Identification of proteins that form specific ...

Biochimica et Biophysica Acta 1844 (2014) 767–777

Contents lists available at ScienceDirect

Biochimica et Biophysica Acta journal homepage: www.elsevier.com/locate/bbapap

Identification of proteins that form specific complexes with the highly conserved protein Translin in Schizosaccharomyces pombe Elad Eliahoo a,1, Phyana Litovco a, Ron Ben Yosef a, Keren Bendalak b, Tamar Ziv b, Haim Manor a,⁎ a b

Department of Biology, Technion—Israel Institute of Technology, Haifa, Israel Smoler Proteomics Center, Technion—Israel Institute of Technology, Haifa, Israel

a r t i c l e

i n f o

Article history: Received 23 October 2013 Received in revised form 18 December 2013 Accepted 22 December 2013 Available online 29 December 2013 Keywords: Protein–protein interactions RNA-binding proteins Mass spectrometry Affinity purification SR proteins

a b s t r a c t Translin is a single-stranded DNA and RNA binding protein that has a high affinity for G-rich sequences. TRAX is a Translin paralog that associates with Translin. Both Translin and TRAX were highly conserved in eukaryotes. The nucleic acid binding form of Translin is a barrel-shaped homo-octamer. A Translin–TRAX hetero-octamer having a similar structure also binds nucleic acids. Previous reports suggested that Translin may be involved in chromosomal translocations, telomere metabolism and the control of mRNA transport and translation. More recent studies have indicated that Translin–TRAX hetero-octamers are involved in RNA silencing. To gain a further insight into the functions of Translin, we have undertaken to systematically search for proteins with which it forms specific complexes in living cells. Here we report the results of such a search conducted in the fission yeast Schizosaccharomyces pombe, a suitable model system. This search was carried out by affinity purification and immuno-precipitation techniques, combined with differential labeling of the intracellular proteins with the stable isotopes 15N and 14N. We identified for the first time two proteins containing an RNA Recognition Motif (RRM), which are specifically associated with the yeast Translin: (1) the pre-mRNA-splicing factor srp1 that belongs to the highly conserved SR family of proteins and (2) vip1, a protein conserved in fungi. Our data also support the presence of RNA in these intracellular complexes. Our experimental approach should be generally applicable to studies of weak intracellular protein–protein interactions and provides a clear distinction between false positive vs. truly interacting proteins. © 2013 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/3.0/).

1. Introduction Translin is an RNA and single-stranded DNA (ssDNA) binding protein, which has a high affinity for G-rich sequences [1,2]. It was first discovered in extracts of human and mouse cells and found to constitute a multimeric complex [3–6]. TRAX (TRanslin Associated protein X) [7] is a Translin paralog which does not bind nucleic acids. However, it forms specific complexes with Translin and these complexes also bind nucleic acids [8,9]. Both Translin and TRAX are highly conserved in many higher and lower eukaryotes, including the fission yeast Schizosaccharomyces pombe, but not the budding yeast Saccharomyces cerevisiae [1,2,10]. TRAX-like proteins are also found in some prokaryotes and in archaea [10,11]. Detailed biochemical studies of recombinant Translins from several organisms indicated that they are assembled into homo-octamers, and that these octamers are the configurations that bind nucleic acids

⁎ Corresponding author. Tel.: +972 4 8294229; fax: +972 4 8225153. E-mail address: [email protected] (H. Manor). 1 Present address: Department of Laboratory Medicine, University of Washington, Seattle, Washington, USA.

[12–14]. The 3D structures of the human and mouse Translins were solved by X-ray crystallography at resolutions of 2.65 Å and 2.2 Å, respectively, and were found to be virtually identical [15,16]. The crystals had a tetramer as an asymmetric unit, but an octameric barrel structure is generated by rotation of the tetramer about a 2-fold crystallographic axis [15,16]. The individual Translin subunits consist of 228 amino acids and contain seven α helices that constitute N70% of the amino acid residues. More recent studies confirmed that the octameric barrel is the structure that exists in solution [17]. Furthermore, although the barrel structure deduced from the crystallographic data is quite compact and does not appear to be capable of binding nucleic acids in solution [15,16], the octameric barrel undergoes conformational transitions into open structures that can bind nucleic acids [17]. Additionally, it was found that the N-terminus and the C-terminus of each Translin subunit, which are exposed to the solvent, play a crucial role in their interactions within the octamers and in their ability to interact with other proteins [18]. Translin is found in many mammalian tissues [1]. Based on its relatively high affinity for some ssDNA sequences, it has been proposed to play a role in chromosomal translocations and in telomere metabolism [6,19]. It has also been suggested that Translin might be involved in processes affecting mitotic cell division [20]. The RNA binding characteristics

http://dx.doi.org/10.1016/j.bbapap.2013.12.016 1570-9639/© 2013 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/3.0/).

768

E. Eliahoo et al. / Biochimica et Biophysica Acta 1844 (2014) 767–777

of Translin and Translin–TRAX complexes, led to the suggestion that both proteins may be involved in the control of mRNA translation [21–23]. More recent studies have indicated that in Drosophila and in mammals hetero-octamers consisting of Translin and TRAX subunits are involved in RNA silencing [24–28]. In particular, the TRAX subunits in these hetero-octamers were reported to have a ribonuclease activity that appears to activate RISC complexes by removing siRNA passenger strand cleavage products and thereby promoting the activity of the RNAi machinery [25,26]. It has also been reported that complexes including Translin, or both Translin and TRAX, are involved in transport of specific mRNAs along dendrites in mammalian neurons [29]. Further evidence for the involvement of Translin in neuronal functions was obtained in studies of mice, in which the Translin gene has been deleted. These mice exhibit sex-specific phenotypic variations in neurological and behavioral tests [30]. Based on the studies described above, it appears that Translin and TRAX are multifunctional proteins involved in various biochemical pathways. In living cells they might perform the various functions in the context of complexes containing either Translin homo-octamers, or Translin–TRAX hetero-octamers, which are associated with other proteins and nucleic acids. We have chosen to use the fission yeast S. pombe, a suitable model system, for biochemical and genetic studies aimed at identification and characterization of such complexes. Previously, we have characterized the interactions of the S. pombe Translin (spTranslin) with nucleic acids and with the protein spTRAX [14,18]. In this article we describe studies in which we searched for additional proteins that associate with spTranslin in the fission yeast cells. We used for this search affinity purification and immuno-precipitation techniques, which yielded a large number (N 75) of proteins in the purified fractions. However, many of these proteins were also recovered in the purified fractions of control extracts obtained from cells that do not synthesize Translin. Therefore, we used differential labeling of the intracellular proteins in the experimental and control cultures with the stable “heavy” and “light” isotopes 15N and 14N to further distinguish between truly associated proteins and “false positive” background proteins. Our improved search protocols revealed for the first time that, in addition to spTRAX, two proteins designated pre-mRNA-splicing factor srp1 and vip1, also associate with spTranslin in the yeast cells, albeit much less tightly than spTRAX. Both proteins contain an RNA Recognition Motif (RRM). Our data also provide support for the presence of RNA in the intracellular complexes including spTranslin and the protein srp1.

150 mM NaCl, 0.1% Igepal. TEV buffer—10 mM Tris–HCl, pH 8.0, 150 mM NaCl, 0.1% Igepal, 0.5 mM EDTA, 1.0 mM DTT. Calmodulin binding buffer—10 mM Tris–HCl, pH 8.0, 150 mM NaCl, 1 mM Mgacetate, 1 mM imidazole, 2 mM CaCl2, 10 mM β-mercaptoethanol, 0.1% Igepal. Calmodulin elution buffer—10 mM Tris–HCl, pH 8.0, 150 mM NaCl, 1 mM Mg-acetate, 1 mM imidazole, 20 mM EGTA, 10 mM β-mercaptoethanol, 0.02% Igepal.

2. Materials and methods

2.6. Immunoprecipitation of spTranslin and associated proteins from S. pombe cell extracts

2.3. S. pombe culture procedures The S. pombe culture procedures described by Laufman et al. were used [14].

2.4. Expression and purification of TAP-tagged spTranslin and associated proteins The preparation of S. pombe cell extracts and the two-step TAP-tag affinity purification procedure were performed as described by Eliahoo et al. [18] (also see the supplementary material), and were based on the protocols of Gould and coworkers [32,35].

