Index of /rozprawy2/10972

Pełen tekst

(1)Doctoral thesis mgr inż. Aleksandra Wandzilak. Selected chemical elements as potential indicators of cancerous brain tissue. Supervisor: prof. dr hab. inż. Marek Lankosz. Krakow 2015.

(2) Declaration of the author of this dissertation: Aware of legal responsibility for making untrue statements I hereby declare that I have written this dissertation myself and all the contents of the dissertation have been obtained by legal means.. data. podpis autora. Declaration of the thesis Supervisor: This dissertation is ready to be reviewed.. data, podpis promotora rozprawy. 2.

(3) I would like to thank my supervisor, Professor Marek Lankosz, for his support, sharing his vast knowledge and for his almost unlimited patience. I would also like to thank Professor Dariusz Adamek and Edyta Radwaoska for contributing their time, expertise and for their cooperation. I would like to thank the members of our research team for inspiring discussions. I also thank all those who have always believed in me. Last but not least, I would like to thank my parents and Pawel without whom this thesis wouldn’t have been completed.. 3.

(4) The research leading to these results received funding from: the European Community’s Seventh Framework Programme (FP7/2007-2013) under grant agreement No. I-20090047, HASYLAB, Hamburg. I acknowledge the European Synchrotron Radiation Facility for provision of synchrotron radiation facilities (projects no.: MD 726 and MD 676). I thank Diamond Light Source for access to beamline I18 (proposal number: sp7553) that contributed to the results presented here. The research leading to these results received funding from the Ministry of Science and Higher Education (Warsaw, Poland) grant No. N N518 377 537.. I would also like to acknowledge the Royal Society of Chemistry for allowing authors to reuse their own published materials in a thesis. (doi: 10.1039/C3MT00158J). This work was supported by “Doctus – Lesser Poland fellowship program for PhD students” (2012-2015).. This PhD thesis has been completed within the framework of the Human Capital Operational Program POKL.04.01.01-00-434/08-02 co-financed by the European Union.. 4.

(5) CONTENTS Abstract................................................................................................................................. 8 Abstrakt .............................................................................................................................. 10 1.. Introduction and goal of the study ........................................................................... 12 1.1. Literature review .................................................................................................. 13. 1.2. Motivation............................................................................................................ 16. 2.. Biological background............................................................................................... 17 2.1. Brain tumours ...................................................................................................... 17. 2.2. Biologically significant chemical elements .......................................................... 21. 2.2.1. Iron ................................................................................................................ 21. 2.2.2. Copper .......................................................................................................... 22. 2.2.3. Zinc ................................................................................................................ 23. 2.2.4. Calcium ......................................................................................................... 24. 2.2.5. Phosphorus ................................................................................................... 24. 2.2.6. Sulphur .......................................................................................................... 25. 2.2.7. Potassium, sodium, chlorine ........................................................................ 25. 3.. Theorethical background of the research methods ................................................. 26 3.1. Synchrotron radiation .......................................................................................... 26. 3.2. X-ray fluorescence................................................................................................ 29. 3.3. Ion beam analysis ................................................................................................. 32. 3.4. X-Ray Absorption Spectroscopy ........................................................................... 33. 4.. Experimental ............................................................................................................. 40 4.1. Sample preparation ............................................................................................. 40. 4.1.1. Freeze-dried samples.................................................................................... 40. 4.1.2. Frozen samples ............................................................................................. 41 5.

(6) 4.2. Measurement methodology ................................................................................ 42. 4.2.1. X-Ray Fluorescence ....................................................................................... 42. 4.2.2. Ion Beam Analysis ......................................................................................... 45. 4.2.3. X-ray Absorption Spectroscopy .................................................................... 51. 4.3. Statistical analysis ................................................................................................ 57. 4.3.1. Comparing two groups ................................................................................. 57. 4.3.2. Multivariate analysis ..................................................................................... 58. 4.4 5.. Mathematical presentation of histological data ................................................. 58 Measurement results and discussion ....................................................................... 60. 5.1. Concentration of selected elements in brain tissue ............................................ 60. 5.1.1 5.1.1.1. Study of homogeneous tissue with a photon beam ................................. 62. 5.1.1.2. Study of homogeneous tissue with an ion beam ..................................... 66. 5.1.1.3. Comparing the results of photon and ion beam studies .......................... 69. 5.1.1.4. Study of calcifications and blood vessels .................................................. 70. 5.1.1.5. Testing sample homogeneity .................................................................... 74. 5.1.1.6. Issues connected with studying concentrations of elements .................. 75. 5.1.2. The effect of neoplastic processes on spatial distribution of elements ...... 76. 5.1.2.1. Study of spatial distribution of elements with a photon beam ................ 77. 5.1.2.2. Study of spatial distribution of elements with an ion beam .................... 80. 5.1.3. 5.2. The effect of neoplastic processes on the concentration of elements ........ 61. Sample classification for the histological type of tumour ............................ 81. 5.1.3.1. Classification of tissues based on photon beam study ............................. 81. 5.1.3.2. Classification of tissues based on ion beam study.................................... 84. Oxidation states and chemical forms of the studied elements ........................... 88. 5.2.1 5.2.1.1. Iron ................................................................................................................ 88 Frozen samples ......................................................................................... 88 6.

(7) 5.2.1.2. Freeze-dried samples ................................................................................ 91. 5.2.1.3. The effect of sample preparation on the results ...................................... 92. 5.2.1.4. Influence of oxygen supply on the oxidation state................................... 94. 5.2.1.5. Changes in chemical environment ............................................................ 96. 5.2.2 5.2.2.1. Study of oxidation state ............................................................................ 98. 5.2.2.2. Changes in chemical environment ............................................................ 99. 5.2.3. 6.. Zinc ................................................................................................................ 98. Copper ........................................................................................................ 100. 5.2.3.1. Study of the oxidation state .................................................................... 100. 5.2.3.2. Forms of copper in brain tissue .............................................................. 102. Bibliography ............................................................................................................ 109. 7.

(8) A BSTRACT The second most common cause of death is cancer, many types of which are brain tumours. These kinds of tumours are among those with the smallest survival rate, with most patients living for less than one year after diagnosis. In order to successfully combat the disease it is extremely important to learn about the mechanism of its formation. It is believed that trace elements play a significant role in neoplastic processes. Iron plays a particularly important role: via Fenton’s reaction it controls the formation of reactive oxygen species which are a major cause of DNA damage. Copper and zinc are important for protection against superoxide radicals as they are present in active centres of CuZn superoxide dismutase (SOD1). Also other elements present in brain tissue are important for its proper functioning. For this reason, information on the differences in their concentrations and forms in healthy and cancerous tissues may significantly contribute to the knowledge concerning biochemical reactions involved in oncogenesis. It may be also useful in supporting histopathological diagnostics. Among the samples analyzed, there were tumours with different grades of malignancy, according to the latest World Health Organization (WHO) classification, and different histopathological types. As a non-cancerous control sample, brain abscess wall and post mortem collected samples were used. The study was approved by the Jagiellonian University Bioethical Committee. For the purpose of mapping, freeze-dried thin tissue sections were used. In studying the oxidation states we used mostly frozen samples in order to slow down biochemical processes, e.g. oxidation. In studying the distribution of chemical elements, µXRF and µPIXE methods were used. For the purpose of determining the oxidation state of the elements and changes in chemical environment, Xray absorption spectroscopy was used. XRF study showed a statistically significant decrease in P, S, Ca and Fe concentrations and an increase in Zn concentrations in cancerous tissue comparing to healthy tissue. On the other hand, PIXE study showed a statistically significant increase in Na, K, Cu and Zn concentration in cancerous tissue. The results of concentration measurements obtained using the PIXE and XRF methods are not consistent for most of the elements studied. This has most probably been caused by the differences in density between the samples. Combining the PIXE and STIM methods enabled the elimination of the influence of sample thickness. Results obtained with both methods are consistent for Zn analysis and show its levels increase in cancerous samples relative to the control. Also, interesting changes were observed in the distribution of elements. It was shown that the higher the malignancy grade the less spatially correlated are the elements. The strongest correlation can be seen for transition metals e.g. copper and zinc. These observations were made in homogeneous areas. For comparison, the study also included various structures present in the tissue. It was shown that in the vicinity of calcifications, the tissue contains elevated amounts of calcium, and in the vicinity of blood vessels – even up 8.

