Regulation of the activity of chloroplast translational initiation factor 3 by NH2- and COOH-terminal extensions.

The mature form of the chloroplast translational initiation factor 3 (IF3chl) from Euglena gracilis consists of an internal region homologous to prokaryotic IF3 flanked by long NH2- and COOH-terminal extensions. Sequences in these extensions reduce the activity of the homology domain in promoting initiation complex formation with chloroplast mRNAs and 30 S ribosomal subunits. A series of deletions of the NH2- and COOH-terminal extensions of IF3chl were constructed and tested for their effects on the activity of the homology domain. About half of the inhibitory effect arises from sequences within 9 residues of the junction between the NH2-terminal extension and the homology domain. The remaining inhibitory effect is the result of sequences in the COOH-terminal extension. The equilibrium constant governing the binding of the homology domain of IF3chl to 30 S subunits is estimated to be 1.3 x 10(7) M-1. Sequences close to the junction of the NH2-terminal extension and the homology domain reduce this binding constant about 10-fold. Sequences in the COOH-terminal extension have a similar negative effect. The negative effects of these two regions are cumulative, resulting in a 100-fold reduction of the binding constant. The 9 residues at the NH2-terminal extension effectively prevent the proofreading activity of IF3chl. The entire COOH-terminal extension reduces the proofreading ability by about half. These results are discussed in terms of the proposed three-dimensional structure of the homology domain of IF3chl.

Three translational initiation factors (IF1, IF2, and IF3) 1 are required for the initiation of protein synthesis in Escherichia coli (1,2). During initiation, IF3 binds to the 30 S subunit and shifts the equilibrium between the ribosome and its subunits toward dissociation (3,4). IF1 and IF2 bind to the 30 S⅐IF3 complex. The initiation factor⅐30 S complex binds the mRNA and fMet-tRNA, resulting in the formation of an unstable preinitiation complex. This complex is converted into a stable initiation complex when the initiator tRNA has been selected and codon-antidocon interaction occurs (5,6). IF3 has three major functions: 1) it binds to the 30 S subunit, preventing the joining of 50 S subunits (1)(2)(3)5); 2) it increases the affinity of IF1 and IF2 for the 30 S subunit and stimulates fMet-tRNA binding to the 30 S subunit by promoting the conversion of the preinitiation complex to the initiation complex (7,8); and 3) it proofreads the selection of fMet-tRNA at an AUG initiation codon (9 -14).
The chloroplast translational initiation factors are postulated to be functionally analogous to their E. coli counterparts. Only IF2 chl and IF3 chl from Euglena gracilis have been purified (15,16). Both of these factors are nuclear-encoded proteins in this organism (17,18). IF3 chl has been resolved into three forms, ␣, ␤, and ␥. The ␣ form has a molecular mass of about 34 kDa, whereas the ␤ and ␥ forms have molecular masses of about 45 kDa (16). In contrast, E. coli IF3 has a molecular mass of 20 kDa. IF3 chl is active on E. coli ribosomes.
A complete cDNA encoding E. gracilis IF3 chl has been cloned and sequenced (17). The molecular mass deduced from the nucleotide sequence is 58 kDa, including a signal peptide of 130 -140 residues required for localization to the chloroplast. The mature form of this factor (IF3 chl M) can be divided into three parts (Fig. 1). An NH 2 -terminal extension termed the head (Hd) region encompasses the first 140 amino acids. This region contains a proline-rich sequence followed by a (GX) 12 motif and a short acidic sequence. A middle region of about 180 amino acids shows homology to prokaryotic IF3 and is referred to as the homology (H) domain (19). Structural analysis of E. coli and Bacillus stearothermophilus IF3 indicates that this region will fold into two highly compact domains separated by a lysine-rich linker (20 -24). The COOH-terminal extension is referred to as the tail (T) region. This 64-amino acid region is rich in glutamic acid residues (17).
Previous studies have shown that both IF3 chl M and the homology domain, IF3 chl H, are active in promoting the dissociation of ribosomal subunits and in promoting initiation complex formation on E. coli ribosomes using poly(A,U,G) as an mRNA (19). However, IF3 chl M is only 10 -20% as active as IF3 chl H in promoting initiation complex formation on chloroplast 30 S ribosomal subunits using mRNAs carrying natural translational start sites for chloroplast mRNAs (19). These observations suggest that sequences in the head and tail regions of IF3 chl down-regulate the activity of this factor in initiation. In the present work, the roles of sequences in the head and tail regions in affecting the activity of IF3 chl have been examined in more detail.

