Reconstitution and biochemical assays of an active human histone pre-mRNA 3′-end processing machinery

Abstract

In animal cells, replication-dependent histone pre-mRNAs are processed at the 3′ end by an endonucleolytic cleavage carried out by the U7 snRNP, a machinery that contains the U7 snRNA and many protein subunits. Studies on the composition of this machinery and understanding of its role in 3′-end processing were greatly facilitated by the development of an in vitro system utilizing nuclear extracts from mammalian cells 35 years ago and later from Drosophila cells. Most recently, recombinant expression and purification of the components of the machinery have enabled the full reconstitution of an active machinery and its complex with a model pre-mRNA substrate, using 13 proteins and 2 RNAs, and the determination of the structure of this active machinery. This chapter presents protocols for preparing nuclear extracts containing endogenous processing machinery, for assembling semi-recombinant and fully reconstituted machineries, and for histone pre-mRNA 3′-end processing assays with these samples.

1. Introduction

Eukaryotic messenger RNA precursors (pre-mRNAs) must undergo processing at the 3′ end before they can be exported to the cytoplasm as mature mRNAs for translation into proteins. The processing events typically consist of cleavage of a pre-mRNA at a specific location followed by the addition of a poly(A) tail. A large machinery with many protein subunits is required for this canonical processing (Shi and Manley, 2015; Sun et al., 2020a; Zhao et al., 1999). The mammalian cleavage factor (mCF) (Chan et al., 2014; Schonemann et al., 2014), with CPSF73, CPSF100 and symplekin as its subunits, catalyzes the cleavage reaction, with CPSF73 as the endonuclease (Mandel et al., 2006).

In animal cells, replication-dependent histone pre-mRNAs are processed through a different mechanism, as they are cleaved at the 3′ end but not polyadenylated (Dominski and Marzluff, 2007; Romeo and Schumperli, 2016). Histone pre-mRNAs contain two distinct sequence elements encompassed within an approximately 60-nucleotide region located less than 100 nucleotides (nts) 3′ of the stop codon (Fig. 1A). The upstream sequence element folds into a conserved stem-loop structure (Table 1), which is recognized by the stem-loop binding protein (SLBP) (Tan et al., 2013; Zhang et al., 2014). Residues 125–200 of human SLBP constitute its RNA binding domain (RBD), which together with residues 201–223 are essential for processing (Dominski et al., 1999). The cleavage site is located between the stem-loop and the second sequence element, the histone downstream element (HDE) (Table 1), which is recognized through base pairing with the 5′ end of U7 snRNA, a component of the U7 snRNP (Bond et al., 1991; Mowry and Steitz, 1987; Schaufele et al., 1986) (Fig. 1A).

Figure 1. The human histone pre-mRNA 3′-end processing machinery.

A. Schematic representation the human of U7 processing machinery determined by biochemical studies using nuclear extracts from mammalian cells. Base pairing between the HDE of histone pre-mRNA (dark blue line) and the 5′ region of U7 snRNA (dark green line) is illustrated by vertical lines. Components of the HCC are indicated in gray and the two U7-specific subunits of the Sm ring are indicated in dark green. B. Processing of wild-type mouse histone pre-mRNA in a complete mouse nuclear extracts or in modified extracts where SLBP or U7 snRNA were pre-bound and hence inactivated by appropriate competitors.

Table 1.

Sequences of RNAs used for structural studies and 3′-end processing assays

RNA	Sequence1
U7 snRNA (1–63)	CAGUGUUACAGCUCUUUUAGAAUUUGUCUAGUAGGCUUUCUGGCUUUUUACCGGAAAGCCCCU
U7 Mut snRNA	CAGUGUUACAGgagaaaUAGAAUUUGUCUAGUAGGCUUUCUGGCUUUUUACCGGAAAGCCCCU
H2a* (52-mer), for structural studies	CCAAAGGCUCUUUUCAGAGCCACCCA↓CUGAAUCAGAUAAAGAGCUGUAACAC
H2a* (60-mer), doubly fluorescently labeled	TAMRA-CCAAAGGCUCUUUUCAGAGCCACCCA↓CUGAAUCAGAUAAAGAGCUGUAACACGGUAGCCA-FAM
H2a* (64-mer), radio-labeled	32P-CAAAAGGCUCUUUUCAGAGCCACCCA↓CUGAAUCAGAUAAAGAGCUGUGACACGGUAGCCGGUCU
H2a* Mut (64-mer), radio-labeled	32P-CAAAAGGCUCUUUUCAGAGCCACCCA↓CUGAAUCAGAUuuucucCUGUGACACGGUAGCCGGUCU
Downstream cleavage product (26-mer)	CUGAAUCAGAUAAAGAGCUGUAACAC-FAM

¹The cleavage sites in the H2a* substrates are indicated with the downward arrow, and the stem-loop is underlined. The HDE in H2a* (bold) can form 15 consecutive Watson-Crick base pairs with the 5′ region of U7 snRNA (bold). In U7 Mut snRNA, 6 nucleotides of this region were substituted with complementary nucleotides (lower case letters). This mutation reduces the length of duplex with H2a* to 9 base pairs and restores 15-base-pair duplex with H2a* Mut containing a compensatory mutation in the HDE (lower case letters). For the 52-mer H2a*, the cleavage reaction produces two 26-mers; hence the 60-mer fluorescently labeled H2a* is used for 3′-end processing assays in order to distinguish the two products.

U7 snRNP is a minor snRNP and exists at low concentrations in most mammalian cells, not exceeding 5×10³ particles per cell. Of the seven proteins in its Sm ring, five are shared with the spliceosomal snRNPs: SmB, SmD3, SmE, SmF and SmG. The two remaining proteins, Lsm10 and Lsm11, are specific for the U7 snRNP, and they replace SmD1 and SmD2 found in the same positions of the spliceosomal ring. The complex of U7 snRNA with the Sm ring is referred to as the U7 Sm core. Lsm11 contains a long N-terminal extension that interacts with an N-terminal segment of the protein FLASH (Yang et al., 2009a; Yang et al., 2011), which forms a coiled-coil dimer (Aik et al., 2017). The two interacting proteins recruit the histone pre-mRNA cleavage complex (HCC) to the U7 snRNP histone pre-mRNA 3′-end processing machinery, also referred to as the U7 machinery in short here. Remarkably, HCC is equivalent to mCF, with CPSF73, CPSF100, symplekin and CstF64 as its subunits. CPSF73 is the endonuclease for the cleavage reaction of histone pre-mRNA 3′-end processing as well (Dominski, 2010; Dominski et al., 2005a), indicating that the cleavage module is shared between the canonical and U7 machineries.

Histone pre-mRNA 3′-end processing has been studied extensively over the past 35 years using nuclear extracts (Gick et al., 1986). Recently, components of the machinery have been expressed recombinantly, purified and studied individually to gain structural information, including SLBP (Tan et al., 2013), FLASH (Aik et al., 2017), and HCC (Zhang et al., 2020). In addition, the U7 Sm core was reconstituted from all seven recombinant Sm/Lsm proteins and synthetic U7 snRNA and shown to support accurate processing of histone pre-mRNA in nuclear extracts by binding endogenous HCC and forming a semi-recombinant machinery (Bucholc et al., 2020). Most importantly, a fully reconstituted machinery, containing human U7 Sm core, FLASH and HCC, has been found to be active in cleaving model histone pre-mRNA substrates in vitro (Sun et al., 2020b; Yang et al., 2020). Such a reconstituted machinery was also crucial in enabling its structure to be determined in complex with a model substrate and SLBP (13 proteins and 2 RNAs), providing the first molecular insights into an active pre-mRNA 3′-end processing machinery (Sun et al., 2020b). This chapter will present protocols for preparing nuclear extracts, semi-recombinant as well as fully reconstituted histone machineries, and for histone pre-mRNA 3′-end processing assays with these samples.

