Molecular Studies of the Protein Complexes Involving Cis-Prenyltransferase in Guayule (Parthenium argentatum), an Alternative Rubber-Producing Plant

Abstract

Guayule (Parthenium argentatum) is a perennial shrub in the Asteraceae family and synthesizes a high quality, hypoallergenic cis-1,4-polyisoprene (or natural rubber; NR). Despite its potential to be an alternative NR supplier, the enzymes for cis-polyisoprene biosynthesis have not been comprehensively studied in guayule. Recently, implications of the protein complex involving cis-prenyltransferases (CPTs) and CPT-Binding Proteins (CBPs) in NR biosynthesis were shown in lettuce and dandelion, but such protein complexes have yet to be examined in guayule. Here, we identified four guayule genes - three PaCPTs (PaCPT1-3) and one PaCBP, whose protein products organize PaCPT/PaCBP complexes. Co-expression of both PaCBP and each of the PaCPTs could complemented the dolichol (a short cis-polyisoprene)-deficient yeast, whereas the individual expressions could not. Microsomes from the PaCPT/PaCBP-expressing yeast efficiently incorporated ¹⁴C-isopentenyl diphosphate into dehydrodolichyl diphosphates; however, NR with high molecular weight could not be synthesized in in vitro assays. Furthermore, co-immunoprecipitation and split-ubiquitin yeast 2-hybrid assays using PaCPTs and PaCBP confirmed the formation of protein complexes. Of the three PaCPTs, guayule transcriptomics analysis indicated that the PaCPT3 is predominantly expressed in stem and induced by cold-stress, suggesting its involvement in NR biosynthesis. The comprehensive analyses of these PaCPTs and PaCBP here provide the foundational knowledge to generate a high NR-yielding guayule.

Keywords: cis-prenyltransferase; guayule (Parthenium argentatum); polyisoprene; protein complex; terpenoid.

FIGURE 1

Phylogenetic analysis of CPT (A) and CBP proteins (B). Phylogenetic trees were created based on the protein sequence similarities between CPTs and CBPs from various prokaryotic and eukaryotic species. Bootstrap values from 1,000 replicates were calculated, and the percentages of replicates are shown in each node only when higher than 50%. Two prokaryotic CPTs (EcCPT and SaCPT) are used as the outgroup and indicated by a gray square. PaCPT1-3 and PaCBP are marked by red boxes. Abbreviations used are: At, Arabidopsis thaliana; Ce, Caenorhabditis elegans; Ec, Escherichia coli; Ha, Helianthus annuus; Hb, Hevea brasiliensis; Hs, Homo sapiens; Ht, Helianthus tuberosus; Ls, Lactuca sativa; Pa, Parthenium argentatum; Pt, Populus tremuloides; Sa, Staphylococcus aureus; Sl, Solanum lycopersicum; Tb, Taraxacum brevicorniculatum; To, Taraxacum officinale; Vv, Vitis vinifera; RER2 and SRT1 are yeast CPTs. NUS1 is yeast homolog of CBP. Accession numbers of the sequences used to construct the phylogeny are given in “Materials and Methods.” Arabidopsis CPTs (AtCPT1-9) and tomato CPTs (SlCPT1-7) were numbered according to the published articles (Surmacz and Swiezewska, 2011; Akhtar et al., 2013). Note that Kera et al. (2012) assigned different numbering of the nine Arabidopsis CPTs, and different nomenclature of AtCPTs can be listed in other publications.

FIGURE 2

Complementation of rer2Δ srt1Δ yeast by PaCPT1-3 and PaCBP. The yeast strain, rer2Δ srt1Δ is lethal but is maintained by expressing RER2 in URA-selectable plasmid. This strain was used to transform plasmids expressing each PaCPT1-3 and PaCBP. The successful transformants were streaked on 5FOA selection plates to remove RER2 containing URA-plasmid. Yeast growth in 5FOA selection was observed only by PaCPT/PaCBP pairs or by retransformed RER2 in TRP-plasmid. No growth was observed when PaCPT alone or PaCBP alone was expressed.

FIGURE 3

Isoprene product separations by reverse phase thin layer chromatography. Microsomes were prepared from the yeast strains selected on 5FOA in Figure 2, and in vitro enzyme assays were performed using the microsomes and ¹⁴C-isopentenyl diphosphate, followed by TLC (C18 reverse-phase silica) of extracted cis-polyisoprene products using a solvent mixed with acetone and water (39:1), and phosphorimager analysis. (A) The cis-polyisoprene product profiles from yeast RER2 as a size calibration standard. (B) The cis-polyisoprene products from PaCPT/PaCBPs are identical to one another and slightly shorter than those from the RER2-containing microsomes. S indicates the starting point of separation and SF indicates solvent front. R_f values and carbon numbers are shown.

FIGURE 4

Split ubiquitin yeast 2-hybrid assays to assess the protein interactions between PaCPT1-3 and PaCBP. The bait protein, Cub, was fused at the C-terminus of PaCBP in all assays. Under the selective growth condition, SC-LTAH, yeast growth was observed only when the bait and prey proteins interacted either through NubI-Cub auto-assembly (A: Ost1-NubI, positive control) or through the PaCPT/PaCBP interaction-mediated NubG-Cub assembly (B: NubG-PaCPT3, NubG-PaCPT2, NubG-PaCPT3). PaCPT/PaCBP interactions occurred only when NubG was fused on the N-terminus of PaCPTs. Under the non-selective growth condition, SC-LT, all yeast strains grew regardless of the protein interaction. As an important negative control, NubG-fused Ost1 protein did not interact with Cub-fused PaCBP (A), demonstrating a specific interaction of PaCBP to PaCPT, but not to Ost1. Yeast growth was observed in a wide range of 5–25 mM AT, indicating a strong activation of the HIS3 reporter gene by the ubiquitin-activated transcription factor. The β-galactosidase activity was measured in the corresponding yeast strains that showed protein interactions in the absence of AT. PaCPT1/PaCBP interaction showed the same level of β-galactosidase activity as the Ost1-NubI control while PaCPT2/PaCBP and PaCT3/PaCBP showed lower activities. Ten μL of OD₆₀₀ yeast culture was placed in the first spot, and yeast cells with 10- and 100-times dilution were placed in the subsequent spots. Asterisks indicate statistically significant differences (p-value < 0.05).

FIGURE 5

Co-immunoprecipitation of recombinant PaCPT/PaCBP proteins. (A) HA-PaCBP could be co-immunoprecipitated by FLAG-PaCPT1-3, and (B) FLAG-PaCPT1-3 could be co-immunoprecipitated by HA-PaCBP. PaCPT1-3 and PaCBP proteins tagged with FLAG- and HA-epitopes, respectively, and were prepared by in vitro transcription and translation. GFP tagged with an HA-epitope was used as a negative control. The information of the protein mixture is detailed on the top panel with the appropriate epitope-tag. The antibodies used in the immunoblot are indicated beside blots (FLAG, FLAG antibody or HA, HA antibody).

FIGURE 6

RNA-seq analysis of PaCBP and PaCPT1-3. FPKM (fragments per kilobase million) values for PaCBP and PaCPT1-3 were calculated from the publicly available data (short read archive number: SRP107961). Gray bars, control; Black bars, cold treatment. L, leaf sample; S, stem sample. Data are means ± SD (n = 3).