key: cord-0324873-afgghhdw
authors: Gurzeler, Lukas-Adrian; Ziegelmüller, Jana; Mühlemann, Oliver; Karousis, Evangelos D.
title: Production of human translation-competent lysates using dual centrifugation
date: 2021-11-26
journal: bioRxiv
DOI: 10.1101/2021.05.27.445927
sha: 35199103d09b4b765f832eea1d29a860cba2430b
doc_id: 324873
cord_uid: afgghhdw

Protein synthesis is a central process in gene expression and the development of efficient in vitro translation systems has been the focus of scientific efforts for decades. The production of translation-competent lysates originating from human cells or tissues remains challenging, mainly due to the variability of cell lysis conditions. Here we present a robust and fast method based on dual centrifugation that allows for detergent-free cell lysis under controlled mechanical forces. We optimized the lysate preparation to yield cytoplasm-enriched extracts from human cells that efficiently translate mRNAs in a cap-dependent as well as in an IRES-mediated way. Reduction of the phosphorylation state of eIF2α using recombinant GADD34 and 2-aminopurine considerably boosts the protein output, reinforcing the potential of this method to produce recombinant proteins from human lysates.

Cell-free translation is the process of synthesizing proteins in extracts without the use of intact cells. During the last five decades, cell-free extracts have been used to recapitulate key steps of translation and have provided valuable insights into this fundamental process (1) .

Bacterial-based lysates have been extensively used to develop, understand, and expand the applications of in vitro translation (2) . Ranging from the historical experiments that led to the determination of the genetic code (3, 4) to recent advances that allow the design of orthogonal in vitro translation systems (5, 6) , cell-free protein synthesis has revolutionized the possibilities of synthetic biology by employing and expanding the properties of in vitro translation.

Moreover, cell-free systems have allowed the production of numerous recombinant proteins with clear advantages compared to production in living cells. The possibility of genetic manipulation, the incorporation of unnatural amino acids and the tolerance to toxic proteins in cell-free extracts have extended the limits of synthetic biology significantly (2) .

Even though the production and use of human lysates is of particular interest, it has not been equally developed. Human cell-derived lysates provide a better environment for the proper folding of human proteins as they facilitate the incorporation of human-specific posttranslational modifications. Moreover, they are an appealing system to study translationrelated processes in small scale or high throughput settings. However, several key constraints like the cost and technical difficulties in the reproducible production of eukaryotic translation systems compared to prokaryotic have limited their industry and research usage (7) .

The relatively low and variable protein synthesis output are the most important reasons that impede the routine application of human in vitro translation systems (1) . To date, the only commercially available mammalian in vitro translation system that allows for cap-dependent translation is the rabbit reticulocyte lysate (RRL) (8) . In contrast, the currently commercially available human in vitro translation systems depend on Internal Ribosome Entry Site (IRES)mediated initiation and perform poorly on normal 7 mG-capped mRNAs (Thermo Fisher Scientific, Cat No. 88890). The RRL system has proven to be a valuable source to study several aspects of mammalian translation. However, it is accompanied by unsurpassed caveats due to reticulocytes being special cells regarding their translation activity and regulation. The requirement of complex manipulation of animal tissue, the increased concentration of endogenous globin mRNAs, and the abundance of RNase M and other tissue-specific nucleases (9) are substantial limitations of this system. Moreover, RRL cannot be genetically modified to deplete or enrich certain protein factors (10) . Previous efforts have advanced the use of mammalian cells to produce translation-competent lysates. Chinese hamster ovary cells (CHO) have been used to develop an in vitro translation system that allows for IRES-dependent translation of reporter constructs and has been shown to produce functional glycosylated proteins of interest (11) . Lysates of human origin that support cap-dependent and independent in vitro translation of reporter mRNAs originate from HeLa cells (12) (13) (14) or from HEK-293 cells for the study of native mRNP complexes in vitro (15) . However, issues concerning the variability of lysis parameters, such as the applied mechanical force, have not been satisfactorily solved.

In our previous efforts to address mechanistic questions about translation termination, we developed an in vitro translation system based on HeLa cells that was used to monitor ribosomal density at the termination codon of reporter mRNAs (16) . This protocol was optimized to prepare lysates using the shearing forces of syringe treatment under hypotonic conditions to break up the cells and release functionally active molecules that can reconstitute the complex translation process in vitro (16) , based on a previously reported protocol (14) .

During this development, we observed that cell lysis is the most crucial step for the success of our in vitro system. To further improve the efficiency and the reproducibility of this step, we applied dual centrifugation (DC) as a novel approach to lyse cells. Dual centrifugation differs from regular centrifugation by combining the simultaneous rotation around two axes. As a consequence, the centrifugal acceleration constantly changes, resulting in powerful agitation of the sample material inside the vials (Fig. 1A) . Thus, in contrast to conventional centrifugation that separates components, DC leads to efficient sample homogenization (17, 18) . Adapting this technique as a cell lysis method, we streamlined the production of translation-competent lysates in various batch sizes in a reproducible manner. The resulting lysates allowed us to dissect the role of Nsp1, a potent virulent factor produced during the early steps of infection by SARS-CoV-2 in human cells (19) .

Here we document that this new, improved protocol using dual centrifugation enables the efficient and reproducible production of translation-competent cytoplasmic cell extracts from HeLa S3 cells in large amounts. We show that dual centrifugation can be used for routine cell lysis under different buffer compositions, and we investigate the conditions that give rise to lysates that translate reporter mRNAs in a cap-and IRES-dependent manner. We perform biochemical characterization of the lysates, provide technical details for their production, and show that by modifying the phosphorylation state of eIF2α we can increase protein synthesis considerably.