2. Preparation of nuclear extracts for histone pre-mRNA 3′-end processing

Nuclear extracts active in 3′-end processing of replication-dependent histone pre-mRNAs can be prepared using a number of mammalian cell lines, including HeLa cells (Gick et al., 1986), K21 mouse mastocytoma cells (Stauber et al., 1990), U2OS cells (ZD, unpublished results), CHO cells (ZD, unpublished results) and 66–2 mouse myeloma cells (Dominski et al., 1995; Skrajna et al., 2018). Nuclear extracts from HeLa cells contain relatively small amounts of U7 Sm core and FLASH and typically process histone pre-mRNAs inefficiently. This low processing activity can be enhanced by the addition of bacterially expressed N-terminal segment of FLASH, which interacts with endogenous Lsm11 and promotes the recruitment of the HCC that is abundant in HeLa nuclear extracts, converting an inactive U7 Sm core into an active U7 machinery (Yang et al., 2011). Besides mammalian cells, nuclear extracts active in 3′-end processing of histone pre-mRNAs were also obtained from two Drosophila cells lines: S2 (Dominski et al., 2002) and Kc (Dominski et al., 2005b).

Mouse myeloma cells may contain 3–5 fold more U7 snRNP than HeLa cells, reflecting natural differences between cell lines and species (Smith et al., 1991). They yield nuclear extracts that are highly active in 3′-end processing, converting as much as 90% of ³²P-labeled pre-mRNA substrate into mature mRNA during 1 h incubation (Fig. 1B). Extracts from mouse myeloma cells were instrumental in identifying CPSF73 as the catalytic component of the 3′-end processing machinery (Dominski et al., 2005a; Yang et al., 2009b) and were successfully used for one-step purification of U7 machinery and determining its composition by mass spectrometry (Skrajna et al., 2019).

2.1. Equipment

1. Shaker (Eppendorf New Brunswick Innova 44) for bacterial and insect cell cultivation

2. Fluorescence microscope (for insect cell culture)

3. Hemocytometer to count cells

4. Sonicator (Misonix S-4000)

5. Centrifuge with different rotors: Beckman TY.JS4.2 rotor for harvesting cells; F15–8×50c carbon fiber rotor for clearing lysates

6. Beckman GS-6R benchtop refrigerated centrifuge and GH 3.8 Swing Bucket Rotor for collecting cells and nuclei

7. Sorvall RC-5B refrigerated centrifuge with rotor SS-34 for high-speed spinning

8. Eppendorf 5424 benchtop mini-centrifuge

9. Geiger counter

10. Corning stirring plate

11. Tube rotator

12. Water bath

13. Power supply

14. Vertical gel electrophoresis apparatus V16–2 for RNA separation

15. Protein purification system: ÄKTA Pure (Cytiva)

16. HiTrap^™ Heparin HP 5 ml column (Cytiva)

17. HiTrap^™ Q HP 5ml column (Cytiva)

18. Size-exclusion chromatography column (Sephacryl S-300; Cytiva)

19. Superose 6 10/300 GL column (Cytiva)

20. Biological safety cabinet (Thermo)

21. Freezer (−80°C) (Thermo)

22. Nanodrop (Thermo)

23. Protein or RNA electrophoresis equipment (Bio-Rad)

24. ChemiDoc MP imaging system (Bio-Rad)

2.2. Reagents or resources

1. Cellfectin^™ II reagent (Gibco)

2. Transfection medium (Expression Systems)

3. Sf9 cells and Hi5 cells (Expression Systems)

4. ESF 921 medium (Expression Systems)

5. Protease inhibitor cocktail tablet (Sigma), PMSF

6. Ni-NTA resin (Qiagen)

7. HEPES, Tris

8. Isopropyl-β-D-thiogalactopyranoside (IPTG)

9. Diethyl pyrocarbonate (DEPC)

10. Filtered buffer stocks: 1 M Tris–HCl (pH 8.0), 5 M NaCl, 1 M Imidazole, 1 M DTT, 2 M KCl, 0.5 M EDTA (pH 8.0)

11. E. coli DH10MultiBac cells

12. E. coli BL21 (DE3) Star cells

13. X-gal

14. RNasin plus ribonuclease inhibitor (Promega-Fisher).

15. Urea

16. Kanamycin, Ampicillin, Gentamycin

17. Ethanol, isopropanol

18. β-mercaptoethanol, DTT

19. Antimycotic solution for insect cell media

20. LB powder, LB agar tablets

21. Plasmid miniprep purification kit (Thermo)

22. Amicon filter concentrators

23. Purified TEV protease

24. NaCl, KCl

25. Glycerol

26. Spermine, spermidine (Sigma)

27. Streptavidin-agarose (Sigma)

28. Membrane tubing MW cutoff 6,000–8,000 (23 mm) for dialysis of nuclear extracts (Spectrapor)

29. Adenosine 5′-triphosphate (γ−³²P) 3,000 Ci/mmol (PerkinElmer)

30. T4 polynucleotide kinase (New England Biolabs)

2.3. Growing and handling mouse myeloma cells

In mammalian cells, SLBP starts accumulating in cells at the G1/S phase transition and reaches the highest level in S phase, concomitant with DNA replication, when histone mRNAs are generated, and it is rapidly degraded by the proteasome pathway during the G2 phase (Zhang et al., 2014). Similar profile of cell cycle-regulated expression was also observed for FLASH, suggesting that the active form of U7 snRNP, i.e. containing the HCC, may be present at the highest concentrations during S phase (Barcaroli et al., 2006). It is therefore critical that cell cultures used for preparation of nuclear extracts active in processing of histone pre-mRNAs are maintained in the exponential growth phase and contain a high percentage of dividing cells.

Large amounts of mouse myeloma cells (e.g. 20 liter culture and more) can be ordered from Cell Culture Company (C3, Minneapolis, MN). The cells are grown at the company’s production site at 37 °C in the presence of 5% CO₂ in Joklik’s Minimal Essential Medium (MEM) containing 10% heat inactivated horse serum at the company’s production site to a density of 1.2–1.4×10⁶ cells/ml. Mouse myeloma cells are exceptionally fragile and their inappropriate handling during harvesting and extract preparation can lead to uncontrolled nuclear lysis and significant reduction of both the volume and protein concentration of the resultant extract. Immediately prior to shipment, the cells are gently harvested in separate 250 ml conical tubes using Beckman GH 3.8 swinging bucket rotor under gentle conditions of centrifugation (5 min at 2,000 rpm or 650 × g), avoiding multiple spinning of the same cell pellet. Cell pellets are softened by tube swirling or tapping, and combined by gentle pipetting using a small amount of growth media. Tubes containing cell suspension are shipped on wet ice. Upon arrival the following day, the cells are used immediately for preparation of nuclear extract. Smaller amounts of mouse myeloma cells (1–5 liters) can be grown at 37 °C in the laboratory using spinner flasks of appropriate size and 5% CO₂ incubator.

2.4. Preparation of nuclear extracts from mouse myeloma cells

Mouse myeloma cells are processed using a modified method of Dignam et al. (Abmayr et al., 2006; Dignam et al., 1983a; Dignam et al., 1983b), with most modifications incorporated into the method serving to stabilize the integrity of the nuclear envelope and to limit the extent of nuclear lysis during high salt extraction (Dominski et al., 1995).

Nuclear extracts with the highest activity in supporting 3′-end processing of histone pre-mRNAs are routinely obtained by adjusting the final KCl concentration during extraction to 0.25 M, significantly lower than the 0.42 M required to yield extracts active in transcription (Dignam et al., 1983a; Dignam et al., 1983b) or splicing (Krainer et al., 1984). This concentration is well tolerated by nuclei isolated from mouse myeloma cells and is sufficient to extract all essential components of the processing machinery. Higher salt concentrations bring a risk of nuclear breakage during extraction and yield extracts with lower processing activity, possibly as a result of extracting uncharacterized inhibitors or irreversibly disrupting U7 processing machinery.

The protocol given below describes how to prepare active extracts from mouse myeloma cells. All steps during the preparation of nuclear extracts are conducted on ice or in a cold room, using pre-cooled solutions, centrifuge tubes and glassware. Note that this protocol is a list of specific recommendations and suggestions, and some critical steps may need to be optimized experimentally to adjust for differences among cell lines, growth conditions and available equipment.

