In-house produced MarathonRT comparison to uMRT and Induro in tRNA sequencing library preparation

This study presents a cost-effective, in-house production and purification protocol for a redesigned MarathonRT enzyme that matches the performance of commercial alternatives (Induro and uMRT) in tRNA sequencing, while also introducing a rapid spin-column extraction method to further reduce the time and cost of library preparation.

The Gist

Imagine your cell is a bustling, high-tech factory. The tRNAs (transfer RNAs) are the hardworking forklifts that move raw materials (amino acids) to the assembly line to build proteins. To understand how this factory is running, or why it might be breaking down in diseases, scientists need to take a "snapshot" of every single forklift in the building. This snapshot is called tRNA sequencing.

However, taking this snapshot is incredibly difficult. These tRNA forklifts are wrapped in tight, knotty ropes (secondary structures) and covered in sticky, weird stickers (chemical modifications). When scientists try to scan them, the scanner (an enzyme called Reverse Transcriptase) gets stuck, trips over the knots, and stops working. This results in a blurry, incomplete photo.

To fix this, scientists use special "super-scanners" (called ngRTs) that are strong enough to untangle the knots and read right through the stickers. But here's the problem: buying these super-scanners from big companies is like buying a Ferrari for a daily commute—it works great, but it costs a fortune.

This paper is about building your own Ferrari in your garage for pennies.

Here is the breakdown of what the researchers did, using simple analogies:

1. The "Garage-Built" Super-Scanner (MarathonRT)

The team took the blueprint for a known super-scanner (MarathonRT) and made two major upgrades to make it easier to build and cheaper to run:

• The "Sandwich" Trick: They added a special tag to the end of the enzyme (a Chitin Binding Domain). Think of this like wrapping the enzyme in bubble wrap. It stops the enzyme from clumping together or falling apart during production, keeping it stable and happy.
• The "One-Step" Wash: Old recipes for making this enzyme were like a complex 10-step car wash that required taking the engine out and reassembling it. The team simplified this to a single "rinse and spin" step. They realized they didn't need to strip off the tags (the bubble wrap) to make the enzyme work.
• The Result: They can now make enough of this enzyme from a small bucket of bacteria to run 26,000 tests for the price of a cup of coffee. Commercial versions would cost thousands of dollars for the same amount.

2. The "Magic Filter" for tRNA

Before scanning, you have to separate the tRNA forklifts from the rest of the factory junk (other RNAs).

• The Old Way (Gel Extraction): This was like trying to pick out specific red marbles from a giant jar of mixed marbles by hand, one by one, under a microscope. It took days, was exhausting, and you often dropped some marbles.
• The New Way (Spin Columns): The team tested a new method using a special filter (a spin column). Imagine pouring the whole jar of marbles through a sieve that only lets the red ones (tRNAs) pass through while holding back the big blue ones (junk).
• The Result: This new filter is fast (30 minutes vs. days), cheap, and captures just as many forklifts as the old hand-picking method.

3. The "Taste Test" (Did it work?)

The researchers put their homemade enzyme and the new filter to the test against the expensive, store-bought versions.

• The Verdict: Their homemade enzyme performed just as well as the expensive ones. It read the tRNAs accurately, didn't get stuck on the knots, and produced clear, high-quality data.
• Bonus: They even checked if the "bubble wrap" (the extra tag) changed how the enzyme worked. It didn't! In fact, the tagged version (MRT-CBD) was slightly more stable and performed very similarly to the top commercial brands.

Why Does This Matter?

For a long time, studying tRNA was reserved for wealthy labs with huge budgets because the tools were too expensive and the methods too difficult.

This paper is like giving everyone a DIY kit to build their own super-tools. By showing how to make the enzyme cheaply and how to filter the RNA quickly, they are removing the financial and technical barriers. Now, almost any biology lab can study these tiny but mighty molecules, potentially leading to new discoveries about how cells work and how to treat diseases.

In short: They turned a $5,000 luxury tool into a $0.05 DIY project without losing any quality.

Technical Summary

1. Problem Statement

Transfer RNA (tRNA) sequencing (tRNA-seq) is critical for understanding cellular physiology and disease, but it faces significant technical hurdles:

• Structural Complexity: tRNAs possess robust secondary structures and extensive post-transcriptional modifications (PTMs) that cause conventional reverse transcriptases (RTs) to stall, leading to truncated reads and quantification biases.
• Enzyme Limitations: Current solutions rely on Next-Generation Reverse Transcriptases (ngRTs) like Induro (NEB) and uMRT (RNAConnect). While effective, these commercial enzymes are prohibitively expensive for large-scale studies.
• In-House Production Barriers: Researchers attempting to produce their own ngRTs (e.g., MarathonRT/MRT) using existing protocols face low yields and high costs. Previous protocols were optimized for structural biology (requiring extreme purity via multi-step chromatography and tag cleavage) rather than high-yield production, often resulting in protein precipitation and low recovery.
• Sample Preparation Bottlenecks: Isolating bulk tRNA from total RNA traditionally requires time-consuming, labor-intensive denaturing polyacrylamide gel (PAA) extraction.