EXPERIMENTAL PROCEDURES
Materials-[ 35 S]fMet-tRNA and [ 14 C]AcPhe-tRNA were prepared as described (25,26). A plasmid carrying the 5Ј untranslated leader region and the translational start site of the E. gracilis chloroplast rbcL gene fused in-frame to an internal coding region of the neomycin phosphotransferase gene was transcribed in vitro providing the mRNA, mRbcN (27). E. coli ribosomes, initiation factors, E. gracilis chloroplast 30 S subunits, IF2 chl , IF3 chl , and IF3 chl H antiserum were prepared as described (16, 19, 28 -31).
Induction and Purification of Various Derivatives of IF3 chl -Qiagen pQE vectors were used to express IF3 chl or its derivatives carrying a His tag at the COOH terminus. The regions of IF3 chl to be expressed were amplified by polymerase chain reaction using the cDNA clone described previously (17) or a derivative of this plasmid as template. Cells were grown and IF3 chl derivatives were induced as described previously (19). Induction times were as follows: 45 min for IF3 chl HdH, 20 min for IF3 chl HT, and 2-3 h for the remaining constructs. IF3 chl HdH was purified as described for IF3 chl M (19). The other forms of IF3 chl were purified using the two-step purification procedure developed for IF3 chl H (19).
Binding of IF3 chl to 30 S Subunits-The indicated concentrations of IF3 chl and chloroplast 30 S subunits were incubated in a total volume of 250 l in 50 mM Tris-HCl, pH 7.8, 10 mM dithiothreitol (DTT), 50 mM NH 4 Cl, and 10 mM MgCl 2 at room temperature for 5 min. The mixture was applied to a 5-ml 10 -30% linear sucrose gradient prepared in the same buffer except that the concentration of Tris-HCl was reduced to 10 mM. Samples were subjected to centrifugation at 48,000 rpm for 2 h in a Beckman SW50.1 rotor. Gradients were fractionated at a flow rate of 1 ml/min. Fractions (100 l) were collected from the region of the gradient containing the 30 S subunits. Aliquots (50 l) of appropriate fractions were analyzed for the amount of IF3 chl present using an ELISA (32). A standard curve for each derivative tested was determined in each experiment to allow the amount of IF3 chl present to be quantified.
Assay for Initiation Complex Formation-The abilities of IF3 chl and its derivatives to promote initiation complex formation with E. coli 70 S ribosomes using poly(A,U,G) were assayed as described (19). The abilities of IF3 chl and its derivatives to promote initiation complex formation with chloroplast 30 S subunits and mRbcN were determined as indicated (19).
Proofreading Assay-This assay has been modified from the method described in Ref. 33 for E. coli IF3. A complex carrying AcPhe-tRNA bound to chloroplast 30 S subunits (AcPhe-tRNA⅐poly(U)⅐30 S) was formed by incubation of chloroplast 30 S subunits (10 pmol) with poly(U) (2.5 g) and AcPhe-tRNA (4 pmol) in a reaction mixture (50 l) containing 50 mM Tris-HCl, pH 7.8, 10 mM dithiothreitol, 50 mM NH 4 Cl, and 15 mM MgCl 2 . After incubation at 37°C for 30 min, the mixture was diluted 2-fold with 50 mM Tris-HCl, pH 7.8, and 50 mM NH 4 Cl in the presence of different concentrations of IF3 chl or its derivatives. Mixtures were incubated for an additional 5 min at 37°C. The destabilization of the complex by IF3 chl was monitored following dilution with 1 ml of prewarmed dilution buffer (50 mM Tris-HCl, pH 7.8, 10 mM dithiothreitol, 50 mM NH 4 Cl, and 7.5 mM MgCl 2 ). These reaction mixtures were incubated at 37°C for 5 min. The amount of initiation complex remaining was determined by a nitrocellulose filter binding assay (19). A similar assay was also carried out using a complex formed with 30 S subunits (10 pmol), poly(A,U,G) (2.5 g) and fMet-tRNA (4 pmol).

