Logo of rnaThe RNA SocietyeTOC AlertsSubscriptionsJournal HomeCSHL PressRNA
RNA. 2003 Sep; 9(9): 1123–1137.
PMCID: PMC1370476

Partitioning and translation of mRNAs encoding soluble proteins on membrane-bound ribosomes


In eukaryotic cells, it is generally accepted that protein synthesis is compartmentalized; soluble proteins are synthesized on free ribosomes, whereas secretory and membrane proteins are synthesized on endoplasmic reticulum (ER)-bound ribosomes. The partitioning of mRNAs that accompanies such compartmentalization arises early in protein synthesis, when ribosomes engaged in the translation of mRNAs encoding signal-sequence-bearing proteins are targeted to the ER. In this report, we use multiple cell fractionation protocols, in combination with cDNA microarray, nuclease protection, and Northern blot analyses, to assess the distribution of mRNAs between free and ER-bound ribosomes. We find a broad representation of mRNAs encoding soluble proteins in the ER fraction, with a subset of such mRNAs displaying substantial ER partitioning. In addition, we present evidence that membrane-bound ribosomes engage in the translation of mRNAs encoding soluble proteins. Single-cell in situ hybridization analysis of the subcellular distribution of mRNAs encoding ER-localized and soluble proteins identify two overall patterns of mRNA distribution in the cell—endoplasmic reticular and cytosolic. However, both partitioning patterns include a distinct perinuclear component. These results identify previously unappreciated roles for membrane-bound ribosomes in the subcellular compartmentalization of protein synthesis and indicate possible functions for the perinuclear membrane domain in mRNA sorting in the cell.

Keywords: mRNA, endoplasmic reticulum, polysome, protein synthesis, secretion


Protein synthesis in eukaryotic cells is compartmentalized, with soluble proteins being synthesized on free ribosomes and secretory/integral membrane proteins on endoplasmic reticulum (ER)-bound ribosomes. Such compartmentalization limits access to the secretory pathway to those proteins bearing signal sequences or other topogenic domains (Blobel and Dobberstein 1975a,b; Palade 1975). As is well documented, the compartmentalization of secretory and integral membrane protein synthesis to the ER occurs via a positive selection process; ribosomes engaged in the synthesis of such proteins are trafficked from the cytoplasm to the ER membrane via the signal recognition particle (SRP) pathway (Blobel and Dobberstein 1975b; Lingappa and Blobel 1979; Rapoport et al. 1989; Walter and Johnson 1994). The cytosol-to-ER ribosome/mRNA trafficking event is thought to be transient, with ribosome binding and release occurring in synchrony with the elongation and termination stages of protein synthesis, respectively (Blobel and Dobberstein 1975b; Lingappa and Blobel 1979; Rapoport et al. 1989; Walter and Johnson 1994).

The role of the SRP pathway in targeting ribosome/mRNA complexes to the ER is well established. Little is known, however, regarding what is thought to be a termination-coupled process of ribosome release from the ER. This topic has recently come under investigation and has yielded several unexpected findings. Foremost, it has been demonstrated, in situ and in vitro, that ribosomes remain bound to the ER membrane following the termination of protein synthesis (Seiser and Nicchitta 2000; Potter et al. 2001; Potter and Nicchitta 2002). This finding raises several fundamental questions regarding the compartmentalization of cellular protein synthesis. For example, can membrane-bound ribosomes engage in de novo protein synthesis? Two, if ribosome release from the ER does not accompany the termination reaction of protein synthesis, how are membrane-bound ribosomes returned to the common cytosolic ribosomal subunit pool? Several insights into these questions have recently become available. In vitro analyses of the protein synthesis activity of membrane-bound ribosomes have demonstrated that posttermination, membrane-bound ribosomes can initiate de novo protein synthesis, and, importantly, do not distinguish between mRNAs encoding soluble or signal-sequence-bearing proteins (Potter and Nicchitta 2000a). Furthermore, in vitro studies have demonstrated that secretory proteins whose synthesis was initiated on membrane-bound ribosomes can undergo protein translocation in the absence of a functional SRP/SRP receptor pathway (Potter and Nicchitta 2000a,b, 2002). Also noteworthy in these in vitro studies was the observation that ER-bound ribosomes engaged in the synthesis of soluble proteins can undergo release from the ER to the cytosol compartment (Potter and Nicchitta 2000a; Potter et al. 2001). It was proposed that this newly identified process of nascent chain elongation-coupled ribosome release (E-CRR) serves as the primary mechanism for ribosome exchange on the ER (Potter and Nicchitta 2000a; Potter et al. 2001).

Although the E-CRR model is consistent with in vitro data (Potter and Nicchitta 2000a, 2002; Potter et al. 2001), evidence that such a pathway functions in situ is lacking. One prominent prediction of this model is that the steady-state distribution of mRNAs on membrane-bound polysomes should include mRNAs encoding soluble proteins. Interestingly, analyses of the mRNA compositions of free and membrane-bound polysomes have not revealed the distinct, bimodal mRNA partitioning that would be expected were the SRP pathway to serve as the sole arbitrator of mRNA localization to the ER (Mechler and Rabbitts 1981; Mueckler and Pitot 1981; Diehn et al. 2000). For example, early studies of the mRNA composition of free and membrane-bound polysomes performed by cDNA–RNA kinetic hybridization analyses indicated a substantial overlap in the composition of the two pools (Mueckler and Pitot 1981). These early observations, which were limited to overall population similarities, were recently confirmed in a genome-wide cDNA microarray analysis (Diehn et al. 2000). There are at least two possible explanations for the existence of mRNAs encoding soluble proteins in the ER compartment. One, cell homogenization/fractionation protocols produce the artifactual partitioning of mRNAs between the cytosol and ER membrane compartments. As insightfully discussed by Palade (1975), it is difficult to unambiguously exclude this possibility. Two, such noncanonical distributions may be a reflection of elongation-coupled ribosome release (Potter et al. 2001). According to the E-CRR model, membrane-bound ribosomes engage in the translation of mRNAs encoding soluble proteins, either because such mRNAs are localized to the ER and/or because ribosomes do not innately distinguish between mRNAs encoding soluble or secretory protein (Potter and Nicchitta 2000a).

Here, we used two cell fractionation methods, and two cell types, to examine the partitioning of mRNAs between the free and membrane-bound polysomes of mammalian cells. In one method, traditional homogenization and isopycnic density flotation protocols were used to separate free and membrane-bound polysomes. In an alternative protocol, cells were permeabilized by addition of digitonin and the free (soluble) and membrane-bound polysome pools were selected by differential centrifugation. Nuclease protection, Northern blot, and cDNA microarray analyses were performed to determine the relative enrichment of different mRNAs in the polysome pools. The data indicate that in addition to the expected canonical distribution of mRNAs encoding signal sequences, mRNAs encoding soluble proteins were well represented on the ER membrane fractions. In addition, evidence is provided that mRNAs encoding soluble proteins can be markedly enriched and translated on membrane-bound ribosomes. Using a different experimental approach, single-cell in situ hybridization studies provided direct evidence for the partitioning of mRNAs encoding soluble proteins to the ER membrane fraction of intact cells. In addition, in situ hybridization studies provided evidence for an unexpected, perinuclear distribution for the mRNAs examined. These data are discussed with respect to present models of the mechanisms governing mRNA localization in somatic cells.


