Characterization of a High Affinity Phytochelatin Synthase From The Cd-Utilizing Marine Diatom Thalassiosira pseudonana

CHARACTERIZATION OF A HIGH AFFINITY PHYTOCHELATIN SYNTHASE FROM THECD-UTILIZING MARINE DIATOM THALASSIOSIRA PSEUDONANA1

Tiffany Gupton-Campolongo

Department of Biological and Environmental Engineering, Cornell University, Ithaca, New York, 14853, USA

Leonardo M. Damasceno2, Anthony G. Hay

Department of Microbiology, Cornell University, Ithaca, New York, 14853, USA

and Beth A. Ahner3

Department of Biological and Environmental Engineering, Cornell University, Ithaca, New York, 14853, USA

Phytochelatin synthase (PC synthase) is the enzymethat catalyzes the production of phytochelatins, peptidesof the structure (c-Glu-Cys)n-Gly, where n = 2–11, fromthe sulfhydryl-containing tripeptide glutathione, inresponse to elevated metal exposure. Biochemicalutilization of Cd in the marine diatom Thalassiosiraweissfloggi, as well as unusually high ratios of PC to Cd insome Thalassiosira species including T. pseudonanaHasle et Heimdal, motivated the characterization ofT. pseudonana PC synthase 1 (TpPCS1). This enzyme isthe product of one of three genes in the T. pseudonanagenome predicted to encode for a PC synthase basedon its homology to canonical PC synthases previouslyexamined. TpPCS1 was cloned, expressed in Escherichiacoli and purified under both aerobic and anaerobicconditions. TpPCS1 exhibits several characteristics thatset it distinctly apart from the well-studied PC synthase,Arabidopsis thaliana PCS1 (AtPCS1). It is extremelysensitive to oxidation, which suppresses activity, and it isreadily inhibited by the addition of Cd in the absence ofthiolate ligands. TpPCS1 also has significantly greateraffinity for one of its key substrates, the bis-glutathionato-Cd complex. TpPCS1 kinetics is bestdescribed by a ternary complex model, as opposed tothe ping-pong model used to describe AtPCS1 kinetics.The findings indicate that although the function ofTpPCS1 is synonymous to that of AtPCS1, its divergentbiochemistry suggests adaptation of this enzyme to thedistinct trace metal chemistry of the marineenvironment and the unique physiological needs ofT. pseudonana.

Key index words: cadmium; metal detoxification; phy-tochelatin; phytochelatin synthase; Thalassiosirapseudonana

Abbreviations: Cd�GS2, bis-glutathionato-cadmium;DTT, dithiothreitol; GSH, glutathione; OPA,o-phthaldialdehyde; PC, phytochelatin; S-methyl-GS,

S-methylglutathione; TCEP, tris(2-carboxyethyl)phos-phine

The enzyme phytochelatin (PC) synthase catalyzesthe synthesis of phytochelatins (PCs), small metal-binding peptides of the structure (c-Glu-Cys)n-Gly,from a glutathione (GSH) precursor. This is donein response to toxic metals such as Cd, Hg, Pb orexcess essential metals, Cu, Zn or Fe, with Cd beingthe strongest elicitor of PCs in vivo and in vitro(Kondo et al. 1984, Grill et al. 1985, Vatamaniuket al. 2000, Loscos et al. 2006). PC synthases areconstitutively transcribed and catalyze a dipeptidyltranspeptidase reaction in which the terminal gly-cine of a GSH (c-Glu-Cys-Gly) precursor is cleavedand released resulting in c-glutamylcysteine (c-Glu-Cys), which is then transferred onto a second GSHpeptide or a PCn chain to form PC2 or PCn+1,respectively (Vatamaniuk et al. 2004).Much of what is known about the activation and

mechanism of PC synthases has been determinedfrom studies on AtPCS1, a PC synthase cloned fromthe higher plant, Arabidopsis thaliana. Structurally,the N-terminal end of AtPCS1 (amino acids 1–221)has been identified as the catalytic domain, sincePC synthesis still occurs in the absence of theC-terminal end (amino acids 222–485), albeit at alower rate (Ruotolo et al. 2004, Romanyuk et al.2006). In addition, site-directed mutagenesis of theAtPCS1 conserved N-terminal residues Cys-56, His-162 and Asp-180, resulted in the termination ofenzyme activity, confirming that these residues arerequired for PC synthesis (Vatamaniuk et al. 2004,Romanyuk et al. 2006). The crystal structure of thePC synthase-like enzyme from the cyanobacterium,Nostoc sp. (NsPCS), later confirmed that the con-served cysteine, histidine, and aspartate residuesform a catalytic triad (Vivares et al. 2005) similar tothat found in the papain superfamily of cysteineproteases (Rea 2006). Kinetic modeling (Vat-amaniuk et al. 2000) and enzyme acylation studies(Vatamaniuk et al. 2004) revealed that the mechanism

1Received 21 May 2012. Accepted 6 August 2012.2Present address: Centro de Pesquisas Ren�e Rachou/FIOCRUZ,

Belo Horizonte, Brazil.3Author for correspondence: e-mail [email protected].

J. Phycol. 49, 32–40 (2013)© 2012 Phycological Society of AmericaDOI: 10.1111/jpy.12022

32

of AtPCS1 is best described by the ping-pong model,in which PC2 synthesis occurs in two steps using twosubstrates, GSH and the bis-glutathionato-Cd(Cd�GS2) complex. In the first step, the low affinitysubstrate GSH is cleaved resulting in an enzyme acylintermediate, AtPCS1-c-Glu-Cys, and a free glycine.In the second step, the c-Glu-Cys is transferred fromthe enzyme to the amino terminus of one GSH mol-ecule of Cd�GS2, the high affinity substrate, result-ing in the release of PC2 and GSH with Cd nowpresumably bound to PC2.

