Prokaryotic organisms can rely on polycistronic transcription (operons), which allows for the expression of needed mRNAs from a single promoter and enables a rapid and robust response to corresponding stimuli (e.g., lactose operon) (Jacob and Monod, 1961). However, this mode of action has not been reported in eukaryotes, except for a few limited examples (e.g., in Caenorhabditis elegans [C. elegans]) (Spieth et al., 1993), and eukaryotic operons have been observed to give rise to dicistronic messages (Blumenthal, 2004). However, it is intriguing to postulate whether eukaryotes have devised functional alternatives to operons. One possibility is that eukaryotic messenger ribonucleoprotein (mRNP) particles, which are composed of multiple RNAs and RNA-binding proteins (RBPs) (Mitchell and Parker, 2014), effectively confer combinatorial gene expression networks similar to prokaryotic operons (Keene and Tenenbaum, 2002). In this case, however, the mRNP particles contain individual mRNAs that undergo co-translational control and may encode proteins involved in the same biological process or molecular complex. The idea that mRNPs might essentially constitute RNA operons (Keene, 2007) or transperons, as termed here, implies several features. First, although mRNAs derived from different chromosomes can reside in the same mRNP, they should have common regulatory elements not only for transcriptional and translational control, but also for interacting at the post-transcriptional level with other mRNAs within the particle. The latter includes motifs/structures to facilitate interactions with shared RBPs, as well as elements that might facilitate RNA-RNA interactions (e.g., base-pairing). Second, some mechanism must confer the recruitment and assembly, in trans (i.e., multiplexing), of the individual mRNAs into single mRNP particles. Third, in order to be functionally relevant, transperons should contain mRNAs encoding proteins involved in the same functional context (e.g., organelle biogenesis, macromolecular complex, or biological process). Although the mechanism remains unknown, studies have shown that genomic DNA folds create sites of transcriptional hot spots (Lieberman-Aiden et al., 2009; Rao et al., 2014) that could facilitate multiplexing by co-localizing messages during transcription and coordinating the subsquent association of RBPs. Moreover, RNA-RNA interactome studies show that extensive interactions can occur directly between RNAs in living cells (Aw et al., 2016; Engreitz et al., 2014; Kudla et al., 2011; Nguyen et al., 2016; Sharma et al., 2016). Thus, the multiplexing of mRNAs into mRNPs to yield transperons should be directly testable.

Although individual yeast RBPs have been shown to bind to numerous (e.g., 10s-1000s) mRNAs (Hogan et al., 2008), these RNA-binding studies are based primarily upon crosslinking and RBP pulldowns that are unable to define the minimal number of mRNA species in a single mRNP particle. To test the idea of RNA multiplexing and define the composition of such particles, we employed a single mRNA species pulldown procedure (RaPID) (Slobodin and Gerst, 2010; Slobodin and Gerst, 2011). This method employs MS2 aptamer tagging of endogenously expressed messages in yeast (Haim et al., 2007) and their precipitation from cell lysates via the MS2 coat protein (MCP) to identify bound non-tagged transcripts using RNA-seq (RaPID-seq). We have previously demonstrated that RaPID-seq identifies MS2-tagged target mRNAs (Haimovich et al., 2016) and now show here that additional messages associate with these transcripts.

By employing MATα yeast in RaPID-seq pulldown experiments, we found that select MS2-labeled target mRNAs (e.g., SRO7, EXO70, OM45) co-precipitated untagged mRNAs encoding secreted proteins involved in cell mating (e.g., STE3, SAG1, MFα1, MFα2). Since a corresponding complex (e.g., STE2, AGA1, AGA2, MFA1, MFA2) could also be precipitated from MATa cells, it suggested to us that mRNAs encoding secreted proteins involved in mating (e.g., pheromones, pheromone receptors, agglutinins, and proteases) multiplex into a functionally selective mRNP complex – the mating transperon. To identify potential factors involved in mating transperon assembly, we employed pulldowns of these mRNAs followed by mass spectrometry (RaPID-MS). We found that the yeast histone H4 paralogs, Hhf1 and Hhf2 (Dollard et al., 1994), interact with these mRNAs to regulate complex assembly. Furthermore, both pheromone responsiveness (e.g., shmooing) and mating were inhibited by either histone H4 depletion or a block in H4 acetylation. When mutated, no other histones had this effect. Moreover, we could identify conserved cis elements in the mating mRNAs and demonstrate that mutations therein had deleterious effects on mRNA multiplexing, pheromone responsiveness, and mating, which were similar to those seen upon histone H4 mutation.

To help elucidate the mechanism by which mRNA multiplexing might occur, we performed chromatin conformation capture (Lieberman-Aiden et al., 2009), as well as genomic locus tagging experiments, to look for evidence of direct allelic interactions. Importantly, a specific interaction between the STE2 and AGA2 genes was identified in MATa cells, which suggests that interallelic coupling may give rise to RNA multiplexing. Furthermore, coupling appeared to occur in both a transcription- and histone H4-dependent manner. To help validate this hypothesis, we examined whether mRNAs encoding heat shock proteins (HSPs), whose genes were previously shown to undergo intergenic interactions during heat shock (Chowdhary et al., 2019), also undergo multiplexing. Importantly, HSP mRNAs were found to multiplex in trans during heat shock to form RNP complexes, and histone H4 was required for both RNA multiplexing and robust post-heat shock cell recovery. Parallel studies revealed the existence of additional RNA multiplexes for mitochondrial outer membrane proteins and MAP kinase pathway proteins in yeast. Together, these results suggest that histone H4-mediated chromatin interactions may lead to the formation of functionally selective transperons in eukaryotic cells and possibly act as a means to co-regulate gene expression in place of polycistronicity.

Prokaryotes coordinate the expression of genes encoding proteins involved in the same biological process/context by sequential placement in chromosomes in order to generate polycistronic messages (operons) that can be translationally controlled. Perhaps due to the greater complexity observed both at the gene and cellular levels, eukaryotes have preferentially relied on the discontiguous distribution of genes over different chromosomes. In this case, the co-regulation of gene expression presumably becomes dependent on shared transcriptional control elements and specific RBPs that interact selectively with certain transcripts and confer trafficking to sites of co-translation. However, given that the number of RBPs is limited and that they can interact with a wide number of potential transcripts, it is unclear if and how defined subsets of mRNAs are packaged together in order to facilitate co-translation.

We hypothesized that mRNAs encoding proteins involved in the same process might assemble into macromolecular complexes composed of multiplexed transcripts and specific RBPs to form discrete RNP complexes. By performing mRNA pulldown and RNA-seq experiments, we first identified a subset of mRNAs that encode secreted components involved in the mating of α-haploid cells. This mRNP complex included mRNAs encoding the a-pheromone receptor (STE3), α-mating factor (MFα1, MFα2), α-agglutinin (SAG1), and a-pheromone blocker (AFB1) (Figures 1A, B and 2A, and Figure 1—figure supplement 1). Correspondingly, we identified a mating mRNP complex from MATa cells that included mRNAs encoding the α-mating factor receptor (STE2), a-mating pheromone (MFA1, MFA2), and a-agglutinins (AGA1, AGA2) (Figure 2B). Thus, we demonstrated that mRNAs coding for proteins of the same cellular process multiplex into mRNP complexes, presumably to allow for their delivery to the same intracellular site for local translation. Additional experiments in which we identified the existence of other mRNA subsets that multiplex, like those encoding HSPs, MOMPs, and MAPK pathway proteins (Figures 8 and 9), clearly support this idea. Thus, mRNA multiplexing in trans appears to form functional transperons (i.e., RNA operons [Keene, 2007; Keene and Tenenbaum, 2002]) relevant to cell physiology (see below). While we do not know the full extent of RNA multiplexing, we might predict that analogous transperons (i.e., containing other transcripts) form within cells to modulate other physical processes. For example, recent works by Ashe and colleagues have shown that mRNAs encoding either translation factors or glycolytic enzymes co-localize and undergo co-translation in yeast (Morales-Polanco et al., 2021; Pizzinga et al., 2019). In addition, mRNAs encoding proteins in hetero-oligomeric/multisubunit complexes (proteasome, exocyst, COPI, FAS, ribosome, etc.) might form multiplexes to facilitate polypeptide complex assembly upon co-translation, as shown at the protein level (Shiber et al., 2018), even if dedicated assembly chaperones are needed. In our study, we may have missed multiplexes in the initial RaPID-seq experiment (Figure 1 and Figure 1—figure supplement 1) for several reasons. First, RNA multiplexing may require elevated gene expression, as in the case of the HSP or MOMP complexes (Figures 8 and 9A–C), which require induction conditions to visualize the RNA granules, as opposed to the normal growth conditions employed initially. Second, the RaPID-seq pulldowns employed only 0.01% formaldehyde in the crosslinking procedure, as initially published (Slobodin and Gerst, 2010), and not the 0.1–0.5% formaldehyde we have employed in the later work (Zabezhinsky et al., 2016). Third, the precipitated complexes were stringently washed and, thus, these different factors may have disrupted weaker transperons or missed them entirely.

Importantly, by using a variety of molecular and single-cell imaging techniques (e.g., Figure 3C; live fluorescence imaging) we show specific and robust physical interactions between genes located on different chromosomes that encode mRNAs that form transperons (Figure 7 and Figure 7—figure supplement 1A and B). This observation of allelic coupling is strengthened by the work of Chowdhary et al., 2019, where the intergenic association of yeast HSP genes was shown. Other chromatin conformation capture studies have revealed intergenic associations that lead to the so-called transcription factories associated with different transcriptional regulators (e.g., promoters, transcription factors). For example, Schoenfelder et al., 2010 showed interchromosomal interactions between Klf4-regulated globin genes in fetal liver cells that allowed for co-regulated gene transcription. Papantonis et al., 2012 revealed that TNF-α responsive coding and miRNA genes in human endothelial cells undergo intrachromosomal interactions regulated by NFκB upon cytokine induction. Thus, specialized transcriptional factories created through allelic coupling may be a common mechanism for co-regulated gene expression in eukaryotes.

