Glycine-rich RNA binding protein of Oryza sativa inhibits growth of M15 E. coli cells

Background Plant glycine-rich RNA binding proteins have been implicated to have roles in diverse abiotic stresses. Findings E. coli M15 cells transformed with full-length rice glycine-rich RNA binding protein4 (OsGR-RBP4), truncated rice glycine-rich RNA binding protein4 (OsGR-RBP4ΔC) and rice FK506 binding protein (OsFKBP20) were analyzed for growth profiles using both broth and solid media. Expression of OsGR-RBP4 and OsGR-RBP4ΔC proteins caused specific, inhibitory effect on growth of recombinant M15 E. coli cells. The bacterial inhibition was shown to be time and incubation temperature dependent. Removal of the inducer, IPTG, resulted in re-growth of the cells, indicating that effect of the foreign proteins was of reversible nature. Although noted at different levels of dilution factors, addition of purified Os-GR-RBP4 and OsGR-RBP4ΔC showed a similar inhibitory effect as seen with expression inside the bacterial cells. Conclusions Expression of eukaryotic, stress-associated OsGR-RBP4 protein in prokaryotic E. coli M15 cells proves injurious to the growth of the bacterial cells. E. coli genome does not appear to encode for any protein that has significant homology to OsGR-RBP4 protein. Therefore, the mechanism of inhibition appears to be due to some illegitimate interactions of the OsGR-RBP4 with possibly the RNA species of the trans-host bacterial cells. The detailed mechanism underlying this inhibition remains to be worked out.


Findings Background
RNA binding proteins (RBPs) are characterized by two important structural features: (1) presence of one or more RNA recognition motif [RRM: also known as RNA binding domain (RBD) or ribonucleotide protein domain (RNPs)] at the N-terminus and (2) a variety of auxillary motifs at the C-terminus such as glycine rich (GR) region, arginine rich region, acidic SR-repeats and RDrepeats [1]. In general, RRM is shown to be composed of two RNA binding consensus sequences namely RNPI [a highly-conserved octameric peptide stretch RGFGFVTF; (K/R)G(F/Y)(G/A)FVX(F/Y)] and RNPII [a relatively lessconserved hexameric peptide stretch CFVGGL; (C/I) (F/Y)(V/I)(G/K)(G/N)L]. Genome sequencing projects have revealed that RRM containing proteins are abundantly present in varied life forms including viruses, prokaryotes as well as eukaryotes. Proteins that contain RRMs at the N-terminus and GR domain (GD) at the C-terminus are referred to as glycine rich-RNA binding proteins (GR-RBPs). GD in general consists of an RGG and three GGYGG boxes (two complete; one incomplete). GR-RBPs have been identified in a variety of monocotyledon and dicotyledon plants including bean, Arabidopsis, tobacco, carrot, maize and rice [2][3][4]. These proteins have largely been implicated in cell functions linked to metabolism of mRNA molecules. These include processing, transport, localization, translation and stability of mRNAs. Strikingly, transcripts of GR-RBPs are noted to be up-regulated in response to number of stress stimuli including cold, water stress, high salinity, UV-radiations and heavy metals [3]. The expression of GR-RBPs has also been known to show diverse response to wounding, flooding, hormone and infection by pathogens [3]. AtGR-RBP designated as RZ-1a has been proposed to play a role in the enhancement of freezing tolerance of Arabidopsis plants [5]. Kim et al. [6] produced transgenic Arabidopsis plants over-expressing RZ-1a using 35S-promoter. This work showed that RZ-1a has a negative impact on seed germination and seeding growth of transgenic Arabidopsis plants under salt or dehydration stress conditions. Kim et al. [6] further reported that GR-RBP2, in Arabidopsis has a positive impact on seed germination and seedling growth of plants under cold stress conditions. Kwak et al. [7] have reported that GR-RBP4 negatively affects seeds germination and seedling growth of Arabidopsis plants under salt and dehydration stress conditions. Recently, it has been shown that Arabidopsis GR-RBP7 has RNA chaperone activity during the process of cold adaptation in E. coli [6].