Despite the fact that PCs have been found in allhigher plants, some fungi, most algae, and a fewnematodes (see review by Rea et al. 2004), only afew PC synthases have been purified and studied.Some marine diatoms evolved to live in the extre-mely low metal concentration environment of theopen ocean have been shown to use Cd as a nutri-ent (Lane et al. 2005) and therefore may contain aPC synthase with special attributes. Indeed, the mar-ine diatom, Thalassiosira pseudonana, exhibits in vivoPC accumulation that is distinct from that of higherplants in that it has a high intracellular cysteinyl-PC:Cd ratio upon Cd exposure (Ahner et al. 2002) andincreases PC accumulation when Zn is limiting andCd is present (Ahner et al. 1998). When cultured inmedia-containing sub-nanomolar Cd concentrations,a maximal 80:1 cysteinyl-PC:Cd ratio was measured inT. pseudonana cells (Ahner et al. 2002). This ratio farexceeds the 4:1 ratio observed in plant cell suspen-sion cultures (Strasdeit et al. 1991), the 2:1 ratio mea-sured in plant cell extracts and extract assays (Grillet al. 1985, 1987) and the ratios found in other mar-ine algae with the exception of T. weissflogii, a closerelative of T. pseudonana (Ahner et al. 1995). PCshave been hypothesized to serve a role in metalhomeostasis in this organism (Xu et al. 2008).

Here we report the cloning and expression ofone of three putative PC synthase genes fromT. pseudonana. The subsequent gene product,TpPCS1, was purified and compared to AtPCS1 viaactivity assays to determine the extent of biochemi-cal similarity between the two enzymes. We foundthat TpPCS1 has a much higher affinity for Cd�GS2than AtPCS1, is highly sensitive to oxidation, whichsubstantially reduces enzyme activity, and is moresensitive to Cd inhibition than AtPCS1. We also

show that TpPCS1 displays kinetics consistent with aternary complex model, in which both substratesmust bind to the enzyme prior to the release of anyproducts. This is in contrast to the ping-pong mech-anism proposed for AtPCS1, which describes a reac-tion in which a stable intermediate is formed.Differences in the primary and secondary structuralpredictions for the enzymes are discussed.

MATERIALS AND METHODS

Algae growth and vector construction for expression of TpPCS1,TpPCS2, TpPCS3 and AtPCS1 in Escherichia coli. Thalassiosirapseudonana (strain CCMP1335; Provasoli Guillard NationalCenter for Culture of Marine Phytoplankton, East Boothbay,Maine, USA) cells were cultured in artificial seawater (Priceet al. 1989) to late exponential growth and harvested usingvacuum filtration. RNA was extracted with Tri Reagent(Sigma-Aldrich, St. Louis, MO, USA). First strand cDNA syn-thesis was performed using the Promega RT System (Pro-mega, Madison, WI, USA) with oligo dT15 primers. TheT. pseudonana genome is predicted to encode for three poten-tial PC synthases having protein IDs 334418, 261250, 263132(http://genome.jgi-psf.org/Thaps3/Thaps3.home.html),which correspond to TpPCS1, TpPCS2, and TpPCS3, respec-tively. Dr. Mak Saito (Woods Hole Oceanographic Institu-tion) kindly provided predicted coding region sequences foreach gene. TpPCS1, TpPCS2, and TpPCS3 were amplified fromthe cDNA template by PCR using gene specific primers(Table 1). These primers were designed to incorporate NdeIand EcoRI restriction sites upstream of the start codon anddownstream of the stop codon, respectively. Similarly, AtPCS1cDNA was generated from leaf tissue (var. Columbia) and wasamplified by PCR using gene specific primers (Table 1). Theseprimers incorporated NheI and EcoRI upstream of the startcodon and downstream of the stop codon, respectively. Reverseprimers added a sequence encoding the FLAG fusion peptide(DYDDDDK) to the C-terminal end of each protein. Followingrestriction digests of the PCR products, PC synthase genes weredirectly inserted into the pET-24a(+) vector (Novagen, Darms-tadt, Germany) using standard procedures. Following cloning,the fidelity of the gene sequences was verified via Sangersequencing. PC synthase genes were expressed in the Escherichi-a coli strain, BL21(DE3)pLysS, as described (Romanyuk et al.2006).

PC synthase purification. Cells were pelleted by centrifuga-tion (12,000 g, 4°C, 10 min) rinsed by resuspension in PBS,and pelleted again. The rinsed pellets were resuspended in1.5 mL lysis buffer containing 50 mM Tris–HCl (pH 7.4),150 mM NaCl, 10% glycerol (v/v), 1% Triton (v/v) and2 mM phenylmethylsulfonyl fluoride pre-chilled to 4°C. A500 lL volume of 0.1 mm glass beads (Biospec Products, Bar-tlesville, OK, USA) was added to each suspension. Additional

TABLE 1. PCR primer sequences for amplification and cloning of TpPCS genes and AtPCS1

Gene name Primer sequence (5′–3′)

TpPCS1 F�TTATCATATGCAACCACCACTAACTpPCS1 R�GATTTGGATATTGATTATAAAGATGATGATGATAAATAATAAGAATTCAAATTpPCS2 F�TTATCATATGGTTGATCAGATAGCTpPCS2 R�CAGAAGAATTGATTATAAAGATGATGATGATAAATAATAAGAATTCAAATTpPCS3 F�TTATCATATGACCTTGTGGAATCCTpPCS3 R�CTAAAAAAGAAGGCGGATTATAAAGATGATGATGATAAATAATAAGAATTCAAAATAtPCS1 F�TTATGCTAGCATGGCTATGGCGAGAtPCS1 R�CTCGCTGCTCCTGCCTATGATTATAAAGATGATGATGATAAATAATAAGAATTCAAAT

F and R denote forward and reverse orientations, respectively.

