- Email: [email protected]
Microscopic FTIR studies of lung cancer cells in pleural fluid
The Science of the Total Environment 204 (1997) 283-287
ELSEVIER
Short communication
Microscopic FTIW studies of lung cancer cells in pleural fluid H.P. Wang”,“, H.-C. Wangb, Y.-J. Huang” aDepatiment bDepamnent
of Environment
of Medicine,
Engineering, Cheng Kung University, Tainan, Veterans General Hospital-Kaoshiung, Kaoshiung,
Taiwan Taiwan
Received 24 February 1997; accepted 11 June 1997
Abstract
Structural changesassociatedwith lung cancer and tuberculouscells in pleural fluid were studiedby microscopic FTIR spectroscopy.Infrared spectra demonstratesignificant spectraldifferences betweennormal, lung cancer and tuberculouscells.The ratio of the peak intensitiesof the 1030and 1080cm-l bands(originated mainly in glycogen and phosphodiestergroupsof nucleic acids)differs greatly between normal and lung cancer samples.Such findings prompt the considerationthat recordinginfrared spectrafrom lung cancer and tuberculousceilsmay be of diagnostic value. Since measurementsof IR spectra of lung cancer cells in the pleural fluid can be a very rapid inexpensive process,our finding warrant exploration of this possibility in the investigation of the mechanism whereby the environmentalpollution related cancersdevelop. Q 1997Elsevier ScienceB.V. Kqwords:
Lung cancer cell; Tuberculouscell; Pleural fluid; STIR
1. Introduction The concept of the association between respiratory illness and Iung cancer and environmental poliution has been well recognized (Brownson et
*Corresponding author. Tel.: + 11 8866 2757575, ext. 54 551, fax.: + 11 8866 2752790; e-mail:wanghp@mail. ncku.edu.tw 004%9697/97/$17.00 PZZ SOO48-9697(97)
al., 1987; Crawford, 1988; Morris, 1995). Air pollution by polycyclic aromatic hydrocarbons (Lewtas, 1993) and volatile organic compounds (Ashley et al., 1995), irrespective of source, present a threat to health which is related to the amount of pollution, duration or exposure and total dose. No one can logically claim that in the presence of the trace carcinogen there is no risk. A negative cannot be proven either. Indeed, cause-effect relationships cannot scientifically be
0 1997 Elsevier Science B.V. All rights reserved. 00180-O
284
H.P. Wang et al. / The Science of the Total Environment
proven by estimated based on the statistics. Thus, fundamental approaches in the molecular level would be of importance. Infrared spectroscopy has been frequently used in elucidating structure and bonding of molecules (Parker, 1983; Yang et al., 1985; Le Gal et al., 1991, 1993; Wang et al., 1991). Because of its high resolution, infrared spectroscopy is also an excellent tool for measuring small frequency shifts. Recent improvements in the infrared spectroscopic techniques have made it possible to measure the infrared spectra of microsamples by the transmission of reflectance method. Currently, methodological and technological advances are greatly enhancing the sensitivity of infrared spectroscopy in studying the highly complex biological materials. Infrared spectroscopy has been used in the studies of cancer cells and some attempts have been made to utilize them as a diagnostic tool (Rigas et al., 1990; Wong et al., 1991; Rigas and Wong, 1992; Le Gal et al., 1993). In order to investigate in greater detail in the lung cancer cells, microscopic FTIR spectroscopy in a reflectance mode was used. We have also applied this method to determine the tuberculous cells separated from the pleural effusion. 2. Materials
204 (1997) 283-287
were frozen in liquid nitrogen until studied spectroscopically. 3. Results
and discussion
