Biospectroscopic analysis of human breast cancer tissue: probing infrared signatures to comprehend biochemical alterations

Journal of Biomolecular Structure and Dynamics ISSN: 0739-1102 (Print) 1538-0254 (Online) Journal homepage: http://www.tandfonline.com/loi/tbsd20 Biospectroscopic analysis of human breast cancer tissue: probing infrared signatures to comprehend biochemical alterations Ranjana Mehrotra, Gunjan Tyagi, Sonika Charak, Bhumika Ray, Geeta Kadayaprath, Harit Chaturvedi, Urmi Mukherjee & Andleeb Abrari To cite this article: Ranjana Mehrotra, Gunjan Tyagi, Sonika Charak, Bhumika Ray, Geeta Kadayaprath, Harit Chaturvedi, Urmi Mukherjee & Andleeb Abrari (2017): Biospectroscopic analysis of human breast cancer tissue: probing infrared signatures to comprehend biochemical alterations, Journal of Biomolecular Structure and Dynamics, DOI: 10.1080/07391102.2017.1298469 To link to this article: http://dx.doi.org/10.1080/07391102.2017.1298469 Accepted author version posted online: 22 Feb 2017. Submit your article to this journal View related articles View Crossmark data Full Terms & Conditions of access and use can be found at http://www.tandfonline.com/action/journalInformation?journalCode=tbsd20 Download by: [FU Berlin] Date: 24 February 2017, At: 23:27 Publisher: Taylor & Francis Journal: Journal of Biomolecular Structure and Dynamics DOI: http://dx.doi.org/10.1080/07391102.2017.1298469 LETTER TO THE EDITOR Biospectroscopic analysis of human breast cancer tissue: probing infrared signatures to comprehend biochemical alterations Ranjana Mehrotra1*, Gunjan Tyagi1, Sonika Charak1, Bhumika Ray1, Geeta Kadayaprath2, Harit Chaturvedi2, Urmi Mukherjee3, Andleeb Abrari3 1 CSIR-National Physical Laboratory, Dr. K. S. Krishnan Marg, New Delhi-110012, India 2 Department of Surgical Oncology, Max Super Speciality Hospital, Press Enclave Road, Saket, New Delhi-110017, India 3 Department of Histopathology, Max Super Speciality Hospital, Press Enclave Road, Saket, New Delhi-110017, India *To whom correspondence should be addressed Email: ranjana@nplindia.org drranjana.mehrotra@gmail.com First Author & Corresponding Author: 1 Ranjana Mehrotra* CSIR-National Physical Laboratory Dr. K. S. Krishnan Marg, New Delhi -110012, India Fax: 91-11-45609310 Phone: 91-11-45608315 Email: ranjana@nplindia.org drranjana.mehrotra@gmail.com Second Author: 1 Gunjan Tyagi CSIR-National Physical Laboratory Dr. K. S. Krishnan Marg, New Delhi -110012, India Phone: 91-11-45609470 Email: gunjantyagi18@gmail.com Third Author: 1 Sonika Charak CSIR-National Physical Laboratory Dr. K. S. Krishnan Marg, New Delhi -110012, India Phone: 91-11-45609470 1 Email: sonu.mbhu@gmail.com Fourth Author: 1 Bhumika Ray CSIR-National Physical Laboratory Dr. K. S. Krishnan Marg, New Delhi -110012, India Phone: 91-11-45609470 Email: ray.bhumika24@gmail.com Fifth Author: 2 Geeta Kadayaprath Department of Surgical Oncology Max Super Speciality Hospital Press Enclave Road, Saket, New Delhi-110017, India Email: geeta.kadayaprath@maxhealthcare.com Sixth Author: 2 Harit Chaturvedi Department of Surgical Oncology Max Super Speciality Hospital Press Enclave Road, Saket, New Delhi-110017, India Email: haritchaturvedi0210@gmail.com Seventh Author: 3 Urmi Mukherjee Department of Histopathology Max Super Speciality Hospital Press Enclave Road, Saket, New Delhi-110017, India Email: urmi.mukherjee@maxhealthcare.com Eighth Author: 3 Andleeb Abrari Department of Histopathology Max Super Speciality Hospital Press Enclave Road, Saket, New Delhi-110017, India Email: andleeb.abrari@maxhealthcare.com Running Title: Molecular fingerprints of normal and malignant breast tissue 2 1. Introduction Breast cancer (BC) is one of the most studied and leading form of malignancy in human females. Currently, studies conducted in the field of breast cancer focuses on its early detection using noninvasive or minimally invasive techniques in lieu of traditional excisional biopsy, as cancer treatment is often simpler and effective, when diagnosed at an early stage. Mammography is the first step, usually performed in diagnosing breast cancer, but at times mammogram may not be able to provide a clear picture. In addition, biopsy is performed to confirm the presence or absence of tumor, which is associated with false-positive results. Consequently, the limitations of current