Agarwal [8] isolated a GR-RBP cDNA from rice heat shock cDNA library constructed using mRNA of heat shocked rice seedlings. Osgr-rbp cDNA isolated in this study was designated as OsGR-RBP4 [4]. Osgr-rbp4 transcript was seen to be up-regulated in response to heat stress in rice. Sahi et al. [4] further reported that recombinant yeast cells over-expressing full-length Osgr-rbp4 showed increased basal high temperature tolerance as compared to the wild type cells or cells transformed with vector backbone. There are indications that heat stress increases the stability of specific mRNAs [9]. Noon et al. [10] reported that post-transcriptional modifications in tRNAs and rRNAs are especially abundant in thermophilic organisms and these modifications appear to play a functional role in structural stabilization of RNA at elevated temperatures. Sahi et al. [4] argued that OsGR-RBP4 may have provided increased high temperature protection to yeast cells through its binding to mRNA molecules. Agarwal [8] as well as Sahi et al. [4] from our laboratory have earlier observed that bacterial cells transformed with full-length Osgr-rbp4 are slower in growth under culturing conditions. We provide detailed analysis of the expression of Osgr-rbp4 in M15 E. coli cells, in this study.

Methods
Construction of recombinant E. coli strains E. coli strain M15 [pREP4] (Qiagen) which permits high levels of protein expression was used in this study. M15 contains a low-copy pREP4 plasmid which confers kanamycin resistance and constitutively expresses the lac repressor protein encoded by lac I gene. E. coli M15 strain does not contains a chromosomal copy of lac I and therefore, pREP4 must be maintained by selection for kanamycin resistance. The cells were cultured in LB broth (Luria-Bertani medium, USB Corporation, USA) with kanamycin (kan, 50 μg mL -1 ) for antibiotic selection. pQE30 (Qiagen) was used for expression of OsGR-RBP4 protein as well as OsFKBP20 protein which was used as a positive control. pQE30 contains β-lactamase gene (bla) which confers resistance for ampicillin. Carbenicillin is a semi-synthetic analog of ampicillin and can be used in place of ampicillin. We used carbenicillin in place of ampicillin because of its stability. PCR was performed to confirm the presence of the foreign insert into E. coli M15 cells and subsequently, sequencing was performed to confirm the fidelity of the cloned products. In order to analyze the expression of foreign insert into the E. coli M15 cells, RNA was isolated and RT-PCR was performed. For RNA isolation, M15 cells containing OsGR-RBP4, OsGR-RBP4ΔC, OsFKBP20/, and pQE30 constructs were grown overnight at 37°C in LB broth with carbenicillin and kanamycin. Secondary cultures were put up after~16 h of incubation of primary cultures by measuring the O.D. of primary culture at 600 nm using Hitachi U-2810 spectrophotometer and diluting the cultures to an O.D. of 0.5. After 3 h of initiation of secondary culturing, cultures were divided into two parts: one part was induced with 1 mM IPTG while the other served as a control. The cultures were pelleted at 6000 rpm for 3 min at 4°C and RNA was isolated as per standard procedures. The RNA so obtained was used to make cDNA. cDNAs were synthesized for semi-quantitative RT-PCR from 2 μg of total RNA using MMLV reverse transcriptase (Promega, USA). The following primers were used for amplifying the genes; for both OsGR-RBP4 and OsGR-RBP4ΔC: Forward 5' cgcctgggccaccgacgac3' and reverse 5'ggagcggcgcgactg-ggcct3'; OsFKBP20:Forward 5'ggaattcatggcagaggttgca-gattt3' and reverse 5'ggggtaccttatttctcttcctccttg3'.
To confirm the protein expression from OsGR-RBP4M15, OsGR-RBP4ΔCM15, OsFKBP20M15, and pQE30/M15 cells, secondary cultures were put up from the primary cultures as described earlier. After 3 h of initiation of secondary culturing, cultures were divided into two parts: one part was induced with 1 mM IPTG while the other served as a control. The two sets were pelleted, resuspended in Laemmli buffer and loaded on SDS-PAGE gel (12% resolving, 3.9% stacking). The gel was stained with Coomassie brilliant blue R (CBB-R).