With a view to determining the structure and bonding of the lung cancer cells, we consider spectra in Fig. 1, which were measured by microscopic FTIR spectroscopy with an optical cross section of 40 ,um. Fig. 1 shows that similar infrared spectra are observed at three selected points of a typical lung cancer cell sample separated from the malignant pleural effusion. A detailed discussion of the infrared spectra of cancer cells was give by Wong and Rigas (Rigas et al., 1990; Wong et al., 1991; Rigas and Wong, 1992). The bands at 1030 and 1050 cm-‘, frequently found in the glycogen-rich tissues, can be asvibrations and the C-O signed as -CH,OH stretching coupled with C-O bending of the C -OH carbohydrates, respectively (Rigas and Wong, 1992; Wong et al., 1991). The band at 1080 cm-’ has been assigned to the symmetry phosphate [PO, (sym)] stretching (Rigas et al., 1990). Glycogen also makes a main contribution to the intensity of this band. The band at 1241 cm-’ can
and methods
Infrared spectra were recorded on a Digilab FTIR spectrometer (ITS-40) with fully computerized data storage and data handling capability. Infrared beam was also directed into a microscopy (LIMA-500) equipped with a liquid nitrogen cooledMCT (mercury cadmium telluride) detector for determining spectra of microsamples. For all spectra reported, a 64-scan data accumulation was used at a resolution of 4 cm-i. Measurement capability below the lop3 absorbance level is achieved with good signal-to-noise ratios. In order to observe small changes in band intensity and frequency, substraction and base line correction procedures were used to obtain spectra that more effectively compare spectra measured. Eight pairs of normal and lung cancer samples were collected from eight patients. Lung cancer and tuberculous samples were also separated from the pleural effusion by centrifugation. Samples
PWsym)
i1 I 5cx
,400
,300
,200
,100
,000
Wavenumbers
Fig. 1. Infrared spectra of a lung cancer microscopic ITIR spectroscopy.
sample
measured
by
H.P. Wang et al. /The
Science of the Total Environment
phosphate be assigned to the asymmetric [PO;(asym)l stretching. The vibrations of the PO; groups are mainly in the phosphodiester groups on nucleic acids. Generally, the PO, groups of phospholipids do not contribute to these bands. It should be notedthat the spectra of DNA and RNA display extremely weak peaks at 1399 and 1457 cm-’ (Wong et al., 1991; Rigas and Wong, 1992). These bands, arising mainly from the vibrational modes of methyl and methylene groups of proteins and lipids and the amide groups (Rigas et al., 1990). The band at 1173 cm-’ can be assigned to the CO stretchings of the C-OH groups of serine, threonine and tyrosine in cell proteins as well as the carbohydrates. The bands at approx. 1396 and 1449 cm-’ are known to be mainly from symmetric and asymmetric CH, bendings of the methyl groups of proteins, respectively (Rigas et al., 1990; Wong et al., 1991). The CM, bending of the acyl chains of lipids is contributed to the band at 1469 cm-‘. Study of the frequency region 1000-1300 cm-’ reveals significant differences in the infrared spectra between normal and malignant lung cells. Fig. 2, illustrating these typical differences, shows the Superimposed infrared spectra of a pair of
normal and malignant tissue samples from one patient. The strong bands at 1241 and 1080 cm-’ (wavenumbers in normal cells, see Fig. 2a) are contributed to the asymmetric and symmetric phosphate (PO,) stretchings, respectively. It is very clear (in all eight cases), the malignant cells display decreased intensity of the PO, (asyrn) band, increased intensity of the PO; (sym) band and the changes in the shape of these bands. Frequency shifts also need to be considered. The PO; (asym) band displays a red shift from 1241 cm-l in the normal tissue to 1238 cm-’ in the lung cancer cells while the PO, (sym) band is shifted from 1080 to 1084 cm-*. It should be noted that by using the infrared microsampling method, the variations in the intensity, wavenumber and band shape, that is due to changes in the proportion of non-malignant and non-epithelial cells present in each tissue section, are highly reduced. Generally, the wavenumbers of the PO; (asym) vibrations are at 1240 cm-i when it is non-hydrogen-bonded and at 1220 cm-’ when it is highly hydrogen-bonded (Rigas and Wong, 1992). Thus, the hydrogen-bonding of the phosphate backbone of nucleic acids in the lung-cancer cell is in-
(b)
,300
1200
285
204 (1997) 283-287
.
