A study on increasing enzymatic stability and activity of Baliospermum montanum hydroxynitrile lyase in biocatalysis

HNL catalysis is usually carried out in a biphasic solvent and at low pH to suppress the non-enzymatic synthesis of racemic cyanohydrins. However, enzyme stability under these conditions remain a challenge. We have investigated the effect of different biocatalytic parameters, i.e., pH, temperature, buffer concentrations, presence of stabilizers, organic solvents, and chemical additives on the stability of Baliospermum montanum hydroxynitrile lyase (BmHNL). Unexpectedly, glycerol (50 mg/mL) added BmHNL biocatalysis had produced >99% of (S)-mandelonitrile from benzaldehyde, while without glycerol it is 54% ee. Similarly, BmHNL had converted 3-phenoxy benzaldehyde and 3,5-dimethoxy benzaldehyde, to their corresponding cyanohydrins in the presence of glycerol. Among the different stabilizers added to BmHNL at low pH, 400 mg/mL of sucrose had increased enzyme’s half-life more than fivefold. BmHNL’s stability study showed half-lives of 554, 686, and 690 h at its optimum pH 5.5, temperature 20 °C, buffer concentration, i.e., 100 mM citrate-phosphate pH 5.5. Addition of benzaldehyde as inhibitor, chemical additives, and the presence of organic solvents have decreased both the stability and activity of BmHNL, compared to their absence. Secondary structural study by CD-spectrophotometer showed that BmHNL’s structure is least affected in the presence of different organic solvents and temperatures.     

Immobilized Baliospermum montanum hydroxynitrile lyase catalyzed synthesis of chiral cyanohydrins

Hydroxynitrile lyase (HNL) catalyzed enantioselective CC bond formation is an efficient approach to synthesize chiral cyanohydrins which are important building blocks in the synthesis of a number of fine chemicals, agrochemicals and pharmaceuticals. Immobilization of HNL is known to provide robustness, reusability and in some cases also enhances activity and selectivity.We optimized the preparation of immobilization of Baliospermium montanum HNL (BmHNL) by cross linking enzyme aggregate (CLEA) method and characterized it by SEM. Optimization of biocatalytic parameters was performed to obtain highest % conversion and ee of (S)-mandelonitrile from benzaldehyde using CLEA-BmHNL. The optimized reaction parameters were: 20 min of reaction time, 7 U of CLEA-BmHNL, 1.2 mM substrate, and 300 mM citrate buffer pH 4.2, that synthesized (S)-mandelonitrile in ∼99% ee and ∼60% conversion. Addition of organic solvent in CLEA-BmHNL biocatalysis did not improve in % ee or conversion of product unlike other CLEA-HNLs. CLEA-BmHNL could be successfully reused for eight consecutive cycles without loss of conversion or product formation and five cycles with a little loss in enantioselectivity. Eleven different chiral cyanohydrins were synthesized under optimal biocatalytic conditions in up to 99% ee and 59% conversion, however the % conversion and ee varied for different products. CLEA-BmHNL has improved the enantioselectivity of (S)-mandelonitrile synthesis compared to the use of purified BmHNL. Nine aldehydes not tested earlier with BmHNL were converted into their corresponding (S)-cyanohydrins for the first time using CLEA-BmHNL. Among the eleven (S)-cyanohydrins syntheses reported here, eight of them have not been synthesized by any CLEA-HNL. Overall, this study showed preparation, characterization of a stable, robust and recyclable biocatalyst i.e. CLEA-BmHNL and its biocatalytic application in the synthesis of different (S)-aromatic cyanohydrins.     