Experimental systems

mRNA partitioning between free and membrane-bound polysomes was first examined by cell fractionation, using two fractionation methods. In one method (Fig. 1A [triangle]), tissue culture cells were physically disrupted by homogenization in hypotonic media and the ER and cytosol fractions were subsequently isolated by isopycnic flotation and velocity sedimentation, respectively. In a second method (Fig. 1B [triangle]), tissue culture cells were treated with digitonin, a β-sterol binding detergent, at concentrations sufficient for selective permeabilization of the plasma membrane and the release of the soluble contents of the cell (Dunn and Holz 1983; Brooks and Treml 1984). This procedure has been extensively used in studies of endosome trafficking as well as regulated nuclear import (Diaz et al. 1989; Adam et al. 1990). Under these conditions, free polysomes are recovered in the cytosol fraction, and ER-bound ribosomes remain in association with the ER membrane, as reported previously (Potter and Nicchitta 2002). Two cell lines were used in these studies; the human T-cell line Jurkat and the murine B-cell line J558. The latter displays the abundant ER typical of highly secretory tissues.

Methods for fractionation of cultured cells. (A) Mechanical homogenization. Tissue culture cells were suspended in hypotonic buffer, homogenized, and either mixed with high density (2.5 M) sucrose and centrifuged in a discontinuous gradient to yield isopycnic ...

Marker protein distribution was used to define the relative purity of the subcellular fractions generated by the two fractionation procedures. The results for fractions prepared by mechanical homogenization are depicted in Figure 2A [triangle]. For both cell lines, TRAPα, a resident ER integral membrane protein, the soluble kinase ERK2, or cytoskeletal component actin and the plasma membrane integral membrane protein Na+/K+ ATPase were present in the homogenate fraction. Following fractionation, the cytosol component was found to contain ERK2 or actin (in Jurkat and J558, respectively) and to be lacking either the plasma membrane marker Na+/K+ ATPase or the ER marker TRAPα. Similarly, the purified ER fraction was found to contain TRAPα, and to be free of plasma membrane or cytosol markers (Fig. 2A [triangle]). In parallel studies, the soluble marker lactate dehydrogenase (LDH) was examined for multiple fractionations of both Jurkat and J558 cells. Consistent with the distribution of ER markers depicted in Figure 2A [triangle], LDH activity was wholly recovered in the cytosol fraction, with <3.5% of total cellular LDH activity being recovered in the ER fraction. Analyses of ribosome distribution by denaturing RNA gel indicated that the cytosol fraction contained abundant ribosomes (28S, 18S rRNA) and tRNA (Fig. 2B [triangle]). The ER fraction contained abundant ribosomes, but lacked tRNA. Similar analyses were conducted for cells fractionated by the digitonin extraction procedure illustrated in Figure 1B [triangle]. As depicted in Figure 2C [triangle], the cytosol and membrane fractions display the expected relative enrichment of cytosol (ERK2, Hsp90) and ER membrane markers (TRAPα), respectively. Additionally, lactate dehydrogenase activity was used to assess the purity of the ER fraction attained by digitonin extraction. As was found following mechanical cell fractionation, LDH activity was enriched in the cytosol fraction, with negligible (<2.5%) LDH activity recovered in the ER fraction.

Marker protein analysis of subcellular fractions obtained by mechanical homogenization or detergent-based fractionation of Jurkat or J558 cells. (A) Samples of homogenate, cytosol, or purified ER membrane fractions, obtained as depicted in Figure 1A [triangle] ...

To examine the degree of mRNA and ribosome cross-contamination between the ER and cytosol fractions prepared by mechanical homogenization, cell homogenates were supplemented with either biosynthetically labeled free ribosomes, radiolabeled in vitro transcribed mRNA, or radiolabeled total J558 poly(A)+ RNA (data not shown), and fractionated as per Figure 1A [triangle]. The purpose of these experiments was to directly examine whether artifactual redistribution of ribosomes or mRNAs accompanied mechanical homogenization and velocity flotation. As detailed in Table 1 [triangle], modest levels of added free ribosomes or mRNAs (<10%) were recovered in the ER membrane fraction isolated by isopycnic flotation, indicating that the described fractionation procedure efficiently preserves the in situ distribution of mRNA and ribosomes between the two polysome fractions (Table 1 [triangle]). These data are consistent with previous analyses of mRNA and ribosomal cross-contamination during subcellular fractionation (Mueckler and Pitot 1981).

Marker ribosome/RNA distribution during cell fractionation

In a second series of studies, the efficiency with which the digitonin extraction procedure yields the segregation of free and membrane-bound polysomes was examined, first by electron microscopy. In these experiments, J558 cells were subject to glutaraldehyde fixation prior to or after digitonin-based fractionation, and the efficiency with which the soluble contents (free ribosomes) were depleted was examined (Fig. 3A–F [triangle]). Low-magnification micrographs of control and digitonin-extracted J558 cells are depicted in Figure 3, A and B [triangle], respectively. As is readily evident (Fig. 3B [triangle]), the detergent extraction procedure efficiently depletes the cells of the electron-dense cytosol component and results in dilation of the ER compartment. These conclusions are better detailed in the high-magnification images presented in Figure 3, C–F [triangle]. Typical high-magnification sections from control (non-digitonin-extracted) cells are shown in Figure 3, C and D [triangle]. In these micrographs, the ER profiles are seen as cisternal compartments bearing surface-associated ribosomes (arrowheads), with a dense, protein-rich lumen (*). Cytoplasmic polysomes are readily evident as electron-dense particles in the protein-rich regions surrounding the ER cisternae. For scalar comparison, the inset (Fig. 3D [triangle]) depicts a single ER profile with bound ribosomes (arrowhead), lumenal contents (*), and cisternal protein aggregates, presumably IgG (arrow). Typical high-magnification micrographs of the digitonin-extracted cells are shown in Figure 3, E and F [triangle]. As in the control cells, the ER profiles are readily apparent as membrane-bound compartments bearing surface-associated ribosomes (arrowheads), dilated cisternae with, presumably, secretory cargo in the process of translocation (*), and the observed large protein aggregates (arrow). These features are again depicted in the inset (Fig. 3F [triangle]), showing the characteristic profiles of the ER of digitonin-extracted cells. Direct comparisons of Figure 3, C versus E [triangle], clearly indicate that the cytosol compartment of the digitonin-treated cells is essentially devoid of free polysomes.

Electron micrographic analysis of digitonin-based cell fractionation. J558 murine plasmacytoma cells were fixed in glutaraldehyde and processed for thin-section electron microscopy either (A) prior to or (B) following digitonin extraction of the cytosol ...

In summary, two independent fractionation procedures were performed, using both a T-cell and a B-cell line, to yield highly enriched ER-membrane-bound and free, soluble polysomes. In subsequent experiments, the partitioning of mRNAs between the polysome fractions was examined.

mRNA partitioning on free and bound polysomes: mechanical homogenization

S1 nuclease protection assays (NPA) were used to analyze the mRNA content of the free and ER-membrane-bound polysome pools obtained by homogenization/density fractionation (Fig. 1A [triangle]). This method provides a quantitative measure of mRNA levels in a given fraction (Malo 1990). Oligonucleotide probes were designed to hybridize with mRNAs encoding representative members of three classes of protein: soluble (GAPDH, Hsp90, and LDH), ER resident membrane (Sec61α and calnexin), and ER resident lumenal (BiP, calreticulin, and GRP94). Analyses of mRNA distribution were made on the basis of total RNA content; as the rRNA component of a polysome fraction provides >95% of the total RNA, this basis of comparison yields the mRNA distribution between an essentially identical quantity of free or membrane-bound ribosomes.

