Guanylyl cyclase activity in plants?

Cyclic nucleotides have defined signal transduction roles in animals, fungi, and other diverse eukaryote kingdoms. However, the situation in plants is still not clear after more than 3 decades of investigations reporting evidence for and against a regulatory role for cyclic nucleotides in higher plants. No sequenced higher plant genome contains homologs of the nucleotidyl cyclase genes that are recognizable in diverse eukaryotic kingdoms, including, curiously, the Chlorophyta phylum of the Viridiplantae kingdom. The report of Qi et al. (1) is a recent example of claims that recombinant plant proteins (or their fragments) from diverse protein families have guanylyl cyclase (GC) activity in vitro—claims that have led to the hope that the cyclic nucleotide regulatory system in plants is not a molecular house of cards. However, the GC activity reported was ≈150 fmol cGMP per μg of 14-kDa protein fragment (AtPepR1-GC) per 30 min. This activity is extraordinarily low. Each molecule of this putative recombinant GC synthesizes one cGMP molecule every 10 d. By contrast, recombinant and native human GC molecules synthesize 10³–10⁴ cGMP molecules per second, 10⁹- to 10¹⁰-fold faster (2). This very low activity is typical of the other putative noncanonical GCs reported in plants, including fragments of other plant receptor kinases cited by Qi et al. (1) and Ludidi and Gehring (AtGC1, ref. 3). GC comprises 0.005–0.1% of bovine lung protein (2), but even if AtPepR1 was 99% of total Arabidopsis protein the GC activity of the plant would be <10⁻⁶ of the activity in some animal tissues.

This GC activity could be caused by the recombinant AtPepR1-GC containing 0.0001% contaminating bacterial GC that binds to the recombinant fragment during the batchwise metal affinity chromatography. An accounting of the partitioning of GC activity in the existing purification steps and further high-resolution separation steps would confirm whether the minute GC activity is associated with the recombinant plant protein or is a host contaminant.

The 118-aa AtPepR1-GC fragment is part of the ≈300-aa protein kinase domain of AtPepR1. The 3D structure of Pto (4), a plant disease resistance protein kinase homolog of AtPepR1, as well as hundreds of homologous animal and fungal protein kinase structures, show that the AtPepR1-GC region is not a discrete domain and is remote from the kinase nucleotide binding site. Because of the missing complementary surfaces present in the complete protein kinase domain, it is unlikely that the recombinant AtPepR1-GC will fold in the same way as it does in the intact AtPepR1 molecule. In the same way, even if the recombinant AtPepR1-GC has GC activity, its in vivo folding will be very much constrained by the rest of the full-length molecule to fold into the 3D structure shown by the many homologous protein kinase structures present in the Protein Data Bank.

In summary, it is not clear that the AtPepR1-GC fragment has intrinsic GC activity, and even if it does it is likely to be insufficient to have a regulatory role in higher plants.