1. Collect cells shipped as a high concentration suspension in growth medium by spinning 5 min at 650 × g and gently aspirate the supernatant using a vacuum pump, making sure that no residual medium is left over the cell pellet.

2. Estimate the volume of the pellet and loosen it by gently tapping the bottom of the tube. This step is important as it eliminates the need for excessive pipetting in the following step.

3. Resuspend cells in 5 volumes of hypotonic buffer (10 mM HEPES-KOH pH 7.9, 1.5 mM MgCl₂, 20 mM KCl, 0.5 mM DTT, 0.75 mM spermidine, 0.15 mM spermine) by slowly pipetting up and down the suspension and occasionally swirling the tube.

   Tip: DTT, spermidine and spermine are added to buffers just before use from the following stock solutions kept frozen or freshly prepared: 1 M DTT (2,000×), 75 mM spermidine (100×), and 15 mM spermine (100×). KCl in all buffers can be replace by NaCl without noticeable loss of activity in 3′-end processing of histone pre-mRNAs.

4. Transfer cell suspension to a Dounce homogenizer of appropriate size and allow cells to swell on ice for 10 minutes.

   Tip: In contrast to the extremely fragile mouse myeloma cells, HeLa cells and some other cell types, including Drosophila cells, typically require one wash with hypotonic buffer prior to the cell swelling step. The cells are gently centrifuged immediately after adding 5 volumes of the hypotonic buffer, and resuspended in 2 volumes of the same buffer for the subsequent 10 min swelling step, as in the original protocol (Abmayr et al., 2006; Dignam et al., 1983a; Dignam et al., 1983b). This extra washing step removes residual amounts of the media trapped in the pellet and enhances cell swelling. Since it is followed by adding only two volumes of hypotonic buffer for cell lysis, it also ultimately results in a more concentrated cytoplasmic fraction that can be used for other purposes. If required, this protocol can be used for preparation of nuclear and cytoplasmic extracts from mouse myeloma cells, although it may compromise the integrity of nuclei during the extraction step.

5. Place a tight pestle in the homogenizer and move it up and down to lyse the cells, avoiding fast and forceful strokes. After 10 initial up and down strokes, check the ratio of unbroken cells to free nuclei under a phase-contrast microscope by placing ~50 μL of cell suspension on a glass slide and covering the drop with a cover slip. Use 5 additional strokes if the percentage of unbroken cells is larger than 20% but do not continue if no additional lysis occurs. Note that this is one of the most critical steps in extract preparation and it should be carefully monitored as different cell batches and different cell lines require different time for swelling and different number of strokes for lysis. As a rule of thumb, use minimal number of strokes to get sufficient percentage of free nuclei in the suspension.

6. Add 1/10 volume of restore buffer (67.5% sucrose in hypotonic buffer) and spin 5 min at 1,500 × g. This step is a modification of the original protocol of Dignam et al. (Dignam et al., 1983a; Dignam et al., 1983b) and is intended to stabilize the nuclei and prevent their lysis during subsequent steps of the protocol.

7. Carefully remove the supernatant containing the cytosolic fraction and mitochondria, and estimate the volume of the darker and more compact nuclear pellet at the bottom of the tube.

8. Soften the nuclear pellet by gently tapping the bottom of the tube with a finger, add ¼ pellet volume of low salt extraction buffer (20 mM HEPES-KOH pH 7.9, 1.5 mM MgCl₂, 20 mM KCl, 0.5 mM DTT, 0.2 mM EDTA, 25% glycerol, 0.75 mM spermidine, 0.15 mM spermine) and resuspend the nuclei by gently swirling the tube.

9. Use a pipette of appropriate size to complete the process of resuspending the nuclei by slowly pipetting the suspension up and down. Transfer the resuspended nuclei to a glass beaker of an appropriate size placed on ice.

10. Use an additional ¼ volume of low salt buffer to rinse the nuclei remaining in the centrifuge tube and in the pipette and combine the recovered volume with the nuclei already placed in the beaker.

    Tip: Make sure to carefully record the volume of the suspension that is being collected in the beaker. For smaller scale preparations, this could be facilitated by using a graduated conical centrifuge tube. Note that the suspension is relatively thick and partially viscous, leaving a sizeable fraction inside the pipette, potentially leading to overestimation of the transferred volume. Note also that the overall volume of low salt extraction buffer added to resuspend nuclei equals 50% of the nuclear pellet. This amount could be reduced to 30–40%, yielding more concentrated and potentially more active nuclear extracts.

11. Insert a magnetic bar of an appropriate size into the suspension of the nuclei, place the beaker on ice over a stirring plate and set up the speed of stirring to achieve visible mixing.

12. Use the following formula to calculate the volume (X) of high salt extraction buffer required to give final salt concentration of 0.25 M: X = 0.25×Y/0.95, where Y is the volume of nuclear suspension in low salt extraction buffer.

    Tip: The final salt concentration during the extraction step will depend on the accuracy of measuring the volume of the transferred nuclear suspension. Overestimation of the volume may result in the actual salt concentration during the extraction being much higher than anticipated, causing excessive lysis of nuclei during the extraction step, and hence reducing protein concentration, processing activity and total volume of the extract. Nuclei of mouse myeloma cells, when properly handled, should survive extraction in 0.25 M salt with only limited nuclear lysis.

13. While constantly stirring the suspension, drop-wise add high salt extraction buffer (20 mM HEPES-KOH pH 7.9, 1.5 mM MgCl₂, 1.2 M KCl, 0.5 mM DTT, 0.2 mM EDTA pH 8.0, 25% glycerol, 0.75 mM spermidine, 0.15 mM spermine). Note that high salt buffer has the same composition as low salt buffer with the exception that 20 mM KCl is replaced with 1.2 M KCl. Avoid adding larger amounts of the buffer in short time as this can create local high salt concentration and accelerate lysis of nuclei.

14. Reduce the intensity of stirring and extract nuclei for 45–60 min. Limited lysis may occur during this time, as indicated by the presence of a viscous clump of released chromatin in the center of the beaker.

15. Spin down the extracted nuclei by centrifugation at 12,000 × g for 30 min and carefully collect the supernatant.

16. Transfer the supernatant to dialysis tube (12–14 kDa MWCO) that was boiled in deionized water for 15 min prior to using or prepared as recommended by the manufacturer. Leave a small air bubble in the tube that will facilitate its flow in a vertical orientation during dialysis. Make sure that the two open ends are securely sealed with dialysis clips by squeezing the liquid in the tube and checking for potential leaks.

17. Place the dialysis tube in 100× volume of ice-cold dialysis buffer (20 mM HEPES-KOH pH 7.9, 100 mM KCl, 0.5 mM DTT, 0.2 mM EDTA, 20% glycerol) in a beaker or cylinder containing a magnetic bar and dialyze on a stirring plate in a cold room for 1.5–2 h, making sure that the tube slowly spins in the buffer. Transfer the tube to a fresh change of dialysis buffer and continue dialyzing for additional 1.5–2 h.

18. Transfer the dialyzed nuclear extract to a centrifuge tube and spin at 12,000 × g for 20 min to remove visible precipitates.

19. Aliquot the nuclear extract, snap freeze in dry ice and store at −80 °C. The extracts can be stored at this temperature for decades and survive multiple thawing-freezing cycles without losing activity in 3′-end processing of histone pre-mRNAs.

Typically, 1 liter myeloma cell culture cultured in suspension to a density of 1×10⁶ cells/ml yields 1 ml of a highly active nuclear extract with a protein concentration approaching 8–10 mg/ml. The yield of the extract and the reproducibility of the protocol can be substantially improved by scaling up the cell culture to 10–20 liters or more. The same protocol can be used to prepare nuclear extracts from other mammalian cell lines and was also successfully used to generate highly active nuclear extracts from two Drosophila cell lines grown in suspension: Schneider’s line S2 (S2) and Kc (Dominski et al., 2003, 2005b; Dominski et al., 2002; Sabath et al., 2013). As indicated above, both HeLa and Drosophila cells are relatively sturdy and their efficient breakage into the nuclear and cytoplasmic fractions typically requires one wash with the hypotonic buffer (see above) after the cells are harvested. Spermine and spermidine, the two related polyamines added to buffers to prevent nuclear breakdown and chromatin leakage, are omitted when working with HeLa and Drosophila cells.