How allelic coupling comes about to form these factories and resulting transperons is not yet clear. However, we identified a role for histone H4 in the assembly of the mating transperon (Figures 3C, E and 4C, D, and Figure 4—figure supplement 1E) and its physiological consequences (i.e., pheromone responsiveness of both mating partners and mating; Figure 4A and B, and Figure 6—figure supplement 2A). RaPID-MS pulldowns of mating mRNAs encoding secreted proteins identified the H4 paralog, Hhf1, as binding to the STE2, MFA1, and MFA2 mRNAs, but not to ASH1 mRNA or STE2 mRNA lacking its 3’UTR (Figure 3—figure supplement 1). Thus, specific H4-RNA interactions are likely to be involved in transperon assembly. Hhf1 pulldowns directly precipitated the mating transperon (Figure 3C and E) and the deletion of either paralog (HHF1 or HHF2) led to defects in RNA multiplexing (Figure 4C and D) and mating (Figure 4A and B). Likewise, a combination of HHF2 deletion and an auxin-induced degradative form of Hhf1 resulted in a complete block in mating and loss of mRNA co-localization under conditions in which cell viability was not affected (Figure 4B and Figure 4—figure supplement 1C and E). Similar results were obtained using the HSP mRNAs, which also form physiologically functional transperons and co-localize in a manner dependent upon histone H4 (Figure 8, Figure 8—figure supplement 1D, and Figure 8—figure supplement 2). This suggests that both Hhf1 and Hhf2 are important for the formation of the mRNA complexes identified here.

Mutations in the amino terminal histone H4 acetylation sites had similar effects as gene deletions, whereby inactivating K-to-R mutations inhibited mating (Figure 5B and C). In contrast, K-to-Q mutations (or H4 overexpression) improved mating (Figure 5B). Thus, histone H4 paralogs, but no other histones (Figure 4A and Figure 8—figure supplement 1C), appear to affect the transmission of physiological signals relevant to RNA multiplexing. More work is required to demonstrate the role histone H4 plays in transperon formation and whether H4 is present in these multiplexes post-assembly and export from the nucleus. Although we found that fluorescent protein-tagged Hhf1 labeled the nucleus, we were unable to observe H4 in the cytoplasm or show co-localization there with STE2 mRNA (data not shown). Thus, we do not know whether this non-canonical interaction of H4 with mRNA is exclusive to the nucleus or not. Canonical histone functions relate mainly to DNA-protein interactions that facilitate nucleosome formation. However, the requirement of acetylation for H4 function in mating (Figure 5A–C) suggests that histone H4 might participate in the formation of a DNA-RNA intermediate upon gene transcription. This mechanism potentially allows for RNA multiplexing, provided that the loci encoding mRNAs intended for assembly in trans also are in close apposition, as evidenced here (Figure 7B and C, and Figure 7—figure supplement 1A,B and F). Results showing that a transcription inhibitor affects allelic coupling imply that gene apposition and transcription may be connected (Figure 7—figure supplement 1E). Moreover, the fact that H4 deletion also affects allelic coupling may implicate an involvement of chromatin architecture in transperon formation (Figure 7—figure supplement 1F).

Our lab is interested in how mRNA localization affects cell physiology, and a previous work has demonstrated that yeast mSMPs (including MFA2) localize to ER in the cell body (Kraut-Cohen et al., 2013). Here, we verified that endogenously expressed STE2 and AGA1 mRNAs also localize to ER in the cell body, both before and after treatment with pheromone (Figure 2C). Moreover, these mRNAs co-localize both prior to and post nuclear export (Figure 2D–F). This finding demonstrates once again that mSMPs are not asymmetrically localized to the polarized extensions of yeast (e.g., bud and shmoo tips), unlike ASH1 and polarity-establishing mRNAs (e.g., SRO7, SEC4, CDC42) in budding cells, which is mediated by She2 (Aronov et al., 2007), or SRO7 and FUS3 mRNAs in shmooing cells, which is mediated by Scp160 (Gelin-Licht et al., 2012). It also demonstrates that RNA multiplexing and transperon formation occur within the nucleus (Figures 2D–F and 8B). Since this work and previous studies have revealed that mRNAs encoding secreted mating pathway components, MOMPs, and pheromone-activated MAPK pathway components localize to the ER, mitochondria, and ER of the shmoo tip, respectively (Gadir et al., 2011; Gelin-Licht et al., 2012; Zabezhinsky et al., 2016), we predict that these mRNAs are targeted as intact transperons. Moreover, as Scp160 binds MAPK pathway mRNAs (e.g., FUS3, STE7, KAR3) and delivers them to the shmoo tip upon pheromone treatment (Gelin-Licht et al., 2012), we predict that Scp160 might traffick this MAPK transperon.

An earlier work has (Aronov et al., 2015) suggested that mRNAs encoding mating components, like MFA2, localize preferentially to P-bodies in the shmoos and shmoo tips of pheromone-treated yeast, and that P-bodies are necessary for transmission of the mating signal and subsequent mating. However, these results were obtained under conditions of mRNA overexpression. In contrast, we found that neither endogenously expressed MS2-tagged MFA2 (Haimovich et al., 2016) nor STE2 mRNA (Figure 6—figure supplement 2C) localize to P-bodies, nor do P-bodies form under normal growth conditions conducive to the expression of these mRNAs and to mating (Figure 6—figure supplement 2C; top panel). In addition, we show here that MFA2 and other mating mRNAs (e.g. MFA1, STE2, and AGA1) do not localize to the shmoos or shmoo tips, but rather to ER present in the cell body (Figure 2C and Figure 2—figure supplement 1). Moreover, the deletion of genes necessary for P-body formation (e.g., PAT1, DHH1) had no significant effect upon mating (Figure 6—figure supplement 2B). As mating pathway mRNAs localize to P-bodies only upon stress conditions (e.g., gene overexpression, glucose starvation) (Haimovich et al., 2016 and Figure 6—figure supplement 2C), it would appear that P-bodies are not sites for mating mRNA localization or transmission of the mating signal under normal growth conditions. Our findings re-emphasize the importance of (i) measuring mRNA localization under native conditions of gene expression; (ii) considering the possibility that RNA overexpression and/or the insertion of RNA aptamers may induce artifacts (e.g., mRNA mislocalization, P-body formation, or both); and (iii) examining mRNA integrity/localization using additional techniques (e.g., Northern analysis, RNA-seq, smFISH) (Garcia and Parker, 2015; Haimovich et al., 2016; Heinrich et al., 2017) for secondary validation.

As cis-acting RNA elements may work in conjunction with specific RNA-binding proteins to confer multiplexing, we analyzed the coding regions and 3’UTRs of the mating mRNAs from the two yeast haplotypes. Sequence analysis revealed the presence of conserved motifs present in the mating mRNAs from either MATa or MATα cells (Figure 6A). We mutated the MATa sequence element in the AGA2 gene and found that it affected transperon assembly similar to the deletion of AGA2 (aga2Δ) (Figure 6B). Importantly, this correlated with an inhibition in cellular responsiveness (i.e., shmooing) toward its mating partner, as well as the responsiveness of its mating partner (Figure 6—figure supplement 2A), resulting in significant defects in mating (Figure 6C). Thus, defects in transperon formation result in significant changes in cell physiology, and we predict that a yet unidentified RBP interacts with this element to facilitate RNA assembly and, possibly, trafficking. The MATa element is unlikely to be the H4-interacting motif itself, since it was not identified among the mating mRNAs from MATα cells, which have other shared motifs (Figure 6—figure supplement 1C). Likewise, removal of the 3’UTR from STE2 appeared to reduce H4 binding (Figure 3—figure supplement 1), possibly indicating its role in histone association. Additional RaPID-MS experiments should help reveal the identity of this protein, as well as those interacting with the motifs in mating mRNAs from MATα cells (Figure 6—figure supplement 1C), and genetic analyses should reveal their contribution to the mating process.

Finally, we note that in a few cases, transcripts from genes that neighbor the target gene were enriched in MS2-tagged target mRNA pulldowns in our RaPID-seq experiment (Figure 1A and Figure 1—figure supplement 1; see Supplementary file 1 for RNA-seq data), as compared to the bead control. For example, we observed that PET122 (YER153C) mRNA precipitated with OXA1 (YER154W); SPE1 (YKL184W) with ASH1 (YKL185W); SNC1 (YAL030W) with MYO4 (YAL029C); SNC2 (YOR327C) with MYO2 (YOR326W); and ALY2 (YJL084C) with EXO70 (YJL085W). It is possible that these genes are co-expressed and that their chromatin-associated RNA transcripts precipitate during RaPID, prior to DNase treatment. We did not examine these interactions further. Nonetheless, this possibility does not negate our hypothesis supporting the existence of RNA multiplexes/transperons, which was validated by a variety of means (e.g., RaPID-qRT-PCR, smFISH, and physiological assays).