Screening hydroxynitrile lyases for (R)-pantolactone synthesis

The application of (R)-hydroxynitrile lyases ((R)-HNLs) enables a simple chemo-enzymatic approach towards (R)-pantolactone synthesis. For the first time, several new recombinant almond (R)-HNL isoenzymes were compared with native HNLs from different Prunus species with respect to cyanohydrin formation from hydroxypivalaldehyde providing the chiral key precursor in HNL based (R)-pantolactone synthesis. Recombinant PaHNL5 (R-selective hydroxynitrile lyase, isoenzyme 5, from Prunus amygdalus) surpasses all other tested natural and recombinant HNL variants. At low pH even very low amounts of crude enzyme catalysed stereoselective hydroxypivalaldehyde cyanohydrin formation in water based reaction systems.     

Characterization of a novel hydroxynitrile lyase from Nandina domestica Thunb

The leaves of Nandina domestica Thunb. exhibited high hydroxynitrile lyase (HNL) activity in (R)-mandelonitrile synthesis. The specific activity of young leaves was significantly higher than that of mature leaves. We isolated two HNLs with molecular mass of 24.9 kDa (NdHNL-S) and 28.0 kDa (NdHNL-L) from the young leaves. Both NdHNLs were composed of two identical subunits, without FAD and carbohydrates. We purified NdHNL-L and revealed its enzymatic properties. The whole deduced amino acid sequence of NdHNL-L was not homologous to any other HNLs, and the specific activity for mandelonitrile synthesis by NdHNL-L was higher than that by other plant HNLs. The enzyme catalyzed enantioselective synthesis of (R)-cyanohydrins, exhibited high activity at pH 4.0, and high stability in the pH range of 3.5–8.0 and below 55°C. Thus, NdHNL-L is a novel HNL with novel amino acid sequence and has a potential for the efficient production of (R)-cyanohydrins.     

Parameters influencing asymmetric synthesis of (R)-mandelonitrile by a novel (R)-hydroxynitrile lyase from Eriobotrya japonica

(R)-Mandelonitrile was successfully synthesized by an enzymatic transcyanation reaction of benzaldehyde and acetone cyanohydrin catalyzed by a hydroxynitrile lyase from Eriobotrya japonica (EjHNL) in an aqueous-organic biphasic system. The effects of pH, temperature, organic solvent, substrate concentration and enzyme concentration on the initial activity and enantioselectivity of the enzyme were studied. Both pH and temperature had a large effect on the initial velocity and enantiomeric excess (e.e.) of the product, (R)-mandelonitrile. High enantiomeric purity of the product was observed at low pH and temperature because the non-enzymatic reaction producing racemates of mandelonitrile was almost suppressed. The optimum pH and temperature to obtain high e.e. were pH 4.0 and 10 °C, respectively. Surprisingly, the organic solvents had a significant influence on the initial velocity of the reaction but less influence on the enantiomeric purity of product. The EjHNL was very stable in ethyl acetate, diethyl ether, methyl-t-butyl ether, diisopropyl ether, dibutyl ether and hexane for 12 h. The best solvent for the highest initial velocity and e.e. was diethyl ether with an optimum aqueous phase content of 50% (v/v). The initial reaction rate increase as the aqueous phase content rose, but when the content was more than 50%, a reduction of e.e. was observed. Increasing the concentration of the substrates accelerated the initial velocity, but caused a slight decrease in the e.e. of the product. Under the optimized conditions, the conversion and e.e. of (R)-mandelonitrile for 3 h were 40 and 99%, respectively. The aqueous phase containing the enzyme also showed considerably efficient reusability for 4 batch reactions.     

Synthesis of internal re-initiation fragments of beta-galactosidase in vitro and in vivo.