HeLa S3 cells (Sigma-Aldrich, ECACC, Cat. No. 87110901) were cultured in DMEM/F-12 (Gibco, Cat. No. 32500-043) containing 10% FCS, 100 U/ml penicillin and 100 μg/ml streptomycin (termed DMEM+/+) at 37 °C and 5% CO2 in humid atmosphere. To start a new culture from a cryopreserved stock (stored in 50% FCS, 40% DMEM +/+ and 10% DMSO), 1-2x10 6 cells were seeded into a 75 cm 2 flask (TPP, Cat. No. 90026). After reaching a confluency of approximately 80% as adherent cells, they were trypsinized and transferred into a suspension culture. The suspension cultures were then grown in volumes between 20 ml and 500 ml under constant agitation (Corning flasks, Cat. No. 431405, 431147 and 431255) within a range of 0.2x10 6 -1x10 6 cells/ml.

Lysates were prepared from non-synchronized HeLa S3 cell cultures grown to a cell density ranging from 0.8x10 6 to 1x10 6 No. EVS-050). The lysate was centrifuged at 13,000 g, 4 °C for 10 min and the supernatant (the lysate) was aliquoted, snap-frozen and stored at -80 °C. The lysate could be thawed and refrozen multiple times for subsequent applications with only a minor loss of translation efficiency.

The DNA template sequences of the different reporters used in this study are depicted in sup. Fig. 3 .The p200-6xMS2 plasmid encoding the RLuc reporter used for in vitro transcription is described in (19) . We created the dual-luciferase reporter pCRIIRLuc-UAA-FLucA80 by ligating NotI-KpnI digested pCRII-hRLuc-200bp-A80 (16) and the PCR-amplified Luc ORFs from the plasmid p2luc-UAA (20) (digested with the same restriction enzymes). The ECMV IRES sequence was subcloned into pCRII-RLuc-UAA-FLuc-A80 from the pT7CFE1-CHis plasmid (Thermo Fisher Scientific, Cat. No. 88860) by In-Fusion cloning (Takara, Cat. No. 639650) to create pCRII-RLuc-IRES-FLuc-A80. A second HindIII restriction site was deleted by site-directed mutagenesis because this enzyme is used for linearization before run-off in vitro transcription. To clone pCRII_SC-LD_SMG6-3xFLAG_A80, the SMG6 coding sequence (CDS) along with a C-terminal 3xFLAG was cloned into pCRII-FL-hRLuc described in (19) by removing the hRLuc sequence and only leaving the SARS-CoV-2 Leader sequence (SC-LD) of the SARS-CoV-2 5΄UTR using InFusion cloning. To obtain pCRII_SC-LD_3xFLAG-HBB-CTE_A80, the beta globin CDS (HBB) with a 70-nucleotide long C-terminal extension (CTE) was cloned into the pCRII_SC-LD_SMG6-3xFLAG_A80, by replacing the SMG6 coding sequence and moving the 3xFLAG tag to the N-terminus using InFusion cloning.

Plasmids used for in vitro transcription were linearized using a restriction site located 

To monitor the protein synthesis output by luciferase assay, a standard 12. Dual-luciferase assays were performed using 40 μl of the substrate in the appropriate buffer for a 10 μl in vitro translation assay (Promega, Cat. No. E1960). Luminescence was measured using TECAN Infinite M1000 and plotted on GraphPad (Version 8.4.2).

For the preparation of control cytoplasmic and nuclear lysates, 1x10 7 HeLa S3 cells were harvested and lysed by resuspension in 1 ml buffer 1 (50 mM HEPES pH 7.3, 140 mM NaCl, 1 mM EDTA, 10% Glycerol, 0.5% NP-40, 0.25% Triton X-100, 1x protease inhibitor cocktail (Bimake, Cat. No. B14002)) and incubation for 10 min at 4 °C rotating. Subsequently, the samples were centrifuged at 500 g, 4 °C for 5 min and the supernatant was collected. The same steps were repeated twice with the pellets, and the supernatants were pooled to obtain the cytoplasmic fraction. The pellets were washed by resuspension in 1 ml buffer 2 (10 mM Tris pH 8.0, 200 mM NaCl, 1 mM EDTA, 0.5 mM EGTA, 1x protease inhibitor cocktail (Bimake, Cat. No. B14002)) and incubation for 10 min at 4 °C rotating. The nuclei were pelleted by centrifugation at 500 g, 4 °C for 5 min, resuspended in 500 µl buffer 3 (1 mM EDTA, 0.5 mM EGTA, 10 mM Tris pH 8.0, 10% SDS, 0.1% deoxycholate, 1% Triton X-100, 1x protease inhibitor cocktail (Bimake, Cat. No. B14002)) and sonicated with 3x 10 s pulses at 45% amplitude (Vibra-Cell, Cat. No. 75186). Finally, samples were centrifuged at 16,000 g, 4 °C for 30 min, and the supernatant was collected to obtain the nuclear fraction.

To assess whether our HeLa S3 lysate is enriched for cytoplasmic factors, samples were collected before lysis by DC (whole cells in hypotonic lysis buffer) and after the final centrifugation step (lysate and pellet). The whole-cell and pellet samples were diluted in three volumes of hypotonic lysis buffer and subjected to sonication for 3x 10 s pulses at 45% amplitude (Vibra-Cell, Cat. No. 75186). Proteins from DC-derived material and the samples deriving from the subcellular fractionation were extracted by chloroform-methanol precipitation. From each sample, 6.8 μg protein (quantification by OD280) were loaded.

To compare the phosphorylation status of eIF2α, HeLa S3 cells were harvested from a culture with a density of 0.8x10 6 to 1x10 6 Densitometry was performed using Image Studio Lite Ver 5.2. The average from 3 biological replicates that were run three times (6.67x10 5 cell equivalents per sample) was calculated by extrapolating the density values of the RLuc standards from the corresponding standard curve that was prepared in Microsoft Excel.

The cDNA encoding the GADD34 protein with a deletion of the amino acids 1-240 was cloned into the pRSET A vector (Thermo Fisher Scientific, Cat. No. V35120) by In-Fusion cloning 

The efficient lysis of eukaryotic cells under well-defined biochemical conditions is a prerequisite for a wide range of biological assays. The lysis process is often challenging, especially when the isolation of active biomolecules is required for downstream applications.