To determine the NPA detection limits and confirm assay linearity, a fixed concentration of oligonucleotide probe was incubated with increasing amounts of total RNA, the reaction was subsequently digested with S1 nuclease, and the nuclease-protected hybridization products were quantified following separation on denaturing acrylamide gels. As expected, the assays displayed high sensitivity, with detection limits at ~5 μg of total RNA and high linearity (correlation coefficients of >0.97; data not shown).

Using the procedures described above, the subcellular distribution of individual mRNAs in the cytosol and rough ER polysome fractions of Jurkat and J558 cells was determined. Samples of 10 μg of total RNA from the two polysome pools were analyzed. This quantity of total RNA yielded readily quantifiable hybridization products for each mRNA assayed and, by virtue of the number of ribosomes present in this quantity of total RNA (~1012; Martin et al. 1969), provides an accurate statistical measure of the distribution of a given mRNA between the two pools. As depicted in Figure 4 [triangle], Sec61α and calnexin were highly enriched in the membrane-bound fraction, as may be predicted for those mRNAs encoding signal sequences. mRNAs for resident proteins of the ER lumen, including BiP, calreticulin, and GRP94, were also highly enriched on membrane-bound polysomes. In present models of signal recognition particle-dependent trafficking of mRNA–ribosome complexes to the ER, mRNAs encoding soluble proteins would be expected to partition on free polysomes. However, and as depicted in Figure 4 [triangle] (Cytosol), mRNAs encoding GAPDH and LDH were present in both free and membrane-bound polysomes, and mRNAs encoding Hsp90 displayed a high enrichment in the membrane-bound fraction (75%–97%). Such partitioning of mRNAs to the ER was observed in 15 independent experiments using ER derived from Jurkat and J558 cells alike, thus confirming that mRNAs encoding signal-sequence-bearing proteins and mRNAs encoding cytosolic proteins are present on the ER. Interestingly, in J558 cells, the relative partitioning of the selected mRNAs between the soluble and ER-bound polysome fractions was skewed to the bound fraction, as compared with Jurkat-derived samples. This distribution correlated with the steady-state partitioning of ribosomes between the cytosol and ER compartments of the two cell lines; in J558 cells, ~70% of total ribosomes are membrane bound, whereas in Jurkat cells, ~55% of total ribosomes are membrane associated (data not shown).

Steady-state mRNA partitioning between cytosolic and ER-bound polysomes. Total RNA was isolated from cytosol and ER membrane fractions of Jurkat and J558 cells, as indicated in Figure 1A [triangle]. For each reaction, 10 μg of RNA was incubated ...

Further analyses were done to assess whether mRNA association with the ER was an indirect consequence of protein synthesis, that is, whether ER-associated mRNAs were “tethered” to the ER via the ribosome/nascent polypeptide chain complex. This question was addressed by using protein synthesis inhibitors that cause polysome breakdown and thus physically separate nascent chains from the ribosome. Inhibitors tested included puromycin and pactamycin, which use different mechanisms to block mRNA translation and induce termination of protein synthesis. Puromycin, a tRNA analog, causes premature termination of protein synthesis, whereas pactamycin blocks a late stage of initiation, thereby allowing for run-off translation. Cells were treated with either puromycin or pactamycin, under conditions sufficient to elicit polysome breakdown. Purified ER fractions were then isolated from these inhibitor-treated cells, and nuclease protection analyses of the eight mRNAs listed above were performed. In these studies, all mRNAs tested were found to maintain their association with the ER in the presence or absence of protein synthesis inhibitors, indicating that the association of mRNA with the ER membrane is not solely dependent on the nascent polypeptide chain (data not shown).

Additional experiments were performed to determine whether the NPA products depicted in Figure 4 [triangle] were derived from intact, functional mRNAs, or, alternatively, represented membrane-bound mRNA degradation intermediates. To examine mRNA integrity, total poly(A)+ RNA was purified from the membrane-bound polysome pool and translated in a reticulocyte lysate translation system. Using antibodies against Hsp90, full-length Hsp90 protein was readily discernible by immunoprecipitation from the total translation products, indicating that the full-length, functional mRNA was present in the ER-bound mRNA pool (data not shown).

mRNA distribution on membrane-bound polysomes: cDNA microarray analysis

To examine further mRNA partitioning on membrane-bound polysomes, the composition of total and membrane-bound polysome RNA pools was compared by cDNA microarray analysis. In these experiments, the total cellular RNA and the total ER-associated RNA fractions obtained from Jurkat cells were used to generate radiolabeled cDNA probes that were subsequently hybridized to Clontech human 1.2K cDNA arrays (1176 genes). By accepted standards, and in recognition of the inherent limitations in strict quantification of cDNA microarray data, a default signal criterion of twofold above background was used to score significant representation on these arrays. Interestingly, if the hybridization products obtained from the ER-RNA pool were examined in a purely qualitative manner—without the arbitrary twofold-over-background restriction—385 of the 389 hybridization products could be readily identified (Fig. 5A [triangle]). These data indicate that mRNAs encoding soluble proteins are, indeed, broadly represented on the ER fraction; accurate quantification of such distributions requires, however, protocols such as S1 nuclease digestion.

Analysis by cDNA microarray of mRNA distribution between free and ER-membrane-bound polysomes of Jurkat cells. Total and ER-membrane-bound RNA fractions were obtained from Jurkat cells, as described in the legend to Figure 1A [triangle], and used to generate ...

To take a more quantitative approach to analysis of these data, mRNAs were classified based on their relative enrichment in the ER-RNA pool, as compared with total RNA (Fig. 5B [triangle]). In this classification method, accepted microarray standards were applied, therefore only mRNAs displaying a signal greater than twofold over background are represented in Figure 5B [triangle]. Thus, whereas a qualitative assessment of the microarray data displayed that 98.9% of all mRNAs represented in the total RNA pool are also found in the ER-RNA pool at some level (Fig. 5A [triangle]), a more quantitative examination demonstrated that 83% of mRNAs from the total mRNA pool are present in the ER-RNA pool (Fig. 5B [triangle]).

In examining the relative distribution of mRNAs between the total and ER-associated RNA pools, mRNAs were categorized into three classes: mRNAs present in the ER-associated pool but below the limit of detection in total RNA (ER-membrane-enriched); mRNAs present in total RNA but absent in ER-associated RNA (cytosol-enriched); and mRNAs present in both pools, but enriched fourfold or greater in ER-associated RNA (ER membrane and cytosol; Fig. 5B [triangle]; Table 2 [triangle]). In Table 2 [triangle], 10 representative mRNAs from each class are listed. mRNAs found exclusively in the ER-associated fraction encode, as expected, secretory and integral membrane proteins (Fig. 5B [triangle]; Table 2 [triangle], left column). mRNAs present in total RNA but absent from ER-associated RNA encode typical soluble proteins such as protein kinases (Table 2 [triangle], middle column). Of particular interest were those mRNAs present in both RNA pools but enriched fourfold or greater in the ER-associated fraction. Present among these messages were mRNAs encoding the expected secretory and integral membrane proteins, as well as noncanonically partitioned mRNAs encoding soluble proteins. A representative sample of such noncanonically partitioned mRNAs is listed (Table 2 [triangle], right column) and includes mRNAs encoding β-catenin, cytochrome p450, protein kinase C β, and rκB DNA-binding protein (Table 2 [triangle]). These data lend further support to the observations reported in Figure 4 [triangle], and are consistent with earlier reports in which cDNA/RNA kinetic hybridization methods indicated that ~90% of mRNAs present on free ribosomes could be detected on membrane-bound ribosomes (Fig. 5A [triangle]; Mueckler and Pitot 1981).

