High-throughput assay and engineering of self-cleaving ribozymes by sequencing

Abstract

Self-cleaving ribozymes are found in all domains of life and are believed to play important roles in biology. Additionally, self-cleaving ribozymes have been the subject of extensive engineering efforts for applications in synthetic biology. These studies often involve laborious assays of multiple individual variants that are either designed rationally or discovered through selection or screening. However, these assays provide only a limited view of the large sequence space relevant to the ribozyme function. Here, we report a strategy that allows quantitative characterization of greater than 1000 ribozyme variants in a single experiment. We generated a library of predefined ribozyme variants that were converted to DNA and analyzed by high-throughput sequencing. By counting the number of cleaved and uncleaved reads of every variant in the library, we obtained a complete activity profile of the ribozyme pool which was used to both analyze and engineer allosteric ribozymes.

Figure 1.

Library construction strategy. First, a partially randomized ribozyme library is transcribed in vitro from a DNA template. The cleaved and uncleaved ribozymes are reverse transcribed into cDNAs using a primer that contains a barcode and an adapter sequence. After removing the RNAs, the 3′ adapter is attached and amplified by PCR to obtain the sequencing library. The core ribozyme sequence is shown in black with the degenerate bases depicted in red. Other sequence elements: T7 promoter (purple), barcode (yellow), adapter sequences for sequencing (green and blue).

Figure 2.

Ribozyme library designs. Degenerate bases are shown in red and the cleavage site is marked by an arrowhead. (A) Lib-Tw based on a twister ribozyme discovered from the environmental DNA (3). The bases involved in the T1 pseudoknot are shaded. (B) Lib-J1/2 based on the HDV-like ribozyme in A. gambiae drz-Agam-2-1(1). The guanine aptamer (purple) was inserted at the J1/2 loop via four random bases. (C) Lib-P4 based on drz-Agam-2-1. The P4 stem was replaced with the guanine aptamer (purple) through a randomized connector sequence. (D) Illustration of aptazymes used to regulate gene expression in mammalian cells. An aptazyme is embedded in the 3′ UTR of a reporter gene (EGFP) mRNA. Ribozyme cleavage induced by the aptazyme results in low EGFP expression due to the scission of the poly (A) sequence from the coding region.

Figure 3.

Correlation between the HTS and in vitro assays. Ribozyme activities (fraction cleaved) of selected variants were assayed individually by gel electrophoresis (in vitro) and plotted against the activities measured by HTS. At least two measurements by gel electrophoresis were performed for each variant. (A) Lib-Tw. (B) Lib-J1/2 (filled circles) and Lib-P4 (open diamonds).

Figure 4.

Characterization of the guanine-responsive aptazymes from Lib-J1/2 and Lib-P4. (A) Comparison of guanine-dependent ribozyme activities of individual aptazyme variants as measured by gel electrophoresis (in vitro, white) and HTS (black). In +guanine reactions, 500-μM guanine was present in the reactions. The error bars (in vitro) indicate the ranges of the two independent measurements. (B) Expression of EGFP in the HEK 293 cells transfected with EGFP-aptazyme expression plasmids in the absence (hatched) and presence (gray) of guanine. EGFP expression is reported relative to the control which contains an inactivated drz-Agam-2-1 ribozyme (1.0), after normalization for transfection efficiency using the cotransfected mCherry expression plasmid. In +guanine samples, 500-μM guanine was added to the medium. The data shown are mean ± SD from three replicate samples.

Figure 5.

Phenotypic distributions of the aptazyme libraries. Ribozyme activities of the library variants (±guanine) were plotted. The solid line has a slope = 1 which implies no change in ribozyme activity in the presence of guanine. The shaded region represents a guanine-activated aptazyme phenotype that satisfies the following arbitrarily determined criteria: (i) activity in the presence of guanine is >0.30, and (ii) ribozyme is activated at least 3-fold in the presence of guanine (on/off ratio >3). These plots provide a visual overview of the phenotypic distributions of aptazyme libraries. (A) Lib-J1/2. (B) Lib-P4.