While many different protocols have been developed to produce biologically active lysates that can be used for in vitro studies, they are often limited in accurately defining the mechanical force applied during lysis. To apply reproducible shearing forces, we opted for dual centrifugation (DC). In this process, the samples undergo a simultaneous rotation around two axes of symmetry (Fig. 1A) . The continuous change of centrifugal forces in the sample vials leads to shearing and homogenization of the processed material (17, 18) . We rationalized that such a mechanical treatment would allow the lysis of cells independently of the buffer conditions. Importantly, this method accurately defines and fine-tunes the lysis parameters by controlling the centrifugation speed while keeping the processed material at a low temperature.

Based on our previously established in vitro translation assays (16) 

Dual centrifugation has not been previously used to disrupt living cells and isolate biologically active lysates. We set out to assess whether cell lysis can, in principle, be performed using this method. We established our system using HeLa S3 cells (spinner cells), which are adapted to grow in suspension culture and reach high cell densities. The cells were harvested at a density of 0.8 -1.0x10 6 cells/ml in suspension, resuspended in a hypotonic lysis buffer to reach a final concentration of 2x10 8 cells/ml and subjected to DC. To allow a comparative analysis, the dual centrifugation experiments were performed for 4 min and the temperature of the rotor was set at -5 °C, taking into consideration a possible increase of the sample temperature during DC. In all experiments, it was confirmed that the samples did not freeze during this step.

First, we assessed under which centrifugation speed cell lysis can be achieved, ideally without disrupting the cell nuclei. To visualize the lysis efficiency, we performed a Trypan Blue staining that assesses cell membrane integrity; living cells are visible as white spheres, whereas dead cells and nuclei as dark blue foci. As shown in Fig. 2A , the lysis efficiency increased proportionally to the applied centrifugal force. Lysis was achieved even under isotonic buffer conditions (in PBS, data not shown), rendering dual centrifugation an appealing method for lysing cells in virtually any buffer conditions without the addition of detergents. This is of particular interest in cases where preserving intact, functional biomolecules is crucial.

Since we wanted to obtain translation-competent lysates, we opted for a mainly cytoplasmic extract and applied dual centrifugation forces in the range of 250 -1000 RPM to titrate the translation efficiency of the resulting lysates.

After DC, we removed residual intact cells, nuclei and insoluble fragments from the lysate by regular centrifugation. To assess whether our lysates are enriched for cytoplasmic components, we assessed cytoplasmic and nuclear markers by immunostaining. We compared our lysates (supernatant after DC using 500 RPM and regular centrifugation, as shown in Fig. 1B) with total cell lysates and cytoplasmic and nuclear lysates obtained using a detergent-based protocol for subcellular fractionation. As shown in Fig. 2B , the supernatant is These results show that dual centrifugation allows for the production of lysates that are highly enriched for cytoplasmic components. Further optimization could allow for the application of this method to separate cytoplasmic from nuclear fractions. 

To assess whether the DC-derived lysates can translate reporter mRNAs, we used an in vitro transcribed and capped Renilla luciferase reporter that contains an 80-nucleotide long poly(A) tail (RLuc reporter). To this end, we set up in vitro translation reactions for lysates derived from DC at 250, 500, 750, or 1000 RPM. The processed lysates (i.e., the supernatant after the final, regular centrifugation) were supplemented with a buffer containing necessary salts, energy source and amino acids needed for in vitro translation and with an equal amount of mRNA for the four reactions. After a 50 min incubation at 33 °C, we measured the enzymatic activity of RLuc, which is proportional to the protein production in the lysates. As shown in Fig. 3A , the luminescence measurements verify the translation capacity of all the tested lysates, with an optimal activity of lysates that were subjected to DC at 500 RPM and underwent, as shown in Fig. 2A , a relatively mild lysis process. No translation activity was detected in the presence of the translation inhibitor puromycin. The fact that milder lysis conditions lead to a higher translation output may be attributed to the optimal balance between disruption of sensitive components due to the mechanical forces and the efficiency of releasing the cytoplasm.

Additionally, the prevention of lysis of nuclei may also contribute to the higher translation efficiency of the lysates.

To test our DC-derived lysates, we applied the in vitro translation conditions as they were previously optimized for our syringe-treated lysates (16) . Among the most crucial factors for efficient translation is the Mg 2+ concentration. Thus, we decided to titrate this parameter to ensure that the concentration is optimal under the new lysis conditions. Setting as reference a concentration of 0.30 mM, we assessed the efficiency of translation under different MgCl2 concentrations in the translation buffer, and we observed that the concentration of choice falls in the optimal range of translation activity (Fig. 3B) . For this reason, we selected a concentration of 0.30 mM MgCl2 for our in vitro experiments. Notably, the range of suitable concentration was wider than for our previous translation system, indicating that lysates prepared by DC are more robust towards small changes of in vitro translation conditions. Next, we titrated the effect of the reporter mRNA concentration on the translation output of the in vitro translation reaction. We added increasing amounts of reporter mRNA, ranging from 0 to 80 fmol/μl of reaction, and as shown in Fig. 3C , we observed a concentrationdependent increase of the reporter protein output. This increase is linear for concentrations up to 40 fmol/μl (Fig. 3C) . Based on this result, we routinely use 5 or 10 fmol/μl for standard translation reactions since these concentrations clearly correspond to the linear range and therefore do not saturate the translation system with mRNA. The concentration of mRNA used can be adjusted depending on the downstream applications and depends on the translation efficiency of the individual mRNA.

To assess how the duration of the incubation affects the production of the recombinant protein and, therefore, the translation efficiency of the system, we performed a time course from 0 to 150 min (Fig. 3D ) and 0 to 4 h (Fig. 3E) . We observed an almost linear increase of the luminescence signal during the entire 150 min incubation. When we prolonged the assay duration, the reporter protein production reached a plateau after 2 h of translation.

Interestingly, there is a non-linear increase of the signal during the first 30 min of incubation, in accordance with the observations that the initial association of translation factors with an mRNA is a rate-limiting step in translation (21) . 

