Received for publication May 9, Since the absolute transmission values at any. disk, the film thickness, scattering of light and

Size: px

Start display at page:

Download "Received for publication May 9, Since the absolute transmission values at any. disk, the film thickness, scattering of light and"

Phoebe Reynolds
5 years ago
Views:

1 INFRARED SPECTROPHOTOMETRY OF ENTERIC BACTERIA SEYMOUR LEVINE, HEBER J. R. STEVENSON, LESLIE A. CHAMBERS, AND BERNARD A. KENNER Environmental Health Center, United States Public Health Service, Cincinnati, Ohio Infrared spectrophotometry is rapidly assuming an important place in the biological sciences. Its value derives from the fact that each organic compound has a characteristic and distinct infrared spectrum that can be used to identify it (Randall et al., 1949). The biologist working with cells or tissues is confronted by a highly complex mixture of substances. The infrared spectrum of a cell, therefore, represents only an over-all chemical composition. It cannot, because of the complexity of the mixture, be interpreted quantitatively in terms of individual compounds. On the other hand, this method is simple, rapid, and versatile. It has been applied to tissues and cells without preliminary treatment other than drying by Barer, Cole, and Thompson (1949), Blout and Mellors (1949), and Schwarz et al. (1951). Randall, Smith, Colm, and Nungester (1951) have demonstrated the reproducibility of infrared spectra of chloroform extracts of tubercle bacilli and have indicated their usefulness in the control and development of fractionation procedures. Whole bacterial cells have been examined by Stevenson and Bolduan (1952) who showed that many species of bacteria were sufficiently varied in chemical composition to be differentiated by their infrared spectra. This paper represents a continuation of their work with emphasis on enteric bacteria. MATERIALS AND METHODS Bacteria are harvested by scraping the surface of agar plates or by centrifuging broth cultures. Since water absorbs strongly in the infrared region, the organisms must be spread and dried on the surface of silver chloride disks. Gla#s and Received for publication May 9, 1952 quartz are not useful as sample holders because of their infrared absorption. Spectra are obtained from a double beam, direct recording infrared spectrophotometer (Baird), using a sodium chloride prism as the dispersing element. In most cases, it is sufficient to record the spectrum from 5 to 12 microns (u) since this is the most distinctive region, although the range from 2 to 16 p& is available on this instrument.' Since the absolute transmission values at any point of the spectrum depend on several factors, including the absorption of the silver chloride disk, the film thickness, scattering of light and specific absorptions, these values are disregarded in favor of a consideration of the shape and relative depths of bands. The shape of the spectrum depends only on specific absorptions and is determined by the relative quantities of the many organic compounds present in the bacterial cells. Disintegration of the cells by sonic vibration does not change the spectrum. RESULTS Figure 1 presents the infrared spectrum from 2 to 15 p of a dried film of Escherichia coli harvested from nutrient agar (Difco). Certain of the absorption bands can be ascribed to definite chemical groupings. These include 0-H and N-H at 3 A, C-H at 3.4 and 6.9 is, C=O at 6.05,X, and C-N at 6.45, (Randall et al., 1949). Since the chief absorbing constituent (by weight) of the bacterial cells is protein, it is likely that the 6.05 and 6.45 bands are largely derived from the peptide linkages of this source. Interpretation of bands at higher wavelengths is more difficult. There is considerable evidence, however, to indicate that nucleic acids make an important contribution to the spectrum in the 8.0 to 8.1, band whereas carbohydrates and nucleic acids together may be largely responsible for the deep band that stretches from 8.6 to 10 p. The spectra of pure proteins, nucleic acids, and carbohydrates are consistent with these interpretations. Working with E. coli and other enteric bacteria, we have accumulated the following additional evidence for the biochemical interpretation of these absorption bands: Carbohydrates. Extraction of bacterial poly- 1 Note in the figures that transmission increases upward so that absorption bands extend downward. 10

2 19531 INFRARED SPECTROPHOTOMETRY 11 Figure 1. Infrared spectrum from 2 to 15 p of Escherichia coli harvested from nutrient agar after 18 hours at 37 C. r I- -r - - -~~~~~~~~~~~ X Wn~~~~~~ e Downloaded from a X I on July 26, 2018 by guest L L 7 e 9 10 I 1 12 WAVE LENGTH IN MICRON$ i Figure 2. Infrared spectra from 7 to 12 paof Escherichia coli: a-control, washed cells harvested from nutrient agar; b-residue after extraction with cold trichloroacetic acid; c-residue after extraction with hot trichloroacetic acid; d-residue after phenol extraction.