HIGH AFFINITY ALGAL PC SYNTHASE 33

lysis buffer was added to the centrifuge tubes containing thesuspensions to minimize air headspace. The cells were lysedby five cycles of beating for 30 s (homogenate speed) using aMini-Beadbeater-8 (Biospec Products). Lysates were chilledon ice for 2 min in between cycles. The cell lysates were clari-fied by centrifugation (12,000 g, 4°C, 30 min). PC synthaseenzymes were purified from the resulting cell extract using aFLAG M Purification Kit (Sigma-Aldrich) at 4°C as described inRomanyuk et al. 2006. The purified enzymes were assayedimmediately or stored at �80°C.

For assays in which small amounts of contaminating metalscould affect the results, TpPCS1 and AtPCS1 were subjectedto dialysis to remove metals loosely associated with the puri-fied enzymes. Enzymes (30 lg in 200 lL volume) were incu-bated on ice for 1 h in buffer containing 1 mM Tris–HCl(pH 7.4) and 5 mM tris(2-carboxyethyl)phosphine (TCEP),which is both a reductant and metal ligand. The enzyme prep-arations were dialyzed in Slide-a-Lyzer mini dialysis tubes, MWcutoff of 7,000 kDa (Pierce, Rockford, IL, USA), against two suc-cessive 300 mL volumes of 50 mM HEPES-NaOH (pH 8) for 6 heach at 4°C. To further minimize oxidation of TpPCS1, the dialy-sis buffer was continually flushed with argon gas during theexperiment. Dialysis buffer was applied to a column of ChelexResin 100 (Bio-Rad, Hercules, CA, USA) to remove metals priorto use. Centrifuge tubes, beakers, and pipette tips used to per-form dialyzed enzyme experiments were washed once with ultra-pure 1 M HCl and then rinsed three times with ultrapure water(Millipore, Billerica, MA, USA) prior to use.

SDS-PAGE and western blot analysis. For SDS-PAGE analysis,proteins were separated on 10% denaturing gels and stainedwith Coomassie Blue as described in Sambrook and Russell(2001). For Western blot analysis, proteins were transferredonto PVDF membranes as described in Sambrook and Russell(2001). The blots were probed with an anti-FLAG antibody(Sigma-Aldrich) and visualized using the WesternBreezeChromogenic Kit (Invitrogen, Carlsbad, CA, USA).

Enzyme assays and PC2 quantification. Purified TpPCS1 orAtPCS1 was incubated in standard substrate buffer containing200 mM HEPES-NaOH (pH 8), 7 mM GSH, and 50 lMCdCl2 for 3 min at 25°C unless stated otherwise. When areductant was included in the assay mixture, it was addedimmediately before the assay was started. Pre-incubation ofenzyme with 5 mM or 10 mM dithiothreitol (DTT) for up to10 min had no effect on assay results (data not shown). A10 lL sample of the assay was removed and immediatelyadded to 990 lL of 10 mM methanesulfonic acid. The assaysamples were stored at �20°C or immediately prepared foranalysis. For PC2 quantification, assay samples were labeledwith monobromobimane (Invitrogen) and separated byHPLC using Method C as described in Wei et al. 2003. Beck-man System Gold 32 Karat software was used to calculate thePC2 peak area. Calibration was with a PC2 standard (Anaspec,Fremont, CA, USA).

For assays in which S-methylglutathione (S-methyl-GS;Sigma-Aldrich) was used as a substrate, dialyzed AtPCS1 orTpPCS1 (0.5 lg � mL�1) was incubated in substrate buffercontaining 200 mM HEPES-NaOH (pH 8), 3 mM S-methyl-GS, and variable CdCl2 at 25°C for 15 min. To terminate thereaction, 5-sulfosalicylic acid was added to 5% (w/v). Result-ing samples (200 lL volume) containing S-methyl-GS andS-methyl-PC were separated on an Alltech Adsorbosphere HSC8 3U (Grace Davison Discovery Sciences, Deerfield, IL,USA) column according to Vatamaniuk et al. 2000. Columnfractions were collected in 30 s intervals by an SC100 FractionCollector (Beckman Coulter, Fullerton, CA, USA). A 100 lLvolume of each fraction was added to the wells of an opaque96-well plate. A Synergy four Multi-Mode Microplate Reader(BioTek, Winooski, VT, USA) was programmed to dispense

275 lL of o-phthaldialdehyde (OPA) solution containing0.5 M sodium tetraborate, 10 mg � mL�1 OPA, and5 lL � mL�1 2-mercaptoethanol to each well. The plate wasgently shaken for 2 min and the fluorescence was read atexcitation 340, emission 455. Calibration was with S-methyl-GS.

Anaerobic procedures. Anaerobic purifications and assayswere performed inside an anaerobic chamber filled withnitrogen and helium gas (Coy Laboratories Inc., Grass Lake,MI, USA). Prior to use, pre-made buffers and water werepurged with nitrogen gas and equilibrated inside the cham-ber for at least 24 h. Assay components DTT, TCEP and GSHwere prepared fresh daily inside the chamber. E. coli cell pel-lets were resuspended inside the anaerobic chamber asdescribed above. Parafilm was used to seal the centrifugetubes, which were then placed on ice and removed from thechamber for the lysis and centrifugation procedures. Theresulting supernatants were returned to the chamber andimmediately applied to the FLAG M affinity column whichwas pre-equilibrated with 20 volumes of nitrogen gas-purgedlysis buffer. Purified protein elutants were assayed immedi-ately inside the chamber or transferred to a sterile glass vialwhich was then sealed with an airtight rubber stopper andcrimped metal cap and stored at �80°C. Protein concentra-tions were measured using the Coomassie Plus-The BetterBradford Assay Kit (Pierce).