screening methods have shifted the focal area of oncological research in applying biospectroscopy techniques for diagnostics (Gajjar et al., 2014). Infrared (IR) and Raman spectroscopy are versatile vibrational spectroscopy methods that have been used to discriminate normal and cancer tissue and/or cell of different kinds , including endometrial cancer, cervical cancer, lung cancer, precancerous lesion and brain tumors (Gajjar et al., 2013). Coupled with some algorithms (Gajjar et al., 2013), these spectroscopic outcomes can deliver an objective, high-throughput and low cost solution to breast cancer diagnosis. Infrared spectroscopy (IR) has expanded its application in the field of human biology, since it was revealed that biological molecules present in a living tissue possess vibrational features that can be studied to derive their molecular information. Thus, the biochemical modification in a normal tissues/cell can be analyzed and compared to its malignant state (Gajjar et al., 2014). Further, several reports have highlighted its advancement in both near- and mid-infrared regions, making it an efficient and convenient 3 method for clinical purposes. From last few years, FTIR spectrophotometer has been exploited to study the molecular and structural characteristics of proteins, carbohydrates, lipids, and nucleic acids. Initially, Chirgadze & Nevskaya in 1976 studied the infrared spectral features of amide I and amide II (Chirgadze & Nevskaya, 1976). Further, in an investigation, Liquier and his colleagues (1977) demonstrated that FTIR spectroscopy could be utilized to identify the different conformations of DNA (Liquier, Taboury, Taillandier, & Brahms, 1977). In the year 2000, using infrared spectroscopic vibrations, Bouchard and his co-researchers, revealed the structure of insulin and described the formation of amyloid fibrils via insulin, which involves substantial unfolding of the native protein (Bouchard, Zurdo, Nettleton, Dobson, & Robinson, 2000). Since then many more complex studies have been conducted on proteins and nucleic acids (DNA/RNA) structures, their conformations and interactions with small ligands. The biochemical changes in a cell/tissue generally lead to nuclear, cytoplasmic and morphological variations and hence, FTIR spectroscopy could detect these alterations during the developmental stages of cancer before morphological and cytological changes are evident under light microscope. Many studies have shown that spectroscopic techniques (with different sampling modes) can differentiate the biochemistry of normal and neoplastic cells. It has been employed to investigate the carcinoma of the breast, esophagus, colon, stomach, prostate and significantly (Pu, Wang, Tang, & Alfano, 2010; Wang et al., 2003). In the present study, we report the infrared signatures of breast normal and cancer tissues that have been acquired using FTIR spectrophotometer to observe the biomolecular changes in breast malignant tissues. The outcomes of this study can be the basis of further multivariate analysis, to develop this approach as complementary screening method. 2. Methods 4 2.1 Tissue sampling: Tissue samples of around 150 cases of BC were obtained from Oncology (Cancer Care), Max Super Speciality Hospital, New Delhi. Post-surgical cancer tissue and normal tissue (2-3 cm away from the tumor) samples were collected. The tissue samples were of grade II and III. For each case two samples were cut, one was put on the glass slide and used for histological review. The other part of the tissue was frozen (-28oC) to obtain cryostat sections (2-4 µm thick). The tissue sections were taken on zinc selenide (ZnSe) crystal plates without any fixative and utilized for infrared spectral acquisition. 2.2 Spectral measurements: Fourier transform infrared (FTIR) spectrophotometer (Varian 660-IR) equipped with deuterated triglycine sulfate (DTGS) detector and KBr beam splitter was employed to record the infrared spectra (normal and cancer tissues) in the mid-IR region (4000-650 cm-1) using transmission mode and 4 cm-1 resolution. Prior to each sample spectral measurement, background spectrum was also collected to nullify atmospheric (CO2) interferences and ambient humidity (45% RH) was also maintained throughout the experiment. For each sample, five spectra were recorded with 256 scans (for each spectral acquisition) at five different positions and then averaged. Further, the raw spectra were ratioed against the background spectrum, linear baseline corrected and mean normalized using Unscrambler 6.0 software before analysis. All the spectra were smoothened using Savitzky-Golay algorithm (9 points) and second order derivative were also estimated. The representative infrared spectra have been shown in figures, where solid line and dotted line corresponds to spectral features of normal and cancer tissue respectively. 