To the best of our knowledge, the results presented here support the concept that eukaryotic mRNP complexes/transperons are the functional equivalents of bacterial operons, as first proposed by Keene, 2007 and Keene and Tenenbaum, 2002. Thus, eukaryotes may circumvent polycistronicity by assembling relevant mRNAs into functional complexes that confer co-translation. This is likely to occur via physical intergenic interactions that couple individual alleles concurrent with transcription to form transperons that can undergo export and translation at different sites within the cell (Figure 10A). A schematic describing mRNA assembly into functional transperons encoding secreted mating pathway components via histone H4 mediation is shown in Figure 10B. While this work implies that there are higher order principles governing the selection of mRNAs for mRNP complex assembly, further study is required to uncover the entire mechanism by which mRNAs are recruited and assembled together into multiplexes, and to elucidate how histones and other proteins link mRNAs to fully formed transperons.

Figure 10

Download asset Open asset

All data is available within the text, figures, and tables of the manuscript.

1. 1. Allis CD
   2. Jenuwein T

   (2016) The molecular hallmarks of epigenetic control

   Nature Reviews Genetics 17:487–500.

   https://doi.org/10.1038/nrg.2016.59

   • Google Scholar
2. 1. Aronov S
   2. Gelin-Licht R
   3. Zipor G
   4. Haim L
   5. Safran E
   6. Gerst JE

   (2007) mRNAs encoding polarity and exocytosis factors are cotransported with the cortical endoplasmic reticulum to the incipient bud in Saccharomyces cerevisiae

   Molecular and cellular biology 27:3441–3455.

   https://doi.org/10.1128/MCB.01643-06

   • PubMed
   • Google Scholar
3. 1. Aronov S
   2. Dover-Biterman S
   3. Suss-Toby E
   4. Shmoish M
   5. Duek L
   6. Choder M

   (2015) Pheromone-encoding mRNA is transported to the yeast mating projection by specific RNP granules

   The Journal of cell biology 209:829–842.

   https://doi.org/10.1083/jcb.201408045

   • PubMed
   • Google Scholar
4. 1. Aw JG
   2. Shen Y
   3. Wilm A
   4. Sun M
   5. Lim XN
   6. Boon KL
   7. Tapsin S
   8. Chan YS
   9. Tan CP
   10. Sim AY
   11. Zhang T
   12. Susanto TT
   13. Fu Z
   14. Nagarajan N
   15. Wan Y

   (2016) In Vivo mapping of eukaryotic RNA interactomes reveals principles of Higher-Order organization and regulation

   Molecular Cell 62:603–617.

   https://doi.org/10.1016/j.molcel.2016.04.028

   • PubMed
   • Google Scholar
5. 1. Bailey TL
   2. Boden M
   3. Buske FA
   4. Frith M
   5. Grant CE
   6. Clementi L
   7. Ren J
   8. Li WW
   9. Noble WS

   (2009) MEME SUITE: tools for motif discovery and searching

   Nucleic acids research 37:W202–W208.

   https://doi.org/10.1093/nar/gkp335

   • PubMed
   • Google Scholar
6. 1. Blumenthal T

   (2004) Operons in eukaryotes

   Briefings in functional genomics & proteomics 3:199–211.

   https://doi.org/10.1093/bfgp/3.3.199

   • PubMed
   • Google Scholar
7. 1. Bray NL
   2. Pimentel H
   3. Melsted P
   4. Pachter L

   (2016) Near-optimal probabilistic RNA-seq quantification

   Nature biotechnology 34:525–527.

   https://doi.org/10.1038/nbt.3519

   • PubMed
   • Google Scholar
8. 1. Chowdhary S
   2. Kainth AS
   3. Pincus D
   4. Gross DS

   (2019) Heat shock factor 1 drives intergenic association of its target gene loci upon heat shock

   Cell Reports 26:18–28.

   https://doi.org/10.1016/j.celrep.2018.12.034

   • PubMed
   • Google Scholar
9. 1. Dollard C
   2. Ricupero-Hovasse SL
   3. Natsoulis G
   4. Boeke JD
   5. Winston F

   (1994) SPT10 and SPT21 are required for transcription of particular histone genes in Saccharomyces cerevisiae

   Molecular and cellular biology 14:5223–5228.

   https://doi.org/10.1128/mcb.14.8.5223

   • PubMed
   • Google Scholar
10. 1. Dovrat D
    2. Dahan D
    3. Sherman S
    4. Tsirkas I
    5. Elia N
    6. Aharoni A

    (2018) A Live-Cell imaging approach for measuring DNA replication rates

    Cell Reports 24:252–258.

    https://doi.org/10.1016/j.celrep.2018.06.018

    • PubMed
    • Google Scholar
11. 1. Engreitz JM
    2. Sirokman K
    3. McDonel P
    4. Shishkin AA
    5. Surka C
    6. Russell P
    7. Grossman SR
    8. Chow AY
    9. Guttman M
    10. Lander ES

    (2014) RNA-RNA interactions enable specific targeting of noncoding RNAs to nascent Pre-mRNAs and chromatin sites

    Cell 159:188–199.

    https://doi.org/10.1016/j.cell.2014.08.018

    • PubMed
    • Google Scholar
12. 1. Gadir N
    2. Haim-Vilmovsky L
    3. Kraut-Cohen J
    4. Gerst JE

    (2011) Localization of mRNAs coding for mitochondrial proteins in the yeast Saccharomyces cerevisiae

    RNA 17:1551–1565.

    https://doi.org/10.1261/rna.2621111

    • PubMed
    • Google Scholar
13. 1. Garcia JF
    2. Parker R

    (2015) MS2 coat proteins bound to yeast mRNAs block 5' to 3' degradation and trap mRNA decay products: implications for the localization of mRNAs by MS2-MCP system

    RNA 21:1393–1395.

    https://doi.org/10.1261/rna.051797.115

    • PubMed
    • Google Scholar
14. 1. Gelin-Licht R
    2. Paliwal S
    3. Conlon P
    4. Levchenko A
    5. Gerst JE

    (2012) Scp160-dependent mRNA trafficking mediates pheromone gradient sensing and chemotropism in yeast

    Cell reports 1:483–494.

    https://doi.org/10.1016/j.celrep.2012.03.004

    • PubMed
    • Google Scholar
15. 1. Genz C
    2. Fundakowski J
    3. Hermesh O
    4. Schmid M
    5. Jansen RP

    (2013) Association of the yeast RNA-binding protein She2p with the tubular endoplasmic reticulum depends on membrane curvature

    The Journal of biological chemistry 288:32384–32393.

    https://doi.org/10.1074/jbc.M113.486431

    • PubMed
    • Google Scholar
16. 1. Haas BJ
    2. Papanicolaou A
    3. Yassour M
    4. Grabherr M
    5. Blood PD
    6. Bowden J
    7. Couger MB
    8. Eccles D
    9. Li B
    10. Lieber M
    11. MacManes MD
    12. Ott M
    13. Orvis J
    14. Pochet N
    15. Strozzi F
    16. Weeks N
    17. Westerman R
    18. William T
    19. Dewey CN
    20. Henschel R
    21. LeDuc RD
    22. Friedman N
    23. Regev A

    (2013) De novo transcript sequence reconstruction from RNA-seq using the Trinity platform for reference generation and analysis

    Nature protocols 8:1494–1512.

    https://doi.org/10.1038/nprot.2013.084

    • PubMed
    • Google Scholar
17. 1. Hafner M
    2. Landthaler M
    3. Burger L
    4. Khorshid M
    5. Hausser J
    6. Berninger P
    7. Rothballer A
    8. Ascano M
    9. Jungkamp AC
    10. Munschauer M
    11. Ulrich A
    12. Wardle GS
    13. Dewell S
    14. Zavolan M
    15. Tuschl T

    (2010) Transcriptome-wide identification of RNA-binding protein and microRNA target sites by PAR-CLIP

    Cell 141:129–141.

    https://doi.org/10.1016/j.cell.2010.03.009

    • PubMed
    • Google Scholar
18. 1. Hagège H
    2. Klous P
    3. Braem C
    4. Splinter E
    5. Dekker J
    6. Cathala G
    7. de Laat W
    8. Forné T

    (2007) Quantitative analysis of chromosome conformation capture assays (3C-qPCR)

    Nature protocols 2:1722–1733.

    https://doi.org/10.1038/nprot.2007.243

    • PubMed
    • Google Scholar
19. 1. Haim L
    2. Zipor G
    3. Aronov S
    4. Gerst JE

    (2007) A genomic integration method to visualize localization of endogenous mRNAs in living yeast

    Nature methods 4:409–412.

    https://doi.org/10.1038/nmeth1040

    • PubMed
    • Google Scholar
20. 1. Haimovich G
    2. Zabezhinsky D
    3. Haas B
    4. Slobodin B
    5. Purushothaman P
    6. Fan L
    7. Levin JZ
    8. Nusbaum C
    9. Gerst JE

    (2016) Use of the MS2 aptamer and coat protein for RNA localization in yeast: A response to "MS2 coat proteins bound to yeast mRNAs block 5' to 3' degradation and trap mRNA decay products: implications for the localization of mRNAs by MS2-MCP system"

    RNA 22:660–666.

    https://doi.org/10.1261/rna.055095.115

    • PubMed
    • Google Scholar
21. 1. Heinrich S
    2. Sidler CL
    3. Azzalin CM
    4. Weis K

    (2017) Stem-loop RNA labeling can affect nuclear and cytoplasmic mRNA processing

    RNA 23:134–141.

    https://doi.org/10.1261/rna.057786.116

    • PubMed
    • Google Scholar
22. 1. Hogan DJ
    2. Riordan DP
    3. Gerber AP
    4. Herschlag D
    5. Brown PO