iooo
ilO Wavenumbers
Fig. 2. Infrared samples.
spectra
of (a) normal
and (b)
lung
cancer
Fig. 3. Infrared absorbance ratio of (a) cancer, and (c) tuberculous samples.
normal,
(b)
lung
286
H.P. Wang et al. / The Science of the Total Environment
creased as indicated by the increase in the intensity of the infrared band at approx. 1220 cm-l (deconvolution of Fig. 2b), that is in contrast to the corresponding normal lung tissue. In addition, there is a broad underlying feature in the 1160 cm-r region which may be the superposition of different components. By using deconvolution method, it is found that the broad feature corresponding to CO stretching band consists of three overlapping bands peaking at 1553, 1160 and 1172 cm-‘. Of these, the first two band display intensity and the third band displays increased intensity when compared with the corresponding bands of normal lung cells. The component bands at 1153 and 1161 cm-i arise from the stretching vibrations of hydrogenbonding C-OH groups, whereas the band at 1172 cm-’ is due to the stretching vibrations of non-hydrogen-bonded C-OH groups. The infrared spectra of all eight-paired (normal and lung cancer) cell samples from eight patients represent prominent differences include the significant changes in intensity and wavenumbers of the bands at 1030, 1080, 1153 and 1241 cm-l in malignant samples. Fig. 3 shows that the ratio of the peak intensities of the 1030 and 1080 cm-r
Fig. 4. Infrared samples.
spectra
of (a) tuberculous
and Cb) lung cnacer
204 (1997) 283-287
bands (originated mainly in glycogen and phosphodiester groups nucleic acids) differs greatly between normal and lung cancer samples. Note that the ratio for all samples measured are within the same range. Fig. 4 depicts superimposed infrared of lung cancer cell and tuberculous cellsamples separated. More specifically, the tuberculous sample shows a blue shift of 6 cm-’ approximately from the lung cancer sample at 1084 cm-‘. The ratio of the peak intensity of the 1030 and 1080 cm-’ bands for tuberculous samples is 0.67 approximately (see Fig. 3). As expected, the loss of hydrogen bonding of the C-OH of amino acid residues of proteins for the lung cancer as well as the tuberculous cells is observed spectroscopically at 1153 and 1160 cm-’ (Fig. 4). 4. Conclusions
By using microscopic FTIR spectroscopy, the structural changes associated with lung cancer and tuberculous cells in the pleural effusion were studied. The presence of glycogen in normal and tuberculous samples and its reduction generally in abnormal samples, including cancer cells, is observed spectroscopically. Infrared spectra also show that hydrogen bonding of phosphodiester groups of nucleic acid increased in lung cancer cells. One can speculate that it may reflect a step of a shared pathway in carcinogenesis or alternatively, a common epiphenomenon of malignant transformation. In addition, a reduction of hydrogen bonding of C-OH groups of amino acid residues of proteins for the lung cancer and tuberculous cells is observed. Our spectroscopic data demonstrate significant spectral differences between normal, lung cancer and tuberculous cells. Such findings prompt the consideration that recording infrared spectra from lung cancer and tuberculous samples separated from the pleural fluid may be of diagnostic value and can be a very rapid and inexpensive method. It may also warrant the exploration of the possibility in the investigation of the mechanism whereby the environmental pollution related cancers develop.
