N-glycosylation on Oryza Sativa Root Germin-like Protein 1 is conserved but not required for stability or activity

Germin and germin-like proteins (GLPs) are a broad family of extracellular glycoproteins ubiquitously distributed in plants. Overexpression of Oryza sativa root germin like protein 1 (OsRGLP1) enhances superoxide dismutase (SOD) activity in transgenic plants. Here, we report bioinformatic analysis and heterologous expression of OsRGLP1 to study the role of glycosylation on OsRGLP1 protein stability and activity. Sequence analysis of OsRGLP1 homologs identified diverse N-glycosylation sequons, one of which was highly conserved. We therefore expressed OsRGLP1 in glycosylation-competent Saccharomyces cerevisiae as a Maltose Binding Protein (MBP) fusion. Mass spectrometry analysis of purified OsRGLP1 showed it was expressed by S. cerevisiae in both N-glycosylated and unmodified forms. Glycoprotein thermal profiling showed little difference in the thermal stability of the glycosylated and unmodified protein forms. Circular Dichroism spectroscopy of MBP-OsRGLP1 and a N-Q glycosylation-deficient variant showed that both glycosylated and unmodified MBP-OsRGLP1 had similar secondary structure, and both forms had equivalent SOD activity. Together, we concluded that glycosylation was not critical for OsRGLP1 protein stability or activity, and it could therefore likely be produced in Escherichia coli without glycosylation. Indeed, we found that OsRGLP1 could be efficiently expressed and purified from K12 shuffle E. coli with a specific activity of 1251±70 Units/mg. In conclusion, we find that some highly conserved N-glycosylation sites are not necessarily required for protein stability or activity, and describe a suitable method for production of OsRGLP1 which paves the way for further characterization and use of this protein.


Introduction
Germins and germin-like proteins (GLPs) belong to a functionally diverse family of proteins called cupins [1]. These proteins have been extensively studied in cereals and legumes due to their historical significance, abundance, expression during biotic and abiotic stresses and potential biotechnological applications [2,3]. Germins were first identified during a search for molecular markers associated with germination in wheat and were the only gene product synthesized de novo during germination; hence the name germin, implying an association with and serving as a signal for germination [4]. Several germin isoforms have been identified in diverse cereals, but not all are associated with germination. GLPs form a broader protein family, with ~30-70% sequence similarity to wheat germins [5].
Germins and GLPs share a well conserved β-barrel fold [1] and are typically glycoproteins. Differences in the extent of glycosylation are responsible for their presence as different isoforms, with variably modified forms associated with different subcellular localization [6,7] and tissue distribution [8]. Oryza sativa root germin like protein 1 (OsRGLP1) contains a putative N-glycosylation site (N-x-S/T; x¹P) [9], but the exact role of glycosylation in OsRGLP1 protein stability, folding, or activity is unknown. Glycosylation affects the folding efficiency and final structure of proteins, and most secreted proteins are glycosylated. Changes in glycosylation can affect how glycoproteins recruit, interact with, and activate other proteins in diverse biological systems [10]. In addition to assisting in protein folding, glycosylation is also important for increasing protein thermostability. Addition of even a single glycosylation event can impact the equilibrium of protein between folded and unfolded states, although in some proteins elimination of all or some glycans does not affect protein folding [11].
Germins and GLPs have been widely exploited in transgenic plants to enhance resilience to biotic and abiotic stresses. The most important biochemical activity reported for true germins is oxalate oxidase activity [12], and for GLPs superoxide dismutase (SOD activity [13]. OsRGLP1 has been successfully transformed into Nicotiana tabacum (tobacco), Solanum tuberosum (potato) and Medicago truncatula (barrelclover) [13,14].
OsRGLP1-transformed transgenic plants have higher SOD activity than wild type plants, and enhanced stress tolerance due to this heat resistant SOD activity [15].
Heterologous expression of OsRGLP1 in potato also induces resistance against infection by Fusarium oxysporum f.sb. tuberosi [14]. Transgenic Medicago truncatula expressing OsRGLP1 are more resistant to F. oxysporum than wild type plants, which was attributed to the heat resistant SOD activity of OsRGLP1 [16].
Here, we investigated the importance of glycosylation for the stability and activity of OsRGLP1. As OsRGLP1 likely required eukaryotic post-translational modifications, we first tested the importance of the single conserved glycosylation site in OsRGLP1 for protein activity and stability by its expression and purification as an MBP-fusion protein in Saccharomyces cerevisiae. Surprisingly, we found that although it contained a conserved N-glycosylation site, this was not required for its stability or function. We therefore also produced OsRGLP1 in Escherichia coli, without glycosylation, and found it was active and could be efficiently expressed and purified, which will be of utility in diverse applications.

In Silico Analysis of Glycosylation in Germins and GLPs
Homologous sequences to OsRGLP1 (UniProtKB accession Q6YZY5) were found using blastP with default parameters. Phylogenetic trees were constructed from alignments by PhyML [17] and ancestral sequence reconstruction was carried with CODEML [18] with a WAG amino acid substitution matrix and molecular clock turned off. The possible instance of glycosylation sites in relation to each other were counted and the intensity of data was displayed as a heat map for visualization of conserved glycosylation sequons. shaking overnight.

Data Analysis
Peptides and proteins were identified using ProteinPilot v5.0.1 (SCIEX) using a custom database containing the OsRGLP1 or MBP-OsRGLP1 fusion protein sequences. ppm were applied to precursor and fragment ions, respectively. Cys-S-betapropionamide was set as a fixed modification and variable modifications were set as mono-oxidised Met and deamidation of Asn. The glycan database contained high mannose glycans (GlcNAc2Man1-15), with one glycan allowed per peptide.

