F] Journal of Microscopy, Vol. 105, Pt 3, December 1975, pp. 293-303. Received 1 July 1975; revision received 22 August 1975
Cathodoluminescence of biological molecules, macromolecules and cells
by W. A. B A R N E T TM. , L. H. W I S Eand E. C. J O N E S * , Centre for Materials Science, The University of Birmingham, Elms Road, North Campus, P.O. Box 363, Birmingham, B15 2 T T and * Department of Anatomy, The University of Birmingham, Medical School, Metchley Park Road, Birmingham, B15 2TJ. SUMMARY
Cathodoluminescence observations on biological compounds are compared with previously established data from ultra violet and visible light excitation studies. The comparison demonstrates that the same molecules are responsible for the luminescence properties of macromolecules independent of the type of exciting radiation. Natural cathodoluminescence was also observed from cells. Moreover, advantages gained by the absorption of strongly cathodoluminiscent dyes into cells are demonstrated. In most studies using the scanning electron microscope (SEM) reflected or emitted electrons are detected (Echlin, 1973) but other effects may also be monitored including the detection of visible radiation; this is referred to as the cathodoluminescent mode of operation (Thornton, 1968). The detection and analysis of fluorescence and phosphorescence following photon excitation has been used to study the structure of biological molecules and macromolecules (Konev, 1967; Vigny ?L Duquesne, 1974; Luisi et al., 1975). Although cathodoluminescence studies of biological molecules have been reported (Muir & Grant, 1974) the technique has not received the attention accorded to photon excitation studies. A Cambridge Scientific Instruments Limited (C.S.I.) ' Stereoscan' Mark I1 SEM operating at accelerating voltages of not more than 10 kV was used throughout these exploratory series of experiments. The SEM was fitted with the C.S.I. cathodoluminescence accessory comprising of a simple plastic focusing lens, photomultiplier, amplifier and second video display. The first video channel was used to display either the backscattered and emissive image or the backscattered image. The latter was compared with the cathodoluminescence image as suggested by Muir & Grant (1974) to check for anomalous cathodoluminescence resulting from the arrival of high energy primaries at the photocathode. The influence of stray cathodoluminescence resulting from high energy electron scintillation of the light guide and lens and filament incandescence was also monitored by attempting to image a non-luminescent substance under the same operating conditions as that used to examine the specimen. In all the examples recorded the influence of these factors was minimal. All specimens were examined uncoated as coating decreased observable luminescence.
299
W. A . Barnett, M . L. H. Wise and E. C.Jones Photon excited spectral characteristics of proteins are determined by two aromatic amino acids, tyrosine and tryptophan (Teale & Weber, 1957; Teale, 1960; Longworth, 1961; Edelhoch, 1967; Edelhoch, Bernstein & Wilchek, 1968; Edelhoch, Perlman & Wilchek, 1969). Other amino acids present in the protein sequence (e.g. glycine, an aliphatic amino acid) make no contribution to the observed luminescence. Powders of tyrosine, tryptophan and glycine were examined in the SEM and cathodoluminescence was observed from tyrosine and tryptophan, but not from glycine. The cathodoluminescence from tyrosine was weaker than that from tryptophan at room temperature, but the emission from tyrosine, unlike that from tryptophan, increased at temperatures below 170°K. This effect is most probably due to an increase in phosphorescence (Dillon, 1971), as even at slow scan speeds there was a high background signal against which it was impossible to resolve a luminescent image. Several crystalline and purified proteins, including ribonuclease, human and bovine serum albumin and lysozyme, were examined in the SEM. The observed cathodoluminescence from these was weaker than that from the equivalent amounts of tryptophan and tyrosine contained within the protein. This observation is consistent with results from photon excitation studies and may be due to inter- or intra-molecular interactions (Teale, 1960; Edelhoch et al., 1969; Brun, Toulmk & Hklkne, 1975). Cathodoluminescence was also observed from the four nucleic acid bases, adenine, thymine, cytosine and guanine; of these, adenine cathodoluminesced most strongly (Fig. 1). Calf thymus DNA (Miles Laboratories Limited, 9;) purity = 89.5 with 0.34% associated protein) was only weakly cathodoluminescent
Fig. 1. Cathodoluminescence micrograph of adenine powder at 10 kV and x 60.