Cap-dependent translation is the primary mechanism of translation initiation for mammalian mRNAs. For this reason, we decided to address the effect of a 5΄ cap on the translation efficiency of mRNA reporters in the in vitro translation system by comparing the RLuc output of in vitro capped and uncapped Renilla luciferase reporter mRNAs. We used equimolar amounts of the two mRNA molecules and as shown in Fig. 4A , we observed a >30-fold increase in the efficiency of translation in the presence of a 5΄ cap on the reporter mRNAs.

This result shows that our system is sensitive to the presence of a cap, in accordance with the cap-dependent translation initiation of most physiological mRNAs in mammalian cells.

After optimizing the translation efficiency of cap-dependent translation, we addressed whether DC-derived lysates can perform IRES-dependent translation, a cap-independent process (22) . IRES usage is of interest because many vectors have been developed for in vitro protein production employing this principle. To address whether we can perform IRESdependent translation in our lysates, we prepared a dual-luciferase reporter mRNA encoding Renilla and Firefly Luciferase. We used two versions of this construct, one where the two ORFs are separated by an IRES originating from the human EMCV virus and a control version that lacks the IRES sequence (23) . The detection of a clear Firefly luciferase signal only in the presence of the IRES demonstrates the capacity of the lysate to perform IRES-dependent translation (Fig. 4B) .

We compared the efficiency of our system to a commercially available rabbit reticulocyte lysate system (RRL). The in vitro translation reaction in the RRL was performed according to the manufacturer's guidelines with an equal amount of RLuc reporter. As shown in Fig. 4C , our lysate is about five-fold more translationally active than the RRL. Thus, our system allows for the reproducible production of translation-competent lysates with high translation potential. 

The lysate described above was prepared using a hypotonic lysis buffer complemented with the necessary salts, amino acids and energy regeneration system before in vitro translation.

We decided to assess whether lysates can also be prepared in a buffer that already contains all the necessary components. When we lysed the cells directly in such a translationcompetent buffer, we obtained translationally active lysates with a similar activity compared to the previous method of lysate preparation (Fig. 5A) . Moreover, we observed that the translation is enhanced if it is not preceded by a 5 min pre-incubation at 33 °C before adding the reporter mRNA under these conditions (Sup. Fig. 1 ).

Additionally, we assessed whether it is possible to generate active lysates using pellets from frozen cells. We compared lysates from the same amount of frozen or fresh cells prepared in a translation-competent buffer. The lysates deriving from frozen cells were translationally active, although with reduced activity compared to unfrozen cells (Fig. 5B) .

The cell lysis in a translation-competent buffer further facilitates the production of translation-competent lysate and improves the reproducibility of the in vitro translation assay by reducing intermediate steps. Such an approach is ideal for structural studies that require large amounts of translationally active lysates. The fact that we can obtain translationally active lysates from frozen cells allows us to decouple cell culturing and harvesting from the lysate production, which further expands the range of possible applications. Under these conditions, we also assessed whether it is possible to translate reporter mRNAs with different 5΄UTRs and ORFs encoding proteins with different molecular masses. We cloned the ORF of human beta globin (HBB) and SMG6, an endonuclease that is implicated in nonsensemediated mRNA decay (24) (25) (26) , and used the corresponding in vitro transcribed capped and polyadenylated mRNAs, which harbor the SARS-CoV-2 leader sequence in the 5´UTR, as templates for in vitro translation. Similarly to RLuc (36 kDa, Fig. 5C ), we observed single distinct bands for 3xFlag-HBB (21 kDa, Fig. 5D) and SMG6-3xFlag (163 kDa, Fig. 5D ).

Comparison of lysates produced using a hypotonic lysis buffer complemented with components required for in vitro translation (hypotonic lysis) versus lysis in a translation-competent buffer (translation comp lysis). B. Comparison of lysates produced using a hypotonic lysis buffer complemented with components required for in vitro translation (hypotonic lysis) from freshly harvested cells versus from frozen cell pellets. For A and B, RLuc activity measurements are depicted as arbitrary luminescence units (AU). Each dot depicts the value of an individual experiment and in vitro translation (biological replicate) that was measured three times (technical replicates). Mean values and SD are shown. C-E. Western blot analysis of time-course of in vitro translations of different mRNA reporters using lysates produced in translation-competent buffer. For C and D 1.6 μl and for E 1.25 μl of the translation reactions were loaded on the gel. In vitro translation experiments were performed using 10 fmol/μl RLuc reporter, a final amino acids concentration of 0.01 mM and without preincubating the translation reactions before adding the reporter mRNAs.

To use our system for the production of recombinant proteins, maximizing the translation efficiency is crucial. The eukaryotic initiation factor 2 (eIF2) is a key regulator of global translation initiation (21) and the dephosphorylation of its α subunit (eIF2α) has been shown to increase the translation efficiency of HeLa lysates (27) . We assessed and modulated the phosphorylation status of eIF2α in our lysates to further increase the protein production capacity. eIF2 is the major Met-tRNAi carrier when it is bound to GTP, however, eIF2 is more stable when it is bound to GDP (28) . For this reason, the activation of eIF2 requires the activity of a guanine exchange factor, eIF2B, which promotes GDP dissociation and facilitates GTP and Met-tRNAi to form a ternary complex with eIF2. eIF2 is subject to tight control as it can be phosphorylated by several protein kinases at Ser 51 of its α subunit (eIF2α). Once phosphorylated, eIF2α does not allow eIF2B to exchange GDP to GTP, and protein synthesis initiation is inhibited (21) .