    (2008) Diverse RNA-binding proteins interact with functionally related sets of RNAs, suggesting an extensive regulatory system

    PLOS biology 6:e255.

    https://doi.org/10.1371/journal.pbio.0060255

    • PubMed
    • Google Scholar
23. 1. Jacob F
    2. Monod J

    (1961) Genetic regulatory mechanisms in the synthesis of proteins

    Journal of molecular biology 3:318–356.

    https://doi.org/10.1016/s0022-2836(61)80072-7

    • PubMed
    • Google Scholar
24. 1. Keene JD

    (2007) RNA regulons: coordination of post-transcriptional events

    Nature reviews. Genetics 8:533–543.

    https://doi.org/10.1038/nrg2111

    • PubMed
    • Google Scholar
25. 1. Keene JD
    2. Tenenbaum SA

    (2002) Eukaryotic mRNPs may represent posttranscriptional operons

    Molecular cell 9:1161–1167.

    https://doi.org/10.1016/s1097-2765(02)00559-2

    • PubMed
    • Google Scholar
26. 1. Kraut-Cohen J
    2. Afanasieva E
    3. Haim-Vilmovsky L
    4. Slobodin B
    5. Yosef I
    6. Bibi E
    7. Gerst JE

    (2013) Translation- and SRP-independent mRNA targeting to the endoplasmic reticulum in the yeast Saccharomyces cerevisiae

    Molecular biology of the cell 24:3069–3084.

    https://doi.org/10.1091/mbc.E13-01-0038

    • PubMed
    • Google Scholar
27. 1. Kudla G
    2. Granneman S
    3. Hahn D
    4. Beggs JD
    5. Tollervey D

    (2011) Cross-linking, Ligation, and sequencing of hybrids reveals RNA-RNA interactions in yeast

    PNAS 108:10010–10015.

    https://doi.org/10.1073/pnas.1017386108

    • Google Scholar
28. 1. Kurdistani SK
    2. Grunstein M

    (2003) Histone acetylation and deacetylation in yeast

    Nature reviews. Molecular cell biology 4:276–284.

    https://doi.org/10.1038/nrm1075

    • PubMed
    • Google Scholar
29. 1. Labun K
    2. Montague TG
    3. Gagnon JA
    4. Thyme SB
    5. Valen E

    (2016) CHOPCHOP v2: a web tool for the next generation of CRISPR genome engineering

    Nucleic acids research 44:W272–W276.

    https://doi.org/10.1093/nar/gkw398

    • PubMed
    • Google Scholar
30. 1. Larson DR
    2. Zenklusen D
    3. Wu B
    4. Chao JA
    5. Singer RH

    (2011) Real-time observation of transcription initiation and elongation on an endogenous yeast gene

    Science 332:475–478.

    https://doi.org/10.1126/science.1202142

    • PubMed
    • Google Scholar
31. 1. Levin JZ
    2. Yassour M
    3. Adiconis X
    4. Nusbaum C
    5. Thompson DA
    6. Friedman N
    7. Gnirke A
    8. Regev A

    (2010) Comprehensive comparative analysis of strand-specific RNA sequencing methods

    Nature methods 7:709–715.

    https://doi.org/10.1038/nmeth.1491

    • PubMed
    • Google Scholar
32. 1. Li C
    2. Wen A
    3. Shen B
    4. Lu J
    5. Huang Y
    6. Chang Y

    (2011) FastCloning: a highly simplified, purification-free, sequence- and ligation-independent PCR cloning method

    BMC biotechnology 11:92.

    https://doi.org/10.1186/1472-6750-11-92

    • PubMed
    • Google Scholar
33. 1. Lieberman-Aiden E
    2. van Berkum NL
    3. Williams L
    4. Imakaev M
    5. Ragoczy T
    6. Telling A
    7. Amit I
    8. Lajoie BR
    9. Sabo PJ
    10. Dorschner MO
    11. Sandstrom R
    12. Bernstein B
    13. Bender MA
    14. Groudine M
    15. Gnirke A
    16. Stamatoyannopoulos J
    17. Mirny LA
    18. Lander ES
    19. Dekker J

    (2009) Comprehensive mapping of long-range interactions reveals folding principles of the human genome

    Science 326:289–293.

    https://doi.org/10.1126/science.1181369

    • PubMed
    • Google Scholar
34. 1. Longtine MS
    2. McKenzie A
    3. Demarini DJ
    4. Shah NG
    5. Wach A
    6. Brachat A
    7. Philippsen P
    8. Pringle JR

    (1998) Additional modules for versatile and economical PCR-based gene deletion and modification in Saccharomyces cerevisiae

    Yeast 14:953–961.

    https://doi.org/10.1002/(SICI)1097-0061(199807)14:10<953::AID-YEA293>3.0.CO;2-U

    • PubMed
    • Google Scholar
35. 1. Mans R
    2. van Rossum HM
    3. Wijsman M
    4. Backx A
    5. Kuijpers NG
    6. van den Broek M
    7. Daran-Lapujade P
    8. Pronk JT
    9. van Maris AJ
    10. Daran JM

    (2015) CRISPR/Cas9: a molecular swiss army knife for simultaneous introduction of multiple genetic modifications in Saccharomyces cerevisiae

    FEMS Yeast Research 15:fov004.

    https://doi.org/10.1093/femsyr/fov004

    • PubMed
    • Google Scholar
36. 1. Mitchell SF
    2. Parker R

    (2014) Principles and properties of eukaryotic mRNPs

    Molecular cell 54:547–558.

    https://doi.org/10.1016/j.molcel.2014.04.033

    • PubMed
    • Google Scholar
37. 1. Morales-Polanco F
    2. Bates C
    3. Lui J
    4. Casson J
    5. Solari CA
    6. Pizzinga M
    7. Forte G
    8. Griffin C
    9. Garner KEL
    10. Burt HE

    (2021) Glycolytic mRNAs localise and are translated in core fermentation (CoFe) granules to fuel glucose fermentation

    iScience 24:102069.

    https://doi.org/10.1016/j.isci.2021.102069

    • Google Scholar
38. 1. Mueller F
    2. Senecal A
    3. Tantale K
    4. Marie-Nelly H
    5. Ly N
    6. Collin O
    7. Basyuk E
    8. Bertrand E
    9. Darzacq X
    10. Zimmer C

    (2013) FISH-quant: automatic counting of transcripts in 3D FISH images

    Nature methods 10:277–278.

    https://doi.org/10.1038/nmeth.2406

    • PubMed
    • Google Scholar
39. 1. Nguyen TC
    2. Cao X
    3. Yu P
    4. Xiao S
    5. Lu J
    6. Biase FH
    7. Sridhar B
    8. Huang N
    9. Zhang K
    10. Zhong S

    (2016) Mapping RNA-RNA interactome and RNA structure in vivo by MARIO

    Nature communications 7:12023.

    https://doi.org/10.1038/ncomms12023

    • PubMed
    • Google Scholar
40. 1. Papantonis A
    2. Kohro T
    3. Baboo S
    4. Larkin JD
    5. Deng B
    6. Short P
    7. Tsutsumi S
    8. Taylor S
    9. Kanki Y
    10. Kobayashi M
    11. Li G
    12. Poh HM
    13. Ruan X
    14. Aburatani H
    15. Ruan Y
    16. Kodama T
    17. Wada Y
    18. Cook PR

    (2012) Tnfα signals through specialized factories where responsive coding and miRNA genes are transcribed

    The EMBO Journal 31:4404–4414.

    https://doi.org/10.1038/emboj.2012.288

    • PubMed
    • Google Scholar
41. 1. Pizzinga M
    2. Bates C
    3. Lui J
    4. Forte G
    5. Morales-Polanco F
    6. Linney E
    7. Knotkova B
    8. Wilson B
    9. Solari CA
    10. Berchowitz LE
    11. Portela P
    12. Ashe MP

    (2019) Translation factor mRNA granules direct protein synthetic capacity to regions of polarized growth

    The Journal of cell biology 218:1564–1581.

    https://doi.org/10.1083/jcb.201704019

    • PubMed
    • Google Scholar
42. 1. Rao SS
    2. Huntley MH
    3. Durand NC
    4. Stamenova EK
    5. Bochkov ID
    6. Robinson JT
    7. Sanborn AL
    8. Machol I
    9. Omer AD
    10. Lander ES
    11. Aiden EL

    (2014) A 3D map of the human genome at kilobase resolution reveals principles of chromatin looping

    Cell 159:1665–1680.

    https://doi.org/10.1016/j.cell.2014.11.021

    • PubMed
    • Google Scholar
43. 1. Robinson MD
    2. McCarthy DJ
    3. Smyth GK

    (2010) edgeR: a Bioconductor package for differential expression analysis of digital gene expression data

    Bioinformatics 26:139–140.

    https://doi.org/10.1093/bioinformatics/btp616

    • PubMed
    • Google Scholar
44. 1. Schoenfelder S
    2. Sexton T
    3. Chakalova L
    4. Cope NF
    5. Horton A
    6. Andrews S
    7. Kurukuti S
    8. Mitchell JA
    9. Umlauf D
    10. Dimitrova DS
    11. Eskiw CH
    12. Luo Y
    13. Wei CL
    14. Ruan Y
    15. Bieker JJ
    16. Fraser P

    (2010) Preferential associations between co-regulated genes reveal a transcriptional interactome in erythroid cells

    Nature genetics 42:53–61.

    https://doi.org/10.1038/ng.496

    • PubMed
    • Google Scholar
45. 1. Sharma E
    2. Sterne-Weiler T
    3. O'Hanlon D
    4. Blencowe BJ