300
Cathodolumznescence of biological materials
Fig. 2. Cathodoluminescence micrographs of cornified vaginal epithelial cells (a) natural cathodoluminescence, 10 kV and x 250; (b) cells treated with quinacrine dihydrochloride 5 IzV and x 220.
when compared. to adenine. A similar result has been obtained from photon excitation studies of DNA, because only one of the four bases participates in the observed luminescence (Konev, 1967). Cathodoluminescence has been observed from cells (Muir et al., 1971; Horl, 1972; Judge, Stubbs & Philp, 1974; Nixon, 1974) and we have found that airdried, unfixed and unstained vaginal epithelial cells from mice in the pro-oestrus, oestrus or di-oestrus phase of the reproductive cycle also possess this property (Fig. 2a). Weak cathodoluminescence was observed froin all cell types and distinct changes in intensity between phases was not observed. However, with the present photon collection system only a small proportion of the emitted light was detected and an improved system (Muir & Grant, 1974; Carlsson & van Essen, 1974; De Mets, 1974; Judge et al., 1974; Brocker et al., 1975) may allow the detection of changes occurring during the cell cycle. An alternative approach to biological studies using naturally occurring cathodoluminescence effects is to introduce specific dyes to enhance cathodoluminescence (Pease and Hayes, 1966; Soni, Kalnins & Haggis, 1975). The dye quinacrine dihydrochloride, previously used as a biological stain for chromosome typing (Allderdice et al., 1973; Distiche & Bontemps, 1974), was found to be more luminescent than natural biological compounds under all SEM operating conditions. On the basis of these initial observations a controlled dosage of quinacrine was introduced into the vaginae of anaesthetized mice in known stages of the oestrous cycle prior to cell collection. SEM examination showed that it was absorbed into all cell types and caused a marked increa;e in observable cathodoluminescence (Fig. 2b). The diminution of the intensity of cathodoluminescence during SEM examination (‘fading’) has previously been observed (Pease & Hayes, 1966; Falk, 1972). This effect caused by electron beam damage, and to a lesser extent contamination of the surface, made observations of weak emission cathodolumin30 1
W. A . Barnett, M . L. H . Wise and E. C.Jones escence difficult. The vaginal epithelial cells were examined for only a short period of time before ‘fading’ became apparent; however the introduction of quinacrine into these cells aided observation by increasing the detectable luminescence signal. This exploratory study shows that some of the molecules and macromolecules which are incorporated into cells and tissues cathodoluminesce. Various cells also luminesce under the influence of the electron beam but fading may make definitive observations difficult. Increased luminescence from biological substances can be successfully achieved using a selective dye and this technique should prove increasingly useful in the study of cells and tissues.
ACKNOWLEDGMENTS
We wish to thank Dr F. W. J. Teale for advice during the course of this work and Mrs J. Heathcote for assistance in the preparation of this paper.