(9) to 100 µm – the concentrations of some elements are clearly elevated. The analysis of simultaneous measurements of concentrations of elements in homogenous areas of the samples and employing multivariate discriminant analysis enabled very effective classification of various types of brain tumours. The method made it possible to accurately classify more than 90% of samples into the histopathological types. Elements of greatest significance in classifying were Cu, K, Fe, Ca, and Zn. Analysis of the oxidation states showed noteworthy changes in the case of iron. In all the samples iron occurred in compounds with both Fe2+ and Fe3+, however, in frozen samples the ratio of Fe2+ to Fe3+ content in the tissue was noticed to increase with the tumour malignancy grade. The results suggest that the lowered oxidation state of iron can be a marker of cancerous tissue. No similar correlation was found for freeze-dried samples. The discrepancy between the results obtained by these two preparation methods shows that cryogenic conditions produce accurate results in studies of biological samples, whose form, under such conditions, is close to their native state, without preparation-caused artefacts. Also, analysis of the oxidation states of copper showed a small increase in the Cu2+ to Cu1+ content ratio with the tumour malignancy grade. The results show a great influence of sample preparation and measurement conditions. Also, the sample preparation and measurement conditions have a great influence on the outcome of the study. The results obtained here emphasise the importance of transition metals in pathological processes. This study may contribute to the understanding of biochemical processes involved in the carcinogenesis and learning more about one of the least known types of tumours. This may also be an important step in developing tumour identification methods, complementary to the histopathological examination.. 9.

(10) A BSTRAKT Drugą co do częstości występowania przyczyną zgonów są choroby nowotworowe, wśród których dużą częśd stanowią nowotwory mózgu. Charakteryzują się one niską przeżywalnością pacjentów, z których większośd żyje krócej niż rok od momentu postawienia diagnozy. Aby móc skutecznie zwalczad choroby niezwykle cenna jest wiedza na temat mechanizmu ich powstawania. Uważa się, że pierwiastki śladowe odgrywają istotną rolę w procesach nowotworzenia. Żelazo jest szczególnie ważnym pierwiastkiem ponieważ poprzez udział w reakcji Fentona, bierze udział w powstawaniu reaktywnych form tlenu mogących uszkadzad DNA. Z kolei miedź oraz cynk, obecne w centrach aktywnych dysmutazy ponadtlenkowej (SOD1), odgrywają rolę przy ochronie przed anionorodnikami ponadtlenkowymi. Pozostałe pierwiastki obecne w tkance mózgu również są ważne dla jej prawidłowego funkcjonowania. Z tego powodu informacja dotycząca różnic ich stężeo oraz form chemicznych w jakich występują w nowotworowej oraz zdrowej tkance może przyczynid się do wzrostu wiedzy na temat reakcji biochemicznych towarzyszących procesowi nowotworzenia. Może byd ona również przydatna we wspomaganiu diagnostyki histopatologicznej. Pośród badanych próbek obecne były próbki nowotworów reprezentujących różne typy histopatologiczne oraz różne stopnie złośliwości wyróżnione zgodnie z klasyfikacją sporządzoną przez Światową Organizację Zdrowia (WHO). Jako próbki kontrolne wykorzystano ścianę ropnia mózgu oraz tkanki pobrane pośmiertnie. Badania zostały zatwierdzone przez komisję bioetyczną Uniwersytetu Jagiellooskiego. Na cele mapowania, próbki przygotowano w postaci cienkich skrawków suszonych w ujemnej temperaturze. W badaniu stopni utlenienia wykorzystano próbki mrożone, ponieważ taki sposób preparatyki zapewnia spowolnienie zachodzenia procesów chemicznych, np. utleniania pierwiastków zawartych w tkance. Badania dystrybucji przestrzennej pierwiastków prowadzono przy użyciu technik µXRF oraz µPIXE. Określanie stopni utlenienia prowadzono natomiast przy użyciu spektroskopii absorpcji promieniowania X. Analiza XRF pokazała istotny statystycznie spadek stężenia Ca, Fe, P oraz S w tkance nowotworowej względem tkanki kontrolnej, natomiast stężenie Zn było wyższe. Z kolei analiza PIXE pokazała istotny statystycznie wzrost stężenia Na, K, Cu oraz Zn w tkance nowotworowej. Wyniki uzyskane w oparciu o te metody nie są spójne dla większości badanych pierwiastków. Prawdopodobnie spowodowane jest to przez różnice w gęstościach poszczególnych próbek. Sprzężenie pomiarów PIXE z pomiarami STIM pozwoliło na uwolnienie się od wpływu tego czynnika. Wyniki uzyskane obiema stosowanymi metodami spójne są tylko dla Zn i pokazują wzrost jego stężenia w tkankach nowotworowych w porównaniu do tkanek kontrolnych. Interesujące zmiany zostały również zaobserwowane w dystrybucji pierwiastków. Zauważono, że wraz ze wzrostem stopnia złośliwości nowotworu spada korelacja przestrzenna pierwiastków. Najsilniejsza zależnośd została zaobserwowana dla pierwiastków przejściowych np. miedzi i cynku. 10.

(11) Zależności te zostały zaobserwowane w jednorodnych obszarach próbek. Ponadto pomiary zostały wykonane na różnych strukturach obecnych w tkance. Pokazano, że w pobliżu zwapnieo tkanka zawiera znacznie podwyższone ilości wapnia, natomiast w pobliżu naczyo krwionośnych, w odległości nawet do 100 µm stężenie niektórych pierwiastków jest wyraźnie wyższe. Wzięcie pod uwagę stężeo wszystkich badanych pierwiastków równocześnie oraz zastosowanie statystycznej analizy wielowymiarowej pozwoliło na skuteczną klasyfikację próbek do poszczególnych typów nowotworów. Zastosowana metoda pozwoliła na poprawne zaklasyfikowanie ponad 90% próbek. Wśród pierwiastków mających największe znaczenie dla poprawnej klasyfikacji znalazły się Cu, K, Fe, Ca oraz Zn. Analiza stopni utlenienia wykazała ciekawe zmiany dla żelaza. W wszystkich badanych tkankach żelazo było obecne zarówno w postaci Fe2+ jak i Fe3+, natomiast w próbkach mrożonych zauważono wzrost zawartości Fe2+ w stosunku do Fe3+ wraz ze wrzostem stopnia złośliwości nowotworu reprezentowanego przez próbkę. Otrzymane wyniki sugerują, że obniżony stopieo utlenienia żelaza może byd wskaźnikiem tkanek nowotworowych. Dla próbek suszonych w ujemnych temperaturach nie zaobserwowano podobnej zależności. Rozbieżności między wynikami uzyskanymi w tych dwóch różnych podejściach pokazują, że warunki kriogeniczne dają dobre rezultaty w badaniu próbek biologicznych. W takich warunkach próbki znajdują się w postaci zbliżonej do tej w której znajdują się w organizmie człowieka, nie zmienionej znacznie w procesie preparatyki. Analiza stopni utlenienia miedzi pokazała natomiast mały wzrost zawartości Cu 2+ w stosunku do zawartości Cu1+ wraz ze wzrostem stopnia złośliwości nowotworu. Wyniki przeprowadzonych badao pokazują jak duży jest wpływ preparatyki próbek oraz warunków pomiarowych na rezultaty. Podkreślają one również wagę pierwiastków przejściowych w procesach patologicznych. Otrzymane wyniki mogą również przyczynid się do zwiększenia wiedzy dotyczącej procesów biochemicznych zachodzących w procesie nowotworzenia oraz do wzrostu wiedzy na temat jednego z mniej poznanych nowotworów. Byd może mogą byd one również kolejnym krokiem w rozwoju komplementarnych do analizy histopatologicznej metod identyfikacji typów nowotworu.. 11.

