We have investigated the phenomenon of particle aggregation in a sample of 71,038 Escherichia coli B/r A cells in balanced exponential growth, during preparation for electron microscopy by agar filtration. The bacteria were photographed in a transmission electron microscope and the dimensions and spatial relationships among all the members of each aggregate were recorded using an interactive image processing system. The proportion of aggregated cells, 22%, is much greater than that found by direct count in a light microscope (7%), implying that most aggregation takes place during the preparation stages. The aggregated cells are about 1% narrower than the free cells, because of mutual compression, and 1.5% longer, because of a selection bias in favor of longer cells. From a statistical analysis of the data, we conclude that the clustering of cells into aggregates in the course of sample preparation is the result of random encounters during the settling on the collodion membrane and of the changing surface tension during the drying process. A method is proposed to correct morphometric measurements for the distortion caused by cellular aggregation of this kind.0 1992 Wiley-Liss, Inc.

ABSTRACT

We have investigated the phenomenon of shape distortion in a sample of 1,552 Escherichia coli Blr A cells in balanced exponential growth, during preparation for electron microscopy by agar filtration. Mixed preparations of bacterial cells and polystyrene latex spheres were shadow cast at low angle and the resulting shadows used to obtain quantitative estimates for the dimensions of the dehydrated cell; these then serve as a basis for a model of its shape in three dimensions. Key terms: Cell distortion, dehydration, dimensional changes, dryingAgar filtration is a common method of preparing Escherichia coli cells for electron microscopy in order to obtain size distributions of various kinds (3,10,17). This technique was first proposed by Kellenberger and Arber (6) and later improved by Woldringh et al. (16). Agar (2%) is made up in distilled water and dried to about 80% of its original weight. A perforated plastic membrane is superimposed by pouring a solution of 0.4% Parlodion (Mallinckrodt, St. Louis, MO) in amylacetate on the agar slant precooled below the dewpoint in order to induce the formation of tiny holes by water vapor condensation (2). Small drops of a suspension of cells fixed in 0.1-0.2% Os04 are deposited on the membrane and allowed to drain away; those that fail to do so within 60 min are rejected. The dimensions of a cell are determined by measuring the length and width of its projection in a transmission electron microscope (TEM).The deformation that must perforce be caused by the dehydration process (9,16) has cast doubt on the justification in representing the dimensions of a hydrated cell by the corresponding projections of its dehydrated counterpart (12), but such skepticism is far from universal (13,161. Moreover, the controversy cannot be settled simply by comparing TEM measurements with those obtained on living cells using light microscopy, even under phase contrast or Nomarski optics, because bacteria are very small (1.5 pm long by 0.5 pm in diameter, nominal dimensions) and any differences are expected to be well below effective optical resolution.In this article we attempt to estimate the effect of drying on bacterial cells, utilizing information inherent in the preparation itself. Low-angle shadow casting is used a t several morphologically defined locations to determine average values of cell height above the cellulose membrane. A model describing the structure of the dehydrated cell is then proposed, based on these results. Thus the difference between the mean width of aggregated and free cells is used to estimate the change in width that occurs during dehydration, whereas a n analysis of the fissures formed between cells aggregated during settling on the copper grid provides a n estimate for the change in length. The copper grids were photographed in focus on 70 mm film (SD-281, Kodak, Rochester, NY) using a TEM (EM300, Philips, Eindhoven, Netherlands) a t 60kV with a final magnification of ~3 , 6 0 0 ; exposure time was 1% s. MATERIALS AND METHODS

No abstract

A RFLP assay which had been developed for distinguishing between a hybrid pepper and its parents, has been converted into a simple applicable PCR assay, suitable for large-scale quality-control assessment of the commercial hybrid . For this conversion the sequences of the ends of the probe used in the previous RFLP assay were determined . From these sequences suitable primers were devised for inverse PCR of heterogeneous DNA fragments derived from the male parent . The inverse-PCR product was cloned and partially sequenced . These sequences, in turn, made it possible to determine primers on both sides of the locus of mutation, and to develop the reported conventional PCR assay .