References Allderdice, P.W., Miller, O.J., Miller, D.A., Warburton, D., Pearson, P.L., Klein, G. & Harris, H. (1973) Chromosome analysis of two related heteroploid mouse cell lines by quinacrine fluorescence. 3. Cell Sci. 12, 263. Brocker, W., Schmidt, E.H., Pfefferkorn, G. & Beller, F.K. (1975 Demonstration of cathodoluminescence in fluorescein marked biological tissues. In : Proc. VZZth Ann. SEM Symposium ZZTRZ, St. Louis, USA, 243. Brun, F., Toulme, J.J. & Helene, C. (1975) Interactions of aromatic residues of proteins with nucleic acids. Fluorescence studies of the binding of oligopeptides containing tryptophan and tyrosine residues to pol ynucleotides. Biochem. 14, 558. Carlsson, L. & van Essen, C.G. (1974) An efficient apparatus for studying cathodoluminescence in the scanning electron microscope.3. Phys. E: Sci. Znstrum. 7, 98. De Mets, M. (1974) Improved cathodoluminescence detection system. J . Phys. E: Sci. Znstrum. 7 , 971. Dillon, M.A. (1971) Electron impact. In: Creation and Detection of the Excited State (Ed. by A.A. Lamola), p. 375. Marcel Dekker Inc., New York. Distkche, C. & Bontemps, J. (1974) Chromosome regions containing DNAs of known base composition, specifically evidenced by 2, 7-di-t-butyl Proflavine. Comparison with Q banding and relation to dye-DNA interactions. Chromosoma, 47, 263. Echlin, P. (1973) The scanning electron microscope and its applications to research. Microscopica Acta, 73, 97 and 189. Edelhoch, H. (1967) Spectroscopic determination of tryptophan and tyrosine in proteins. Biochem. 6, 1948. Edelhoch, H., Bernstein, R.S. & Wilchek, M. (1968) The fluorescence of tyrosyl and tryptophanyl diketopiperazines. J. Biol. Chem. 243, 5985. Edelhoch, H., Perlman, R.L. & Wilchek, M. (1969) Tyrosine fluorescence in proteins. Annals N . Y . Acad. Sci. 158, 391. Falk, R.H. (1972) Cathodoluminescence-its potential for biology. In: Proc. Vth Ann. SEM Symposium IITRI, Chicago, USA, 35. Horl, E.M. (1972) Scanning electron microscopy of biological material using cathodoluminescence. Micron, 3, 540. Judge, F.J., Stubbs, J.M. & Philp, J. (1974) A concave mirror, light pipe photon collecting system for cathodoluminescent studies on biological specimens in the JSM2 scanning electron microscope. 3. Phys. E: Sci. Znstrum. 7, 173. Konev, S.V. (1967) Fluorescence and Phosphorescence of Proteins and Nucleic Acids. Plenum Press, New York. Longworth, J.W. (1961) Tyrosine phosphorescence of proteins. Biochem. 3.81, 23,. Luisi, P.L., Baici, A., Bonner, F.J.& Aboderin, A.A. (1975) Relationship between fluorescence and conformation of ENAD bound to dehydrogenases. Biochem. 14,362. Muir, M.D., Grant, P.R., Hubbard, G. & Mundell, J. (1971) Cathodoluminescence spectra. In: Proc. ZVth Ann. SEM Symposium ZZTRZ, Chicago, USA, 401. +
302
Cathodolumznescence of biological materials Muir, M.D. & Grant, P.R. (1974) Cathodoluminescence. In: Quantitative Scanning Electron Microscopy (Ed. by D.B. Holt, M.D. Muir, P.R. Grant and I.M. Boswarva), p. 287. Academic Press, London. Nixon, W.C. (1974) Dynamic scanning electron microscopy of human skin in situ.J. Phys. E: Sci. Instrum. 7, 685. Pease, R.F.W. & Hayes, T.L. (1966) Scanning electron microscopy of biological material. Nature, 210, 1049. Soni, S.L., Kalnins, V.J. & Haggis, G.H. (1975) Localisation of caps on mouse B lymphocytes by scanning electron microscopy. Nature, 255, 717. Teale, F.W.J. & Weber, G. (1957) Ultraviolet fluorescence of the aromatic amino acids. Biochem. J . 65, 476. Teale, F.W. J. (1960). The ultraviolet fluorescence of proteins in neutral solutions. Biochem. 3. 76, 381. Thornton, P.R. (1968) Scanning Electron Microscopy. Chapman and Hall Limited, London. Vigny, P. & Duquesne, M. (1974) A spectrophotofluorometer for measuring very weak fluorescences from biological molecules. Photochern. Photobiol. 20, 15.