    (2016) Global Mapping of Human RNA-RNA Interactions

    Molecular cell 62:618–626.

    https://doi.org/10.1016/j.molcel.2016.04.030

    • PubMed
    • Google Scholar
46. 1. Shiber A
    2. Döring K
    3. Friedrich U
    4. Klann K
    5. Merker D
    6. Zedan M
    7. Tippmann F
    8. Kramer G
    9. Bukau B

    (2018) Cotranslational assembly of protein complexes in eukaryotes revealed by ribosome profiling

    Nature 561:268–272.

    https://doi.org/10.1038/s41586-018-0462-y

    • PubMed
    • Google Scholar
47. 1. Slobodin B
    2. Gerst JE

    (2010) A novel mRNA affinity purification technique for the identification of interacting proteins and transcripts in ribonucleoprotein complexes

    RNA 16:2277–2290.

    https://doi.org/10.1261/rna.2091710

    • PubMed
    • Google Scholar
48. 1. Slobodin B
    2. Gerst JE

    (2011) RaPID: an aptamer-based mRNA affinity purification technique for the identification of RNA and protein factors present in ribonucleoprotein complexes

    Methods in molecular biology 714:387–406.

    https://doi.org/10.1007/978-1-61779-005-8_24

    • PubMed
    • Google Scholar
49. 1. Spieth J
    2. Brooke G
    3. Kuersten S
    4. Lea K
    5. Blumenthal T

    (1993) Operons in C. elegans: polycistronic mRNA precursors are processed by trans-splicing of SL2 to downstream coding regions

    Cell 73:521–532.

    https://doi.org/10.1016/0092-8674(93)90139-h

    • PubMed
    • Google Scholar
50. 1. Tsanov N
    2. Samacoits A
    3. Chouaib R
    4. Traboulsi AM
    5. Gostan T
    6. Weber C
    7. Zimmer C
    8. Zibara K
    9. Walter T
    10. Peter M
    11. Bertrand E
    12. Mueller F

    (2016) smiFISH and FISH-quant - a flexible single RNA detection approach with super-resolution capability

    Nucleic acids research 44:e165.

    https://doi.org/10.1093/nar/gkw784

    • PubMed
    • Google Scholar
51. 1. Wu J
    2. Grunstein M

    (2000) 25 years after the nucleosome model: chromatin modifications

    Trends in biochemical sciences 25:619–623.

    https://doi.org/10.1016/s0968-0004(00)01718-7

    • PubMed
    • Google Scholar
52. 1. Zabezhinsky D
    2. Slobodin B
    3. Rapaport D
    4. Gerst JE

    (2016) An essential role for COPI in mRNA localization to mitochondria and mitochondrial function

    Cell Reports 15:540–549.

    https://doi.org/10.1016/j.celrep.2016.03.053

    • PubMed
    • Google Scholar
53. 1. Zipor G
    2. Haim-Vilmovsky L
    3. Gelin-Licht R
    4. Gadir N
    5. Brocard C
    6. Gerst JE

    (2009) Localization of mRNAs coding for peroxisomal proteins in the yeast, Saccharomyces cerevisiae

    PNAS 106:19848–19853.

    https://doi.org/10.1073/pnas.0910754106

    • PubMed
    • Google Scholar

Acceptance summary:

This work reveals that yeast regulate cellular processes by packaging mRNAs encoding proteins that function on the same pathway into single ribonucleoprotein particles that can be co-trafficked and presumably co-translated. Interestingly, this form of gene regulation is dependent upon histone H4 function and interchromosomal contacts that allow the genes encoding these mRNAs to be closely apposed during transcription. As opposed to prokaryotic cells that employ polycistronic operons, eukaryotic cells can employ monocistronic messages as RNA operons (transperons) to achieve gene co-expression.

Decision letter after peer review:

[Editors’ note: the authors submitted for reconsideration following the decision after peer review. What follows is the decision letter after the first round of review.]

Thank you for submitting your work entitled "Multiplexed mRNA assembly into ribonucleoprotein particles plays an operon-like role in controlling yeast physiology" for consideration by eLife. Your article has been reviewed by 3 peer reviewers, one of whom is a member of our Board of Reviewing Editors, and the evaluation has been overseen by a Senior Editor. The reviewers have opted to remain anonymous.

Our decision has been reached after consultation between the reviewers. Based on these discussions and the individual reviews below, we regret to inform you that your work will not be considered further for publication in eLife.

All three reviewers agreed that your manuscript contains very interesting information, and a description of functional mRNP packaging could be of great importance to the field. However, all three reviewers raise a number of important technical concerns that prevent publication of your manuscript in its current form. Given the extensive nature of the requested revisions, the decision was to 'reject' the paper, in line with the editorial policy of eLife. However, we would like to encourage you to perform the requested control experiments and to resubmit a revised manuscript. Although such a revised manuscript would be treated like a new submission, we would attempt to consult the same editors and reviewers in the review process.

Reviewer #1:

In this manuscript, Nair et al., propose that functionally related mRNAs can be packaged and transported in an operon-like manner. The discovery of such functional mRNP packages, termed transperons, could have wide-ranging implications for our understanding of the regulation of gene expression. Unfortunately, there are significant technical problems with the results, and while the manuscript contains a large amount of information, it lacks the required rigor, and key controls are missing. Most importantly, additional experiments are needed to demonstrate the existence of specific transperons.

1. Introduction of aptamer-binding sites can lead to the accumulation of decay intermediate that accumulate in P-body-like structures (see results from the Parker and Weis labs 2017). It is therefore critical that the authors perform extensive FISH experiments with untagged alleles (using probes against the body of the mRNAs). In situs should be performed for the different groups of mRNAs involved in mating, heat shock etc. The experiments need to be carefully quantified, statistically evaluated and all experiments need to include negative controls (i.e. mRNAs that do not co-localize). FISH experiments should be performed in wt and in H4 mutants. Of note, it is important to differentiate between nuclear spots, which likely present sites of transcription, and cytoplasmic spots. Only significant cytoplasmic-localization would lend further support for the transperon hypothesis.

2. The RAPID results lack a statistical evaluation. Not all interactions 'look' convincing, are these results significant?

3. The fact that deletion of DHH1 or PAT1 does not affect mating is not an argument against (1). However, the authors should colocalize Dhh1 and Pat1 in their aptamer-tagged strains. Note that this experiment does not replace the need of the FISH experiments described in (1).

4. The nuclear co-localization experiments with Hhf1 protein and the STE2 mRNA need negative controls. Hhf1 seems to fill the entire nucleus and thus is likely to "interact' with every mRNA that would be tested in such a way.

Reviewer #2:

In this manuscript the authors present a detailed set of experiments to describe and mechanistically characterize potential multiplexing of functionally related mRNAs. The focus is mainly the mating pathway mRNAs and the physiology of the mating pathway in yeast. The work supports models where the fate of functionally related mRNAs is coordinated and the authors show that this coordination is first seeded at the site of mRNA transcription which may involve a histone protein. The mechanistic advances supporting colocalized mRNA biogenesis at the site of transcription in the nucleus leading to a coordination of the subsequent cytosolic fate of this set of mRNAs are exciting and particularly noteworthy. –

Reviewer #3:

The observations presented are potentially interesting. But many of the data presented do not match, in my opinion, the quality standards that one would expect for such experiments in 2020.

A few examples. Only one RaPID-seq experiment is presented for each bait factor (Figure 1, S1 and Table S1). It is only stated "The experiment was performed twice and gave similar results"…… The standard for this type of experiment is rather to perform the experiments in triplicates and perform statistical analyses with programs such as DeSEQ2. The authors say that the replicate gives "similar results", why not provide them in Table S1 then?

Figure 2 A and B provide the data for the key experiment, which is at the basis of the whole manuscript. One should expect that this would be very robust data. In fact, this is not the case at all. The individual mRNA enrichments are not quantified by qRT-PCR, as one would expect, but by a single ethidium bromide stained agarose gel. Even worse, I am actually unable to see what the authors claim, i.e. a signal for the STE3, SRO7, OM45 and EXO70 PCR products in the IP fractions. The only signals one can see are for MFalpha1 and MFalpha2, but the signals in the input fraction are much more abundant for those genes that the signals seen for the control transcripts, making it impossible to say if these mRNAs are actually truly enriched. The control used for Figure A is the signal from the input, while for B, the only control provided is "-RT". How one would expect to get an idea of an enrichment level without the input data and the proper control mRNAs. Frankly, I don't understand. These experiments should be analyzed by qRT-PCR with the proper replicates and control mRNAs to be able to quantify enrichment values and calculate standard deviations. These remarks stand for Figure 3AandB, Figure 4C etc.

For the Fish experiments, some field view, with several cells, should be provided in the supplemental data.

Figure 6A: "Mating mRNAs and histone H4 co-localize only in the nucleus." I don't understand the meaning of this. Hhf1 localizes to the whole nucleus (see the yeast-GFP localization data base). Hence, in this type of experiment, Hhf1 will merely localize the nucleus (pretty much as dapy staining). How can one say that an mRNA fish signal, appearing as a dot, "co-localize" with Hhf1. It only says that the nuclear fraction of this mRNA is…… nuclear!

I will stop here. I think it is pretty clear that the authors should repeat properly quite a number of experiments before one would consider publishing this work.

[Editors’ note: further revisions were suggested prior to acceptance, as described below.]

Thank you for submitting your article "Multiplexed mRNA assembly into ribonucleoprotein particles plays an operon-like role in yeast cell physiology" for consideration by eLife. Your article has been reviewed by 3 peer reviewers, one of whom is a member of our Board of Reviewing Editors, and the evaluation has been overseen by Kevin Struhl as the Senior Editor. The reviewers have opted to remain anonymous.