2. Methodology

The authors developed a streamlined workflow addressing both enzyme production and tRNA extraction:

A. In-House Enzyme Production (MRT & MRT-CBD)

• Construct Design: They redesigned the MarathonRT construct by adding a C-terminal Chitin Binding Domain (CBD) to create MRT-CBD, effectively "sandwiching" the protein between an N-terminal SUMO tag and a C-terminal CBD tag to enhance solubility.
• Expression Optimization: Protein expression was induced at a lower cell density (OD600 ~0.4–0.5) in E. coli Rosetta2(DE3)pLysS at 16°C overnight, preventing the precipitation issues seen at higher densities.
• Simplified Purification: They eliminated the enzymatic tag-cleavage step and subsequent ion-exchange/gel-filtration chromatography. Instead, they utilized a single-step Ni-NTA affinity purification (gravity flow) followed by buffer exchange and concentration.
• Activity Assay: A colorimetric malachite green assay was established to measure pyrophosphate (Ppi) release, defining enzyme units (U) and allowing for batch-to-batch normalization without expensive reagents.

B. tRNA Extraction Optimization

• Spin Column vs. Gel: The authors benchmarked a rapid silica-based spin column protocol (using NEB Monarch or Macherey-Nagel NucleoSpin RNA XS columns) against the traditional denaturing PAA gel extraction.
• Library Preparation: They followed the mim-tRNAseq protocol, comparing libraries generated using in-house MRT/MRT-CBD against commercial Induro and uMRT.
• Validation: Libraries were sequenced on an Illumina NovaSeqX platform. Additionally, Liquid Chromatography-Mass Spectrometry (LC-MS) was used to verify that the extraction method did not alter PTM profiles.

3. Key Contributions

1. High-Yield, Low-Cost Enzyme Production: A robust, one-step purification protocol yielding milligram quantities of active MRT and MRT-CBD.
2. Cost Reduction: Drastically reduced the cost per reaction from commercial prices to approximately €0.0023–€0.0034 (a ~5,000–8,000-fold reduction).
3. Stability Data: Comprehensive data on enzyme stability over 12 months and through freeze-thaw cycles, identifying optimal storage conditions.
4. Extraction Workflow: Validation of a rapid (30 min), scalable spin-column method for tRNA enrichment that replaces days-long gel extraction.
5. Benchmarking: Rigorous comparison showing that in-house enzymes perform equivalently to commercial standards in complex tRNA-seq workflows.

4. Key Results

Enzyme Production & Stability

• Yield: Produced 7.5 mg of MRT and 5.3 mg of MRT-CBD per 0.5 L culture.
• Activity: Specific activity was ~1.79 U/µg. This yields ~26,957 reactions per 0.5 L culture for MRT and ~18,841 for MRT-CBD.
• Stability: Enzymes stored at –70°C retained full activity for 12 months. Storage at –20°C showed a slight decline (~30% loss) over 12 months. MRT-CBD demonstrated slightly better stability and resistance to freeze-thaw cycles than native MRT, likely due to the CBD tag.
• Optimization: The optimal enzyme-to-template ratio was determined to be 0.5 U per 100 ng of tRNA. Excess enzyme (>0.75 U) caused the formation of inactive high-molecular-weight complexes, reducing cDNA yield.

tRNA Extraction Comparison

• Efficiency: Spin column extraction achieved a 48% recovery rate of the tRNA fraction, compared to 37% for gel extraction.
• Purity: Both methods showed minor 5S rRNA contamination, but spin columns effectively excluded 5.8S rRNA.
• Bias: Principal Component Analysis (PCA) showed distinct clustering between gel and column methods, but read count correlations were high (r = 0.93–0.96) across all 42 cytosolic tRNA isoacceptors.
• PTM Integrity: LC-MS analysis revealed a Pearson correlation of 0.99 in PTM profiles between column and gel extracts, confirming that the spin column method does not bias modification detection.

Sequencing Performance (MRT vs. Commercial)

• Mapping: All libraries achieved >97% uniquely mapped reads.
• Correlation: MRT-CBD showed the highest correlation with commercial enzymes (r = 0.99 with uMRT; r = 0.97 with Induro). Native MRT was slightly more divergent (r = 0.92 with Induro).
• Bias & Errors: All ngRTs (MRT, MRT-CBD, uMRT, Induro) showed similar misincorporation patterns at known modification sites (e.g., positions 9, 26, 58). Induro showed a slight bias toward higher initiator tRNA-Met-CAT counts, while MRT enzymes showed a slightly higher 3' bias, but overall performance was comparable.

5. Significance

This study provides a cost-effective, accessible, and reproducible platform for advanced tRNA research.

• Democratization of tRNA-seq: By reducing the enzyme cost by nearly 8,000-fold and simplifying the purification to a single step, high-quality tRNA sequencing is now feasible for laboratories with limited budgets.
• Workflow Efficiency: Replacing gel extraction with spin columns reduces sample preparation time from 1–2 days to ~30 minutes, significantly increasing throughput.
• Scientific Rigor: The data confirms that in-house produced ngRTs (specifically MRT-CBD) are functionally equivalent to expensive commercial alternatives, enabling reliable quantification of tRNA isoacceptors and PTMs without the need for prior enzymatic demodification.
• Broader Applications: The availability of low-cost, high-activity ngRTs opens doors for their use in other challenging applications, such as RT-PCR on structured templates, RNA structure probing, and direct RNA sequencing.