Enrichment of mRNAs in free and ER-bound polysomes

Subcellular mRNA partitioning: in situ hybridization

One potential weakness in the results reported above is that the cell fractions were derived by protocols requiring mechanical dissolution of the cell architecture. We thus sought additional methodologies, not requiring cell fractionation, for examining mRNA partitioning in the cell. Single-cell in situ hybridization allows identification of subcellular mRNA distributions. Moreover, because the method can be performed on intact, fixed cells, it represents an important control for the fractionation studies reported above. For these experiments, three mRNAs were examined: that encoding GRP94, which is highly enriched in the ER membrane fraction of fractionated cells; GAPDH, which displays a predominately (but not uniquely) soluble enrichment; and Hsp90, which displays noncanonical partitioning on the ER. Experiments were performed using NIH-3T3 fibroblasts and digoxygenin-labeled antisense RNA probes, with optical sections obtained by widefield microscopy. To aid in the analysis of mRNA distribution patterns, the ImageJ software surface plot algorithm was used to generate 3D renderings of the probe distribution. As expected, mRNAs encoding GRP94 display a typical, perinuclear and reticular ER pattern (Fig. 6A [triangle]), with the ImageJ rendering providing a graphical display of the perinuclear, ER partitioning of this mRNA species (Fig. 6A [triangle], far right panel). In contrast, mRNAs encoding GAPDH, a soluble protein, were distributed rather homogeneously throughout the cell, including the perinuclear region (Fig. 6B [triangle]). A representative surface plot of these data depicts a pancellular mRNA distribution (Fig. 6B [triangle], far right panel). Significantly, transcripts for Hsp90, which encode a soluble protein lacking a signal sequence, were clearly concentrated in the perinuclear region of the cell, yielding a pattern similar to that seen for GRP94 (Fig. 6, cf. A and C [triangle]). These data were not unique to murine (NIH-3T3) cells; identical distribution patterns were observed in HEK293 cells, a human embryonic cell line (T. Zheng and C.V. Nicchitta, unpubl.).

Single-cell in situ hybridization analysis of mRNA distribution in intact cells. NIH-3T3 cells were fixed and processed for in situ hybridization, as described in Materials and Methods. Cells were probed with digoxygenin-labeled riboprobes against mRNAs ...

We performed optical sectioning after in situ hybridization to examine further the partitioning of Hsp90, GAPDH, and GRP94 mRNAs in the cell. Using this technique, it was clear that a subpopulation of all three mRNAs occupy a discrete, perinuclear location in the cell. This is shown in Figure 7 [triangle] for mRNA encoding GAPDH, which, in the majority of the focal planes, is dispersed throughout the cell (Fig. 6 [triangle]). In Figure 7 [triangle], panels A and B represent optical sections nearing the central plane of the nucleus, where a perinuclear concentration of mRNAs can be detected, as can the more disperse staining pattern illustrated in Figure 6 [triangle]. The micrographs in Figure 7 [triangle], C and D, represent the central nuclear plane, where a very distinct perinuclear staining pattern is evident. In these panels, the contributions of the mRNA hybridization products that are not within the focal plane are evident as a pale, diffuse signal throughout the cell. From these results, it appears reasonable to suggest that the ER component of the GAPDH distribution seen by cell fractionation (Fig. 4 [triangle]) derives from the perinuclear GAPDH mRNA subpopulation detected by in situ hybridization.

Perinuclear localization of GAPDH mRNAs in NIH-3T3 cells. NIH-3T3 cells were processed for single-cell in situ hybridization analyses of mRNA encoding GAPDH, as described in the legend to Figure 6 [triangle]. Optical sections (A,B) nearing the central ...

mRNA partitioning on membrane-bound polysomes: digitonin cell fractionation protocol

To examine further mRNA partitioning between ER-bound and free polysomes, tissue culture cells were fractionated by selective release of the cytosol contents upon digitonin-mediated disruption of the plasma membrane. Following isolation of the total RNA, ER-bound RNA, and soluble RNA populations, Northern blot analyses were performed. The results of a series of experiments performed in Jurkat cells are depicted in Figure 8A [triangle]. Consistent with results obtained by mechanical homogenization, polysome fractions obtained by digitonin extraction displayed an enrichment of mRNAs encoding ER resident lumenal proteins (GRP94, BiP) in the ER-membrane-bound fraction B (Fig. 8A [triangle]). Significantly, the relative partitioning of GAPDH and Hsp90 mRNAs between the ER-membrane bound (B) and free (F) polysome populations of digitonin-extracted cells directly mirrored that seen in fractions obtained from mechanically homogenized cells (Fig. 5 [triangle]), further indicating that mRNAs encoding proteins that lack signal sequences or other topogenic determinants are partitioned on membrane-bound ribosomes.

Membrane-bound ribosomes are engaged with mRNAs encoding cytosolic proteins. (A) Analysis of mRNA distribution in the cell fractions obtained by detergent fractionation. Total cell (T), membrane-bound (B), and cytosolic, or free (F), fractions were obtained ...

We sought to determine whether those mRNAs displaying noncanonical partitioning to the ER were polysome associated, and thus translationally active. To this end, the cytosol-depleted cells (Fig. 3B,E [triangle]) were solubilized and fractionated by velocity sedimentation on a continuous sucrose gradient under conditions sufficient to resolve free, monosome-associated, and polysome-associated mRNAs (Fig. 8B [triangle]). As depicted in Figure 8B [triangle], the monosome fraction was identified by the sequential appearance (low to high gradient density) of the small (18S), large (28S), and couplet (18S + 28S) rRNAs, with the polysome component evident in the faster-migrating gradient fractions. Northern blot analyses of total RNA from each fraction demonstrated that mRNAs encoding GAPDH and Hsp90 were present in the polysome fraction of membrane-bound ribosomes. From these data, and those presented in Figures 4 [triangle] and 5 [triangle], it is apparent that mRNAs encoding soluble proteins can be translated on membrane-bound ribosomes. For a subset of such mRNAs (i.e., Hsp90), translation on membrane-bound ribosomes may represent the preferred subcellular site of synthesis. As further evidence of the polysome association of these mRNAs, their distribution on density gradients was analyzed following induction of polysome breakdown, as occurs in response to addition of the protein synthesis initiation inhibitor pactamycin (Seiser and Nicchitta 2000). In the experiments depicted in Figure 8, C and D [triangle], Jurkat cells were treated with either cycloheximide (Fig. 8C [triangle]), to stabilize polysomes, or pactamycin (Fig. 8D [triangle]), to elicit polysome breakdown, and the mRNA and ribosome distributions were analyzed. In control cells (cycloheximide-treated) mRNAs encoding GAPDH, LDH, and GRP94 were recovered almost entirely in the faster-migrating polysome fractions, consistent with their functional assembly into translating ribosomes. Following induction of polysome breakdown, a substantial monosome/ribosomal subunit peak was observed, immediately preceding a minor disome/trisome region. In all cases, the relative migration of the assayed mRNAs shifted accordingly to the slower-migrating monosome/disome peak. That a population-wide shift in the mRNA distribution occurred upon induction of polysome breakdown strongly supports the argument that the indicated mRNAs exist predominately in functional association with ribosomes.