3. Results and Discussion 3.1 Analysis of spectral peak variations The infrared spectra of normal and cancer breast tissue from different patients were recorded in the mid-IR spectral region 4000-650 cm-1. The spectral assignments were in concordance to the literature (Wang et al., 2003). As vibrations accredited to functional groups of various 5 biomolecules indicate distinct spectral features in breast normal and cancer tissues in all the cases studied. Therefore, this region is of particular interest in probing biochemical changes occurring in a cell during diseased state. Figure 1 shows the overlaid IR spectra of normal and cancer tissue in 1300-700 cm-1 region. This region harbors infrared vibrations related to nucleic acid content of a cell. The infrared spectrum of normal tissue demonstrates strong bands in this region, while peaks of lower intensity with a shift in position are apparent in the infrared spectrum of malignant tissue. The band at 1083 cm-1, arises due to the symmetric stretching vibrations of phosphate groups (mainly contributed by nucleic acids) (Banyay, Sarkar, & Gräslund, 2003). This band shows a shift of 3 cm-1 (to 1080 cm-1) in position along with decrease in intensity in the spectrum of malignant tissue. In addition, an infrared vibration at 1243 cm-1, attributed to anti-symmetric phosphate stretching in normal tissue (of breast), appears as less intense band with a shift of 6 cm-1 (to 1249 cm-1) in the spectrum of malignant tissue. These bands mainly originate in the phosphodiester backbone in cellular nucleic acids and similar pattern showing higher frequency for phosphate groups vibrations in normal tissue than malignant tissue, have been observed in earlier studies attempting to characterize cancer tissues (Cohenford et al., 2012). In terms of qualitatively identifying the type of DNA impairment; disturbance in DNA phosphodiester backbone as an indicator of DNA damage in malignant tissue can be attributed to the role of ROS (reactive oxygen species) in cancer progression. Notably, oxidation leads to changes in the phosphodiester backbone of duplex that are identified by alterations in the vibrational frequencies in the molecule (Malins et al., 1995). Therefore, the present alterations in the bands associated with DNA backbone might have resulted from the ROS induced damage and subsequent DNA strand breaks. This feature is a characteristic of malignant phenotype and similar observations have also been reported in studies characterizing structural disorders in cancer DNA (Kondepati et al., 2008). The spectral 6 changes associated with symmetric and anti-symmetric stretching vibrations of phosphate bands are further evident in the second order derivative spectra as illustrated in Figure 2. Besides this, a significant spectral signature is observed at 722 cm-1 in the spectrum of normal breast tissue, whereas, this band loses its intensity in malignant tissue that could be allied to the grade of malignancy. It has been reported that this band either appears as of very low intensity or gets diminished with higher grade of malignancy. This is also consistent with our results, where around 90% of the studied sample followed the same. Biochemically, the 722 cm-1 vibration is attributed to rocking vibrations of carbonyl (C=O) group present in cellular NADPH (Nicotinamide adenine dinucleotide phosphate) (Pu et al., 2010). Cancer cells are usually devoid of energy due to their exhaustive multiplication, the loss of band (or its intensity) related to this energy molecule can be employed as a marker of a malignant condition. Moreover, certain reports are also available, which testifies that cancer cells have less amount of NADH and NADPH in cancerous conditions, are being treated with extracellular supply of NADH molecule (Pu et al., 2010). The infrared