Inhibitory Effects of the Head and the Tail Regions on the Activity of the Homology Domain of IF3 chl -Previous studies
have shown that the mature form of IF3 chl (IF3 chl M) is almost as active as the homology domain (IF3 chl H) in an assay that measures the ability of IF3 chl to promote the binding of fMet-tRNA to E. coli 70 S ribosomes using poly(A,U,G). However, IF3 chl M shows very poor activity in promoting the binding of fMet-tRNA to chloroplast 30 S subunits in the presence of an mRNA carrying the translational initiation region of a natural mRNA (19). This observation indicates that the head, the tail, or both have a negative effect on the activity of the homology domain of IF3 chl . To investigate which region or regions of IF3 chl M are responsible for this inhibitory effect, the activities of a series of derivatives of IF3 chl containing different parts of IF3 chl (Fig. 1) were tested. These derivatives were designed based on the structures of the two best characterized prokaryotic IF3s (from E. coli and B. stearothermophilus). The IF3s from these organisms have very similar overall three-dimensional structures (20 -22). However, B. stearothermophilus IF3 is 9 residues shorter than E. coli IF3 at the NH 2 terminus.
Chloroplast homologues of both these prokaryotic IF3s were prepared (Fig. 1). IF3 chl srH is the homologue of B. stearothermophilus IF3, whereas IF3 chl rH is the homologue of E. coli IF3. These two forms of IF3 chl differ by 9 residues at the NH 2 terminus. To test the effects of sequences in the NH 2 -terminal extension, a derivative, IF3 chl HdH, encompassing the homology domain and the entire head region was prepared. To test the effects of sequences in the COOH-terminal extension, a derivative, IF3 chl sHT, covering the homology domain and the entire tail region was prepared. Note that this derivative of IF3 chl begins at the position corresponding to the start of the B. stearothermophilus factor.
The induction of IF3 chl sHT, like IF3 chl M, results in a significant decrease in cell growth, indicating that the tail of IF3 chl is quite toxic to the cell. The induction of IF3 chl HdH has less effect on cell growth, whereas the expression of IF3 chl srH does not affect cell growth to an appreciable extent (data not shown). Each derivative of IF3 chl was purified; the derivatives were estimated to be 90 -95% pure in all cases (Fig. 2).
The activity of each construct in promoting the binding of fMet-tRNA to E. coli 70 S ribosomes was examined. As indicated in Fig. 3A, all of these forms of IF3 chl were quite active in this assay. IF3 chl HdH and IF3 chl sHT had slightly less activity than IF3 chl srH but slightly more activity than IF3 chl M. These results indicate that the head and tail had little effect on the activity of IF3 chl when E. coli 70 S ribosomes and a synthetic mRNA, poly(A,U,G), were used. These derivatives of IF3 chl were then tested for the ability to promote the binding of fMet-tRNA to chloroplast 30 S subunits using an mRNA carrying the translational initiation region of the rbcL gene. As shown previously and as indicated in Fig. 3B, IF3 chl M had only 15-20% of the activity of the homology domain of IF3 chl in this assay. The effect of sequences in the head was assessed by comparing the activity of IF3 chl HdH with IF3 chl srH (Fig. 3B). The head region reduced the activity of the homology domain by 2-fold, indicating that the head accounts for about half of the reduction in activity seen with IF3 chl M. To assess the effect of sequences in the tail, the activity of IF3 chl sHT was tested. IF3 chl sHT had about 30% of the activity seen with IF3 chl srH (Fig. 3B), indicating that sequences in the tail account for a little over half of the inhibitory effect seen in IF3 chl M.
A Small Region of the Head Is Sufficient to Confer Its Full Inhibitory Effect-The results presented above indicate that sequences in both the head and the tail of IF3 chl M have a negative effect on the ability of the homology domain to promote initiation complex formation. Additional constructs were then prepared to narrow down the inhibitory region in the head. IF3 chl erH (Fig. 1) covers the homology region and 15 residues of the head from the NH 2 terminus of IF3 chl srH (the B. stearothermophilus homologue) to the edge of (GX) 12 -acidic motif (19). IF3 chl rH, the E. coli homologue, is 9 residues longer than IF3 chl srH at the NH 2 terminus (Fig. 1). The induction of either IF3 chl rH or IF3 chl erH retards cell growth indicating that their expression is toxic to E. coli (data not shown). Both IF3 chl rH and IF3 chl erH were purified (Fig. 2, lanes 3 and 6).