Circular Dichroism Spectroscopy
CD spectra were measured from 195 to 280 nm with a Jasco J-710 spectrometer (Jasco) in a 1 mm path length cuvette.

SOD Assays
SOD assays were performed as described [25]. In-solution SOD activity was determined by the difference between the absorbance of the samples in light and dark. The % inhibition was calculated by the difference between OD595 of blank and OD595 of sample, divided by OD595 of blank. 1 unit of SOD was defined as the amount of SOD required to inhibit the photochemical reduction of NBT by 50% [25]. In-gel SOD activity was remaining SOD activity measured as described above. The data was analyzed in Graphpad Prism 7 using an unpaired t-test.

Conservation of Glycosylation in Germins and Germin-Like Proteins
Germins and GLPs are typically N-glycosylated. These post-translational modifications may be important in determining protein structure and activity. Although glycosylation is common in germins and GLPs, their glycosylation site conservation has not been carefully studied. We therefore examined the presence and location of glycosylation sites in proteins with at least 60% sequence homology to OsRGLP1. Multiple sequence alignment and phylogenetic analysis revealed that germins and GLPs in rice typically have one or two glycosylation sites. In order to elucidate evolutionary relationships between the different sequon positions we created a correlation heat map ( Fig. 1) showing the probability of any two sequons occurring in the same sequence. This   We next tested the impact of glycosylation at the single conserved N-glycosylation site in OsRGLP1 on protein stability and activity. We first performed thermal glycoprotein stability profiling of purified MBP-OsRGLP1, with LC-MS/MS measurement of the glycosylation profile of protein that remained soluble after incubation at temperatures from 30-90 °C (Fig. 2C). No pronounced differences were observed in the relative abundance of glycosylated and unmodified forms in the soluble fraction after incubation at any temperature. As an alternative method for measuring the impact of glycosylation on the structural integrity of OsRGLP1, we created a non-glycosylated MBP-OsRGLP1-N76Q variant, expressed and purified this variant and MBP-OsRGLP1, and measured their secondary structure by CD spectroscopy. This analysis showed equivalent CD spectra for MBP-OsRGLP1 and MBP-OsRGLP1-N76Q (Fig. 2D). Together, these thermal glycoprotein profiling and secondary structural analysis results showed that the presence or absence of glycosylation at the single conserved glycosylation site of MBP-OsRGLP1 did not have a strong influence on the thermal stability of the protein.
Overexpression of OsRGLP1 has been reported to increase the SOD activity of transgenic plants [26]. To test if N-glycosylation affected the SOD activity of MBP-OsRGLP1, we expressed and purified MBP-OsRGLP1, non-glycosylated MBP-OsRGLP1-N76Q variant, and MBP from S. cerevisiae, and measured their activities using an in-solution SOD assay (Fig. 2E). This assay showed clear SOD activity for both the wild-type and non-glycosylated variant proteins, which were not significantly different. This is particularly surprising, given the conservation of its single N-  Figure 4). Expression at 16°C reduced, but did not remove, GroL contamination ( Figure 3A). Interactions between molecular chaperones such as GroL and heterologously expressed proteins can result in their copurification. To remove the bacterial chaperones from OsRGLP1, we washed the His-OsRGLP1 bound talon resin with MgATP and a misfolded bacterial protein extract as described [22]. This resulted in essentially pure OsRGLP1 ( Figure 3B)  We tested the SOD activity of OsRGLP1 purified from E. coli using an in-gel assay.
This assay showed that OsRGLP1 had equivalent activity to E. coli SODA ( Figure 3D-F), confirming that OsRGLP1 purified from K12 shuffle E. coli was properly folded and active. OsRGLP1 expressed in plants is reported to confer high-temperature resistant SOD activity [15,16,26]. To test the thermal stability of OsRGLP1 expressed from E.
coli, purified protein was incubated at different temperatures and then tested for residual SOD activity ( Figure 3G). No significant reduction in residual OsRGLP1 activity was observed after incubation at 30°C or 40°C, whereas incubation at 50°C or 60°C caused a significant reduction in residual activity. We therefore concluded that OsRGLP1 retained activity after incubation over the range of temperature tested, but lost activity after incubation at particularly high temperatures. The specific SOD activity of purified OsRGLP1 was estimated to be 1251 U/mg (Fig. 3G). This is higher than the activity for SOD from Cicer arietinum (157.5 U/mg) [27], but less than that of Cu-Zn SOD (SOD_C1) from citrus lemon (7456 U/mg) [28], Cu/Zn SOD from Cucurbita moschata (1794 U/mg) and CuZn SOD from Allium sativum (2867 U/mg) of SOD [29].
In summary, we successfully used two different host systems, S. cerevisiae and K12 shuffle E. coli, for expression and purification of OsRGLP1. This represents the first report of successful expression and purification of germins or GLPs in soluble and active form in these host systems. We provide the first direct biochemical evidence that OsRGLP1 has SOD activity. Finally, we found that although there was a highly conserved N-glycosylation sequon in OsRGLP1 that can be glycosylated, modification of this site does not affect protein stability, and was not required for enzymatic activity.
The possibility of expressing OsRGLP1 in E. coli opens possibilities for efficient production of this protein for fundamental research and a variety of biotechnological applications.