Trans-species microRNA loci in the parasitic plant Cuscuta campestris have a U6-like snRNA promoter

Abstract

Small regulatory RNAs can move between organisms and regulate gene expression in the recipient. Whether the trans-species small RNAs being exported are distinguished from the normal endogenous small RNAs of the source organism is not known. The parasitic plant Cuscuta campestris (dodder) produces many microRNAs that specifically accumulate at the host-parasite interface, several of which have trans-species activity. We found that induction of C. campestris interface-induced microRNAs is similar regardless of host species and occurs in C. campestris haustoria produced in the absence of any host. The loci-encoding C. campestris interface-induced microRNAs are distinguished by a common cis-regulatory element. This element is identical to a conserved upstream sequence element (USE) used by plant small nuclear RNA loci. The properties of the interface-induced microRNA primary transcripts strongly suggest that they are produced via U6-like transcription by RNA polymerase III. The USE promotes accumulation of interface-induced miRNAs (IIMs) in a heterologous system. This promoter element distinguishes C. campestris IIM loci from other plant small RNAs. Our data suggest that C. campestris IIMs are produced in a manner distinct from canonical miRNAs. All confirmed C. campestris microRNAs with documented trans-species activity are interface-induced and possess these features. We speculate that RNA polymerase III transcription of IIMs may allow these miRNAs to be exported to hosts.

Figure 1.

Improved annotations of C. campestris interface-induced MIRNA (IIM) loci. A) Volcano plot showing differential accumulation of miRNAs comparing interface vs. parasite stem. Each dot represents small RNA accumulation from a single IIM locus. Data are from Johnson et al. (2019). Additional volcano plots from other comparisons are shown in Supplemental Fig. 1. B) Length and 5′ nucleotides (“FirstNT”) of all mature miRNAs, including both strands of the miRNA/miRNA* duplexes. C) Evidence for trans-species targeting of host mRNAs by C. campestris IIM families. See Supplemental Data Set 1 for details. D) Length and 5′ nucleotides (“FirstNT”) of the subset of miRNAs with confirmed trans-species activity.

Figure 2.

IIMs are detectable during the adhesive phase. Toluidine blue stained sections of haustoria growing on A. thaliana  A to C) and S. lycopersicum  D to F). The text at the top right in each image A to F) denotes the stage of haustorium organogenesis, the color correlates with the phases labeled in G to J. EP, endophyte primordium; P, parasite; H, host; SH, searching hyphae; HX, host xylem; XB, xylem bridge; PX, parasite xylem (scale bars: 200 µm). G and H) Detection of IIMs and secondary siRNAs I and J) using A. thaliana or S. lycopersicum as hosts. Boxplots show medians (central line), 25th to 75th percentiles (box boundaries), up to 1.5 × the interquartile range (whiskers), and outliers (dots). RPMM, reads per million mapped. Italic letters indicate significantly distinct groups with adjusted P-values ≤ 0.05 (Dunn's test with Holm's correction for multiple testing). Histological sections from the full time course, including the images shown in A) to F), are shown in Supplemental Figs. 2 and 3.

Figure 3.

IIM production does not require a host. A) Comparison of IIM abundance between hosts. Each dot represents the median value of conductive phase samples for an IIM from the sRNA-seq time course. A linear regression analysis is shown (line and equation). B) Graphical overview of in vitro haustoria experiment. Shoot tips were illuminated for two hours with far-red light and pressured with a stack of glass microscope slides. C) IIM accumulation from in vitro haustoria. Boxplots show medians (central line), 25th to 75th percentiles (box boundaries), up to 1.5 × the interquartile range (whiskers), and outliers (dots). Italic letters indicate significantly distinct groups with adjusted P-values ≤ 0.05 (Dunn's test with Holm's correction for multiple testing). D) Comparing interface-induced accumulation between the Host Free and With Host in vitro haustoria. Each dot represents the median value across all replicates for a single IIM from sRNA-seq of in vitro haustoria. A linear regression analysis is shown (line and equation). IIM, interface-induced microRNA; RPMM, reads per million mapped. Linear regression is shown in red.

Figure 4.

Interface-induced MIRNA loci share a cis-regulatory element with snRNAs and have primary transcript features indicating Pol III transcription. A) Schematic showing a MIRNA primary transcript (top) and the corresponding genomic locus (bottom). Upstream and downstream regions were anchored by the lower DCL cut sites of the primary transcript. B) MEME sequence logo of the USE found upstream of interface-induced MIRNA loci. C) Frequency of the presence of the USE at interface-induced MIRNA loci and canonical MIRNA loci in C. campestris. D) Schematic of RNA 5′ end modifications and enzymatic treatments used in full-length, strand-specific RNA-seq experiments. E) Distributions of abundances of full-length primary transcripts of interface-induced MIRNAs from full-length, stranded RNA-seq with the indicated enzymatic treatments. Boxplots show medians (central lines), 25th to 75th percentiles (boxes), 1.5 × the interquartile ranges (whiskers), and outliers (dots). Italic letters indicate significantly distinct groups with adjusted P-values ≤ 0.05 (Dunn's test with Holm's correction for multiple testing). Note that the y axis of the plot was adjusted to emphasize the central distributions and does not show some high outliers (25/378 RppH outliers and 6/378 Dcp outliers were omitted). F) As in E), except that data are for 2,2,7mGppp capped snRNAs, and all data are shown. G) As in E), except that data are for pre-tRNAs. Note that the y axis of the plot was adjusted to emphasize the central distributions and does not show some high outliers (45/1,448 RppH outliers, 3/1,448 None outliers, and 3/1,448 Dcp outliers were omitted). H) Schematics of TSS and 3′ ends expected for the indicated transcript types. I) Frequency distribution of USE–TSS distances for interface-induced MIRNA primary transcripts. J) Frequency distribution of distances from the 3′ ends and polyT regions (defined as four or more Ts on the coding strand) of interface-induced MIRNA primary transcripts. Canon., canonical microRNA loci; IIM, interface-induced microRNA loci; TSS, transcriptional start site; Dcp, decapping enzyme; RppH, RNA 5′ pyrophosphohydrolase; Pol II, RNA polymerase II; Pol III, RNA polymerase III.

Figure 5.

The USE is required for accumulation of IIMs during transient expression in N. benthamiana. A) Graphical representation of constructs used during transient expression. LB:,left border; USE, upstream sequence element; X, scrambled USE; pT, polyT stretch; RB, right border. B, C) Relative expression of miR12463a or miR12497f with scrambled USE sequence (NoUSE), wild-type USE sequence (USE), and un-infiltrated negative control N. benthamiana leaves (Nb). Expression was normalized to that of miR159. Each dot represents a biological replicate (RNAs from a distinct infiltrated leaf). Italic letters indicate significantly distinct groups with adjusted P-values ≤ 0.05 (Dunn's test with Holm's correction for multiple testing).