We examined the distribution of mRNAs between endoplasmic-reticulum-bound and free ribosomes and report that mRNAs encoding soluble proteins are broadly represented on ER-membrane-bound ribosomes. In addition, a subset of mRNAs encoding soluble proteins were found to be both partitioned to, and translated on, ER-membrane-bound ribosomes. cDNA microarray analysis of the membrane-bound mRNA population identified the predicted abundance of mRNAs encoding secretory/integral membrane proteins. The microarray data also indicated that the majority of mRNAs, regardless of the compartmental fate of their translation product, were present at detectable levels on membrane-bound ribosomes. This unexpected observation was supported by data obtained (1) by traditional cell fractionation, in which cells were mechanically homogenized and the ER membrane fraction was recovered by isopycnic flotation; (2) by selective detergent fractionation, in which the cytosol contents of cultured cells are selectively released by permeabilization of the plasma membrane with digitonin, to yield highly enriched free and ER-membrane-bound polysomes; and (3) by single-cell in situ hybridization. In situ hybridization studies also demonstrated that mRNAs encoding soluble proteins are present in a discrete perinuclear region of the cell, regardless of their relative steady-state partitioning between free and bound ribosome populations.

The mRNA partitioning data obtained in the present study are consistent with earlier studies demonstrating, by poly(A)+ RNA–cDNA mass hybridization, that the mRNA cohort of membrane-bound ribosomes bear significant population similarities to those mRNAs found on free polysomes (Mueckler and Pitot 1981). They are also consistent with more recent studies, in which cDNA microarray screens for novel secretory and membrane-protein-encoding mRNAs demonstrated the presence of mRNAs encoding soluble proteins in the membrane fraction (Diehn et al. 2000). In all studies to date, however, the conclusion that mRNAs encoding soluble proteins are present on the endoplasmic reticulum suffered from the inescapable concern that such noncanonical distributions reflected artifacts associated with mechanical homogenization (Palade 1975). In the present study, the results obtained by two independent cell fractionation procedures, combined with analyses of the subcellular distribution of mRNAs by in situ hybridization studies, indicate that mRNAs encoding soluble proteins are present on, and in some cases partitioned to, the endoplasmic reticulum. In addition, these results demonstrate that membrane-bound ribosomes can, indeed, participate in the translation of mRNAs encoding soluble proteins. Furthermore, the observation that mRNAs encoding soluble proteins can be highly enriched on membrane-bound polysomes, again, as obtained by multiple cell fractionation procedures and in situ hybridization studies, supports the argument against redistribution/fractionation artifacts as an explanation for noncanonical mRNA distributions between the cytosol and endoplasmic reticulum fractions.

Although present views on the mechanism of mRNA partitioning to the ER membrane do not accommodate the observation that mRNAs encoding soluble proteins are broadly represented on ER polysomes, an alternative explanation for this finding is provided by recent studies on the compartmental fate of ER-bound ribosomes following the termination of protein synthesis (Potter and Nicchitta 2000a, 2002; Seiser and Nicchitta 2000; Potter et al. 2001). In these studies, it was reported that membrane-bound ribosomes remain in association with the protein translocation machinery following the termination stage of protein synthesis. In addition, membrane-bound ribosomes, residing in stable association with the protein translocation machinery of the ER, can reinitiate the translation of mRNAs encoding soluble as well as signal-sequence-bearing proteins. Moreover, translation of an mRNA encoding a soluble protein yielded the release of ribosomes from the ER (Potter and Nicchitta 2000a, 2002; Seiser and Nicchitta 2000; Potter et al. 2001). This directed trafficking event, which we term E-CRR (elongation-coupled ribosome release), can be viewed as serving a complementary function to that established for SRP. Whereas the SRP pathway serves to direct soluble ribosomes engaged in the synthesis of secretory/membrane proteins to the ER, E-CRR provides a means for mRNAs encoding soluble proteins, whose translation is initiated on the ER membrane, to be partitioned to the cytosol.

In analyzing the partitioning of mRNAs between free and membrane-bound polysomes, two primary observations are apparent. First, mRNAs encoding soluble proteins are broadly represented in the population of mRNAs present on membrane-bound polysomes (albeit many at low levels). Second, a subset of mRNAs encoding soluble proteins is highly partitioned onto membrane-bound polysomes. These observations suggest additional functions for membrane-bound ribosomes, beyond their established role in the translation and translocation of secretory and integral-membrane proteins. Such functions may include mRNA sorting between the ER membrane and cytosol compartments of the cell, as noted above. In addition, other aspects of RNA metabolism, such as regulated mRNA degradation and/or processing may be involved. The Hac1p mRNA, which displays noncanonical partitioning to the endoplasmic reticulum, is one useful example of an mRNA whose localization may serve important roles in metabolic regulation. In Saccharomyces cerevisiae, the mRNA encoding the soluble transcription factor Hac1p is enriched on ER polysomes (Diehn et al. 2000). Hac1p mRNA undergoes endonucleolytic processing in response to the induction of the unfolded protein response. In an early step in this unusual processing reaction, an ER resident endonuclease, Ire1p, cleaves an intron present in the polysome-assembled HAC1 mRNA (Sidrauski and Walter 1997; Patil and Walter 2001). Through localization of HAC1 mRNA to the ER, a regulatory processing step, intron cleavage, is confined to the site of residence of the partner endonuclease, Ire1p. With HAC1 mRNA/Ire1p as a model, perhaps other important regulatory processes would involve localization of mRNAs, such that the mRNA, an associated RNA-binding protein(s), or the translation product, would be the target of modification of an ER-localized enzyme. That Hac1p encodes a bZIP transcription factor implies that other transcription factors may undergo localized translation and/or regulatory processing on the ER.

The appearance of mRNAs in the perinuclear membrane domain may also reflect a broader process of mRNA sorting between the membrane and cytosol compartments of the cell. In this model, mRNAs, upon export from the nuclear pore complex, reside in association with the ER membrane and undergo translation on membrane-bound ribosomes. Should they encode a protein that lacks a signal sequence or transmembrane domain, such mRNAs would dissociate from the ER to complete translation in the cytosol. Such a mechanism, termed E-CRR, has been previously observed and could function in the regulation of mRNA partitioning between the cytoplasm and the ER membrane (Potter et al. 2001). It can also be considered that mRNAs, upon exiting the nuclear pore complex, are associated with the ER membrane and from this locale can undergo assembly into ribonucleoprotein particle/motor protein complexes for transport to distal compartments of the cell (Bassell et al. 1999; Oleynikov and Singer 2003). This possibility is consistent with recent data demonstrating that the ER serves as a site of localization of Staufen, an RNA-binding protein that has been implicated in mRNA transport to the cell periphery (Mallardo et al. 2003). As well, β-actin mRNA is known to undergo cytoskeleton-dependent transport to the leading edge of cells and, as demonstrated by in situ hybridization studies, displays both a leading edge and a distinct, perinuclear localization, similar to that depicted in Figures 6 [triangle] and 7 [triangle] (Sundell and Singer 1990; Kislauskis et al. 1994).

In summary, we provide multiple, complementary lines of evidence indicating that the mRNAs encoding soluble proteins can be translated on, and in some cases, partitioned to, the endoplasmic reticulum. We postulate that such noncanonical localization patterns reflect novel roles for ER-bound ribosomes in the genesis of mRNA partitioning in the cell, in metabolic regulation, such as the ER-localized processing of mRNAs or their translation products, and in the processes governing mRNA sorting between cellular compartments. These and related questions are under investigation at present.