2.5. Coupling of polyclonal anti-spTranslin antibodies to protein A beads 300 μl of suspended protein A Sepharose beads (Sigma) were centrifuged down in the Eppendorf microfuge. The beads were resuspended in 1 ml of PBS buffer, centrifuged and resuspended again in 2 ml of PBS buffer containing affinity-purified anti-spTranslin antibodies at a concentration of 25 μg/ml. The suspension was slowly agitated 1 h at 25 °C and the beads were centrifuged again. Next, the beads were resuspended in 1.5 ml of 0.20 M sodium borate (pH 9.0); then, dimethyl pimelimidate (DMP) was added to a final concentration of 20 mM and the suspension was agitated 35 min at 25 °C. The coupling reaction was quenched by addition of 1.0 ml ethanol amine (pH 8.0) and agitation was resumed for 2.0 h at 25 °C. To remove uncoupled antibodies, the beads were washed by two cycles of resuspension and centrifugation, first in 1.0 ml 0.10 M glycine pH 3.0, then in 1.0 ml PBS buffer. Finally, the beads were resuspended in 1.0 ml PBS buffer that also contained 2 mg/ml of Bovine Serum Albumin (BSA) and 0.02% Naazide and the suspension was stored at 4 °C.

2.1. Plasmids The following plasmids were constructed and used in this study: the plasmids pREP41-TSN and pREP81X-TSN, both expressing spTranslin, were constructed by cloning the spTranslin open reading frame into the plasmids pREP41 and pREP81, respectively [31]. The plasmid pREP41-TAP-TSN expressing the spTranslin linked to the TAP-tag at the N-terminus was constructed by cloning the spTranslin open reading frame into the plasmid pREP41-NTAP [32]. The plasmid pSLF172-Vip1 expressing the HA-tagged vip1 protein was constructed by cloning the vip1 open reading frame into the plasmid pSLF172 [33]. We also used for this study the plasmid pRH42S expressing the HA-tagged srp1 [34]. In all plasmids the proteins were expressed under the control of the nmt1 promoter. 2.2. Buffers PBS buffer—137 mM NaCl, 2.7 mM KCl, 10 mM Na2HPO4, 1.8 mM KH2PO4, pH 7.2. NP-40 lysis buffer—6 mM Na2PO4, 4 mM NaH2PO4·H20, 150 mM NaCl, 2 mM EDTA, 50 mM NaF, 0.1 mM Na3VO4, 1% Igepal CA-630 (Sigma-Aldrich). IPP150 buffer—10 mM Tris–HCl, pH 8.0,

75 μl of protein A beads coupled to anti-spTranslin antibody were mixed with 2.0 ml of cell extract. The suspension was slowly swirled 2 h at 4 °C, the beads were centrifuged down and the supernatant was removed. The beads were then resuspended in 1 ml of NP-40 lysis buffer to which were added 120 units/ml ribonuclease inhibitor (porcine liver) and 0.05 mg/ml heparin, as well as the protease inhibitors phenylmethylsulfonyl fluoride (PMSF; 1 mM), leupeptin (4 μg/ml) and one Complete Mini EDTA-free protease inhibitor cocktail tablet (Roche). The beads were then centrifuged, resuspended in 1.0 ml PBS buffer and centrifuged again. This was followed by 5 similar cycles of resuspension in 1.0 ml PBS buffer and centrifugation. The bound spTranslin and associated proteins were eluted by resuspension of the beads in 200 μl of 0.10 M glycine, pH 2.5. The suspension was incubated 5 min at 25 °C and the beads were centrifuged down. The supernatant was collected and brought to pH 8.0 by addition of 1 M Trizma base.

2.7. Western blot analysis Western blot analyses were performed as previously described [14].


2.8. Analysis of the affinity-purified, or the immuno-purified proteins by mass-spectrometry

769

3. Results 3.1. Identification of proteins that associate with spTranslin in S. pombe cells

The volume of the purified proteins was reduced by evaporation to 40 μl and the proteins were resolved by electrophoresis in SDS-PAGE (6% polyacrylamide). The gels were stained with Coomassie Blue, then were destained and each lane was cut into three pieces. The pieces were incubated with trypsin and the resulting peptides were processed and analyzed by mass-spectrometry in an ion-trap mass spectrometer (Orbitrap, Orbitrap XL; Thermo), as previously described [36]. 2.9. Calculation of the normalized 15N/14N ratios for proteins identified by mass spectrometry For each protein, let: a = No. of MS2 peptides b = Median of the 15N/14N ratios of the MS2 peptides c = No. of MS1 peptides d = Median of the 15N/14N ratios of the MS1 peptides x = Weighted average of 15N/14N ratios for MS2 and m MS1 peptides y = Median of the 15N/14N ratios of all the recovered proteins n = Normalized weighted average of 15N/14N ratios for MS2 and MS1 peptides. Then, x = (a · b + c · d) / (a + c) n = x / y. Note that if a = 0, then x = d and n = d / y; if c = 0, then x = b and n = b / y.

A

3.1.1. The TAP-tag affinity purification approach To carry out this part of the project, we constructed a line of S. pombe cells expressing wild type (WT) spTranslin linked to a TAP-tag through its N-terminus by transfecting into the cells the expression plasmid shown in Fig. S1A in the supplementary material. We also constructed another line of S. pombe cells expressing TAP-tag alone by transfecting into the cells a plasmid expressing just the tag. Fig. 1A shows a scheme of our experimental procedure. In the first step of this procedure we grew cells of the line expressing TAP-tag alone, or a TAP-tagged spTranslin variant, for at least five generations in the presence of 15 NH4Cl, such that the vast majority of the intracellular protein molecules were uniformly labeled with the “heavy” 15N isotope. In parallel, cells expressing the TAP-tagged WT spTranslin were grown in the presence of the “light” 14NH4Cl. Next, the cells labeled with the 15N isotope were mixed with the cells labeled with the 14N isotope in a 1:1 ratio, and the mixed cells were disrupted. Then, the Tandem Affinity Purification protocol was employed to isolate the tagged spTranslin (or the tag alone) and associated proteins under physiological conditions [37] (see Fig. S1B in the supplementary material for details). Next, the proteins in the purified complexes were resolved by SDS-PAGE and were digested with trypsin. The resulting “heavy” and “light” peptide pairs were recovered and identified by mass spectrometry (MS). The MS analysis also provided the 15N/14N ratio for each peptide pair, from which data the 15 N/14N ratios of the corresponding proteins were deduced. As described below, determination of these ratios allowed ranking the proteins as to the likelihood of being true or false binding partners of spTranslin. We also carried out a “reciprocal” type of experiment, in which two such S. pombe cultures were similarly grown and differentially labeled, except that the isotopes were switched. This experiment

B

cells expressing TAP-tag cells expressing (TAP-tag)spTranslin alone, or (TAP-tag) labeled with N14 spTranslin variant, 15 labeled with N

the cells are combined

preparation of cell extracts by followed tandem affinity purification of proteins

C

cells expressing cells expressing no spTranslin untagged spTranslin labeledwith N15 labeledwith N14

the cells are combined

preparation of cell extracts followed by Immunoaffinity purification of proteins

cells expressing no spTranslin labeled with N15

cells expressing untagged spTranslin

preparation of cell extracts followed by Immuno-affinity purification of proteins

preparation of cell extracts followed by Immuno-affinity purification of proteins

labeled with N14

the purified proteins are combined

SDS-PAGE of the purified proteins followed by their cleavage with Trypsin

quantitative mass spectrometry of the peptides

calculations of the N15/N14 ratios of the peptides and the purified proteins Fig. 1. Schemes of experimental procedures for identification of proteins that are specifically associated with spTranslin in S. pombe cells. A. The TAP-tag affinity purification procedure. B. The immuno-affinity purification procedure. C. The modified immuno-affinity purification procedure. Detailed explanations of these procedures are presented in the text.