region from 1700-1500 cm-1, as illustrated in Figure 3, is comprised of vibrations owing to functional groups of amide I, amide II and amide III of proteins. In the normal tissue, the band at 1654 cm-1 (amide I) arises primarily due to C=O stretching vibrations of amide group. While, the band at 1596 cm-1 is assigned to amide N-H bending vibrations of protein backbone (Banyay et al., 2003). Both of these bands show a shift of 3 cm-1 in their position along with major intensity variation in the infrared spectrum of malignant tissue. Furthermore, the shifts in band position and changes in intensity can be more apparent in the second order derivative spectra (Figure 4). As proteins, the building blocks, are associated with several cellular processes (Yu, Zhou, Conrads, & Veenstra, 2004), these visible disturbances derived from vibrational features of the protein signifies exhausted demands of energy in a cell during malignant state. Moreover, Figure 5 represents the 7 overlaid infrared spectra of normal and malignant breast tissues in the region 3100-2700 cm-1, which also consists of bimolecular signatures of interest. This region mainly includes stretching vibrations of phospholipids, cholesterol and creatine. The infrared vibrations at 2855 cm-1 and 2925 cm-1 present in this region are attributed to the stretching vibrations of CH2 and CH3 acyl group of lipids respectively. As compared to normal tissue, these bands show a decrease in intensity with negligible shift in wavenumber in the spectrum of malignant breast tissue, which is more visible in the second order derivative spectra (Figure 6) as pathological state of cell/tissue can be evaluated by considering the variations in band intensities as well. Additionally, an infrared peak at 3008 cm-1 due to =C-H groups (related to olefins or unsaturated fatty acids), observed in breast normal tissues, gets diminished in malignant tissues (Banyay et al., 2003). From the intensity variations found in this region, it can be inferred that lipid content is greatly reduced in malignant cells and there is a severe alteration in the distribution of lipids and proteins in diseased condition. Similar outcomes have been reported by Colagar and his colleagues, where with the advent of Fourier transform infrared microspectroscopy, malignant human gastric tissues is differentiated from its normal counterpart (Colagar, Chaichi, & Khadjvand, 2011). 3.2 Analysis of different wavenumber ratios The pathological state of tissues could be analyzed by observing the changes in band intensities in certain key regions. We have identified certain wavenumber ratios, whose variations could be related to the malignancy. Amide I/Amide II (1654 cm-1/1539 cm-1) ratio derived from the N-H stretching vibrations of amide bonds was found to be increased in case of malignant breast tissue (Mehrotra, Gupta, Kaushik, Prakash, & Kandpal, 2007). This also implies a change in the intensity of vibrations originating from the alpha and beta protein structure. This could be attributed to variation in the secondary structure of proteins due to malignant transformation. Lipid/Protein (2850 cm-1/1654 cm-1) ratio in case of malignant 8 tissue was found to be less when compared to normal. Various other studies have also pointed towards the effectiveness of lipid-to-protein ratio as a spectroscopic marker to discriminate between normal and malignant tissue as well as between low- and high-grade cancer (Beljebbar & Manfait, 2011). It has been reported that this ratio is found to be large for normal tissue and decreases with the progression of the disease. This lipid reduction in malignant tissues could be related to the fast growth of tumor cells, which require more energy. Indeed, it is known that structural and functional cell changes take place in developing tumors where lipids play a crucial role. Yet, various aspects of lipid changes in tumors of different degree of malignancy are still the subjects of research. Moreover, decrease of protein to lipid ratio as well as increase of amide I/amide II ratio have also been observed for certain other pathological conditions other than cancer. The absorbance ratio 1654 cm-1/1230 cm-1 yields information on the relative content of collagen and is observed to increase with the increased grade of the malignancy. Since, a cancerous cell is known