(12) 1. I NTRODUCTION AND GOAL OF THE STUDY The development of human civilisation is accompanied by increasing influence of adverse conditions on our health. These mainly involve the lifestyle-related factors, such as rich in synthetic components diet, stress and the lack of physical activity. Environmental factors related to increasing pollution are also significant. All this contributes to the emergence of new diseases and increases the incidence of those already known. Cancer is the second (after cardiovascular diseases) most common cause of death in highly developed countries.1,2 Among all types of cancer, brain tumours have one of the smallest survival rates.3 Although thanks to continuous development in diagnostics, neoplastic diseases are detected earlier and earlier, and more efficient treatment methods are being developed, cancer mortality is still very high. In order to successfully combat the disease, it is extremely important to learn about the mechanism of its development. It is believed that trace elements such as Fe, Cu and Zn play a significant role in neoplastic processes. Although these elements are present in the human body at low concentrations, they are vital for its proper functioning. They are components of many enzymes and hormones and thus affect metabolic processes in cells. For this reason, the information about the differences in their levels, oxidation states and chemical environment in healthy and cancerous tissues with various grades of malignancy may significantly contribute to the understanding of the biochemical reactions involved in oncogenesis. The ideal tools for the study of elements present in tissues at very low concentrations are synchrotron-generated or ion microbeams. These methods make it possible to study very precisely selected areas of a sample. Another advantage is that they do not require complex sample preparation. X-Ray Absorption Spectroscopy (XAS) has proved very useful in studying the chemical state of elements present in disordered matter, such as tissues. Thanks to the non-destructive character of the measurement, valuable samples of human brain tumour can be re-tested.. 12.

(13) 1.1 L ITERATURE. REVIEW. PubMed – the largest biomedical database – indexes more than three million references to carcinomas out of the overall number of 23 million publications. These numbers show the importance of the study of carcinomas. The first publication in this field indexed in PubMed dates back to 1818. Changes in the elemental composition of neoplastic tissues have been studied since 1960’s and still attract considerable interest.4–6 These studies focus on changes in both the concentration and oxidation states of elements. X-ray Fluorescence (XRF) and Particle Induced X-ray Emission (PIXE) are most commonly used to measure elements concentrations, but Atomic Absorption Spectroscopy (AAS) is also often used. In studies of the oxidation states of elements present in tissues, the XAS method is employed. This topic of changes in the concentration of elements in brain tumours is addressed in detail in a paper by Szczerbowska-Boruchowska et al. The authors showed the average concentrations of elements present in the samples representing various types of tumours and outlined the possibility to classify tumour samples into various histopathological types on the basis of elements they contain. It was found that S, Cl, Cu, Fe, K, Br, and Zn are of the highest significance in discriminating between tumour types.7 Andrasi et al. studied brain tumours and found that Zn concentration in brain tumours is lower than in control tissues. In some areas of the brain they also observed a decrease in the concentrations of Cu and Fe.8 Concentrations of Cu and Zn in blood serum and in brain tissue were also studied by Yoshida et al. who demonstrated a significant increase in the concentrations of Cu and Zn and in the Cu to Zn ratio in malignant glioma tissues as compared to control tissue, whereas the differences between meningioma and control tissue were insignificant. No significant differences in the concentrations of these elements were found in blood serum.9 Al.-Chalabi et al. also studied blood sera from patients with brain tumours. They observed a decrease in the concentration of Zn and Se and an increase in Fe and Cu in these patients as compared to the control group. They also noticed the significant changes in the concentrations of some electrolytes: in cancerous tissues the concentrations of Na and Cl significantly increased relative to the control. No changes were found in the concentrations of K.10 An interesting investigation with MR imaging was done on brain cancer tissues. It showed an increase in sodium concentration 13.

(14) in tumours, relative to that in normal brain tissues.11 The studies on the concentration of elements in brain tissues are not numerous, which is due to the limited availability of such samples. On the other hand, the results for blood serum analysis are not consistent for all elements, so it is difficult to draw unambiguous conclusions on this issue. Other types of tumour were also studied, including cancer of breast, kidneys, prostate, liver and many others. Farquharson et al. compared the concentrations of Fe, Cu, Zn and K in healthy and cancerous liver tissues. They observed a decrease in the levels of all these elements in cancerous tissues relative to the control.12 Al.-Ebraheem et al. observed that in breast cancer the concentrations of Ca, Fe, Cu and Zn increased relative to the control tissue, with the greatest increase observed for Zn. They also noticed that increased concentrations of Ca, Fe and Zn are associated with oestrogen receptor positive status.13 Silva et al. studied the concentration of selected elements as a tool for brain tumour prognostication. They noticed that increased concentrations of Fe and Zn are associated with an increased risk of death. For calcium an opposite trend was observed.14 Vatankhah et al. also studied this type of cancer. They found that in malignant cancer tissue the concentration of Zn was elevated as compared to the benign one. Also, Cu/Zn ratio was lower for malignant cancer tissues.15 Raju et al. used the PIXE method to compare the concentration of elements in human breast cancer. This study revealed that the concentrations of Cl, Ca, Cr, Fe, Cu, Zn, As, Se, Br, Rb and Sr were elevated in cancerous tissue. No significant difference was observed for K.16 Kwiatek et al. studied kidney cancer and observed a decrease in the concentrations of Cu, Zn and Se, whereas the concentration of Fe increased in cancerous tissue relative to the controls. No statistically significant differences were found for Mn.17 Sera collected from colon cancer patients were studied by Al Faris et al. They found that patients suffering from this type of tumour had significantly lower concentration of Zn. No significant differences were found for Ca, Cu and Zn.18 An interesting study was conducted by Abnet et al. They correlated concentrations of elements in the sample with the patient’s subsequent medical history. They found that high tissue Zn concentration was strongly associated with a reduced risk of developing oesophageal cancer. No such association was found for Cu, Fe and Ni.19 Feldstein et al. performed experiments on mice inoculated with various types of cancers. The experiment showed changes in the concentration of trace elements as a function of time elapsed from inoculation.20 They noted a considerable increase in the concentration 14.

(15) of some elements (ten-fold in the case of rubidium) and a decrease in the case of Fe with time elapsed from inoculation. In the case of breast cancer, most of the studies showed an increase in the concentrations of Fe and Zn, but it is difficult to identify clear trends when comparing the concentrations of healthy and cancerous tissue in the case of other organs. Nevertheless, the elements most often showing a change in the concentration include Fe, Cu and Zn, which emphasises a particular role of transition metals in oncogenesis. In addition to concentrations, oxidation states of these elements were also studied. In their study of iron speciation in brain gliomas, Szczerbowska-Boruchowska et al. showed relationships between three samples showing an increase in Fe3+ in neoplastic tissues as compared to the control. The Fe absorption edges in samples representing highly malignant tumours were shifted to higher energies by 1.5 eV and 1.8 eV as compared to the control.21 Kwiatek et al. in their study of prostate cancer, compared two spectra: for cancerous tissue and for healthy tissue; they found that the absorption edge of the latter is shifted by about 1 eV towards lower energies.22 Al.-Ebraheem et al. showed that for average spectra of all normal and tumour tissues, absorption edges are located near the energy corresponding to Fe3+, whereas for the normal samples are shifted towards lower energies, which may indicate a higher proportion of Fe2+ in the total Fe content of these tissues.23 The latter of the cited papers also showed that zinc is present in a bounded form (Zn2+) in human brain tissue, whereas copper is present in the brain mainly in the form of Cu1+, in both healthy and control tissues. The fact that the results of many of the cited papers are divergent may be due to the various sample preparation procedures. Some samples used for concentration measurements were freeze-dried7, which is a generally accepted preparation method. In some studies samples were freeze-dried, then ground and pelletised16,17, whereby although elements are not leached, the procedure makes it impossible to select areas of interest. Some of the samples were just frozen9,12 or dried8,20. Using formalin-fixed, paraffin-embedded samples is also common13,14,19, but this method causes leaching of elements from the tissue. In studies on the oxidation states, samples were dried22, freezedried21 or fixed in formalin and then embedded in parafin23. Among these methods, the latter affects the oxidation state of elements the most, but the other methods do not protect elements present in the sample against oxidation either.. 15.

