Symbiosis between plants and cyanobacteria
Overview

The need for nitrogen-fixing symbioses
Nitrogen is a major limiting nutrient in agriculture (Fixon & West, 2002). The need may be met with chemical fertilizer, but that comes at a steep environmental cost (Vitousek et al, 1997) and is also a major expense that is too high for farmers in many underdeveloped countries (Dixon et al, 2001).

Biological nitrogen fixation addresses both of these concerns. Soil bacteria -- Rhizobia -- in symbiosis with legumes fix atmospheric N2 and provide all the nitrogen needs of the host plants. However these bacteria form symbioses with only narrow group of plants (Wang et al, 2018), not with the cereals (wheat, corn, rice) that make up the bulk of agriculture. In contrast, Nostoc punctiforme, a single strain of N-fixing cyanobacteria, forms symbioses with representatives of a broad range of plants (Santi et al, 2013) (Figure 1).

Why is it that Rhizobia are confined to a thin slice of plant hosts while Nostoc punctiforme is able to interact with representatives from a much more diverse fraction of life? One difference is readily apparent: unlike Rhizobia, Nostoc fixes nitrogen without the need of a special plant-provided environment. It does so by differentiating specialized cells, heterocysts, that protect the nitrogen-fixing apparatus from oxygen. But symbiosis is initiated by the plant, and few choose to do it. What makes these few plants special?


Figure 1. Hosts for cyanobacterial symbionts. (A) Lichen/fungus (from George H Daniel). (B) Azolla/fern (from Onyemeh Nwamarabia). (C) Cycas revoluta/gymnosperm (from Horticultural Impex). (D) Gunnera/angiosperm.
Specialization by Gunnera
If enough nitrogen is present to support growth, the angiosperm Gunnera manicata does not associate with cyanobacteria (Figure 1A). However under N-deprived conditions, the plant forms red, anthocyanin-rich stem glands (Figure 1B) that attract Nostoc (Figure 1C). Glands are formed solely in response to nitrogen-deprivation -- the presence of cyanobacteria is not necessary (Chiu et al, 2005). The cyanobacteria are taken into cells to form dense colonies that can provide all the fixed nitrogen needed for growth.

We know very little about how the gland differs from the rest of the plant. It induces Nostoc to differentiate into its motile state and then attracts the motile Nostoc, possibly using a simple sugar as the attractant (Khamar et al, 2010). Despite the presence of other bacteria, only Nostoc is taken up into gland cells. The gland is highly vascularized, perhaps to facilitate the dissemination of fixed nitrogen to the rest of the plant.

The gland also alters the state of the internalized Nostoc. The level of fixed nitrogen within gland cells is far higher than the amount that would normally shut down heterocyst differentiation and nitrogen fixation by Nostoc, yet the Nostoc within Gunnera, as within other host plants, differentiates heterocysts at a rate greater than normal and fixes prodigious amounts of nitrogen (Meeks & Elhai, 2002).

We therefore would like to know how the gland differs from nonglandular stem, with and without Nostoc present, and how Nostoc within the gland differs from free-living Nostoc. The interaction between the plant and cyanobacterium promises to be much simpler than the complex give and take between Rhizobia and legumes, making it easier to imagine how the trick of hosting a nitrogen-fixer, once we understand it, may be taught to crop plants.


Figure 2. Symbiotic gland of Gunnera manicata. Seedlings grown with (A) high nitrogen or (B) low nitrogen. (C) Expanded view of gland, showing green Nostoc on its surface. (D) Cross-section of gland, showing dark internal colonies of Nostoc. From Chiu et al (2005).