H.P. Wang et al. / The Science of the Tot& Environment
Acknowledgements
Financial support by the Veterans General Hospital-Kaoshiung and the National Science Council (NSC84-2211-E-006-022, NSC852621-P006-003 and NSC86-2113-M-006-020), Taiwan, ROC are gratefully acknowledged. References Ashley DL, Ronin MA, Hamar B, McGeehin MA. Removing the smoking confounder from blood volatile organic compounds measurements. Environ Res 1995;41:39. Brownson RC, Reif JS, Keefe TJ, Ferguson SW, Pritzl JA. Risk factors for adenocarcinoma of the lung. Am J Epidemiol 1987;125:25. Crawford WA. On air pollution, environmental tobacco smoke, radon, and lung cancer. International Journal of Air Pollution Control and Hazardous Waste Management 1988;38:1386. Le Gal J-M, Manfait M, Theophanides T. Applications of FT-IR spectroscopy in structural studies of cells and bacteria. J Mol Struct 1991;242:397. Le Gal J-M, Morhani H, Manfait M. Ultrastructural appraisal of the multidrug resistance in K562 and LR73 cell lines from Fourier Transform infrared spectroscopy. Cancer Res 1993;53:3681.
204 (1997) 283-287
287
Lewtas J. Complex mixtures of air pollutants. Characterizing the cancer risk of polycyclic organic matter. Environ Health Perspect 1993;100:211. Morris PD. Lifetime excess risk of death from lung cancer for a U.S. female never-smoker exposed to environmental tobacco some. Environ Res 1995;68:3. Parker FS. Application of infrared spectroscopy in biochemistry, biology and medicine. New York: Plenum Publishing, 1983. Rigas B, Wong PTT. Human colon adenocarcinoma cell lines display infrared spectroscopic features of malignant colon tissues. Cancer Res 1992;52:84. Rigas B, Morgello S, Goldman IS, Wong P’IT. Human colorectal cancers display abnormal Fourier-transform infrared spectra. Proc Nat1 Acad Sci USA 1990;76:8140. Wang HP, Eyring EM, Huai HA. Photoacoustic enhancement of surface IR modes in zeolite channels. Appl Spect 1991;14:883. Wong PTT, Wong RK Caputo TA, Godwin TA, Rigas B. Infrared spectroscopy of exfoliated human cervical cells. Evidence of extensive changes during carcinogenesis. Proc Nat1 Acad Sci USA 1991;88:10988. Yang PW, Baudais FL, Mantsch HH, Wong PTT, Moffatt DJ. Crystalline quartz as an internal pressure calibrant for high-pressure infrared spectroscopy. Appl Spect 1985:41:539.
ELSEVIER
Short communication
Microscopic FTIW studies of lung cancer cells in pleural fluid H.P. Wang”,“, H.-C. Wangb, Y.-J. Huang” aDepatiment bDepamnent
of Environment
of Medicine,
Engineering, Cheng Kung University, Tainan, Veterans General Hospital-Kaoshiung, Kaoshiung,
Taiwan Taiwan
Received 24 February 1997; accepted 11 June 1997
Abstract
Structural changesassociatedwith lung cancer and tuberculouscells in pleural fluid were studiedby microscopic FTIR spectroscopy.Infrared spectra demonstratesignificant spectraldifferences betweennormal, lung cancer and tuberculouscells.The ratio of the peak intensitiesof the 1030and 1080cm-l bands(originated mainly in glycogen and phosphodiestergroupsof nucleic acids)differs greatly between normal and lung cancer samples.Such findings prompt the considerationthat recordinginfrared spectrafrom lung cancer and tuberculousceilsmay be of diagnostic value. Since measurementsof IR spectra of lung cancer cells in the pleural fluid can be a very rapid inexpensive process,our finding warrant exploration of this possibility in the investigation of the mechanism whereby the environmentalpollution related cancersdevelop. Q 1997Elsevier ScienceB.V. Kqwords:
Lung cancer cell; Tuberculouscell; Pleural fluid; STIR
1. Introduction The concept of the association between respiratory illness and Iung cancer and environmental poliution has been well recognized (Brownson et
*Corresponding author. Tel.: + 11 8866 2757575, ext. 54 551, fax.: + 11 8866 2752790; e-mail:wanghp@mail. ncku.edu.tw 004%9697/97/$17.00 PZZ SOO48-9697(97)