to have a disturbed extracellular matrix (ECM), increase in the content of collagen, which is an important part of ECM, is widely associated with the cellular malignancy grade. The relative intensity ratio of the bands at 2855 cm-1 and 2925 cm-1 due to CH2 and CH3 stretching vibrations respectively has increased along with the progression of the disease. The variation in this ratio may suggest the change in distribution of lipids and proteins in malignant tissues with respect to normal one. The ratio of the intensity of the peak at 1045 cm-1 to that at 1545 cm-1 give an estimate of the carbohydrate concentration (Colagar et al., 2011). Carbohydrate levels, which also include glycogen, in particular are known to be reduced in the presence of cancer as compared to healthy tissue. The absorption or metabolism of carbohydrate may be disturbed in cancerous tissue in the advanced stages of the disease. Usually, carbohydrate levels change in a more complex manner with the progression of the disease. Large variation in carbohydrate intensity ratio has been observed for patients with grade III of cancer 9 (Colagar et al., 2011). Observation of intensity ratio 1080 cm-1 to 1539 cm-1 is found to be decreased in the cancerous tissue. This ratio of phosphate to amide II may provide an estimation of the distribution of protein and phosphate in the malignant tissue. 4. Conclusions This study substantiates that FTIR spectroscopy has the potential of providing the biochemical information related to malignant transformation of breast tissue complementary to histopathology. Our results suggest that this technique is sensitive in analyzing the diseased state using a small panel of significant wavenumbers (allied to biomolecules). Apart from several other advantages, it is time and cost ineffective, does not require special sample preparation or biochemical reagents, leaving adequate material for other clinical investigations. It also holds tremendous possibility to be converted into a minimally invasive diagnostic method, when coupled with fiber-optic probe and several statistical analysis, which could be a real leap forward in the area of cancer diagnosis. Acknowledgements The authors are thankful to the Director, CSIR-National Physical Laboratory, New Delhi110012, for granting permission regarding publication of this work. Department of Science and Technology (DST/TSG/PT/2012/129), Govt. of India, New Delhi is thankfully acknowledged for providing financial support. Abbreviations: BC, breast cancer; FTIR, Fourier transform infrared; IR, infrared; ATR, attenuated total reflectance; DNA, deoxyribonucleic acid; RNA, ribonucleic acid; ECM, extracellular matrix; DTGS, deuterated triglycine sulfate; NADPH, nicotinamide adenine dinucleotide phosphate Authors’ Contributions RM is responsible for the concept and design of the study. RM, GT and SC searched the databases. UM and AA prepared the cryostat sections of breast tissue. GT, BR and SC 10 collected the samples and analyzed the data. GT and BR wrote the manuscript. RM, GK and HC reviewed and edited the manuscript extensively. Manuscript is revised and approved for the final version by all the authors. Disclosure Statement The authors declare that they have no competing interests. Ethics Approval and Consent to Participate The study has been performed with the standard protocol and approval of the Institutional Review Board (RS/MSSH/ONCO/10-5), (RS/MSSH/BMDRC/ONCO/ISC/15-11) and Institutional Institutional Scientific Committee Ethics Committee (RS/MSSH/BMDRC/ONCO/15-09) of Max Super Speciality Hospital, New Delhi, India. Written and informed consent has been obtained from all individual participants included in the study. Thereafter, the cryostat sections of tissues (normal and malignant) were collected and analyzed. 