3. Reconstitution of an active human histone pre-mRNA 3′-end processing machinery

For the U7 Sm core, the expression of SmB, SmD3, SmE, SmF and SmG in E. coli as SmB-SmD3 and SmE-SmF-SmG sub-complexes, and the purification conditions, have been established in previous reports on spliceosomal snRNPs (Grimm et al., 2013; Kambach et al., 1999; Leung et al., 2010; Leung et al., 2011; Pomeranz Krummel et al., 2009; Raker et al., 1999). FLASH N-terminal segment can be readily expressed in E. coli as well (Aik et al., 2017). However, SLBP produced from E. coli is not stable on its own and readily precipitates when concentrated. HCC and Lsm10-Lsm11 heterodimer are uninduced or insoluble when expressed in E. coli. Therefore, expression in insect cells using MultiBac technology (Sari et al., 2016) (Geneva Biotech) is used for SLBP, HCC and Lsm10-Lsm11. For HCC, symplekin must be co-expressed with CPSF73, CPSF100 and CstF64, but the protein yield is quite low. To improve the yield, adding a SUMO tag at the N terminus of symplekin is necessary. Similarly, an MBP tag is added at the N terminus of Lsm10 to increase solubility. After cell lysis, the proteins are purified by nickel affinity followed by either ion exchange or size-exclusion chromatography. A detailed list of the constructs, expression and purification conditions is given in Table 2. The first step of U7 machinery reconstitution is the formation of U7 Sm core (Bucholc et al., 2020), which is then combined with the other components (Sun et al., 2020b). The active fully reconstituted machinery contains 13 recombinant proteins and 2 RNAs.

Table 2.

List of constructs, expression and purification conditions

Protein(s)	Residue ranges	Tags*	Vectors	Expression conditions	Purification conditions
Lsm10-Lsm11	Lsm11 1–360 (∆211–322) Lsm10 1–123	N-6×His on Lsm11 N-MBP on Lsm10	pFL	Hi5 cell	Nickel affinity HiTrap™ Heparin HP column
SmB-SmD3	SmB 1–95 SmD3 1–126	N-6×His on SmD3	SmB in pCDF Duet SmD3 in pET-28a	Co-express in E. coli BL21 Star (DE3) strain	Nickel affinity HiTrap™ Heparin HP column
SmE-SmF-SmG	SmE 1–92 SmF 1–86 SmG 1–76	C-6×His on SmG	SmG in pET-26b SmE and SmF in pCDF Duet	Co-express in E. coli BL21 Star (DE3) strain	Nickel affinity HiTrap™ Heparin HP column
FLASH	51–137 C54S/C83A mutant	C-6×His	pET-26b	E. coli BL21 Star (DE3) strain	Nickel affinity Size-exclusion chromatography
HCC	Symplekin 30–1101 CPSF73 1–684 CPSF100 1–782 CstF64 1–597	N-6×His-SUMO on symplekin	Symplekin in pFL CstF64 in pSPL fused with pFL containing CPSF73 and CPSF100	Co-infect Hi5 cell	Nickel affinity HiTrap™ Q HP column
SLBP	125–223	N-6×His	pFL	Hi5 cell	Nickel affinity HiTrap™ Q HP column

^*N is short for N-terminal; C is short for C-terminal.

3.1. Expression and purification from bacteria (SmB-SmD3, SmE-SmF-SmG and FLASH)

1. Inoculate 100 ml LB containing appropriate antibiotic(s) with a colony of bacteria harboring construct(s) of interest and grow overnight with vigorous shaking at 37 °C, as starter culture.

2. Inoculate 6 liter LB containing appropriate antibiotic(s) with the starter culture (10 ml/l) and incubate with vigorous shaking at 37 °C until OD₆₀₀ reaches 0.8–1.

3. Induce protein expression by adding isopropyl β-D-1-thiogalactopyranoside (IPTG) to a final concentration of 0.4 mM and grow at 20 °C for 18 h.

4. Harvest cells by centrifugation at 4,000 rpm in a Beckman TY.JS4.2 rotor for 15 min at 4 °C and discard the supernatant.

5. Resuspend cell pellet in 100 ml lysis buffer containing 20 mM Tris (pH 7.5), 500 mM NaCl, 10 mM imidazole, 5% (v/v) glycerol, 17.8 μg/ml PMSF, 10 mM β-mercaptoethanol.

6. Sonicate the cell suspension at 70% output for 6 min with 1 s on and 2 s off (Misonix S-4000).

7. Centrifuge at 13,000 rpm in a F15–8×50c carbon fiber rotor at 4 °C for 30 min.

8. Equilibrate 250 μl Ni-NTA resin in each of two 50 ml Falcon tubes with 4 volumes of Ni wash buffer containing 20 mM Tris (pH 7.5), 500 mM NaCl, 40 mM imidazole, and 10 mM β-mercaptoethanol.

9. Transfer 50 ml of the supernatant to equilibrated Ni-NTA resin in each tube and slowly rotate for 1 h in a cold room.

10. Transfer the mixture to a 14 cm high, gravity flow column (Bio-Rad).

11. Wash resin with 3×15 ml Ni wash buffer.

12. Elute protein with 4 ml Ni elution buffer containing 20 mM Tris (pH 7.5), 500 mM NaCl, 500 mM imidazole, 5% (v/v) glycerol, and 10 mM β-mercaptoethanol.

    Tip: For SmB-SmD3 complex, the elution buffer needs 600 mM NaCl.

13. Dilute with 6 ml buffer A (20 mM HEPES (pH 7.5) and 5 mM DTT) to ~200 mM NaCl.

14. Load onto HiTrap^™ Heparin HP 5 ml column (Cytiva), and elute the protein complex with a salt gradient starting with 80% buffer A and 20% buffer B (20 mM HEPES (pH 7.5), 1 M NaCl, 5 mM DTT) to 100% buffer B.

    Tip: For SmB-SmD3 complex, load onto Heparin column without dilution, since dilution will cause protein precipitation. The complex is eluted with 40–100% buffer B.

15. For FLASH, skip steps 13 and 14. Instead, load onto size-exclusion chromatography column (Sephacryl S-300; Cytiva) with a buffer containing 20 mM Tris (pH 8.5), 250 mM NaCl, and 5 mM DTT.

16. Run SDS-PAGE gel to analyze the peak fractions.

17. Concentrate the factions of interest to 5~10 mg/ml. Divide into aliquots, snap freeze in liquid nitrogen and store at −80 °C.

3.2. Expression and purification from insect cells (SLBP, Lsm10-Lsm11 and HCC)

1. Bacmid preparation: Transform the final plasmids into competent E. coli DH10MultiBac cells and grow on plates at 37 °C for 48 h with appropriate antibiotics, IPTG and X-gal for color selection. Select positive/white clones for bacmid extraction. For more detailed information refer to the MultiBac protocol (Sari et al., 2016).

2. P1 baculovirus production: Pre-incubate 1.5 μg bacmid DNA and 8 μl Cellfectin^™ II reagent (Gibco) in 200 μl Transfection medium (Expression Systems) at room temperature for 30 min, then supplement with Transfection medium to 1 ml. For every bacmid DNA, seed 0.8–1×10⁶ freshly diluted Sf9 cells (Expression Systems) in a 6-well tissue culture plate. Use the 1 ml Cellfectin-DNA suspension to replace the supernatant from seeded cell. Incubate overnight (~16 h) at 27 °C. Change the supernatant to 2 ml ESF 921 medium (Expression Systems). Incubate at 27 °C for 4~5 days. Collect supernatant (P1 virus) by centrifugation for 5 min at 5,000 rpm and store in a 2 ml sterile Eppendorf tube at 4 °C.

   Tip: The standard protocol only requires incubating the 1 ml Cellfectin-DNA suspension and the seeded cell for 5 h at 27 °C. However, we have found that in many cases incubation overnight can give viruses that produce more recombinant protein.