(16) 1.2 M OTIVATION The aim of this work was to study changes occurring in brain tissue as a result of oncogenesis. The study investigates changes concerning most of the elements detectable in human brain tissue, except for those most common, such as C, O and H. Among the studied elements the focus is in particular on trace metals, such as Fe, Cu and Zn due to their presence in the active centres of many important enzymes. This work deals with two main issues: 1.. Finding out the differences in elemental concentration between healthy and. cancerous brain tissue. In particular, the aim is to assess if there are also perceptible differences in the concentration and distribution of elements between samples representing various tumour malignancy grades and various histopathological types. The next step may be an attempt to assess whether the differences in concentrations make it possible to classify tumour types. 2.. Determining the chemical form of transition metals present in the tissue in order. to expand the knowledge on their role in cancerous tissues. The aim is to determine their average oxidation state in the compounds present in the healthy brain and in the tissue areas affected by tumour. Average distances between atoms located near the selected elements present in cancerous and control tissues will also be studied. For copper, spectra obtained for samples will also be compared with organic standards, to determine the proportion of its selected forms in the total content of copper.. Concentrations will be studied in freeze-dried samples to avoid leaching of the elements of the sample during preparation, whereas the oxidation states will be studied in both freeze-dried and frozen samples, which makes it possible to maintain the tissues in their native state. The goal of this study is also to determine the effect of preparation on the measurement results. The information obtained from this work may be another step towards understanding biochemical processes accompanying carcinogenesis. Another aim is to answer the question whether it is possible to identify tumour types on the basis of their elemental composition.. 16.

(17) 2. B IOLOGICAL BACKGROUND 2.1 B RAIN. TUMOURS. Brain tumours are rated between the first and second ten of most common cancers.3,24 Most of the cases are diagnosed in people in their seventies.25 In children aged 0-14 years, however, they are the second most frequent cancer, and the third one in people aged 1524 years.26,27 Although they are not a very common type of cancer, they rank high in mortality statistics. Brain tumours rank in the first ten (about seventh) causes of death among all types of cancer and account for about 3% of cancer deaths.3 For brain tumours, the 5-year survival rates – the percentage of people who live at least 5 years after being diagnosed – are below 20%, whereas the average value for all cancers is above 50%. This makes them one of the cancers with the worst prognosis. A brain tumour is an abnormal growth of tissue in the brain. Cancer formation is the effect of disturbed gene control mechanism. People who had inherited mutations in certain genes are more likely to develop cancer. Genetic predisposition and lifestyle have also some significance.28,29 Brain tumour symptoms may be connected with the damage of the centre where the tumour is located or with an increase in intracranial pressure. Brain tumours can be divided into two categories: primary brain tumours, which start in the brain, and secondary or metastatic brain tumours, which start in another part of the body and spread to the brain. There are about 130 different types of brain tumours. Primary brain tumours can be classified according to of the type of cells from which they come or the area of brain where they developed. In adults, gliomas and meningiomas are the most common types of brain cancer. I will briefly describe each of these groups.30–33. GLIOMAS – glial brain tumours are cancers that originate from glial cells. Glial cells are the most abundant cell types in the central nervous system. These cells do not directly participate in electrical signalling, but help maintain the signalling abilities of neurons. Their role is to maintain homeostasis, form myelin, and provide support and protection for neurons.34 Several types of glial cells can be identified: astrocytes, oligodendrocytes and ependymal cells. Glial tumours can be classified according to the type of cells from which they arise: 17.

(18) . Astrocytomas Astrocytoma is a tumour that arises from the star-shaped cells (astrocytes) that form the supportive tissue of the brain. Astrocytomas are a main category of the gliomas. A resection of astrocytomas often results in their recurrence in more malignant forms. o Multiform glioblastoma (Glioblastoma multiforme) – is the most common primary brain tumour affecting adults. It accounts for about 50% cases among malignant glial tumours. When growing, it infiltrates the neighbouring tissues, which makes difficult its complete surgical resection. Intense angiogenesis provides numerous blood vessels which can cause cerebral haemorrhage. This type of tumour is also characterised by fast growth and intense migration. What singles it out from other glial tumours is the possibility of developing metastasis within the central nervous system through the blood system or cerebrospinal fluid. It represents the highest possible grade of malignance and carries the worst prognosis. It occurs in adulthood with peak incidence between 65 and 75 years of age. Research data show that primary GBM occurs more often in men than in women – probably due to the protective effect of estrogens.35 o Anaplastic astrocytoma (Astrocytoma anaplasticum) – is a locally aggressive form of astrocytoma. It also tends to infiltrate normal tissue. This type of tumour is considered malignant. Unlike glioblastomas, it does not undergo necrosis or vascular proliferation. It occurs in adulthood with peak incidence between 40 and 50 years of age. o Diffuse astrocytoma (Astrocytoma diffusum) – is also known as low grade infiltrative astrocytoma. Although it seems to be well delimited in imaging, this type of tumour has no identifiable border between neoplastic and the normal brain tissue. It tends to occur in children or young adults. o Fibryllary astrocytoma – is a subtype of diffuse astrocytoma rich in neuroglial fibrils. It is considered low-malignant. Its growth is not fast, thus it does not carry very bad prognosis.. 18.

(19) . Oligodendrogliomas Oligodendroglioma is a tumour that arises from the oligodendrocytes. These are cells that wrap around nerve cells and act as electrical insulation for axons conducting impulses. It grows slowly and usually does not spread into surrounding brain tissue, but has the potential to turn into more aggressive tumours. It is characterized by the presence of microcalcifications and numerous blood vessels. It tends to occur in middle-aged adults. o Anaplastic oligodendroglioma (Oligodendroglioma anaplasticum) – is an aggressive form of oligodendroglioma. It is characterized by increased proliferation and necrosis.. . Ependymomas Ependymoma is a tumour that arises from ependymal cells. These are cells lining the ventricular system of the brain and the central canal of the spinal cord. It grows slowly and infiltrates the neighbouring tissues. It tends to affect children.. MENINGIOMAS – tumours arising from meninges, which are membranes that cover the brain and surround the central nervous system. Meningiomas usually grow slowly and are benign in nature; they can rarely be malignant. As they grow slowly, the brain adapts to their increasing dimensions, which delays the symptoms. Meningioma is the most common benign brain tumour in adults. . Meningothelial meningioma (Meningioma meningotheliale) – is probably the most common type of Meningioma. It is considered benign and carries a very good prognosis.. . Atypical mieningioma (Meningioma atypicum) – is neither malignant nor benign, but it may become malignant. It tends to recur and grows faster than ordinary Meningioma.. . Anaplastic meningioma (Meningioma anaplasticum) – is an aggressive type of Meningioma that tends to invade surrounding brain tissue.. In Table 2.1, 5-year survival rates are given for some of the tumours presented above.36. 19.

