A cooling chain for studies of cryofixed biological specimens by scanning transmission electron microscopy and X-ray microanalysis

Ultramicroscopy 6 (1981) 181-186 North-Holland Publishing Company

181

A COOLING CHAIN FOR STUDIES OF CRYOFIXED BIOLOGICAL SPECIMENS BY SCANNING TRANSMISSION ELECTRON MICROSCOPY AND X-RAY MICROANALYSIS K. ZIEROLD, R. K()NIG, K.-H. OLECH, D. SCH)~FER, D.W. LUBBERS, K.-H. MULLER * and H. WINTER * Max-Planck-bTstitut ftlr Systernphysiologie, Rheinlanddamm 201, D-4600 Dortmund 1, West Germany

Received 18 December 1980

A cooling chain is described which enables the transfer of frozen hydrated biological specimens (ultrathin cryosections as well as about 1 tam thick cultured ceils) from a cryoultramicrotome into a scanning transmission electron microscope with a field emission gun. Transfer is done at I 18 K, specimen temperature in the microscope is 165 K. Sublimation processes are controlled visually and by mass spectrometry. Electron micrographs and X-ray microanalytical spectra of cryofixed unstained tissue culture ceils and rat liver tissue sections are described and discussed. Contamination of the specimen is much reduced by use of the cold stage.

1. Introduction Living cells and tissues cannot be studied in an electron microscope. The well-known success of electron microscopy for cell biology is based on techniques of chemical fixation, staining and dehydration. Nevertheless alterations of ultrastructure and composition due to this conventional preparation procedure cannot be excluded [1,2], especially physiologically important diffusible substances as water and ions are washed out. One approach to bypass these preparative difficulties consists in the following concept: the biological object (tissue, single cells, cell culture) is quickly frozen (cryof'txation) without any pretreatment and then, if necessary, sectioned at low temperature (<193 K). The so obtained cryospecimen is transferred into the electron microscope in the cold state and studied there by an electron beam in the frozen-hydrated or at least freeze-dried state. In this paper we will not discuss such aspects of this preparation method as cryofixation [ 3 - 6 ] or cryosectioning [7,8] but concentrate on the transfer of the frozen-hydrated specimen into a scanning transmission * Siemens AG. 0 304-3991/81/0000-0000/$02.50 © North-Holland

electron microscope and on the analysis in the cold state. First cryotransfer systems for transmission electron microscopes were developed some years ago [ 9 11 ] but due to the difficulties of cryosectioning this preparation conception was mainly applied in scanning electron microscopy [ 1 2 - 1 5 ] . In the last few years cryotransfer systems were described and discussed for scanning transmission electron microscopes [ 1 6 - 1 8 ] . By means of this paper we want to describe a cooling chain between a cryoultrarnicrotome and a scanning transmission electron microscope (STEM) with a field emission gun and a cryostage for ultrastructural as well as X-ray microanalytical studies.

2. Transfer procedure The transfer o f the frozen-hydrated speeirnen into the Siemens Elmiskop ST 100 F scanning transmission electron microscope developed by us is illustrated in figs. 1 - 6 : it starts in a modified FC4 cryochamber o f the Reichert cryoultramicrotome, as shown in fig. 1. The specimen support consists of a fiat and elastic metal tongue with a cavity for the reception of the grid and a screw thread to be screwed to the transfer rod. It is placed in a dovetail-shaped

182

K. Zierold et al. / Cooling chain to study cryofixed biological specimens

Fig. 1. Preparation for specimen transfer in the FC 4 cryoultramicrotome chamber. Notation: b = copper braid, br = bearing points of cooling rod, c = cooling ring, ch = transfer chamber, ct = cold trap, g = grid holder, gr = grid, i = glass components for thermal insulation, k = knife of cryoultramicrotome, n = vessel for liquid nitrogen, p = preparation block, pp = pole piece, r = cooling rod, s = specimen support, st = goniometer stage, t = transfer rod, vc = slide valve of transfer chamber, vm = slide valve of microscope, x = X-ray detector.

Fig. 3. Transfer chamber on a special transfer car. For notation, see fig. 1.

metal block. The grid is fastened in the specimen support by a snap ring. The temperature in the gaseous nitrogen atmosphere is below 173 K. After m o u n t i n g the grid with the specimen on the specimen support, the transfer rod is screwed on and the specimen support is w i t h d r a w n into the transfer chamber (fig. 2). In the transfer chamber the specimen temperature is maintained below 118 K by contact with metal jaws cooled by liquid nitrogen. After the

slide valve is closed, the transfer chamber (fig. 3) is brought to the Elmiskop ST 100 F with a m o d i f i e d specimen entry (fig. 4). The transfer c h a m b e r is connected to the microscope e n t r y and evacuated along with the intermediate space b e t w e e n b o t h slide valves o f microscope and transfer chamber (fig. 5). If necessary the specimen can be coated with a carbon a n d / o r metal layer in an evaporation e q u i p m e n t with a fixed stage, before the entry into the electron microscope. A f t e r evacuation the slide valves are o p e n e d and the specimen support is slid into the cold stage by the transfer rod. The transfer rod is screwed o f f and withdrawn, the slide valve o f the microscope is closed and the specimen can be studied by means o f the electron beam.