3. P2 baculovirus production: Culture 50 ml Sf9 cells to a density of 1.5~2×10⁶ cells/ml and add all P1 virus to the cells, then incubate at 27 °C at 120 rpm. After 3 days, harvest the cells by centrifugation for 10 min at 2,000 rpm, then transfer the supernatant (P2 virus) into a 50 ml sterile Falcon tube and store at 4 °C.

   Tip: The quality of the viruses, in terms of how much recombinant protein they can produce, will decay over time. P1 virus can be stored at 4 °C for less than 3~4 weeks or at −80 °C for one year. P2 virus is usually stored at 4 °C for 1–2 months, but the virus for HCC can be stored for much shorter time (~2 weeks) before it loses the ability to produce recombinant protein. It is best to immediately use the P1 virus for amplification and P2 virus for protein production. The cell pellets can be stored at −80 °C for a long time.

4. Protein production in Hi5 cells: Infect or co-infect 1 liter Hi5 cells (1.8×10⁶ cells/ml, Expression Systems) with 25 ml corresponding P2 virus(es) (2.5% (v/v)) in a 3 liter flask, then incubate at 27 °C for 48 h, shaking at 120 rpm.

   Tip: Steps 2–4 should be performed in a sterile hood. P1 and P2 viruses need to be protected from light. Florescent proteins had been co-integrated into the MultiBac baculoviral genome (Sari et al., 2016), so its expression can be monitored by fluorescence spectroscopy and become the marker for the expression analysis of the respective proteins. The strength of P2 viruses can vary between different batches. Therefore, the amount of virus for infection can vary from 1 to 5% (v/v).

5. Harvest cells by centrifugation at 2,000 rpm in a Beckman TY.JS4.2 rotor for 13 min at 4 °C, discard the supernatant, flash freeze in liquid nitrogen and store at −80 °C.

6. Resuspend cell pellet in 100 ml lysis buffer containing 20 mM Tris (pH 7.5), 500 mM NaCl, 10 mM imidazole, 5% (v/v) glycerol and 1 protease inhibitor cocktail tablet (Sigma). For HCC complex, the lysis buffer is 25 mM Tris (pH 8.0), 300 mM NaCl and 1 protease inhibitor cocktail tablet.

7. Sonicate the cell suspension at 70% output for 6 min with 1 s on and 2 s off.

8. Centrifuge at 13,000 rpm in a F15–8×50c carbon fiber rotor at 4 °C for 45 min.

9. Equilibrate 250 μl Ni-NTA resin in each of two 50 ml Falcon tubes with 4 volumes of Ni wash buffer.

10. Transfer 50 ml of the supernatant to equilibrated Ni-NTA resin in each tube and slowly rotate for 1 h in a cold room.

11. Transfer the mixture to a 14 cm high, gravity flow column (Bio-Rad).

12. Wash resin with 3×15 ml Ni wash buffer containing 20 mM Tris (pH 7.5), 500 mM NaCl and 40 mM imidazole. For HCC complex, the Ni wash buffer is 25 mM Tris (pH 8.0), 150 mM NaCl and 20 mM imidazole.

13. Elute protein with 4 ml Ni elution buffer containing 20 mM Tris (pH 7.5), 500 mM NaCl, 500 mM imidazole, and 5% (v/v) glycerol. For HCC complex, the Ni elution buffer is 25 mM Tris (pH 8.0), 150 mM NaCl and 250 mM imidazole. 10 mM β-mercaptoethanol can also be included in the wash and elution buffers.

14. Dilute with 6 ml buffer A (20 mM HEPES (pH 7.5) and 5 mM DTT) to ~200 mM NaCl.

    Tip: It is necessary for Lsm10-Lsm11 to be eluted in a buffer with high concentration of salt first, then slowly diluted a little bit for ion exchange. Lsm10-Lsm11 tends to precipitate if eluted directly in a buffer with ~200 mM NaCl.

15. Load onto HiTrap^™ Heparin HP 5 ml column (Cytiva), and elute the protein with a salt gradient starting with 80% buffer A and 20% buffer B (20 mM HEPES (pH 7.5), 1 M NaCl, 5 mM DTT) to 100% buffer B. For SLBP, load onto HiTrap^™ Q HP column. For HCC, skip step 14. Instead, load onto HiTrap^™ Q HP column and elute with a gradient starting with 15% to 60% buffer B (25 mM Tris pH 8.0, 1 M NaCl, 5 mM DTT).

16. Concentrate the factions of interest. Aliquot, snap freeze and store at −80 °C.

3.3. Reconstitution of U7 Sm core in complex with FLASH, SLBP and histone pre-mRNA

1. Anneal an equimolar mixture of human U7 snRNA and modified mouse H2a pre-mRNA (H2a*) in 100 μl reconstitution buffer A containing 20 mM HEPES (pH 7.5), 500 mM NaCl, 5 mM EDTA, and 5 mM DTT by heating to > 90 °C for 5 min. The sequences of the RNAs are given in Table 1.

2. Snap cool on ice for 10 min.

3. Add equimolar Lsm10-Lsm11, SmE-SmF-SmG, SmB-SmD3, the annealed RNAs, and 2 molar equivalents of FLASH into appropriate volume of reconstitution buffer A. The total volume should be under 900 μl.

   Tip: In most assembly experiments, N-terminal FLASH segment contained two point mutations, C54S/C83A, to prevent potential formation of disulfide bridges between two FLASH molecules in its dimer (Aik et al., 2017). However, wild-type FLASH segment is equally suitable.

4. Incubate the mixture at 30 °C for 30 min, followed by 37 °C for 15 min (Leung et al., 2010).

5. Cool on ice for 10 min.

6. Add equimolar SLBP to the mixture and incubate on ice for 5 min.

7. Add appropriate volume of 2.5 mg/ml TEV protease (at a ratio of 1:7 (w/w)) to remove the MBP tag on Lsm10. Incubate overnight at 4 °C. The total volume should be under 1 ml. The SUMO tag on symplekin is not removed.

8. Purify U7 Sm core-FLASH-SLBP-H2a* complex by gel filtration using a Superose 6 10/300 GL column (Cytiva), in reconstitution buffer B containing 20 mM HEPES (pH 7.5), 100 mM NaCl, 10 mM EDTA, and 5 mM DTT.

   Tip: Some components of U7 Sm core are unstable and prone to precipitation in low salt buffer, e.g. Lsm10-Lsm11 and SmB-SmD3. Therefore, to reconstitute U7 Sm core, the buffer must have high concentration of salt, otherwise the proteins will precipitate or aggregate. The high salt condition also promotes the spontaneous assembly of the three Sm sub-complexes on the U7 snRNA to form the heptameric ring. However, the high salt concentration will prevent interaction with HCC. After U7 Sm core is formed, the complex can tolerate low salt buffer, allowing a change to the low salt buffer during SP6 purification.

9. Fractions of interest are concentrated and stored at −80 °C. The presence of RNA in the complex is confirmed with an A₂₆₀/A₂₈₀ ratio of ~1.7.

3.4. Reconstitution of the active processing machinery in complex with pre-mRNA substrate and SLBP

The two processing signals required for 3′-end processing of histone pre-mRNAs, the stem-loop and the HDE, are encompassed within an approximately 60-nucleotide RNA segment (Fig. 1A). This model pre-mRNA substrate can be chemically synthesized by commercial sources (for example Horizon Discovery/Dharmacon, Lafayette, CO or Integrated DNA Technologies, Inc., Coralville, IA) or prepared by in vitro transcription. Multiple assembly experiments demonstrated that chemical synthesis and T7/SP6-mediated transcription are equally suitable methods of generating U7 snRNA for U7 snRNP assembly (Bucholc et al., 2020; Sun et al., 2020b; Yang et al., 2020).

1. Mix purified HCC and U7 Sm core-FLASH-SLBP-H2a* complex at a molar ratio of 1:1.3, in reconstitution buffer B.

   Tip: It is important that all manipulations with RNA, its handling and storage are carried out in an RNase-free environment, paying particular attention to avoid contaminations with the powerful RNase A routinely used for DNA isolation from bacterial cells and other general laboratory techniques.

2. Incubate on ice for 1 h.

3. Purify the histone pre-mRNA 3′-end processing machinery substrate complex using a Superose 6 10/300 GL column, in reconstitution buffer B.