770


A kDa

1

2

3

4

5

6

7

8

250 150 75 50 37 25 20 15

B

kDa

1

2

3

4

5

6

250 150 100 75 50 37 25 20 Fig. 2. Western blot analyses of fractions collected during purification of the spTranslin complexes. A. Samples withdrawn from various fractions collected during the TAP-tag affinity purification were analyzed by western blotting. The blots were probed with polyclonal anti-spTranslin antibodies [14]. Lane 1: total cell extract (50 μg protein). Lane 2: flow through fraction (50 μg protein). Lane 3: wash with the buffer IPP150. Lane 4: wash with TEV buffer. Lane 5: TEV elution fraction. Lane 6: wash with TEV buffer. Lane 7: wash with calmodulin binding buffer. Lane 8: calmodulin elution fraction. B. Samples withdrawn from various fractions collected during the immuno-affinity purification were analyzed by western blotting, as described in A. Lane 1: total cell extract (50 μg protein). Lane 2: flow through fraction (50 μg protein). Lane 3: wash with the NP-40 lysis buffer. Lane 4: wash with PBS buffer. Lane 5: wash with PBS buffer. Lane 6: fraction eluted with 0.10 M glycine, pH 2.5.

provided an independent set of data allowing a similar ranking of the proteins. To test the TAP-tag affinity purification procedure, we carried out a western blot analysis of an extract obtained from S. pombe cells expressing the tagged spTranslin, and of fractions collected during affinity purification of the tagged protein from the extract (Fig. 2A). We used as a probe a polyclonal anti-spTranslin antibody to detect bands containing spTranslin in the blots (see Ref. [14] and the supplementary material for control experiments that prove the specificity of this antibody for spTranslin). As indicated, the crude extract contained a major band having a mobility of about 50 kDa, which is the expected mobility of spTranslin fused to the TAP-tag (lane 1). Other bands apparently contained degradation products of the protein. Most of the spTranslin was bound by the IgG coated beads (compare lane 1 to lanes 2–4). Cleavage of the bound protein with the protease TEV released a shorter protein consisting of spTranslin linked to the calmodulin binding peptide, which had the expected faster mobility (lane 5). This protein also displays some heterogeneity which remains after the second purification step (lanes 6–8). To examine the feasibility of the entire experimental scheme illustrated in Fig. 1A, we performed a control experiment in which two identical S. pombe cultures expressing TAP-tagged WT spTranslin were grown in parallel in media containing 15NH4Cl and 14NH4Cl, respectively. The cells of the two cultures were mixed before being disrupted in a single tube. Then, the tagged spTranslin and the associated proteins were purified from the crude extract, as described above. This procedure ensured that the efficiencies of protein extraction and purification were the same for the cells of the two cultures. To identify the proteins

recovered after the two-step affinity purification, we ran the purified fraction in SDS-PAGE and treated all the proteins in the gel with trypsin. The resulting peptides were eluted and analyzed by mass-spectrometry. Thus, the “heavy” and the “light” peptide pairs and the corresponding proteins recovered after the two-step affinity purification were identified and their 15N/14N ratios were determined (Table S1 in the supplementary material). It can be seen that the list of recovered proteins includes, in addition to spTranslin and spTRAX, also ribosomal proteins and other proteins associated with ribosomes, such as translation initiation factors, elongation factors and tRNA ligases. Another group of proteins found in this fraction included abundant enzymes involved in several key metabolic pathways, such as pyruvate kinase, glyceraldehyde-3-phosphate dehydrogenase, 6-phosphofructokinase, fatty acid synthase subunit beta and sulfite reductase [NADPH] subunit beta. In addition, we identified actin, some chaperones and proteasome subunits, as well as the premRNA-splicing factor srp1. Note that the proteins spTranslin, spTRAX and srp1 are specifically marked. Table S1 also shows the number of peptides from each protein that were identified, the average area of the peptide peaks, which is a measure of the recovered protein mass, and the isotopic ratio 15N/14N for the proteins calculated from the 15N/14N ratios of the peptides. As expected, the recovered mass of spTranslin was found to be the highest. The mass of spTRAX (which was not tagged) was also found to be exceptionally high, in line with previous observations of its strong association with spTranslin. The recovered masses of the other proteins were 2–3 orders of magnitude smaller. All the 15 N/14N ratios were smaller than 1.0 and their median value was 0.642. The latter observation indicated that a substantial amount of 14 N atoms were incorporated into proteins synthesized in the cells grown in the presence of 15NH4Cl. Most of the “light” nitrogen atoms in these proteins were probably derived from the adenine and/or the essential vitamin molecules that were included in the media. Table 1 (Exp. no. 1) presents the 15N/14N data collected in this experiment on the proteins spTranslin, spTRAX and srp1. It should be noted that the second stage of the MS/MS (MS2) analysis (the random scission of selected peptides and analysis of the cleaved products [38]) might have neglected to identify and analyze peptides whose yields were low. Clearly, this could be the case for the srp1 peptides whose yield was 2–3 orders of magnitude smaller than the yields of the spTranslin and spTRAX peptides. For this reason, we also performed a manual search for additional srp1 peptides (MS1 peptides in Table 1), that were not identified by the MS2 analysis. We used the MS1 peptide data, as well as the MS2 peptide data, for the calculations of the normalized 15N/14N ratios for srp1 presented in Table 1. It can be seen that in this experiment the normalized ratios for all three proteins varied between 1.076 and 1.313. The log2 values of the ratios of the three proteins are plotted in Fig. 3 (Exp. no. 1) and observed to be near the zero baseline of the graph. Next, we carried out an experiment, in which S. pombe cells expressing the TAP-tag alone were grown in the presence of 15NH4Cl and cells expressing the TAP-tagged WT spTranslin were grown in the presence of 14NH4Cl (Exp. no. 2). The proteins identified in this experiment in the purified fraction are listed in Table S2 in the supplementary material. Like the previous experiment, the list of proteins recovered at the end of the purification procedure included many ribosomal proteins and abundant enzymes, but the total number of recovered proteins was considerably smaller than the number found in the previous experiment. Such variability in the total number of proteins recovered in the purified fraction was typical for this series of experiments. As expected, the 15N/14N ratio in spTranslin (0.010) was very low compared to the median value of 0.58 for the whole protein population (Table S2). A similar ratio was also observed for the untagged spTRAX, thereby confirming its strong association with spTranslin. Other proteins, such as the 60S ribosomal protein L33-B and the heat shock factor sks2, as well as srp1, also had relatively low ratios, which could indicate their association with spTranslin. However, additional experiments only supported the srp1–spTranslin association (see below).


771

Table 1 Demonstration of specific association of the protein srp1 with spTranslin by the TAP-tag affinity purification procedure. Exp. S. pombe culturesa no.

20 20 5

spTranslin 2.483e9 spTRAX 1.192e8 srp1 7.672e6

14 14 4


10 10 1

(14N) TAP-tagged WT spTranslin + (15N) TAP-tagged y227 mutant spTranslin spTranslin 8.143e9 spTRAX 1.208e9 srp1 1.480e6

16 16 1

0.842 0.832 0.722 (0.641)b 0.01 0.01 0.114 (0.575)b 100 64.187 1.662 (0.625)b 0.01 0.01 0.252 (0.464)b 100 100 1.480 (0.567)b 0.569 0.334 0.275 (0.553)b 0.979 1.090 1.521 (0.531)b

14

15

( N) TAP-tag alone + ( N) TAP-tagged WT spTranslin

4

14

5

6

15

( N) TAP-tagged y227 mutant spTranslin + ( N) TAP-tagged WT spTranslin spTranslin 1.841e9 spTRAX 9.480e7 srp1 2.773e7

11 10 7

(14N) TAP-tagged WT spTranslin + (15N) TAP-tagged K106 mutant spTranslin spTranslin 8.416e8 spTRAX 3.651e7 srp1 5.249e6

18 13 5

14

7

c


(14N) TAP-tagged WT spTranslin + (15N) TAP-tag alone

3

b

b

15

( N) TAP-tagged WT spTranslin + ( N) TAP-tagged WT spTranslin

2

Peak area No. MS2 Median 15N/14N No. MS1 Median 15N/14N Normalized peptides of MS2 peptides peptides of MS1 peptides 15N/14N a

14

1

a

Protein recovered

15

( N) TAP-tagged K106 mutant spTranslin + ( N) TAP-tagged WT spTranslin spTranslin 5.009e9 spTRAX 1.790e8 srp1 2.517e7

21 21 10

c

1

1

5

3

1

–

–

d

nc

0.53

1.313 1.298 1.076

0.19

0.017 0.017 0.224

2.00

160 102.70 1.943

0.22

0.021 0.021 0.491

0.08

176.37 176.37 2.301

–

1.029 0.604 0.497

–

1.844 2.053 2.864

S. pombe cultures labeled with either the “heavy” or the “light” nitrogen isotope were mixed before cell disruption. In each experiment the median (y) of the 15N/14N ratios of all the recovered proteins is shown in parentheses. The calculation of n is described in the section of Materials and methods.