Kinetic modeling. Excel Solver (Microsoft, Redmond, WA,USA) was used to simultaneously fit kinetic data points toeither a 2-substrate non-sequential (ping-pong) model or a2-substrate sequential (ternary complex) to calculate kineticconstants. Ultimately, TpPCS1 kinetic curves were fit to thelatter model using the equation below:

Activity ¼ Vmax½GSH�KGSHð1þ KCd�GS2

½Cd�GS2�Þ þ ½GSH�ð1þ KCd�GS2½Cd�GS2�Þ

RESULTS

TpPCS1 is an active PC synthase, TpPCS2 and TpPC3are not. Of the three genes predicted to encode forPC synthases, only TpPCS1 (JN368431) encodes fora full length PC synthase. PC synthase activity incrude cell extracts of E. coli expressing recombinantTpPCS1-FLAG (hereafter referred to as TpPCS1)was confirmed using standard assay conditions(Gupton 2007). TpPCS1 was immunoaffinity puri-fied and a single band corresponding to � 51 kDawas observed by SDS-PAGE and immunoblotting,consistent with the predicted molecular weight of49.8 kDa plus the FLAG tag (0.88 kDa; Fig. 1). Simi-larly, in agreement with Vatamaniuk et al. (1999), asingle band corresponding to 58 kDa was observedfor AtPCS1-FLAG (hereafter referred to as AtPCS1).An amino acid sequence alignment of TpPCS1

(deduced from the nucleic acid sequence) withAtPCS1, revealed that TpPCS1 contains many of thefeatures common to all PC synthases: the conservedcatalytic triad amino acids, the highly conservedN-terminal domain and the divergent C-terminaldomain (Fig. 2). In the N-terminal domain, TpPCS1shares 48% sequence identity with AtPCS1 and con-tains all but three of the strictly conserved residues

34 TIFFANY GUPTON-CAMPOLONGO ET AL.

of PC synthases. These three divergent amino acidsare all in the vicinity of the catalytic aspartate (Asp-180 in AtPCS1) (Romanyuk et al. 2006), a regionidentified as the fourth B-loop in NsPCS (Vivareset al. 2005). The isoleucine at position 178 isreplaced by a valine; immediately adjacent to thecatalytic aspartate (Asp-180) is a threonine in placeof a valine (Val-181); and a proline (Pro-187) isreplaced by a glycine in TpPCS1 (Fig. 2). Further-more, all but one of the predicted phosphorylationsites in AtPCS1 (Thr-139; Wang et al. 2009) arepresent in the N-terminal of TpPCS1. In TpPCS1,the C-terminal domain, which is cysteine-rich incanonical PC synthases, contains only four cysteinescompared to 10 in AtPCS1 (Fig. 2). None of thepredicted phosphorylation sites in the C-terminal ofAtPCS1 (Wang et al. 2009) align with sites in the C-terminal of TpPCS1.

TpPCS1 has a pH optimum at pH 8, which is typi-cal for PC synthases (data not shown). Its tempera-ture optimum is 45°C (Fig. 3), which is 10–15°Chigher than previously published values for PC

synthases (Grill et al. 1989, Chen et al. 1997, Ovenet al. 2002) and confirmed here.In contrast to TpPCS1, PC2 synthesis was not

detected in assays of crude cell E. coli extractsexpressing recombinant TpPCS2-FLAG or TpPCS3-FLAG, or in assays of the subsequently immunopuri-fied proteins. The amino acid sequences, deducedfrom the nucleic acid sequences, of TpPCS2(JX418033) and TpPCS3 (JX418034) contain theresidues comprising the catalytic triad, but are trun-cated in comparison to canonical PC synthases, lack-ing the C-terminal domain. Their N-terminaldomains are 47% (TpPCS2) and 21% (TpPCS3)identical to those of TpPCS1. Multiple PC synthasegenes and truncated genes are not exclusive toT. pseudonana. A. thaliana contains a second PC syn-thase gene in addition to AtPCS1, called AtPCS2,which encodes for a PC synthase that is only func-tional in vitro and has been hypothesized to be tis-sue specific (Cazal�e and Clemens 2001). Inaddition, a cyanobacterium, Nostoc sp., contains agene that encodes for a truncated PC synthase-likeenzyme, NsPCS, that can cleave glycine from GSH,and that was reported to weakly synthesize PCs inone study (Harada et al. 2004, Tsuji et al. 2004).But despite our knowledge of these PC synthasegene variants, the role of multiple PC synthasegenes in T. pseudonana is not known.Purified TpPCS1 is highly sensitive to oxidation. Puri-

fied TpPCS1 and AtPCS1 were both assayed immedi-ately after purification in standard substrate buffer.Although activity was observed for AtPCS1(11 � 1 lmol � mg�1 � min�1), no activity was mea-sured for TpPCS1 unless DTT was also added to thesubstrate buffer. Both purified enzymes were subse-quently assayed with a range of DTT concentrationsand a non-thiol containing reductant, TCEP.TpPCS1 activity increased with increasing concentra-tions of both DTT and TCEP and both reductants

FIG. 1. SDS-PAGE and immunoaffinity detection of purifiedTpPCS1 and AtPCS1. Purified enzymes, (a) TpPCS1 and(b) AtPCS1, each 1 lg, were separated by PAGE on a 10% geland stained with Coomassie Blue. Purified enzymes (2.5 lg) weretransferred onto a PDVF membrane and probed with a polyclonalanti-FLAG antibody (c) TpPCS1 and (d) AtPCS1. Sizes of amolecular weight standard run in parallel are indicated.

FIG. 2. Amino acid sequence alignment of TpPCS1 (JN368431) and AtPCS1 (AF085230). Sequences were aligned using ClustalW2(Chenna et al. 2003) and the figure was prepared using Jalview (Waterhouse et al. 2009). Identical amino acids are shaded in light gray.Asterisks denote conserved catalytic residues. C-terminal domain cysteines of TpPCS1 and AtPCS1 are in bold and underlined. Aminoacids in TpPCS1 that are different from strictly conserved residues are in white text with dark gray background. Conserved N-terminalcysteines are in white text with light gray background.


elicited similar activities at each concentrationtested (Fig. 4a). In contrast, AtPCS1 was active inthe absence of DTT. Activity remained the same inthe presence of 5 mM DTT and decreased atgreater concentrations (Fig. 4b). Conversely,the addition of TCEP to assays more than doubledAtPCS1 activity at all concentrations tested. SinceDTT is a reductant as well as a strong ligand for Cd,it is likely that any positive effect of its action on At-PCS1 as a reductant is eclipsed by a simultaneousdecrease in the concentration of Cd2+ and there-fore, the substrate Cd�GS2, resulting in lower activ-ity. Given that decreased activity in the presence ofDTT was only observed with AtPCS1, it is likely thatTpPCS1 activity is saturated with respect to Cd�GS2at the concentration range used in this experiment.