IF3 chl rH and IF3 chl erH were as active as IF3 chl srH when tested on E. coli 70 S ribosomes (Fig. 4A). However, IF3 chl rH and IF3 chl erH, like IF3 chl HdH, had half the activity of IF3 chl srH when tested on chloroplast 30 S subunits (Fig. 4B).
These results indicate that only 9 residues in the NH 2 -terminal extension measured from the B. stearothermophilus factor are required to give the inhibitory effect of the entire head region. This observation is quite surprising because IF3 chl rH is the same length at the NH 2 terminus as E. coli IF3. The activity of E. coli IF3 decreases markedly without the NH 2 -terminal hexapeptide (34).
Effect of the Tail on IF3 chl srH and Additive Effects of Sequences in the Head and Tail Regions-As indicated in Fig. 3B, the tail region contributed about half of the negative regulatory effect seen with the mature form of IF3 chl . Secondary structure analysis indicated that the tail probably contains two long helices that have a high probability of forming a coiled-coil. To gain further insight into which sequences in the tail might be responsible for this result, a derivative was prepared (IF3 chl sHT/3) that contained about 1 ⁄3 of the sequences in the tail encompassing residues through the first helix (Fig. 1). The induction of IF3 chl sHT/3 had little effect on the growth of E. coli. IF3 chl sHT/3 was purified to greater than 95% purity (Fig.  2, lane 7). IF3 chl sHT/3 had essentially the same activity as IF3 chl srH when E. coli 70 ribosomes are used. When chloroplast 30 S subunits and mRbcN were used, IF3 chl sHT/3 was as active as IF3 chl srH (Fig. 5B). This result indicates that the last 2 ⁄3 of the tail region are essential for the inhibitory effect of the tail.
To test the effects from the short NH 2 -and the entire COOHterminal extension, IF3 chl HT, consisting of the homology domain surrounded by the tail and 15 residues of the head (Fig.  1), was prepared and purified (Fig. 2, lane 8). The induction of IF3 chl HT resulted in a significant decrease in cell growth and eventually appeared to cause cell lysis. IF3 chl HT had activity slightly lower than that of IF3 chl srH but the same as that of IF3 chl sHT and IF3 chl M when tested on E. coli 70 ribosomes with poly(A,U,G) (Fig. 5A). When chloroplast 30 S subunits and mRbcN were used, IF3 chl HT had the same low activity observed with IF3 chl M (Fig. 5B). These results indicate that IF3 chl HT contains all the negative regulatory elements present in IF3 chl M and that the negative effects due to sequences in the head and tail are additive.
Basis for the Inhibitory Effect of Sequences in the Head and Tail on the Activity of IF3 chl -In an attempt to understand whether the low activity of IF3 chl HT could be overcome by raising the concentrations of 30 S subunits, mRNA, or IF2, assays were carried out using different amounts of each component, separately. As indicated in Fig. 6, increasing the concentration of chloroplast 30 S subunits, mRbcN, or IF2 did not allow IF3 chl HT to increase its activity relative to the activity of IF3 chl srH. Similar results were obtained when the levels of either IF2 chl or E. coli IF2 were varied. These observations suggest that the low activity of IF3 chl HT is a complex phenomenon involving the interplay of IF3 chl with multiple components of the initiation machinery.
The activities of several derivatives of IF3 chl in promoting initiation complex formation on chloroplast 30 S subunits were tested in the presence of either the ␣ or the ␤ form of IF2 chl (data not shown) and E. coli IF2. The results of this study indicated that all of the negative regulatory effects from the head and tail regions are seen in the presence of either form of IF2 chl or E. coli IF2. The natural mRNA used above (mRbcN) carries the initiation region of the rbcL gene. This region does not have a Shine/Dalgarno sequence. Indeed, about half of the chloroplast mRNAs in E. gracilis lack a Shine/Dalgarno sequence (35,36). The negative effects of the head and tail were also tested with an mRNA carrying the translational start site for the atpH gene, which has a Shine/Dalgarno sequence just upstream of the start codon. The head and tail also inhibited the activity of the homology domain when this mRNA was used (data not shown).