Internal re-initiation polypeptides which are products of the lacZ gene of Escherichia coli have been synthesized in a DNA-directed cell-free protein synthesis system. Some of the properties of these protein fragments have been characterized. The de novo synthesized re-initiation proteins, unlike both in vitro synthesized wild-type β-galactosidase and nonsense termination fragments, are insoluble when synthesized at 37 °C, but soluble if synthesis takes place at 25 °C. The same re-initiation proteins which are made in vivo have been detected in vitro. Unlike their in vivo counterparts, which are degraded rapidly, the in vitro synthesized restart proteins are completely stable for at least one hour following their synthesis. Both in vivo and in vitro, the re-initiation proteins are not synthesized from DNA containing a wild-type Z gene, but require a specific nonsense mutation in order to be expressed. Furthermore, the location of the mutation within the Z gene is very important in determining whether or not re-initiation will occur at a given site.Several nonsense mutations which do not result in the synthesis of detectable amounts of restart protein in vivo produce specific re-initiation polypeptides in vitro. These restart proteins display many of the same properties as do those which are made both in vivo and in vitro: they are not made from wild-type DNA, and they are only made from DNA containing a specific nonsense mutation. One of these mutations is 118, which is an extreme polar mutation in vivo. Another is 545, which synthesizes a restart protein in vivo if, and only if, it is coupled with a secondary mutation, π(1). π(1) thus appears to be necessary for the synthesis of a particular re-initiation protein in vivo but unnecessary for the synthesis of the same protein in vitro. The efficiencies of re-initiation vary at the different sites, but in all cases are less than the initiation frequency at the start of the gene. The experiments thus show that when complicating factors, such as polarity and protein degradation, are eliminated, translational re-initiation can be detected at many sites in the lacZ gene.     

Arabidopsis dynamin-related protein 1A polymers bind, but do not tubulate, liposomes

The Arabidopsis dynamin-related protein 1A (AtDRP1A) is involved in endocytosis and cell plate maturation in Arabidopsis. Unlike dynamin, AtDRP1A does not have any recognized membrane binding or protein-protein interaction domains. We report that GTPase active AtDRP1A purified from Escherichia coli as a fusion to maltose binding protein forms homopolymers visible by negative staining electron microscopy. These polymers interact with protein-free liposomes whose lipid composition mimics that of the inner leaflet of the Arabidopsis plasma membrane, suggesting that lipid-binding may play a role in AtDRP1A function. However, AtDRP1A polymers do not appear to assemble and disassemble in a dynamic fashion and do not have the ability to tubulate liposomes in vitro, suggesting that additional factors or modifications are necessary for AtDRP1A’s in vivo function.     

Structural and functional analysis of hydroxynitrile lyase from Baliospermum montanum with crystal structure,molecular dynamics and enzyme kinetics

Hydroxynitrile lyases (HNLs) catalyze degradation of cyanohydrins to hydrogen cyanide and the corresponding ketone or aldehyde. HNLs can also catalyze the reverse reaction, i.e., synthesis of cyanohydrins. Although several crystal structures of S-selective hydroxynitrile lyases (S-HNLs) have been reported, it remains unknown whether and how dynamics at the active site of S-HNLs influence their broad substrate specificity and affinity. In this study, we analyzed the structure, dynamics and function of S-HNL from Baliospermum montanum (BmHNL), which has an α/β hydrolase fold. Two crystal structures of BmHNL, apo1 and apo2, were determined at 2.55 and 1.9 Å, respectively. Structural comparison between BmHNL (apo2) and S-HNL from Hevea brasiliensis with (S)-mandelonitrile bound to the active site revealed that hydrophobic residues at the entrance region of BmHNL formed hydrophobic interactions with the benzene ring of the substrate. The flexible structures of these hydrophobic residues were confirmed by a 15 ns molecular dynamics simulation. This flexibility regulated the size of the active site cavity, enabling binding of various substrates to BmHNL. The high affinity of BmHNL toward substrates containing a benzene ring was also confirmed by comparing the kinetics of BmHNL and S-HNL from Manihot esculenta. Taken together, the results indicated that the flexibility and placement of the residues are important for the broad substrate specificity of S-HNLs.     