Fig. 2. Transfer chamber connected to the FC 4 cryoultramicrotome. For notation, see fig. 1.

Fig. 4. Modified specimen entry of the Elmiskop ST 100 F. For notation, see fig. 1.

183

K. Zierold et al. / Cooling chain to study co, ofixed biological specimens (3 r3 ,,...~--T = 12 0 K

e

°

--:ql L--- c, s,__q-F_

Fig. 5. Transfer chamber connected to the Elmiskop ST 100 F. For notation, see fig. 1. The cold stage of the Elmiskop ST 100 F (figs. 6 and 7) consists of a goniometer stage with a hollow coolb~g rod which receives the specimen support through a dovetailed channel. The temperature of 165 K at the specimen is achieved by contact of the cooling rod with a cooling ring which is connected to a liquid nitrogen Dewar outside the microscope over a copper braid and rod. Thermal insulation of the cooling rod against the warm microscope environment is performed by glass components. The specimen can be tilted to -+45° and can be moved in the three perpendicular dimensions by means of a specimen support which is inserted in the compound table. A ring shaped cold trap is situated above the specimen in order to reduce contamination of the specimen. Schematical drawings in figs. 7a and

,.

/ J 7" -

t /

41' i

Fig. 6. Cold stage with specimen support of the Elmiskop ST 100 F. For notation, see fig. 1.

-

.s,

Fig. 7. Schematic drawings of the cold stage with regard to thermal contact and thermal insulation, as seen from above (a) and from side (b). Hatched areas are glass components for thermal insulation. For notation, see fig. 1.

7b display the geometrical arrangement and thermal situation of the cold stage. Temperatures were measured at different points by means of thermocouples.

3. STEM of cryofixed specimens Tissue culture cells and cryosections of rat liver tissue were transferred in the way described above into the scanning transmission electron microscope. The sublimation of ice from the specimen was monitored by direct observation of bright and dark field signals of transmitted electrons and by measuring the partial water vapor pressure in the microscope vacuum with a quadrupole mass spectrometer. No sublimation of water from the frozen-hydrated specimen was observed in an electron beam of I00 kV and approximately 10- 8 A. The scanned area varied between 1 mm and about 50 run in diameter. Water sublimation from the specimen occurred by irradiation of the specimen with ultraviolet light [19] for 1 - 2 h or by removing the specimen from the cooling stage for some minutes, which could be determined by visual control and mass spectrometry. Whereas contrast and X-ray spectra of fully hydrated biological material were always low, they both were enhanced after at least partial dehydration. Image contrast was improved by mixing (partly addition and division) of bright field and dark field signal.

184

K. Zierold et al. / Coolhlg chain to stud), co,otT.xed biological specimens

I ,

_

S lOOkV

•

,f4

32s

2,Sum

80

1423

Fig. 8. Glioma tissue culture cells in the frozen-hydrated state. X-ray spectra a and b are obtained from sites a and b in the micrograph.

Figs. 8 and 9 show micrographs of glioma tissue culture cells with intracellular and extracellular X-ray spectra which were recorded by an energy dispersive X-ray spectrometer (USC nuclear semiconductor) connected to a multichannel analyser {Link Systems) before and after dehydration in the microscope. The cells were grown on Pioloform-coated [20] golden grids, frozen in liquid nitrogen cooled propane and

transferred to the microscope as described above. Au peaks in the X-ray spectra result from the gold grids and interfere sometimes with pkosphorus in the ceils. Unstained cryofixed cells tip to about 1/am thickness can be studied in this way. Figs. 10 and 11 show cryosections of rat liver pieces of about 1 mm in diameter shock frozen in liquid propaue. The 90 nm thick sections were obtained at

Fig. 9. Glioma tissue culture cells in the freeze-dried state. X-ray spectra a and b are obtained from sites a and b in the micrograph, c = cytoplasm, n = nucleus.

K. Zierold et al. / Cooling chain to study co'ofixed biological specimens

185

Fig. 10. Cryosection of rat liver tissue, fixed with glutaraldehyde, unstained. Inset shows X-ray spectrum of cytoplasm, c = cytoplasm, ER = endoplasmatic reticulum, m = mitochondrium, n = nucleus. X-ray spectrum obtained from the cytoplasm.

I

m

m

m

l

Fig. 11. Cryosection of rat liver tissue, unfixed, unstained. X-ray spectra a and b are obtained from the sites indicated in the micrograph, c = cytoplasm, ER = endoplasmatic reticulum, h = holes in the section, presumably by cracking during sectioning, m = mitochondrium, n = nucleus.