This purified complex is used primarily for structural studies. For cleavage assays, the active U7 machinery can be assembled in situ, by combining its various components in the reaction buffer (see section 4).

3.5. An alternative protocol for assembling U7 Sm core-FLASH complex

This protocol differs in some details from the protocol described above and was successfully used to reconstitute active processing machinery for functional studies with radio-labeled substrate.

1. Mix 75 μl of the assembly buffer (600 mM KCl, 15 mM HEPES pH 7.9, 15% glycerol, 0.1 μg/μl yeast tRNA and 20 mM EDTA) with 2,000 pmols each of Lsm10-Lsm11, SmE-SmF-SmG, and U7 snRNA, and incubate 90 min at 32 °C.

2. Add 2,000 pmols of SmB-SmD3 and, if necessary, bring up the volume to 100 μl with the assembly buffer. Incubate at 32 °C for an additional 90 min.

3. Add 4,000 pmols (2-fold molar excess compared to the other subunits) of bacterially expressed FLASH (amino acid residues 51–137) and incubate overnight at 4 °C.

4. Separate U7 Sm core-FLASH complex from the unbound components by size-exclusion chromatography using Superose™ 6 Increase 3.2/300 (Cytiva) or other suitable column. To equilibrate the column, use buffer compatible with in vitro 3′-end processing reaction: 75 mM KCl, 15 mM HEPES-KOH pH 7.9, 5% glycerol and 20 mM EDTA pH 8. Note that compared to processing buffer, the concentration of glycerol was reduced from 15% to 5% to improve liquid flow during purification.

5. Combine fractions containing the assembled complex, evaluate its concentration and analyze a fraction by SDS-PAGE to confirm the presence of all subunits. Store at −80 °C.

4. Histone pre-mRNA 3′-end processing assays using radio-labeled substrate

4.1. Labeling of synthetic histone pre-mRNA at the 5′ end with ³²P

1. Set up a 5′-end labeling reaction in a final volume of 30 μl by mixing the following components: 3 pmol of synthetic histone pre-mRNA substrate containing 5′ OH group (Table 1), 3 μl of 10× buffer (New England Biolabs, NEB, Ipswich, MA), 1.5 μl of T4 polynucleotide kinase (NEB), 25 μCi of γ−³²P-ATP, and water to total volume of 30 μl.

   Tip: Chemically synthesized RNAs contain a 3′-end hydroxyl group and are ready for 5′-end labeling without prior treatment with calf intestinal phosphatase.

2. Incubate 1 h at 37 °C.

3. Run through G50 spin column (Cytiva) to remove unincorporated radioactive ATP, following manufacturer’s protocol.

4. Use Geiger counter to measure the amount of radiation (counts per minute, cpm) in 1 μl of the purified probe. Depending on the efficiency of labeling, the probe can be used for as long as 4–6 weeks.

Tip: An RNA segment containing the two processing signals can be alternatively generated by in vitro transcription using a bacteriophage RNA polymerase (SP6 or T7) and an appropriate DNA template (a PCR-generated segment or a linearized plasmid). This method is recommended for generating longer pre-mRNA substrates that exceed the maximum length limit suitable for chemical synthesis (typically over 120 nucleotides). The generated RNA transcript is next treated with calf intestinal phosphatase (NEB) to remove the 5′ triphosphate and labeled at the 5′ end with ³²P using T4 polynucleotide kinase. The quality of labeled RNAs generated by these procedures should be checked at each step by electrophoresis in a denaturing gel.

4.2. Assays using nuclear extract

In vitro processing of histone pre-mRNAs is not inhibited by the presence EDTA, with a concentration as high as 50 mM having no major effect on the efficiency of the reaction (Kolev et al., 2008) (ZD, unpublished results). Cleavage of histone pre-mRNAs is catalyzed by CPSF73, a member of the metallo-β-lactamase family of nucleases (Callebaut et al., 2002; Dominski, 2007) and contains two zinc ions required for catalysis (Mandel et al., 2006). Structural and biochemical studies indicate that the zinc ions are tightly coordinated by a number of amino acids of the catalytic center and are not accessible for binding EDTA (Kolev et al., 2008). Routinely, in vitro processing reactions are carried out in the presence of 20 mM EDTA (Gick et al., 1986), a concentration that effectively inhibits the activity of metal-dependent nonspecific nucleases and phosphatases.

1. Set up a processing reaction on ice by combining 7.5 μl of nuclear extract and 2.5 μl of 80 mM EDTA (pH 8) containing pre-mRNA substrate labeled at the 5′ end with ³²P. A typical 10 μl reaction contains 5,000 counts per minute of the radioactive substrate, which equals, depending on the efficiency of 5′ labeling, 0.01 to 0.025 pmole (1 to 2.5 nM) or 0.2 to 0.5 ng of single-stranded RNA (60 nucleotides of length and 20,000 kDa of molecular weight).

   Tip: If desired, the reaction can be scaled up, or lower amounts of nuclear extracts can be tested, with the remaining volume made up using dialysis buffer. To achieve high ratio of final product relative to input substrate, use the smallest amount of the substrate possible. For older probes that have lost most of their radioactive signal due to short half-life of ³²P isotope (~14 days), keep the amount of substrate at constant level, extending instead the time of subsequent exposure by autoradiography.

2. Mix by gentle vortexing and incubate at 32 °C for 60 min.

   Tip: Carry out a time course reaction to establish the shortest time of incubation necessary to achieve maximum processing efficiency (Dominski et al., 1995; Gick et al., 1986). Cleavage of histone pre-mRNAs proceeds rapidly at 32 °C without a significant lag time and in some extracts most of the cleavage product is generated within 20–30 min after the start of incubation. Once the reaction reaches the plateau phase, further extension of this time does not significantly increase the amount of the final product and may instead cause its non-specific degradation, in particular in samples where EDTA was omitted or used at low concentration.

3. Stop the reaction by adding 2 μl of proteinase K solution (5 μg/μl) in 50 mM Tris pH 7.5 and 2.5% SDS, and incubate 1 h at 37 °C.

   Tip: Alternatively, each sample can be processed by adding 200 μl of 0.3 M sodium acetate, followed by extraction with the same volume of phenol followed by precipitation of the aqueous phase with 500 μl ethanol. RNA precipitated at −20 °C is recovered by centrifugation, rinsed with ethanol, air dried and dissolved in 10–15 μl of loading dye (8 M urea, 0.01% bromophenol blue and 0.01% xylene cyanol). This approach takes longer time, may result in partial loss of RNA and is relatively laborious when multiple samples are processed, but it produces samples of much higher purity and is recommended for longer RNA substrates (>150 nucleotides) that may run during electrophoresis as a smear rather than a sharp band if proteinase K is used for their purification.

4. Add 40 μl of loading dye (8 M urea, 0.01% bromophenol blue and 0.01% xylene cyanol), and boil 5 min to denature RNA.

5. Cast an 8% polyacrylamide/7 M urea denaturing gel (30:1 acrylamide to bisacrylamide) using 20 cm × 20 cm gel plates and 0.75 mm spacer and pre-run the gel for at least 15–20 min to increase its temperature and remove contaminating salt ions.

6. Rinse the wells from excess of urea using a syringe and load 5 μl of each sample.

7. Apply appropriate voltage to keep the gel at 45–50 °C (for a gel 0.75 mm thick this will typically require 400–500 Volts) and run until desired separation of the labeled input pre-mRNA and the 5′ cleavage product is achieved. For a 60 nucleotide RNA and its product of 45 nucleotides (Fig. 1A), this takes between 20 and 25 min.

8. Manually transfer the gel to Whatman paper, place in a cassette with an X-ray film and intensifying screen and expose at −80 °C.

The 5′-labeled substrate is cleaved in the extract yielding two products. The upstream product that terminates with the stem-loop retains the label at the 5′ end and is visible upon autoradiography. The downstream product lacks radioactive label and its fate cannot be followed with this substrate. Studies with uniformly or site-specifically labeled substrates demonstrate that this product is degraded in a U7-dependent manner by a 5′−3′ exonuclease activity of CPSF73 (Dominski et al., 2005a; Yang et al., 2009b; Yang et al., 2020).