We also performed a “reciprocal” experiment with two similar cultures, in which the isotope labels have been switched (Exp. no. 3). The list of proteins identified in this experiment is included in Table S3 in 10 8

1

2

3

4

5

6

7

log215N/14N

6 4 2 0

spTranslin spTRAX srp1

-2 -4 -6 -8

Fig. 3. Graphical representation of quantitative MS data for Translin, TRAX and srp1 that 14 were purified by the TAP-tag procedure. Each group of three bars depicts the log15 2 N/ N ratios determined in an experiment performed with two cultures that were differentially labeled with the 15N and the 14N isotopes. Exp. 1: two identical cultures of cells expressing wild type (WT) TAP-tagged spTranslin. One culture was labeled with 15N and the other culture was labeled with 14N. Exp. 2: 15N-labeled cells expressing TAP-tag alone + 14Nlabeled cells expressing WT TAP-tagged spTranslin. Exp. 3: 15N-labeled cells expressing WT TAP-tagged spTranslin + 14N-labeled cells expressing TAP-tag alone. Exp. 4: 15Nlabeled cells expressing the TAP-tagged Y227stop spTranslin variant + 14N-labeled cells expressing WT TAP-tagged spTranslin. Exp. 5: 15N-labeled cells expressing WT TAPtagged spTranslin + 14N-labeled cells expressing the TAP-tagged Y227stop spTranslin variant. Exp. 6: 15N-labeled cells expressing the TAP-tagged K106A spTranslin variant + 14N-labeled cells expressing WT TAP-tagged spTranslin. Exp. 7: 15N-labeled cells expressing WT TAP-tagged spTranslin + 14N-labeled cells expressing the TAPtagged K106A spTranslin variant.

the supplementary material, which also presents a statistical analysis of the data obtained in both this and the previous experiment. The analysis was performed, using the Perseus software [39]. In particular, we have calculated the significance B parameter for each protein recovered in the two “reciprocal” experiments. It can be seen in Table S3 that of these proteins, only spTranslin, spTRAX, srp1 and sks2 exhibited Bsignificant 15N/14N ratios marked by two + signs. These results further support the association of spTRAX and srp1 with spTranslin. However, the apparent association of sks2 with spTranslin was not confirmed by other experiments described below. The data regarding spTranslin, spTRAX and srp1 in Exp. nos. 2 and 3 are also summarized in Table 1 and the results are graphically illustrated in Fig. 3. It can be seen that the log2 values of the 15N/14N ratios for these proteins have been inverted in the “reciprocal” experiment (Exp. no. 3) vs. the previous experiment (Exp. no. 2). This inversion confirms the earlier observation of the association between spTranslin and spTRAX and also supports the specificity of association of srp1 with spTranslin. Note however, that judging from the lengths of the bars in Fig. 3, the interaction of spTranslin with srp1 is much less tight than its interaction with spTRAX. Subsequently, we have carried out two additional “reciprocal” experiments, in which the same line of S. pombe cells expressing the wild-type (WT) spTranslin was employed. However, instead of the cells synthesizing the TAP-tag alone, we have used other S. pombe cells expressing a TAP-tagged spTranslin variant designated Y227stop. This variant protein lacks the ability to form spTranslin octamers and to interact with spTRAX [18]. As in the previous two “reciprocal” experiments, the cells expressing the TAP-tagged variant spTranslin, and the cells expressing the TAP-tagged WT spTranslin, were differentially labeled with the “heavy” 15N isotope and the “light” 14N isotope. Table S4 in the supplementary material shows the list of proteins recovered in these two experiments and the statistical analysis of the 15N/14N data. Table 1 and Fig. 3 show the data regarding spTranslin, spTRAX and srp1 obtained in these two experiments that were designated 4 and 5,

772


Table 2 Demonstration of specific association of the proteins srpi and vipl with spTranslin by the immune-affinity purification procedure. Exp. No.

No. MS2 peptides a

Median 15N/14N of MS2 peptides b

2.064e9 4.746e7 1.208e6 4.673e7

16 18 2 18

spTranslin spTRAX srp1 vip1

4.556e8 5.045e7 2.57e5 2.771e7

8 9 19

(14N) spTranslin+ (15N) no spTranslin (separately processed)b


4.517e9 1.944e8 2.410e7 1.426e7

19 20 9 17

(14N) no spTranlin+ (15N) spTranslin (separately processed)b


6.890e8 2.825e7 7.672e6 4.045e6

10 7 4

0.01 0.01 0.372 0.568 (0.565)c 92.986 50.245 0.550 (0.594)c 0.01 0.01 0.01 0.199 (0.358)c 100 100 4.297 (2.125)c

S. pombe cultures

Protein recovered

Peak area


(14N) no spTranslin+ (15N) spTranslina

(14N) spTranslin+ (15N) no spTranslina

1

2

3

4

No. MS1 peptides c

-

1 -

-

4 -

Median 15N/14N of MS1 peptides d

-

0.76 -

-

100 -

Normalized N/14N nd 15

0.017 0.017 0.658 1.005 100 84.587 1.279 0.926 0.027 0.027 0.027 0.556 47.05 47.05 47.05 2.022

a

S. pombe cultures labeled with either the “heavy” or the “light” nitrogen isotope were mixed before cell disruption. The (14N)-labeled cells and the (15N)-labeled cells were separately lysed; then spTranslin and associated proteins were immuno-precititated from each extract. The purified proteins from the two extracts were combined for the mass spectrometry analysis. c In each experiment the median of the 15N/14N ratios of all the recovered proteins is shown in parentheses. d The calculation of n is described in the section on Materials and methods. b

which clearly resemble the data obtained in Exp. nos. 2 and 3. These results indicated that the spTranslin variant lacking the ability to form homo-octamers, or hetero-octamers with spTRAX, either turned over rapidly, or lost its ability to interact with srp1. In either case, the inversion of the bars in these two reciprocal experiments (Fig. 3) provided independent evidence for the strong association of spTranslin with spTRAX and its weak, but specific, association with srp1. Furthermore, the significance B calculations shown in Table S4, which gave two + signs to the 15N/14N ratios for spTranslin, spTRAX and srp1 also supported the association of the latter two proteins with spTranslin. A possible

8

1

2

3

4

6

log215N/14N

4 2

spTranslin spTRAX

0 srp1 -2

vip1

-4 -6 -8

Fig. 4. Graphical representation of quantitative MS data for Translin, TRAX, srp1 and vip1 that were isolated by the immuno-affinity purification procedure. Each group of four bars 14 illustrates the log15 2 N/ N ratios determined in an experiment performed with two cultures that were differentially labeled with the 15N and the 14N isotopes. In experiments 1 and 2 the cells of the two cultures were combined before the extraction and the immuno-purification steps. In the experiments 3 and 4 the extraction and the immunopurification of the proteins were performed separately from each culture and then the purified proteins were mixed, resolved by SDS-PAGE and analyzed by mass-spectrometry, as described in Materials and methods. Experiments 1 and 3: 15N-labeled cells expressing no spTranslin + 14N-labeled cells expressing spTranslin. Experiments 2 and 4: 15N-labeled cells expressing spTranslin + 14N-labeled cells expressing no spTranslin.