3 12 SEYMOUR LEVINE ET AL. [VOL. 65 saccharide by hot dilute acetic acid (method of Freeman, 1942) or extraction of Boivin antigen by cold trichloroacetic acid, in each case followed by alcohol precipitation of the extract, yields a polysaccharide whose infrared spectrum is chiefly characterized by a deep band from 8.6 to 10 p. and protein) can be extracted from bacterial cells with water (Smith, 1938), and such a loss could account for the observed attenuation. Following the procedure of Schneider (1945), cold trichloroacetic acid has been used to remove polysaccharides and acid soluble phosphorus Figure S. Representative enteric bacteria incubated 18 hours at 37 C on nutrient agar: a-aerobacter aerogenes; b-escherichia coli; c-salmonella montevideo; d-shigella dy8enteriae, type 2. Similarly, phenol extraction (method of Palmer and Gerlough, 1940) removes proteins and leaves a polysaccharide residue with a very similar spectrum (figure 2d). Washing bacterial suspensions with water causes a slight attenuation of the 8.6 to 10 ; band. This effect is increased by heating. It is known that polysaccharide (unlike nucleic acid compounds, followed by hot trichloroacetic acid to extract nucleic acids. Infrared spectra (figure 2b) of the residues left by these treatments show, after the first stage, an attenuation of the 8.6 to 10 p band due to removal of polysaccharides and other acid soluble compounds. Nucleic acids. After the second stage of the Schneider procedure (to remove nucleic acids)

4 19531 INFRARED SPECTROPHOTOMETRY 13 there is marked attenuation of the band at 8.0 to The absorption band at 8.0 to 8.1 p shows a 8.1, and loss of the remainder of the band at definite progressive attenuation with increasing 8.6 to 10 p (figure 2c). This suggests that nucleic age of culture. This has been demonstrated for acid absorption is largely responsible for the E. coli and Aerobacter aerogenes grown on nu- 8.0 to 8.1 p band and partly responsible for the trient agar and glucose agar, and for numerous Downloaded from WAVE LENGTH IN MICRONS Figure 4. Effect of glucose in the growth medium on infrared spectra: a-salmonella paratyphi incubated 18 hours on nutrient agar; b-salmonella typhosa incubated 18 hours on nutrient agar; c-salmonella paratyphi incubated 18 hours on nutrient agar with 1 per cent glucose; d-salmonella typhosa incubated 18 hours on nutrient agar with 1 per cent glucose; e-salmonella paratyphi incubated 3 days on nutrient agar with 1 per cent glucose; f-salmonella typhosa incubated 3 days on nutrient agar with 1 per cent glucose. 8.6 to 10 p band. Similarly, extraction of nucleoproteins with 0.2 per cent NaOH destroys the 8.0 to 8.1 p band and attenuates the 8.6 to 10 I, absorption. Extraction with water does not influence the 8.0 to 8.1 p band. This is in accord with the experience that nucleic acids are not readily extractable by water (Belozersky, 1947). Salmonella species grown on nutrient agar, glucose agar, and SS agar. Belozersky (1947), Caspersson (1950), and others have presented data to indicate that this is exactly the fate of nucleic acid as a culture ages. Differentiation of enteric bacteria by means of infrared spectra. Stevenson and Bolduan (1952) on July 26, 2018 by guest

14 SEYMOUR LEVINE ET AL. [VoL. a5 in examining the spectra of various gram positive and gram negative bacteria have demonstrated considerable differences that may be useful for identification.