All tissue culture reagents were obtained from Invitrogen/Life Technologies. Molecular biology reagents were acquired from Promega and New England Biolabs. Radioactive reagents were obtained from Amersham Pharmacia Biotech unless otherwise noted. Other reagents were from Sigma-Aldrich unless otherwise indicated.

Cell culture

Human T lymphocytes (Jurkat E6-1) and murine plasmacytoma (J558) were cultured in RPMI 1640 medium supplemented with 10% fetal calf serum or Dulbelcco’s Modified Eagle’s Medium (DMEM) supplemented with 10% equine serum, respectively. Cells were maintained at 37°C and 5% CO2, and were subcultured at 2–3-d intervals. Jurkat and J558 cells were harvested at a density of 8 × 105 cells/mL and 1.5 × 106 cells/mL, respectively.

Cell fractionation

Mechanical homogenization

Jurkat and J558 cells were collected by centrifugation (5 min, 500g) and resuspended in a homogenization buffer (HB) containing 10 mM KOAc, 10 mM K-HEPES (pH 7.5), 1.5 mM Mg(OAc)2, 2 mM DTT, 1 mM PMSF, and 200 U/mL RNAsin (Promega). After cooling at 4°C for 5 min, the cell suspension was removed to a Dounce homogenizer and homogenized on ice until >95% of cells were ruptured, as assayed by vital dye staining. For isolation of ER microsomes, the homogenate was mixed with 2 volumes of buffer containing 2.5 M sucrose in 150 mM KOAc, 50 mM K-HEPES (pH 7.5), and 5 mM Mg(OAc)2 (HKM); and 2-mL aliquots were overlaid with 0.75 mL of 1.3 M sucrose-HKM and 0.5 mL of 0.25 M sucrose-HKM. Gradients were centrifuged for 45 min at 500,000g in a TLA-100.3 rotor (Beckman). After centrifugation, the rough endoplasmic reticulum (ER) membrane fraction, banding at the 1.3 M/0.25 M sucrose interface, was removed, HKM buffer was added to a total volume of 3 mL, and the suspension was centrifuged for 20 min at 85,000g. The resulting membrane pellet was resuspended in 0.25 M sucrose-HKM. To generate the cytosol fraction, the cell homogenate was centrifuged in a TLS-55 rotor (Beckman) for 5 min at 1500g to sediment nuclei, mitochondria, and unbroken cells. The resulting supernatant was centrifuged for 20 min at 65,000g in a TLA-100.2 rotor (Beckman) to remove any contaminating membranes.

Detergent fractionation

Jurkat and J558 cells were collected by centrifugation as above and resuspended in a cytosol buffer containing 150 mM KOAc, 20 mM K-HEPES (pH 7.5), 2.5 mM Mg(OAc)2, 2 mM DTT, 1 mM PMSF, and 200 U/mL RNAsin on ice. Cell suspensions were immediately permeabilized by addition of digitonin (Calbiochem) to 100 μg/mL. Detergent-treated cells were incubated on ice for 5 min and promptly collected by centrifugation (5 min, 500g). Following centrifugation, the supernatant was saved, and the cell pellet was resuspended in cytosol buffer to the original volume, lacking detergent, and the centrifugation step was repeated. This cell pellet obtained in the second centrifugation step was resuspended in cytosol buffer and reserved as the ER-bound polysome fraction. The original supernatant fraction was centrifuged for 10 min at 7500g to remove contaminating cellular components, and this supernatant served as the cytosol (free polysome) fraction.

Characterization of subcellular fractions

The protein composition of the cell homogenate, cytosol, and ER membrane fractions was assessed by immunoblot using antibodies directed against TRAPα (ER resident protein); ERK2, Hsp90, and actin (cytosol); and Na+/K+ ATPase (plasma membrane). Antibodies against TRAPα were previously described (Migliaccio et al. 1992). Antibodies against ERK2 and actin were kindly provided by Vann Bennett (Duke University Medical Center), and antibodies against Na+/K+ ATPase were obtained from the Iowa Hybridoma Bank. Antibodies against Hsp90 were kindly provided by Sara Felts and David Toft (Mayo Clinic, Rochester, MN). For rRNA analyses, total RNA was separated on 1% agarose gels containing 3% formaldehyde and visualized with ethidium bromide staining.

Quantification of ribosome and mRNA cross-contamination in the final ER membrane and cytosol fractions prepared by mechanical homogenization was performed by supplementation of Jurkat and J558 homogenates with radiolabeled ribosomes or mRNA and subsequent fraction isolation, as described above. Radiolabeled Jurkat ribosomes were prepared as follows: Jurkat cell cultures were supplemented with 10 μCi/mL [3H]5,6-uridine (Moravek Biochemicals) for 18 h, harvested, and resuspended in the homogenization buffer described above for the mechanical homogenization procedure. Cells were Dounce homogenized, and centrifuged in a TLS-55 rotor (Beckman) for 5 min at 1500g to recover a postmitochondrial supernatant (PMS). To remove membranes, this PMS was centrifuged over a cushion of 0.5 M sucrose-HKM, and centrifuged for 15 min at 60,000g in a Beckman TLA-100.2 rotor To recover free ribosomes, the resulting supernatant was layered onto a cushion of 0.75 M sucrose-HKM and centrifuged for 40 min at 250,000g in a TLA-100.2 rotor. The ribosome pellet was resuspended in 0.2M sucrose-HKM, and the [3H]5,6-uridine content was determined by liquid scintillation spectrometry. Radiolabeled preprolactin (pPL) mRNA was obtained by in vitro transcription in the presence of [α-32P]UTP using the MegaScript kit (Ambion).

Electron microscopy

To J558 cells, in suspension, fixative containing 4% glutaraldehyde and 0.2% tannic acid (Electron Microscopy Sciences) in buffered mammalian Ringer’s (10 mM MOPS, 150 mM NaCl, 5 mM KCl, 1 mM CaCl2, 1.5 mM MgCl2 at pH 7.0) was added. Cells were pelleted after 30 min at room temperature and rinsed by centrifugation through two changes of buffered Ringer’s and two changes of 100 mM phosphate/10 mM MgCl2 buffer. Postfixation was accomplished with 1% OsO4 in 100 mM PO4, plus 10 mM MgCl2 (pH 6.0) on ice for 30 min. Cells were subsequently rinsed and pelleted three times in deionized water. Cells were block stained in aqueous 2% uranyl acetate for 30 min, rinsed, and pelleted twice with deionized water. Cells were then dehydrated in a graded ethanol series (50%–75%–100%) and embedded in Araldite 506 (Electron Microscopy Sciences). Gray-to-silver sections were cut with a Diatome diamond knife and mounted on thin carbon films. Sections were stained with aqueous 2% KMnO4, followed by Sato Lead stain (Electron Microscopy Sciences; Reedy and Reedy 1985).

RNA analysis

RNA isolation

Total RNA was isolated from cells by addition of 5 mL of Trizol reagent (Invitrogen/Life Technologies) to ~1 × 107 cells in a packed volume of 0.5 mL. To obtain RNA from subcellular fractions, Trizol was added to membranes or cytosol in a ratio of 10:1 (v/v), and the RNA was isolated by isopropanol precipitation, washed in 75% ethanol, and resuspended in DEPC H2O. RNA concentrations were determined by UV spectrometry.

