Posttranscriptional control is a strong factor enabling exclusive expression of surface antigens in Paramecium tetraurelia

Abstract

Variable antigens are large proteins located on the outer membrane of parasitic but also of free-living protists. Multigene families encoding surface antigens demonstrate an exclusive expression of proteins. The resulting presence of just one protein species on the cell surface is required for surface antigen function; therefore, the molecular mechanism of exclusive expression is of main interest. Regulation of gene expression and mechanisms establishing switching of antigens are hardly understood in any organism. Here we report on the reaction of Paramecium to the artificial knock down of surface antigen 51A expression by bacteria-mediated RNAi. This technique involves the feeding of dsRNA-producing bacteria. We analyzed different fragments of the target gene for dsRNA template regarding their specific knock down efficiency and found great differences. Treatment of Paramecia with RNAi against the 51A antigen demonstrated that although a massive amount of mRNA was present, the protein was not detected on the cell surface. Moreover, a minor abundance of 51D transcripts resulted in an exclusive presence of 51D proteins on the cell surface. This posttranscriptional regulation was confirmed by the transcript ratio (51A/51D) determined by real-time (RT) PCR of single cells. Because we were able to document unexclusive transcription also in wild-type cells our results indicate that this posttranscriptional regulation is a main factor of enabling exclusive gene expression. The comparison of serotype shifts, caused by efficient and inefficient knock down, indicates an involvement of full-length transcripts in regulation of gene expression. Thus, our study gives new insights into the mechanism of exclusive expression on the molecular level: (i) exclusive transcription does not occur, (ii) posttranscriptional regulation is a powerful factor enabling exclusive antigen expression, and (iii) surface antigen mRNA is shown to be involved in this mechanism in a regulating way.

Figure 1

Total nucleic acids were isolated from synchronous bacteria, meaning simultaneous induction with IPTG and lysis (see Materials and Methods). DNAse digested samples were loaded on a 1.8% agarose gel: 1 kb ladder; lane 1, HT115DE3 bacteria containing the 340 plasmid; lane 2, bacteria containing the 647 plasmid; lane 3, bacteria containing the 700 plasmid. The size of lanes is according to the distance between T7 promotors of each plasmid. The higher abundance of the 340 dsRNA compared to other samples is obvious. Lanes 4–6 demonstrate the negative control using the same plasmids (lane 4, 340; lane 5, 647; lane 6, 700) in the E. coli SURE® strain showing no dsRNA enrichment.

Figure 2

Parallel indirect immunofluorescence staining of surface antigens using antibodies against the 51A and 51D serotype to detect the silencing phenotype. Cultures under RNAi treatment were tested for the presence of antigens on the surface and showed three different types of antigen expression: pure A-expressing cells (a), cells showing a shifting state where remaining A protein (indirectly labeled with FITC, green), and upcoming D protein (indirectly labeled with TexasRed) coexist on the surface (b), and cells that have already completely shifted to serotype D (c). Specificity of the staining is demonstrated in (d) showing a serotype B-expressing cell treated in the same manner. There were 150–200 cells analyzed per sample to represent the culture composition as demonstrated in Figure 3.

Figure 3

Expressed surface antigen in silenced cultures determined by the indirect immunofluorescence staining: (a) 340 plasmid, (b) 647 plasmid, (c) 700 plasmid. Serotype A is indicated as black, serotype D is indicated as white. Coexpressive cells showing serotype A and D on the cell surface are indicated by black and white stripes and unknown serotypes by gray. The comparison demonstrates that only plasmid 340 is able to induce a nearly complete serotype shift from A to D. On the first day after feeding the phenotypes of every plasmid showed a high percentage of coexpressive cells; a small percentage kept expressing serotype A. Cells showed 3.45 ± 0.3 divisions per day and all performed controls did not shift to another serotype but continued expressing serotype A.

Figure 4

Quantitative analysis of surface antigen transcript level by real-time PCR. For each fragment [(a) 340, (b) 647, (c) 700] total RNA was isolated every 24 h (x-axis) during the feeding procedure, reverse transcribed, and cDNA was quantified from serotype genes A and D but also from GAPDH, which was demonstrated to be expressed at a constant level. The demonstrated data display mRNA abundances of surface antigens A (black) and D (white) as percentage of the expression level of GAPDH (y-axis). Multiple amounts of A relative to D transcripts are indicated in the diagram. Fragment 340 shows an efficient downregulation of A transcripts; both other fragments only show a slight reduction followed by re-increase. D transcripts accumulated in all cultures but only fragment 340 showed an efficient A downregulation and an efficient D upregulation.

Figure 5

Cells silenced with fragment 700 were isolated 24 h after first feeding and transcript levels of A, D, and GAPDH were determined from single cells by real-time PCR. Abundance of serotype mRNA is plotted against GAPDH, each dot indicating a single cell (y-axis: A mRNA; x-axis: D mRNA). The distribution shows that in one group increased D abundance correlated with decreased A abundance, assuming that these cells will completely shift to pure serotype D. Another group of cells showed fewer D transcripts; hence, higher levels of A transcription seem to inhibit D transcription and vice versa. Please note that dots plotted directly on the y-axis, indicating the lack of D transcripts, do not necessarily imply the absence of transcripts. Abundances may have been below the detection limit because RNA was isolated from one single cell.