Direct measurements of the abilities of various derivatives of IF3 chl to bind to chloroplast 30 S subunits were carried out using sucrose density gradient centrifugation. For these experiments, the appropriate derivatives of IF3 chl were incubated with chloroplast 30 S subunits. The bound factor was separated from the free factor by sucrose gradient centrifugation. The amount of IF3 chl bound to the 30 S subunit was quantified FIG. 4. Small regions of the head are sufficient to confer the full inhibitory effect of the head. A, activities of IF3 chl erH (q) and IF3 chl rH (ƒ) compared with IF3 chl HdH (E) and IF3 chl srH () in promoting initiation complex formation on E. coli ribosomes. A blank (0.05 pmol) representing the amount of fMet-tRNA bound in the absence of IF3 chl has been subtracted from each value. B, stimulation of initiation complex formation on chloroplast 30 S ribosomal subunits in the presence of 10 pmol of mRbcN (27). A blank (0.1 pmol) representing the amount of fMet-tRNA bound in the absence of IF3 chl has been subtracted from each value.

FIG. 5. Effect of the tail on IF3 chl srH and additive effects of sequences in the head and tail.
A, activity of IF3 chl sHT/3 (E) and IF3 chl HT (Ⅺ) compared with IF3 chl sHT (ƒ), IF3 chl srH (q), and IF3 chl M () in promoting initiation complex formation on E. coli ribosomes. A blank (0.05 pmol) representing the amount of fMet-tRNA bound in the absence of IF3 chl has been subtracted from each value. B, stimulation of initiation complex formation on chloroplast 30 S ribosomal subunits in the presence of 10 pmol of mRbcN (27). A blank (0.1 pmol) representing the amount of fMet-tRNA bound in the absence of IF3 chl has been subtracted from each value.
using an enzyme-linked immunosorbent assay. The amount of IF3 chl bound was calculated based on a standard curve providing a measure of the response of each IF3 chl derivative to the antibody. The standard curves for each of the derivatives are quite similar (Table I). This observation was expected because the antibodies were raised against the homology domain. The total amount of each derivative of IF3 chl bound to 30 S subunits and the estimated K obs are indicated in Table I. IF3srH had the highest affinity for 30 S subunits, with a K obs ϭ 1.3 ϫ 10 7 M Ϫ1 . This value is similar to the affinity of E. coli IF3 for E. coli 30 S subunits (K ϭ 2.5 ϫ 10 7 M Ϫ1 ) (37). IF3 chl erH and IF3 chl sHT bound to 30 S subunits with about 10-fold lower affinity than IF3 chl srH. IF3 chl HT showed the lowest ability to bind, with a K obs approximately 100-fold lower than that of IF3 chl srH. These results suggest that the small NH 2 -terminal extension region and the full tail interfere with the ability of IF3 chl to bind to 30 S subunits. IF3 chl HT, which contains both regions, showed the lowest affinity for chloroplast 30 S subunits. Because the low activity of IF3 chl HT was not overcome by raising the concentration of 30 S subunits (Fig. 6), the head and tail must still down-regulate the activity of IF3 chl after this factor binds to 30 S subunits.
Proofreading Ability of IF3 chl and Its Derivatives-In E. coli, IF3 is believed to proofread the selection of the initiator tRNA and the AUG start codon (10 -12, 38). One procedure for monitoring this function is to examine the ability of IF3 to destabilize preformed initiation complexes consisting of 30 S ribosomal subunits carrying poly(U) and AcPhe-tRNA (33,38). This destabilization affects all initiation complexes with the exception of those containing the initiator fMet-tRNA at an AUG  codon, which remains resistant to the destabilization induced by IF3 (39).
The ability of IF3 chl to proofread in the chloroplast system was examined by testing its ability to promote the dissociation of a preformed 30 S⅐poly(U)⅐AcPhe-tRNA complex. IF3 chl srH has the greatest ability to destabilize the 30 S⅐poly(U)⅐AcPhe-tRNA complex (Fig. 7A). This observation is in agreement with its greater ability to bind to 30 S subunits and to promote initiation complex formation. IF3 chl HT shows the least activity in this assay, however, it still has some ability to proofread. Surprisingly, all of the reduced proofreading ability seen with IF3 chl HT is also observed with IF3 chl erH. IF3 chl sHT has more than half of the proofreading ability of IF3 chl srH. These observations suggest that sequences in the head region interfere with proofreading to a greater extent than those in the tail region. The ability of derivatives of IF3 chl to discriminate between initiation complexes containing fMet-tRNA was also tested (Fig. 7B). None of the derivatives that were examined destabilized the binding of fMet-tRNA to 30 S subunits. Indeed, some stimulation of fMet-tRNA binding was observed with IF3 chl srH even under the dilute conditions used in this assay. This stimulation presumably reflects the high activity of this derivative in promoting initiation complex formation. DISCUSSION IF3 chl from E. gracilis is the first organellar IF3 that has been cloned and over-expressed. The results presented here indicate that a 9-residue sequence in the head region of IF3 chl and sequences in the tail play a negative regulatory role in promoting initiation complex formation on chloroplast ribosomes. Structural studies on E. coli and B. stearothermophilus IF3 (20 -22) indicate that both factors fold into two compact domains separated by a lysine-rich linker (Fig. 8). These two domains are formed by the independent folding of sequences in the NH 2 -terminal and COOH-terminal halves of the protein.