To assess the phosphorylation status of eIF2α in our cells and lysate, we immunoblotted total and Ser51-phosphorylated eIF2α. We observed that eIF2α was considerably phosphorylated in the lysates compared to intact cells before lysis (Fig. 6A) , probably due to the activity of kinases during and/or after lysis. To modulate the phosphorylation status of eIF2α in the lysates during in vitro translation, we used GADD34 and 2-aminopurine. GADD34 forms a complex with the catalytic subunit of protein phosphatase 1 (PP1) and promotes the dephosphorylation of eIF2α in vivo (29) . It was shown that treatment with a 240 amino acids N-terminal truncation of recombinant GADD34 (GADD34 Δ240) can reduce the phosphorylation state of eIF2α and increase translation efficiency of HeLa lysates (27) . 2-aminopurine (2-AP) is an inhibitor of the protein kinase R (PKR) and other serine/threonine protein kinases and therefore represses the phosphorylation of eIF2α at Ser51 (30, 31) .

To reduce the phosphorylation of eIF2α in our lysates, we expressed and purified recombinant GADD34 Δ240 (Sup. Fig. 2A) , titrated 2-AP in the presence of the optimal GADD34 Δ240 concentration, and assessed the effect on translation using RLuc reporter assays (Sup. Fig. 2B ). The addition of 2-AP alone led to a decrease in eIF2α phosphorylation but did not affect the translation output. The addition of GADD34 Δ240 led to a distinct decrease in eIF2α phosphorylation and a two-fold increase in the translation output ( Figure   6B and C). Interestingly, the addition of both 2-AP and GADD34 Δ240 (3 mM and 0.5 mM, respectively) to the translation reactions led to almost complete dephosphorylation of eIF2α (Fig. 6B) and a 6-fold increase of protein synthesis compared to the untreated lysates (Fig.   6C ).

We quantified the protein output of RLuc under these conditions by comparing the immunoblot signal of the produced protein to standard dilutions of recombinant RLuc and estimated the production of approximately 1 ng of recombinant protein per μl of lysate after 50 min οf translation (Sup. Fig. 2C ). This output can be further increased by prolonging the translation reaction, increasing the concentration of the reporter mRNA, and using more concentrated lysates of higher cell equivalent density (data not shown). In other words, the translation-competent lysates can be used to produce potentially toxic proteins in vitro, yielding more than 1 μg protein per milliliter of translation reaction.

In conclusion, we provide a new protocol for producing translation-competent lysates from human cells in a fast, scalable, and highly reproducible manner. Our system is highly active and enables cap-dependent translation with unprecedented protein yields, which not only facilitates fundamental research on translation regulation but also paves the way for the future development of large-scale in vitro protein synthesis in a human system. 