Growth analysis of transformed E. coli cells E. coli M15 cells containing OsGR-RBP4, OsGR-RBP4ΔC, OsFKBP20, and pQE30 constructs were grown overnight at 37°C in LB broth containing kanamycin and carbenicillin. After approximately 16 h of incubation of primary culture, secondary culture was put up and induced with 1 mM IPTG as mentioned above. After every 30 min, O.D. (at 600 nm) of the culture was determined and growth curve was plotted. For the plate assays of E. coli cells, secondary cultures for E. coli M15 cells containing OsGR-RBP4, OsGR-RBP4ΔC, OsFKBP20 and pQE30 constructs were put up after approximately 16 h of incubation of primary culture and induced with IPTG as mentioned earlier. After every 30 min, 10 μl culture was diluted in 90 μl of water to make a dilution of 10 -1 which was further diluted 2 times to obtain 10 -2 and 10 -3 dilutions. 5 μl of these dilutions were spotted on LB agar plates with carbenicillin and kanamycin as selection and were allowed to be absorbed by the plates. The plates were made for -IPTG, +IPTG 30 min, +IPTG 60 min, +IPTG 90 min and +IPTG 120 min treatments.
Plate assay of E. coli cells exogenously provided with OsGR-RBP4, OsGR-RBP4ΔC and OsFKBP20 proteins Secondary cultures for E. coli M15 cells were initiated after approximately 16 h of incubation of primary culture and induced with IPTG as mentioned earlier. Purified OsGR-RBP4, OsGR-RBP4ΔC and OsFKBP20 proteins as of above were quantified using spectrophotometer at 595 nm and equal amount of proteins (120 μg) were added to a 3 mL culture containing equal number of M15 cells (using equal O.D. values at A 600 ). Cultures were kept on continuous shaking at 37°C for approximately 8 h. After 8 h, 10 μl of each culture were diluted in 90 μl of water to make a dilution of 10 -1 which was further diluted 11 times to obtain 10 -2 to 10 -12 dilutions. 5 μl of these dilutions (10 -5 to 10 -12 ) were spotted on LB agar plates with kanamycin as selection and plates were incubated at 37°C.

Analysis of the sequences and homology search
Protein sequences were analysed using DNA analysis software DNASTAR. Predicted protein sequences were determined using DNASTAR software and deduced amino acid sequences were searched for their homology with the previously existing sequences in the NCBI database using the BlastP program [11].

Constructs, transformation and expression analysis
This study aimed at the analysis of the OsGR-RBP4 expression in E. coli M15 cells. To accomplish this, OsGR-RBP4 cloned in pQE30 vector was used. We also used OsGR-RBP4ΔC (truncated version of the Osgr-rbp4 with intact RRM but lacking the glycine rich domain sequences) cloned in pQE30. In addition, we used OsFKBP20 cDNA cloned in pQE30 as control ( Figure 1a). As a plasmid control, pQE30 was also transformed in E. coli M15 cells. Four types of recombinant E. coli cells thus analyzed were as per the following constitution: OsGR-RBP4 in M15 cells (OsGR-RBP4/M15), OsGR-RBP4ΔC/M15, OsFKBP20/M15 and pQE30/M15. Selection of the transformed bacterial cells on antibiotic provided first line of evidence that the cells were transformed. Further, expression of the foreign inserts in the E. coli M15 host cells were examined at the transcript level. RNAs isolated from OsGR-RBP4/M15, OsGR-RBP4ΔC/M15 and OsFKBP20/M15 cells were used for cDNA preparation. cDNAs were used for RT-PCR reactions and the products of these reactions were analyzed on 1% agarose gel. As shown in Figure 1b, clear expression was noted for OsGR-RBP4, OsGR-RBP4ΔC and OsFKBP20 RNAs in RT-PCR reaction from the IPTG-induced cells. It is noteworthy that low level expression was also apparent for the IPTG un-induced samples of various transformants (particularly for the OsGR-RBP4ΔC/pQE30 lane). Subsequently, expression of the foreign protein was examined in the E. coli M15 host cells. Total proteins isolated from OsGR-RBP4/ M15, OsGR-RBP4ΔC/M15 and OsFKBP20/M15 cells were loaded on 12% SDS-gel. As shown in Figure 1c There was no effect of IPTG addition to pQE30/M15 cell type. While OsFKBP20/M15 cells did grow upon IPTG induction, its growth was slightly less than pQE30/M15 cell type.