Functional expression and sub-cellular localization of the early aflatoxin pathway enzyme Nor-1 in Aspergillus parasiticus

Aflatoxin biosynthesis in Aspergillus parasiticus requires at least 17 enzyme activities (from acetate). Although the activities of most aflatoxin biosynthetic enzymes have been established, the mechanisms that govern transport and sub-cellular localization of these enzymes are not clear. We developed plasmid constructs that express Nor-1 fused to a green fluorescent protein reporter (EGFP) to monitor transport and localization of this early pathway enzyme in real time in Aspergillus parasiticus. Plasmids expressing EGFP fused to Nor-1 were introduced into A. parasiticus B62 (carries non-functional Nor-1). Transformants were screened for increased aflatoxin accumulation (restored Nor-1 activity) on coconut agar medium and for EGFP expression using fluorescence microscopy. Increased aflatoxin accumulation was confirmed by TLC and ELISA. Nor-1 fused to EGFP at either the N- or C- terminus functionally complemented non-functional Nor-1 in B62 and increased aflatoxin synthesis to wild-type (N-terminus) or lower levels (C-terminus). We detected full-length Nor-1 fusion proteins in transformants with increased aflatoxin accumulation (Western blot) and determined that the expression plasmid integrated at the nor-1 locus in these cells (Southern blot). Confocal laser scanning microscopy (CLSM) demonstrated that Nor-1 fusion proteins localized in the cytoplasm and vacuoles of fungal hyphae grown on aflatoxin-inducing solid media for 48 h; control EGFP (no Nor-1) did not localize to vacuoles until 72 h. The highest rate of aflatoxin synthesis coincided with the highest rate of transport of Nor-1 fusion proteins to the vacuole strongly suggesting that Nor-1 is synthesized in the cytoplasm and transported to the vacuole to carry out an early step in aflatoxin synthesis.     

Catabolism of Cyanogenic Glycosides by Purified Vicianin Hydrolase from Squirrel's Foot Fern (Davallia Trichomanoides Blume)

Vicianin hydrolase, which catalyzes the hydrolysis of vicianin (K_m, 4.9 millimolar) to (R)-mandelonitrile and vicianose at an optimum pH of 5.5, was extensively purified from the young fronds and fiddleheads of the squirrel's foot fern (Davallia trichomanoides Blume) using DEAE-cellulose and Ultrogel HA chromatography. The native molecular weight of the enzyme was 340,000, and the isoelectric point was 4.6 to 4.7. SDS-PAGE analysis yielded three polypeptides with molecular weights of 56,000, 49,000, and 32,500. The enzyme hydrolyzed only a narrow range of glycosides and, among cyanogenic glycosides, exhibited a strict requirement for (R)-epimers and a preference for disaccharides over monosaccharides. (R)-Amygdalin, (R)-prunasin and p-nitrophenyl-β-d-glucoside were hydrolyzed at 27, 14, and 3%, respectively, of the rate of vicianin hydrolysis. Mixed substrate studies showed that (R)-vicianin, (R)-prunasin, and p-nitrophenyl-β-d-glucoside competed for the same active site. The enzyme was significantly inhibited by castanospermine, δ-gluconolactone, and p-chloromercuriphenylsulfonate. Failure to recognize concanavalin A-Sepharose 4B and to stain with periodic acid-Schiff reagent indicated that the enzyme was not a glycoprotein.     

Optimisations and Challenges Involved in the Creation of Various Bioluminescent and Fluorescent Influenza A Virus Strains for In Vitro and In Vivo Applications