Sup. Fig. 1 : Comparison of lysates produced using a hypotonic lysis buffer complemented with components required for in vitro translation (hypotonic) versus lysis in a translationcompetent buffer (translation comp lysis) with or without a 5΄ pre-incubation at 33 °C before in vitro translation. RLuc activity measurements are depicted as normalized values of luminescence to the hypotonic buffer condition with pre-incubation. Each dot depicts the value of an individual experiment and in vitro translation (biological replicate) that was measured three times (technical replicates). In vitro translation experiments were performed using 10 fmol/μl RLuc reporter. Mean values and SD are shown. Sup. Fig. 3 Nucleotide sequences of the DNA templates used for in vitro transcription followed by the colour code for every sequence. Start and stop codons are in bolt and underlined. (from pCRII-RLuc-IRES-FLuc-A80)   TAATACGACTCACTATAGGGCGAATTGGGCCCTCTAGATGCATGCTCGAGCGGCCGCATGACTTCGAAAGTTTATGATCCAGA  ACAAAGGAAACGGATGATAACTGGTCCGCAGTGGTGGGCCAGATGTAAACAAATGAATGTTCTTGATTCATTTATTAATTATT  ATGATTCAGAAAAACATGCAGAAAATGCTGTTATTTTTTTACATGGTAACGCGGCCTCTTCTTATTTATGGCGACATGTTGTG  CCACATATTGAGCCAGTAGCGCGGTGTATTATACCAGACCTTATTGGTATGGGCAAATCAGGCAAATCTGGTAATGGTTCTTA  TAGGTTACTTGATCATTACAAATATCTTACTGCATGGTTTGAACTTCTTAATTTACCAAAGAAGATCATTTTTGTCGGCCATG  ATTGGGGTGCTTGTTTGGCATTTCATTATAGCTATGAGCATCAAGATAAGATCAAAGCAATAGTTCACGCTGAAAGTGTAGTA  GATGTGATTGAATCATGGGATGAATGGCCTGATATTGAAGAAGATATTGCGTTGATCAAATCTGAAGAAGGAGAAAAAATGGT  TTTGGAGAATAACTTCTTCGTGGAAACCATGTTGCCATCAAAAATCATGAGAAAGTTAGAACCAGAAGAATTTGCAGCATATC  TTGAACCATTCAAAGAGAAAGGTGAAGTTCGTCGTCCAACATTATCATGGCCTCGTGAAATCCCGTTAGTAAAAGGTGGTAAA  CCTGACGTTGTACAAATTGTTAGGAATTATAATGCTTATCTACGTGCAAGTGATGATTTACCAAAAATGTTTATTGAATCGGA  CCCAGGATTCTTTTCCAATGCTATTGTTGAAGGTGCCAAGAAGTTTCCTAATACTGAATTTGTCAAAGTAAAAGGTCTTCATT  TTTCGCAAGAAGATGCACCTGATGAAATGGGAAAATATATCAAATCGTTCGTTGAGCGAGTTCTCAAAAATGAACAAATGTCG  ACGGTCTCCCTCCACTGCTGTAGTAACCCGGGTCCGGGGCCTCGGTGGTGCTAAGCGAATTAATTCCGGTTATTTTCCACCAT  ATTGCCGTCTTTTGGCAATGTGAGGGCCCGGAAACCTGGCCCTGTCTTCTTGACGAGCATTCCTAGGGGTCTTTCCCCTCTCG  CCAAAGGAATGCAAGGTCTGTTGAATGTCGTGAAGGAAGCAGTTCCTCTGGAAGCAACTTGAAGACAAACAACGTCTGTAGCG  ACCCTTTGCAGGCAGCGGAACCCCCCACCTGGCGACAGGTGCCTCTGCGGCCAAAAGCCACGTGTATAAGATACACCTGCAAA  GGCGGCACAACCCCAGTGCCACGTTGTGAGTTGGATAGTTGTGGAAAGAGTCAAATGGCTCACCTCAAGCGTATTCAACAAGG  GGCTGAAGGATGCCCAGAAGGTACCCCATTGTATGGGATCTGATCTGGGGCCTCGGTGCACATGCTTTACATGTGTTTAGTCG  AGGTTAAAAAACGTCTAGGCCCCCCGAACCACGGGGACGTGGTTTTCCTTTGAAAAACACGATGATAATATGGAAGACGCCAA  AAACATAAAGAAAGGCCCGGCGCCATTCTATCCTCTAGAGGATGGAACCGCTGGAGAGCAACTGCATAAGGCTATGAAGAGAT  ACGCCCTGGTTCCTGGAACAATTGCTTTTACAGATGCACATATCGAGGTGAACATCACGTACGCGGAATACTTCGAAATGTCC  GTTCGGTTGGCAGAAGCTATGAAACGATATGGGCTGAATACAAATCACAGAATCGTCGTATGCAGTGAAAACTCTCTTCAATT  CTTTATGCCGGTGTTGGGCGCGTTATTTATCGGAGTTGCAGTTGCGCCCGCGAACGACATTTATAATGAACGTGAATTGCTCA  ACAGTATGAACATTTCGCAGCCTACCGTAGTGTTTGTTTCCAAAAAGGGGTTGCAAAAAATTTTGAACGTGCAAAAAAAATTA  CCAATAATCCAGAAAATTATTATCATGGATTCTAAAACGGATTACCAGGGATTTCAGTCGATGTACACGTTCGTCACATCTCA  TCTACCTCCCGGTTTTAATGAATACGATTTTGTACCAGAGTCCTTTGATCGTGACAAAACAATTGCACTGATAATGAATTCCT  CTGGATCTACTGGGTTACCTAAGGGTGTGGCCCTTCCGCATAGAACTGCCTGCGTCAGATTCTCGCATGCCAGAGATCCTATT  TTTGGCAATCAAATCATTCCGGATACTGCGATTTTAAGTGTTGTTCCATTCCATCACGGTTTTGGAATGTTTACTACACTCGG  ATATTTGATATGTGGATTTCGAGTCGTCTTAATGTATAGATTTGAAGAAGAGCTGTTTTTACGATCCCTTCAGGATTACAAAA  TTCAAAGTGCGTTGCTAGTACCAACCCTATTTTCATTCTTCGCCAAAAGCACTCTGATTGACAAATACGATTTATCTAATTTA  CACGAAATTGCTTCTGGGGGCGCACCTCTTTCGAAAGAAGTCGGGGAAGCGGTTGCAAAACGCTTCCATCTTCCAGGGATACG  ACAAGGATATGGGCTCACTGAGACTACATCAGCTATTCTGATTACACCCGAGGGGGATGATAAACCGGGCGCGGTCGGTAAAG  TTGTTCCATTTTTTGAAGCGAAGGTTGTGGATCTGGATACCGGGAAAACGCTGGGCGTTAATCAGAGAGGCGAATTATGTGTC  AGAGGACCTATGATTATGTCCGGTTATGTAAACAATCCGGAAGCGACCAACGCCTTGATTGACAAGGATGGATGGCTACATTC  TGGAGACATAGCTTACTGGGACGAAGACGAACACTTCTTCATAGTTGACCGCTTGAAGTCTTTAATTAAATACAAAGGATATC  AGGTGGCCCCCGCTGAATTGGAATCGATATTGTTACAACACCCCAACATCTTCGACGCGGGCGTGGCAGGTCTTCCCGACGAT  GACGCCGGTGAACTTCCCGCCGCCGTTGTTGTTTTGGAGCACGGAAAGACGATGACGGAAAAAGAGATCGTGGATTACGTCGC  CAGTCAAGTAACAACCGCGAAAAAGTTGCGCGGAGGAGTTGTGTTTGTGGACGAAGTACCGAAAGGTCTTACCGGAAAACTCG  ACGCAAGAAAAATCAGAGAGATCCTCATAAAGGCCAAGAAGGGCGGAAAGTCCAAATTGTAAGGTACCAAAAAAAAAAAAAAA SC-LD_SMG6-3xFlag_A80)   TAATACGACTCACTATAGATTAAAGGTTTATACCTTCCCAGGTAACAAACCAACCAACTTTCGATCTCTTGTAGATCTGTTCTCT  AAACGATGGCGGAAGGGCTGGAGCGTGTGCGGATCTCCGCGTCGGAGCTGCGCGGGATCCTGGCTACTCTGGCCCCGCAGGCCGG  GAGCAGAGAAAACATGAAGGAATTAAAGGAGGCCAGGCCGCGCAAAGATAACAGGCGTCCAGATCTGGAAATCTATAAGCCTGGC  