The center of mass of the two domains are separated by about 45 Å (22). The crystal structures of the NH 2 -domain and COOH-domain of B. stearothermophilus IF3 (Fig. 8) indicate that both the NH 2 and COOH termini are oriented toward the central linker. The linker is highly basic and could play an important role in interacting with 16S rRNA, tRNA, or mRNA (21). Both domains of IF3 are thought to be involved in ribosome binding (22). Because the head and tail of IF3 chl extend toward the linker region, they may interact with specific residues in the linker region, which are important for the binding of IF3 to the 30 S subunit and for its function after it binds.
The negative influence of sequences in the tail and of residues near the junction between the head and the homology domain on the activity of IF3 chl suggests that there must be a mechanism by which these effects may be modulated in vivo.
One attractive hypothesis is that these regions down-regulate the intrinsic activity of IF3 chl and that other factors in the chloroplast alleviate this inhibition under appropriate conditions. This idea is based on numerous observations that indicate that chloroplast protein synthesis is regulated in response FIG. 7. Proofreading abilities of IF3 chl derivatives. A, the abilities of IF3 chl srH (), IF3 chl erH (q), IF3 chl sHT ([itrio), and IF3 chl HT (E) to promote the dissociation of a preformed 30 S⅐poly(U)⅐ AcPhe-tRNA complex were analyzed in the presence of the indicated amount of IF3 chl as described under "Experimental Procedures." The value for 100% complex remaining is 1.0 pmol. B, the abilities of derivatives of IF3 chl to destabilize the initiation complexes containing fMet-tRNA was tested under similar conditions. The value of 100% remaining bound represents 0.12 pmol. to light and mRNA-specific trans-acting factors (40 -47). Because IF3 chl is required for the translation of all mRNAs, it could play a key role in modulating the activity of the chloroplast translational system as a whole, for example, in response to light or developmental signals. In addition, trans-acting factors bound to specific chloroplast mRNAs could interact with IF3 chl to recruit this factor for the translation of a specific mRNA. The most logical region of IF3 chl to interact with such putative regulatory proteins is the head. The rationale for this idea is as follows. The head has an unusual amino acid sequence and, presumably, structure. It contains a Pro-rich region reminiscent of many protein-protein interaction sites and their flanking regions (48 -54). Prominent examples of such sites include proteins recognized by SH3 domains or the WW motif found in many proteins participating in regulatory cascades. The (GX) n motif (glycine-X motif, where X indicates a large basic hydrophobic residue) following the Pro-rich region would be expected to have significant structural flexibility and could function as a flexible hinge region.
In a working model (Fig. 9), IF3 chl is visualized as being in a low activity state due to the negative effects from the extensions on the homology domain (Fig. 9). In this low activity state, the activity of IF3 chl would limit the rate of translation in the chloroplast to some basal amount. This level would, presumably, allow the chloroplast to maintain the amounts of critical proteins at minimum required levels. In the presence of appropriate environmental signals (for example, in conditions promoting photosynthesis), a regulatory factor interacts with the head on IF3 chl relieving the inhibitory effects and allowing the homology domain to become fully active. A protein affecting the activity of IF3 chl could potentially act either in general, increasing the overall rate of chloroplast protein synthesis, or more specifically, promoting the translation of specific mRNAs. In the latter case, IF3 chl can be envisioned as playing a role in tying mRNA-specific trans-acting factors to the general translational machinery. Current efforts are designed to gain insight into the factors that modulate the activity of IF3 chl and, thus, the rate of chloroplast protein synthesis.