5 14 SEYMOUR LEVINE ET AL. [VoL. a5 in examining the spectra of various gram positive and gram negative bacteria have demonstrated considerable differences that may be useful for identification. We have examined nlmerous strains of Enterobacteriaceae grown on nutrient agar (Difco) and find that the spectral differences in this group are comparatively small but reproducible. The extent of such differences is illustrated in figure 3 in which is shown the 7 to 12 p region of A. aerogenes, E. coli, Salmonella montevideo, and Shigells dysnteriae. The most important differences are in the shape of the curve at 8.7 and at 10.3 p and in the relative depths of the 8.6 to 10 p and 8.0 to 8.1 p bands. Most strains of A. aerogeses and E. coli conform to the examples shown. However, intermediate spectral types do occur. The spectra of SalmwoneUa strains are quite constant. All of eighteen strains examined conform to the pattern of the illustration, which, it should be noted, is not distinguishable from the spectrum of A. aerogenes. Twelve strains of Shigella of various species have shown much more spectral variation, many of them resembling E. coli or assuming intermediate shapes. Further studies are in progress aimed at achieving both genus and species identification within the enteric group by infrared spectrophotometry. Effect of changes in the growth medium on the infrared spectrum. The incorporation into the growth medium of a utilizable carbohydrate often results in small but reproducible changes in the bacterial spectrum. Figure 4 shows the spectra (from 7 to 12 p) of Salmonella paratyphi and S. typhosa grown on plain nutrient agar and on the same medium with 1 per cent glucose. The spectra of both organisms grown on nutrient agar are identical. On glucose agar, however, changes appear which are more marked in the case of S. paratyphi. Since the type and extent of the ensuing changes are not the same in all organisms, they can be used for the study of species differences and presumably for identification. These changes vary with age of culture as demonstrated by the last two spectra of figure 4. The small absorption band at 8.65 p and the changes of shape between 9.2 and 9.7 p are considerably more evident after 3 days. The biochemical explanation of these observations is not immediately evident. It appears unlikely that these spectral effects are due to the inclusion of glucose in the bacterial growth either by simple diffusion or by mechanical means during the harvesting procedure. The different rates of occurrence of these changes among different organisms and the slowly increasing spectral changes with age could not be explained by diffusion or mechanical inclusion. In addition, almost identical effects are produced by concentrations of glucose ranging from 0.1 to 3 per cent, and these effects are not removed by washing the organisms with water. The addition to the growth medium of utilizable carbohydrates other than glucose also produces changes in the spectrum of bacteria. On the other hand, a nonutilizable sugar (like lactose in the case of SalmoneUa species) never produces any effect on the spectrum. DISCUSSION Interpretation of infrared spectra of complex mixtures like bacterial cells is exceedingly difficult. The fact that pure nucleic acids absorb strongly at 8 p and that nucleic acids and polysaccharides both have strong bands between 9 and 10 p does not necessarily prove that they are responsible for the corresponding bands of the spectrum of whole bacterial cells. For example, it has been pointed out that the absorption spectrum of a substance contained in a cell may be influenced by interaction with neighboring molecules, and that several different substances may coincidently absorb at the same wavelength (Danielli, 1946). Nevertheless, the hypothesis is made likely by the behavior of these bands with aging of the culture and after extraction or fractionation. It is apparent that the absorption of substances present in small concentration can be masked by the absorption of other more plentiful compounds, or be below the sensitivity of the instrument. Consequently, a negative result (no detectable change in the spectrum) after a particular alteration of growth conditions or a particular chemical procedure carries little significance. On the other hand, a definite change in the spectrum must be significant. The biochemical interpretation of such changes depends on future research. The simplicity and rapidity of the procedure make it possible to examine many species and many strains of the same species grown on numerous media. Bio. chemical differences due to strain or to medium thus have been found which would have escaped notice unless exhaustive chemical analyses had