Bioluminescent and fluorescent influenza A viruses offer new opportunities to study influenza virus replication, tropism and pathogenesis. To date, several influenza A reporter viruses have been described. These strategies typically focused on a single reporter gene (either bioluminescent or fluorescent) in a single virus backbone. However, whilst bioluminescence is suited to in vivo imaging, fluorescent viruses are more appropriate for microscopy. Therefore, the idea l reporter virus varies depending on the experiment in question, and it is important that any reporter virus strategy can be adapted accordingly. Herein, a strategy was developed to create five different reporter viruses in a single virus backbone. Specifically, enhanced green fluorescent protein (eGFP), far-red fluorescent protein (fRFP), near-infrared fluorescent protein (iRFP), Gaussia luciferase (gLUC) and firefly luciferase (fLUC) were inserted into the PA gene segment of A/PR/8/34 (H1N1). This study provides a comprehensive characterisation of the effects of different reporter genes on influenza virus replication and reporter activity. In vivo reporter gene expression, in lung tissues, was only detected for eGFP, fRFP and gLUC expressing viruses. In vitro, the eGFP-expressing virus displayed the best reporter stability and could be used for correlative light electron microscopy (CLEM). This strategy was then used to create eGFP-expressing viruses consisting entirely of pandemic H1N1, highly pathogenic avian influenza (HPAI) H5N1 and H7N9. The HPAI H5N1 eGFP-expressing virus infected mice and reporter gene expression was detected, in lung tissues, in vivo. Thus, this study provides new tools and insights for the creation of bioluminescent and fluorescent influenza A reporter viruses.     

Fusion of a Short Peptide that Binds Immunoglobulin G to a Recombinant Protein Substantially Increases Its Plasma Half-Life in Mice

We explore a strategy to substantially increase the half-life of recombinant proteins by genetic fusion to FcIII, a 13-mer IgG-Fc domain binding peptide (IgGBP) originally identified by DeLano and co-workers at Genentech [DeLano WL, et al. (2000) Science 287∶1279–1283]. IgGBP fusion increases the in vivo half-life of proteins by enabling the fusion protein to bind serum IgG, a concept originally introduced by DeLano and co-workers in a patent but that to the best of our knowledge has never been pursued in the scientific literature. To further investigate the in vitro and in vivo properties of IgGBP fusion proteins, we fused FcIII to the C-terminus of a model fluorescent protein, monomeric Katushka (mKate). mKate-IgGBP fusions are easily expressed in Escherichia coli and bind specifically to human IgG with an affinity of ∼40 nM and ∼20 nM at pH 7.4 and pH 6, respectively, but not to mouse or rat IgG isotypes. mKate-IgGBP binds the Fc-domain of hIgG1 at a site overlapping the human neonatal Fc receptor (hFcRn) and as a consequence inhibits the binding of hIgG1 to hFcRn in vitro. High affinity binding to human IgG also endows mKate-IgGBP with a long circulation half-life of ∼8 hr in mice, a 75-fold increase compared to unmodified mKate. Thus, IgGBP fusion significantly reduces protein clearance by piggybacking on serum IgG without substantially increasing protein molecular weight due to the small size of the IgGBP. These attractive features could result in protein therapies with reduced dose frequency and improved patient compliance.     

Luciferase-based reporter system for in vitro evaluation of elongation rate and processivity of ribosomes

The elongation step of translation is a key contributor to the abundance, folding and quality of proteins and to the stability of mRNA. However, control over translation elongation has not been thoroughly investigated. In this study, a Renilla–firefly luciferase fusion reporter system was further developed to investigate the in vitro elongation rate and processivity of ribosomes independent of the initiation and termination steps. The reporter mRNA was constructed to contain a single ORF encoding in-frame Renilla luciferase, a specific domain moiety and firefly luciferase. Such a reporter structure enables the quantitative and individual evaluation of the synthesis of a specific domain. As a proof of principle, the synthesis of three protein domains of different lengths and structures was analyzed. Using a cell-free translation assay, both the elongation rate and processivity of ribosomes were shown to vary depending on the domain synthesized. Additionally, a stalling sequence consisting of ten rare arginine codons notably reduced the elongation rate and the processivity of the ribosomes. All these results are consistent with the previously known dynamics of elongation in vivo. Overall, the methodology presented in this report provides a framework for studying aspects that contribute to the elongation step of translation.     