To minimize the observed oxidation of TpPCS1,the enzyme was mostly purified and assayed in ananaerobic chamber. TpPCS1 prepared in this man-ner and assayed without reductants was nearlyfour times more active than the oxidized enzymeassayed in substrate buffer containing the highestconcentrations of reductant (128 � 8, Table 2 vs.35 � 7 lmol � mg�1 � min�1, Fig. 4a). This suggeststhat some fraction of the oxidation which occurredduring aerobic purification may not be totallyreversible. Activity was further increased by 47% and21%, when 5 mM DTT or TCEP was added toanaerobic assays, respectively (Table 2). In contrast,the activity of AtPCS1 prepared in the anaerobicchamber without reductants was similar to oxidizedenzyme assayed in the presence of TCEP (22 � 2 vs.� 25 lmol � mg�1 � min�1). The addition of 5 mMDTT had little effect on anaerobically purified At-PCS1 activity, whereas the addition of 5 mM TCEPincreased activity by 72% (Table 2). The increasedactivity of both anaerobically purified TpPCS1 andAtPCS1 in the presence of reductants suggests thatsome oxidation occurred during one of the proce-

dures performed outside of the anaerobic chamberduring the purification process.Standard assays performed on anaerobically puri-

fied TpPCS1 and AtPCS1 following their removalfrom the anaerobic chamber revealed that bothwere susceptible to oxidation over time, but thatTpPCS1 lost activity more rapidly (Fig. 5). After only2 h, the activity of TpPCS1 had decreased by 30%and a linear decrease was observed thereafter with86% of the original activity lost after 9 h. The addi-tion of 5 mM TCEP resulted in 90% recovery of theoriginal activity (data not shown). In contrast,AtPCS1 was more resistant to oxidation. Its activityremained constant for 4 h following removal fromthe chamber and after 9 h, 54% of its original activ-ity was lost.TpPCS1 activity is inhibited by direct Cd-binding. Pre-

viously, S-methyl-GS has been substituted for GSHin PC synthase assays in attempts to determine theeffect of direct Cd-binding to PC synthases. TheS-methyl-GS substrate is identical to GSH with theexception that the sulfhydryl side chain is blockedby a methyl group, thereby preventing the formation

FIG. 3. Effect of temperature on TpPCS1 and AtPCS1 activities.(a) TpPCS1 and (b) AtPCS1, each 1 lg � mL�1, were assayed in pre-warmed standard substrate buffer. Error bars represent means � SE(n = 3–4). Where not shown, error bars are within symbol.

FIG. 4. Effect of varying reductant concentrations on TpPCS1and AtPCS1 activities. (a) TpPCS1 and (b) AtPCS1, each1 lg � mL�1, were assayed in standard substrate buffer containingeither DTT or TCEP. Error bars represent means � SE (n = 3–4).Where not shown, error bars are within symbol.


of the Cd�GS2 substrate. In a previous study,AtPCS1 was able to synthesize S-methyl-PCs fromS-methyl-GS and was active in the absence of addedCd. Activity was augmented by the addition of Cd to0.5 lM and was decreased at higher concentrations(Vatamaniuk et al. 2000). We repeated these assayswith dialyzed TpPCS1 and AtPCS1. Like AtPCS1,TpPCS1 was active in the absence of Cd, however,the addition any Cd (0.25 lM or higher concentra-tions) resulted in an inhibition of activity. Theresponse of AtPCS1 to Cd was confirmed (Fig 6).TpPCS1 kinetics are best described by a ternary complex

model. Dialyzed TpPCS1 was assayed in substrate buf-fer containing varying concentrations of GSH andCdCl2 (Fig. 7a). Enzyme activity increased with bothincreasing Cd�GS2 and GSH. The regression lines of aHanes-Wolff plot of the data, [Cd�GS2]/m versus[Cd�GS2] (Fig. 7b), collectively intersected the y-axisat one focal point, which is characteristic of a ternarycomplex model described by the following reaction:

TpPCS1þGSHþ Cd � GS2 ! TpPCS1þ Cd � PC2

þ GSHþ Gly

In this model, both substrates must be bound tothe enzyme in order for the reaction to proceed.

The substrate binding may be ordered or random. Fit-ting our data to this model (R2 = 0.95), we obtained aKmCd�GS2 for TpPCS1 of 0.087 � 0.02 lM, two orders ofmagnitude less than values previously reported for At-PCS1. A value of 9.2 � 2.3 lM was reported for At-PCS1 (Vatamaniuk et al. 2000) using a nonlinear fit tothe ping-pong model equation and recently a value of5.14 � 1.22 lM was reported (Ogawa et al. 2011). Thecalculated KmGSH for TpPCS1 was 85.5 mM, which wassignificantly higher, but in the same order of magni-tude as those reported for AtPCS1, 13.6 � 3.3 mM(Vatamaniuk et al. 2000) and 18 mM (Ogawa et al.2011). The vmax calculated for TpPCS1 was808 lmol � mg�1 � min�1, which is higher, but in thesame order of magnitude as the range of values (from152 � 22 to 302 � 25 lmol � mg�1 � min�1) estimatedfor AtPCS1 (Ogawa et al. 2011).We note that there is the distinct possibility that

the kinetics observed here, best modeled by theternary complex model, is instead a manifestationof a special case of the ping-pong model indistin-guishable from the ternary complex model (Cookand Cleland 2007). In this special case, the hypo-thetical TpPCS1-c-Glu-Cys acyl intermediate isextremely unstable, which is manifested in thekinetic analysis as the absence of the intermediatecomplex.