association with spTranslin was also indicated in these experiments by the significance B calculations (two + signs) for the 60S ribosomal protein L21-A and the mitochondrial ATP synthase subunit alpha. However, these results were not reproducible in other experiments. Finally, we carried out two similar “reciprocal” experiments for which we used the cells expressing WT TAP-tagged spTranslin and cells expressing a TAP-tagged spTranslin variant protein designated K106A. This protein cannot bind RNA [18]. The results of these experiments are presented in Table S5 in the supplementary material, as well as Table 1 and Fig. 3 (Exp. nos. 6 and 7). It can be seen that, unlike the spTranslin Y227stop variant, the K106A spTranslin variant was recovered in the purified fraction along with the wild-type spTranslin indicating that it forms stable octameric complexes. However, as shown in Table S5, the significance B calculations indicated that srp1 was preferentially associated with the wild-type spTranslin (two + signs for the 15 N/14N ratio of srp1). This preferential association, which is also revealed by the data presented in Table 1 and Fig. 3, supports the notion that the interaction of srp1 with spTranslin is enhanced by the presence of RNA in the complexes (see Discussion). Note that the calculations for spTRAX also resulted in two + signs (Table S5). This observation is consistent with the notion that the assembly, or the stability, of spTranslinspTRAX hetero-octamers is also dependent on the association of such hetero-octamers with RNA. The dependence of the interaction between spTranslin and srp1 on the presence of RNA is also supported by the experiment presented in Table S6 in the supplementary material. In this experiment the protocol used for Exp. no. 3 was used, except that the extract was treated with RNase A before the TAP-tag purification of the complexes. It can be seen in Table S6 that this treatment led to removal of srp1 from the purified fraction. 3.1.2. The immuno-affinity purification approach As described in the Introduction, we have recently identified protein–protein interaction surfaces near the N-terminus and the C-terminus of each subunit of the spTranslin octamers [18]. In view of these findings, it appeared plausible that the large (21 kDa) TAP-tag, which was attached to the N-terminus of the protein (as shown in


A

B

Anti-HA AB

kDa 250 150 100 75

1

2

3

4

5

6

7

8

50 37 25 20 15

HA-srp1

D

Anti-HA AB

250 150 100 75 50 37

Anti-Tsn AB

kDa 250 150 100 75 50 37

1

2

spTranslin

25 20 15

C kDa

773

1

2

3

4

Anti-Tsn AB

kDa 250 150 100 75 HA-srp1 50 37

25 20 15

25 20 15

1

2

3

4

spTranslin

HA Fig. 5. Western blot analyses of intracellular S. pombe complexes including the HA-tagged srp1 + spTranslin, or the HA tag alone. S. pombe strains expressing the HA-tagged srp1, or just the HA tag, were constructed by transformation of appropriate expression plasmids into S. pombe cells, in which the spTRAX gene has been deleted. Complexes including the HA-tagged srp1 (or the HA tag alone), and associated proteins, were isolated from extracts prepared from these cells by immuno-precipitation with a monoclonal anti HA antibody coupled to agarose beads (Sigma). A. Western blot of extract prepared from the cells expressing the HA-tagged srp1 and of fractions collected during the immuno-precipitation procedure. This blot was probed with a monoclonal anti-HA antibody. Lane 1: extract. Lane 2: flow-through fraction. Lane 3: first wash with IP buffer (Sigma). Lane 4: second wash with IP buffer. Lane 5: third wash with IP buffer. Lane 6: fourth wash with IP buffer. Lane 7: fifth wash with 0.10× IP buffer. Lane 8: sample of material eluted from the beads by incubation with 0.10 M glycine buffer, pH 2.5. The positions attained by protein markers in the SDS-polyacrylamide gel are indicated on the left. B. Other samples of the extract and the immuno-purified fraction used for the assay shown in panel A were also analyzed by the western blotting technique, but probed with a polyclonal anti-spTranslin antibody14. Lane 1: sample of extract. Lane 2: sample of eluted material. C and D: western blots of extracts and immuno-purified fractions prepared from either the cells expressing the HA-tagged srp1, or the cells expressing the HA tag alone. The blots in panel C were probed with the anti-HA antibody, while the blots in panel D were probed with the anti spTranslin antibody. The lane designations for each of the two panels are: lane 1: extract from the HA-tagged srp1-expressing cells, lane 2: extract from the HA tag-expressing cells, lane 3: immuno-purified fraction from the HA-tagged srp1-expressing cells, and lane 4: immuno-purified fraction from the HA tag-expressing cells.

Fig. S1A in the supplementary material), interfered with the binding of some proteins that might bind the native untagged spTranslin in the cells. To investigate this possibility, we used the experimental procedure illustrated schematically in Fig. 1B. As indicated, we differentially labeled S. pombe cells expressing untagged spTranslin, and control cells expressing no spTranslin, with the 14N and 15N isotopes. In the initial two “reciprocal” experiments carried out according to this scheme, the cells from the two cultures were first combined and then disrupted, as was done in the TAP-tag purification experiments described in the previous section. Subsequently, the complexes including untagged spTranslin (or no spTranslin) and associated proteins were isolated from these extracts by immuno-affinity purification with the polyclonal anti-spTranslin antibody [14]. Finally, the 15N/14N ratios of the proteins in these complexes were determined by mass spectrometry and analyzed, as described in the previous section. Note that on account of the less stringent one-step purification employed in these experiments, the total number of proteins that were recovered in the purified fraction (up to 600 proteins; not shown) was considerably larger than the number of proteins recovered in the TAP-tag experiments. We present in Table 2 and Fig. 4 (Exp. nos. 1 and 2) just the data obtained for the proteins spTranslin, spTRAX, srp1 and one additional protein designated vip1. It should be noted that in the many TAP-tag affinity purification experiments that we have carried out the protein vip1 has never been identified in the purified complexes. In contrast, it has been found in relatively large quantities in the immunoprecipitates of both “reciprocal” immunoaffinity purification Exp. nos. 1 and 2. Further inspection of these data also reveals that the 15N/14N ratios for both spTranslin and spTRAX were very small in Exp. no. 1 and very large in Exp. no. 2, thereby confirming the strong association between the two proteins. However, the differences

between the 15N/14N ratios for the protein srp1 have been quite small, while the ratios for the protein vip1 have not been substantially changed in the two experiments. In view of these results, we considered the possibility that during the extraction and immuno-precipitation steps proteins that are weakly associated with spTranslin can readily dissociate from the “heavy” 15N-labeled complexes and re-associate with the “light” 14N-labeled complexes, and vice versa. Such an exchange could mask relatively small differences between the 15N/14N ratios in the “reciprocal” experiments. To examine this possibility, we have carried out two additional “reciprocal” experiments, in which rather than mixing the 15 N-labeled cells with the 14N-labeled cells before the extraction and immuno-precipitation steps, we have separately extracted and purified the proteins from the “heavy” and the “light” cultures (procedure shown in Fig. 1C). Only subsequent to these steps, we combined the proteins from the two cultures and analyzed their tryptic peptides, as previously described. As seen in Table 2 and Fig. 4, in this “reciprocal” pair of experiments the 15N/14N ratios of both srp1 and vip1 have been inverted. These results confirm that both proteins are specifically associated with native untagged spTranslin. Moreover, the exchange that apparently occurred between the “heavy” and the “light” complexes in the Exp. nos. 1 and 2 of this series supports the notion that the association between these proteins and spTranslin is rather weak in comparison to the association between spTranslin and spTRAX. 3.2. Further support for the specific association of spTranslin with srp1 and vip1 in S. pombe cells If the proteins srp1 and vip1 are specifically associated with spTranslin in S. pombe cells, then immuno-precipitation of either of

774


A

HA-srp1

+

B

-

HA-srp1

+

-

anti-HA AB

anti-HA AB extract

extract

anti-Tsn AB

anti-Tsn - AB

anti-HA AB

anti-HA AB IP

IP anti-Tsn AB

anti-Tsn AB

C

HA-vip1

+

-

anti-HA AB extract anti-Tsn AB

anti-HA AB IP anti-Tsn AB

Fig. 6. Western blot analyses of intracellular complexes including the HA-tagged srp1, or the HA-tagged vip1, and of complexes prepared from control strains. A. Left side: western blots of extracts and immuno-precipitates of proteins derived from S. pombe cells that express both the endogenous spTranslin and spTRAX, and also express HA-srp1. Only the HA-srp1 and spTranslin bands are shown. A. Right side: western blots of extracts derived from cells of a control strain that do not express HA-srp1. The positions in the blot corresponding to the bands found on the left side of the panel are shown. B. Left side: western blots of proteins derived from cells that express spTRAX and HA-srp1, and also over-express spTranslin under the control of the nmt1 promoter. B. Right side: western blots of extracts derived from cells of a control strain that do not express HA-srp1. The positions in the blot corresponding to the bands found on the left side of the panel are shown. C. Left side: similar western blots of proteins derived from S. pombe cells that express spTRAX and the HA-tagged vip1 protein, and also over-express spTranslin under the control of the nmt1 promoter. C. Right side: western blots of extracts derived from cells of a control strain that express the endogenous spTRAX and over-express spTranslin under the control of the nmt1 promoter, but do not express the HA-tagged vip1 protein.