The biosynthesis of alkaline phosphatase with a particulate fraction of Escherichia coli. The mechanism of inhibition by inorganic phosphate

1. By digitonin lysis of penicillin spheroplasts of Escherichia coli a particulate fraction P₁ was previously obtained that supported the sustained synthesis of alkaline phosphatase when supplied with amino acids, nucleotide triphosphates and other cofactors. This P₁ fraction, when subjected to mild ultrasonic treatment in the presence of sucrose and Mg²⁺, yielded the P₁(S) fraction, consisting of integrated particulate subcellular particles containing DNA and RNA. 2. The P₁(S) fraction from E. coli K10 wild type (R⁺₁R⁺₂P⁺) grown under repressed conditions supported the immediate synthesis of alkaline phosphatase in vitro. The synthesis occurred in phases. The first was followed by a lag, and then there was a linear rapid phase that continued for at least 3hr. Actinomycin D inhibited the appearance of the second phase. It was concluded that the particles are programmed to synthesize enzyme even when prepared from repressed cells, and therefore that synthesis of the specific messenger RNA for alkaline phosphatase in vivo was not inhibited when the bacteria were grown in an excess of inorganic phosphate. 3. Phosphate inhibited synthesis of enzyme to the same extent with the P₁(S) fractions of two constitutive strains as with the P₁(S) fraction of the wild-type strain. 4. Inorganic phosphate inhibited amino acid incorporation with the P₁(S) fraction and also inhibited enzyme synthesis in vitro. The effect on amino acid incorporation could be partially overcome by adding Mn²⁺ to the incubation mixtures. However, Mn²⁺ inhibited the synthesis of alkaline phosphatase. Also, inhibition of the incorporation of [³²P]CTP into RNA was overcome by Mn²⁺. The effect of phosphate on amino acid uptake was most probably due to a phosphorolysis of RNA by polynucleotide phosphorylase, also present in the P₁(S) fraction. This phosphorolysis may be responsible for the instability of messenger RNA in vitro and in vivo. 5. Phosphate also specifically inhibited the formation of alkaline phosphatase, since it did not affect markedly the induced formation of β-galactosidase by the same P₁(S) fraction. The specific effect is attributed to the prevention of formation of the enzymically active dimer from precursors, a Zn²⁺-dependent reaction. It is suggested that the repression of the synthesis of alkaline phosphatase in vivo in the wild-type strain was the sum of these two effects.     

Purification and characterization of the nitrilase from Alcaligenes faecalis ATCC 8750 responsible for enantioselective hydrolysis of mandelonitrile

A nitrilase that converts racemic mandelonitrile to R-(—)-mandelic acid was purified to apparent homogeneity from a cell extract of Alcaligenes faecalis ATCC 8750. The molecular weight of this enzyme was estimated to be 32,000±2,000 from SDS-PAGE and that of the native enzyme 460,000±30,000 from HPLC gel filtration. The enzyme preferentially hydrolyzed substituted aliphatic nitriles, in particular benzyl cyanide and its p-substituted compounds, but hydrolyzed aromatic nitriles only with difficulty. The amino-terminal amino acids were sequenced and their sequences compared with those of other nitrilases. The purified enzyme had a pH optimum of 7.5 and an optimum temperature range of 40 to 45°C. The enzyme was inhibited by various thiol reagents. It hydrolyzed racemic mandelonitrile, producing optically pure R-(—)-mandelic acid and ammonia without the concomitant production of mandelamide, evidence that this nitrilase is highly enantioselective for R-mandelonitrile.     