The reviewers have discussed their reviews with one another, and the Reviewing Editor has drafted this to help you prepare a revised submission.

Essential Revisions:

The reviewers are concerned about the reproducibility of some of the experiments that are shown in the paper. The concerns are described in their detailed comments below but key issues include:

1. Figure 2

Given the high variability the results remain difficult to interpret.

2. AID experiments

The AID experiments are problematic as the mRNA levels go down and tagging alone in the absence of auxin leads to a protein reduction.

3. Colocalization microscopy data

The mRNA localization results vary dramatically between different experiments, which makes it difficult to interpret these results.

Reviewer #1:

In this revised paper, the authors have added additional controls and statistical analyses. This has improved the quality of the manuscript. However, there are still significant technical issues.

1. Figure 2A/B

Based on the figure legend it remains unclear what the authors call significant versus not significant?

2. Figure 2, 3, 8, 9

The mRNA localization results vary dramatically between different experiments (e.g., STE2 foci in different conditions). Why? Irrespectively, this makes it very hard to compare the experiments and the controls. Furthermore, the signal overlap in hs between SSA2 and HSP104 are not convincing. It could be a consequence of the enhance HSP104 expression.

3. The AID experiments in hs are not convincing. mRNA levels go down which might be sufficient to reduce co-localization (see 2). AID-tagging alone (without auxin) also leads to a reduction, which is potentially problematic.

4. Figure 9 (A-C) is also problematic since no convincing negative controls are included.

Reviewer #2:

In this manuscript the authors present a detailed set of experiments to describe and mechanistically characterize potential multiplexing of functionally related mRNAs. The focus is mainly the mating pathway mRNAs and the physiology of the mating pathway in yeast. The work supports models where the fate of functionally related mRNAs is coordinated and the authors show that this coordination is first seeded at the site of mRNA transcription which may involve a histone protein. The mechanistic advances supporting colocalized mRNA biogenesis at the site of transcription in the nucleus leading to a coordination of the subsequent cytosolic fate of this set of mRNAs are exciting and particularly noteworthy.

The authors present a revised version of a manuscript describing mRNA multiplexing and going on to study the mechanism of mRNA localization and physiological role of this process. The manuscript is very much improved over the previous version, but I do have some further comments for the authors – detailed below.

1. I wasn't convinced that the IST2 mRNA was significantly enriched in the pulldown in

Figure 2A relative to Figure 2B- as claimed by the authors. Plus many of the error bars On Figure 2B are very high across this analysis- so it is difficult to interpret.

2. Are the images in Figure 7C from a single focal plane? I'm guessing so since some of the cells don't have a spot in their nucleus. If this is the case the authors should add this information to the legend to prevent confusion.

3. Have the authors tested the expression levels of the Hhf1K-R and K-Q- as these mutations a differential effect on protein stability could explain the observations?

4. Similar to above- have the authors measured the RNA levels for AGA2 47-MUT or AGA2 27-MUT relative to wild type, as effects on RNA stability or expression could explain the observed effects.

Reviewer #3:

In this novel version of the manuscript, the authors have, I think, satisfactorily answered to most reviewer concerns about the first version.

Considering the suggestion to provide the results of the replicate of the Rapid-RNAseq experiment, the authors rather preferred to remove all mentions of it. Considering that all analyzed and discussed results of the now single documented experiment are validated by independent RaPID followed by qRT-PCR experiments, one can consider this as acceptable.

I have one suggestion, thought. When examining Figure 1 and the associated Table S1, one can see that some of the transcripts most highly enriched together with the bait MS2 tagged mRNAs are actually the mRNAs from the directly adjacent genes. This is true for PET122 (YER153C) pulled down by OXA1 (YER154W), or SPE1 (YKL184W) / ASH1 (YKL185W), SNC1 (YAL030W) / MYO4 (YAL029C), SNC2 (YOR327C) / MYO2 (YOR326W) or ALY2 (YJL084C) / EXO70 (YJL085W). This striking observation, which cannot have escaped the authors, is never mentioned nor discussed, which, I think, is a pity. These genes are most likely without any functional link with the bait mRNAs. Moreover, they are indifferently in tandem, divergent of convergent orientations, which preclude the fact that these co-purifications might result from the existence of some kind of polycistronic mRNAs. The most likely explanation that comes to mind is thus that the RaPID procedure is able to enrich not only MRNP particles but also chromatin fragments with their associated nascent transcripts.

I think that this is important to discuss because it suggests that a substantial fraction of the transcripts enriched in the genomic experiment might actually represent such chromatin-associated transcripts. This observation would not be in conflict with the suggestion in the following sections of the manuscript that co-purified mRNAs are synthesized within "transcription factories" in which the genomic locations of the corresponding scattered genes are brought to vicinity.

Anyhow, this observation, which is readily apparent, should be discussed.

[Editors’ note: the authors resubmitted a revised version of the paper for consideration. What follows is the authors’ response to the first round of review.]

Reviewer #1:

In this manuscript, Nair et al.et al., propose that functionally related mRNAs can be packaged and transported in an operon-like manner. The discovery of such functional mRNP packages, termed transperons, could have wide-ranging implications for our understanding of the regulation of gene expression. Unfortunately, there are significant technical problems with the results, and while the manuscript contains a large amount of information, it lacks the required rigor, and key controls are missing. Most importantly, additional experiments are needed to demonstrate the existence of specific transperons.

We thank the reviewer for their comments and have revised the paper accordingly with new added experiments that validate our previous observations.

1. Introduction of aptamer-binding sites can lead to the accumulation of decay intermediate that accumulate in P-body-like structures (see results from the Parker and Weis labs 2017). It is therefore critical that the authors perform extensive FISH experiments with untagged alleles (using probes against the body of the mRNAs). In situs should be performed for the different groups of mRNAs involved in mating, heat shock etc. The experiments need to be carefully quantified, statistically evaluated and all experiments need to include negative controls (i.e. mRNAs that do not co-localize). FISH experiments should be performed in wt and in H4 mutants. Of note, it is important to differentiate between nuclear spots, which likely present sites of transcription, and cytoplasmic spots. Only significant cytoplasmic-localization would lend further support for the transperon hypothesis.

We appreciate the reviewer’s comment and, per their suggestion, performed smFISH experiments (with statistical analyses) on strains expressing the untagged endogenous alleles in both WT and in H4 deletion mutants, and using specific probes against the mRNAs (Figures 3A, 8B, S3E, and S8C) along with different negative controls. The results fully support our previous observations made using the endogenously expressed aptamer-tagged mRNAs. RNA colocalization was observed both in the nucleus and cytoplasm, further supporting our hypothesis of the existence of these RNA multiplexes/transperons.

2. The RAPID results lack a statistical evaluation. Not all interactions 'look' convincing, are these results significant?

Since this is a critical observation of our paper, and taking into consideration the reviewer’s comment, we re-performed the RaPID pulldown experiments (shown previously in the original Figure 2A and B) using both MATa and MATalpha yeast strains and now used qRT-PCR to obtain quantitative results regarding the pulldown of the mating pathway mRNAs. These are shown in the new Figure 2A and B. We found that the results matched our original RT-PCR results and show that the quantitative coprecipitation of mating RNAs is significant compared to the negative controls (Figure 2, and B). We note that we previously used qRT-PCR to demonstrate the multiplexing of the HSP, MOMP, MAPK RNAs (Figures 8A, 9A and D, S8A and B) in the original version of the manuscript.

3. The fact that deletion of DHH1 or PAT1 does not affect mating is not an argument against (1). However, the authors should colocalize Dhh1 and Pat1 in their aptamer-tagged strains. Note that this experiment does not replace the need of the FISH experiments described in (1).

As per the reviewer’s suggestion we looked for the colocalization of Dhh1 protein and MS2-tagged STE2 mRNA under conditions of normal growth (and mating), as compared to conditions that induce P-body formation (i.e. no glucose – 10min; Figure S5C). Under normal growth conditions (2% glucose), we found that only ~2% of cells had Dhh1-mCherry-labeled puncta and which did not colocalize with the mating pathway mRNA examined (STE2). As a positive control for P-body formation, we exposed the cells to no glucose conditions for 10min, which resulted in all cells having well-formed Dhh1-mCherry puncta (P-bodies). Under these conditions, we now observed around 41% colocalization of puncta colocalizing with STE2 mRNA (Figure S5C). Our results confirm the idea that under conditions conducive to normal growth or mating (i.e. in the presence of glucose) that mating pathway mRNAs do not colocalize with P-bodies (if even formed) and, therefore, conclude that P-bodies cannot be a major site for mating pathway signaling events. Thus, our work contrasts with the earlier study of Aronov et al.et al. 2015 (JCB) in a number of critical points.

4. The nuclear co-localization experiments with Hhf1 protein and the STE2 mRNA need negative controls. Hhf1 seems to fill the entire nucleus and thus is likely to "interact' with every mRNA that would be tested in such a way.

We concur with the reviewer in that an mRNA in the nucleus is likely to colocalize with Hhf1 protein because of the high concentration of protein therein. The point we endeavored to show is that we cannot observe colocalization between RFP-Hhf1 with the mRNAs in the cytoplasmic transperons, either due to lack of RFP-Hhf1 presence in the cytoplasm or that the level in transperons is below the threshold of detection. Since the experiment is inconclusive, we have removed this figure in order to prevent any confusion.