TABLE 2. Activity of anaerobically purified TpPCS1 and AtPCS1 with and without reductant. Purified enzymes were assayedin standard substrate buffer with and without the addition of 5 mM DTT or TCEP. Assay results are reported as activity aswell as a percentage of enzyme activity in the absence of reductant.

Reductant TpPCS1 activity (lmol � mg�1 � min�1) % Control AtPCS1 activity (lmol � mg�1 � min�1) % Control

None 128 � 8 100 22 � 2 100DTT 188 � 8 147 20 � 1 91TCEP 155 � 18 121 38 � 2 172

FIG. 5. Time dependent oxidation of TpPCS1 and AtPCS1.Enzymes (1 lg � mL�1) were assayed inside an anaerobic chamber instandard substrate buffer (t = 0). The enzymes were removed fromthe chamber and assayed at the indicated times. Activities are pre-sented as the percentage of activities measured for each respectiveenzyme when assayed inside the chamber which were122 � 8 lmol � mg�1 � min�1 and 39 � 3 lmol � mg�1 � min�1 forTpPCS1 and AtPCS1, respectively. Error bars represent means � SE.(n = 3). Where not shown error bars are within symbol.

FIG. 6. S-methyl-PC2 synthesis by AtPCS1 and TpPCS1 at varyingCd concentrations. Dialyzed enzymes, AtPCS1 and TpPCS1, wereassayed in 3 mM S-methyl-GS and varying concentrations of CdCl2.Activities are presented as the percentages of activities measured forAtPCS1 and TpPCS1 assayed in the absence of cadmium, which were1.5 and 3.9 lmol � mg�1 � min�1, respectively. Error bars representmeans � SE (n = 3).


DISCUSSION

Demonstrated biochemical utilization of Cd as anutrient by marine Thalassiosira sp. (Lane et al.2005) as well as unusually high intracellular ratiosof PC to Cd in a subset of these eukaryotic algae,including T. pseudonana, motivated this study ofTpPCS1. As described here, TpPCS1 exhibits severalcharacteristics that set it apart from the well charac-terized A. thaliana enzyme, AtPCS1. Oxidationdecreases the activity of TpPCS1 far more rapidlythan that of AtPCS1. Activity can be preserved byanaerobic purification of the enzymes or recoveredupon addition of strong reductants. In addition,TpPCS1 has a significantly greater substrate affinityfor the Cd�GS2 complex and is more sensitive to Cdinhibition than AtPCS1. Both these findings areconsistent with the very low trace metal concentra-tions typical of surface seawater. Furthermore,TpPCS1 kinetics is best described by a ternary com-plex model as opposed to the ping-pong model thatdescribes AtPCS1 kinetics.

Previous studies of AtPCS1 have shown that theenzymatic synthesis of PCs from GSH can be sepa-rated kinetically into two steps (Vatamaniuk et al.2000). The first step is mechanistically understoodand involves cleavage of the peptide bond betweencysteine and glycine resulting in the formation ofan acyl intermediate (via a thioester on the catalyti-cally essential Cys-56) and the release of glycine.The second step involves a transpeptidation reactionduring which the c-Glu-Cys is transferred from theenzyme intermediate onto the amino terminal ofanother GSH molecule or a longer PC oligomer. Inthis model, the cysteinyl sulfhydryl of the secondarysubstrate must have a blocked thiol either as a resultof a metal complex or an alkylated sulfhydryl, as inexperimental in vitro assays. In TpPCS1, these twosteps are kinetically indistinguishable, eitherbecause a ternary complex must form prior to thesynthesis of any product or because the acyl-enzymeintermediate of TpPCS1 is less stable than that ofAtPCS1. A ternary complex mechanism would implythat there is no deglycylation activity independentof PC synthesis whereas an extremely unstable inter-mediate would result in rapid GSH degradation inaddition to PC accumulation. This distinction hasyet to be experimentally confirmed.Adding to the complexity of metal-PC synthase

interactions, Cd can also be inhibitory at concentra-tions that are only slightly greater than those thatstimulate maximal activity. The cysteine-rich C-termi-nal of AtPCS1 has been shown to protect it frommetal poisoning in assays in which S-methyl-GS isthe substrate (Romanyuk et al. 2006). TpPCS1 isinhibited by lower concentrations of Cd comparedto AtPCS1 (Fig. 6). This greater sensitivity ofTpPCS1 to Cd may be due to the decreased numberof cysteines in its C-terminal. In AtPCS1, Cd has alsobeen shown to have an influence on phosphoryla-tion which in turn has been shown to influenceactivity in vitro (Wang et al. 2009) as well as in vivo(Lima et al. 2012); we did not examine the influ-ence of phosphorylation on PC synthase activity.As shown in Figure 2, the TpPCS1 amino acid

sequence contains the highly conserved N-terminaldomain (1–221 for AtPCS1; 1–238 for TpPCS1)within which it retains nearly all of the strictly con-served amino acids identified in Romanyuk et al.(2006). In the immediate vicinity of the catalyticallycritical aspartate (Asp-180 in AtPCS1), however,there are three substitutions unique to TpPCS1.In particular, the substitution of valine (Val-181) bythreonine at the site immediately adjacent to thecatalytic aspartate may have an influence on activity.Secondary structure models constructed by the Phy-re server (Kelley and Sternberg 2009) reveal thatthe hydroxyl group of the threonine side chain isadjacent to the substrate binding site and thereforemay interact with GSH, perhaps destabilizing theacyl intermediate or altering the affinity of thesecondary substrate. Furthermore, in the crystal

FIG. 7. TpPCS1 kinetic modeling. (a) Dialyzed TpPCS1(0.5 lg � mL�1) was assayed in substrate buffer containing varyingconcentrations of CdCl2 and GSH. (b) Hanes-Woolf plot of datapoints from (a).


structure of the cyanobacterial NsPCS, a secondbinding site or pocket adjacent to the c-Glu-Cysbound acyl intermediate was identified (Vivareset al. 2005). Three of the five residues lining thispocket are located in the stretch of peptide identi-fied as the fourth B-loop which also includes thecatalytically necessary aspartate. These changesunique to TpPCS1 and located adjacent to theactive site or lining the opening through which thesubstrate must enter, may contribute to observeddifferences between TpPCS1 and AtPCS1. For exam-ple, a larger opening might leave the catalyticallycritical cysteine more vulnerable to Cd inhibition oroxidative damage. It may also leave the active sitemore vulnerable to the diffusive influx of water,which can act as a nucleophile hydrolyzing the thio-ester in the enzyme intermediate (Rea 2006), poten-tially explaining noted differences in the kineticmodeling.