11 REFERENCES Banyay, M., Sarkar, M., & Gräslund, A. (2003). A library of IR bands of nucleic acids in solution. Biophysical Chemistry, 104(2), 477-488. doi: 10.1016/S0301- 4622(03)00035-8 Beljebbar, A., & Manfait, M. (2011). Fourier Transform Infrared Microspectroscopy for Cancer Diagnostic of C6 Glioma on Animal Model. Fourier Transforms-New Analytical Approaches and FTIR Strategies, InTech, ISBN, 978-953. doi: 10.5772/15361 Bouchard, M., Zurdo, J., Nettleton, E. J., Dobson, C. M., & Robinson, C. V. (2000). Formation of insulin amyloid fibrils followed by FTIR simultaneously with CD and electron microscopy. Protein Science, 9(10), 1960-1967. doi: 10.1110/ps.9.10.1960 Chirgadze, Y. N., & Nevskaya, N. (1976). Infrared spectra and resonance interaction of amide‐I vibration of the antiparallel‐chain pleated sheet. Biopolymers, 15(4), 607-625. doi: 10.1002/bip.1976.360150402 Cohenford, M. A., Lim, S., Brown, C., Chaudhry, M. A., Sigdel, S., Beckelhimer, E., & Rigas, B. (2012). FT-IR Microspectroscopy of Mouse Colon Tissues: Insight into the Chemistry of Carcinogenesis and Diagnostic Potential. The American Journal of Pathology, 181(6), 1961-1968. doi: 10.1016/j.ajpath.2012.08.039 Colagar, A. H., Chaichi, M. J., & Khadjvand, T. (2011). Fourier transform infrared microspectroscopy as a diagnostic tool for distinguishing between normal and malignant human gastric tissue. Journal of biosciences, 36(4), 669-677. doi: 10.1007/s12038-011-9090-5 Gajjar, K., Ahmadzai, A. A., Valasoulis, G., Trevisan, J., Founta, C., Nasioutziki, M., . . . Karakitsos, P. (2014). Histology verification demonstrates that biospectroscopy 12 analysis of cervical cytology identifies underlying disease more accurately than conventional screening: removing the confounder of discordance. PLoS One, 9(1), e82416. doi: 10.1371/journal.pone.0082416 Gajjar, K., Heppenstall, L. D., Pang, W., Ashton, K. M., Trevisan, J., Patel, I. I., . . . Martin, F. L. (2013). Diagnostic segregation of human brain tumours using Fourier-transform infrared and/or Raman spectroscopy coupled with discriminant analysis. Analytical Methods, 5(1), 89-102. doi: 10.1039/C2AY25544H Kondepati, V. R., Heise, H. M., Oszinda, T., Mueller, R., Keese, M., & Backhaus, J. (2008). Detection of structural disorders in colorectal cancer DNA with Fourier-transform infrared spectroscopy. Vibrational Spectroscopy, 46(2), 150-157. doi: 10.1016/j.vibspec.2007.12.010 Liquier, J., Taboury, J., Taillandier, E., & Brahms, J. (1977). Infrared linear dichroism investigations of deoxyribonucleic complexes with histones H2B and H3. Biochemistry, 16(14), 3262-3266. doi: 10.1021/bi00633a034 Malins, D. C., Polissar, N. L., Nishikida, K., Holmes, E. H., Gardner, H. S., & Gunselman, S. J. (1995). The etiology and prediction of breast cancer. Fourier transform‐infrared spectroscopy reveals progressive alterations in breast DNA leading to a cancer‐like phenotype in a high proportion of normal women. Cancer, 75(2), 503-517. doi: 10.1002/1097-0142(19950115)75:2<503::AID-CNCR2820750213>3.0.CO;2-0 Mehrotra, R., Gupta, A., Kaushik, A., Prakash, N., & Kandpal, H. (2007). Infrared spectroscopic analysis of tumor pathology. Indian journal of experimental biology, 45(1), 71. doi: http://nopr.niscair.res.in/handle/123456789/5234 Pu, Y., Wang, W., Tang, G., & Alfano, R. R. (2010). Changes of collagen and nicotinamide adenine dinucleotide in human cancerous and normal prostate tissues studied using 13 native fluorescence spectroscopy with selective excitation wavelength. Journal of biomedical optics, 15(4), 1-5. doi: 10.1117/1.3463479 Wang, J.-S., Shi, J.-S., Xu, Y.-Z., Duan, X.-Y., Zhang, L., Wang, J., . . . Wu, J.-G. (2003). FT-IR spectroscopic analysis of normal and cancerous tissues of esophagus. World Journal of Gastroenterology, 9(9), 1897-1899. doi: 10.3748/wjg.v9.i9.1897 Yu, L.-R., Zhou, M., Conrads, T. P., & Veenstra, T. D. (2004). Diagnostic proteomics: Serum proteomic patterns for the detection of early stage cancers. Disease markers, 19(4), 209-218. doi: 10.1155/2004/612071 14 TBSD-2016-0593.R2 The Figure numbers have been incorporated in front of Figure Captions. The revised Figure Captions are listed below. FIGURE CAPTIONS Figure 1 Overlaid infrared spectra of normal and malignant breast tissue in the region 1300-700 cm-1. Figure 2 Overlaid second order derivative infrared spectra of normal and malignant breast tissue in the region 1300-700 cm-1. Figure 3 Overlaid infrared spectra of normal and malignant breast tissue in the region 1700-1500 cm-1. Figure 4 Overlaid second order derivative infrared spectra of normal and malignant breast tissue in the region 1700-1500 cm-1. Figure 5 Overlaid infrared spectra of normal and malignant breast tissue in the region 3100-2700 cm-1. Figure 6 Overlaid second order derivative infrared spectra of normal and malignant breast tissue in the region 3100-2700 cm-1. 15 16 17 18 19 20 21