these two proteins by antibodies directed against it should cause coimmuno-precipitation of spTranslin. To test this prediction with regard to srp1, we used several S. pombe strains that express the srp1 protein linked to an HA tag, as well as control lines that do not express the tagged protein. Extracts prepared from cells of each of these lines were mixed

Fig. 7. Schematic drawing of spTranslin complexes in S. pombe cells. See the text for details.

with beads coated with a monoclonal anti HA-antibody. After extensive washing of the beads, the HA-srp1 protein was eluted along with proteins that remained attached to it. Western blot analysis was then performed on the extracts and the eluted fractions. Fig. 5A shows such a western blot of HA-srp1-expressing S. pombe cells that also express the endogenous spTranslin, but do not express spTRAX. The blot includes a sample of the crude extract and samples of various fractions collected during the immuno-precipitation procedure. The probe used for this analysis was a soluble monoclonal anti HA-antibody. It can be seen that the HA-linked srp1 was found in a single band; based on the mobility of this band, the apparent molecular weight of the protein was calculated and found to be 47.5 kDa. This analysis also revealed that most of the HA-linked srp1 was strongly bound to the beads (compare lane 1 to the lanes 2–7). As shown in lane 8, the bound HA-linked srp1 was efficiently eluted with 0.10 M glycine, pH 2.5. Fig. 5B shows a western blot of other samples withdrawn from the same extract and eluted fraction. This blot was probed with the polyclonal anti-spTranslin antibody [14]. It can be seen that the antibody reacted with several protein bands in the extract. Two of these bands were highly enriched by the immunoprecipitation with the antibody directed against the HA-linked srp1 (compare lanes 1 and 2 in Fig. 5B). The slower band contains a protein with an apparent molecular weight of 26 kDa, which is expected for spTranslin. The faster band, which has an apparent molecular weight of 23 kDa, is probably a degradation product of spTranslin. The authenticity of spTranslin in these bands has been further confirmed by massspectrometry of a corresponding piece of a gel run in parallel, in which 7 specific tryptic peptides derived from spTranslin were identified (not shown). Fig. 5C and D show western blot analyses of extracts and immunoprecipitates obtained in another independent assay of the same strain


and of the corresponding control strain expressing just the HA tag. The blot shown in Fig. 5C was probed with an anti-HA antibody. As indicated, the HA-tagged srp1 band was found in both the extract and the immuno-precipitate of the experimental line (lanes 1 and 3), while much faster migrating bands contained the free HA tag derived from cells of the control line (lanes 2 and 4). The blot shown in Fig. 5D, which contained other samples of the same extracts and immunoprecipitates, was probed with the anti-spTranslin antibody. As indicated, this antibody reacted with several bands in the extracts obtained from both the experimental and the control lines (lanes 1 and 2). However, immuno-precipitation of the HA-tagged srp1 with the anti-HA antibody led to a co-immuno-precipitation of spTranslin only in the samples from the experimental line and not in the samples from the control line (compare lanes 3 and 4). These results strongly support the specificity of association between spTranslin and the HA-tagged srp1 in the S. pombe cells expressing this protein, while the results obtained with the control strain revealed that this association is not due to an interaction with the HA tag. It is also noteworthy that since these cells do not express spTRAX, they only contain homogenous spTranslin octamers with which the srp1 protein apparently associated. Fig. 6A shows the results of a similar analysis performed on S. pombe cells which do express the endogenous spTRAX protein, as well as the endogenous spTranslin. Shown on the left side of this panel are the HA-tagged srp1 and spTranslin bands found in the extract and the immuno-precipitate obtained from cells of this line that also express the HA-tagged srp1. The right side of this panel shows blots of the corresponding gel positions of an extract and an immuno-precipitate obtained from cells of a control line that does not express the HA-tagged srp1. Evidently, the results obtained in this experiment resemble the results obtained in the analysis of the line that does not express spTRAX shown in Fig. 5, except that spTranslin was barely detectable in this immunoprecipitate (compare the spTranslin bands in the immunoprecipitate of Fig. 6A to the corresponding bands in lane 3 of Fig. 5D). This quantitative difference has been reproduced in several independent experiments, and its significance will be discussed in the Discussion section. Fig. 6B shows a similar western analysis performed on cells that express the endogenous spTRAX and spTranslin, but in addition, also overexpress exogenous spTranslin. It can be seen that, as might be expected, the amount of spTranslin observed in the immuno-precipitate in this panel was larger than the amount observed in Fig. 6A. Taken together, the data shown in Figs. 5 and 6 indicate that in S. pombe cells with different genetic backgrounds srp1 specifically associates with homogeneous spTranslin octamers and possibly also with mixed spTranslin-spTRAX octamers. Fig. 6C shows similar data obtained in western analyses of extracts and immuno-precipitates obtained from cells expressing HA-tagged vip1 protein and overexpressing spTranslin, and from the corresponding control cells that do not express the HA-tagged vip1 protein. Clearly, spTranslin was co-immunoprecipitated with the HA-tagged vip1 protein, confirming the association between the vip1 protein and spTranslin. It should be noted, however, that a similar western analysis of cells expressing just the endogenous spTranslin failed to detect spTranslin in the immunoprecipitate (not shown). 4. Discussion The first attempt to identify intracellular binding partners of Translin was made by use of the yeast two-hybrid technique. In this early study, the cloned human Translin was used as bait and the search for protein partners was made among clones of a whole human cDNA library. This search led to the discovery of the human TRAX, but no other binding partners of Translin were identified [7]. The affinity purification methods that we employed in the present study, combined with the sensitive mass spectrometry analyses, were expected to be more suitable for identification of proteins that have relatively low affinities for

775

spTranslin. These methods were also expected to identify proteins whose binding to spTranslin might be dependent on the presence of RNA in the complexes. However, whereas the strong association between spTRAX and spTranslin was readily confirmed by this approach, identification of additional protein partners of spTranslin proved to be a very difficult task. For as indicated in the Results section, co-purification with spTranslin in affinity columns, or in immunoprecipitation assays, was found to be an insufficient criterion for designating a protein as a true binding partner of spTranslin. Clearly, the additional step of differential labeling of the experimental and control cultures with the 15N and 14N isotopes was essential for accomplishment of this task. Using this improved technique, we identified the protein srp1 as a likely component of complexes including spTranslin, and probably also RNA molecules. However, in some of the assays we encountered technical difficulties stemming from the limitations of the MS2 massspectrometry analysis. For apart from spTRAX, the masses of the rest of the proteins recovered in the purified fractions, including srp1, were rather small. Hence, the mass/charge peaks of the peptides derived from these proteins were very small relative to the instrument background and were not always detectable. Furthermore, even when these peaks were detectable in the first stage of the analysis, they were not always picked up by the apparatus for fragmentation and MS2 identification of the fragments. Therefore, we also performed a manual search for additional srp1 peptides that have not been detected by the automatic MS2 analysis. This search led to identification of such peptides (MS1 peptides in Tables 1 and 2). The 15N/14N ratios of these peptides further supported our conclusions regarding the specificity of the association between srp1 and spTranslin. In none of the many TAP-tag affinity purification assays that we have carried out we found the protein vip1 to co-purify with the tagged spTranslin. However, vip1 was found to co-purify with the untagged spTranslin in the immuno-purification assays. The TAP-tag was attached to the N-terminal region of spTranslin, which had been implicated by our previously published bioinformatics calculations to be a binding surface for proteins [18]. Hence, our inability to detect vip1 in the TAP-tag purification assays is consistent with the notion that the binding site for vip1 resides at the N-terminus of spTranslin and that it was masked by the large TAP-tag. Clearly, the interactions of spTranslin with either srp1, or vip1, are rather weak. As a result, exchange of molecules of both proteins between the 15N- and the 14N-labeled complexes occurred during the extraction and purification procedures (see Table 2 and Fig. 4). This exchange caused difficulties in their identification as specific binding partners of spTranslin. These observations underlie the difficulties expected in the characterization of many other intracellular multi-protein complexes, in which the protein–protein interactions are rather weak. As shown in this article, it is possible to distinguish specific weak-binding partners in such complexes from false positive partners by using differential “heavy” and “light” isotope labeling and a careful statistical analysis of the data. srp1 belongs to the evolutionarily conserved SR family of proteins that have been initially identified as RNA splicing regulators. Subsequently, these proteins were also found to be involved in various other functions, including mRNA nuclear export and translation, as well as micro RNA processing [40,41]. Interestingly, some of these functions have been previously attributed to Translin [1,2]. Like several of the metazoan SR proteins, the S. pombe srp1 contains a single RNA Recognition Motif (RRM) and an arginine/serine-rich RS domain. In srp1, these two domains are separated by a glycine-rich hinge [34]. In addition, srp1 contains a C-terminal domain of unknown function. Unlike a second SR protein family member identified in S. pombe, which was designated srp2 [42], srp1 is not essential for the fission yeast growth. We have not detected srp2 among the proteins found to be associated with spTranslin, even though srp1 and srp2 have been reported to interact [43]. On the other hand, we did present evidence for the dependence