4.3. Assays using semi-recombinant machinery

Semi-recombinant U7 machinery is assembled by adding purified recombinant U7 Sm core-FLASH N-terminal segment complex to a nuclear extract. Nuclear extracts contain relatively large amounts of free HCC (Yang et al., 2013) that spontaneously assembles with U7 Sm core-FLASH, giving rise to semi-recombinant U7 machinery (Bucholc et al., 2020). Studying processing activity of semi-recombinant U7 machinery requires that its activity is not obscured by the activity of the endogenous machinery from the extract. One way of achieving this objective is to assemble U7 Sm core on a mutant U7 snRNA (U7 Mut) that contains an extensive mutation within its 5′-end region and analyze the activity of semi-recombinant U7 machinery together with a mutant histone pre-mRNA containing a compensatory mutation within the HDE (HDE Mut) that restores the base pairing interaction with the mutant U7 snRNA (Table 1), as previously described (Bond et al., 1991; Schaufele et al., 1986). This mutant substrate is not recognized by the endogenous machinery, with the entire processing detected in the extract being contributed by the semi-recombinant machinery.

1. Add 1 pmol of recombinant U7 Sm core that was assembled on U7 Mut snRNA and bound to N-terminal FLASH to a 10 μl processing reaction containing 7.5 μl of nuclear extract, 20 mM EDTA and 0.025 pmol (2.5 nM) of radio-labeled HDE Mut pre-mRNA substrate.

   Tip: Titration experiments demonstrated that this amount of nuclear extract contains sufficient amount of free HCC to assemble 1 pmol of U7 Sm core into an active semi-recombinant U7 machinery. Note that using too much of U7 Sm core-FLASH complex relative to the amount of the HCC present in the extract may result in a significant fraction of the machinery lacking HCC. This form of the U7 machinery is capable of binding histone pre-mRNA substrate but is inactive, inhibiting processing through a dominant negative effect.

2. Incubate 60 min at 32 °C.

   Tip: To assure that the mutant histone pre-mRNA is not processed by the endogenous machinery, set up a control reaction containing nuclear extract but lacking recombinant Mut U7 Sm core-FLASH complex.

3. Process samples and separate RNA by electrophoresis, as described in section 4.2 (steps 3–8).

U7 Sm core assembled from recombinant Sm/Lsm subunits and U7 snRNA interacts with the HCC from nuclear extract and supports accurate processing of histone pre-mRNAs (Bucholc et al., 2020). One important conclusion from these studies is that the ring proteins and U7 snRNA do not require any essential modifications to function in processing. No inhibitory effect on processing activity of semi-recombinant U7 machinery was observed when synthetic U7 snRNA was replaced with T7-generated RNA that contained additional nucleotides at both the 5′ and 3′ ends. Thus, in vitro transcription with bacteriophage RNA polymerases provides a cost-effective alternative to chemical RNA synthesis, allowing for generation of multiple U7 snRNA mutants of choice in a short time. For example, these studies demonstrated that the 3′ terminal stem-loop in U7 snRNA can be deleted without causing any adverse effects on the reaction, and that the spliceosomal type Sm binding site in U7 snRNA supports formation of the U7-specific ring but the resultant semi-recombinant U7 machinery displays reduced accuracy and efficiency of processing (Bucholc et al., 2020). The MBP tag on Lsm10 had no negative effect on processing activity of semi-recombinant U7 machinery and therefore it was left in functional assays. By altering the recombinant components of the ring, studies with semi-recombinant U7 machinery demonstrated that the long unstructured C-terminal extension present in SmB and the large loop separating motifs Sm1 and Sm2 in Lsm11 are dispensable for in vitro processing. With this simple approach, the importance of other regions in the ring proteins and U7 snRNA for 3′-end processing can be readily assessed.

4.4. Single-step purification of semi-recombinant U7 machinery via a photo-cleavable linker in the U7 snRNA

In this method recombinant U7 Sm core is assembled on chemically synthesized U7 snRNA containing biotin and a photo-cleavable linker at the 5′ end, which is then bound to N-terminal segment of FLASH and the complex immobilized on streptavidin beads and thoroughly washed to remove unincorporated subunits. In the next step, the beads are incubated with a nuclear extract to recruit endogenous HCC and the assembled semi-recombinant U7 machinery is eluted from the beads by exposure to long-wavelength UV (Skrajna et al., 2018; Skrajna et al., 2019).

1. Assemble U7 Sm core bound to N-terminal FLASH, as described in section 3.5, using chemically synthesized U7 snRNA containing biotin and photocleavable linker at the 5′ end (commercially available from Horizon Discovery/Dharmacon or Integrated DNA Technologies, Inc.).

   Tip: Synthesis of RNA of this length (60 nucleotides) and containing the two modifications may present a challenge, resulting in a relatively low yield of the full-length product. It is therefore recommended to order a larger scale synthesis and request gel purification to remove shorter products.

2. Add 100 pmol of the U7 Sm core-FLASH complex to 1 ml of processing buffer (75 mM KCl, 15 mM HEPES-KOH pH 7.9, 15% glycerol and 20 mM EDTA pH 8), vortex and incubate on ice 5 min.

3. Spin down potential precipitates in a microcentrifuge (10 min at 10,000 × g)

4. Leaving a small amount at the bottom, transfer the supernatant over 30–40 μl of streptavidin beads that have been washed twice with processing buffer prior to use.

5. Rotate 1 h at 4 °C to immobilize recombinant U7 Sm core-FLASH complex.

6. Gently spin down the beads in a microcentrifuge (3 min at 30 × g) and aspirate the supernatant.

7. Briefly rinse the pellet of beads three times with processing buffer using the same spinning conditions and rotate 1 h with 1 ml of the same buffer.

8. Remove the supernatant and place the tube on ice.

9. Add 250 μl of 80 mM EDTA to 750 μl of a nuclear extract (to final EDTA concentration of 20 mM) and spin down 10 min in cold microcentrifuge at 10,000 × g to remove potential precipitates.

10. Transfer pre-cleared nuclear extract to the tube containing pellet of streptavidin beads and immobilized complex of U7 Sm core-FLASH (see step 8).

11. Rotate 1 h at 4 °C to reconstitute semi-recombinant U7 machinery by binding endogenous HCC.

12. Gently spin down the beads in a microcentrifuge (3 min at 30 × g) and aspirate the supernatant.

13. Apply three brief rinses of the pellet using each time 1 ml of processing buffer, as described above (step 7).

14. Rotate the beads with 1 ml of processing buffer for 1 h at 4 °C

15. Spin as above (3 min at 30 × g), aspirate the supernatant and add 75 μl of processing buffer.

16. Transfer the suspension to a 500 μl tube, place on ice 0.5 inch away from the surface of a lamp emitting high intensity long-wavelength UV.

17. Irradiate 30 min, frequently vortexing and inverting the tube to avoid overheating and ensure equal exposure, as described in detail (Skrajna et al., 2018; Skrajna et al., 2019).

    Tip: For most efficient elution, pre-warm the lamp prior to exposing the samples until the full brightness and intensity of UV irradiation are achieved (typically 5 min).

18. Spin 5 min at 60 × g and collect the supernatant.

19. Re-spin to remove residual beads and collect the supernatant, which contains purified semi-recombinant U7 machinery.

20. Analyze a fraction of the UV eluted material by SDS-PAGE and silver staining to confirm the presence of the four subunits of endogenous HCC: symplekin, CPSF100, CPSF73 and CstF64. If necessary, confirm the identity of visualized proteins by Western blotting and/or mass spectrometry.

21. Test processing activity of the UV-eluted semi-recombinant U7 machinery by mixing together 5 μl of processing mix (see above) containing 0.1 μg/μl yeast tRNA (Invitrogen), 0.5–2.5 μl of the eluted solution, SLBP (1–5 pmol), 0.025 pmol (0.5 ng) of 5′-labeled pre-mRNA substrate. Bring up the volume to 10 μl using processing/tRNA mix.

22. Incubate 60 min at 32 °C, inhibit the reaction by using proteinase K solution and separate radioactive RNA, as described in section 4.2 (steps 3–8).

