,tap water) show highly characteristic dependence on the excitation wavelength:emission energy decreases by excitation energy decrease, i.e.,so called, red shift (RS) - Fig.1a, b, c. We reproduce here tentatively also the BF spectra of “inorganic” oxidized graphene (GO) from [10] to demonstrate the typicality of the BF RS phenomenon.
Fig.1. Fluorescence spectra at different excitation energies: a) CKD Pt’s urine; b) oxidized graphene [10]; c) boiled and raw tap water. Raman scattering lines are present at the short wavelength tails of water spectra.
Equations y=ax+b can be written for RS, where x is the excitation energy and y – the main maxima energy of emission bands (x and y are expressed in eV).
In Fig.2 are depicted the linear graphs for BF RS empirical equations forsome substances investigated. We see that there are two groups of samples with equal or proximate RS. The first group consists of HD, urine, CND and oxidized graphene. For the second group with smaller slopes of graphs are represented: CKD Pt’s serum, FA rich natural yellow water from a dystrophic lake and solution of FA 2S103F.
Fig.2. Red shift graphs for subjects investigated. Example equations are given for CND (in blue) and for serum (red). Note different y axis for different groups. Some lines are extrapolated beyond the measurements limits. For GO data from [10] is used.
RS phenomenon itself and proximity of RS coefficients give the “first signal” about possible common nature of BF in these different substances.
2) Featureless absorption.Fig.3together with the BF excitation spectra for urine (NC) shows the absorption spectrum which is also very much characteristic for all substances investigated. There are usually no selective absorption bands in the region where the BF can be excited most effectively. Featureless absorption in the range 300-400 nm is typical also for CND and FA solutions.
Sometimes, it is possible to seea peculiarity in the first derivative spectrum as it is the case for urine at 330 nm in Fig. 3.The BF excitation spectra in other biofluids, CND and FA are like that in Fig.3.
Fig.3. Excitation spectra of a NC’s urine for marked emission wavelength. Blue and red bold lines reflect the absorption and first derivative of absorption.
3)Spectral overlapping.We have observed, for the first time, the coincidence of emission spectra in biofluids and in CND solutions. In Fig.4a the coincidence takes place in two aspects: 1) the band shapes overlap as a whole and 2) both spectra show substructures at 380 and 430 nm. We will pay special attention to the structure 380 nm
Fig.4. Fluorescence spectra at excitation 325 nm. a) Hemodialysate and CND; b) CKD Pt’s serum and boiled tap water.
(see below).
In Fig.4b it is shown the coincidence of emission spectra in serum and in boiled tap water. The overlapping is almost perfect, deviations occur only in the short wavelength region with the Raman scattering replica for water and pronounced 380 nm hump for serum.
4) FWHM change. There is a parallel trend - the integral FWHM(full width at half maximum) of all emission bands in matters investigated decreases by decreasing of excitation energy (Fig.5). We do not consider here the subbands structure in integral emission spectra (see Fig.1a,c)
Fig.5. FWHW decrease by decrease of excitation energy.
5) Aluminum ions as an indicator of similarity.We have found that after adding of the aluminum salts Al2(SO4)3, Al(N03)3 and AlCl3 into samples investigated the selective BF intensity increase,peaked at 380 nm,takes place. Fig.6 demonstrates this effect for serum (AlCl3 1.28 mg/mL was added).
Fig.6. Change of diluted serum fluorescence (exc 320 nm) after reaction with AlCl3.
In Fig. 7 are drawn two groups of difference spectra obtained (as in Fig.6) after adding 10 mM of Al ions into samples HD and “sugar” CND solution (a) or into biofluids (b). A full match and cross-match of thesespectra strongly point to the identity of the reaction agents in both groups of substances.
a) b)
1
Fig.7. Coincidence of difference spectra induced by adding of Al ions (10mM) into: a) serum and CND solution or b) biofluids as marked.
In the case of biofluids we have used the samples taken from patients as well as from healthy people. Qualitatively the effects are very similar and the difference is only in the quantities: for the Pts the BF intensities are always higher (2-10 times) than for NCs regardless were aluminum ions added or not. The BF intensity increase is the basic momentin this subject from the clinical point of view.
Aluminum salts react also with FA and yellow water with the analogous result – a new fluorescence component at 380 nm arises (see Fig.8).
Fig.8. Dependence of fluorescence 320/380 nm in HD, serum and FA dissolved in blank dialysate on concentration of Al ions.