776


of the spTranslin-srp1 association on the ability of spTranslin octamers to bind RNA (Table 1, Fig. 3 and Table S6 in the supplementary material). It is interesting to recall, in this connection, the observation that in the cells that do not express spTRAX, the endogenous spTranslin was much more readily detectable in the HA-tagged srp1 immunoprecipitates than in the cells that express both spTranslin and spTRAX (compare Fig. 5D, lane 3 to Fig. 6A, bottom panel). A likely interpretation of these data is that the spTRAX nuclease activity found in spTranslinspTRAX hetero-octamers might hydrolyze RNA molecules found in complexes that also contain srp1 and thereby weaken the association of this protein with such hetero-octamers. Hence, the association between the two proteins is more readily detectable in the cells that do not synthesize spTRAX and contain only spTranslin homo-octamers. This interpretation is in line with the data presented in Table 1 and Fig. 3 (Exp. nos. 6 and 7), which indicate that the presence of RNA in the complexes stabilizes the association between srp1 and spTranslin. However, an alternative interpretation is that the presence of spTRAX per se in the complexes weakens the association between spTranslin and srp1. Like srp1, vip1 contains an RNA Recognition Motif (RRM). Hence, we suggest that complexes containing spTranslin and vip1 may contain RNA as well. vip1, like srp1, is also non-essential for growth of S. pombe cells under normal growth conditions. However, unlike srp1, vip1 has only been conserved in fungi [44]. In conclusion, the data reported in this article support the existence in S. pombe cells of an ensemble of complexes containing just spTranslin, or both spTranslin and spTRAX, bound to the RNA-binding proteins srp1 and vip1, as well as to RNA (Fig. 7). We propose that this ensemble may also include complexes containing additional proteins that we failed to detect by the methods used in this study, and that, depending on their protein and RNA composition, each of the complexes might have a different physiological role. Moreover, the various complexes in this ensemble can readily exchange partners due to the rather weak interactions between their components. Finally, we suggest that the experimental approaches used for these studies should be applicable to studies of numerous other RNA-protein and protein–protein complexes, in which the interactions between the protein components are rather weak. Supplementary data to this article can be found online at http://dx. doi.org/10.1016/j.bbapap.2013.12.016. Acknowledgements The authors thank (1) Dr. K.L. Gould of the Department of Cell and Developmental Biology, Vanderbilt University School of Medicine, for sending us the plasmid pREP41-NTAP, (2) Dr. M. Lützelberger of the Institut für Genetik-Biozentrum, Technische Universitat Braunschweig, for sending us the plasmid pRHA42S, and (3) Dr. S.L. Forsburg of the Department of Molecular & Computational Biology, University of Southern California, for sending us the plasmid pSLF172. This study was supported by the Israel Science Foundation (grants 378/03 and 967/08). References [1] Z. Li, Y. Wu, J.M. Baraban, The Translin/Trax RNA binding complex: clues to function in the nervous system, Biochim. Biophys. Acta 1779 (2008) 479–485. [2] A. Jaendling, R.J. McFarlane, Biological roles of translin and translin-associated factor-X: RNA metabolism comes to the fore, Biochem. J. 429 (2010) 225–234. [3] A. Aharoni, N. Baran, H. Manor, Characterization of a multisubunit human protein which selectively binds single stranded d(GA)n and d(GT)n sequence repeats in DNA, Nucleic Acids Res. 21 (1993) 5221–5228. [4] M. Kasai, K. Aoki, Y. Matsuo, J. Minowada, R.T. Maziarz, J.L. Strominger, Recombination hotspot associated factors specifically recognize novel target sequences at the site of interchromosomal rearrangements in T-ALL patients with t(8;14)(q24;q11) and t(1;14)(p32;q11), Int. Immunol. 6 (1994) 1017–1025. [5] J.R. Han, W. Gu, N.B. Hecht, Testis-brain RNA-binding protein, a testicular translational regulatory RNA-binding protein, is present in the brain and binds to the 3′ untranslated regions of transported brain mRNAs, Biol. Reprod. 53 (1995) 707–717.