al., 1987; Crawford, 1988; Morris, 1995). Air pollution by polycyclic aromatic hydrocarbons (Lewtas, 1993) and volatile organic compounds (Ashley et al., 1995), irrespective of source, present a threat to health which is related to the amount of pollution, duration or exposure and total dose. No one can logically claim that in the presence of the trace carcinogen there is no risk. A negative cannot be proven either. Indeed, cause-effect relationships cannot scientifically be
0 1997 Elsevier Science B.V. All rights reserved. 00180-O
284
H.P. Wang et al. / The Science of the Total Environment
proven by estimated based on the statistics. Thus, fundamental approaches in the molecular level would be of importance. Infrared spectroscopy has been frequently used in elucidating structure and bonding of molecules (Parker, 1983; Yang et al., 1985; Le Gal et al., 1991, 1993; Wang et al., 1991). Because of its high resolution, infrared spectroscopy is also an excellent tool for measuring small frequency shifts. Recent improvements in the infrared spectroscopic techniques have made it possible to measure the infrared spectra of microsamples by the transmission of reflectance method. Currently, methodological and technological advances are greatly enhancing the sensitivity of infrared spectroscopy in studying the highly complex biological materials. Infrared spectroscopy has been used in the studies of cancer cells and some attempts have been made to utilize them as a diagnostic tool (Rigas et al., 1990; Wong et al., 1991; Rigas and Wong, 1992; Le Gal et al., 1993). In order to investigate in greater detail in the lung cancer cells, microscopic FTIR spectroscopy in a reflectance mode was used. We have also applied this method to determine the tuberculous cells separated from the pleural effusion. 2. Materials
204 (1997) 283-287
were frozen in liquid nitrogen until studied spectroscopically. 3. Results
and discussion
With a view to determining the structure and bonding of the lung cancer cells, we consider spectra in Fig. 1, which were measured by microscopic FTIR spectroscopy with an optical cross section of 40 ,um. Fig. 1 shows that similar infrared spectra are observed at three selected points of a typical lung cancer cell sample separated from the malignant pleural effusion. A detailed discussion of the infrared spectra of cancer cells was give by Wong and Rigas (Rigas et al., 1990; Wong et al., 1991; Rigas and Wong, 1992). The bands at 1030 and 1050 cm-‘, frequently found in the glycogen-rich tissues, can be asvibrations and the C-O signed as -CH,OH stretching coupled with C-O bending of the C -OH carbohydrates, respectively (Rigas and Wong, 1992; Wong et al., 1991). The band at 1080 cm-’ has been assigned to the symmetry phosphate [PO, (sym)] stretching (Rigas et al., 1990). Glycogen also makes a main contribution to the intensity of this band. The band at 1241 cm-’ can
and methods
Infrared spectra were recorded on a Digilab FTIR spectrometer (ITS-40) with fully computerized data storage and data handling capability. Infrared beam was also directed into a microscopy (LIMA-500) equipped with a liquid nitrogen cooledMCT (mercury cadmium telluride) detector for determining spectra of microsamples. For all spectra reported, a 64-scan data accumulation was used at a resolution of 4 cm-i. Measurement capability below the lop3 absorbance level is achieved with good signal-to-noise ratios. In order to observe small changes in band intensity and frequency, substraction and base line correction procedures were used to obtain spectra that more effectively compare spectra measured. Eight pairs of normal and lung cancer samples were collected from eight patients. Lung cancer and tuberculous samples were also separated from the pleural effusion by centrifugation. Samples
PWsym)
i1 I 5cx
,400
,300
,200
,100
,000
Wavenumbers
Fig. 1. Infrared spectra of a lung cancer microscopic ITIR spectroscopy.