Reviewer #3:

The observations presented are potentially interesting. But many of the data presented do not match, in my opinion, the quality standards that one would expect for such experiments in 2020.

We thank the reviewer for their comments and have revised the paper with new experiments that validate our previous observations.

A few examples. Only one RaPID-seq experiment is presented for each bait factor (Figure 1, S1 and Table S1). It is only stated "The experiment was performed twice and gave similar results"…… The standard for this type of experiment is rather to perform the experiments in triplicates and perform statistical analyses with programs such as DeSEQ2. The authors say that the replicate gives "similar results", why not provide them in Table S1 then?

Since the RaPID-seq experiments performed were similar, but non-identical (i.e. in that they did not use all the same baits), we did not put them in the table nor perform statistical comparisons. We have now removed this sentence (i.e. about the experiment being performed twice) from the text in order to prevent any confusion. We concur with the reviewer that if these experiments would be performed today they would be done differently, with replicates, and under different conditions of stringency in order to identify (and not eliminate) other potential multiplexes. We are in the process of setting this up and, in parallel, will perform quality Hi-C experiments for the yeast genome under different conditions, since allelic coupling appears to be part of the mechanism leading to RNA multiplexing. In this revised version of the paper we now validate the existence of mating pathway RNA multiplexes (as observed in Figure 1A and B for MATalpha cells), in both MATa and MATalpha yeast using RaPID and qRT-PCR (Figure 2A and B). These results verify the existence of the multiplex shown in Figure 1 as well as the original RT-PCR results obtained via RaPID pulldowns in the original Figure 2. In addition, we now add smFISH experiments (new Figure 3A and B) that also validate our original finding that aptamer-tagged mating pathway mRNAs colocalize (Figure 2D-F). Together with the original data, our new experimentation further validates the hypothesis that these mRNAs form functional multiplexes in yeast.

Figure 2 A and B provide the data for the key experiment, which is at the basis of the whole manuscript. One should expect that this would be very robust data. In fact, this is not the case at all. The individual mRNA enrichments are not quantified by qRT-PCR, as one would expect, but by a single ethidium bromide stained agarose gel. Even worse, I am actually unable to see what the authors claim, i.e. a signal for the STE3, SRO7, OM45 and EXO70 PCR products in the IP fractions. The only signals one can see are for MFalpha1 and MFalpha2, but the signals in the input fraction are much more abundant for those genes that the signals seen for the control transcripts, making it impossible to say if these mRNAs are actually truly enriched. The control used for Figure A is the signal from the input, while for B, the only control provided is "-RT". How one would expect to get an idea of an enrichment level without the input data and the proper control mRNAs. Frankly, I don't understand. These experiments should be analyzed by qRT-PCR with the proper replicates and control mRNAs to be able to quantify enrichment values and calculate standard deviations. These remarks stand for Figure 3AandB, Figure 4C etc.

We concur with the reviewer, and as multiplexing is a critical observation, we reperformed the RaPID pulldowns of select target mRNAs from both MATa and MATalpha cells and quantitated the results by qRT-PCR (versus a wild-type strain expressing the coat protein alone). The results verify our previous RT-PCR experiments (now removed) showing quantitative and statistically significant results regarding the specific coprecipitation of mating RNAs, as compared to negative controls (new Figure 2A and B).

For the Fish experiments, some field view, with several cells, should be provided in the supplemental data.

We have included Author response image 1 and Author response image 2 field views of some experiments, as requested. We did not include these in the manuscript, so as not to further increase its length. As can be seen, we observed similar results in all cells scored and the representative cells presented in the figures are accurate.

Author response image 1

Download asset Open asset

Author response image 2

Download asset Open asset

Figure 6A: "Mating mRNAs and histone H4 co-localize only in the nucleus." I don't understand the meaning of this. Hhf1 localizes to the whole nucleus (see the yeast-GFP localization data base). Hence, in this type of experiment, Hhf1 will merely localize the nucleus (pretty much as dapy staining). How can one say that an mRNA fish signal, appearing as a dot, "co-localize" with Hhf1. It only says that the nuclear fraction of this mRNA is…… nuclear!

We concur with the reviewer’s comment. We endeavored to show whether Hhf1 could colocalize with a mating mRNA in the cytoplasm, but were unable to do so, perhaps, due to the low intensity of the signal in the cytoplasm. Given that the experiment was inconclusive in this regard we have now removed it from the manuscript. Obviously, colocalization in the nucleus was pretty much a given.

I will stop here. I think it is pretty clear that the authors should repeat properly quite a number of experiments before one would consider publishing this work.

We thank the reviewer for the critique and have taken their comments (as well as those of the other reviewers) to heart, and have reperformed a number of experiments using different techniques (e.g. qRT-PCR, smFISH). Importantly, using all manners of approach we could validate our previous findings that show the formation of these mRNA complexes, their presence in the cytoplasm, and their importance to cell physiology.

[Editors’ note: what follows is the authors’ response to the second round of review.]

Essential Revisions:

The reviewers are concerned about the reproducibility of some of the experiments that are shown in the paper. The concerns are described in their detailed comments below but key issues include:

1. Figure 2

Given the high variability the results remain difficult to interpret.

As one of the original pulldown experiments showed variability, we have repeated this RaPID experiment and added the data to a new graphic presented in the Figure 2A and B. This new figure shows improved significance in the pulldown results for the mating pathway mRNAs and supports the idea that each yeast haplotype creates its own mating RNA multiplex.

2. AID experiments

The AID experiments are problematic as the mRNA levels go down and tagging alone in the absence of auxin leads to a protein reduction.

We apologize for any confusion regarding the AID experiments. As shown previously in Figure S3D (now referred as Figure 4—figure supplement 1D), the mRNA levels for both STE2 and AGA1 do not go down upon auxin addition (either with or without α factor). In fact, the levels for STE2 mRNA do not change, while those of AGA1 actually go up a bit. Thus, the depletion in histone H4 does not greatly affect the amounts of the mating pathway mRNAs. Therefore, the reduced colocalization of STE2 and AGA1 mRNA (Figure S3E; now referred as Figure 4—figure supplement 1E) is independent of RNA levels.

In new Figure 8—figure supplement 2D, we now show that upon histone depletion the mRNA levels of HSP104 and SSA2 do not go down, either with or without heat shock. This shows that the reduced amount of HSP104 and SSA2 mRNA colocalization (S8C; now referred as Figure 8—figure supplement 2C) is also independent of RNA levels. We also note that AID tagging alone does not lead to a reduction in histone H4 (Hhf1), as shown previously in Figure S3B; now referred to as Figure 4—figure supplement 4B.

3. Colocalization microscopy data

The mRNA localization results vary dramatically between different experiments, which makes it difficult to interpret these results.

We concur that some variability in the number of fluorescent labeled spots/foci was observed in the different smFISH experiments. However, it is important to note these results did not change our overall conclusions regarding RNA colocalization. Nevertheless, we have now provided a new smFISH experiment (new Figure 2E and F) for STE2 and AGA2 mRNA using probes to native STE2 and PP7 aptamer-based probes for AGA2, which show greatly increased numbers of both RNA spots (foci) and colocalized RNAs in the nucleus and in the cytoplasm. This replaces the original and much older Figure 2E and F, which differed in spot number from the other smFISH

Reviewer #1:

In this revised paper, the authors have added additional controls and statistical analyses. This has improved the quality of the manuscript. However, there are still significant technical issues.

We thank the reviewer for their helpful comments and suggestions, and have addressed the issues raised.

1. Figure 2A/B

Based on the figure legend it remains unclear what the authors call significant versus not significant?

We concur with the reviewer’s comment and now better illustrate the significance, statistical tests done, and number of experiments performed in the figure and legend of Figure 2A and B. We apologize for their omission in the previous version of the manuscript. In addition, we have added new data that shows increased statistical significance of the mating RNAs pulled down by the target RNA.

2. Figure 2, 3, 8, 9

The mRNA localization results vary dramatically between different experiments (e.g., STE2 foci in different conditions). Why? Irrespectively, this makes it very hard to compare the experiments and the controls. Furthermore, the signal overlap in hs between SSA2 and HSP104 are not convincing. It could be a consequence of the enhance HSP104 expression.

We concur with the reviewer’s comment regarding variability (in the number of fluorescent-labeled foci) used to score colocalization. However, it is important to note these results did not change our overall conclusions regarding RNA colocalization. Nevertheless, we repeated the smFISH experiment in Figure 2E and F now using probes for endogenous STE2 mRNA and PP7 probes for AGA2 mRNA, and observed much greater numbers of RNA foci and colocalization. The inclusion of this new data greatly reduces the variability between Figure 2E and F and Figures 3A and B and 8B and C.

Regarding the greater level of RNA colocalization between SSA2 and HSP104 upon heat shock (Figure 8B and C), we concur that this could result due to enhanced gene expression, as we now show to occur Figure 8—figure supplement 2D. However, we also showed previously (Figure 8A) that heat shock increases SSA2 and HSP104 multiplexing. Moreover, we show in Figure 8—figure supplement 2C that colocalization is dependent on histone H4 levels, which is independent of mRNA levels as shown in Figure 8—figure supplement 2D. Therefore, the increased colocalization is most likely a consequence of increased multiplexing and not increased gene expression.

3. The AID experiments in hs are not convincing. mRNA levels go down which might be sufficient to reduce co-localization (see 2). AID-tagging alone (without auxin) also leads to a reduction, which is potentially problematic.