186

K. Zierold et al. / Cooling chain to stud), cryofixed biological specimens

173 K and sandwiched between two Pioloform-coated and carbon-evaporated copper grids. The specifnen of fig. 10 was prefixed with 2.5% glutaraldehyde before cryofixation, and therefore ice crystal distortions are very small. Nucleus, mitochondria and endoplasmatic reticulum can be identified. As shown by the X-ray spectrum (inset of fig. 10) diffusible substances are washed out by chemical fixation. Therefore analytical studies on an ultrastructural level need cryofixation without any pretreatment [21]. Fig. 11 shows a cryosection of chemically untreated fresh liver tissue. Ice crystals up to about 0.2 ~tm in diameter have partly damaged the ultrastructure. Nevertheless main subcellular details can be recognized. X-ray spectra demonstrate the inhomogeneous distribution of elements in different compartments o f the cell. Presumably diffusible ions like chlorine, potassium and calcium are detected. In this case it is not proved that no water has sublimed from the section. Partial dehydration o f sections in figs. 10 and 11 cannot be excluded. Finally it should be mentioned that specimen contamination due to electron exposure can be greatly reduced by use of the described cold stage, but cannot be prevented as seen by contamination spots after high electron exposures used for X-ray microanalysis.

4. Conclusion The described cooling chain enables the transfer o f frozen hydrated biological specimens from a cryoultramicrotome into a scanning transmission electron microscope. In the STEM mode ultrathin cryosections can be studied as well as cultured ceils of about 1 pan thickness without chemical fixation, cryoprotection or staining. Electron microscopical contrast and X~ray microanalytical spectra are poor in the frozen hydrated state, but they are improved by at least partial freeze-drying o f the specimen. Whereas morphological contrast and resolution can be optimized b y electronic manipulation and combination o f bright and dark field signal, quantitative X-ray microanalysis of liquid spaces in tissues is not yet possible at the present time. Nevertheless this aim will be pursued in the future on the basis o f the described cooling chain technique. The reduction o f contamination o f the specimen by use of the cold stage is very promising for studies o f electron-sensitive objects at low temperature.

Acknowledgement The authors want to thank Dr. F. Pietruschka for the tissue culture cells and Miss S. Dongard, Mr. U. Babst, Mr. W. Briiseke, Mr. F. Sieland and Mr. F.K. Ziegler for technical assistance.

References [1] N. Weissenfels, D. Sch~er and M. Hfindgen, Cytobiol. 3 (1971) 188. [2] M. Hiindgen, D. Sch~iferand N. Weissenfels, Cytobiol. 3 (1971) 202. [3] M. Moor, Z. Zellforsch. 62 (1964) 546. [4] M.J. CosteUo and J.M. Corless, J: Microsc. 112 (1978) 17. [5] H. Plattner, Arzneimittelforsch./Drug Res. 29 (1979) 1809. [6] K. Zierold, Microsc. Acta 83 (1980) 25. [7] L. Sev~us, J. Microsc. 112 (1978) 269. [8] H. Sitte, K. Neumann, H. Hffssig,H. Kleber and G. Kappl, in: Proc. 7th Europ. C0ngr. on Electron Microscopy, The Hague, 1980, Vol. 1, p. 540. [9] H.G. Heide and S. Grund, J. Ultrastruct. Res. 48 (1974) 259. [10] Th. E. Hutchinson, M. Bacaner, J. Broadhurst and J. LiUey, Rev. Sci. Instr. 45 (1974) 252. [11] A. Saubermann and P. Echlin, J. Microsc. 105 (1975) 155. [12] P. Echlin, Low Temperature Biological Scanning Electron Microscopy, in: Advanced Techniques in Biological Electron Microscopy II, Ed. J.K. Koehler (Springer, Berlin, 1978) pp. 89-122. [13] W. Fuchs and B. Lindemann, J. Microsc. Biol. Cell. 22 (1975) 227. [14] A.T. Marshall, Microsc. Acta 79 (1977) 254. [15 ]. K. Zierold, in: Proc. 6th Europ. Congr. on Electron Microscopy, Jerusalem, 1976, Vol. 2, p. 233. [16] R. Freeman, F. Booy and K. Leonard, in: Proc. 7th Europ. Congr. on Electron Microscopy, The Hague, 1980, Vol. 2, p. 650. [17] S. Lichtenegger and W.M.A. Hax, in: Proc. 7th Europ. Congr. on Electron Microscopy, The Hague, 1980, Vol. 2, p. 652. [18] K. Zierold, R. KSnig, K.-H. Olech, D.W. L/ibbers, K.-H. M~ller, D. ScNi'fer and H. Winter, in: Proc. 7th Europ. Congr. on Electron Microscopy, The Hague, 1980, Vol. 2, p. 786. [19] B.W. Payne and E.D.F. Williams, in: Proc. Electron Microscopy Soc. of Southern Africa, Vol. 8 (Pretoria, 1978) p. 9. [201 W. Stockem, Mikroskopie 26 (1970) 185. [21] K. Zierold, D. Scl~'fer and J. Gullasch, Microsc. Acta Suppl. 2 (1978) 92.