Nuclease protection assays

Nuclease protection assays were performed using DNA oligonucleotide probes consisting of 28 to 45 complementary bases, with 10 noncomplementary bases at the 3′-ends. The sequence of each oligonucleotide was unique for the target mRNA and common to both human and murine coding regions. The synthesized DNA probes were 5′-end-labeled with [32P]ATP using T4 polynucleotide kinase (Promega) and purified by electrophoresis in 8 M urea/12% acrylamide gels.

Total RNA from each cell fraction was mixed with 30 fmole of labeled probe and 20 μL of hybridization buffer containing 80% formamide, 100 mM sodium citrate, 300 mM NaOAc, 1 mM EDTA, 20 mM NaCl, and 10 mM HEPES (pH 7.5). The solution was incubated at 30°C overnight. To degrade unhybridized probe, 200 μL of a nuclease digestion mix containing 280 mM NaCl, 50 mM NaOAc, 4.5 mM ZnSO4, and 25 U of S1 nuclease (Promega) was added. After incubation at 30°C for 20 min, the digestion reaction was stopped by placing the tubes on ice and adding 100 μL of buffer containing 4 M NH4OAc, 50 mM EDTA (pH 8.0), and 100 μg/mL salmon sperm DNA. The intact DNA–RNA complexes were precipitated in 2 volumes of 100% ethanol, recovered by centrifugation, and washed in 75% ethanol. The final pellet was resuspended in formamide/SDS sample buffer and heated to 95°C for 5 min. Samples were resolved on 8 M urea/12% acrylamide gels, and subsequently fixed and dried. Radiolabeled hybridization products were viewed on a BAS 1000 phosphorimager (FujiFilm Medical Systems) and quantified using the MacBAS v.2.0 image analysis software (FujiFilm Medical Systems).

cDNA array analysis

Microarray analyses were performed with Atlas Human 1.2 Arrays (Clontech) as previously described (Tenenbaum et al. 2000). Briefly, RNA was extracted from purified Jurkat ER microsomes and used to produce reverse-transcribed probes. A pooled set of primers complementary to the genes represented on the array was used for the reverse-transcription probe synthesis. After hybridization, the array membrane was washed, visualized by phosphorimager, and analyzed using ATLASIMAGE 1.01 software (Clontech). The signal for any given gene was calculated as the average of the signals from the two duplicate cDNA spots. A default external background setting was used in conjunction with a background-based signal threshold to determine gene signal significance. For all data reported herein, a minimum intensity of twofold over background was the significance threshold. Comparisons between multiple cDNA array experiments were performed by normalization against the average composite array intensity (global normalization).

Northern blots

Northern blot analyses were conducted by standard protocols (Sambrook et al. 1989). First, 10 μg of total RNA per sample was separated on agarose gels containing 3% formaldehyde and transferred to Hybond charged nitrocellulose membranes (Amersham Pharmacia Biotech). Membranes were probed with a KpnI fragment of a canine GRP94 cDNA (975 bp), a BamHI/KpnI fragment of human GAPDH cDNA (498 bp), or an EcoR1 fragment of Hsp90 (623 bp), all of which were internally labeled with [32P]CTP using a random hexanucleotide primer kit (Boehringer-Mannheim). Hybridization products were detected by phosphorimager analysis.

Polysome analysis

Jurkat cells were treated with either 200 μM cycloheximide, to stabilize polysome structure, or with 200 nM pactamycin (the kind gift of Pharmacia/Upjohn, Kalamazoo, MI), to break down polysomes, at 37°C for 15 min. Cells were then harvested, and free polysomes were released by digitonin extraction, as described above. Permeabilized cells were lysed in buffer containing 150 mM KOAc, 20 mM K-HEPES (pH 7.5), 2.5 mM Mg(OAc)2, 200 μM cycloheximide, 2 mM DTT, 1 mM PMSF, 80 U/mL RNasin, 1% Nikkol, and 0.5% deoxycholate. For polysome analyses, 2 × 107 cell equivalents of the free-polysome-depleted cells (ER-membrane-bound polysomes) were loaded onto continuous 15%–50% sucrose gradients containing 400 mM KOAc, 25 mM K-HEPES (pH 7.5), 15 mM Mg(OAc)2, 200 μM cycloheximide, 0.1% Nikkol, and 5 U/mL RNasin. The gradients were centrifuged for 3 h at 150,000g in the SW41 rotor (Beckman). Gradient fractions were collected manually, and the absorbance of each fraction at 260 nm was measured. RNA was isolated from gradient fractions using Trizol reagent.

Single-cell in situ hybridization

NIH-3T3 cells were cultured on 18-mm acid-washed glass coverslips in DMEM, 10% FBS at 37°C to 50%–80% confluency. Cells were rinsed in PBS and fixed by addition of 4% paraformaldehyde in PBS at room temperature for 10 min. All solutions were prepared using DEPC-treated water. Following fixation, cells were washed three times in PBS at room temperature and subsequently treated at room temperature for 10 min in triethanolamine/acetic anhydride (0.1 M triethanolamine at pH 8.0, containing a 1:400 dilution of reagent-grade acetic anhydride). Cells were then washed in 1× SSC at room temperature for 5 min. Cells were permeabilized by addition of 0.2 M HCl at room temperature for 10 min, and washed twice in PBS. Following permeabilization, cells were prehybridized in hybridization buffer (10% Dextran, 2 mM vanadyl-ribonucleoside complex, 0.02% RNase-free BSA, 0.1% Escherichia coli tRNA, 2× SSC, 50% formamide) at room temperature for 5 h, followed by probe addition (final concentration of 50 ng/50 μL hybridization buffer) and hybridization at 55°C overnight. Cells were rinsed with 0.2× SSC and washed in 0.2× SSC at 55°C for 1 h. For probe detection, cells were blocked in 3% BSA/PBS-T at room temperature for 1 h, followed by addition of anti-digoxygenin-alkaline phosphatase conjugate (Roche), diluted to a final concentration of 1:500 in 3% BSA in PBS-T, at room temperature for 2 h. Cells were rinsed in PBS-T, and washed four times in PBS-T for 10 min. Following rinsing, cells were then washed in 200 mM Tris (pH 8.2), 1.5 mM levamisole, and incubated in Vector Red substrate solution at room temperature for 30 min, following the manufacturer’s instructions (Vector Laboratories). Cells were rinsed and mounted in Difco FA mounting media. For all experiments, control assays with detection system alone (anti-digoxygenin-alkaline phosphatase conjugate + substrate), nonsense probe, or sense probe were performed. No specific staining was observed for any control conditions. Images were collected on a Zeiss Axiophot microscope using a Photometrics CCD camera and processed using IPLab software.

Probe synthesis

For probe synthesis, 2 μL of 10× transcription buffer, 2 μL of 0.1 M DTT, 2 μL of digoxygenin nucleotide mix (Roche), 0.5 μL of RNase inhibitor (100 U/μL), 1 μL of T7 or SP6 polymerase, and 2 μg of linearized DNA were combined in a final volume of 20 μL, and incubated at 37°C for 3 h. Reactions were terminated by addition of DNase and incubation at 37°C for 15 min. The following probe sizes were used: 3′-UTR-Hsp90, ~350 bp; GAPDH, ~500 bp; 5′-UTR-GRP94, ~90 bp.