As shown previously in Figure S3D (now referred as Figure 4—figure supplement 1D), the mRNA levels for both STE2 and AGA1 do not go down upon auxin addition (either with or without α factor). In fact, the levels for STE2 mRNA do not change, while those of AGA1 actually go up a bit. Thus, the auxin-induced depletion of histone H4 does not greatly affect the levels of the mating pathway mRNAs and, therefore, the reduced colocalization of STE2 and AGA1 mRNAs (Figure S3E; now referred as Figure 4—figure supplement 1E) is independent of RNA levels. In Figure 8—figure supplement 2D, we now show that the mRNA levels of HSP104 and SSA2 also do not go down upon histone depletion, either with or without heat shock. This shows that the reduced amounts of HSP104 and SSA2 mRNA colocalization (S8C; now referred as Figure 8—figure supplement 2C) are also likely to be independent of RNA levels.

As shown in Figure 8—figure supplement 2D, AID tagging alone (e.g. without auxin) does not affect the levels of HSP104 and SSA2 mRNA. Likewise AID tagging (e.g. with α factor) did not affect STE2 mRNA levels, but slightly increased those of AGA1 slightly (Figure S3D; now referred as Figure 4—figure supplement 1D). Thus, AID tagging did not affect gene expression in the case of either transperon. We also showed previously that AID tagging alone did not affect Hhf1 protein levels (Figure S3B; now referred as Figure 4—figure supplement 1B). We did observe a decrease in RNA colocalization in the tagged strains, but this more likely reflects the effect of the single histone H4 deletion (hhf2delta), which we have demonstrated to affect RNA multiplexing (Figure 4C and D).

4. Figure 9 (A-C) is also problematic since no convincing negative controls are included.

As per the reviewer’s suggestion we have now included negative controls (e.g. HSP104 vs. TOM6 and HSP104 vs. FUS3 RNA colocalization) for the mitochondrial outer membrane and MAPK pathway transperons, respectively. Our results fully support our previous observations regarding the existence of these RNA multiplexes/transperons.

Reviewer #2:

In this manuscript the authors present a detailed set of experiments to describe and mechanistically characterize potential multiplexing of functionally related mRNAs. The focus is mainly the mating pathway mRNAs and the physiology of the mating pathway in yeast. The work supports models where the fate of functionally related mRNAs is coordinated and the authors show that this coordination is first seeded at the site of mRNA transcription which may involve a histone protein. The mechanistic advances supporting colocalized mRNA biogenesis at the site of transcription in the nucleus leading to a coordination of the subsequent cytosolic fate of this set of mRNAs are exciting and particularly noteworthy.

The authors present a revised version of a manuscript describing mRNA multiplexing and going on to study the mechanism of mRNA localization and physiological role of this process. The manuscript is very much improved over the previous version, but I do have some further comments for the authors – detailed below.

We thank the reviewer for their helpful comments and suggestions, and have addressed the issues raised.

1. I wasn't convinced that the IST2 mRNA was significantly enriched in the pulldown in

Figure 2A relative to Figure 2B- as claimed by the authors. Plus many of the error bars On Figure 2B are very high across this analysis- so it is difficult to interpret.

As one of the original pulldown experiments showed variability, we repeated this RaPID experiment and have added the data to a new graphic presented in the Figure 2A and B. This new figure shows improved significance in the pulldown results for the mating pathway mRNAs and supports the idea that each yeast haplotype creates its own mating RNA multiplex. We now better describe the significance, statistical tests performed, and number of experiments performed in the figure and legend of Figure 2A and B. With regards to IST2, we could only observe significant enrichment in the SAG1 pulldown from MATalpha cells. Since we have not done any other experiments with IST2 this result is inconclusive at this point.

2. Are the images in Figure 7C from a single focal plane? I'm guessing so since some of the cells don't have a spot in their nucleus. If this is the case the authors should add this information to the legend to prevent confusion.

For all microscopy figures a single focal plane is shown. We concur with the reviewer and now mention this in the various figure legends.

3. Have the authors tested the expression levels of the Hhf1K-R and K-Q- as these mutations a differential effect on protein stability could explain the observations?

We concur with reviewer’s suggestion and performed qRT-PCR to examine the levels of the endogenously expressed HHF11K-R mutant used in Figure 5C. We observed no statistical difference in HHF1 mRNA levels between the WT and HHF1K-R mutant strains, and now include mention of this observation in the Results section. We did not examine protein levels (due to a lack of specific antibody), so it’s possible that protein instability could also contribute to the effect. A statement to this effect has been added to the text. As for the HHF1K-Q mutation (used in Figure 5B), this gene was expressed using a multicopy plasmid, but we did not examine protein stability.

4. Similar to above- have the authors measured the RNA levels for AGA2 47-MUT or AGA2 27-MUT relative to wild type, as effects on RNA stability or expression could explain the observed effects.

We performed qRT-PCR to measure AGA2 47-MUT or AGA2 27-MUT RNA levels in mutant backgrounds (used in Figure 6B and C) and observed no statistical difference between WT AGA2 and the AGA2 47-MUT mutant strain, but did see an increased expression of AGA2 27-Mut. This may explain why this mutant had a lesser effect on multiplexing and mating. A note regarding this observation has now been added to the Results section.

Reviewer #3:

In this novel version of the manuscript, the authors have, I think, satisfactorily answered to most reviewer concerns about the first version.

We thank the reviewer for their helpful comment and have addressed the issue raised.

Considering the suggestion to provide the results of the replicate of the Rapid-RNAseq experiment, the authors rather preferred to remove all mentions of it. Considering that all analyzed and discussed results of the now single documented experiment are validated by independent RaPID followed by qRT-PCR experiments, one can consider this as acceptable.

I have one suggestion, thought. When examining Figure 1 and the associated Table S1, one can see that some of the transcripts most highly enriched together with the bait MS2 tagged mRNAs are actually the mRNAs from the directly adjacent genes. This is true for PET122 (YER153C) pulled down by OXA1 (YER154W), or SPE1 (YKL184W) / ASH1 (YKL185W), SNC1 (YAL030W) / MYO4 (YAL029C), SNC2 (YOR327C) / MYO2 (YOR326W) or ALY2 (YJL084C) / EXO70 (YJL085W). This striking observation, which cannot have escaped the authors, is never mentioned nor discussed, which, I think, is a pity. These genes are most likely without any functional link with the bait mRNAs. Moreover, they are indifferently in tandem, divergent of convergent orientations, which preclude the fact that these co-purifications might result from the existence of some kind of polycistronic mRNAs. The most likely explanation that comes to mind is thus that the RaPID procedure is able to enrich not only MRNP particles but also chromatin fragments with their associated nascent transcripts.

I think that this is important to discuss because it suggests that a substantial fraction of the transcripts enriched in the genomic experiment might actually represent such chromatin-associated transcripts. This observation would not be in conflict with the suggestion in the following sections of the manuscript that co-purified mRNAs are synthesized within "transcription factories" in which the genomic locations of the corresponding scattered genes are brought to vicinity.

Anyhow, this observation, which is readily apparent, should be discussed.

We duly appreciated the reviewer’s suggestion. This was indeed observed, but since we had not examined the interactions could not prove or disprove the findings. However, we agree with the reviewer’s suggestion and explanation, and briefly discuss the issue in the Discussion section.

Author details

1. Rohini R Nair

   Department of Molecular Genetics, Weizmann Institute of Science, Rehovot, Israel

   Contribution

   Conceptualization, Data curation, Formal analysis, Validation, Investigation, Visualization, Methodology, Writing - original draft, Writing - review and editing

   Contributed equally with

   Dmitry Zabezhinsky

   Competing interests

   No competing interests declared

2. Dmitry Zabezhinsky

   Department of Molecular Genetics, Weizmann Institute of Science, Rehovot, Israel

   Contribution

   Conceptualization, Data curation, Formal analysis, Investigation, Visualization, Methodology, Writing - review and editing

   Contributed equally with

   Rohini R Nair

   Competing interests

   No competing interests declared

3. Rita Gelin-Licht

   Department of Molecular Genetics, Weizmann Institute of Science, Rehovot, Israel

   Contribution

   Data curation, Formal analysis, Investigation

   Competing interests

   No competing interests declared

4. Brian J Haas

   Broad Institute of MIT and Harvard, Cambridge, United States

   Contribution

   Data curation, Software, Formal analysis, Writing - original draft

   Competing interests

   No competing interests declared

5. Michael CA Dyhr

   Department of Molecular Genetics, Weizmann Institute of Science, Rehovot, Israel

   Contribution

   Formal analysis, Investigation

   Competing interests

   No competing interests declared

6. Hannah S Sperber

   Department of Molecular Genetics, Weizmann Institute of Science, Rehovot, Israel

   Contribution

   Data curation, Formal analysis

   Competing interests

   No competing interests declared

7. Chad Nusbaum

   Broad Institute of MIT and Harvard, Cambridge, United States

   Present address

   Cellarity Inc, Cambridge, United States

   Contribution

   Resources, Funding acquisition, Writing - original draft

   Competing interests

   Chad Nusbaum is affiliated with Cellarity Inc. The author has no financial interests to declare.

8. Jeffrey E Gerst

   Department of Molecular Genetics, Weizmann Institute of Science, Rehovot, Israel

   Contribution

   Conceptualization, Supervision, Funding acquisition, Visualization, Writing - original draft, Project administration, Writing - review and editing

   For correspondence

   jeffrey.gerst@weizmann.ac.il

   Competing interests

   No competing interests declared

   "This ORCID iD identifies the author of this article:" 0000-0002-8411-6881

• 2,061

  Page views

• 274

  Downloads

• 6

  Citations

Article citation count generated by polling the highest count across the following sources: Crossref, PubMed Central, Scopus.