Rationale: The rapid domestication of novel crops will permit development of sustainable primary production systems adapted to a wider variety of environments. This approach has recently become feasible and now permits to exploit botanic diversity, including that of seed-free plants.
Azolla species are seed free-ferns and heterosporous. Moreover, they constitute symbioses dominated by the cyanobacteria Nostoc azollae. They, therefore, present a formidable challenge for domestication. To tackle this challenge we will need conceptually novel research in development, physiology, biochemistry, genetics and ecology.
Nevertheless, domestication of Azolla ferns, or transfer of its dinitrogen fixating capabilities to crop plants, is important because the ferns hold the secret to nitrogen fertilizer-independent plant protein production. Moreover, their prolific floating mats could be used to build substrate in subsiding delta regions, or to re-cycle phosphate complexed in ditch sediments. They may also be employed for re-circularization of nutrients on farm, or where ever large amounts of liquid nutrient-rich fractions accumulate.
Below a few aspects of the Azolla biology that have kept us preoccupied towards Azolla domestication since 2013.
An Aquatic symbiosis model organism that colonized the water surface
How to culture? Cultures of Azolla filiculoides and A. pinnata on defined medium can be maintained by continuous harvest and addition of nutrients for long periods of time. They require no nitrogen and reach 50 t DW h-1 a-1; plant protein makes up some 20% by dry weight of the biomass. They require far-red light in the light spectrum for optimum growth (Brouwer et al., 2018 https://doi.org/10.1002/jsfa.9016).
How to store? Clumps of pre-dried mega- and microsporocarps from A. filiculoides may be cryopreserved (Brouwer et al., 2014 https://doi.org/10.1111/nph.12708).
How to cross? For A. filiculoides, gently shake megasporocarp and microsporocarp containing filtrate in destilled water, transfer clumps to destilled water of shallow depth, and keep under light at ambient temperature (Brouwer et al., 2014 https://doi.org/10.1111/nph.12708).
Characteristic lipids and waxes. The symbiosis accumulates characteristic long-chain lipids (Brouwer et al., 2016 https://doi.org/10.1007/s12155-015-9665-3). Long chain mid-chain ω20-hydroxy compounds (ω20-alkanols, 1,ω20-diols, ω20-hydroxy fatty acids) and structurally related ω9,ω10-dihydroxy compounds (ω9,ω10-diols, 1,ω9,ω10-triols and ω9,ω10-dihydroxy fatty acids) are typically found in Azolla species. These very long chain fatty acid (VLCFA) derivatives occur in the ferns’ waxes in free and esterified form (Nierop et al., 2018 https://doi.org/10.1016/j.orggeochem.2018.09.014). Azolla ferns synthesize hopenes with hopene synthases from cyanobacterial origin (Li et al., 2018, supplement https://doi.org/10.1038/s41477-018-0188-8).
Azolla ferns with their microbiome dominated by Nostoc azollae enter the age of meta- and pangenomics
The fern transcripts and genome. A first assembly and annotation of the A. filiculoides genome was published in 2018. (Li et al., 2018 https://doi.org/10.1038/s41477-018-0188-8), an improved one in 2024 (Güngör et al. 2024 https://doi.org/10.1111/pce.14907).
The metagenomes of six species have been sequenced using short read technology. Persistent bacteria are found associated with the Azolla species in addition to the N. azollae; they belong to the Rhizobiales and their genomes encode complementing parts of the denitrification pathway (Dijkhuizen et al., 2018 https://doi.org/10.1111/nph.14843). The metagenomes are now used for pangenome comparisons.
A by-product of sequencing efforts has further been the realization that taxonomic assignments for the Azolla fern species may need to be revised, as in the case of some accessions sampled from the Anzali lagoon (Dijkhuizen et al., 2020 https://doi.org/10.1101/2020.09.09.289736).
To understand the functions of the genes discovered in the Azolla metagenomes, we now need to develop genome editing tools for all members of the symbiosis, we developped RNA-guided transposition for cyanobacteria (Arevalo et al., 2024 https://doi.org/10.1021/acssynbio.3c00583) and are testing the method for N. azollae from the shoot apical colonies of the fern hosts, carrying out conjugation experiments on modified shoot apices.
Control over the Azolla symbiosis life cycle for dissemination, storage and pre-breeding
How to control the induction of mega and microsporocarps? Mat density and far-red light are key to transit into sexual reproduction for A. filiculoides (Dijkhuizen et al., 2020 https://doi.org/10.1101/2020.09.09.289736), but did not work for A. anzali, and A. pinnata.
Towards the domestication of seed-free plants. Gene expression profiling complemented by small RNA profiling uncovered networks active during Azolla sporocarp induction that are also active when sporangia of other ferns are initiated; the genes encoding the network components are phylogenetically related to those from the transition to sexual reproduction in seed plants. Further characterization of the networks and how they may be connected to mechanisms sensing the environments will help to more generally understand the evolution of phase transitions in seed-free plants.
Domestication for applications in the circular economy
Can the wildtype biomass be used for protein extraction? We first tested the potential yields for the ferns: without any nitrogen fertilizer the A. filifculoides and A. pinnata could yield up to 50 ton dry weight per hectare annually, the biomass contained some 20% w/w protein with an essential amino acid content fit for use as feed (Brouwer et al., 2018 https://doi.org/10.1002/jsfa.9016). Tannins interfere with protein digestibility, low tech extraction of protein is possible (Brouwer et al., 2019 https://doi.org/10.1016/j.btre.2019.e00368).
What tannins accumulate in A. filiculoides and A. pinnata, and how are they biosynthesized? Tannins do not differ fundamentally in the two species, but A. pinnata accumulates more than A. filiculoides. Azolla ferns encode a functional LAR-enzyme which is very highly expressed. Gungor et al., 2020 https://doi.org/10.1111/nph.16896).
What is the distribution of abundant secondary metabolites in Azolla grown under ambient and cold temperatures? Metabolite imaging revealed the distinct spacial distribution of highly related secondary metabolites which implies their different functions (Gungor et al., 2024 https://doi.org/10.1111/pce.15010).
How are physiological states of the ferns related to protein quality? When stressed, the ferns are known to synthesize deoxyanthocyanins and turn into a red carpet, at the cost of protein content.
Is anoxia rapidly produced under the thick mats and what biogeochemical processes operate within the thick mat and underneath? Ecology studies found that Azolla mats promote anoxia and thence mine phosphate in flooded phosphate-rich agricultural soils (Vroom et al., 2024 https://doi.org/10.1016/j.watres.2024.121411).
Fundamental research on the communication between symbionts and fern host
Given the methodological advances in single-cell profiling and metabolite imaging we are now able to address this in ongoing work. We stumbled on the compound cornicinine, discovered in a cranefly from wetlands where Azolla ferns also thrive, Nephrotoma cornicinina. Cornicinine applied to Azolla fern fronds interfers with N. azollae cell specification in the leaf so as to turn the filaments into single celled akinetes, and when applied to germinating Azolla sporelings it stops the N. azollae akinetes from germinating (Güngör et al. 2024 https://doi.org/10.1111/pce.14907). Nevertheless, a real botleneck for this work, are the lacking genetic tools to adequately test the function of genes in the fern host...we are pursuing this but any collaboration on this would help greatly to develop Azolla as fully fledged models for floating plants exhibiting nitrogen autotrophy at even the highest growth rates.