The authors gratefully acknowledge S. Tenenbaum for stimulating discussion and members of the Nicchitta laboratory, particularly Rebecca Dodd, Robyn Reed, and Samuel Stephens, for helpful comments and criticism. Brigid Hogan is also acknowledged for her insightful analysis, critique, and editing of the manuscript. This work was supported by NIH Grants DK47897 (C.V.N.), CA79907 (J.D.K.), and AI46451 (J.D.K.). The publication costs of this article were defrayed in part by payment of page charges. This article must therefore be hereby marked “advertisement” in accordance with 18 USC section 1734 solely to indicate this fact.


  • Adam, S.A., Marr, R.S., and Gerace, L. 1990. Nuclear protein import in permeabilized mammalian cells requires soluble cytoplasmic factors. J. Cell Biol. 111: 807–816. [PMC free article] [PubMed]
  • Bassell, G.J., Oleynikov, Y., and Singer, R.H. 1999. The travels of mRNAs through all cells large and small. FASEB J. 13: 447–454. [PubMed]
  • Blobel, G. and Dobberstein, B. 1975a. Transfer of proteins across membranes. I. Presence of proteolytically processed and unprocessed nascent immunoglobulin light chains on membrane-bound ribosomes of murine myeloma. J. Cell Biol. 67: 835–851. [PMC free article] [PubMed]
  • ———. 1975b. Transfer of proteins across membranes. II. Reconstitution of functional rough microsomes from heterologous components. J. Cell Biol. 67: 852–862. [PMC free article] [PubMed]
  • Brooks, J.C. and Treml, S. 1984. Effect of trifluoperazine and calmodulin on catecholamine secretion by saponin-skinned cultured chromaffin cells. Life Sci. 34: 669–674. [PubMed]
  • Diaz, R., Wileman, T.E., Anderson, S.J., and Stahl, P. 1989. The use of permeabilized cells to study the ion requirements of receptor–ligand dissociation in endosomes. Biochem. J. 260: 127–134. [PMC free article] [PubMed]
  • Diehn, M., Eisen, M.B., Botstein, D., and Brown, P.O. 2000. Large-scale identification of secreted and membrane-associated gene products using DNA microarrays. Nat. Genet. 25: 58–62. [PubMed]
  • Dunn, L.A. and Holz, R.W. 1983. Catecholamine secretion from digitonin-treated adrenal medullary chromaffin cells. J. Biol. Chem. 258: 4989–4993. [PubMed]
  • Kislauskis, E.H., Zhu, X., and Singer, R.H. 1994. Sequences responsible for intracellular localization of β-actin messenger RNA also affect cell phenotype. J. Cell Biol. 127: 441–451. [PMC free article] [PubMed]
  • Lingappa, V.R. and Blobel, G. 1979. Early events in the biosynthesis of secretory and membrane proteins: the signal hypothesis. Rec. Prog. Horm. Res. 36: 451–475. [PubMed]
  • Mallardo, M., Deitinghoff, A., Muller, J., Goetze, B., Macchi, P., Peters, C., and Kiebler, M.A. 2003. From the Cover: Isolation and characterization of Staufen-containing ribonucleoprotein particles from rat brain. Proc. Natl. Acad. Sci. 100: 2100–2105. [PMC free article] [PubMed]
  • Malo, M.S. 1990. A rapid and efficient method for transcript mapping. Nucleic Acids Res. 18: 6159. [PMC free article] [PubMed]
  • Martin, T.F., Rolleston, F.S., Low, R.B., and Wool, I.G. 1969. Dissociation and reassociation of skeletal muscle ribosomes. J. Mol. Biol. 43: 135–149. [PubMed]
  • Mechler, B. and Rabbitts, T.H. 1981. Membrane-bound ribosomes of myeloma cells. IV. mRNA complexity of free and membrane-bound polysomes. J. Cell Biol. 88: 29–36. [PMC free article] [PubMed]
  • Migliaccio, G.N., Nicchitta, C.V., and Blobel, G. 1992. The signal sequence receptor, unlike the signal recognition particle receptor, is not essential for protein translocation. J. Cell Biol. 117: 15–25. [PMC free article] [PubMed]
  • Mueckler, M.M. and Pitot, H.C. 1981. Structure and function of rat liver polysome populations. I. Complexity, frequency distribution, and degree of uniqueness of free and membrane-bound polysomal polyadenylate-containing RNA populations. J. Cell Biol. 90: 495–506. [PMC free article] [PubMed]
  • Oleynikov, Y. and Singer, R.H. 2003. Real-time visualization of ZBP1 association with β-actin mRNA during transcription and localization. Curr. Biol. 13: 199–207. [PubMed]
  • Palade, G. 1975. Intracellular aspects of the process of protein synthesis. Science 189: 347–358. [PubMed]
  • Patil, C. and Walter, P. 2001. Intracellular signaling from the endoplasmic reticulum to the nucleus: The unfolded protein response in yeast and mammals. Curr. Opin. Cell Biol. 13: 349–355. [PubMed]
  • Potter, M.D. and Nicchitta, C.V. 2000a. Regulation of ribosome detachment from the mammalian endoplasmic reticulum membrane. J. Biol. Chem. 275: 33828–33835. [PubMed]
  • ———. 2000b. Ribosome-independent regulation of translocon composition and Sec61α conformation. J. Biol. Chem. 275: 2037–2045. [PubMed]
  • ———. 2002. Endoplasmic reticulum-bound ribosomes reside in stable association with the translocon following termination of protein synthesis. J. Biol. Chem. 277: 23314–23320. [PubMed]
  • Potter, M.D., Seiser, R.M., and Nicchitta, C.V. 2001. Ribosome exchange revisited: A mechanism for translation-coupled ribosome detachment from the ER membrane. Trends Cell Biol. 11: 112–115. [PubMed]
  • Rapoport, T.A., Wiedmann, M., Kurzchalia, T.V., and Hartmann, E. 1989. Signal recognition in protein translocation across the endoplasmic reticulum membrane. Biochem. Soc. Trans. 17: 325–328. [PubMed]
  • Reedy, M.K. and Reedy, M.C. 1985. Rigor crossbridge structure in tilted single filament layers and flared-X formations from insect flight muscle. J. Mol. Biol. 185: 145–176. [PubMed]
  • Sambrook, J., Fritsch, E.F., and Maniatis, T. 1989. Molecular cloning: A laboratory manual, 2nd ed. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY.
  • Seiser, R.M. and Nicchitta, C.V. 2000. The fate of membrane-bound ribosomes following the termination of protein synthesis. J. Biol. Chem. 275: 33820–33827. [PubMed]
  • Sidrauski, C. and Walter, P. 1997. The transmembrane kinase Ire1p is a site-specific endonuclease that initiates mRNA splicing in the unfolded protein response. Cell 90: 1031–1039. [PubMed]
  • Sundell, C.L. and Singer, R.H. 1990. Actin mRNA localizes in the absence of protein synthesis. J. Cell Biol. 111: 2397–2403. [PMC free article] [PubMed]
  • Tenenbaum, S.A., Carson, C.C., Lager, P.J., and Keene, J.D. 2000. Identifying mRNA subsets in messenger ribonucleoprotein complexes by using cDNA arrays. Proc. Natl. Acad. Sci. 97: 14085–14090. [PMC free article] [PubMed]
  • Walter, P. and Johnson, A.E. 1994. Signal sequence recognition and protein targeting to the endoplasmic reticulum membrane. Annu. Rev. Cell Biol. 10: 87–119. [PubMed]

Articles from RNA are provided here courtesy of The RNA Society
PubReader format: click here to try


Related citations in PubMed

See reviews...See all...

Cited by other articles in PMC

See all...


Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...