(20) Table 2.1 Survival rates for selected adult brain and spinal cord tumours by age. 5-Year Relative Survival Rate (by age) Type of Tumour. 20-44. 45-54. 55-64. Glioblastoma. 17%. 6%. 4%. Anaplastic astrocytoma. 49%. 29%. 10%. Diffuse astrocytoma. 65%. 43%. 21%. Anaplastic oligodendroglioma. 67%. 55%. 38%. Oligodendroglioma. 85%. 79%. 64%. Ependymoma. 91%. 86%. 85%. Meningioma. 92%. 77%. 67%. More functional from the viewpoint of prognosis is brain tumour classification according to of their malignancy grades. This classification gives information on further pathological development of a tumour. Tumours with a low malignancy grade are built of mature cells; they are well differentiated and show great similarity to normal tissue, whereas tumours with a high malignancy grade grow fast, are not differentiated, have tight packing of cells in neoplastic arrangement and bear no similarity to normal tissue. They tend to infiltrate the surrounding tissues and are characterised by excessive proliferation of blood vessels and the presence of numerous necrosis centres. Their malignancy grade is assessed on the basis of the histopathological image of the tumour section. Additional information can be provided by immunohistochemical studies. Brain tumours have been classified by the World Health Organization.37 The most common types brain tumour types and their malignancy grades are specified in Table 2.2.. Table 2.2 Classification of brain tumours based on of their malignancy grades according to the WHO scale. Malignancy grade. Tumour type. I. Meningioma. II. Diffuse astrocytoma, Oligodendroglioma, Atypisal meningioma. III. Anaplastic astrocytoma, Anaplastic oligodendroglioma, Anaplastic meningioma. IV. Glioblastoma multiforme. 20.

(21) 2.2 B IOLOGICALLY. SIGNIFIC ANT CHEMICAL ELEMENT S. 2.2.1 I R ON Iron is an element very often mentioned in the context of brain diseases, both those involving neurodegradation and cancers. It plays an essential role in normal neurological function of the brain, but on the other hand carries risks associated with the production of free radicals. This is why changes in iron concentration occurring in pathological processes can provide important information. Iron plays a very important role in oxygen transport as it is an essential component of haemoglobin. It also takes part in cell breathing as is a component of cytochromes, cytochrome oxidase and the iron-sulphur complexes of the oxidative chain. For this reason it plays an important role in ATP production, which is very important for a very energyconsuming organ such as brain. Iron is also a cofactor of enzymes involved in the synthesis of neurotransmitters. It also takes part in DNA replication, because ribonucleoside reductase is also an Fe dependent enzyme. Finally, iron plays a very important role in the synthesis and metabolism of myelin, by taking part in biosynthesis of lipids and cholesterol, which are main substrates for this process.38 On the other hand, iron plays a considerable role in the creation of free radicals. 39– 41. It controls the formation of reactive oxygen species (ROS) via Fenton’s reaction: 2.1 2.2. Iron can therefore participate in controlling the level of reactive oxygen species, which are a major source of lipid peroxidation and damage to DNA and many other parts of the cell.42,43 The brain is one of the organs most vulnerable to reactive oxygen species because, although it constitutes only 2% of body weight, it utilizes 20% of oxygen consumed by the body.44 On the one hand, a deficiency of iron causes significant changes in the functioning of many proteins, but its excess, on the other hand, causes oxidative stress. For this reason, iron concentration is controlled by the body, which – depending on the current needs – assimilates it from the food, or releases. A considerable role in controlling iron concentration plays the blood-brain barrier. In brain tissues, iron can be stored in an. 21.

(22) intracellular pool of iron in ferritin, which is the main iron storage protein. It can also be transported by some iron chaperon proteins or can be stored by transferrin. Brain iron concentration increases with age. It is also reported to increase in many neurodegenerative diseases such as Alzheimer’s disease, Parkinson’s disease and Friedreich’s disease. 2.2.2 C O P P E R Copper plays a considerable role in proper functioning of brain via its presence in many cuproenzymes. It is primarily involved in redox reactions, where it acts as electron receptor. Copper is an important component of cytochrome c oxidase, which is part of the mitochondrial electron transport chain. This enzyme catalyzes the reduction of molecular oxygen to water. Cytochrome c oxidase is a major cuproenzyme in the brain and contains ca. 20% of the total brain copper.45 Another important cuproenzyme is CuZn superoxide dismutase (SOD1). It takes part in the dismutation of the superoxide radical. : 2.3. This enzyme contributes to a decrease in the oxidative stress. It is necessary for all oxygenmetabolizing cells. Decreased amounts of SOD1 were found in many tumours.46 This is also a major cuproenzyme of the brain and accounts for ca. 25% of the total brain copper.45 Copper can also be found in reactive centres of ceruloplasmin which carries more than 95% of the total copper found in human plasma. 47 It also has an antioxidant function as it oxidizes iron from Fe2+ to Fe3+ thus preventing Fe3+ from taking part in Fenton’s reaction.48 Copper is also present in many other cuproenyzmes such as dopamine -monooxygenase which takes part in the formation of norepinephiryne from dopamine and lysyl oxidase which is important for collagen synthesis.49 On the other hand, copper – like iron – can take part in a Fenton-like reaction:50,51,52 2.4 Free radicals produced in this reaction can also cause lipid and DNA damage. 53,54 Copper can also take part in promoting angiogenesis. This favours tumour growth, because blood vessels provide the tumour with nutrients and facilitate metastasis.55 Copper is particularly important in the phase of brain development and its deficiency can lead to pathological changes in the central nervous system. Its 22.

(23) concentration increases from the birth and adult concentration, which is ca. 24 µg/g in dry tissue, is attained at ca. 11 years of age and is higher in gray matter than in white matter. A decrease in copper concentration is noted in such diseases as, for instance, the Menkes disease, whereas in Wilson’s disease the concentration of copper increases and its excess is toxic.56,49 In many neurodegradative diseases, such as Alzheimer’s disease or Parkinson’s disease, also changes the level of some cuproenzymes, e.g., ceruloplazmine.57,58 2.2.3 Z I N C Zinc is the second most abundant transition metal in the body. Its highest concentrations can be found in the brain.59 An important role of zinc is its participation in synaptic neurotransmission through its presence in synaptic vesicles. About 90% of zinc is bound in zincoproteins. Zinc binds mostly to nitrogen and sulphur.60 Zinc may play catalytic, structural or regulatory role. Zinc is the only metal that is a component of all classes of enzymes: oxidoreductases, ligases, lyases, isomerases, transferases and hydrolases.56 As a component of enzymes taking part in transcription and replication of DNA and synthesis of proteins, it strongly influences the division and differentiation of cells. In cells it occurs in the form of zinc finger proteins, where it forms a complex with cysteine and histidine residues.61 A significant portion of zinc is contained in w membrane-bound proteins, where it causes changes in the conformation of protein structure necessary for their proper functioning. Zinc can be also found in ionic pool in the cytoplasm. Zinc takes part in many processes important for oncogenesis. Like copper, it is present in the active centre of SOD1 which catalyses superoxide radical dismutation (equation 2.4). On the other hand, zinc ions are considered to be toxic to mitochondira, which can result in disturbances to/malfunctioning of the breathing cycle.62 An excess of zinc can cause apoptosis of neural cells, but on the other hand, its deficiency can lead to cell death. 63 Zinc is thus an element whose proper balance in the organism is vital for its good functioning. The average zinc concentration in the brain is about 10 µg/g.56 Zinc is supplied to the brain by blood-brain and blood-cerebrospinal barriers. It was reported that brain zinc level is usually not affected by dietary zinc, in contrast to other organs, such as e.g., liver.64 Its concentration in the brain is not that sensitive to its temporary deficiency in the diet, but a prolonged deficiency can disturb human growth and development 65 and cause brain dysfunctions, e.g., mental disorders.66 Zinc deficiency in the hipokampal structures can 23.