Towards Multiparametric Fluorescent Imaging of Amyloid Formation: Studies of a YFP Model of α-Synuclein Aggregation

Misfolding and aggregation of proteins are characteristics of a range of increasingly prevalent neurodegenerative disorders including Alzheimer's and Parkinson's diseases. In Parkinson's disease and several closely related syndromes, the protein α-synuclein (AS) aggregates and forms amyloid-like deposits in specific regions of the brain. Fluorescence microscopy using fluorescent proteins, for instance the yellow fluorescent protein (YFP), is the method of choice to image molecular events such as protein aggregation in living organisms. The presence of a bulky fluorescent protein tag, however, may potentially affect significantly the properties of the protein of interest; for AS in particular, its relative small size and, as an intrinsically unfolded protein, its lack of defined secondary structure could challenge the usefulness of fluorescent-protein-based derivatives. Here, we subject a YFP fusion of AS to exhaustive studies in vitro designed to determine its potential as a means of probing amyloid formation in vivo. By employing a combination of biophysical and biochemical studies, we demonstrate that the conjugation of YFP does not significantly perturb the structure of AS in solution and find that the AS-YFP protein forms amyloid deposits in vitro that are essentially identical with those observed for wild-type AS, except that they are fluorescent. Of the several fluorescent properties of the YFP chimera that were assayed, we find that fluorescence anisotropy is a particularly useful parameter to follow the aggregation of AS-YFP, because of energy migration Förster resonance energy transfer (emFRET or homoFRET) between closely positioned YFP moieties occurring as a result of the high density of the fluorophore within the amyloid species. Fluorescence anisotropy imaging microscopy further demonstrates the ability of homoFRET to distinguish between soluble, pre-fibrillar aggregates and amyloid fibrils of AS-YFP. Our results validate the use of fluorescent protein chimeras of AS as representative models for studying protein aggregation and offer new opportunities for the investigation of amyloid aggregation in vivo using YFP-tagged proteins.     

Synthesis of Optically Active Cyanohydrins Using R-Oxynitrilase in a Liquid-Liquid Biphasic System: Part 1: An Industrially useful Procedure

An improved method is presented for preparing optically active cyanohydrins in high yield and high enantioselectivity, using R-oxynitrilase from almonds as the catalyst in a biphasic water-organic solvent system. Reaction conditions for two representative substates have been studied with respect to conversion, optical purity and enzyme stability.

The suitability of the procedure for industrial use was substantiated by the possibility of reusing the enzyme-containing aqueous phase for the synthesis of R-mandelonitrile.     

Photoreactivating enzyme from Euglena and the control of its intracellular level

Photoreactivating (PR) enzyme activity has already been demonstrated by us in cell-free extracts of Euglena gracilis var. bacillaris Pringsheim using the Hemophilus transformation assay. This activity can also be detected in extracts using a direct non-biological assay for the photorepair of thymine dimers in DNA. PR enzyme is found in extracts of both wild-type cells and cells of an aplastidic mutant, W₃BUL, lacking detectable chloroplast DNA, indicating that the PR enzyme is neither coded nor translated exclusively in the chloroplast, but is probably coded in the nucleus and translated in the cytoplasm. Growing cultures of wild-type cells manifest a large increase in PR enzyme activity in vitro upon entering stationary phase. This correlates with the increased photoreactivability of chloroplast inheritance in vivo in stationary phase cells, previously found for Euglena, and suggests that a substantial part of the newly synthesized PR enzyme is available to repair plastid DNA. When dark-grown nondividing wild-type cells are exposed to light, there is a large increase in the specific activity of PR enzyme measured in vitro. This increase is prevented by cycloheximide but not by chloramphenicol or streptomycin, indicating that the enzyme is synthesized on 87s cytoplasmic ribosomes rather than 68s chloroplast ribosomes. Wavelengths of light effective for PR of chloroplast DNA in vivo are also effective for the light induction of PR enzyme. A brief illumination (45 min) of dark-grown nondividing wild-type cells triggers the synthesis of PR enzyme which continues in the absence of light. Growing cultures of W₃BUL also exhibit a preferential synthesis of PR enzyme in the staionary phase of growth, but the specific activity in vitro is consistently ten times higher than that of wild-type. Dark-grown non-dividing cultures of W₃BUL also show a cycloheximide-sensitive light induction of PR enzyme synthesis which, however, is dependent on the continued presence of light. The light induction of PR enzyme synthesis can be regarded as the induction of an enzyme by one of its substrates.