303
Cathodoluminescence of biological molecules, macromolecules and cells
by W. A. B A R N E T TM. , L. H. W I S Eand E. C. J O N E S * , Centre for Materials Science, The University of Birmingham, Elms Road, North Campus, P.O. Box 363, Birmingham, B15 2 T T and * Department of Anatomy, The University of Birmingham, Medical School, Metchley Park Road, Birmingham, B15 2TJ. SUMMARY
Cathodoluminescence observations on biological compounds are compared with previously established data from ultra violet and visible light excitation studies. The comparison demonstrates that the same molecules are responsible for the luminescence properties of macromolecules independent of the type of exciting radiation. Natural cathodoluminescence was also observed from cells. Moreover, advantages gained by the absorption of strongly cathodoluminiscent dyes into cells are demonstrated. In most studies using the scanning electron microscope (SEM) reflected or emitted electrons are detected (Echlin, 1973) but other effects may also be monitored including the detection of visible radiation; this is referred to as the cathodoluminescent mode of operation (Thornton, 1968). The detection and analysis of fluorescence and phosphorescence following photon excitation has been used to study the structure of biological molecules and macromolecules (Konev, 1967; Vigny ?L Duquesne, 1974; Luisi et al., 1975). Although cathodoluminescence studies of biological molecules have been reported (Muir & Grant, 1974) the technique has not received the attention accorded to photon excitation studies. A Cambridge Scientific Instruments Limited (C.S.I.) ' Stereoscan' Mark I1 SEM operating at accelerating voltages of not more than 10 kV was used throughout these exploratory series of experiments. The SEM was fitted with the C.S.I. cathodoluminescence accessory comprising of a simple plastic focusing lens, photomultiplier, amplifier and second video display. The first video channel was used to display either the backscattered and emissive image or the backscattered image. The latter was compared with the cathodoluminescence image as suggested by Muir & Grant (1974) to check for anomalous cathodoluminescence resulting from the arrival of high energy primaries at the photocathode. The influence of stray cathodoluminescence resulting from high energy electron scintillation of the light guide and lens and filament incandescence was also monitored by attempting to image a non-luminescent substance under the same operating conditions as that used to examine the specimen. In all the examples recorded the influence of these factors was minimal. All specimens were examined uncoated as coating decreased observable luminescence.
299
W. A . Barnett, M . L. H. Wise and E. C.Jones Photon excited spectral characteristics of proteins are determined by two aromatic amino acids, tyrosine and tryptophan (Teale & Weber, 1957; Teale, 1960; Longworth, 1961; Edelhoch, 1967; Edelhoch, Bernstein & Wilchek, 1968; Edelhoch, Perlman & Wilchek, 1969). Other amino acids present in the protein sequence (e.g. glycine, an aliphatic amino acid) make no contribution to the observed luminescence. Powders of tyrosine, tryptophan and glycine were examined in the SEM and cathodoluminescence was observed from tyrosine and tryptophan, but not from glycine. The cathodoluminescence from tyrosine was weaker than that from tryptophan at room temperature, but the emission from tyrosine, unlike that from tryptophan, increased at temperatures below 170°K. This effect is most probably due to an increase in phosphorescence (Dillon, 1971), as even at slow scan speeds there was a high background signal against which it was impossible to resolve a luminescent image. Several crystalline and purified proteins, including ribonuclease, human and bovine serum albumin and lysozyme, were examined in the SEM. The observed cathodoluminescence from these was weaker than that from the equivalent amounts of tryptophan and tyrosine contained within the protein. This observation is consistent with results from photon excitation studies and may be due to inter- or intra-molecular interactions (Teale, 1960; Edelhoch et al., 1969; Brun, Toulmk & Hklkne, 1975). Cathodoluminescence was also observed from the four nucleic acid bases, adenine, thymine, cytosine and guanine; of these, adenine cathodoluminesced most strongly (Fig. 1). Calf thymus DNA (Miles Laboratories Limited, 9;) purity = 89.5 with 0.34% associated protein) was only weakly cathodoluminescent
Fig. 1. Cathodoluminescence micrograph of adenine powder at 10 kV and x 60.