sample
measured
by
H.P. Wang et al. /The
Science of the Total Environment
phosphate be assigned to the asymmetric [PO;(asym)l stretching. The vibrations of the PO; groups are mainly in the phosphodiester groups on nucleic acids. Generally, the PO, groups of phospholipids do not contribute to these bands. It should be notedthat the spectra of DNA and RNA display extremely weak peaks at 1399 and 1457 cm-’ (Wong et al., 1991; Rigas and Wong, 1992). These bands, arising mainly from the vibrational modes of methyl and methylene groups of proteins and lipids and the amide groups (Rigas et al., 1990). The band at 1173 cm-’ can be assigned to the CO stretchings of the C-OH groups of serine, threonine and tyrosine in cell proteins as well as the carbohydrates. The bands at approx. 1396 and 1449 cm-’ are known to be mainly from symmetric and asymmetric CH, bendings of the methyl groups of proteins, respectively (Rigas et al., 1990; Wong et al., 1991). The CM, bending of the acyl chains of lipids is contributed to the band at 1469 cm-‘. Study of the frequency region 1000-1300 cm-’ reveals significant differences in the infrared spectra between normal and malignant lung cells. Fig. 2, illustrating these typical differences, shows the Superimposed infrared spectra of a pair of
normal and malignant tissue samples from one patient. The strong bands at 1241 and 1080 cm-’ (wavenumbers in normal cells, see Fig. 2a) are contributed to the asymmetric and symmetric phosphate (PO,) stretchings, respectively. It is very clear (in all eight cases), the malignant cells display decreased intensity of the PO, (asyrn) band, increased intensity of the PO; (sym) band and the changes in the shape of these bands. Frequency shifts also need to be considered. The PO; (asym) band displays a red shift from 1241 cm-l in the normal tissue to 1238 cm-’ in the lung cancer cells while the PO, (sym) band is shifted from 1080 to 1084 cm-*. It should be noted that by using the infrared microsampling method, the variations in the intensity, wavenumber and band shape, that is due to changes in the proportion of non-malignant and non-epithelial cells present in each tissue section, are highly reduced. Generally, the wavenumbers of the PO; (asym) vibrations are at 1240 cm-i when it is non-hydrogen-bonded and at 1220 cm-’ when it is highly hydrogen-bonded (Rigas and Wong, 1992). Thus, the hydrogen-bonding of the phosphate backbone of nucleic acids in the lung-cancer cell is in-
(b)
,300
1200
285
204 (1997) 283-287
.
iooo
ilO Wavenumbers
Fig. 2. Infrared samples.
spectra
of (a) normal
and (b)
lung
cancer
Fig. 3. Infrared absorbance ratio of (a) cancer, and (c) tuberculous samples.
normal,
(b)
lung
286
H.P. Wang et al. / The Science of the Total Environment
creased as indicated by the increase in the intensity of the infrared band at approx. 1220 cm-l (deconvolution of Fig. 2b), that is in contrast to the corresponding normal lung tissue. In addition, there is a broad underlying feature in the 1160 cm-r region which may be the superposition of different components. By using deconvolution method, it is found that the broad feature corresponding to CO stretching band consists of three overlapping bands peaking at 1553, 1160 and 1172 cm-‘. Of these, the first two band display intensity and the third band displays increased intensity when compared with the corresponding bands of normal lung cells. The component bands at 1153 and 1161 cm-i arise from the stretching vibrations of hydrogenbonding C-OH groups, whereas the band at 1172 cm-’ is due to the stretching vibrations of non-hydrogen-bonded C-OH groups. The infrared spectra of all eight-paired (normal and lung cancer) cell samples from eight patients represent prominent differences include the significant changes in intensity and wavenumbers of the bands at 1030, 1080, 1153 and 1241 cm-l in malignant samples. Fig. 3 shows that the ratio of the peak intensities of the 1030 and 1080 cm-r
Fig. 4. Infrared samples.