CTTTCTCGGCTAAGGAACAAGCCCAAAATCAAGGAACCCCCTGGGAGTGAGGAATTCAAAGATGAAATTGTTAATGACCGAGATT  GCTCTGCTGTTGAAAATGGTACACAGCCCGTTAAAGATGTCTGCAAGGAACTGAACAACCAAGAGCAGAATGGTCCTATAGACCC  AGAAAATAATCGGGGACAAGAATCCTTTCCTAGGACTGCTGGACAAGAGGATCGTAGTCTAAAAATTATCAAAAGAACAAAGAAA  CCCGACCTGCAGATCTATCAGCCTGGACGACGTTTGCAGACTGTTAGCAAAGAATCCGCCAGTCGGGTGGAGGAGGAAGAAGTCC  TCAACCAGGTAGAACAACTGAGAGTAGAGGAAGATGAGTGTAGGGGAAATGTTGCGAAGGAGGAAGTTGCGAATAAACCAGACAG  GGCCGAGATAGAAAAGAGCCCAGGTGGTGGGAGAGTAGGGGCTGCAAAAGGAGAAAAAGGAAAGAGGATGGGAAAAGGGGAGGGG  GTGAGGGAAACCCACGACGACCCGGCCCGCGGGAGGCCGGGCTCCGCAAAGCGCTACTCCCGCTCAGACAAACGAAGGAATCGCT  ACCGCACGCGCAGCACCAGCTCAGCTGGCAGCAACAACAGCGCTGAGGGAGCTGGCCTGACGGATAATGGATGTCGCCGCCGCCG  ACAGGATAGGACCAAGGAGAGGCCACCACTGAAGAAGCAAGTGTCTGTGTCCTCAACCGATTCCTTAGACGAGGACAGAATTGAT  GAGCCTGATGGATTAGGACCCAGGAGAAGTTCAGAAAGGAAGAGACATTTAGAAAGAAACTGGTCTGGCCGTGGGGAGGGTGAGC  AGAAAACCAGTGCTAAAGAATATCGAGGCACTCTTCGTGTCACTTTCGATGCAGAAGCCATGAACAAAGAGTCTCCCATGGTGAG  GTCAGCCAGGGATGATATGGATAGAGGAAAGCCTGACAAAGGCTTGAGCAGTGGGGGCAAAGGCTCTGAGAAGCAGGAGTCCAAA  AACCCGAAACAAGAACTTCGGGGTCGTGGTCGTGGCATTCTGATTTTGCCTGCCCATACCACCCTATCTGTCAATTCAGCAGGTT  CTCCAGAGTCCGCGCCTTTGGGACCTCGGCTTTTGTTTGGATCTGGTAGTAAGGGATCTCGGAGTTGGGGCCGTGGAGGCACCAC  ACGCCGATTGTGGGACCCAAACAATCCTGATCAGAAACCTGCTCTAAAGACTCAGACGCCCCAGCTACATTTCTTGGACACTGAT  GATGAAGTCAGCCCTACATCTTGGGGTGACTCACGCCAGGCTCAGGCATCTTACTATAAGTTTCAAAACTCTGACAACCCCTATT  ATTACCCCCGGACACCAGGCCCTGCCTCCCAGTATCCCTATACGGGCTATAACCCTCTACAGTACCCAGTGGGCCCTACGAATGG  TGTGTACCCAGGGCCTTACTACCCAGGCTACCCGACTCCGTCAGGACAGTATGTGTGTAGCCCTCTACCTACCAGCACCATGAGT  CCCGAGGAGGTAGAGCAGCACATGAGGAACCTGCAGCAACAGGAGCTGCACAGGCTTCTCCGGGTGGCTGACAACCAGGAACTGC  AGCTCAGCAACCTGCTCTCCAGGGACCGCATCAGTCCGGAGGGCCTGGAGAAGATGGCGCAACTCAGAGCTGAACTGCTGCAGCT  ATATGAGCGCTGTATTCTATTAGATATTGAGTTCTCTGATAATCAGAATGTGGATCAGATCCTGTGGAAGAATGCTTTCTATCAG  GTGATTGAGAAGTTCAGGCAACTTGTCAAGGATCCGAATGTTGAGAACCCAGAACAGATTCGGAACAGACTTTTGGAGCTCTTGG  ATGAGGGTAGTGACTTCTTTGATAGTTTGCTTCAGAAGCTCCAAGTTACTTATAAGTTCAAACTGGAAGACTACATGGATGGTCT  TGCCATTCGCAGCAAGCCATTACGCAAGACAGTAAAATATGCCTTGATCAGTGCCCAGCGATGCATGATATGCCAAGGAGATATT  GCTAGGTACCGGGAGCAAGCCAGTGATACAGCGAATTATGGGAAAGCACGCAGTTGGTACCTGAAGGCCCAGCACATTGCTCCCA  AGAATGGGCGCCCCTATAACCAGTTGGCTTTGCTGGCAGTGTATACGAGGAGGAAGCTTGACGCTGTCTATTACTATATGCGCAG  TTTAGCTGCCAGCAACCCTATCCTGACTGCCAAGGAGAGTCTCATGAGCTTGTTTGAAGAGACCAAGCGGAAGGCAGAACAGATG  GAAAAGAAGCAACATGAGGAATTTGACCTGAGCCCTGACCAGTGGCGGAAAGGAAAGAAGTCTACTTTCCGGCATGTTGGAGATG  ACACCACTCGCCTGGAGATCTGGATTCATCCATCCCATCCACGGTCTTCCCAGGGCACTGAGTCTGGGAAGGATTCTGAGCAAGA  GAATGGGCTGGGCAGCCTGAGTCCCAGTGATCTGAACAAAAGGTTCATCCTCAGTTTTCTCCATGCCCATGGGAAGCTGTTTACC  CGGATTGGGATGGAGACATTCCCTGCAGTGGCTGAGAAGGTCCTCAAGGAGTTCCAGGTGTTACTGCAGCACAGCCCCTCTCCCA  TTGGAAGTACCCGCATGCTGCAGCTTATGACCATCAATATGTTTGCAGTACACAACTCCCAGCTGAAAGACTGCTTCTCGGAGGA  GTGCCGCTCTGTGATCCAGGAACAAGCCGCAGCTCTGGGCTTGGCCATGTTTTCTCTACTGGTCCGCCGCTGCACCTGCTTACTT  AAGGAGTCCGCCAAAGCTCAGCTGTCCTCTCCTGAGGACCAGGATGACCAAGACGACATCAAGGTGTCTTCCTTTGTCCCGGACC  TGAAGGAGCTGCTCCCCAGTGTCAAAGTCTGGTCAGATTGGATGCTCGGCTACCCGGACACCTGGAATCCTCCTCCCACATCCCT  GGATCTGCCCTCGCATGTTGCTGTGGATGTATGGTCGACGCTGGCTGATTTCTGTAACATACTGACTGCAGTGAATCAGTCTGAG  GTGCCACTGTACAAGGACCCGGATGATGACCTCACCCTTCTTATCCTGGAAGAGGATCGGCTTCTCTCGGGCTTTGTCCCCTTGC  TGGCTGCCCCTCAGGACCCCTGCTACGTGGAGAAAACCTCGGATAAGGTTATTGCAGCTGACTGCAAAAGGGTCACAGTGCTGAA  GTATTTTCTGGAAGCCCTTTGTGGACAAGAAGAGCCTCTGCTGGCATTCAAGGGTGGAAAGTATGTGTCAGTGGCACCCGTCCCA  GACACCATGGGAAAGGAAATGGGAAGCCAAGAGGGAACACGACTGGAAGATGAGGAGGAGGATGTGGTGATTGAAGACTTTGAGG  AAGATTCAGAGGCTGAAGGCAGCGGAGGCGAGGATGACATCAGGGAGCTTCGGGCCAAGAAGCTGGCTCTGGCCAGGAAGATAGC  TGAGCAGCAGCGTCGCCAGGAAAAGATCCAGGCTGTCCTGGAGGACCACAGTCAGATGAGGCAGATGGAGCTCGAAATCAGACCT  TTGTTCCTCGTACCAGACACCAACGGCTTCATTGACCACCTGGCCAGTCTGGCGCGGCTGCTGGAGAGCAGGAAGTACATCCTGG  TGGTGCCCCTCATCGTGATCAATGAGCTGGACGGCCTGGCCAAGGGGCAGGAGACAGACCACCGGGCTGGGGGCTACGCCCGTGT  GGTACAAGAGAAGGCCCGCAAGTCCATCGAGTTCCTCGAGCAGCGATTCGAGAGTCGGGACTCTTGCCTGCGAGCCCTGACCAGC  CGTGGCAATGAACTCGAATCCATCGCCTTCCGCAGTGAGGACATCACTGGCCAGCTGGGTAACAACGATGATCTCATCCTGTCCT  GCTGCCTCCACTACTGCAAAGACAAGGCTAAGGACTTCATGCCCGCCAGCAAAGAGGAGCCAATCCGGCTACTGCGGGAGGTGGT  GCTGTTGACGGATGACCGGAACCTGCGTGTGAAGGCGCTCACAAGGAATGTTCCTGTACGGGACATCCCAGCCTTCCTCACGTGG  GCCCAGGTGGGCGCGGCCGCCGACTACAAGGACCACGACGGTGACTACAAGGACCACGACATCGACTACAAGGACGACGACGACA  AGTGATTCTAGAGGATCATAATCAGCCATACCACATTTGTAGAGGTTTTACTTGCTTTAAAAAACCTCCCACACCTCCCCCTGAA  CCTGAAACATAAAATGAATGCAATTGTTGTTGTTAACTTGTTTATTGCAGCTTATAATGGTTACAAATAAAGCAATAGCATCACA  AATTTCACAAATAAAGCATTTTTTTCACTGCCTCGAGCTTCCTCATCGCCGGTACCAAAAAAAAAAAAAAAAAAAAAAAAAAAAA 