[6] K. Aoki, K. Suzuki, T. Sugano, T. Tasaka, K. Nakahara, O. Kuge, A. Omori, M. Kasai, A novel gene, Translin, encodes a recombination hotspot binding protein associated with chromosomal translocations, Nat. Genet. 10 (1995) 167–174. [7] K. Aoki, R. Ishida, M. Kasai, Isolation and characterization of a cDNA encoding a Translin-like protein, TRAX, FEBS Lett. 401 (1997) 109–112. [8] V.M. Chennathukuzhi, Y. Kurihara, J.D. Bray, J. Yang, N.B. Hecht, Altering the GTP binding site of the DNA/RNA-binding protein, Translin/TB-RBP, decreases RNA binding and may create a dominant negative phenotype, Nucleic Acids Res. 29 (2001) 4433–4440. [9] P.M. Finkenstadt, M. Jeon, J.M. Baraban, Trax is a component of the Translincontaining RNA binding complex, J. Neurochem. 83 (2002) 202–210. [10] G.D. Gupta, A. Kale, V. Kumar, Molecular evolution of translin superfamily proteins within the genomes of eubacteria, archaea and eukaryotes, J. Mol. Evol. 75 (2012) 155–167. [11] E.A. Parizotto, E.D. Lowe, J.S. Parker, Structural basis for duplex RNA recognition and cleavage by Archaeoglobus fulgidus C3PO, Nat. Struct. Mol. Biol. 20 (2013) 380–386. [12] S.P. Lee, E. Fuior, M.S. Lewis, M.K. Han, Analytical ultracentrifugation studies of translin: analysis of protein–DNA interactions using a single-stranded fluorogenic oligonucleotide, Biochemistry 40 (2001) 14081–14088. [13] M.K. Han, P. Lin, D. Paek, J.J. Harvey, E. Fuior, J.R. Knutson, Fluorescence studies of pyrene maleimide-labeled translin: excimer fluorescence indicates subunits associate in a tail-to-tail configuration to form octamer, Biochemistry 41 (2002) 3468–3476. [14] O. Laufman, Y.R. Ben, N. Adir, H. Manor, Cloning and characterization of the Schizosaccharomyces pombe homologs of the human protein Translin and the Translin-associated protein TRAX, Nucleic Acids Res. 33 (2005) 4128–4139. [15] J.M. Pascal, P.J. Hart, N.B. Hecht, J.D. Robertus, Crystal structure of TB-RBP, a novel RNA-binding and regulating protein, J. Mol. Biol. 319 (2002) 1049–1057. [16] I. Sugiura, C. Sasaki, T. Hasegawa, T. Kohno, S. Sugio, H. Moriyama, M. Kasai, T. Matsuzaki, Structure of human translin at 2.2 A resolution, Acta Crystallogr. D Biol. Crystallogr. 60 (2004) 674–679. [17] L. Pe'rez-Cano, E. Eliahoo, K. Lasker, H.J. Wolfson, F. Glaser, H. Manor, P. Bernado, J. Ferna' ndez-Recio, Conformational transitions in human translin enable nucleic acid binding, Nucleic Acids Res. 41 (2013) 9956–9966. [18] E. Eliahoo, R. Ben Yosef, L. Perez-Cano, J. Fernandez-Recio, F. Glaser, H. Manor, Mapping of interaction sites of the Schizosaccharomyces pombe protein Translin with nucleic acids and proteins: a combined molecular genetics and bioinformatics study, Nucleic Acids Res. 38 (2010) 2975–2989. [19] E. Jacob, L. Pucshansky, E. Zeruya, N. Baran, H. Manor, The human protein translin specifically binds single-stranded microsatellite repeats, d(GT)n, and G-strand telomeric repeats, d(TTAGGG)n: a study of the binding parameters, J. Mol. Biol. 344 (2004) 939–950. [20] R. Ishida, H. Okado, H. Sato, C. Shionoiri, K. Aoki, M. Kasai, A role for the octameric ring protein, Translin, in mitotic cell division, FEBS Lett. 525 (2002) 105–110. [21] X.Q. Wu, W. Gu, X. Meng, N.B. Hecht, The RNA-binding protein, TB-RBP, is the mouse homologue of translin, a recombination protein associated with chromosomal translocations, Proc. Natl. Acad. Sci. U. S. A. 94 (1997) 5640–5645. [22] X.Q. Wu, N.B. Hecht, Mouse testis brain ribonucleic acid-binding protein/translin colocalizes with microtubules and is immunoprecipitated with messenger ribonucleic acids encoding myelin basic protein, alpha calmodulin kinase II, and protamines 1 and 2, Biol. Reprod. 62 (2000) 720–725. [23] C.R. Morales, S. Lefrancois, V. Chennathukuzhi, M. El Alfy, X.Q. Wu, J.X. Yang, G.L. Gerton, N.B. Hecht, A TB-RBP and Ter ATPase complex accompanies specific mRNAs from nuclei through the nuclear pores and into intercellular bridges in mouse male germ cells, Dev. Biol. 246 (2002) 480–494. [24] Z. Yu, N.B. Hecht, The DNA/RNA-binding protein, translin, binds microRNA122a and increases its in vivo stability, J. Androl. 29 (2008) 572–579. [25] Y. Liu, X. Ye, F. Jiang, C. Liang, D. Chen, J. Peng, L.N. Kinch, N.V. Grishin, Q. Liu, C3PO, an endoribonuclease that promotes RNAi by facilitating RISC activation, Science 325 (2009) 750–753. [26] X. Ye, N. Huang, Y. Liu, Z. Paroo, C. Huerta, P. Li, S. Chen, Q. Liu, H. Zhang, Structure of C3PO and mechanism of human RISC activation, Nat. Struct. Mol. Biol. 18 (2011) 650–657. [27] Y. Tian, D.K. Simanshu, M. Ascano, R. az-Avalos, A.Y. Park, S.A. Juranek, W.J. Rice, Q. Yin, C.V. Robinson, T. Tuschl, D.J. Patel, Multimeric assembly and biochemical characterization of the Trax–translin endonuclease complex, Nat. Struct. Mol. Biol. 18 (2011) 658–664. [28] F. Philip, Y. Guo, O. Aisiku, S. Scarlata, Phospholipase Cbeta1 is linked to RNA interference of specific genes through translin-associated factor X, FASEB J. 26 (2012) 4903–4913. [29] C. Chiaruttini, A. Vicario, Z. Li, G. Baj, P. Braiuca, Y. Wu, F.S. Lee, L. Gardossi, J.M. Baraban, E. Tongiorgi, Dendritic trafficking of BDNF mRNA is mediated by translin and blocked by the G196A (Val66Met) mutation, Proc. Natl. Acad. Sci. U. S. A. 106 (2009) 16481–16486. [30] J.M. Stein, W. Bergman, Y. Fang, L. Davison, C. Brensinger, M.B. Robinson, N.B. Hecht, T. Abel, Behavioral and neurochemical alterations in mice lacking the RNA-binding protein translin, J. Neurosci. 26 (2006) 2184–2196. [31] G. Basi, E. Schmid, K. Maundrell, TATA box mutations in the Schizosaccharomyces pombe nmt1 promoter affect transcription efficiency but not the transcription start point or thiamine repressibility, Gene 123 (1993) 131–136. [32] J.J. Tasto, R.H. Carnahan, W.H. McDonald, K.L. Gould, Vectors and gene targeting modules for tandem affinity purification in Schizosaccharomyces pombe, Yeast 18 (2001) 657–662. [33] S.L. Forsburg, D.A. Sherman, General purpose tagging vectors for fission yeast, Gene 191 (1997) 191–195.

E. Eliahoo et al. / Biochimica et Biophysica Acta 1844 (2014) 767–777 [34] T. Gross, K. Richert, C. Mierke, M. Lutzelberger, N.F. Kaufer, Identification and characterization of srp1, a gene of fission yeast encoding a RNA binding domain and a RS domain typical of SR splicing factors, Nucleic Acids Res. 26 (1998) 505–511. [35] K.L. Gould, L. Ren, A.S. Feoktistova, J.L. Jennings, A.J. Link, Tandem affinity purification and identification of protein complex components, Methods 33 (2004) 239–244. [36] K.K. Rosenfeld, T. Ziv, S. Goldin, F. Glaser, H. Manor, Mapping of DNA binding sites in the Tetrahymena telomerase holoenzyme proteins by UV cross-linking and mass spectrometry, J. Mol. Biol. 410 (2011) 77–92. [37] G. Rigaut, A. Shevchenko, B. Rutz, M. Wilm, M. Mann, B. Seraphin, A generic protein purification method for protein complex characterization and proteome exploration, Nat. Biotechnol. 17 (1999) 1030–1032. [38] H. Steen, M. Mann, The ABC's (and XYZ's) of peptide sequencing, Nat. Rev. Mol. Cell Biol. 5 (2004) 699–711. [39] J. Cox, M. Mann, MaxQuant enables high peptide identification rates, individualized p.p.b.-range mass accuracies and proteome-wide protein quantification, Nat. Biotechnol. 26 (2008) 1367–1372.

777

[40] J.C. Long, J.F. Caceres, The SR protein family of splicing factors: master regulators of gene expression, Biochem. J. 417 (2009) 15–27. [41] G. Risso, F. Pelisch, A. Quaglino, B. Pozzi, A. Srebrow, Regulating the regulators: serine/arginine-rich proteins under scrutiny, IUBMB Life 64 (2012) 809–816. [42] M. Lutzelberger, T. Gross, N.F. Kaufer, Srp2, an SR protein family member of fission yeast: in vivo characterization of its modular domains, Nucleic Acids Res. 27 (1999) 2618–2626. [43] Z. Tang, N.F. Kaufer, R.J. Lin, Interactions between two fission yeast serine/arginine-rich proteins and their modulation by phosphorylation, Biochem. J. 368 (2002) 527–534. [44] D.U. Kim, J. Hayles, D. Kim, V. Wood, H.O. Park, M. Won, H.S. Yoo, T. Duhig, M. Nam, G. Palmer, S. Han, L. Jeffery, S.T. Baek, H. Lee, Y.S. Shim, M. Lee, L. Kim, K.S. Heo, E.J. Noh, A.R. Lee, Y.J. Jang, K.S. Chung, S.J. Choi, J.Y. Park, Y. Park, H.M. Kim, S.K. Park, H.J. Park, E.J. Kang, H.B. Kim, H.S. Kang, H.M. Park, K. Kim, K. Song, K.B. Song, P. Nurse, K.L. Hoe, Analysis of a genome-wide set of gene deletions in the fission yeast Schizosaccharomyces pombe, Nat. Biotechnol. 28 (2010) 617–623.