To reaffirm the above observations, 5 μl cells from the IPTG un-induced and induced culture were spotted after serial dilutions (10 -1 -10 -3 ) on LB agar plates with antibiotic selection and allowed to be absorbed by the medium. Similar to the observations made with broth culture, plate assay showed that induction with IPTG has detrimental effect on the viability of OsGR-RBP4/ M15 and OsGR-RBP4PΔC/M15 cells. These cells upon culturing on plates showed reduced viability. On the other hand, OsFKBP20/M15 and pQE30/M15 cell types showed no such inhibitory effects ( Figure 2B.a).
Next we tested the effect of temperature on bacterial growth. Secondary culture (0.5 O.D. of the overnight grown primary culture used to set the secondary culture) was set up for all the four transformed E. coli M15 cell types. After 3 h, total culture volume was divided into two parts: one part was induced by IPTG and other was left un-induced. After every 30 min interval, cells from the un-induced and induced cultures were serially diluted and spotted on LB agar plates with carbenicillin and kanamycin. Three sets of plates were prepared and one set of each of these was incubated for 16 h at 20°C, 37°C and 45°C in separate incubators. At 20°C ( Figure  2B. b) and 37°C ( Figure 2B. a), all the four cells types showed comparable growth profile in the sets made without induction by IPTG ( Figure 2B. b). In sets prepared with IPTG induced cell types, OsGR-RBP4/M15 and OsGR-RBP4ΔC/M15 cells showed lower growth profiles than OsFKBP20/M15 and pQE30/M15 cell types ( Figure 2B. c). At both the temperatures of plate incubation, this effect was more pronounced after at least 60 min of IPTG induction. At 45°C ( Figure 2B.c), increased lethality to OsGR-RBP4/M15 and OsGR-RBP4ΔC/M15 cells as against OsFKBP20/M15 and pQE30/M15 cells was apparent even at 30 min of IPTG induction. In this stress regime, it was further noted that OsFKBP20M15 and pQE30/M15 cells were able to grow gradually with time following induction by IPTG. However, OsGR-RBP4/M15 and OsGR-RBP4ΔC/M15 cells failed to overcome the lethality effects following IPTG induction (Figure 3).
In subsequent experiment, IPTG was washed off from the cultures of all four cell types after inducing it with IPTG for 30 min. The recovered cell cultures were serially diluted and spotted on LB agar plates with antibiotic selection (plates incubated at 37°C for 16 h). The lethality noted for OsGR-RBP4/M15, OsGR-RBP4PΔC/M15 cells was alleviated to a significant extent upon IPTG washing ( Figure 3).
Further, growth profiles of E. coli M15 cells were monitored in response to addition of purified OsGR-RBP4, OsGR-RBP4PΔC and OsFKBP20 proteins. The respective proteins were partially purified from the corresponding cultures using affinity column chromatography (respective proteins shown by arrows in Figure 4a). Equal amounts of proteins (120 μg) were added directly into the E. coli M15 cultures. It may be noted here that because of the variations in the molecular weights of the purified OsGR-RBP4, OsGR-RBP4PΔC and OsFKBP20 proteins, the number of individual protein molecules in 120 μg of protein may vary. After~8 h of growth, cultures were spotted on LB agar plates with selection (kan). Exogenous addition of the OsGR-RBP4 and OsGR-RBP4PΔC proteins showed a distinct toxic effect on the growth of E. coli M15 cells. Out of these two, OsGR-RBP4PΔC appeared more toxic to cells than OsGR-RBP4. This again may be possible because of higher ability of OsGR-RBP4PΔC protein to enter into E. coli owing to its lower molecular weight. On the other hand, exogenous addition of OsFKBP20 protein didn't result in growth inhibition of bacterial cells (Figure 4b).