(24) impair learning ability.67 Studies have also revealed changes in zinc concentration in patients with Alzheimer’s disease. It is believed that β-amyloid forms insoluble aggregates (which are characteristic of this disease) in the presence of zinc.68,69 In turn, developing Amyotrophic Lateral Sclerosis is connected in 2-3% cases with a zinc-containing mutation of SOD1.70 2.2.4 C A L CI U M Calcium is the most abundant mineral in the human body and fulfils many functions. It is a component of bones, teeth and nails; it controls pH and blood clotting capability and is vital for the proper performance of the heart muscle. Calcium is also a very important element for the proper functioning of the nervous system as it is crucial in conducting nerve signals.71 At rest, its concentration outside the neuron cell membrane is higher than inside the cell by several orders of magnitude. This concentration gradient is maintained by ion pumps. Nerve signals are transmitted through local depolarisation of membranes, which is made possible by voltage-dependent Ca channels in the membrane.72 Calcium takes also part in releasing neurotransmitters in nerve endings. Calcium is also a significant component of the extracellular fluid. In cells, more than 99% of Ca is bound to molecules in cytoplasm or in membrane-bound organelles.73 The remaining portion of free Ca2+ ions is used for intracellular signalling and control, e.g. in cell death mechanisms.74 A decrease in calcium concentration may induce spontaneous excitation of neurons, whereas an increase – their reduced excitability.75 Also, it is believed to play a role in learning by synaptic modulation.76 Calcium homeostasis was found to be impaired in Alzheimer’s disease, which may be one of the causes of memory loss. 77 In some disorders, such as brain cancer or Fahr’s disease, an excess of calcium can be accumulated in brain tissue in the form of calcium salts.78 Such deposits are called calcifications. 2.2.5 P HO SP H OR U S Phosphorus is the second most abundant mineral in the body. Like calcium, it is an important component of bones and teeth. But its most important role is in storing energy. As a component of ATP, it plays a role in energy production. Energy – needed for all processes occurring in living organisms – is released in the hydrolysis of ATP.79 Phosphorus is also an important component of phospholipids of which cell membranes are built. The most important role of cell membranes is establishing the permeability barriers for cells 24.

(25) and cell organelles.80 Phosphorus is also present in DNA and RNA. With the pentose sugar it builds the structure of these molecules. Genetic information is stored in the cell nucleus. It was shown that cancer cells are characterized by a higher nucleus-cytoplasm ratio.81,82 Also, an increased number of mitoses are characteristic of malignant cells.82 2.2.6 S U LP HU R Sulphur is the fourth most abundant mineral in the body. It is most often mentioned in the context of creatine, which is present, among other structures, in hair and collagen strands. But it also has other important functions: it is a component of two amino acids, methionine and cysteine.83 Another important sulphur compound is glutathione which is involved in protection against reactive oxygen species. A particular role among brain cells have astroglial cells which supply glutathione precursors to neighbouring cells.84 Sulphur is also present in disulfide bonds which play an important role in the folding and stability of proteins. These bonds are important for proper biological function of proteins. Also, disulfide bonds protect many enzymes and hormones present in extracellular environments from denaturation by improving their thermodynamic stability.85 This is why these disulfide bonds, present in many cancer-related proteins, can be a target for new cancer therapies.86 2.2.7 P OT A SS I U M ,. S OD I U M , C HL OR I NE. Potassium, sodium and chlorine are electrolytes - substances responsible for conducting electricity in the body. They make possible the conduction of electrical signals in neurons. Potassium, sodium and chlorine are all present in both extracellular and intracellular fluids, but Na+ and Cl- ions are present in the extracellular fluid in much higher concentration than in the intracellular fluid, and K+ is present in higher concentration in the intracellular fluid. A concentration gradient of these elements is maintained by, e.g., the Na+/K+-ATPase pump, whereas highly selective ion channels make possible the flow of ions through cell membranes and are crucial for hyperpolarisation. Sodium and potassium also work to maintain normal water balance in the body and thus blood volume and pressure.. 25.

(26) 3. T HEORETICAL BACKGROUN D OF THE RESEARCH ME THODS 3.1 S YNCHROTRON. RAD IATION. Synchrotron radiation is the electromagnetic radiation emitted by charged particles moving at near-light speed. The power of the radiation released as a result of particle acceleration is given by the formula for radial acceleration of a particle: (. ). 3.1. where: E – particle total energy R – particle trajectory curvature m0 – rest mass of the accelerated particle c – speed of light The released radiation is tangent to the particle’s trajectory and its angular distribution depends on the particle velocity (Figure 3.1). At velocities small relative to the speed of light, the angular distribution of the emitted light is close to that generated by a vibrating electric dipole. For relativistic velocities, the angular distribution of the radiation has a shape similar to a cone, with an angle of: 3.2 where γ is the Lorentz factor equal to: 3.3 For this reason, to produce highly collimated radiation, the particle used to generate it must have the lowest possible mass: most commonly these are electrons.. Figure 3.1 Angular distribution of emitted radiation generated by a non -relativistic (left) and a relativistic (right) particle.. 26.

(27) Synchrotron radiation was discovered by chance, when carrying out experiments in the field of particle physics. To observe the interaction of particles, they were accelerated, but losses limited the maximum energy attainable by particles. This is why initially this radiation was considered parasitic, but thanks to its properties, synchrotron radiation has become one of the most important tools among today’s scientific research methods.87 So far, four generations of synchrotron radiation sources can be identified. The first one includes synchrotrons, whose construction is based on known particle accelerators; the second one includes dedicated equipment: built for the purpose of generating synchrotron radiation. What differs the third generation from the second one is the presence of insertion devices whose role is to increase the brilliance of the emitted radiation. A new class of equipment that generates light brighter by several orders of magnitude is free-electron lasers (Figure 3.2).. 88. Figure 3.2 Peak brilliance of x-ray sources. (after: PSI FEL ). Figure 3.2 shows how important a step in the history of development of synchrotron radiation was the introduction of insertion devices. They consist of an array of magnets producing magnetic fields of alternating polarity, whereby the trajectory of a charged molecule passing through such a structure is repeatedly bent. At each curvature the particle emits radiation. In undulators, constructive interference of the radiation takes 27.

(28) place, resulting in a large increase in the brightness of selected energies of the radiation. The action of free-electron lasers is based on the SASE principle (self-amplified spontaneous emission) involving mutual interaction between the emitted radiation and electrons (Figure 3.3).. Figure 3.3 Comparison of the brilliance and beam geometry of radiation produced by a bending magnet, wiggler, undulator and that produced in a free -electron laser.. The widespread use of synchrotron radiation in research conducted in many fields of physics results from numerous properties that distinguish it from other photon sources. These properties include the wide range of the radiation spectrum, high brilliance, strong beam collimation, linear polarisation in the accumulation ring plane and circular polarisation outside the plane. Radiation sources producing such small beam sizes and such high brilliance enable studies of very small biological structures. Without synchrotrons, it would not be possible to carry out research presented in this work. The enormous significance of synchrotrons in the development of modern science is made evident by the fact that almost 90% of all the protein crystal structures submitted to the global Protein Data Bank have been deciphered with the use of synchrotron radiation.87. 28.

(29) 3.2 X- RAY. FLUORESCENCE. X-ray Fluorescence is a research method in which the energy of fluorescence radiation emitted by a sample is used to determine its elemental composition. Radiation incident on the sample is absorbed by it in accordance with the Lambert-Beer law. The radiation beam is attenuated by the interaction of photons with matter. In the range of energies used in XRF, absorption is associated with the occurrence of the photoelectric effect, incoherent (Compton) scattering and coherent (Rayleigh) scattering: 3.4 The XRF method is based on the first of the phenomena mentioned above. Radiation incident on a sample ejects a photoelectron from its shell. Consequently, in order to study elements we must use energies sufficient for their ionisation. For X-rays, the greatest cross-section for absorption is for electron shells closest to the atom nucleus. Ionisation results in the formation of a vacancy in the shell left by the electron. When the atom returns to its ground state, the vacancy is filled by an electron from a shell with higher energy. This is accompanied by emission of a quantum of energy equal to the difference in electron energies on the higher shell and the one with the vacancy: 3.5 where: ν – frequency of the emitted photon E2 – binding energy of the electron from the higher shell E1 – binding energy of the electron from the higher shell where the vacancy existed Not all electron transitions between any orbitals are allowed: the transitions must follow quantum selection rules. Analysis focuses mainly on dipole transitions as the corresponding spectra lines are more intense than for quadrupole transitions. For dipole transitions, the selection rules take the form: 3.6 |. |. 3.7. |. |. 3.8. The names of the spectral lines are derived from the shell in which the vacancy existed (K, L, M...) and from the difference Δn of the main quantum numbers between the shells. 29.