300
Cathodolumznescence of biological materials
Fig. 2. Cathodoluminescence micrographs of cornified vaginal epithelial cells (a) natural cathodoluminescence, 10 kV and x 250; (b) cells treated with quinacrine dihydrochloride 5 IzV and x 220.
when compared. to adenine. A similar result has been obtained from photon excitation studies of DNA, because only one of the four bases participates in the observed luminescence (Konev, 1967). Cathodoluminescence has been observed from cells (Muir et al., 1971; Horl, 1972; Judge, Stubbs & Philp, 1974; Nixon, 1974) and we have found that airdried, unfixed and unstained vaginal epithelial cells from mice in the pro-oestrus, oestrus or di-oestrus phase of the reproductive cycle also possess this property (Fig. 2a). Weak cathodoluminescence was observed froin all cell types and distinct changes in intensity between phases was not observed. However, with the present photon collection system only a small proportion of the emitted light was detected and an improved system (Muir & Grant, 1974; Carlsson & van Essen, 1974; De Mets, 1974; Judge et al., 1974; Brocker et al., 1975) may allow the detection of changes occurring during the cell cycle. An alternative approach to biological studies using naturally occurring cathodoluminescence effects is to introduce specific dyes to enhance cathodoluminescence (Pease and Hayes, 1966; Soni, Kalnins & Haggis, 1975). The dye quinacrine dihydrochloride, previously used as a biological stain for chromosome typing (Allderdice et al., 1973; Distiche & Bontemps, 1974), was found to be more luminescent than natural biological compounds under all SEM operating conditions. On the basis of these initial observations a controlled dosage of quinacrine was introduced into the vaginae of anaesthetized mice in known stages of the oestrous cycle prior to cell collection. SEM examination showed that it was absorbed into all cell types and caused a marked increa;e in observable cathodoluminescence (Fig. 2b). The diminution of the intensity of cathodoluminescence during SEM examination (‘fading’) has previously been observed (Pease & Hayes, 1966; Falk, 1972). This effect caused by electron beam damage, and to a lesser extent contamination of the surface, made observations of weak emission cathodolumin30 1
W. A . Barnett, M . L. H . Wise and E. C.Jones escence difficult. The vaginal epithelial cells were examined for only a short period of time before ‘fading’ became apparent; however the introduction of quinacrine into these cells aided observation by increasing the detectable luminescence signal. This exploratory study shows that some of the molecules and macromolecules which are incorporated into cells and tissues cathodoluminesce. Various cells also luminesce under the influence of the electron beam but fading may make definitive observations difficult. Increased luminescence from biological substances can be successfully achieved using a selective dye and this technique should prove increasingly useful in the study of cells and tissues.
ACKNOWLEDGMENTS
We wish to thank Dr F. W. J. Teale for advice during the course of this work and Mrs J. Heathcote for assistance in the preparation of this paper.