Next, we were interested to see if there is any protein in E. coli genome homologous to OsGR-RBP4. OsGR-RBP4 sequence was searched for their homology with E. coli in the NCBI database using BlastP program. We didn't observe any significant homology of OsGR-RBP4 to any E. coli protein in this analysis. there are instances in which expression of selective eukaryotic proteins in E. coli have been shown to exert toxic effects on the growth of host E. coli cells. In such cases, the expression yields of the proteins are dramatically diminished or sometimes abolished. The striking examples of this class include (1) aspartic protease from human immunodeficiency virus type I [12], (2) recombinant feline TNF [13], (3) malarial oritidine-5'-monophosphatedecarboxylase [14] and (4) phage HK 022 nun protein [15]. There are indications that such genes interfere with the physiology of E. coli. A few plant proteins have also been reported to have deleterious effect on the growth of the trans-host. Habuka et al. [16] observed that overexpression of Mirabilis antiviral protein (MAP) weakly-inhibits protein synthesis in E. coli which further results in a decrease in the growth of the bacterial cells. Girbes et al. [17] found that Ebulin1(a type 2 ribosome inactivating protein) from Sambucus ebulus strongly inhibits protein synthesis in rabbit reticulocyte lysate, rat brain and rat liver cell free systems but does not affect plant or bacterial protein synthesis in vitro. In this study, we noticed that expression of OsGR-RBP4 in E. coli M15 cells adversely affects the growth of the host cells. This effect was specifically caused by OsGR-RBP4 protein as (1) growth inhibition of the host E. coli cells was apparent only when cells were induced by IPTG and (2) there was no inhibition in growth of E. coli cells when either OsFKBP20 or backbone vector sequence pQE30 were expressed. OsGR-RBP4 protein consists of RBD at N-terminus and GD at C-terminus. We removed GD sequence from the full-length OsGR-RBP4 sequence to construct a GD-deleted strain (i.e. OsGR-RBP4ΔC). When OsGR-RBP4ΔC was expressed in E. coli cells, a similar decrease in the growth was observed (Figures 2 and 3), suggesting that the deleterious effect of OsGR-RBP4 in E. coli cells is possibly contributed by the RBD. It appears that the inhibitory effect of OsGR-RBP4 and OsGR-RBP4ΔC proteins is a timedependent process i.e. there was not much difference in the growth of the OsGR-RBP4/M15 and OsGR-RBP4ΔC/M15 cells up to 60 min of IPTG addition but beyond 60 min time interval of IPTG addition there was a significant percent decrease in the growth of the above cell types. This may be related to the expression/accumulation of the requisite proteins. We have used proteins with an N-terminal 6-histidine tag. Although it has been observed that it is important to maintain native protein structure for protein folding/structural studies, we did not come across any report mentioning proteotoxicity resulting because of histidine tag.
On altering the incubation temperature of the plates, it was further noted that while plates incubated at 20°C and 37°C showed drastic effects with 60 min of IPTG treatment, plates incubated at 45°C showed growth inhibitory effects even with 30 min of IPTG treatment. It is possible that this inhibitory effect has a correlation with the metabolic state of the cells: cells with higher metabolic status/ growth rates at increased temperature appear more vulnerable to inhibitory effects.
The principal activity of the OsGR-RBP4 appears to be its binding to RNA species. The toxic effects of expression of OsGR-RBP4 (as well as OsGR-RBP4ΔC) may be due to its binding to some vital bacterial transcripts. As we didn't find any natural homolog of the OsGR-RBP4 protein in E. coli, it is possible that expression of this  OsGR-RBP4 C 10 -10 10 -12 10 -11 10 -7 10 -9 10 -8 10 -6 10 -5 b. eukaryotic protein may be involved in some illegitimate binding with bacterial transcripts. It also needs to be considered that this binding may be of flux nature because when IPTG is washed off, re-growth of the OsGR-RBP4M15 and OsGR-RBP4ΔC/M15 cells was noticed in this study. It is generally considered that higher accumulation of the foreign proteins may lead to protein toxicity due to higher demand for energy and molecular machinery for the degradation. We also noted that OsFKBP20 expression resulted in lesser growth of the host cells compared to pQE30/M15 cells. However, the increased arrest of growth noted in OsGR-RBP4/ M15 and OsGR-RBP4ΔC/pQE30/M15 cells as compared to OsFKBP20/M15 cells suggests that expression of GR-RBP causes specific interference in RNA metabolism and thus growth of bacterial cells expressing this protein is drastically affected.
This study was undertaken with only the E. coli M15 cells. Testing of the other strains of E. coli as well as other gram-negative or gram-positive bacteria would be interesting. If OsGR-RBP4 turns out to be broad spectrum bactericidal protein, it may have applied values. It would be worth to assess the toxicity of rice OsGR-RBP4 protein on disease causing bacterial strains. In future we plan to do this kind of analysis using Nicotiana tabacum transgenics raised in our laboratory overexpressing OsGR-RBP4 protein