Cell-free gene expression: an expanded repertoire of applications

Strategies for in vitro engineering of the translation machinery

RNA codewords and protein synthesis, VII. On the general nature of the RNA code

The dependence of cell-free protein synthesis in E. coli upon naturally occurring or synthetic polyribonucleotides

Cell-free protein synthesis from genomically recoded bacteria enables multisite incorporation of noncanonical amino acids

A fully orthogonal system for protein synthesis in bacterial cells

Cell-Free Protein Synthesis: Pros and Cons of Prokaryotic and Eukaryotic Systems

An Efficient mRNA-Dependent Translation System from Reticulocyte Lysates

Characterization of Novel Ribosome-Associated Endoribonuclease SLFN14 from Rabbit Reticulocytes

Cell-free protein synthesis: Applications come of age

Cell-free protein expression based on extracts from CHO cells

Picornavirus IRESes and the poly(A) tail jointly promote cap-independent translation in a mammalian cell-free system

An efficient mammalian cell-free translation system supplemented with translation factors

An efficient factor-depleted mammalian in vitro translation system

Highly efficient in vitro translation of authentic affinitypurified messenger ribonucleoprotein complexes. RNA

Human NMD ensues independently of stable ribosome stalling

Dual centrifugation -A new technique for nanomilling of poorly soluble drugs and formulation screening by an DoE-approach

Dual Centrifugation -A Novel "in-vial

SARS-CoV-2 Nsp1 binds the ribosomal mRNA channel to inhibit translation

A dual-luciferase reporter system for studying recoding signals

Protein synthesis initiation in eukaryotic cells

Internal initiation of translation of eukaryotic mRNA directed by a sequence derived from poliovirus RNA

A segment of the 5' nontranslated region of encephalomyocarditis virus RNA directs internal entry of ribosomes during in vitro translation

SMG6 promotes endonucleolytic cleavage of nonsense mRNA in human cells

SMG6 is the catalytic endonuclease that cleaves mRNAs containing nonsense codons in metazoan

Nanopore sequencing reveals endogenous NMD-targeted isoforms in human cells

Nterminally truncated GADD34 proteins are convenient translation enhancers in a human cell-derived in vitro protein synthesis system

GTP-dependent Recognition of the Methionine Moiety on Initiator tRNA by Translation Factor eIF2

Feedback inhibition of the unfolded protein response by GADD34-mediated dephosphorylation of eIF2α

Protein synthesis in pseudorabies virus-infected cells: decreased expression of protein kinase PKR, and effects of 2-aminopurine and adenine

2-Aminopurine Inhibits the Double-Stranded RNA-Dependent Protein Kinase Both In Vitro and In Vivo

We thank Prof. Ulrich Massing and his team at Hettich AG for fruitful discussions about the possible applications of DC and their introduction to the ZentriMix 380 R system. We also thank Andrea Eberle for her technical advice on the project and the critical reading of the manuscript, Nicole Kleinschmidt for her technical advice, and Sofia Nasif for her remarks on the manuscript. 

The authors declare no conflicts of interest.