Given the exceedingly low metal concentrationsin the open ocean, it is perhaps not too surprisingthat the affinity of TpPCS1 for Cd�GS2 is greaterthan that of AtPCS1. PC synthases were originallyhypothesized to be the synthesizers of PCs, primarilyas agents of metal detoxification. However, this viewhas been re-examined given that the evolution ofPC synthases appears to have occurred long beforeanthropogenic environmental Cd contamination(Clemens and Per�soh 2009). PCs and PC synthasesmay yet be shown to play a larger role than previ-ously thought in metal homeostasis. Although pre-sumed deleterious to all other organisms, Cd hasbeen shown to have a biological significance in bothT. pseudonana and T. weissfloggi (Lane et al. 2005).These algae contain a carbonic anhydrase that uti-lizes Cd specifically in its active site. In addition, ithas been shown in vitro that PCs are capable ofshuttling Cd and Zn to this carbonic anhydrase(Xu et al. 2008). It has been hypothesized thatthese specific carbonic anhydrases have evolved inresponse to the low availability of metals in seawater,an environmental condition, which results in com-petition between members of algal blooms to obtainand retain essential metal nutrients, including Cdfor these particular algae. Given the unique physiol-ogy of T. pseudonana, buffering of Cd in the cell isfunctionally both a process of metal homeostasisand detoxification. Therefore, T. pseudonana can beviewed as a model system providing a raison d’etrefor the evolutionary presence of PC synthases ingeneral.

We thank Olena Vatamaniuk (Cornell University) for assis-tance in preparation of the manuscript, Carl Batt (CornellUniversity) for providing technical support. This work was par-tially supported by NSF grant OCE 0451781. TGC acknowl-edges a Cornell University Provost’s Diversity Fellowship.

Ahner, B. A., Kong, S. & Morel, F. M. M. 1995. Phytochelatin pro-duction in marine algae. 1. An interspecies comparison. Lim-nol. Oceanogr. 40:649–57.

Ahner, B. A., Lee, J. G., Price, N. M. & Morel, F. M. M. 1998.High phytochelatin in the Equatorial Pacific. Deep Sea Res. I45:1779–96.

Ahner, B. A., Wei, L., Oleson, J. R. & Ogura, N. 2002. Glutathi-one and other low molecular weight thiols in marine phyto-plankton under metal stress. Mar. Ecol. Prog. Ser. 232:93–103.

Cazal�e, A.-C. & Clemens, S. 2001. Arabidopsis thaliana expresses a sec-ond functional phytochelatin synthase. FEBS Lett. 507:215–9.

Chen, J., Zhou, J. & Goldsbrough, P. B. 1997. Characterization of phy-tochelatin synthase from tomato. Physiol. Plant. 101:165–72.

Chenna, R., Sugawara, H., Koike, T., Lopez, R., Gibson, T. J., Hig-gins, D. G. & Thompson, J. D. 2003. Multiple sequence align-ment with the Clustal series of programs. Nucleic Acids Res.31:3497–500.

Clemens, S. & Per�soh, D. 2009. Multi-tasking phytochelatin synth-ases. Plant Sci. 177:266–71.

Cook, P. F. & Cleland, W. W. 2007. Enzyme Kinetics and Mechanism.Garland Science, London and New York.

Grill, E., L€offler, S., Winnacker, E.-L. & Zenk, M. H. 1989. Phyto-chelatins, the heavy-metal-binding peptides of plants, are syn-thesized from glutathione by a specific c-glutamylcysteinedipeptidyl transpeptidase (phytochelatin synthase). Proc. Natl.Acad. Sci. USA 86:6838–42.

Grill, E., Winnacker, E.-L. & Zenk, M. H. 1985. Phytochelatins:the principal heavy-metal complexing peptides of higherplants. Science 230:674–6.

Grill, E., Winnacker, E.-L. & Zenk, M. H. 1987. Phytochelatins, aclass of heavy-metal-binding peptides from plants, are func-tionally analogous to metallothioneins. Proc. Natl. Acad. Sci.USA 84:439–43.

Gupton, T. L. 2007. Cloning, Expression and Activity of PhytochelatinSynthase from Thalassiosira pseudonana. MS thesis, Cornell Uni-versity, Ithaca, NY, 43 pp.

Harada, E., von Roepenack-Lahaye, E. & Clemens, S. 2004. A cy-anobacterial protein with similarity to phytochelatin synthas-es catalyzes the conversion of glutathione to c-glutamylcysteine and lacks phytochelatin synthase activity.Phytochemistry 65:3179–85.

Kelley, L. A. & Sternberg, M. J. E. 2009. Protein structure predic-tion on the Web: a case study using the Phyre server. Nat.Protoc. 4:363–71.

Kondo, N., Imai, K., Isobe, M., Goto, T., Murasugi, A., Wada-Nak-agawa, C. & Hayashi, Y. 1984. Cadystin a and b, major unitpeptides comprising cadmium binding peptides induced in afission yeast-separation, revision of structures and synthesis.Tetrahedron Lett. 25:3869–72.

