X-ray structure of PD-L4, a Ribosome Inactivating Protein from Phytolacca dioica L. leaves

Ribosome-inactivating proteins (RIPs) are rRNA N-β-glycosidases (EC number 3.2.2.22) widespread in plants, where they play a role in defence against external pathogens. They are able to inhibit protein biosynthesis in pathogens by inactivating their ribosomes through a site-specific deadenylation of the large ribosomal RNA [1]. In recent years, their cytotoxicity has been exploited for the development of immunotoxins for tumor therapy [2-3] as well as for applications as abortifacients and anti-human immunodeficiency virus (HIV) agents [1]. RIPs target a universally conserved sequence of the ribosome alpha-sarcin loop, which is also involved in the binding of elongation factors to the ribosome. Despite the large amount of structural information on the ribosome accumulated recently [4], no detailed information is still known on the mode of interaction of RIPs with ribosomes. RIPs are also capable of inactivating many nonribosomal nucleic acid substrates and hence can be more generally considered as polynucleotide:adenosine glycosidases [5]. Given the interest of RIPs in pharmaceutical applications [1], efforts have been devoted to understand how these proteins act. Crystal structures of various RIPs have been so far determined [6-7]. However, the mechanism by which they inhibit cell growth or pathogen infection is not well understood. Different mechanisms of action have been proposed based on x-ray studies and extensive site-directed mutagenesis. On the other hand, only a few studies have been focused on the possible catalytic role of residues which are not located in the adenine binding pocket. Therefore, we have started a site-directed mutagenesis study using PD-L4, a type 1 RIP from Phytolacca dioica leaves, as a model system [8]. Leaves of Phytolacca dioica produce several type 1 RIPs, named PD-L1, PD-L2, PDL3, and PD-L4. Of these, PD-L4 (AC: P84854) is the sole isoform which is not glycosilated. In this framework, it has been evidenced that a role in catalysis is exerted by the invariant residue S211 [8]. The polynucleotide-adenosine glycosidase activity of the S211A mutant is significantly lower compared to the wild type protein, with the major extent of reduction for poly(A) substrates [8]. With the aim of gaining insights on the reaction mechanism of RIPs and to understand the structural bases of the reduced activity of the PD-L4 S211A mutant we undertook a crystallographic study [9]. Crystals of wtPD-L4 (Figure1A) were obtained using a protein concentration of 8 mg mL-1 and in the presence of 20% (w/v) PEG 4000, 10% (v/v) iso-propanol, 0.1 M Hepes (pH 7.5). Similar conditions, with a protein concentration of 10 mg mL-1, produced high quality crystals of S211APD-L4 with a slightly different morphology than those of wtPD-L4 in about one week (Figure 1). Atomic resolution diffraction data were collected for wtPD-L4 and S211APD-L4 both in their unliganded state and in complex with adenine using the DESY synchrotron radiation (BW7A beamline). Data sets were indexed and processed with the HKL2000 suite of programs [11]. Statistics of data processing are reported in Table 1. Matthews coefficient calculations (Matthews, 1968) suggested the presence of one molecule in the asymmetric unit. VM values and solvent content for wtPD-L4 were 2.17 Å3 Da-1 and 43.2 %. Similar values (VM=2.11 and solvent content 41.6%) were observed for S211APD-L4. Molecular replacement using the program “Il Milione” [11] and the structure of the ribosome inactivating protein from Phytolacca americana (PDB code 1qci) as a starting model resulted in a clear solution. Preliminary electron density maps are of an excellent quality and clearly show the adenine location in both wtPD-L4 and S211APD-L4 adenine complexes (Figure 1B). Manual and automatic model-building sessions, aimed at defining the complete structures, are in progress using the programa ARP/wARP [12] and O [13]. Refinement of the four crystal structures, in progress,