Localization of arabinogalactan protein 6 fused with Sirius ultramarine fluorescent protein in Arabidopsis pollen and pollen tubes

Abstract

Following studies on the transcriptome of pollen tubes of an agp6 agp11 Arabidopsis double null mutant, together with the knowledge that arabinogalactan protein (AGP) 6 is important for male gametogenesis and pollen germination, we sought to know whether AGP6 could be present in the vegetative cell wall or in the generative cell wall or in both. To that end, Arabidopsis plants were transformed with AGP6 gene sequence fused with Sirius fluorescent protein. Fluorescent blue light emission could be detected in the vegetative cell wall only. This result supports the model by which AGP6 and other pollen AGPs are probably important signaling molecules at the pollen tube apex.

Arabinogalactan proteins (AGPs) are massively glycosylated plant proteins whose polysaccharide chains are composed predominantly but not exclusively of arabinose and galactose, and are O-linked to Hyp residues of the polypeptide chain. An added distinguishing feature of this class of proteins is the presence of a complex lipid anchor, the glycosylphosphatidylinositol (GPI) anchor, which holds these molecules to the exoplasmic side of the plasma membrane. AGPs can be found in virtually every cell throughout the plant kingdom. Their cytological localization, besides the plasma membrane, is mainly the cell wall, extracellular space, and sometimes they are also found abundantly in secretions. They have been functionally correlated with several fundamental biological phenomena, such as cell expansion and division, cell death, plant defense, pollen tube growth, and guidance, among others, but a definitive mode of action has not yet been established.¹

As a result of studies in our group on Arabidopsis pistil and pollen development with anti-AGP monoclonal antibodies,^2,3 it has become evident that AGPs could be used as fine molecular markers of gametophytic cell differentiation. For instance, JIM8, an anti-AGP monoclonal antibody, was found to label the generative cell only, with high specificity and precision, in bicellular pollen.³ Anti-AGP monoclonal antibodies, however, bind to sugar epitopes present in different AGP molecules, and consequently an analysis of the role of individual arabinogalactan gene products in a particular cell type or developmental stage is not possible by such approach.

AGP6 and AGP11 are Arabidopsis pollen-specific AGPs, which, among others, have been a major interest in our laboratory. These proteins were found to be important for male gametophytic viability and timely germination.^4,5 We were therefore keen on knowing whether AGP6 epitopes were present in the generative cell wall or in the vegetative cell wall only.

To that end, a custom-made synthetic construct based on a fusion between AGP6 and Sirius fluorescent protein⁶ was purchased from GenScript USA Inc. The construct included, from 5′ to 3′ direction: 1) AGP6 promoter region plus 5′ untranslated region (771 nt in total), 2) start codon followed by the predicted signal peptide of AGP6 (63 nt), 3) the Sirius fluorescent protein coding sequence with codon optimization for plant expression, 4) a 6 amino acid spacer (ASGGGA), 5) the AGP6 mature peptide coding sequence (393 nt). The whole DNA sequence was flanked by Gateway attL sites for insertion into a suitable plant destination vector. A shorter version of the sequence described but without the promoter region, was isolated by PCR and the resultant DNA fragment was inserted into a pET30a plasmid, for IPTG-induced protein expression in E. coli. Bacterial colonies were bright blue under the fluorescence microscope indicating conformational viability of the Sirius::AGP6 fusion polypeptide.

A Gateway expression vector containing the whole DNA sequence was introduced into wild type Arabidopsis by the floral dip method. Seeds from basta-resistant plants were subsequently grown and the respective pollen was collected and analyzed under the confocal fluorescence microscope. Blue fluorescence was detected in approximately half of the inspected pollen grains, as would be expected for hemizygous individuals (Fig. 1C and D). Higher magnifications of germinated pollen showed blue fluorescence along the vegetative cell wall, with no indications whatsoever of blue light emission associated with the generative cell (Fig. 1E).

Figure 1. Localization of AGP6 in Arabidopsis pollen (A-D) and pollen tube cell wall (E), using a sirius fluorescence protein-based fusion (F). Blue fluorescent light was found associated with the vegetative cell wall only. (A) Bright field microscopy image; (B and E) Confocal optical microscopy images; (C) Bright field and confocal microscopy images combined; (D) Higher magnification of combined image; attL – Gateway recombination sequences; CDS – coding sequence; SP – signal peptide; UTR – untranslated sequence. Scale bars in (A), (B), and (C), 150 μm, in (D), 80 μm, and in (E), 40 μm.

Very recently, we performed microarrays with the double mutant agp6 agp11 pollen tube as well as yeast two-hybrid assays with the AGP6 and AGP11.⁷ Efforts to understand the biological function of AGPs have been persistent because many members of the class seem to regulate many important aspects of plant growth, development, and resistance to different types of stress. The results obtained in those studies provided possible interactors for these AGPs that have reinforced their roles as important structural proteins for pollen grain and pollen tube, knowing that AGPs are complex supramolecular scaffolds which can contribute to the reinforcement of cell walls. We also propose that AGP6 and AGP11 are most likely important signaling molecules at the pollen tube apex, being recycled back again to the apex as the pollen tube grows, by a clathrin mediated endocytosis pathway. Besides, GPI-anchored proteins are widespread cell sensors in mammals, especially during egg-sperm communication.

The choice of Sirius fluorescent protein, a GFP variant obtained after random mutagenesis of the amino acids surrounding the chromophore,⁶ was used in the present work for its alleged pH insensitivity, and as far as we know was used for the first time in plant cells or tissues. This should potentially be advantageous in studies of the plant cell wall, which is known to be an acidic environment. However, a major drawback for its wider use is its extremely low absorption peak (355 nm), which is not compatible with most light sources of confocal fluorescence microscopes.

Disclosure of Potential Conflicts of Interest

No potential conflicts of interest were disclosed.