Discussion.A striking similarity observed in the spectra of different substances (biofluids, CND and FA solutions) can be explained on the basic role which play aromatic carbon moieties.
Already in early thirties last century (1931) there was proposed by W.Fuchs a model for humic substances in coal [11] (Fig.9a) which is very much like the model intensively used in modern studies of CND, especially of blue fluorescent graphene [10] (Fig.9b).
Fig.9. Fuchs’s model for coal humic substance (a) [11] and model of blue luminescent oxidized graphene (b) [10].
Existing theoretical calculations [12] done for flat one layer aromatic carbon fragments like (b) support and explain one more our observation, i.e. the existence of quasi resonance excitation energy (QREE). QREE is the excitation energy when y = x in the RS equations y = ax + b.
We have obtained 11 empirical QREEsfor all kind of matters touched in this study with mean value 2.050.23 eV. According to [12] the energy of 2 eV is the limit (minimal) energy of - transitions in the large clusters of 20 or more aromatic rings (Fig. 11).
Fig.10. The graphene clusters and the energy gap equal to minimal excitation energy[12].
The gap energy of 2 eV in Fig.10 can be compared with observed by us QREEs.
How the BF substance can arise in the body? The first possibility is evident and trivial: FA substances accumulate in blood as a result of drinking FA waters. We remind the proximity of BF RS just for serum and waters (Fig. 2) and the emission spectra as well (Fig. 4b).
The other reason and source can be of greater interest. Intrinsic normal and inflammation stimulated decay processes (phagocytosis, necrosis etc.)occur in the body of patients and healthy people and can lead, in principle, to formation of FA like substances. As a measure of inflammation level one can use the concentration of C-reactive protein (CRP) in blood. Fig. 11 shows the positive correlation between the BF intensity and concentration of CRP in blood of CKD Pts.
Fig.11. The dependence of the normalized blue fluorescence intensity in HDs on the concentrationof CRP in the blood of CKD Pts. Normalization was done in relation to the intensity of the UV fluorescence of proteins and peptides in the same samples. N=53. The analogouscorrelation BF in CKD Pts sera versus CRP has also been found.
The correlation in the scales “CRP concentration – BF intensity” we take as an impetus for elaborating of hypothesis concerning the genesis of BF substance in the human body. Normal and pathological organic matter decay in the body can be the main reason of existence of BF substance.
Conclusions. Wecan state that BF phenomenon manifests itself in the variety of different substances (biofluids, CNDs and FA solutions) as a rather universal wavelength dependent autofluorescence, most intense in the biofluids of sick people, e.g., chronic kidney disease patients. The phenomenon can be investigated further from the united physical position, i.e., carbon nanodots ensembles fluorescence approach. It should be further tested as a tool for sensing of inflammation, degradation and ageing processes in the body.
The investigation was supported by Archiemedes project 3.2.1101.12-0011.
References
[1] D. McKenzie, A.R. Henderson, Clin Chem Acta 70 (1976) 333-36.
[2] A. O. Vladutiu, E. E. Cunningham, J. Walshe, Clin Chem 24 (1978) 1084-1085.
[3] H. A. Schwertner, S. B. Hawthorne, Clin Chem 26 (1980) 649-652.
[4] H. A. Schwertner, S. T. Weintraub, S. B. Hawthorne, Clin Chem 26 (1980) 1927-1928.
[5] G. Digenis, A. G. Hadjivassiliou, D. Mayopoulou-Symvoulidis, N. Tsaparas, A. Symvoulidis, Clin Chem 27 (1981) 1618-1619.
[6] H. A. Schwertner, Nephron 31 (1982) 209-211.
[7] H. Mabuchi, H. Nakahashi, Clin Chem 29 (1983) 675-677.
[8] M. Shaykh, N. Bazilinski, D. S. McCaul, S. Ahmed, A. Dubin, T. Musiala, G. Dunea, Clin Chem 31 (1985) 1988-1992.
[9] X. Wang, K. Qu, B. Xu, J. Ren, X. Qu, J Mater Chem 21 (2011) 2445-2460.
[10] D. Pan, J. Zhang, Z. Li, M. Wu, Adv Mater 22 (2010) 734-738.
[11] W. Fuchs, Die Chemie der Kohle, Springer, Berlin, 1931.
[12] G. Eda, Y. Y. Lin, C. Mattevi, H. Yamaguchi, H. A. Chen, I. S. Chen, C. W. Chen, M.
Chhowalla, Adv Mater 22 (2010) 505-509.