This is the first example of an in vitro reconstituted processing reaction. This reaction, while using some components purified from nuclear extract, is essentially free of endogenous U7 snRNP so wild-type histone pre-mRNA can be used as a substrate. In addition, with this system it is possible to make various mutations within U7 Sm core, FLASH and SLBP and analyze their effects on the efficiency of processing.

4.5. Assays using reconstituted machinery

One disadvantage of assembling semi-recombinant U7 machinery is that it requires endogenous HCC and therefore it can be isolated only in limited quantities, and even when purified using photo-cleavable RNA, it contains small amount of other polyadenylation factors (CPSF160, WDR33, Fip1 and CPSF30) and nuclear proteins that tend to associate with the HCC and may play unknown role in processing. In addition, semi-recombinant U7 machinery is relatively heterogeneous in size due to partial proteolysis of individual HCC components. The assembly of fully recombinant U7 machinery became possible following successful expression of the four components of HCC (symplekin, CPSF100, CPSF73 and CstF64) as a soluble complex in insect cells (Zhang et al., 2020). Below we describe how to use recombinant components to reconstitute a U7 machinery that accurately cleaves a radio-labeled histone pre-mRNA. All components of the processing reaction are mixed directly prior to the start of the incubation, resulting in rapid assembly of catalytically active U7 machinery and accurate cleavage of the histone pre-mRNA.

1. To 7.5 μl of processing mix containing yeast tRNA at 0.1 μg/μl add the following recombinant components of the processing reaction: 2.5 pmol of U7 Sm core-FLASH (amino acid residues 51–137) complex, 2.5 pmol of the HCC containing either near full-length or N-terminally truncated symplekin, 2.5 pmol SLBP (amino acid residues 125–223) and 0.025 pmol (0.5 ng, 2.5 nM) of 5′-labeled pre-mRNA substrate. Bring up the volume to 10 μl with processing mix containing yeast tRNA at 0.1 μg/μl.

2. Incubate 1 h at 32 °C

3. Process the samples and electrophoretically separate radioactive RNA, as described for 3′-end processing with nuclear extracts (section 4.2, steps 3–8)

Functional studies based on this approach demonstrated that recombinant U7 machinery accurately cleaves histone pre-mRNAs and yielded a number of unexpected findings that explained the molecular mechanism of cleavage by U7 snRNP (Sun et al., 2020b; Yang et al., 2020). With the successful generation of fully recombinant and active U7 machinery, various mutations were incorporated within the HCC components and tested for their effects on processing. Using this approach, CstF64 was shown to be dispensable in vitro, suggesting that its association with the machinery may reflect a role in coupling transcription with processing in vivo. Most surprisingly, functional studies with the reconstituted U7 machinery demonstrated that processing of histone pre-mRNAs critically depends on the N-terminal domain (NTD) of symplekin (amino acids 30–360) (Kennedy et al., 2009; Xiang et al., 2010). This NTD was shown to support processing not only as a part of full-length symplekin and an integral component of the HCC but also as a separate polypeptide added to the processing reaction in trans even though it is unable to form a stable complex with the HCC (Sun et al., 2020b). Among the most important findings, the D75N/H76A mutation within the active site of CPSF73 abolished processing activity, providing ultimate evidence that CPSF73 is the catalytic component of the U7 machinery (Sun et al., 2020b; Yang et al., 2020), supporting previous, more circumstantial conclusions from UV-cross linking studies (Dominski et al., 2005a). The mutant version of U7 machinery was also shown to be unable to degrade the downstream cleavage product in the 5′−3′ direction, supporting the notion that CPSF73 is both an endonuclease and a 5′−3′ exonuclease, at least in the U7 machinery (Yang et al., 2020).

5. Histone pre-mRNA 3′-end processing assays using fluorescently labeled substrate

Fluorescence-based reporter assay is a non-radioactive and rapid technique to monitor RNA processing and degradation. For this assay, a 60-mer modified mouse histone H2a pre-mRNA (H2a*) substrate (Sun et al., 2020b) is labeled with two fluorophores: 5′-end TAMRA and 3′-end FAM (Table 1). Two products are generated by the endonuclease activity of CPSF73 in the U7 machinery, an upstream product with TAMRA label and a downstream product with FAM label. While the upstream product is stable, the downstream product is degraded to mononucleotides, providing direct evidence for the 5′−3′ exonuclease activity of CPSF73. This exonuclease activity can also be demonstrated with an RNA substrate corresponding to the downstream cleavage product, labeled at the 3′ end with FAM (Table 1).

1. Set up a cleavage reaction on ice by mixing 1 μM U7 Sm core-FLASH, SLBP, HCC protein mixture and 0.1 μM labeled substrate in a 10 μl reaction containing 15 mM HEPES (pH 8.0), 75 mM KCl, 15% (v/v) glycerol, 20 mM EDTA (pH 8.0) and RNasin plus ribonuclease inhibitor (1U/reaction, Promega-Fisher).

   Tip: Protease inhibitors are often added to protein purification buffers. Some of the commercial protease inhibitor cocktails contain the metallo-protease inhibitor phosphoramidon, which appears to also inhibit CPSF73 at high concentrations. The chemical structure of this compound bears some similarity to that of activated JTE-607, a known inhibitor of CPSF73 (Kakegawa et al., 2019; Ross et al., 2020). Therefore, care should be taken with the presence of protease inhibitor cocktails in cleavage assays.

2. Incubate at 30 °C for 1 h.

3. Add 10 μl 2× urea denaturing RNA loading buffer (8M urea, 1×TBE) to each reaction and boil the sample for 10 min.

4. Load the boiled samples directly to 15% denaturing urea polyacrylamide gel (8M urea).

5. Run the gel at 280 V for 25 min.

6. Image the gel by the ChemiDoc MP imaging system (Bio-Rad) or other systems. Typical exposure time is 10 s.

The assay clearly demonstrates the accumulation of the upstream cleavage product (Fig. 2). On the other hand, the downstream cleavage product does not accumulate, and is readily degraded to mononucleotides, with apparently no indication of intermediates, suggesting a processive 5′−3′ exonuclease activity for CPSF73 (Yang et al., 2020). A 3′-end FAM-labeled downstream cleavage product RNA can also be degraded. However, the reaction is much slower on ice, and a mutation in the active site of CPSF73 abolishes the activity with both substrates.

Figure 2. Histone pre-mRNA 3′-end processing assays with fluorescently labeled substrates.

The substrate for lanes 1–4 is a 60-mer RNA (H2a*) labeled at the 5′ end with TAMRA (red color) and 3′ end with FAM (green), and that for lanes 5–8 is a 26-mer downstream cleavage product (DCP) labeled at the 3′ end with FAM. Lanes 1 and 5: RNA alone. Lanes 2 and 6: RNA incubated with the reconstituted machinery at 30 °C for 1 h. For the H2a* RNA, the 5′ cleavage product contains 26 nts and the 3′ product contains 34 nts, which is degraded to mononucleotides by the 5′−3′ exonuclease activity of CPSF73. Lanes 3 and 7: RNA incubated with the reconstituted machinery on ice for 1 h. The cleavage activity is much weaker. Lanes 4 and 8: RNA incubated with the reconstituted machinery containing the catalytically inactive D75N/H76A mutant CPSF73 at 30 °C for 1 h. The symplekin N-terminal domain was provided in trans for these two reactions (Sun et al., 2020b).

The fluorescence assay is not as sensitive as the radio-labeled assay, requiring 50–100 nM of fluorescently labeled substrate as compared to 2 nM radio-labeled substrate. However, the two labels on the substrate confer a great advantage by readily allowing both products to be monitored in the assay and providing direct evidence for the processive degradation of the downstream cleavage product.

6. Summary

The successful reconstitution of an active human histone pre-mRNA 3′-end processing machinery has already led to the determination of its structure, as well as some detailed characterizations of the functional roles of its various components. For example, various mutations were incorporated in the HCC components and tested for their effects on processing. Many additional experiments should be possible, for example assessing the functional importance of interaction interfaces among the subunits observed in the structure as well as the functional roles (if any) of many other components that have been identified to be associated with the machinery.