A Simple and User-Friendly Method for High-Quality Preparation of Pollen Grains for Scanning Electron Microscopy (SEM).

Abstract

Pollen is becoming an increasingly important subject for molecular researchers in genetic engineering, plant breeding, and environmental monitoring. To broaden the scope of these studies, it is essential to develop accessible methods for scientists who are not specialized in palynology. The article presents a simplified technical procedure for preparing pollen grains for scanning electron microscopy (SEM). The protocol is convenient for any molecular laboratory due to its small set of reagents, ease of execution, low cost, does not require special equipment, and takes only one hour to complete. The high penetrating ability of formaldehyde and the final delicate dehydration using hexamethyldisilazane (HMDS) instead of critical point drying allow for sufficient preservation of the architecture of the aperture, which is considered a gateway for the passage of biomolecules. The method was successfully applied to pollen grains of representatives of dicotyledons (beetroot, petunia, radish, tomato and tobacco) and monocotyledons (lily, onion, corn, rye and wheat). Species studied included insect-pollinated (entomophilous) and wind-pollinated (anemophilous) species. A comparative analysis of the sizes of fresh living pollen grains under a light microscope and those prepared for SEM showed some shrinkage. Quantitative analysis of the degree of pollen grain shrinkage showed that this process depends on the initial shape of dry pollen grains, and the number and structure of apertures. The results support the theoretical model of the folding/unfolding pathways of pollen grains.

Plant phenotyping relevance

植物花粉のSEM観察用試料調製法そのものを開発し、複数植物で適用・比較検証しているため、形態計測に関する中心的な方法論研究である。

Code and data availability

The paper's quantitative pollen shrinkage measurements (Table S1) and light microscopy images (Figures S3–S4) are contained in the publicly downloadable MDPI Supplementary Materials, which directly reproduce this paper's phenotyping measurements. No author analysis code or trained models are mentioned.