References Allderdice, P.W., Miller, O.J., Miller, D.A., Warburton, D., Pearson, P.L., Klein, G. & Harris, H. (1973) Chromosome analysis of two related heteroploid mouse cell lines by quinacrine fluorescence. 3. Cell Sci. 12, 263. Brocker, W., Schmidt, E.H., Pfefferkorn, G. & Beller, F.K. (1975 Demonstration of cathodoluminescence in fluorescein marked biological tissues. In : Proc. VZZth Ann. SEM Symposium ZZTRZ, St. Louis, USA, 243. Brun, F., Toulme, J.J. & Helene, C. (1975) Interactions of aromatic residues of proteins with nucleic acids. Fluorescence studies of the binding of oligopeptides containing tryptophan and tyrosine residues to pol ynucleotides. Biochem. 14, 558. Carlsson, L. & van Essen, C.G. (1974) An efficient apparatus for studying cathodoluminescence in the scanning electron microscope.3. Phys. E: Sci. Znstrum. 7, 98. De Mets, M. (1974) Improved cathodoluminescence detection system. J . Phys. E: Sci. Znstrum. 7 , 971. Dillon, M.A. (1971) Electron impact. In: Creation and Detection of the Excited State (Ed. by A.A. Lamola), p. 375. Marcel Dekker Inc., New York. Distkche, C. & Bontemps, J. (1974) Chromosome regions containing DNAs of known base composition, specifically evidenced by 2, 7-di-t-butyl Proflavine. Comparison with Q banding and relation to dye-DNA interactions. Chromosoma, 47, 263. Echlin, P. (1973) The scanning electron microscope and its applications to research. Microscopica Acta, 73, 97 and 189. Edelhoch, H. (1967) Spectroscopic determination of tryptophan and tyrosine in proteins. Biochem. 6, 1948. Edelhoch, H., Bernstein, R.S. & Wilchek, M. (1968) The fluorescence of tyrosyl and tryptophanyl diketopiperazines. J. Biol. Chem. 243, 5985. Edelhoch, H., Perlman, R.L. & Wilchek, M. (1969) Tyrosine fluorescence in proteins. Annals N . Y . Acad. Sci. 158, 391. Falk, R.H. (1972) Cathodoluminescence-its potential for biology. In: Proc. Vth Ann. SEM Symposium IITRI, Chicago, USA, 35. Horl, E.M. (1972) Scanning electron microscopy of biological material using cathodoluminescence. Micron, 3, 540. Judge, F.J., Stubbs, J.M. & Philp, J. (1974) A concave mirror, light pipe photon collecting system for cathodoluminescent studies on biological specimens in the JSM2 scanning electron microscope. 3. Phys. E: Sci. Znstrum. 7, 173. Konev, S.V. (1967) Fluorescence and Phosphorescence of Proteins and Nucleic Acids. Plenum Press, New York. Longworth, J.W. (1961) Tyrosine phosphorescence of proteins. Biochem. 3.81, 23,. Luisi, P.L., Baici, A., Bonner, F.J.& Aboderin, A.A. (1975) Relationship between fluorescence and conformation of ENAD bound to dehydrogenases. Biochem. 14,362. Muir, M.D., Grant, P.R., Hubbard, G. & Mundell, J. (1971) Cathodoluminescence spectra. In: Proc. ZVth Ann. SEM Symposium ZZTRZ, Chicago, USA, 401. +
302
Cathodolumznescence of biological materials Muir, M.D. & Grant, P.R. (1974) Cathodoluminescence. In: Quantitative Scanning Electron Microscopy (Ed. by D.B. Holt, M.D. Muir, P.R. Grant and I.M. Boswarva), p. 287. Academic Press, London. Nixon, W.C. (1974) Dynamic scanning electron microscopy of human skin in situ.J. Phys. E: Sci. Instrum. 7, 685. Pease, R.F.W. & Hayes, T.L. (1966) Scanning electron microscopy of biological material. Nature, 210, 1049. Soni, S.L., Kalnins, V.J. & Haggis, G.H. (1975) Localisation of caps on mouse B lymphocytes by scanning electron microscopy. Nature, 255, 717. Teale, F.W.J. & Weber, G. (1957) Ultraviolet fluorescence of the aromatic amino acids. Biochem. J . 65, 476. Teale, F.W. J. (1960). The ultraviolet fluorescence of proteins in neutral solutions. Biochem. 3. 76, 381. Thornton, P.R. (1968) Scanning Electron Microscopy. Chapman and Hall Limited, London. Vigny, P. & Duquesne, M. (1974) A spectrophotofluorometer for measuring very weak fluorescences from biological molecules. Photochern. Photobiol. 20, 15.
303