6 1953] INFRARED SPECTROPHOTOMETRY 15 been performed. These results point up the inadequacy of orthodox chemical analysis performed on only one or few strains and without rigorous control of the growth medium. The spectra shown in figure 4 demonstrate that this technique reveals metabolic changes induced by utilization of a carbohydrate. It is impossible to say at this time whether these spectral effects are due to certain carbohydrate intermediates or end products, but the evidence points to their being a manifestation of some phase of bacterial metabolism. The data of other workers and some of the results given above indicate the usefulness of infrared spectrophotometry in fractionation and extraction procedures on bacteria. Stevenson and Bolduan (1952) have indicated the possbility of using this technique for diagnostic purposes, and this has been illustrated by our spectra of enteric bacteria. In addition, the results presented above demonstrate that problems of bacterial metabolism and biochemistry may be studied advantageously using infrared spectra of whole organisms. ACKNOWLEDGM1ENTS We wish to express our appreciation to Dr. Paul W. Kabler for his guidance and suggestions and to Robert H. Bordner and Paul A. Schmitz for their technical assistance. Dr. P. R. Edwards and Dr. W. H. Ewing of the Communicable Disease Center kindly furnished many of the cultures and diagnostic sera required for this work. SUMMARY The value of infrared spectrophotometry in the study of bacteria is demonstrated. Evidence is presented, based on the infrared spectra of bacterial extracts and fractions, to indicate that the 8.0 to 8.1 band of the whole bacterial spectrum is due in large part to nucleic acid. The broad band from 8.6 to 10.0 is apparently due to both nucleic acids and carbohydrates. The infrared spectra of Aerobacter, Escherichia, SalmoneUa, and Shigella species are presented and the possibilities for differentiating them are indicated. The effect on the spectrum of variation in the growth medium is demonstrated. Evidence is presented to indicate that the spectral changes due to growth on glucose agar are related to a metabolic process. REFERENCES BARER, R., CoiE, A. R. H., AND THOMPSON, H. W Infrared spectroscopy with the reflecting microscope in physics, chemistry and biology. Nature, 163, BELOZERSKY, A. N On the nucleoproteins and polynucleotides of certain bacteria. Cold Spring Harbor Symposia Quant. Biol., 12, 1-6. BLOUT, E. R., AND MELLORS, R. C Infrared spectra of tissues. Science, 110, CASPERSSON, T Cell growth and cell function. W. W. Norton & Co., New York, N. Y. DANIELLI, J. F Establishment of cytochemical techniques. Nature, 157, FREEMAN, G. G The preparation and properties of a specific polysaccharide from Bact. typhosum Ty2. Biochem. J., 36, PALMER, J. W., AND GERLOUGH, T. D A simple method for preparing antigenic substances from the typhoid bacillus. Science, 92, RANDALL, H. M., FOWLER, R. G., FUSON, N., AND DANGL, J. R Infrared determination of organic structures. D. Van Nostrand Co., Inc., New York, N. Y. RANDALL, H. M., SmITH, D. W., CoLm, A. C., AND NUNGESTER, W. J Correlation of biologic properties of strains of Mycobacterium with infrared spectrums. I. Reproducibility of extracts of M. tuberculo8i8 as determined by infrared spectroscopy. Am. Rev. Tuberc., 63, SCHNEIDER, W. C Phosphorus compounds in animal tissues. I. Extraction and estimation of desoxypentose nucleic acid and of pentose nucleic acid. J. Biol. Chem., 161, SCHWARZ, H. P., RIGGS, H. E., GLICK, C., CAM- ERON, W., BEYER, E., JAFFE, B., TROMBETTA, L Infrared spectroscopy of tissues. Effect of insulin shock. Proc. Soc. Exptl. Biol. Med., 76, SMITH, E Heat stability and serologic activity of toxic extracts of the typhoid bacillus. J. Infectious Diseases, 63, STEVENSON, H. J. R., AND BOLDUAN, 0. E. A Infrared spectrophotometry as a means for identification of bacteria. Science, 116,

STUDIES OF CRYPTOCOCCUS POLYSACCHARIDES BY INFRARED SPECTROPHOTOMETRY

STUDIES OF CRYPTOCOCCUS POLYSACCHARIDES BY INFRARED SPECTROPHOTOMETRY SEYMOUR LEVINE, E. EDWARD EVANS, AND PAUL W. KABLER From the St. Francis Hospital, Jersey City, New Jersey; the University of Alabama