Lane, T. W., Saito, M. A., George, G. N., Pickering, I. J., Prince,R. C. & Morel, F. M. M. 2005. Biochemistry: a cadmiumenzyme from a marine diatom. Nature 435:42.

Lima, A. I. G., Da Cruz e Silva, E., Paula, E. M. & Figueira, A.2012. Cd-induced signaling pathways in plants: possible regu-lation of PC synthase by protein phosphatase 1. Environ. Exp.Bot. 79:31–6.

Loscos, J., Naya, L., Ramos, J., Clemente, M. R., Matamoros, M. A.& Becana, M. 2006. A reassessment of substrate specificity andactivation of phytochelatin synthases from model plants byphysiologically relevant metals. Plant Physiol. 140:1213–21.

Ogawa, S., Yoshidomi, T. & Yoshimura, E. 2011. Cadmium (II)-stimulated enzyme activation of Arabidopsis thaliana phyto-chelatin synthase 1. J. Inorg. Biochem. 105:111–7.

Oven, M., Page, J. E., Zenk, M. H. & Kutchan, T. M. 2002. Molec-ular characterization of the homo-phytochelatin synthase ofsoybean Glycine max: relation to phytochelatin synthase.J. Biol. Chem. 277:4747–54.

Price, N. M., Harrison, G. I., Hering, J. G., Hudson, R. J., Nirel,P. M. V., Palenik, B. & Morel, F. M. M. 1989. Preparationand chemistry of the artificial algal culture medium Aquil.Biol. Oceanogr. 6:443–61.

Rea, P. A. 2006. Phytochelatin synthase, papain’s cousin, in ste-reo. Proc. Natl. Acad. Sci. USA 103:507–8.

Rea, P. A., Vatamaniuk, O. K. & Rigden, D. J. 2004. Weeds,worms, and more. Papain’s long-lost cousin, phytochelatinsynthase. Plant Physiol. 136:2463–74.


Romanyuk, N. D., Rigden, D. J., Vatamaniuk, O. K., Lang, A., Cah-oon, R. E., Jez, J. M. & Rea, P. A. 2006. Mutagenic definitionof a papain-like catalytic triad, sufficiency of the N-terminaldomain for single-site core catalytic enzyme acylation, and C-terminal domain for augmentative metal activation of a eukary-otic phytochelatin synthase. Plant Physiol. 141:858–69.

Ruotolo, R., Peracchi, A., Bolchi, A., Infusini, G., Amoresano, A.& Ottonello, S. 2004. Domain organization of phytochelatinsynthase: functional properties of truncated enzyme speciesidentified by limited proteolysis. J. Biol. Chem. 279:14686–93.

Sambrook, J. & Russell, D. W. 2001. Molecular Cloning: A laboratory Man-ual, 3rd ed. Cold Spring Harbor Laboratory Press, New York.

Strasdeit, H., Duhme, A.-K., Kneer, R., Zenk, M. H., Hermes, C.& Nolting, H.-F. 1991. Evidence for discrete Cd(SCys)4 unitsin cadmium phytochelatin complexes from EXAFS spectros-copy. J. Chem. Soc. Chem. Comm. 16:1129–30.

Tsuji, N., Nishikori, S., Iwabe, O., Shiraki, K., Miyasaka, H., Tak-agi, M., Hirata, K. & Miyamoto, K. 2004. Characterization ofphytochelatin synthase-like protein encoded by alr0975 froma prokaryote, Nostoc sp. PCC 7120. Biochem. Biophys. Res. Com-mun. 315:751–5.

Vatamaniuk, O. K., Mari, S., Lang, A., Chalasani, S., Demkiv, L. O.& Rea, P. A. 2004. Phytochelatin synthase, a dipeptidyltransfer-ase that undergoes multisite acylation with c-glutamylcysteineduring catalysis: stoichiometric and site-directed mutagenicanalysis of Arabidopsis thaliana PCS1-catalyzed phytochelatinsynthesis. J. Biol. Chem. 279:22449–60.

Vatamaniuk, O. K., Mari, S., Lu, Y.-P. & Rea, P. A. 1999. AtPCS1,a phytochelatin synthase from Arabidopsis: isolation and in vi-tro reconstitution. Proc. Natl. Acad. Sci. USA 96:7110–5.

Vatamaniuk, O. K., Mari, S., Lu, Y.-P. & Rea, P. A. 2000. Mecha-nism of heavy metal activation of phytochelatin (PC) syn-thase: blocked thiols are sufficient for PC synthase-catalyzedtranspeptidation of glutathione and related thiol peptides.J. Biol. Chem. 275:31451–9.

Vivares, D., Arnoux, P. & Pignol, D. 2005. A papain-like enzymeat work: native and acyl–enzyme intermediate structures inphytochelatin synthesis. Proc. Natl. Acad. Sci. USA 102:18848–53.

Wang, H.-C., Wu, J.-S., Chia, J.-C., Yang, C.-C., Wu, Y.-J. & Juang,R.-H. 2009. Phytochelatin synthase is regulated by proteinphosphorylation at a threonine residue near its catalytic site.J. Agric. Food Chem. 57:7348–55.

Waterhouse, A. M., Procter, J. B., Martin, D. M. A. M. C. & Bar-ton, G. J. 2009. Jalview Version 2-a multiple sequence align-ment editor and analysis workbench. Bioinformatics 25:1189–91.

Wei, L., Donat, J. R., Fones, G. & Ahner, B. A. 2003. Interactionsbetween Cd, Cu, and Zn influence particulate phytochelatinconcentrations in marine phytoplankton: laboratory resultsand preliminary field data. Env. Sci. Technol. 37:3609–18.

Xu, Y., Feng, L., Jeffrey, P. D., Shi, Y. & Morel, F. M. M. 2008.Structure and metal exchange in the cadmium carbonic an-hydrase of marine diatoms. Nature 452:56–61.


Documents

Characterization of a High Affinity Phytochelatin Synthase From The Cd-Utilizing Marine Diatom Thalassiosira pseudonana