spectra
of (a) tuberculous
and Cb) lung cnacer
204 (1997) 283-287
bands (originated mainly in glycogen and phosphodiester groups nucleic acids) differs greatly between normal and lung cancer samples. Note that the ratio for all samples measured are within the same range. Fig. 4 depicts superimposed infrared of lung cancer cell and tuberculous cellsamples separated. More specifically, the tuberculous sample shows a blue shift of 6 cm-’ approximately from the lung cancer sample at 1084 cm-‘. The ratio of the peak intensity of the 1030 and 1080 cm-’ bands for tuberculous samples is 0.67 approximately (see Fig. 3). As expected, the loss of hydrogen bonding of the C-OH of amino acid residues of proteins for the lung cancer as well as the tuberculous cells is observed spectroscopically at 1153 and 1160 cm-’ (Fig. 4). 4. Conclusions
By using microscopic FTIR spectroscopy, the structural changes associated with lung cancer and tuberculous cells in the pleural effusion were studied. The presence of glycogen in normal and tuberculous samples and its reduction generally in abnormal samples, including cancer cells, is observed spectroscopically. Infrared spectra also show that hydrogen bonding of phosphodiester groups of nucleic acid increased in lung cancer cells. One can speculate that it may reflect a step of a shared pathway in carcinogenesis or alternatively, a common epiphenomenon of malignant transformation. In addition, a reduction of hydrogen bonding of C-OH groups of amino acid residues of proteins for the lung cancer and tuberculous cells is observed. Our spectroscopic data demonstrate significant spectral differences between normal, lung cancer and tuberculous cells. Such findings prompt the consideration that recording infrared spectra from lung cancer and tuberculous samples separated from the pleural fluid may be of diagnostic value and can be a very rapid and inexpensive method. It may also warrant the exploration of the possibility in the investigation of the mechanism whereby the environmental pollution related cancers develop.
H.P. Wang et al. / The Science of the Tot& Environment
Acknowledgements
Financial support by the Veterans General Hospital-Kaoshiung and the National Science Council (NSC84-2211-E-006-022, NSC852621-P006-003 and NSC86-2113-M-006-020), Taiwan, ROC are gratefully acknowledged. References Ashley DL, Ronin MA, Hamar B, McGeehin MA. Removing the smoking confounder from blood volatile organic compounds measurements. Environ Res 1995;41:39. Brownson RC, Reif JS, Keefe TJ, Ferguson SW, Pritzl JA. Risk factors for adenocarcinoma of the lung. Am J Epidemiol 1987;125:25. Crawford WA. On air pollution, environmental tobacco smoke, radon, and lung cancer. International Journal of Air Pollution Control and Hazardous Waste Management 1988;38:1386. Le Gal J-M, Manfait M, Theophanides T. Applications of FT-IR spectroscopy in structural studies of cells and bacteria. J Mol Struct 1991;242:397. Le Gal J-M, Morhani H, Manfait M. Ultrastructural appraisal of the multidrug resistance in K562 and LR73 cell lines from Fourier Transform infrared spectroscopy. Cancer Res 1993;53:3681.
204 (1997) 283-287
287
Lewtas J. Complex mixtures of air pollutants. Characterizing the cancer risk of polycyclic organic matter. Environ Health Perspect 1993;100:211. Morris PD. Lifetime excess risk of death from lung cancer for a U.S. female never-smoker exposed to environmental tobacco some. Environ Res 1995;68:3. Parker FS. Application of infrared spectroscopy in biochemistry, biology and medicine. New York: Plenum Publishing, 1983. Rigas B, Wong PTT. Human colon adenocarcinoma cell lines display infrared spectroscopic features of malignant colon tissues. Cancer Res 1992;52:84. Rigas B, Morgello S, Goldman IS, Wong P’IT. Human colorectal cancers display abnormal Fourier-transform infrared spectra. Proc Nat1 Acad Sci USA 1990;76:8140. Wang HP, Eyring EM, Huai HA. Photoacoustic enhancement of surface IR modes in zeolite channels. Appl Spect 1991;14:883. Wong PTT, Wong RK Caputo TA, Godwin TA, Rigas B. Infrared spectroscopy of exfoliated human cervical cells. Evidence of extensive changes during carcinogenesis. Proc Nat1 Acad Sci USA 1991;88:10988. Yang PW, Baudais FL, Mantsch HH, Wong PTT, Moffatt DJ. Crystalline quartz as an internal pressure calibrant for high-pressure infrared spectroscopy. Appl Spect 1985:41:539.