Orthology of NFP genes in nodulating plant species is consistent with either single or multiple origins of nodulation symbiosis

A recent paper (1) seeks to test the competing hypotheses of a single origin vs. multiple convergent origins of nodulation—nitrogen fixing symbiosis between flowering plants and bacteria housed in structures (nodules) produced by some members of the “Nitrogen Fixing Clade” (NFC)—by reconstructing an updated phylogeny of NFP, a key gene family in many nodulating species. The authors make four claims: 1) The NFC ancestor had a single copy NFP gene, whose duplications in legumes and in the Cucurbitales-Rosales (CuRo) clade were “nonhomologous” (Fig. 1); 2) therefore, “MtNFP and the NFP2-like group are not orthologous”; 3) this provides “evidence for molecular convergence of NFP”; and 4) “Most importantly, these results indicate convergent evolution of NFP-based signaling with nitrogen-fixing bacteria … in contrast to the single-origin hypothesis of nodulation, for which orthology of symbiotic signaling and nodule initiation genes is the most compelling body of evidence.”

Fig. 1.

NFP gene tree, modified from figure 1B of Finegan et al. (1) reduced to show only major NFC clades created by the species phylogeny and three independent gene duplications (“D1”–“D3”). Duplications “D1” and “D2” in the Fabales family Fabaceae (orange highlighting) are unique to subfamilies Papilionoideae (producing paralogs NFPp1 and NFPp2) and Caesalpinioideae (producing paralogs NFPc1 and NFPc2), respectively; because they are independent of one another they meet the definition of co-orthologs (2): “Two or more genes in one lineage that are, collectively, orthologous to one or more genes in another lineage due to a lineage-specific duplication(s).” Duplication “D3” occurs in the common ancestor of the orders Cucurbitales and Rosales, producing paralogs NFP1 and NFP2 (1). Because duplication “D3” is independent of both duplications “D1” and “D2,” NFP1 and NFP2 in Cucurbitales/Rosales are co-orthologous with Fabaceae NFP genes.

Their phylogeny, however, clearly shows legume and CuRos NFP genes as orthologs (Fig. 1) and they refer to Parasponia paralogs as co-orthologs of a legume NFP (MtNFP) (2). In their third claim, “molecular convergence” refers to the independence of the tandem duplication in CuRo from duplications in legumes. NFP function would be convergent only if their unproved fourth claim is assumed: that recruitment of NFP for nodulation was independent in CuRo and legumes, and thus nodulation symbioses evolved convergently in the NFC.

Additionally, if gene orthology was “the most compelling body of evidence” for a single origin of nodulation, then their demonstration of NFP orthology would in fact support the single origin hypothesis over multiple origins. However, recruitment of orthologous genes is consistent with either hypothesis (Fig. 2), and paralogy of recruited genes is not an infallible criterion for independent origins (3), as has been demonstrated recently in grass species that likely share a common origin of C₄ photosynthesis, but “use different paralogous loci to achieve cell-type-specific expression” (4).

Fig. 2.

Paralogy/orthology as a criterion for inferring nodulation origins. (A) Phylogeny of five species, two of which (blue text) nodulate. Nodulation symbiosis evolved either a single time in a shared ancestor or independently (convergently), involving assembly of various modules through recruitment of genes for novel functions in the new symbiosis. In (B–D), an ancestral duplication has produced two paralogs, A and B. In (B) nodule-expressed copies are paralogous (boldface A in species 3 and B in species 5), consistent with independent gene recruitments (black arrows pointing to different nodule types) during convergent assemblies of nodulation. In (C and D) orthologous copies are expressed in nodules of the two species, consistent either with a single recruitment during a single assembly of nodulation in an ancestor [arrow in (C)] and nodule divergence (colored arrows), or independent recruitments of the same ortholog during convergent assemblies of nodulation (D).

In fact, stronger evidence for a single origin of nodulation is the loss or inactivation of nodulation genes in nonnodulating taxa. In the single origin hypothesis, such taxa inherited the ability to nodulate from the NFC common ancestor but ceased to engage in this expensive, optional, plastic symbiosis when decreasing global atmospheric CO₂ increased the cost of carbon (5, 6). Finegan et al. illustrate precisely such a “massive pseudogenization of NFP2 in Rosales” that do not nodulate (1), although they suggest that these genes were inactivated independently in many lineages because they “would have been redundant without co-option and gain of nodulation.” The single origin explanation is simpler: When nodulation ceased so did NFP2 expression, leading to relaxation of purifying selection and eventual decay (5, 6).

Finegan et al. (1) have produced an updated phylogeny of a key nodulation gene, but they have not provided evidence that tests the competing hypotheses of single vs. multiple assemblies of nodulation symbiosis (5–7). The evolution of this complex gene family, critical in many nodulation symbioses, deserves further elucidation, and as more species are studied, these genes may yet shed light on the question of nodulation origin(s).

Data, Materials, and Software Availability

There are no data underlying this work.