(30) between which the transition takes place (α,β...). A diagram of transitions between subshells and their spectral names is presented in Figure 3.4.. Figure 3.4 Electron transitions and characteristic radiation lines.. Electron binding energies on individual shells are characteristic of every element: consequently, the energies of spectral lines are precisely defined. This makes it possible to precisely identify each element based on the fluorescence radiation it emits. The X-ray Fluorescence method is based on the fact that the intensity of the fluorescence radiation from a given element is in direct proportion to its concentration in the sample: ( ). (. ) (. ) (. ( (. ) (. ). ). (. (. ). ). ). where: (. ). (. ). G – geometry factor ( ) – intrinsic detector efficiency for recording a photon of energy (. ) – number of incident photons of energy 30. per second per steradian. 3.9.

(31) and. – effective incidence and takeoff angles. ρ – density of specimen in g/cm3 d – sample thickness in cm (. ) and ( ) – total mass attenuation coefficients in cm2/g at energies. and. weight fraction of element i (. ) – total photoelectric mass absorption coefficient for element i at the energy. in. cm 2 /g – fluorescence yield of element i – transition probability of the kth line of element i - absorption jump at the K-edge of photoelectric absorption in element i. The mass attenuation coefficient in this formula depends on the proportion of individual elements in the sample: ∑. 3.10. where: – mass attenuation coefficient of element i in cm2/g – weight fraction of element i in % n – total number of elements in the absorber. For thin samples where total mass attenuation coefficient (µ in equation 3.10) is small and the mass per unit area (M=ρd) of a studied sample is small, equation 3.9 can be simplified to: ( ). ( ) (. ) (. ). 3.11. It can be noted that the intensity of fluorescence radiation depends linearly on the concentration of the studied element. As all the other factors on the right-hand side of formula 3.11 do not depend on this concentration, it is possible to assess the concentrations of elements by making a comparison with a standard measured in the same geometry. A very important phenomenon to be taken into account when evaluating the measurement data is the matrix effect which involves attenuation of the fluorescence lines originating from the studied element through their absorption by other elements 31.

(32) present in the sample. In addition, in analysing the data it may be useful to assess the peaks of the Compton scattering (incoherent) and Rayleigh scattering (coherent), for which the mass coefficients are given by the formulae: ∑. 3.12. ∑. 3.13. where: k, k’ – constants depending on the geometry of measurement and energy of the radiation Zi, Ai – atomic and mass numbers of element i wi – weight fraction of element i, in % We can thus conclude about the effective atomic number of the sample on the basis of the ratio of the intensities of these two peaks. In X-ray Fluorescence Spectroscopy, atoms can be excited by X-rays from an X-ray tube, from radioisotopes, or that generated by a synchrotron.. 3.3 I ON. BEAM ANALYSIS. Atoms can be excited not only by X-rays, as shown in the previous sub-section, but also by charged particles. This measurement method is known as Particle Induced X-ray Emission (PIXE). When atoms are excited by particles, the effectiveness of excitation decreases with an increase in the atomic number in the range including the elements studied in this work, whereas it increases for photon excitation. For this reason the XRF method is better suited for studying heavier elements, e.g., Fe, Cu and Zn, whereas the PIXE method gives better results for light elements. Various particles can be used for excitation, but the most commonly used are protons, because of their high cross section for inner shell ionisation. As a result of excitation, as in the XRF method, characteristic fluorescent radiation is emitted, whose total K-shell yield is: ( ). (. ). where: Np – number of protons hitting the sample Ai, wi – atomic mass and concentration of element i SM – stopping power of matrix σi – K-shell ionization cross section of element i 32. ∫. ( ) ( ) ( ). 3.14.

(33) ωKi, bKi – fluorescence yield and line intensity fraction of element i on K-shell Ti – transmission factor through the sample of element i tKi – transmission through absorbers interposed between specimen and detector εi – detector intrinsic efficiency Thus obtained spectra differ from those obtained using the XRF method by mostly the shape of the background. Particles can be accelerated by linear accelerators, e.g., a Van de Graaf accelerators or tandem accelerators or by cyclotrons. Very often PIXE measurements are combined with the registration of back scattered particles, the technique known as Rutherford Back Scattering (RBS), which provides information on the matrix composition. Registration of the energy of particles after passing through the sample, which is performed using Scanning Transmission Ion Microscopy (STIM), provides information on the surface density.. 3.4 X-R AY A BSORPTION S PECTROSCOPY Radiation incident on the sample can be absorbed, scattered or can pass through it without interacting. The radiation absorption coefficient depends, among others, on the energy of the radiation and the composition of the sample: 3.15 where: Z – atomic number A – mass number E – incident radiation energy The absorption coefficient can be measured in two ways: in the transmission mode or in the fluorescence mode. In samples containing high concentrations of the studied element, radiation measurements are performed in the transmission mode, that is after passing through the sample. The absorption coefficient is then calculated from formula 3.16. In samples with lower concentrations of the studied element, of the order of several ppm, measurements are performed in the fluorescence mode. In this case the absorption coefficient is described by formula 3.17. During measurement, the absorption coefficient is registered as a function of energy.. 33.

(34) 3.16 ( ). 3.17. where: I0 – the intensity of the radiation incident on the sample I - the intensity of the radiation transmitted by the sample If – the intensity of the fluorescent radiation emitted by the sample atoms. If the energy of the incident radiation is equal to the binding energy of an electron in the shell, an abrupt increase in absorption is observed. This is seen in the graph showing the absorption coefficient as a function of radiation energy (Figure 3.5) as an absorption edge. Every element can have several absorption edges corresponding to the transfers of electrons from successive shells (K,L,M, ...). The energies at which absorption edges occur are characteristic of each element and can be used for its identification. X-ray Absorption Spectroscopy involves studying the positions and shapes of the absorption edges and studying absorption coefficient oscillations immediately after the edge.. Figure 3.5 Mass absorption coefficients for photoelectric absorption of Fe, Zn and Au .. Two ranges can be distinguished in a XAS spectrum: the X-ray Absorption Near Edge Structure (XANES) and Extended X-ray Absorption Fine Structure (EXAFS). It is assumed that the XANES range extends from about 10 eV below the edge to 50 eV above the edge, whereas still further, lies the EXAFS range, up to energies of about 1000 eV. This is, 34.

(35) however, an arbitrary limiting value as the physical basis of both phenomena is the same. Considering, however, that in the XANES range, at low energies of the ejected electron, non-elastic scattering is predominant, whereas in the EXAFS range elastic scattering is predominant, various approximation are used to simplify the mathematical description of the phenomena. The XANES range is particularly sensitive to the oxidation state and the spatial structure of the studied compound. A typical XANES spectrum is shown in Figure 3.6 whereas Figure 3.7 presents a diagram of electron transitions between shells responsible for individual components of a XANES spectrum.. Figure 3.6 Elements of a XANES spectrum: pre-edge peak, absorption edge and scattering peaks.. Figure 3.7 Electron transfers between orbitals in individual parts of the XANES spectrum.. 35.

(36) The pre-edge peak is associated with the transfer of an electron from an inner shell of the atom absorbing radiation to an empty molecular state. The presence of the pre-edge peak can provide information about the geometric structure of the studied compound: it occurs for system with tetrahedral symmetry, e.g., Na2CrO4 (Figure 3.8: purple), whereas it does not occur for octahedral symmetry, e.g., Cr2O3 (Figure 3.8: pink). The positions of the preedge peak on the energy axis can be modified by the oxidation state of the studied atom. The absorption edge corresponds to electron transitions from an inner shell of the atom absorbing radiation to empty states above the Fermi level, that is the highest occupied state. The height of the absorption jump at the edge is in direct proportion to the concentration of the corresponding element in the sample. We can conclude about the element’s oxidation state from the absorption edge position on the energy axis (Figure 3.8). The higher the element’s oxidation state, the less electrons are on its shell and the weaker is the screening of the charge.. Figure 3.8 Spectrum energy shift for chromium compounds in various oxidation states.. EXAFS analysis is focused on oscillations of the absorption coefficient above the absorption edge. It can provide information about the number and kind of atoms directly surrounding the atom absorbing a photon, and about the distances at which these atoms are positioned. If the energy of the incident photon is greater than the electron binding energy, the kinetic energy of the emitted photoelectron is equal to the difference between the photon’s energy and the electron’s binding energy. A de Broglie wave of a. 36.

(37) specific length is associated with the photoelectron, so the electron’s wavelength is related to its energy by: 3.18 where: k – the wave number m – electron mass E – exciting electron energy E0 – electron binding energy of the atom A photoelectron wave propagates towards neighbouring atoms and is scattered by them (Figure 3.9). The scattered waves interfere with the original wave, which results in constructive or destructive interference, depending on the distance from the neighbouring atoms. With increasing energy, the wavelength decreases, which causes changes in the positions of maxima and minima of the interference wave. This causes oscillations of the absorption coefficient as a function of energy.. Figure 3.9 Photoelectron scattering on the neighbouring atoms. The primary wave (gray) and scattered waves (blue).. The waves scattered by atoms surrounding the excited atoms, which are positioned in subsequent coordination shells (groups of atoms positioned at the same average distance Rj) are summed up. After introducing corrections modifying the amplitude of the oscillations, we obtain the EXAFS equation: ( ). ∑. ( ). ( ). where: j – the number of subsequent co-ordination shell 37. (. ( )). 3.19.

(38) – the distance of atoms of the jth coordination shell from the central atom – the number of scattering atoms in the jth coordination shell – the amplitude of back scattering ( ). – factor accounting for the limited lifetime of the system: excited atom –. photoelectron. It is associated with the limited lifetime of a hole in the central atom’s shell and the limited lifetime of the ejected electron (λ(k) – the average path of the scattered electron) – the factor accounting for the deviation of the distance of individual atoms of the jth coordination shell from the average distance Rj from the central atom and thermal oscillations of atoms. ( (. - Debye-Waller factor). ( )) – interference factor (δj(k) – phase shift). To determine such structural parameters as Rj or Nj, the EXAFS function resulting from theoretical considerations is compared to that calculated on the basis of a normalised measured spectrum (Figure 3.10) according to formula: ( ). ( ). ( ) ( ). where: μ(E) – the measured value of the absorption coefficient at energy E μ0(E) – the absorption coefficient for an isolated central atom. 38. 3.20.

(39) a.. b. Figure 3.10 An EXAF function (b) calculated on the basis of a normalised absorption spectrum (a).. An advantage of XAS over many other research methods is that the sample does not need to have an orderly structure. The method can be used for studying amorphous substances, glass, quasi-crystals, membranes, solutions, liquids, metalloproteins, multi-atom gases, etc. The method also enables tracing the course of chemical reactions.. 39.

(40) 4. E XPERIMENTAL 4.1 S AMPLE. PREPARATION. All the samples were obtained from the material collected during resection of brain tumours for diagnostic purposes. The samples came from the Department of Neuropathology, Chair of Pathomorphology, Faculty of Medicine, Jagiellonian University in Krakow, Poland. The study was approved by the Jagiellonian University Bioethical Committee (KBET/101/B/2010). Two types of samples were prepared: thin ones, used for analysis at selected points, which were freeze-dried, and thick ones, used for bulk analysis, which were frozen to slow down chemical processes e.g. oxidation. The tissues studied in this work represented various types of brain tumours whose malignancy grades were determined in accordance with the World Health Organisation (WHO) classification.37 These included glioblastoma multiforme (WHO IV), anaplastic astrocytoma (WHO III), anaplastic oligodendroglioma, (WHO III), diffuse astrocytoma (WHO II), oligodendroglioma (WHO II), atypical meningioma (WHO II) and meningothelial meningioma (WHO I). A benign tumour, containing numerous calcifications, was also analysed. As controls for the purpose of studying the concentration of elements, we used samples taken from patients who died of causes other than cancer: in each of these cases the cause of death was the rupture of aneurism. As a control sample in the study of the chemical state of elements, a brain abscess wall was used, which is non-cancerous brain tissue. 4.1.1 F R E E Z E - D R I E D. SA M P LE S. The tissues were cryo-sectioned and specimens were prepared for both histological examination and elemental analysis. One section of each sample was cut to a thickness of 5 μm and stained with hematoxylin and eosin to determine the type and grade of malignancy of the tumour. The adjacent piece of tissue was cut to a thickness of 20 μm, and then placed on 4 µm thick X-ray-transparent Ultralene foil stretched on a polymer disc (Figure 4.1). Next the samples were freeze-dried. They were dried at gradually increased. 40.

(41) temperature: from -80°C to room temperature. This way of sample preparation preserves the element distribution close to its native state.. Figure 4.1 Freeze-dried sample placed on a foil stretched on a polymer ring.. The choice of sample thickness was primarily dictated by factors connected with measurement capability, that is what part of radiation is transmitted through the sample. From the calculations presented by M. Szczerbowska-Boruchowska89 it results that brain samples with a thickness of 20 µm are thin samples for the study of Ca, Fe, Cu and Zn, to mention only the elements studied here, whereas in the case of P, S, Cl and K they are samples of intermediate thickness and thus self-absorption of radiation may occur in them. On the other hand, the samples should not be too thin, as it would excessively lengthen the measurements. Consequently, 20 µm seems to be the optimal thickness, because for many of the samples it suffices to collect the spectrum for 1 s per pixel whereas for such elements as Fe, Cu and Zn, self-absorption does not yet occur. Also, in samples with thickness 20 µm, at most several cell layers can be seen in the rectangular cross-section which makes it easier to identify areas for elemental analysis. 4.1.2 F R O Z E N. SA M P LE S. As in the case of freeze-dried samples, the specimens were cut for histological examination to determine the type and grade of malignancy of the tumour studied. From the remaining tissue, samples of about 0.1 cm3 were cut and placed in specially prepared polymer measurement containers (Figure 4.2). The front of the container, which was exposed to synchrotron radiation, was covered with X-ray-transparent Ultralene film. The samples were then immediately frozen at -80˚C and stored in a deep freezer to minimize biological and chemical processes, e.g. oxidation. During transport to the synchrotron facility and between measurements, the samples were stored in a container cooled with 41.

(42) liquid nitrogen. This way of sample preparation preserves the chemical form of elements close to their native state.. 1cm. Figure 4.2 A frozen sample in a polymer container.. 4.2 M EASUREMENT. METHODOLO GY. 4.2.1 X-R A Y F LU OR E SCE N CE The XRF measurements were carried out at beamline I18 of Diamond Light Source (Didcot, UK). The synchrotron radiation was monochromatized by liquid nitrogen-cooled double crystal Si (111) monochromator. The beam was focused using a pair of Kirkpatrick-Baez mirrors, and then formed to a size of 2 x 4 μm2 using a pair of slits. The energy of the exciting radiation was set to 16.5 keV. The fluorescence measurements were performed with the detector positioned at 90˚ with respect to the beam direction and the sample surface positioned at 45˚ both to the beam and to the detector. The characteristic fluorescence radiation was detected by a 4-element SDD detector. For each pixel the spectrum was collected for 1 to 2 seconds. Normalisation to the beam current was achieved by monitoring the beam impinging on the sample using an ionisation chamber. The measurements were conducted in air at room temperature. As reference materials, certified NIST SRM 1832 and SRM 1833 standards, were used.90 Both the samples and reference materials were measured in the same geometry. The measurement set-up is shown in Figure 4.3.. 42.